USF Libraries
USF Digital Collections

Structural diversity in crystal chemistry :

MISSING IMAGE

Material Information

Title:
Structural diversity in crystal chemistry : rational design strategies toward the synthesis of functional metal-organic materials
Physical Description:
Book
Language:
English
Creator:
Cairns, Amy
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Metal-organic frameworks
Gas storage
Isosteric heats of adsorption
Supermolecular building blocks
Dissertations, Academic -- Chemistry -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Metal-Organic Materials (MOMs) represent an important class of solid-state crystalline materials. Their countless attractive attributes make them uniquely suited to potentially resolve many present and future utilitarian societal challenges ranging from energy and the environment, all the way to include biology and medicine. Since the birth of coordination chemistry, the self-assembly of organic molecules with metal ions has produced a plethora of simple and complex architectures, many of which possess diverse pore and channel systems in a periodic array. In its infancy however this field was primarily fueled by burgeoning serendipitous discoveries, with no regard to a rational design approach to synthesis. In the late 1980s, the field was transformed when the potential for design was introduced through the seminal studies conducted by Hoskins and Robson who transcended the pivotal works of Wells into the experimental regime. The construction of MOMs using metal-ligand directed assembly is often regarded as the origin of the molecular building block (MBB) approach, a rational design strategy that focuses on the self-assembly of pre-designed MBBs having desired shapes and geometries to generate structures with intended topologies by exploiting the diverse coordination modes and geometries afforded by metal ions and organic molecules. The evolution of the MBB approach has witnessed tremendous breakthroughs in terms of scale and porosity by simply replacing single metal ions with more rigid inorganic metal clusters whilst preserving the inherent modularity and essential geometrical attributes needed to construct target networks for desired applications. The work presented in this dissertation focuses upon the rational design and synthesis of a diverse collection of open frameworks constructed from pre-fabricated rigid inorganic MBBs (i.e. M(CO2)4, M2(RCO2)4, M3O(RCO2)6, MN3O3, etc), supermolecular building blocks (SBBs) and 3-, 4- and 6-connected organic MBBs. A systematic evaluation concerning the effect of various structural parameters (i.e. pore size and shape, metal ion, charge, etc) on hydrogen uptake and the relative binding affinity of H2-MOF interactions for selected systems is provided.
Thesis:
Dissertation (PHD)--University of South Florida, 2010.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Amy Cairns.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0004537
usfldc handle - e14.4537
System ID:
SFS0027852:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Structural Diversity in Crystal Chemistr y: Rational Design Strategies Toward the Synthesis of Functiona l Metal-Organic Materials by Amy J. Cairns A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Chemistry College of Arts and Sciences University of South Florida Major Professor: Mohamed Eddaoudi, Ph.D. Michael J. Zaworotko, Ph.D. Brian Space, Ph.D. Juergen Eckert, Ph.D. Date of Approval: June 4, 2010 Keywords: metal-organic frameworks, gas storage, isosteric heats of adsorption, supermolecular building blocks Copyright 2010, Amy J. Cairns

PAGE 2

Dedication To my family for their continuous support throughout this journey

PAGE 3

Acknowledgements First and foremost, I would like to take this opportunity to express my sincere appreciation towards my Ph.D. advisor, Dr Mohamed Eddaoudi for providing me with the opportunity to conduct res earch under his supervision and for his continuous financial support, inspiration, and guidance thr oughout my graduate career at USF. I would also like to acknowledge the rest of my committee members – Drs. Michael J. Zaworotko, Brian Space, and Juer gen Eckert for their countless discussions, positive input, and helpful suggestions concerning the research presented in this dissertation and for taking valuable time out of their busy schedules to serve on my dissertation examination committee. Special thanks to Dr. Anthony Coleman for agreeing to chair my dissertation defe nse and taking the time to fly all the way from France. In particular, many thanks to Dr. Juergen Eckert for his help with the inelastic neutron scattering (INS) experiments and for his expertise, patience, and limitless hours of discussion during our beam time at the Institu t Laue-Langevin (ILL) in Grenoble, France. Special thanks to Drs. Lukasz Wojtas, Victor Kravtsov, Gregory McManus, and Mohamed Alkordi for single-crystal data collection and Dr Matthias Thommes (Quantachrome Instruments) fo r assisting with sorption expe riments. Last but certainly not least, I would like to extend my heartfelt gr atitude to all the past and present students and post-doctoral fellows in the research groups of Drs. Eddaoudi, Zaworotko, and Space, as well as, all members of the SMMARTT team.

PAGE 4

i Table of Contents List of Tables ................................................................................................................ ..... ix List of Figures ............................................................................................................... ..... xi List of Abbreviations ..................................................................................................... xxi x Abstract ...................................................................................................................... .. xxxiii Chapter 1. Introduction to Metal-Organic Materi als: Historical Perspectives, Design Principles, and Po tential Applications ..................................................................1 1.1. Premable ...........................................................................................................1 1.1.1. Nanoscale Materials and Devices ...........................................................1 1.1.2. Solid State Chemistry .............................................................................3 1.1.2.1. Crystalline versus Amorphous Solids .......................................3 1.1.2.2. Single-Crystal X-ray Diffraction (SCD) ...................................3 1.1.2.3. Crystal Packing: Role of Intermolecular Interactions ...............5 1.1.2.4. The Cambridge Structural Database .........................................6 1.2. Metal-Organic Materials (MO Ms): Historical Perspectives .............................8 1.2.1. Inclusion Compounds .............................................................................8 1.3. Rational Assembly of MO Ms from Expanded NitrogenDonor Ligands ...............................................................................................12 1.3.1. Supramolecular Polygons .....................................................................16 1.3.2. Angular Nitrogen-Donor Ligands ........................................................18 1.3.3. Nitrogen-Based Metal Clusters ............................................................18

PAGE 5

ii 1.4. Construction of MOMs from Carboxylate-Based Ligands .............................19 1.5. Hybrid Ligand Design: MO Ms Constructed from HeteroFunctional Linkers ......................................................................................... .25 1.6. Identification and Classifi cation of MOMs: Topological Descriptors ......................................................................................................29 1.6.1. Point Symbols and Schlfli Notation ...................................................29 1.6.2. Extended Schlfli Nota tion and Vertex Symbols .................................33 1.6.3. Coordination Sequence .........................................................................35 1.6.4. Tiling and Transitivity .........................................................................37 1.6.5. Reticular Chemistry Structure Resource (RCSR) Database ...............................................................................................39 1.7. Properties and Potential Applications of MOMs ............................................40 1.8. Potential Clean Energy Alternativ e: Bridging the Interface Between Hydrogen Storage and On-board Applications ..............................................42 1.8.1. Conventional Hydrogen Storage Technologies: Advantages and Limitations ................................................................45 1.8.2. Potential Hydrogen Storage Materials .................................................45 1.8.2.1. Chemical versus Physical Adsorption ....................................46 1.8.2.2. Proposed Chemisorption-Based Adsorbents ...........................47 1.8.2.3. Hydrogen Storage in Phys isorption-based Materials ..............48 1.9. Key Factors which Govern H2 Binding in MOFs: Literature Summary and Future Outlook ........................................................................49 1.9.1. Importance of High Surface Area ........................................................50 1.9.2. Design Strategies to Incorp orate Potential Open-Metal Binding Sites into MOFs .....................................................................52 1.9.2.1. Extra-Framework Approaches: Ion-Exchange and Metal Doping ...................................................................55

PAGE 6

iii 1.9.2.2. Experimental Techniques: Detection of Metal-H2 Binding in MOFs ....................................................................56 1.9.2.2.1. SCD, FT-IR, and Low-temperature Powder Neutron Diffraction Studies ...................................57 1.10. Inelastic Neutron Scattering (INS) ................................................................59 1.11. Other Applications: MOFs as Platforms for the Capture, Sequestration and Separation of Gases .........................................................62 1.11.1. Methane Storage ...............................................................................62 1.11.2. Storage and/or Sequestra tion of Carbon Dioxide .............................63 1.12. Characterization of Porous Solids via Physical Adsorption Methods.........................................................................................................68 1.12.1. Exploiting the Porosity of MOFs: Protocols for Sample Activation .............................................................................69 1.12.2. Classification of Adsorption Isotherms ............................................71 1.12.3. Surface Area Analysis .......................................................................74 1.12.3.1. Adsorbate Selection ...........................................................74 1.12.3.2. Langmuir and BET Theories .............................................75 1.12.4. Pore volume and Pore size Analysis .................................................79 1.13. References .....................................................................................................82 Chapter 2. Design and Synthesis of Por ous Isostructural MOFs with “soc” Topology: Insights into Adsorption Characterization and Adsorbate-MOF Interactions ................................................................................................................. .....92 2.1. Introduction .....................................................................................................92 2.2. Results and Discussion ...................................................................................97 2.2.1. Low Pressure Adsorption Measurements for 1 ..................................100 2.2.2. Inelastic Neutron Scatte ring (INS) Studies for 1 ................................103

PAGE 7

iv 2.2.3. High Resolution Physical Adsorption Measurements: A Systematic Benchmark Study on 1 ................................................108 2.2.3.1. Adsorption Characterization using H2 and CH4 as Adsorptives ......................................................................109 2.2.3.2. Polar and Non-polar Adsorptives: CO2 versus CH4 Adsorption Studies .......................................................112 2.2.4. Isostructural Series of so c-MOF Analogs: Prototypical Platform to Conduct Systematic H2-MOF Interaction Studies ..........118 2.2.4.1. Iron-Based TMBBs ...............................................................119 2.2.4.2. Decorated TMBBs: Incorporation of Coordinated Halide Anions .......................................................................122 2.2.4.3. Low Pressure Gas Adsorption Measurements ......................124 2.2.4.3.1. Evaluating the Effect of Open Metal Binding Sites on H2-MOF Interactions ................125 2.2.4.3.2. Contributions from Di fferent Extra-framework Anions: Clversus NO3 .......................................129 2.2.4.3.3. Polarizability Effects via Decorated Indium-based TMBBs ..........................................134 2.2.5. Validity of the Isosteric Heats of Adsorption via Computational Studies .......................................................................139 2.2.6. INS Studies: Insights into the H2-MOF Preferential Binding Sites in 2 and 4 .....................................................................140 2.3. Experimental Section ....................................................................................145 2.3.1. Materials and Methods .......................................................................145 2.3.2. Instrumentation and Software ............................................................146 2.3.2. Synthesis and Characterization ..........................................................149 2.4. Summary and Conclusions ...........................................................................152 2.5. References .....................................................................................................155

PAGE 8

v Chapter 3. Structural Diversity of 4-Connected Nodes: Serendipity versus Predictability in Crystal Chemistry ...............................................................................159 3.1. Synthesis of 2-Periodic Metal-Organic Materials .........................................159 3.1.1. Introduction ........................................................................................159 3.1.1.1. Topological Perspective: Regular and Semi-Regular Plane Tilings ..................................................160 3.1.1.2. Design Principles and Applications: Nanoscale Kagom and Square Grid Lattices ........................................162 3.1.1.3. Taxonomy of Calixarene-like MOMs ...................................166 3.1.2. Results and Discussion .......................................................................167 3.1.2.1. Decorated Kagom Lattice ...................................................167 3.1.2.2. Synthesis of an Anioni c Tetragonal Square Grid Lattice from Tetrahedral MBBs ...........................................171 3.1.3. Experimental Section .........................................................................174 3.1.3.1. Materials and Methods ..........................................................174 3.1.3.2. Synthesis and Characterization .............................................174 3.2. From 2-Periodic Layers to 3-Pe riodic Frameworks: Pillaring as a Design Strategy ............................................................................................175 3.2.1. Introduction ........................................................................................175 3.2.1.1. Pillaring Strategy: Design Principles ....................................176 3.2.2. Results and Discussion .......................................................................179 3.2.2.1. Isoreticular MOFs w ith nbo Topology: Pillared Kagom Layers .....................................................................180 3.2.2.2. Enroute to Larger Func tionalized Cavities and Higher Surface Area: Pillared Kagom Lattices from Expanded Tetracarboxylate Ligands ...........................186 3.2.2.3. Pillared Tetragonal Square Grid Networks ...........................190

PAGE 9

vi 3.2.2.4. Potential Gas Storage Applications ......................................193 3.2.2.5. Lanthanide-Based lvt-MOMs ...............................................200 3.2.2.6. Default versus Non-Default Structures: MOMs Assembled from Distorted Tetrahedral Building Units .....................................................................................204 3.2.2.6.1. MOFs Constructed from p -Block Metal Centers .......................................................205 3.2.2.6.2. Compound 16: Prospective Properties .................211 3.2.2.7. Strategy to Incorporate Unsaturated Metal Centers into MOFs via Organic Building Blocks: PorphyrinBased Metalloligands ...........................................................213 3.2.2.8. MOFs with pts Topology Constructed from Monoand Bimetallic Transition Metal Centers ..................219 3.2.3. Experimental Section .........................................................................225 3.2.3.1. Materials and Methods ..........................................................225 3.2.3.2. Synthesis and Characterization .............................................225 3.3. Summary and Conclusions ...........................................................................230 3.4. References .....................................................................................................231 Chapter 4. From Molecular Building Bl ocks (MBBs) to Supermolecular Building (SBBs): The Pursuit of Highly Connected Metal-Organic Frameworks (MOFs) that Possess a Superior Level of Hierarchical Complexity ...............................................................................................236 4.1. Covalent Cross-Linking of Nanoscale Faceted Polyhedra ...........................236 4.1.1. Introduction ........................................................................................236 4.1.1.1. Classification of Metal-Organic Polyhedra ..........................240 4.1.2. Results and Discussion .......................................................................244 4.1.2.1. Structural Analysis: 12-connected fcu nets ...........................247 4.1.2.2.1. Properties ...........................................................................252

PAGE 10

vii 4.1.2.2. Structural and Topological Analysis: (3,24)-connected rht nets from Hexatopic Organic Ligands ............................252 4.1.2.2.1. Gas Sorption Measurements for Compound 25 .......................................................260 4.1.3. Experimental Section .........................................................................268 4.1.3.1. Materials and Methods ..........................................................268 4.1.3.2. Synthesis and Characterization .............................................268 4.2. Rational Directed Assembly of Finite Metal-Organic Cubes (MOCs): A Viable Pathway to Target Zeolitelike Metal-Organic Frameworks (ZMOFs) .........................................................271 4.2.1. Introduction ........................................................................................271 4.2.2. Results and Discussion ......................................................................278 4.2.3. Experimental Section .........................................................................282 4.2.3.1. Materials and Methods ..........................................................282 4.2.3.2. Synthesis and Characterization .............................................283 4.3. Summary and Conclusions ...........................................................................283 4.4. References .....................................................................................................285 Chapter 5. Tuning Pore Size via Charge d Metal-Organic Frameworks: Pathway to Enhance Adsorbate-MOF Interactions .....................................................................289 5.1. Introduction ...................................................................................................289 5.2. Results and Discussion .................................................................................292 5.2.1. Porous Anionic MOFs with ctn Topology: Structural Description ........................................................................292 5.2.1.1. Properties: Preliminary Gas Sorption and Metal Ion Exchange Studies .................................................296

PAGE 11

viii 5.2.2. Novel Quaternary Net form ed by SelfAssembly of Triangles, Squares and Trigonal Prismatic Building Units ...................................................................................................302 5.2.2.1. Structural Description of 36 ..................................................304 5.2.2.2. Properties: Preliminary Gas Sorption Studies.......................309 5.3. Experimental Section ....................................................................................311 5.3.1. Materials and Methods .......................................................................311 5.3.2. Synthesis and Characterization ..........................................................311 5.4. Summary and Conclusions ...........................................................................313 5.5. References .....................................................................................................315 Chapter 6. Conclusions and Prospects for Future Studies ...............................................318 6.1. Summary .......................................................................................................318 6.2. Future Prospects ............................................................................................321 Appendices .................................................................................................................... ...323 Appendix A. X-ray Crystallographic Data and Structure Refinement ................323 Appendix B. Experimental a nd Calculated XRPD Patterns ................................351 Appendix B. Thermal Gravimet ric Analysis (TGA) Patterns .............................361 About the Author ................................................................................................... End Page

PAGE 12

ix List of Tables Table 1.1. Comparison of so me intermolecular forces10 ..............................................6 Table 1.2. Coordination sequence computed up to k = 10 for the cubic and hexagonal diamond nets117. .................................................................35 Table 1.3. Select “revised” technical targets (September 2009) set forth by the US Department of Energy (DoE) for on-board hydrogen storage systems in light-duty vehicles135 ...................................44 Table 1.4. Select examples of MOFs with high surface areas ....................................52 Table 1.5. Select examples of MOFs to highlight the relationship between Q st and accessible open-metal centers ........................................53 Table 1.6. CO2 adsorption capacities ( e.g. low and high) for selected MOFs ............66 Table 2.1. Tentative rotational transitions (meV) assigned for the hydrogen adsorbed at different sites on 1 ................................................... 106 Table 2.2. Selected Sorption data for compounds 1 – 5 ..........................................153 Table 3.1. Comparison of the cages and window dimensions between compounds 8 11 and 12 having nbo topology .......................................196 Table 3.2. Sorption data for compounds 8 11 and 12 .............................................197 Table 3.3. Select examples of binodal edge-transitive nets90 ...................................204 Table 3.4. Coordination sequence up to k = 10 determined for 16 and 17 ...............208 Table 3.5. Summary of the metal cation exchange studies for 16 {[In(ABTC)]DMA} ................................................................................ 212 Table 5.1. Summary of the metal cati on exchange studies carried out on 31 which is tentatively formulated as [In12(BTC)16]-1212DMA+ .................302 Table. 5.2. Selected structural features of compound 36 ...........................................307

PAGE 13

x Table. 5.3. Coordination sequen ce for the independent nodes in 36 .........................308 Table. 5.4. Vertex symbols for the independent nodes in 36 .....................................308

PAGE 14

xi List of Figures Figure 1.1. Histogram to illustrate the growth of the CSD since 197012 .......................7 Figure 1.2. Ball and stick re presentation of two PB an alogues. (left) Fragment of the first crystal structure of a PB analogue. Mn = orange; Co = magenta; N = blue; and C = gr ay; and (right) Decorated PB analogue comprised of [Nb6(Cl)12(CN)6]4and Mn(II) ions bridged by CN ligands.25 Guest molecules are omitted or clarity. Mn = orange; Nb = green; Cl = yello w; N = blue; and C = gray. .......................10 Figure 1.3. Ball and stick representation of a Hofmann-type clathrate: (left) Tetragonal sheet comprised of octahedral and square-planar metal centers, CdN4(NH3)2 and Ni(CN)4, respectively; and (right) Piperizine molecules trapped be tween the layers. Hydrogen atoms and water molecules have been omitted for clarity.31 Cd = yellow; Ni = green; N = blue; and C = gray ...........................................................11 Figure 1.4. Examples of MX2A4 Werner-type complexs: (a) X=Cl-, A=pyridine;34 (b) X= NCS-, A=3,5-lutidine; 35 (c) X=NCS-, A=4,4’-azo-bis(4-pyridyl) 36 ......................................................................12 Figure 1.5. (Left) Non-interp enetrating anionic MOM with dia topology constructed from Zn(II) and Cu(I) tetrahedral building blocks.TMA cations, shown in sp ace-filling representation, reside in each of the adamantane cavitie s to balance the charge. Hydrogen atoms have been omitted for clarity. Zn = yellow; Cu = green; N = blue; and C = gray. (Right) Schema tic representation of the cubic diamond topology39 ....................................................................................15 Figure 1.6. Schematic representation the first dia MOA constructed from complex organic MBBs: (a) Prot otypal tetrahedral metal ion coordination environment, Cu(I); (b) Pre-designed tetrahedral MBB, 4,4',4'',4'''-tetracyanotetraphenylme thane; and (c) Fragment of the {CuI[C(C6H4 .CN)4]}+ MOA All hydrogen atoms and solvent molecules are omitted for clarity. Color code: Cu = green; C = gray; N = blue.39 ..................................................................................................15

PAGE 15

xii Figure 1.7. Examples of polytopic nitr ogen-donor ligands: (l eft to right) imidazole (IMI), hexamethylenetetramine (HMTA), 4,4'-bipyridine (BIPY), 9, 10-bis(4-pyridyl)anthracene, 1,2,4,5-tetra(4-pyridyl)benzene, and 2,4,6-tris(4-pyridyl)-1,3,5-triazine ......................................................16 Figure 1.8. (left) First metal-organi c molecular square, [Pd(BIPY)(En)]4 8+;51 (right) an example of a mol ecular triangle,[Pd(BIPY)(Tmen)]3 6+.55 Hydrogen atoms and counteri ons are omitted for clarity. Color code: Pd = green; C = gray; N = blue ..............................................17 Figure 1.9. (a) 3-oxygen-centered trinuclear cluster, [M3O(N3CR)3(H2O)3]; (b) dinuclear paddlewheel-like cluster, [M2(N2CR)3(H2O)6]; (c) M4Cl(N4CR)8L4 tetranuclear cube-like cluster; (d) [M3(N4CR)6(H2O)6 trinuclear cluster; (e) [M3O(N2CR)3] trinuclear cluster; (f) protot ypal example of a 3-periodic sodalite-type MOM constructed from a M4Cl(N4CR)8L4 (M= Cu or Mn) clusters and a triangular ligand, H3TPT-3 tz Hydrogen atoms and solvent molecules have been omitted for Clarity. Color Code: M = green, C = gray, N = blue, Cl = yellow, and O = red.76 Note: The yellow sphere located inside the cavity represents the largest sphere that can fit insi de taking into account van der Waals radii ....................................................................................19 Figure 1.10. Select examples of pol ytopic carboxylate-based organic ligands used to construct MOMs ...............................................................20 Figure 1.11. Examples of rigid meta l carboxylate-based MBB clusters used to construct MOMs, each of which can be translated into geometrical shapes [SBU(s)]: (a) Dimetal tetracarboxylate cluster, often regarded as the “paddlewheel” M2(RCO2)4L2 forms either a square or octahedral SBU; (b) Basic chromium acetate trimetal cluster M3(RCO2)6L3 represents a trigonal prismatic SBU; and (c) Basic zinc acetate cluster, Zn4O(RCO2)6 forms an octahedral SBU. Color Code: C = gray; O = red; M= green .......................................21 Figure 1.12. Two examples of IRMO Fs constructed from linear carboxylate-based ligands and the 6connected basic zinc acetate, [Zn4O(CO2)6] metal cluster: (left) MOF-5 (IRMOF-1); and (right) IRMOF-11. Hydrogen atoms and solvent molecules have been omitted for clarity; Color Code: Zn = green, C = gray, and O = red .......................................................................................................22

PAGE 16

xiii Figure 1.13. (a) Ball-and-stick repres entation of the neutral nanoball or MOP-1, formulated as, [Cu2( m -BDC)4]12; (b) Schematic representation reveals a small rhomibhexahedron consisting of vertex-linked square SB Us. The 5-position of the m -BDC unit lies on the vertices of the SBB and is highlighted in orange. Color Code: C = gray; O = red; and Cu = green. All hydrogen atoms and solvent molecules are omitted for clarity .................................24 Figure 1.14. Select fragments from the crystal structures of (a) MIL-100; and (b) MIL-101 whereby the assembly of chromium-based TMBBs form a supertetrahedral building block that can be rationalized as a tetrahed ral building unit (center) ....................................25 Figure 1.15. Select examples of hetero -functional ligands with the potential directionality indicate d by the black arrow. Top (from left to right): 4,6-pyrimidinedicarboxylic acid ; pyridine-3,5-bis(phenyl-4carboxylic acid); and 4,5-imidazoledicarboxylic acid. Bottom (from left to right): 2,3 -pyrazinedicarboxylic acid; 4-pyrimidinecarboxylic acid; 2,5-pyridinedicarboxylic acid; 3-pyridinecarboxylic acid; a nd 4-imidazolecarboxylic acid65, 99-104 ..........26 Figure 1.16. Illustration of the (3,24)-connected rht -MOF: (a) Heterofunctional ligand yields a trigonal [Cu3O(N4CR)3] MBB and [Cu2(RCO2)4] paddlewheel MBBs; (b) Tiling representation ...................27 Figure 1.17. Ball-and-stick and schema tic representations of (left) rho -ZMOF comprised of 8-coordinate InN4O4 MBBs (= InN4 TBUs) and (right) sod -ZMOFs comprised of 6-connected InN4O2 MBBs (= InN4 TBUs). Both compounds were constructed using an angular ditopic hetero-functional ligand, 4,5-imidazoledicarboxylic acid (H3-ImDC)64 ...............................................................................................28 Figure 1.18. (left) Tetravalent ba sic zinc acetate cluster, [Zn4O(RCO2)6] simplified into a 6-connected node (right) via the points of extension ....................................................................................................30 Figure 1.19. Representation of the ( n,p ) classification system for 2-periodic lattices: (a) (4,4) Tetragonal sheet; and (b) ( 6,3) Honeycomb lattice ..........................................................................................................31 Figure 1.20. Schematic representation of two Platonic solids: (left) Tetrahedron; and (right) Cube ...................................................................32

PAGE 17

xiv Figure 1.21. A schematic to illustrate the difference between strong rings, rings, and cycles: (left) 4-cycle that is a strong ring, (middle) 12-cycle that is the sum of a 6and 8-cycle, thus is not a ring, and (right) A 12-cycle that is th e sum of a 6-, 8-, and 4-cycle but not of just 2 cycles and theref ore is a ring (not a strong ring) .............34 Figure 1.22. Tiling representation and f ace symbols for two zeolite nets: (left) lta 3[46] + [46.68] + [412.68.86]; and (right) rho 3[48.82] + [412.68.86] ................................................................................................37 Figure 1.23. A plot of pore volume versus surface area for select MOFs to illustrate the correlation between these factors ..........................................51 Figure 1.24. Schematic illustration of se lect elements and the deuterium isotope to emphasize the cross sec tions due to neutron scattering (coherent = blue; incoherent = purple) and absorption (green)190 .............60 Figure 1.25. Schematic of the rotatio nal energy level diagram for a dihydrogen molecule with two ro tational degrees of freedom, under the influence of a hindering potential (Courte sy of Dr. Eckert at UCSB) ........................................................................................61 Figure 1.26. IUPAC classificat ion of sorption isotherms224 ..........................................72 Figure 1.27. A semi-logarithmic plot of the low pressure sorption data collected for indium soc -MOF using nitrogen (red) and argon (blue) as adsorptives at 77.3 K and 87.3 K, respectively shown to highlight the differences in pore filling ranges. ( i.e. 10-7 < P/P0 <10-5 for N2 and 10-5 < P/P0 <10-3 for Ar)174 ....................75 Figure 1.28. BET plots for indium soc -MOF derived from the argon adsorption data collected at 87K: (a) a plot of versus where n is the amount adsorbed; (b) BET plot in the classical re lative pressure range; and (c) a linear BET plot generated by deleting all point above 0.04 atm174 ...........79 Figure 1.29. (a) Semi-logarithmic plot of the experimental argon (87.3 K) isotherm for indium socMOF with cylindrical-like pores compared to the NLDFT-Fit usi ng both the oxidic/zeolite and carbonaceous models and (b) the corresponding pore size distribution174 .............................................................................................82

PAGE 18

xv Figure 2.1. Commonly employed inorga nic and organic MBBs and their corresponding augmented forms: (a) BTEC organic MBB, square BU; (b) [Zn4O(RCO2)6] MBB, octahedral SBU; (c) BTC organic MBB; triangular BU; and (d) [M3O(RCO2)6(H2O)3] trimer MBB (TMBB), trigonal prismatic SBU ..........................................93 Figure 2.2. Select fragments from the X-ray crystal structure of 1 : (a) In-TMBB, generated in situ as [In3O(CO2)6(H2O)3] and the organic ligand, H4-ABTC; (b) Inorganic and organic MBBs can be viewed as a 6-connected node having trigonal prismatic geometry and a 4-connected rectangular-plana r geometry, respectively; (c) Ball-and-stick representation of the cage that houses the partially occupied [NO3]ions; and (d) Polyhedral building unit representation of th e cage. (center) Optical image of 1 to show the crystal morphol ogy of the as-synthesized compound. Color code: In = green; C = gray; N = blue; O = red. Hydrogen atoms and solvent molecules are omitted for clarity ................98 Figure 2.3. Select fragment from the crystal structure of 1 to emphasize the packing of the cages and highlight the two types of intersecting channels. The green tubular rods run through the hydrophobic channels, while the pink tubular rods run through the hydrophilic channels. Color code: In = green; C = gray; N = blue; O = red. Hydrogen atoms, solvent molecules, and [NO3]anions are omitted for clarity .......................................................100 Figure 2.4. Sorption data collected on 1 : (a) Nitrogen sorption isotherm at 77 K; (b) Hydrogen sorption is otherm at 77 K and 87 K; and (c) Isosteric heat of adsorption for H2 ......................................................102 Figure 2.5. INS spectra of 1 obtained at 15 K with loadings of 1, 3, and 5 H2 per indium ........................................................................................105 Figure 2.6. Difference spectra calculated for 1 between 3H2/In – 1H2/In and 5H2/In and 3H2/In ..............................................................................107 Figure 2.7. (a) Low pressure volumet ric hydrogen adsorpti on isotherms for 1 collected at 87, 97, and 107 K and (b) Isosteric heat of adsorption for H2 ......................................................................................110 Figure 2.8. Adsorption isotherms for 1 collected at 107 K using methane and hydrogen as adsorptives illustrated in the linear display (left) and semi-logarithmic plot to highlight the adsorption behavior at low pressures (right) .......................................................................................111

PAGE 19

xvi Figure 2.9. (a) Surface excess and absolute amounts of CO2 adsorbed for 1 at 273, 298, and 323 K; and (b) Surface excess and absolute amounts Of CH4 adsorbed for 1 over the same temperature range ........................115 Figure 2.10. (a) A comparison between the absolute CO2 and CH4 adsorption isotherms for 1 as a function of pressure to illustrate the differences in uptake capacity; and (b) Isos teric heat of adsorption for CO2 and CH4 ....................................................................................................116 Figure 2.11. Evaluation of 1 to separate CO2 and CH4 as a function of temperature and pressure : (a) Loading ratio of CO2/CH4 up to pressures of 3 MPa (30 atm); and (b) Zoomed in plot up to 0.3 atm to emphasize the behavior of the loading ratio at low loadings ...............117 Figure 2.12. (a) Ball-and-stick representati on of the nitrate ions encapsulated in the of 2 ; and (b) Ball-and-stick representation of the chloride ions encapsulated in the cages of 3 The nitrate and chloride ions are highlighted in space-filling representation. Color code: Fe = green; C = gray; N = blue; O = re d; Cl = light green. Hydrogen atoms and solvent molecules are omitted for clarity ...............................122 Figure 2.13. Select fragments fr om the crystal structure of 5 : (a) Ball-and-stick representation of the indium-based TMBB, [In3O(RCO2)6(Br)(H2O)2]; (b) Ball-and-stick representation of the anion-free cuboidal cage. Color code: In = green; C = gray; N = blue; O = red; Br = purple. Hydrogen atoms and solvent molecules are omitted for clarity. .............124 Figure 2.14. (a) Argon sorpti on isotherm at 87 K for 2 {Fe(NO3)3 soc -MOF}, after evacuation at 135oC; and (b) Pore size distribution in 2 calculated from the Ar isotherm (~6.42 ) ..............................................126 Figure 2.15. Sorption studies for 2 {Fe(NO3)3 soc -MOF},after evacuation at 135oC: (a) Hydrogen sorption isotherm at 77 K and 87 K; and (b) Isosteric heat of ad sorption comparison between 1 {In(NO3)3 soc -MOF}, (red) and 2 (green) ................................................................127 Figure 2.16. A comparison of the Argon sorption isotherms at 87 K for 2 {Fe(NO3)3 soc -MOF},after evacuation at 135oC (blue) and 175oC (green) ...........................................................................................128 Figure 2.17. (a) Hydrogen sorption isotherm at 77 K and 87 K for 2 {Fe(NO3)3 soc -MOF}, after evacuation at 175oC; and (b) Isosteric heat of adsorption for 2 after evacuation at 135oC and 175oC compared to the Q st for 1 {In(NO3)3 soc -MOF}, after evacuation at 135oC ..............129

PAGE 20

xvii Figure 2.18. Sorption studies for 3 {FeCl3 soc -MOF}, after evacuation at room temperature: (a) Hydrogen sorption isotherm at 77 K and 87 K; and (b) Isosteric heat of adsorption for H2 .....................................131 Figure 2.19. Sorption studies for 3 {FeCl3 soc -MOF}, after evacuation at 165oC: (a) Argon sorption isotherm at 87 K; and (b) Pore size Distribution calculated from th e Ar isotherm (~6.42 ) by applying a cylindrical NLDFT pore model assuming an oxidic/zeolite surface ...............................................................................132 Figure 2.20. (a) Hydrogen sorption isotherm at 77 K and 87 K for 3 {FeCl3 soc -MOF}, after evacuation at 175oC; and (b) Isosteric heat of adsorption for 3 after evacuation at 165oC compared to the Q st for 3 after evacuation at 175oC ....................................................133 Figure 2.21. Sorption studies for 4 {InCl3 soc -MOF},after evacuation at 135oC: (a) Argon sorption isotherm at 87 K; (b) Pore size distribution obtained from a cylindrical NLDFT pore model assuming an oxidic/zeolite surface (~6.12 and 7.15 ) .......................135 Figure 2.22. (a) Hydrogen sorption isotherm at 77 K and 87 K for 4 {InCl3 soc -MOF}, after evacuation at 135oC; and (b) Isosteric heat of adsorption of H2 for 4 after evacuation at 135oC and 200oC .................................................................................................136 Figure 2.23. Sorption studies for 5, {InBr3 soc -MOF},after evacuation at 145oC: (a) Argon sorption isotherm at 87 K; (b) Pore size distribution obtained from a cylindrical NLDFT pore model assuming an oxidic/zeolite surface (~6.12 and 7.16 ) .......................137 Figure 2.24. (a) Hydrogen sorption isotherm at 77 K and 87 K for 5 {InBr3 soc -MOF}, after evacuation at 145oC; (b) Isosteric heat of adsorption for 5 after evacuation at 145oC and 200oC; and (c) Isosteric heat of adsorption of H2 for 5 compared with {InCl3 soc -MOF}, 4 ........................................................................................ 138 Figure 2.25. Isosteric heats of adsorp tion curves for compounds 2 – 5: (left) unzoomed and (right) zoomed computed by fitting the experimental hydrogen sorption data at 77 K and 87 K to a function ...................................................................................................140 Figure 2.26. INS spectra (T = 4.3 K) corresponding to 1 H2 per trimer of 2 {Fe(NO3)3 soc -MOF}, (red) and 4 {InCl3 soc -MOF}, (green) in neutron energy loss over the range: (a) From 1 – 17 meV; and (b) enlarged range below 12 meV .....................................................141

PAGE 21

xviii Figure 2.27. INS spectra (T = 4.3 K) of 3 H2 per trimer in 2 (black) and 3/2 H2 per trimer in 4 (yellow) in neutron energy loss over the range from 1 to 12 meV ..............................................................................................143 Figure 2.28. INS spectra (T = 4.3 K) to emphasize the differences at higher loading in neutron energy loss over the range from 1 to 12 meV: (a) Compound 2 shown after loadings of 2, 6, 18, and 30 mmol which are represented in red, green, blue, and yellow respectively; (b) Compound 4 at loadings corresponding 1.3, 2.7, 12, and 20 mmol of H2 represented in red, blue, green, and yellow, respectively ...........................................................................144 Figure 2.29. Synthetic strategy follo wed for the preparation of 3,3',5,5'-azobenzenetetracarboxylic acid (H4-ABTC). Reagents and Conditions: (i) H2O / NaOH / stir for 30 min; (ii) slow addition of hot glucose ( aq ) and bubble air through the solution for 24 h; (iii) filter the precipitate; dissolve in a minimal amount of H2O; acidify to pH = 1 using 12 M HCl .................................150 Figure 3.1. Tiling representation of the three exclusive regular [111] plane nets: (left) (3,6) hexagonal lattice ( hxl ); (middle) (4,4) square grid lattice ( sql ); and (right) (6,3 ) honeycomb lattice ( hcb )32 .................161 Figure 3.2. Tiling representation of the semiregular plane nets constructed from two or more regular pol ygons. Top (left to right): (4.82) fes or sql-a ; (3.122) hca ; (32.4.3.4) tts ; (33.42) cem Bottom (left to right): (3.4.6.4) htb ; (4.6.12) fxt ; (34.6) fsz ; (3.6.3.6) kgm32 ......162 Figure 3.3. Schematic illustration of th e two nanoscale SBUsthat can be generated by linking molecular squares through m -BDC type ligands: (a) Square n SBU comprised of four square SBUs; and (b) Triangular n SBUs comprised of three square SBUs ..........................163 Figure 3.4. (left) Dimeta l tetracarboxylate, [M2(RCO2)4L2]n, “paddlewheel” building block, which can be regard ed as (right) a square building unit when viewed down the 4-fold axis ...................................................164 Figure 3.5. Examples of Kagom a nd Square Grid MOMs constructed from m -BDC and [Cu2(RCO2)4L2] MBBs: (a-b) Space-filling representation of a single 2-pe riodic layer in a Kagom and tetragonal square grid structur e, respectively, viewed down the z -axis; (c) Schematic illustration of th e Kagom lattice which is comprised of 6and 3-membered rings; and (d) Schematic illustration of the tetragonal square grid lattice comp rised of solely 4-memebered rings. Color Code: C = gray; O = red; hyd rogen = white; and Cu = green. The 5-position of the BDC moiety is highlighted in purple. ...................165

PAGE 22

xix Figure 3.6. The four types of atropiso mers of calix[4]arene: (a) Cone; (b) Partial cone; (c) 1,2-alternate; an d (d) 1,3-alternate ...........................167 Figure 3.7. 5,5'-{benzene-1,4-diylbis[( E )methylylidenenitr ilo]}benzene-1,3 dicarboxylic acid, H4-ImTC .....................................................................169 Figure 3.8. X-ray crys tal structure of 6 : (a) Ball-and-stic k representation of the Kagom lattice viewed down the z -axis; (b) Schematic illustration to show the assembly of triangular n SBUs; (c) Balland-stick representation of the AAA layers in viewed down the x -axis; and (d) Schematic illustrati on of the arrangement of the undulating triangular n SBUs (highlighted in green). Color Code: C = gray; O = red; N = blue; hydrogen = white; and Cu = green ............170 Figure 3.9. N-H…O hydrogen bonding interactions in 6 : (a) Viewed down the z -axis; and (b) The y -axis. The independent layers are represented in purple and yello w to highlight each layer ........................171 Figure 3.10. A general reaction mechanis m for the direct synthesis of amides via a condensation reac tion between a carboxylic acid and an amine65 ..........................................................................................172 Figure 3.11. X-ray crys tal structure of 7 : (a) In(RCO2)4 MBB which can be viewed as a 4-connected node; (b ) Ball-and-stick representation of the tetragonal sheets which pack in an ABAB fashion down the z -axis; and (c) The layers in teract via offset face-to-face interactions through the phenyl ri ngs of 5-AmBDC. Color Code: C = gray; O = red; N = blue; hyd rogen = white; and In = green. The channels are represented by the undulating yellow columns ............173 Figure 3.12. Schematic representation to illustrate the three possible design strategies to pillar 2D layers into 3D frameworks: the purple, blue, and yellow pillars independently represent the axial-to-axial, axial-to-ligand, and ligand-to-liga nd strategies, respectively ..............................................................................................177 Figure 3.13. Select examples of bipyridyl-b ased organic ligands (left to right): pyrazine; 1,4-diazabicyclo[ 2.2.2.]octane; 4,4'-bipyridine; 1,2-di(pyridine-4-yl)ethane-1,2-diol; and N,N'-di(4-pyridyl)1,4,5,8-napht halenetetracarboxyldiimide ...................178 Figure 3.14. Bifunctional organic MBB, 5-(4-pyridinylmethoxy)-isophthalate (PMOI2-) surrounded by three [Cu2(RCO2)4N2] MBBs. Color Code: C = gray; O = red; N = blue; a nd Cu = green. All hydrogen atoms are omitted for clarity ...............................................................................179

PAGE 23

xx Figure 3.15. Tetracarboxylate organic ligands utilized for the ligand-to-ligand pillaring strategy: (left) H4-ABTC; (middle) H4-BIPATC; and (right) H4-BAYTC ...................................................................................180 Figure 3.16. Space-filled views from the x-ray crystal structure of 8 : (a) Viewed along the x -axis to highlight the undulating 2-D layers (shown in black); (b) Select fr agment of one Kagom layer which propagates in the xy -plane (the triangular and hexagonal cavities are highlighted with black circles) ; (c) Layers are pillared along the z -axis; (d) Two metal-organic ca lixarene-like conformations observed in 1: cone and 1,3,5-alte rnate. Color Code: C = gray; O = red; N = blue; hydrogen = white; and Cu = green. All solvent molecules and axial aqua ligands are removed for cl arity, as well as, hydrogen atoms in figures (a) and (c) ......................................................182 Figure 3.17. Schematic representati on of the arrangement of the n SBUs in the 2D Kagom layer: (t op) Triangle bowl-shaped n SBU are formed by the assembly of three vertex-linked molecular squares, which are formed by linking togeth er the centroids of the BDC units; (bottom) the natural curvat ure of the BDC ligand leads to the formation of undulating layers .................................................................183 Figure 3.18. Two types of cages whose di mensions are predetermined by the relative size of the ABTC4ligand: (a) Spheroid-like cage, [Cu12(ABTC)12]; (b) Elongated cylindrical-like cage, [Cu24(ABTC)6]; (c) Schematic n SBU representation of the spheroid cavity; (d) The elongated cavity (highlighted in blue) results from the packing arrangem ent of the spheroid-like cavity. Color Code: C = gray; O = red; N = blue; hydrogen = white; and Cu = green .........................................................................................184 Figure 3.19. X-ray crys tal structure of 11 : (left) Space-filling view down the y -axis to illustrate the covalent cross-linking between undulating layers; and (right) Space-filling view down the z -axis. Color Code: C = gray; O = red; and Cu = green. Hydrogen atoms, axial ligands, and solvent molecules are omitted for clarity ....................................................................................187 Figure 3.20. Two types of cages in 11 : (left) Spheroid-like cage I, [Cu12(BAYTC)12]; and (right) Elongated cylindrical-like cage II, [Cu24(BAYTC)6]. Color Code: C = gray; O = red; and Cu = green ........187 Figure 3.21. X-ray crys tal structure of 12 : (top) Each BIPATC4coordinates to four paddlewheel MBBs and the naphthalene-based moiety twists out of plane; (botto m) Space-filling view down the y -axis. Color Code: C = gray; O = red; N = blue; and Cu = green ......................189

PAGE 24

xxi Figure 3.22. Two types of cages in 12 : (left) Spheroid-like cage, [Cu12(BIPATC)12]; and (right) Elongate d cylindrical-like cage, [Cu24(BIPATC)6]. Color Code: C = gray; O = red; and Cu = green ........190 Figure 3.23. (left) “1,2-D” configuration associated with the BDC-linked SBU in 13 which can be interpreted as a square building unit by choosing the centroids of the benzene ring as the points of extension; (right) A cluster of f our squares vertex-linked at 120o through 1,3-BDC reveals nanoscale square n SBUs. Color Code: C = gray; O = red; and Cu = green ..........................................................191 Figure 3.24. (a) Schematic representation of the corrugated square grid layers in 13 ; (b) Naphthalene moiety in BIPATC4reduced to a linear spacer; and (c) Schematic represen tation of the pillared square grid layers. Color Code: C = gray; O = red; N = blue .............................192 Figure 3.25. Space-filling views from the crystal structure of 13 : (a) Viewed along the x -axis; (b) Tetragonal square grid layer assembled from nanosized building blocks having the 1,2-alternate conformation; (c) Laye rs are pillared along the y -axis; (d) Select fragment viewed down the y -axis to highlight the conformation (twisting) of the BIPATC ligand. Color Code: C = gray; O = red; N = blue; and Cu = green. Aqua ligands and guest molecules are omitted for clarity ....................................................................................193 Figure 3.26. Sorption data for 8 : (a) Nitrogen sorption is otherm at 77 K; and (b) Hydrogen sorption isotherm at 77 K ..................................................195 Figure 3.27. Sorption data for 11 : (a) Nitrogen sorption is otherm at 77 K; and (b) Hydrogen sorption isotherm at 77 K ..................................................197 Figure 3.28. Sorption data for 13 : (a) Argon isotherm at 87 K; (b) Hydrogen isotherms at 77 K and 87 K; and (c ) Isosteric heat of adsorption for hydrogen; and (d) CO2 sorption isotherm measured at 0oC and 25oC ...................................................................................................199 Figure 3.29. Select fragments from the X-ray crystal structure of 14 : (left) 4-connected inorganic MBB, [Yb2(ABTC)(NO3)2(DMF)4]n; and (right) Each ABTC4ligand is connected to four inorganic MBBs. C = gray; O = red; N = blue; a nd Yb = green. All hydrogen atoms are omitted for clarity ...............................................................................202

PAGE 25

xxii Figure 3.30. Select fragments from the X-ray crystal structure of 14 : (left) 4-connected inorganic MBB, [Yb2(ABTC)(NO3)2(DMF)4]n; and (right) Each ABTC4ligand is connected to four inorganic MBBs and vice versa C = gray; O = red; N = blue; and Yb = green. All hydrogen atoms a nd coordinated DMF molecules are omitted for clarity ...............................................................................203 Figure 3.31. Schematic representation of 16 and 17 : (center) In(RCO2)4 and H4-ABTC MBBs which can be vi ewed as tetrahedral and rectangular building units, respect ively; (left) ba ll-and-stick and augmented representation of 16 ; and (right) ball-and-stick and augmented representation of 17 Color Code: C = gray; O = red; N = blue; and In = green. The infi nite channels are represented by the green and purple columns ..................................................................207 Figure 3.32. Select fragments fr om the crystal structure of 17 : (a-b) Illustration of the two 4-connected indium MBBs, [In(RCO2)4]; (c) Ball-andstick presentation viewed along the x -axis to show the three types of channels as highlighted by th e green, purple, blue, and yellow column; (d) CPK presentation of the 2-D undulating tetragonal layer viewed down the z -axis. Color Code: C = gray; O = red; N = blue; and In = green. Hydr ogen atoms and guest molecules are omitted for clarity ...............................................................................209 Figure 3.33. Select fragments fr om the crystal structure of 16 : (a) Illustration of the indium MBB, [In(RCO2)4] whereby two ABTC ligands are directed upward and two dow nward; (b) Ball-and-stick presentation to show the three t ypes of channels, as highlighted by the green, pink, and yellow columns; (c) CPK view of the 2-D tetragonal layer viewed down the z -axis. Color Code: C = gray; O = red; N = blue; and In = gr een. All hydrogen atoms and guest molecules are omitted for clarity .............................................................210 Figure 3.34. Porphyrin-based linkers used in this study: (left) 5,15-bis(2,6 dibromophenyl)-10,20-bis( 3,5-dicarboxyphenyl)porphyin (H4-TCBrPP); (right) 5,10,15,20-tetrakis(4-carboxy)-21 H ,23 H -porphine (H4-TCPP) ..................................................................214 Figure 3.35. Select fragments fr om the crystal structure of 18 : (a) “3-D” conformation; and (b) Upon de leting the pillars, TCBrP, 2-periodic tetragonal square grid layers are revealed, as viewed down the z -axis. Color Code: C = gray; O = red; N = blue; Br = brown; and In = green. Th e purple atom located in the 5-position of the BDC unit represen ts the point of extension of the pillar. All hydrogen atoms a nd guest molecules are omitted for clarity ..................................................................................................216

PAGE 26

xxiii Figure 3.36. X-ray cr ystal structure of 18 : (a) TCBrPP4organic MBB which coordinates to four [In(RCO2)4(H2O)] inorganic MBBs and thus can be viewed as 4-connected r ectangular and distorted tetrahedral building units, respectively; (b) Ball-and-stick and schematic representation viewed along the y-ax is to highlight the two types of infinite channels shown with gr een and purple columns; (c) Balland-stick and schematic representa tion viewed along the x-axis to show the undulating conformation of the 2-D layers. Color Code: C = gray; O = red; N = blue; Br = brown; and In = green. All hydrogen atoms and guest molecules are omitted for clarity ..................217 Figure 3.37. X-ray crys tal structure of 19 : (top-center) Ball-and-stick representation of the tw o “1,3-D” inorganic MBBs: [M(ABTC)4(DMF)(EtOH)] and [M(ABTC)4(DMF)(EtOH)2], where M = Co or Cd. The MBBs can be simplified into 4-connected distorted tetrahedra l and rectangular-planar nodes; (left) Ball-and stick and schema tic representation viewed down the z -direction; and (r ight) ball-and stic k and schematic representation viewed down the x -direction. All hydrogen atoms and solvent molecules are omitted for clarity. Color Code: C = gray; O = red; N = blue; and Co = green ..........................................220 Figure 3.38. Select fragments fr om the crystal structure of 21 : (a)[Co(ABTC)4(DMF)2], which can be viewed as a 4-connected node with a see-saw or highly di storted tetrahedral geometry; (b) Ball-and-stick representation of the infinite rectangular channels that run along the z -axis; (c) Schematic representation of the 4-connected building units down the z-axis. All hydrogen atoms and solvent molecules are omitted for clarity. Color Code: C = gray; O = red; N = blue; a nd Mn = green. The open channels are highlighted with the green columns ...................................................224 Figure 3.39. A schematic to illustrate the cocrystal contro lled solid-state synthesis (C3S3) of benzoimidephenanthro line tetracarboxylic acid (H4-BIPATC). Reagents and Conditions : (i) DMF, (ii) Grind the solids together (uniform purple color), and (iii) = 180oC for 1 hr .......225 Figure 4.1. (a) Ball-and-stick repres entation of the sulfonated nanoball; (b) Schematic representation of the bcc packing of nanoballs, which is facilitated by double cros s-linking; (c) Select fragment from pcu net of nanoballs viewed along the b c axes, which is sustained by quadrupule covalent cross-linking to six nanoballs; (d) Schematic to illustrate the pcu packing of nanoballs .........................239

PAGE 27

xxiv Figure 4.2. Schematic representation of the five regular Platonic Solids (left to right): Tetrahedron, hexahedron (cube), octahedron, dodecahedron, and icosahedrons .............................................................241 Figure 4.3. Examples of MOPs: (a) molecular tetrahedron, [(Fe3O)4(SO4)12(BPDC)6(Py)12]8(IRMOP-51);11 (b) molecular cube, [Ni8(HImDC)12]8(MOC-1);4 and (c) molecular octahedron, [Pd6(2,4,6-tri(4-pyridyl)-1,3,5-triazene)4(2,2'-BIPY)4]12+.12 Hydrogen atoms and guest molecules are omitted for clarity. Color code: M = green; C = gray; O = red; N = blue; S = yellow ...............................242 Figure 4.4. Schematic representations of the 13 semi-regular Archimedean Solids. Top (left to right): Tr uncated tetrahedron, truncated octahedron, truncated cube, c uboctahedron, rhombicuboctahedron (small rhombicuboctahedron). Midd le (left to right): Truncated cuboctahedron (great rhombicuboctahedron), snub cube, icosidodecahedron, truncated dodeca hedron. Bottom (left to right): Truncated icosahedron (bucky ball), rhombicosidodecahedron (small rhombicosidodecahedron) truncated icosidodecahedron (great rhombicosidodecahedron, and snub dodecahedron13,14 .................243 Figure 4.5. Nine faceted polyhedra. Top (left to right): Tetrahemihexahedron, cubohemioctahedron, octahemioctahedron, small rhombihexahedron, small cubicubocta hedron. Bottom (left to right): Small dodecahemidodecahedron, small icosihemidodecahedron, small dodecaicosidodecahedron, small rhombidodecahedron13,16 ...........244 Figure 4.6. Three uniform faceted polyhedr a that can be generated via linking of molecular squares only: (left) cubohemioctahedron (middle) small rhombihexahedron (right) small rhombidodecahedron ................245 Figure 4.7. (left) Tiling representation of the (3,24)-connected rht net; and (right) augmentation of the 24and 3-connected nodes reveals rhombicuboctahedron and tr iangular vertex figures ................................246 Figure 4.8. (left) rht -1 was constructed using a heterofunctional ligand, TZI, to promote the formation of two different inorganic MBBs, [Cu2(O2CR)4] and [Cu3O(N4CR)3]; (right) Isoreticular analogues can also be generated by substi tuting the inorganic trimer for hexacarboxylate organic ligands that contains three m -BDC moieties ....................................................................................................247

PAGE 28

xxv Figure 4.9. (left) Schematic representation of the cubohemioctahedron assembled from square polygons sustained by a 90o vertex connection; Prototypical meta l-organic cubohemioctahedron [M6(bdc)12]12SBB in ball and stick (middle) and space-filling modes (right). Color code: M (Ni, Co) = green; C = gray; O = red. All hydrogen atoms and the decorate d metal site have been omitted for clarity ..................................................................................................248 Figure 4.10. Ball and stick represen tation of the inorganic MBB in 23 : (left) octahedral coordination envi ronments about Ni1 and the symmetrically disordered Ni2 cation; (right) Ni2 metal center shown without disorder ............................................................................249 Figure 4.11. (a) Fragment of the single-crystal structure of 23 comprised of (b) cubohemioctahedral; (c) te trahedral; and (d) truncated octahedral cavities. Hydrogen at oms, decorated metal site, and solvent molecules are omitted for clarity. Color code: Ni = green; C = gray; O = red .....................................................................................250 Figure 4.12. A topological evaluation of 23 and 24 reveals the structure can be interpreted in two ways : (a) Augmented 12-connected fcu topology; (b) Tiling representation of the fcu net comprised of two types of tiles; (c) Schematic of the augmented nbo network; and (d) Tiling view of the nbo net consisting of one type of tile ...................251 Figure 4.13. Select fragments from th e crystal structure and schematic representation of 25 : Top. m -BDC naturally accommodates the 120o angle needed to generate metal-organic truncated cuboctahedral SBBs and rhombi cuboctahedral TeBUs. Bottom. Cross-linking of the nanoscale polyehdra is facilitated via decoration at the 5-position of the m -BDC unit by employing PTMOI6to afford a (3,24)-connected MOF with rht topology. Hydrogen atoms and solvent molecules are omitted for clarity. Color code: Cu = green; C = gray ; O = red; the 5-position of the m -BDC ligand is highlighted in orange ...................................................254 Figure 4.14. Top. (left to right): Ball-a nd-stick and tiling re presentation of the cages C, A, and B, respectively in 25 Bottom (left to right): Tiling representation showing th e assembly of the smaller polyhedral cages around the largest cage in the rh t net. Color code: Cu = green; C = gray; O = red. Hydrogen atoms and solvent molecules are omitted for clarity .............................................................256

PAGE 29

xxvi Figure 4.15. Tritopic hexacarboxylate organic ligands (L6-) employed in this study: (left) PTMOI6spans ~13.6 these dimensions can be substantially increased to ~24.6 as demonstrated in ABPTMOI6(right) ..................................................................................257 Figure 4.16. Ball-and-stick representati on of two of the three cages in 29 : (a) Tetrahedrallike cage can accommodate a van der Waals sphere that measures approximately 17.138 ; and (b) Largest cage is estimated to 27.199 in diameter. Color code: Zn = green; C = gray; O = red; N = blue. Hydrogen atoms and solvent molecules are omitted for clarity. ............................................................258 Figure 4.17. (left) Novel topological assessment of 25-29 rationalized as a trinodal ternary net and its corresponding augmented conformation (right). Topologica l Terms for Node 1: VS = 6.6.8.8.8(2).8(2); CS = 4,8,18, 29, 52, 61, 106, 120, 170, 187; TD10 = 755. Topological Terms for N ode 2: VS = 6.8(3).8(3); CS = 3, 8, 15, 29, 40, 69, 81, 131, 146, 206; TD10 = 728. Topological Terms for Node 2: VS = 8(3).8(3).8(3); CS = 3, 6, 18, 24, 42, 63, 84, 93, 183, 175; TD10 = 691. ................................................259 Figure 4.18. Argon sorption isotherm for 25 collected at 87 K ...................................261 Figure 4.19. (a) CO2 sorption isotherms measured at 0oC and 25oC on compound 25 after exchange in methanol; and (b) CO2 sorption isotherm measured at 25oC to emphasize the steep rise at low loading and filling in the largest cages that approach the mesoporous range at higher pressures .....................................................263 Figure 4.20. Sorption measurements for 25 following activation in methanol: (a) Hydrogen sorption isotherms at 77 K and 87 K; and (b) Isosteric heat of adsorption for H2 ..........................................................................264 Figure 4.21. Sorption data collected on 25 after supercritical CO2 activation: (a) Argon sorption isotherm at 87 K; (b) Hydrogen sorption isotherms at 77 K and 87 K; and (c) Comparison between the isosteric heat of adsorption for 25 and rht -1 ...........................................266 Figure 4.22. Synthetic strategy follo wed for the preparation of 1,3,5-tris(5-methoxy-1,3-benzene di carboxylic acid)benzene (H6-PTMOI). Reagents and Conditions: (i) DMF / KI / K2CO3 / = 100oC for 1h, (ii) add H2O and filter, (iii) NaOH / H2O / MeOH / = 50oC for 12h / HCl, pH = 1 ..................268

PAGE 30

xxvii Figure 4.23. Ball and stick repres entation of corner-sharing [AlO4]5and [SiO4]4tetrahedra bridged via O2at a T-O-T angle of ~145o. Color Code: Al = green; Si = yellow; O = red .........................................272 Figure 4.24. Schematic illustration of some naturally occurrin g and synthetic zeolite frameworks which exhibit versatile pore structures that render them suitable materials for many commercial and industrial applications: (left) Faujasite, FAU (middle) Linde Type A, LTA ; and (right) ZSM-5, MFI ..........................................................................273 Figure 4.25. An example of a ZMOF cons tructed using the single-metal-ionbased MBB approach: (right) Fragment of rhoZMOF shown in ball and stick and node-and-spacer representations; composed of (left) InN4O4 MBBs which can be viewed as InN4 TBUs. Color code: In = green; C = gray; N = blue; O = red57. .....................................275 Figure 4.26. Examples of ZMOFs closely re lated to edge-transitive 8-connected nets synthesized using H3ImDC as an organic linker: (a) Single metal-ion-based fac -MN3(CO2)3 MBBs facilitate the assembly of (b) finite MOCs which can be utilized as 8-connected SBBs to generate (c) left to right: LTA -, ACO -, and AST like topologies which are closely related to the reo -, bcu -, and flu like edgetransitive nets, respectively ......................................................................277 Figure 4.27. (a) Ball and stick representati on of the single-crystal structure of 30 which reveals edge-to-edge c onnections through octahedrally coordinated Mn2+ ions; (b) vertex-to-vert ex connections occur through charge-assisted H-bonding be tween guanidinium ions which link four MOCs through (c) a supram olecular tetrahedron represented in yellow; (d) six MOCs (red tile) assemble to generate the AST-cage (green tile). Hydrogen atoms and solvent molecules are omitted for clarity Color code: Mn, orange; C, gray; N, blue; O, red ........................280 Figure 4.28. (a) MOC and tetrahedron repr esented as 8and 4-connected nodes, respectively which self-assembly to form (b) a binodal edgetransitive flu -like net ................................................................................281 Figure 5.1. Examples of targeted extr a-framework alkylammonium organic cations (left to right) : Dimethylammonium (DMA+), diethylammonium (DEA+), triethylammonium (TEA+), and tetrabutylammonium (Bu4N+) ..................................................................290

PAGE 31

xxviii Figure 5.2. Examples of common coor dination modes generated by BTC ...............292 Figure 5.3. Ball-and-stick and sc hematic representation of the ctn analogs, 31 and 32 Top: the 3and 4-connected organic and inorganic MBBs can be rationalized as triangular and tetrahedral BUs, respectively. Bottom: MBBs self-assemble to yiel d anionic 3-periodic MOFs with a complex intersecting channel system. Color Code: Ga = yellow; In = green; C = gray; O = red. All hydrogen atoms and guest molecules are omitted for clarity .............................................................293 Figure 5.4. Select fragments fr om the crystal structure of 31 : (a) Triangularshape cage viewed down x -axis and formulated as [In9(BTC)11]6-; and (b) Alternate view of the cage to highlight the four type of intersecting channels with window dimensions in the range of 5.473 to 7.880 Color Code: In = green; C = gray; O = red. Hydrogen atoms and solvent molecules are omitted for clarity ..............296 Figure 5.5. N2 and Ar sorption isotherm for 31 collected at 77 K and 87 K, respectively ..............................................................................................298 Figure 5.6. Gas sorption studies for the indium ctn -MOF, 31 : (a) N2 and Ar sorption isotherm at 77 K and 87 K, respectively; (b) H2 sorption data at 77 K and 87 K; and (c) isosteric heat of adsorption for H2 ..........299 Figure 5.7. Gas sorption studies for a gallium ctn -MOF, 32 : (a) N2 and Ar sorption isotherm at 77 K and 87 K, respectively; (b) H2 sorption data at 77 K and 87 K; and (c) Is osteric heats of adsorption for H2 in 32 compared with the indium analog ( 31 ) ...........................................300 Figure 5.8. Optical image and select fr agments from the crystal structure of 36 including the corresponding vertex figures when augmented. (Left) The four MBBs include BTC, [In(RCO2)4], [In3O(RCO2)3(Br)], and [In3O(RCO2)6(Br)]. (Right) SBUs for each of the inorganic and organic MBBs. (Middle) the assembly of the n -connected nodes (where n = 3,4, and 6) yields a porou s 3-periodic MOF comprised of 3 types of cages. Color Code : In = green, C = gray, O = red, Br = pink ..................................................................................................305 Figure 5.9. Tiling representation of 36 5[6^6] + 2[6^9] ...........................................309 Figure 5.10. Gas sorption studies for compound 36 : (a) Ar sorption isotherm at 87 K; (b) H2 sorption data at 77 K and 87 K; and (c) Isosteric heats of adsorption for H2 ........................................................................310

PAGE 32

xxix List of Abbreviations Acronym Full Name H4-ABTC .............................................................. 3,3',5,5'-azobenzenetetracarboxylic acid H4-BIPATC ................................................ Benzoimidephenanthroline tetracarboxylic acid H3ImDC ............................................................................... 4,5-imidazoledicarboxylic acid BTC ..................................................................................... 1,3,5-benzenetricarboxylic acid BTB .............................................................................. 1,3,5-tris(4-carboxyphenyl)benzene BDC ................................................................................................... Benzenedicarboxylate 5-NH2BDC ...................................................................................... 5-aminoisophthalic acid H3TZI ........................................................................................... Tetrazolylisophthalic acid CH3CN ............................................................................................................... Acetonitrile H2O ............................................................................................................................. Water EtOH ......................................................................................................................... Ethanol MeOH .................................................................................................................... Metha nol DMF ............................................................................................. N,N' -Dimethylformamide

PAGE 33

xxx DMA+ ................................................................................................... Dimethylammonium DMA ............................................................................................. N,N' -Dimethylacetamide DEF ................................................................................................. N,N' -Diethylformamide HNO3.....................................................................................................................Nitric aci d HCl ............................................................................................................Hydrochloric ac id CH2Cl2........................................................................................................ Dichloromethane CH3Cl .................................................................................................................. Chlorofor m NMP .................................................................................................... N -Methylpyrrolidone Py ............................................................................................................................ Pyridine TEA ................................................................................................................ Triethyla mine DEA ................................................................................................................. Diethylam ine TeaBr.................................................................................... Tetraethylammonium Bromide TeaCl .................................................................................... Tetraethylammonium Chloride TMAN ................................................................................... tetramethylammonium Nitrate HMTA ........................................................................................... Hexamethylenetetramine IMI .........................................................................................................................Im idazole PIP ......................................................................................................................... P iperizine

PAGE 34

xxxi MOM............................................................................................... Metal-Organic Material MOF ............................................................................................Metal-Organic Framework CP ....................................................................................................... Coordination Polymer MOP .............................................................................................. Metal-Organic Polyhedra MOC ..................................................................................................... Metal-Organic Cube ZMOF ..................................................................... Zeolitelike Metal-Organic Framework ZIF...................................................................................... Zeolitic Imidazolate Framework COF ....................................................................................... Covalent-Organic Framework MBB ............................................................................................. Molecular Building Block SBB ..................................................................................... Supermolecular Building Block SBU ................................................................................................ Secondary Building Unit BU ................................................................................................................... Building Unit TBU.............................................................................................. Tetrahedral Building Unit TeBU ..................................................................................................Tertiary Building Unit nMR ............................................................................................................ n Member Rings FT-IR......................................................................................... Fourier Transform Infrared PXRD ........................................................................................... Powder X-ray Diffraction

PAGE 35

xxxii SCD .............................................................................................. Single Crystal Diffraction TGA ........................................................................................ Thermogravimetric Analysis UV-Vis ..................................................................................................... Ultraviolet-visibl e INS ............................................................................................ Inelastic Ne utron Scattering RCSR ..................................................................... Reticular Chemistry Structure Resource

PAGE 36

xxxiii Structural Diversity in Crystal Chemistr y: Rational Design Strategies Toward the Synthesis of Functional Met al-Organic Materials (MOMs) Amy J. Cairns ABSTRACT Metal-Organic Materials (MOMs) represen t an important class of solid-state crystalline materials. Their countless attractiv e attributes make them uniquely suited to potentially resolve many present and future utilitarian societal challenges ranging from energy and the environment, all the way to include biology and medicine. Since the birth of coordination chemistry, the self-assembly of organic molecules with metal ions has produced a plethora of simple and complex architectures, many of which possess diverse pore and channel systems in a periodic arra y. In its infancy however this field was primarily fueled by burgeoning serendipitous discoveries, with no regard to a rational design approach to synthesis. In the late 1980s, the field was transf ormed when the potential for design was introduced through the seminal studies conducted by Hoskins and Robson who transcended the pivotal works of Wells into the experimental regime The construction of MOMs using metal-ligand directed assembly is often regarded as the origin of the molecular building block (MBB) approach, a rational design strategy that focuses on the self-assembly of pre-designed MBBs having de sired shapes and geometries to generate

PAGE 37

xxxiv structures with intended topologies by e xploiting the diverse c oordination modes and geometries afforded by metal i ons and organic molecules. The evolution of the MBB approach has witnessed tremendous breakthroughs in terms of scale and porosity by simply repl acing single metal ions with more rigid inorganic metal clusters whilst preservi ng the inherent modularity and essential geometrical attributes needed to construct target networks for desired applications. The work presented in this dissertation focuse s upon the rational design and synthesis of a diverse collection of open frameworks constr ucted from pre-fabricated rigid inorganic MBBs ( i.e. [M(CO2)4], [M2(RCO2)4], [M3O(RCO2)6], MN3O3, etc ), supermolecular building blocks (SBBs) and 3-, 4and 6connected organic MBBs. A systematic evaluation concerning the effect of various structural parameters ( i.e. pore size and shape, metal ion, charge, etc ) on hydrogen uptake and the re lative binding affinity of H2-MOF interactions for selected systems is provided.

PAGE 38

1 Chapter 1. Introduction to Metal-Organic Materials: Historical Perspectives, Design Principles, and Potential Applications 1.1 Preamble “What could we do with layered structures wi th just the right layers? What would the properties of materials be if we could really arrange the atoms the way we want them? They would be very interesting to investigate th eoretically. I can’t see exactly what would happen, but I can hardly doubt that when we have some contro l of the arrangement of things on a small scale we will get an enormously greater range of possible properties that substances can have, and of di fferent things that we can do” Richard P. Feynman1 1.1.1. Nanoscale Materials and Devices Nanoscience represents a highl y diverse and interdisciplin ary field of science that is concerned with the study and manipulation of chemical and biological structures and/or devices within limited dimensi ons, more specifically in the range of 1 to 100 nanometers (nm). The ability to have a certain degree of control over matte r at the atomic and molecular level often leads to the emerge nce of attractive properties and unique phenomena that are fundamentally different fr om that observed in bulk systems. This notion of extreme miniaturization is nothing short than ast ounding when one considers that one nanometer is equi valent to one billionth, 10-9, of a meter. To put this size scale into perspective, a common reference would be to compare the rela tive size of marble versus that of the Earth!!

PAGE 39

2 The prolific “ nano” vision provided by physicist Richard Feynman in his legendary talk at the meeting of the American Physical So ciety in 1959 entitled “There’s Plenty of Room at Bottom” is widely rega rded as the birthplac e of the fundamental concepts featured within the realm of na notechnology or nanoscie nce, i.e. concepts pertaining to the controllable manipulation of matter at the atomic level. Several significant discoveries toward the latter half of the 20th century helped to jump start this every growing field. These include but are no t limited to, the invention of the scanning tunneling (STM) and atomic force micros copes (AFM), development of cluster chemistry, and the synthesis of carbon nanotubes and fullerenes.2, 3 The assortment of top-down design met hods employed by engineers to construct miniaturized devices has undoubtedly led to ex ceptional results but th is strategy becomes less viable and more cumbersome in the na noscale regime. Rational top-town design, bottom-up synthetic approaches developed by materials scientists, chemists, and biologists is an attractive alternative to generate prefabricated nanostructures which exhibit fine-tunable features and properties. They therefor e have demonstrated immense potential to be utilized in a my riad of pertinen t applications ( e.g. gas storage and separation, magnetism, drug delivery, etc ). All of the structures presented in this dissertation in fact, i.e. metal-organic mate rials (MOMs), fall into the second category whereby rational design strategies were em ployed to target f unctional solid-state crystalline materials with intended topologi es based on the self-assembly judiciously chosen molecular build ing blocks (MBBs).

PAGE 40

3 1.1.2. Solid State Chemistry 1.1.2.1. Crystalline versus Amorphous Solids All matter is capable of forming a solid phase in the presence of the appropriate conditions ( i.e. sufficiently cooled). The vast majori ty of these solids adhere to one or more crystalline phases, while others can be classified as being amorphous solids ( e.g. glass, polystyrene, etc ). The terms crystal and amorphous in fact come from the Greek words meaning “ice or clear ice” (krust allos) and “without form” (amorphos), respectively. Crystalline mate rials are defined as having l ong range order and therefore the solid consists of a regular repeating array of atoms, molecules, or ions arranged in an orderly periodic fashion. Amorphous materi als on the other hand consist of random molecules with no long range order.4 1.1.2.2. Single-Crystal X-ray Diffraction (SCD) The discovery of X-radiation also known as X-rays by W. C. Rntgen (1895) was a pivotal scientific discovery and paved th e way for fundamental developments in many fields, particular ly crystallography.5, 6 X-rays are a form of el ectromagnetic radiation with a sufficiently small wavelength; that is, in the range of 10 to 0.01 nm. This limit can be further subdivided into soft ( = 10 – 0.10 nm) and hard X-rays ( = 0.10 – 0.01 nm) based upon their penetrating abilities. Hard X-rays are ideally su ited for atomic level structural characterization ( i.e. crystal structures) because this range is comparable to the distance of chemical bonds. Crystallography was developed shortly afte r the discovery of X-rays as a result of the colle ctive contributions of many inspirational scientists including M. V. Laue, R. J. Hauy, W. H. Miller, J. He ssel, A. Bravais, W. H. Bragg, W. L. Bragg

PAGE 41

4 and many others.7, 8 In a typical experiment, a beam of X-rays were passed through a crystalline sample to produce a diffraction pa ttern on a photographic pl ate. It was shown that the observable spots coul d only be caused by diffraction of short wavelengths due to the resultant spatial arrangement of the atom s in the crystal. Diffraction patterns were significantly improved over the years as a re sult of the development of better X-ray sources. The field was truly revolutionized however when the CCD detector and computers were introduced. This is evidenced by the fact that data acquisition and structure solution for small molecule organics has become rather standard practice for present day crystallographers. I would like to highlight a few relevant crystallogra phic terms that are commonly used in solid state chemistry, as several will be used throughout this dissertation. A diffraction pattern having a sufficient number of useable spots is needed to deduce the position of the atoms relative to each other in three spatial directions. A crystal structure is therefore delimited by a point lattice which is characterized by spec ific lengths () and angles (o) referred to as, a, b, c, and , respectively. The sma llest possible repeating unit that can be used to c onstruct the entire lattice using these parameters and only translation is denoted as the unit cell. The un it cell of a crystal can fall into one of seven crystal systems and these include (1) cubi c, (2) tetragonal, (3) orthorhombic, (4) monoclinic, (5) triclinic (6) hexagonal, and (7) trigonal. They can be further subdivided into 14 Bravais lattices and 230 sp ace groups and thus represent the only ways identical objects can be arranged in an infinite lattice.9

PAGE 42

5 1.1.2.3. Crystal Packing: Role of Intermolecular Interactions Molecular recognition of complimentary molecular building blocks (MBBs) is largely responsible for the self-assembly of atoms to yield functional molecular assemblies. An excellent example to illustra te this concept is base pairing of the nucleotides in the double helix of DNA. Note th at molecular synthesi s is concerned with the making and breaking of intr a-molecular covalent bonds ( e.g. organic chemistry), while the manipulation of weak non-covalent in termolecular interact ions between two or more molecular entities ( e.g. in solution or the solid state) is encompassed within the realm of supramolecular chemistry, defined as “chemistry beyond the molecule” by Jehn Marie Lehn who won the Nobel prize (1987) fo r his significant contributions to this area.10 Crystal packing and in many cases the resultant properties of a material are indeed governed by the existence of attractive and re pulsive intermolecular interactions. It is therefore imperative to have a fundamental unde rstanding of the nature of these forces in order to target poten tial made-to-order molecular assemb lies. A comparison of some of the commonly encountered intermolecular fo rces and their approximate energies is provided in Table 1.1. Note that the weak energetics exhibited by these forces often makes it difficult to predict the outcome of the assembly. The hydrogen bond however is certainly recognized as the most important fo rce due to its robustness and directionality ( i.e. predictable nature). Th is is not to say that other forces are not present, as it is rare to observe a single force in a single molecular system.

PAGE 43

6 Table 1.1. Comparison of some intermolecular forces.10 Interaction Energy (kJ/mol) Description Example Ion-Ion 100 – 350 Electrostatic interaction between two oppositely charged ions Na+ ClIon-Dipole Coordinate covalent bond 50 – 200 Bonding of an ion with a lewis base [Na(H2O)6]+ Hydrogen Bond D-H…A 4 – 120 Attraction of a hydrogen atom on an electronegative atom to a dipole on a neighboring atom DNA double helix – stacking < 50 Occurs between elect ron-delocalized systems graphite Dipole-dipole 5 – 50 Alignment of one polar molecule with another acetone London dispersion 1 – 10 Always present ( i.e. arises from polarization) nobel gases 1.1.2.4. The Cambridge Structural Database The Cambridge Structural Database (CSD ) is available to users as a systems software package provided by The Cambridge Crystallographic Data Center (CCDC).11 It includes user friendly programs that are capab le of search and informational retrieval (ConQuest), structure visualizat ion (Mercury), numerical anal ysis (Vista), and database creation (PreQuest). The CSD is comprised of over 500,000 structures which have been solved using X-ray and neutron diffraction methods (Figure 1.1.). As of February 1st, 2010 in fact the database in cludes bibliographic, chem ical, and crystallographic information for 503,268 organic and metal-orga nic compounds from a collection of open publications in the literatur e and private submissions.

PAGE 44

7 Figure 1.1. Histogram to illustrate the growth of the CSD since 1970.12 Each entry deposited into the CSD is assigned a unique reference code consisting of six alphabetical letters ( e.g. MABJUV) and is sometimes accompanied by a two-digit number ( e.g. MABJUV01). This indicates either data collection using a different temperature(s), radiation source, experimental conditions, or the same structure reported in a different journal or publis hed by another authors. The CSD is widely regarded as vital research tool for crystal engineers. Mo st notably perhaps because (1) it is extremely useful at providing a current inventory of pub lished crystal structures and is a critical feature as the number of reporte d structures continues to e xpand, (2) each en try provides pertinent structural information concerni ng the preferred metal-ligand coordination environments, and (3) when used correctly it can be used to gain insights into the role of intermolecular interac tions in different systems and how it governs the overall crystal packing. It is important to mention that the CSD does not store information related to the following, polypeptides and polysaccharides wi th more than 24 units, oligonucleotides, inorganic structures, or metal alloys. Researchers can theref ore refer to the RCSB Protein

PAGE 45

8 Data Bank to obtain experimentally determined structures of protei ns, nucleic acids, and complex assemblies.13, 14 While the Inorganic Crystal Stru cture Database (ICSD) features over 125,000 data entries accompanied by the 3D atomic coordinates of inorganic crystal structures dating back as far as 1913.15 Crystallographic data fo r metals, including alloys, intermetallics and minerals can be obtained from the CRYSTMET database.16 1.2. Metal-Organic Materials (MOM s): Historical Perspectives Metal-organic assemblies (MOAs) sustaine d by coordinate covalent bonds have been constructed over the years from the react ion of neutral or charged organic ligands with an assortment of meta l cations. This process has re sulted in the formation of a diverse collection of discrete and extended st ructures. In its infancy however little was understood about this field and no rational de sign strategies were implemented to construct functional solid stat e materials possessing targeted properties that would make them suitable for applications. This all change d in the late 1980s with the pivotal works by Hoskins and Robson. The following section will provide a scope of the history in this area and highlight select examples of significan t developments that helped to shape this area to what is today. 1.2.1. Inclusion Compounds Metal-cyanide compounds are amongst the earliest report s of MOAs, as evidenced by Prussian blue (PB) compounds and Hofmann clathrates. PB is indeed the oldest synthetic coordination compound. The blue pigmen t was synthesized in Berlin in the early 1700s17 but initially its composition was unknown but it was believed to be a

PAGE 46

9 mixed valence complex consisting of either iron(III) hexacyanoferra te (II) or iron(II) hexacyanoferrate (III). The simplest a nd most reliable approach to deduce this information would be to grow single crystals and obtain the crystal st ructure. PB however has a low solubility which hampers crysta l growth. Keggin and Miles therefore used powder X-ray diffraction (PXRD) to deduce the first cubic structural model (1936).18 Ludi and co-workers (1972) la ter reported the first crystal structure of PB and confirmed that it is indeed a mixed-valance ir on(III) hexacyanoferrate (II) complex.19 The framework is comprised of octahedral meta l centers bridged through linear cyano-based ligands with a general formula of, Fe4[Fe(CN)6]3 .x H2O ( x = 14-16). The intense blue color of PB naturally le d to its utility in dye-related applications ( e.g. inks).20 Host-guest chemistry applications also became apparent after it was discovered that it could act as a molecular sieve; that is, the framework can be fully dehydrated and reversibly adsorb small mo lecules without comprising the structural integrity of the material. Accordingly, PB became known as a prototypical framework because the octahedral metal could be readily substituted for an array of mixed valence transition metal ions and clusters. This ge nerated a multitude of compounds represented by the general formula, Mx[My(CN)6]n .x H2O (Figure 1.2.).21-23 Note that these analogues constitute a family of complex-based ma gnetic materials, many of which possess interesting high-temperature molecular magnetic properties that can be fine tuned through judicious choice of pre-designe d molecular building blocks.24

PAGE 47

10 Figure 1.2. Ball-and-stick representation of two PB analogu es. (left) Fragment of the first crystal structure of a PB analog. Color Code: Mn = orange; Co = magenta; N = blue; and C = gray; and (right) Decorated PB analog comprised of [Nb6(Cl)12(CN)6]4and Mn(II) ions bridged by CN ligands.25 Guest molecules are omitted for clarity. Color Code: Mn = orange; Nb = green; Cl = yellow; N = blue; and C = gray. Research interest pertaini ng to another family of metal-cyanide compounds with unique inclusion properties occurr ed parallel to studies on PB and its analogs, that is, the Hofmann clathrate (1897) named after it s founder Karl Andreas Hofmann.26 The layered network is built-up from altern ating square-planar and octahedr al Ni(II) metal ions. Each cyano group of [Ni(CN)]4is coordinated to the octahedral Ni cation along the equatorial plane through the nitrogen to reveal a layered network comp rised of tetragonal sheets formulated as Ni(NH3)2Ni(CN)4 .2C6H6. The optimal interlayer separation coupled with the relative position of the coordinated NH3 ligands permits the encapsulation of benzene molecules and an assortment of other suitab le guest molecules. Iwamoto and co-workers and others explored an assortme nt of synthetic strategies ( e.g. amine and metal substitutions) to construct novel Hofmann-type analogs (Figure 1.3.).27-29 The proposed strategies were successful and yielded isostructural compounds having the general formula M(NH3)2M’(CN)4 .2G (M = Co, Ni, Cu, Zn, Cd, M n, or Fe; M’= Ni, Pd, or Pt; G = C4H4S, C4H4N, C6H6, C6H5NH2, or C12H10).30

PAGE 48

11 Figure 1.3. Ball-and-stick representations of a Hofmanntype clathrate: (left) Tetragonal sheet comprised of octahedral and s quare-planar metal centers, CdN4(NH3)2 and Ni(CN)4, respectively; and (right) Piperizine molecules trapped betw een the layers. Hydrogen atoms and water molecules have been omitted for clarity. Color C ode: Cd = yellow; Ni = green; N = blue; and C = gray.31 The repertoire of modular solid state ma terials also led to the development of a specific class of inclusion compounds regarded as Werner-type complexes, named after Alfred Werner. Note that he proposed the oc tahedral configuration of transition metal complexes and is largely responsible for de veloping the basis of modern coordination chemistry. He in fact received the Nobel Pr ize in Chemistry in 1913 for his outstanding contributions. This family of com pounds are generally expressed as MX2A4 .2G, where M represents a divalent metal cation which ty pically assumes an octahedral coordination environment ( e.g. Zn, Ni, Cd, Co, Cr, Cu, Fe, Mn, etc ), X represents an anionic ligand ( e.g. CN-, Cl-, Br-, I-, NCS-, NCO-, NO3 -, NO2 -, etc ), A refers to a neutral pyridine-based ligand ( e.g. 4,4-bipyridine, pyrazine, 4picoline, 4-phenylpyridine, etc ), and lastly G refers to a guest molecule trapped in the framework (Figure 1.4.).32, 33

PAGE 49

12 Figure 1.4. Examples of MX2A4 Werner-type complexs: (a) X=Cl-, A=pyridine;34 (b) X= NCS-, A=3,5-lutidine;35 (c) X=NCS-, A=4,4’-azo-bis(4-pyridyl).36 Early serendipitous discoveries c oncerning metal-cyanide compounds and Werner-type complexes clearly exposed unique modular prototypes that are amenable to design and therefore paved the way for fruitf ul opportunities concerning the development of functional MOAs. Nevertheless, the desi gn and synthesis of coordination polymers assembled from pre-designed molecular building blocks (MBBs) was not envisioned until the late 1980s through the re volutionary discoveries repor ted by Robson and Hoskins. 1.3. Rational Assembly of MOMs fr om Expanded Nitrogen-Donor Ligands “One approach to crystal engineering th at we have been developing is first to choose as a geometrical/topological m odel one of a number of simp le 3D nets such as diamond (all centers tetrahedral), -Po (all centers octahedral), rutile (octahedral and trigonal centers in a 1:2 proportions) and so on, and then tr y to revise ways of chemically linking together molecular building blocks w ith a functionality and a stereochemistry appropriate to the chosen net.” Richard Robson37 Robson et al. embarked on a clever sequence of experiments to demonstrate the possibility of rational construc tion of infinite scaffoldinglike structures ( e.g. coordination polymers) built-up from pre-select ed inorganic and organic MBBs.38, 39 The overall concept is rather simple and in hindsight it is somewhat su rprising that it took (a) (b) (c)

PAGE 50

13 experimentalists so long to prove this no tion, given the amount of knowledge available concerning the simplification and topological assessments of nets provided by A. F. Wells.40 That notwithstanding, the node and spacer approach clearly marked the origin of the MBB approach and laid the foundations fo r the discovery of a plethora of predicted novel MOMs of unprecedented scale and functionality.41-44 Simple MBBs ( i.e. regular polygons and polyhedr al) offer a high degree of structural diversity but in th e absence of structure direc ting agents (SDAs) their selfassembly predominately yields high-symmetry networks, termed default nets i.e. preferred by nature. For example, cubic and hexagonal diamond nets are both uninodal 4connected nets comprised of tetrahedra l nodes bridged by linear carbon-carbon rodlike bonds ( i.e. spacers). The default net for the assembly of tetrahedral building units (TBUs) is cubic diamond ( dia ) and thus it is more cumberso me to form hexagonal diamond ( lon ) because it is less symmetric. It was envisioned that a hybrid network assembled from an assortment of predesigned MBBs could genera te similar nets with appreciably larger cavities and windows and consequently render them well suited fo r many applications ( e.g. as catalysts or molecular sieves). Several different synthetic strategies we re proposed to mimic the MBBs in an attempt yield expanded nets with the same underlying topologies. These include, (1) reaction of simple bif unctional linear spacers ( e.g. CN-, 4,4’-bipyridine, etc ) with metal centers that have a preference to adopt tetrahedral coordination environments, and/or (2) vice versa wherein more complex molecules represent the tetrahedral center ( e.g. tetrasubstituted tetraphenyl methane and si lane analogues) connected through linear centers ( e.g. Ag+,Cu+). The versatile coordination geom etries afforded by metal ions

PAGE 51

14 coupled with the repertoire of organic molecu les offers unlimited possibilities for a large collection of discrete and extended MOMs. Meta l-cyanide chemistry in fact offers one of the simplest bifunctional linear molecules, CN-. Reaction of CNwith a divalent metal cation ( i.e. Zn2+, Cd2+) indeed resulted in the formation of the dia network. Despite the modest length of the carbon-nitrogen bond the framework was still interpenetrated and therefore the available free volume was signi ficantly reduced. Cubi c diamond is a selfdual net and therefore catenati on is commonly observ ed, particularly in instances where longer or flexible organic linkers are em ployed. Depending on the desired application however this feature may be regarded as an advantage or a limitation. Studies have shown that the degree of interpenetration can be controlled, to some extent, or completely avoided by fine-tuning the experimental pa rameters. These incl ude, (1) introducing neutral or charged SDAs ( e.g. metal cations, organic templates, counter ions, etc ) into the reaction medium (Figure 1.5.), (2) varying th e length and functionality of the organic spacers, and (3) temperature control in th e case of solvothermal synthesis.

PAGE 52

15 Figure 1.5. (Left) Non-interpenetrating anionic MOM with dia topology constructed from Zn(II) and Cu(I) tetrahedral building bloc ks.TMA cations, shown in spa ce-filling representation, reside in each of the adamantane cavities to balance th e charge. Hydrogen atoms have been omitted for clarity. Color Code: Zn = yellow; Cu = green; N = blue; and C = gray. (Right) Schematic representation of the cubic diamond topology.39 The success of this approach in terms of its feasibility and versatility was extended to include complex buildin g blocks. The deliberate design and in situ synthesis of a cationic dia MOA constructed from the a ssembly of 4,4',4'',4'''tetracyanotetraphenylmethane ligands with Cu(I) cations was a significant achievement by Robson and co-workers (Figure 1.6.). Figure 1.6. Schematic representation the first dia MOA constructed a complex organic MBBs: (a) Tetrahedral metal ion coordination environm ent, Cu(I); (b) Tetrahed ral molecular building block, 4,4',4'',4'''-tetracyanotetraphenylmethane; and (c) Fragment of the {CuI[C(C6H4 .CN)4]}+

PAGE 53

16 MOA All hydrogen atoms and solvent molecul es are omitted for clarity. Color code: Cu = green; C = gray; N = blue.39 The modularity, mild synthetic conditions, and ease of crystallization presented by this class of crystalline materials therefore prompted a wide spread research interest in this area. The innovative crystal engineers s ubsequently expanded thei r studies to include linear neutral nitrogen-donor lig ands such as 4,4'-bipyridine and its derivatives (Figure 1.7.).45-49 The resultant networks in this case are pr edisposed to be cati onic due to the fact that the organic ligands are neutral and the metal cations are positively charged. Thus, charge balance is often provided by ani onic counter ions in the form of BF4 -, PF6 -, NO3 -, SiF6 -, etc located in the framework lattice. Figure 1.7. Examples of polytopic nitrogen-donor ligands: (left to right) imidazole (IMI); hexamethylenetetramine (HMTA); 4,4'-bipyridine (BIPY); 9, 10-bis(4-pyridyl)anthracene; 1,2,4,5-tetra(4-pyridyl)benzene; and 2,4,6-tris(4-pyridyl)-1,3,5-triazine. 1.3.1. Supramolecular Polygons The utilization of linear sp acers with monodentate coordi nation modes dictate that the underlying topology will be governed by th e preferred coordination environment around the metal center. It is possible however to employ chelating moieties ( e.g. EN, 1,10-Phen, 2,2'-BIPY, etc ) to cap the metal centers and thereby control the coordination

PAGE 54

17 environment and direct the topology. Thes e arrangements commonly form molecular squares, triangles, etc The first metal-organic molecular square was in fact reported by Fujita and co-workers in 199050 and since then metal-or ganic polygons akin to an equilateral triangle, square, pentagon and he xagon have become well documented in the literature. Initial studies focused upon reacting linear bipyridinetype ligands with an assortment of transition metals that adopt predictable coordination geometries. The ligand therefore constitutes the edges of the polygon, while the inorganic MBB is positioned at the vertices to define the angle and contro l the overall geometry (Figure 1.8.). Studies were also conducted by Stang and co-workers to pursue the modular nature of this system.51-53 The field therefore expanded to in clude a series of analogous molecular squares, as well as, other molecular polygons sy nthesized using an assortment of diamine capping ligands and ditopic nitrogen-based ligands.54-57 Figure 1.8. (left) First metal-organic molecular square, [Pd(BIPY)(En)]4 8+;51 (right) an example of a molecular triangl e, [Pd(BIPY)(Tmen)]3 6+.55 Hydrogen atoms and counterions are omitted for clarity. Color code: Pd = green; C = gray; N = blue.

PAGE 55

18 1.3.2. Angular Nitrogen-Donor Ligands A tremendous amount of structural dive rsity resulted by intr oducing angularity into the organic compon ents via ligand design, i.e non-linear polytopic pyridyl-based ligands.58-63 Angularity not only imparts directionali ty but also allows for the formation of n -connected nodes where n typically ranges between 2 and 6 and thus building units with different shapes are feasible targets. Nitrogen-based ligands such as pyrimidine, IMI, HMTA and others have proven to facil itate the formation of nets having topologies and properties akin to traditional inorganic zeolites. Substitution of the oxide anions (O2-) for angular organic ligands whilst ma intaining the optim al M-L-M angle ( i.e. close to 145o) permits edge expansion and decorati on and has therefore generated zeolitelike MOFs (ZMOFs) with enlarg ed cavities and channels.62, 64-67 The construction of MOMs from simple TBUs commonly forms dia nets and therefore devising pathways to overcome has proven challenging.68 A possible explanation is likely associated with the labile nature (flexibi lity) of the metal-nitr ogen coordinate bond. 1.3.3. Nitrogen-Based Metal Clusters MOAs constructed from single-metal i ons that are flexible in nature, i.e. bipyridine-type coordination polymers, are of ten less robust than metal clusters and therefore exploiting their porosity for desired applications can be ch allenging. Alterative pathways have therefore been recently de vised to alleviate th is issue by employing polytopic nitrogen-based ligands that are capable of coordinating to metal ions in a bismonodentate fashion to generate rigid metal clusters (Figure 1.9.). This ligand class belongs to the azole family a nd is therefore identified as 5-membered hetercyclic nitrogen

PAGE 56

19 containing systems. These include pyrazoles (N2C3R), triazoles (N3C2R), and tetrazoles (N4CR).69-75 Figure 1.9. (a) 3-oxo-centered trinuclear cluster, [M3O(N3CR)3(H2O)3]; (b) Dinuclear paddlewheel-like cluster, [M2(N2CR)3(H2O)6]; (c) M4Cl(N4CR)8L4 tetranuclear cube-like cluster; (d) [M3(N4CR)6(H2O)6 trinuclear cluster; (e) [M3O(N2CR)3] trinuclear cluster; (f) Prototypical example of a 3-periodic sod -type MOM constructed from a M4Cl(N4CR)8L4 (M= Cu or Mn) clusters and a triangular ligand, H3TPT-3 tz Hydrogen atoms and solvent molecules have been omitted for clarity. Color Code: M = green; C = gray; N = blue; Cl = yellow; and O = red.76 Note: The yellow sphere located inside the cavity represen ts the largest sphere that can fit inside taking into account van der Waals radii. 1.4. Construction of MOMs fr om Carboxylate-Based Ligands The last decade has witnessed an expl osive increase in the synthesis and characterization of MOMs, particularly fr ameworks constructed from carboxylate-based organic ligands (Figure 1.10.).77-87 The wide spread interest in this class of linkers is due (a) (b) (c) (d) (e) (f)

PAGE 57

20 in part to (1) their ability to accommodate various coordination modes e.g. bidendate, monodentate, bis-monodentate, et c; a feature that is not pe rmitted with nitrogen-based bipyridineor cyano-type ligands, (2) ability to form rigid and directional metal clusters with fixed geometries by conforming to th e bis-monodentate coordination mode, and (3) the negatively charged carboxylate groups ( i.e. RCO2 -) often preclude the need for charge balancing counter ions. Figure 1.10. Select examples of polytopic carboxylate-based organic ligands used to construct MOMs. A few of the commonly employed metal carboxylate clusters ( i.e. generated in situ ) that have proven to be usef ul in the construction of MOFs are illustrated in Figure 1.11. Each metal cluster can be simplified into a geometric entity by connecting the points of extension, i.e. via th e carboxylate carbon atoms, to reveal a so-called secondary building unit, which is a term borrowed from zeolites. A large number of discrete metal clusters are in fact available in the chemis t toolbox but the organi c linker must have the appropriate functionalities built -in to facilitate the requisite geometry and rigidity upon

PAGE 58

21 coordination. Prior to isolating a cluster in an extended structure, the next step is to identify the appropriate synthe sis conditions which c onsistently lead to its formation once the MBBs have been modified, i.e. to produce structures having the same underlying topology but different pore metrics and functiona lity. This step has indeed proven to be a contemporary challenge but once this step is accomplished then the design of target networks can truly be envisioned. Figure 1.11. Examples of rigid metal carboxylate-based MBB clusters used to construct MOFs, each of which can be translated into geometri cal shapes [SBU(s)]: (a) Dimetal tetracarboxylate cluster, often regarded as the “paddlewheel” M2(RCO2)4L2 forms either a square or octahedral SBU; (b) Basic chromium acetate trimetal cluster M3(RCO2)6L3 represents a trigonal prismatic SBU; and (c) Basic zinc acetate cluster, Zn4O(RCO2)6 forms an octahedral SBU. Color Code: C = gray; O = red; M= green. The versatile and predictabl e nature of the MBB approach is still preserved in the case of carboxylate-based ligands whereby differe nt ligands can be utilized to construct discrete and extended structures with fine-t unable pore sizes, shapes, and functionality. An excellent example to illustrate the concept of reticular chemistry is represented by the

PAGE 59

22 archetypical MOF-5 (or IRMOF-1) synthe sized by Yaghi and co-workers from prefabricated basic zinc acetate,[Zn4O(RCO2)6], and ditopic 1,4-BDC MBBs that reticulate to form a primitiv e cubic structure (Figure 1.12.a).84 Figure 1.12. Two examples of IRMOFs constructed from linear carboxylate-based ligands and the 6-connected basic zinc acetate, [Zn4O(RCO2)6] metal cluster: (a) MOF-5 (IRMOF-1); and (b) IRMOF-11. Hydrogen atoms and solvent molecule s have been omitted for clarity; Color Code: Zn = green; C = gray; and O = red. The attractive structural features of MOF5; that is, being exceptionally rigid and highly porous provided a unique prototype to embrace a conglomeration of novel properties through the synthesi s of isoreticular analogs (Figure 1.12b). Note that hydrogen sorption studies conducted on MOF-5 were the first to be reported for this class of materials. It therefore paved the way for the development of sixteen additional IRMOFs through variation of the chemical composition, functionality, and dimensions. Kitagawa and co-workers on the other hand we re the first to report the adsorption of small gaseous molecules such as CH4, N2, and O2 gases at ambient temperature (1997),88 while Eddaoudi and co-workers were th e first to confirm and study permanent (a) (b)

PAGE 60

23 microporosity with MOF-2 using N2 and CO2 gas adsorption measurements at 78 K and 195 K, respectively.89 The dimetal tetracarboxylate “paddlewheel” MBB, [M2(RCO2)4L2],90 is a ubiquitous inorganic MBB in crystal chemistr y (see chapters 3 and 4). The first example of a polymeric material assembled from linked paddlewheel MBBs was reported by O’Connor and Malsen (1966) whereby a 1-pe riodic coordination polymer containing succinic acid of formula, [Cu2(succinate)2(H2O)2]n.91 Yaghi and co-worke rs later reported the synthesis of MOF-289 and then in 1999 the synthesi s of HKUST-1 was reported by Williams and co-workers.87 From this point onward the field of MOF chemistry concerning paddlewheel MBBs was transformed ( i.e. in extended structures), due in part to the outstanding contributions from O. M. Yaghi and M. J. Zaworotko. In 2001 these groups independently repor ted on the design and synthesis of a discrete spheroidlike metal-organic polyhedral (MOP) bu ilt-up from copper paddlewheel MBBs and m -BDC units, coined the nanoball and MOP-1 by Zaworotko et al.80 and Yaghi et al.,92 respectively (Figure 1.13.). The modular nature afforded by this prototype lead to the development of a series of decorated nanoscale analogs. This was accomplished by functionalizing the 5-position of the m -BDC moiety with various substituents ( i.e. SO3 -, -OH, -OCH3, -OC12H25, etc ).93-95 It was also demonstrated that extended suprasupermolecular networks coul d be isolated by employing the metalorganic small rhomibhexahedra as nanoscal e nodes (supermolecular building blocks, SBBs).94, 96, 97

PAGE 61

24 Figure 1.13. (a) Ball-and-stick representation of the neutral nanoball or MOP-1, [Cu2( m BDC)4]12; (b) Schematic representation reveals a small rhombihexahedron consisting of vertexlinked square SBUs. The 5-position of the m -BDC unit lies on the vertices of the SBB and is highlighted in orange. Color Code: C = gray; O = red; and Cu = green. All hydrogen atoms and solvent molecules are omitted for clarity. The 3-oxygen-centered TMBB, [M3O(RCO2)6(L)3], is an ideal rigid and directional MBB used to target nets whos e vertex figures indicate the need for a 6connected node having trigonal prismatic ge ometry. Note that oxobridged trimers are common in transition-metal chemistry, i.e. for discrete complexes, but are less documented in the case of exte nded networks, particularly for p -block metal ions (see Chapter 2).98 Frey and co-workers have in fact e xploited the diversity of the TMBB with an assortment of transition metal cations and organic struts to isol ate robust MOFs with interesting properties, e.g. high porosity a nd breathing effects as a consequence of framework flexibility. Perhaps most notab ly, are the two highly porous MIL compounds with zeotypic giant pores and augmented mtn topology, coined MIL-100 and MIL-101 (M = Materials of Institut Lavoisier).85, 86 In the crystal struct ure of the former the chromium-based TMBBs are linked through 1,3,5-benzenetricarboxylic acid (BTC), while in the latter BTC is substituted for 1,4-benzenedicarboxylic acid (BDC). This arrangement therefore reveals a supertetraheral building block (Figure 1.14.). Both MIL-

PAGE 62

25 100 and -101 have remarkably large cage dimensions of 25 and 34 respectively and huge Langmuir surface areas of 3100 m2/g and 5600 m2/g, respectively and thus are good candidates for many applications. Figure 1.14. Select fragments from the crystal structures of (a) MIL-100; and (b) MIL-101, which form a supertetrahedral building block th at can be rationalized as a super tetrahedral building unit (center). 1.5. Hybrid Ligand Design: MOMs Construc ted from Hetero-Functional Linkers The previous sections highlighted th e modularity offered by MOMs as a function of ligand design, i.e. independently em ploying nitrogenand oxygen-based organic ligands. A natural transition is therefore to combine these functional groups into a single entity to form a hetero-functional organic linker (Figure 1.15.). The existence of more than one type of heteroatom or functiona l group is considered advantageous because these moieties offer the potential to facilitate the formation of rigid and directional mixed metal-ligand coordination environments and t hus different geometri es and shapes can theoretically be isolate d. Monoor polytopics ligands comp rised solely of either nitrogenor oxygen-donor atoms predominately bond to metal cations yiel ding one type of coordination environment and therefore uninodal and binodal networks are commonly observed.

PAGE 63

26 Figure 1.15. Select examples of hetero-functional lig ands with the possible directionality indicated by the black arrow. Top (from left to right): 4,6-pyrimidinedicar boxylic acid; pyridine3,5-bis(phenyl-4-carboxylic acid); and 4,5-imidazoledicarboxylic acid. Bottom (from left to right): 2,3-pyrazinedicarboxylic acid; 4-pyrimidinecarboxylic acid; 2,5-pyridinedicarboxylic acid; 3-pyridinecarboxylic acid; and 4-imidazolecarboxylic acid.65, 99-104 Hetero-functional ligands can sustain a va riety of building blocks and therefore can be used to target more complex architectures ( e.g. tertiary or even quaternary nets). Eddaoudi and co-workers in fact isolated a novel 3-perioidic tertia ry net by reacting a bifunctional tetrazolate ligand, 5-tetraz oleisophthalic acid (TZI) with Cu(NO3)22.5H2O. The cationic 3-periodic MOF is built-up from two types of inorganic MBBs and an organic MBB (Figure 1.16.). Viewing the st ructure as the assembly 24-connected rhombicuboctahedral building units linked together thr ough 3-connected trigonal nodes results in a (3,24)-connected MOF with rht topology (see chapter 4).97

PAGE 64

27 Figure 1.16. Illustration of the (3,24)-connected rht -MOF: (a) Hetero-functional ligand yields a trigonal [Cu3O(N4CR)3] MBB and [Cu2(RCO2)4] paddlewheel MBBs; (b) Tiling representation. Hetero-functional ligands possessing chel ating moieties, i.e. carboxylate and nitrogen groups, can also be used to cont rol the coordination e nvironment around singlemetal ions. This concept involves saturation of the metal center through the formation of hetero-chelating rings. This t ype of arrangement precludes the coordination of unwanted solvent/guest molecules and therefore renders the single-metal node rigid and directional. Design strategies pertaining to the single-me tal-ion-based MBB approach was introduced by Eddaoudi and co-workers as a reliable to target r obust metal-organic assemblies (MOAs) with single metal ions located at the vertices.105 This concept is based on the fact that the aromatic nitrogen atoms w ill direct the framework topology, while the carboxylate moieties located in the -position relative to the nitrogen will complete the coordination sphere and lock the metal ion into position through the formation of 5membered chelating rings. The resultant si ngle-metal-ion based M BBs are therefore of the type MNx(CO2)y where M represents a 6 to 8 coordi nate metal cation, x refers to the number of heter-chelating moie ties, and y refers to the num ber of ancillary (bridging) (a) (b)

PAGE 65

28 donor atoms. The success of this approach is evidenced by the large number of structures reported by Eddaoudi and co-workers, which ha ve been deliberately constructed from predesigned rigid and directional MBBs usi ng hetero-chelating ligands. For example, a discrete octahedron (M6L12), metal-organic cubes (M8L12), 2-periodic kagom lattice, and a 3-periodic diamondoid net.99, 106 A particular subset of MOMs ( i.e. ZMOFs) deserves sp ecial atte ntion. The versatility of the single-metal-ion-based M BB approach permitted the assembly of nondefault networks with topologies akin to inorganic zeolitic ( e.g. rho -, sod -, usf -, ast -, aco -, and lta -ZMOFs).64-67, 100 This was carried out by employing judiciously chosen MBBs that translate into 4-c onnected tetrahedral building unit s. The rigid and directional TBUs adhere to the desired M-L-M angle of ~145o observed in zeolites. Decoration and/or expansion via the organi c functionality affords extra-large cavities that are finetunable (Figure 1.17.). Figure 1.17. Ball-and-stick and schematic representations of (left) rho -ZMOF comprised of 8coordinate InN4O4 MBBs (= InN4 TBUs) and (right) sod -ZMOFs comprised of 6-connected InN4O2 MBBs (= InN4 TBUs). Both compounds were cons tructed using an angular ditopic hetero-functional ligand, 4,5imidazoledicarboxylic acid (H3-ImDC).64 +

PAGE 66

29 1.6. Identification and Classification of MOMs: Topological Descriptors Several notations have been implemented over the years to classify nets, many of which are widely reported to date. These te rms however are used interchangeably and thus misused by researchers, as exemplif ied in hundreds of manuscripts published in ACS and RSC journals since 2000.107 The following sub-chapters will provide an overview of the commonly used topological notations and symbols. An outline of their advantages and limitations will be provided accompanied by relevant examples. 1.6.1. Point Symbols and Schlfli Notation The topological descriptors used to descri be a net are derived from graph theory and therefore in a mathematical sense a ne t can be rationalized as a special kind of graph.108 In the context of MOMs, researchers commonly report the growth of a net as being 0-dimensional (discrete) 1-dimensional, 2-dimenstional, or 3-dimensional. For example, metal-organic polyhedra (MOPs) are often regarded as 0-dimensional architectures, while the metal-organic analog of the dia net is viewed as a 3-dimensional structure. This convention is in fact mi sleading because all atoms and molecules are indeed 3-dimensional. It is theref ore more precise to use the term n -periodic to refer to the independent directions of a net and t hus MOP would be regarded as 0-periodic structures. Early topological assessment s of nets were primarily developed by A. F. Wells, who analyzed and classified a great number of nets based on their connectivity using a simple ( n, p ) notation. This description is rela ted to geometrical principles and mathematical descriptors.40, 109-114 Accordingly, Wells dissected a net into the assembly of

PAGE 67

30 nodes (vertices) connected through spacers (edges) and tried to identify the polygons that were apparent as a consequence of these linkage s. Note that a metal ion is often regarded as a node and the organic ligand represents the edge (bond) to link a pair of vertices and therefore permits the growth of the ne t. In cases of higher coordination ( e.g. metal clusters) it is often easier to simplify the group of nodes into a single node (Figure 1.18.). The connectivity of the vertices is depe ndent upon the node and spacer assignment and therefore more than one topological outcome may be possible for a given structure. Figure 1.18. (left) Tetravalent basic zinc acetate cluster, [Zn4O(RCO2)6] simplified into a 6connected node (right) via the points of extension. In Wells’ notation, the n term refers to the number of edges of the faces of a polygon in the net and p refers to the number of edges th at meet at each of the vertices. For example, a 2-perioidic tetragonal sheet could be given the symbol (4,4) which implies a 4-connected net comprised of 4-membered rings (MRs) that meet at each vertex (Figure 1.19a.). A honeycomb lattice is therefore assigned a symbol of (6,3) because it is a 3-connected net with the shortest ring at each angle being a 6-MR (Figure 1.19b.).

PAGE 68

31 Figure 1.19. Representation of the ( n,p ) classification system for 2-periodic lattices: (a) (4,4) Tetragonal sheet; and (b) (6,3) Honeycomb lattice. The ( n,p ) system is quite efficient for simple nets but it has limitations in the sense that it does not uniquely describe a ne t. It also becomes less reliable and more complicated for higher connected nets. Wells wa s aware of this shortcoming and in some instances assigned a letter to follow the ( n,p ) nomenclature as a way of providing a topological distinction. For example, Wells refe rred to the 3-connected nets of the Si atoms in SrSi2 ( srs ) and ThSi2 ( ths ) structures as (10,3)-a and (10,3)-b, respectively. Although the use of a letter descriptor is useful, it does not pr ovide any topological distinction between the nets and thus more descriptive approaches were needed. Wells therefore introduced a topologi cal descriptor that he called a point symbol In this sense, each net is comprised of a certain number of nodes and each n -connected node has N=[ n ( n -1)/2] angles surrounding it. The shortest cycle between the first and last vertex defined as (x1, x2, …, xn-1, x1) must be identified at each angle. A point symbol is expressed in the form of Aa.Bb.Cn…. etc where A < B < C, and so forth. The upper-case letter denotes the size of the shortest cycle circuits contained at each angle and the superscripts refer to the number of shortest circuits at that specific angle. For example, the uninodal 4-connected dia net has 6 equivalent angles at each node and the shortest (a) (b)

PAGE 69

32 cycle circuit is 6 and therefore the point symbol is 66. Likewise, the aforementioned (10,3)-a net would be assigned a point symbol of 103 because it has 3 equivalent angles at each node and the shortest cycle circuit is 10. Now is an appropriate time to introdu ce another symbol widely used in the literature, Schlfli symbols which appear in the form of a { p q r …} notation and are applied to regular 2and 3-periodic tilings.57, 115 In its simplest form, a p -sided regular polygon is represented as { p }. For example regular polygons such as triangles, squares, and pentagons would be assigned a Schlfli symbol of {3}, {4}, and {5}, respectively. Whereas, a regular polyhedron would be represented by the notation { p q } because it is assembled from q -regular p -gon faces that meet at each vertex. An excellent example to illustrate this notation is represented by the five Platonic solids which include: tetrahedron {3,3}, cube {4,3}, octahedron {3,4}, icosahedron {3,5}, and dodecahedron {5,3}. A tetrahedron accommodates three triangles at its vertices, while three squares meet at each of the vertices in a cube a nd therefore have Schlfli symbols of {3,3} and {4,3}, respectively (Figure 1.20.). Figure 1.20. Schematic representation of two Platonic solids: (left) Tetrahedron; and (right) Cube. Similarly, a tetragonal sheet and honeyco mb lattice would be assigned a symbol of {4,4} and {6,3} because each plane is assembled from tiles of squares and hexagons,

PAGE 70

33 respectively. This notation can also be applied to regular 3-periodic ti lings in the form of a { p q r } Schlfli symbol whereby r { p q }’s meet at each edge. For instance, {4,3,4} would symbolize a cubic lattice comprise d of four cubes around each edge and corresponds to the primitive cubic ( pcu ) net. Unfortunately, point symbols and Schlfli symbols are often applied interchangeably in the literature which le ads to confusion and misrepresentation. A recent article by O’Kee ffe and co-workers addresses this issue and the authors urge researchers to restrict the usage of Schlfli sy mbol to regular tilings in 2and 3-perioidc nets, while point symbols s hould only be applied to 3-periodic nets and not used interchangeably with Schlfli symbol or circuit symbol.116 Although the aforementioned notations are quite useful, they are accompanied by limitations. The problem arises when one must distinguish between two or more nets having the same point or Schlfli symbol. 1.6.2. Extended Schlfli Notation and Vertex Symbols In an attempt to resolve the ambiguity between similar nets, a more descriptive group of terms were introduced, i.e. long (exten ded) Schlfli symbols and vertex symbols (O’Keeffe). Both notations are widely reported in the literature but the key distinction lies in the use of the term “ cycle” versus “ ring” In contrast to a cycl e, a ring is a more specific type of cycle which is not the sum of two smaller cycles and does not contain any shortcuts (Figure 1.21.). 116

PAGE 71

34 Figure 1.21. A schematic to illustrate the difference betw een strong rings, rings, and cycles: (left) 4-cycle that is a strong ring, (middle) 12-cycle that is the sum of a 6and 8-cycle, thus is not a ring, and (right) A 12-cycle that is the sum of a 6-, 8-, and 4-cycle but not of just 2 cycles and therefore is a ring (not a strong ring). The sequence is expresse d in the form of Aa.Bb.Cn…. etc in both cases where the upper-case letter refers to the size of the “shor test cycle” (extended Schlfli symbol) or “ring” (O’Keeffe) contained at each angle and the subscrip ts refer to the number of shortest cycles / rings at the respective a ngle. The key distinction to be made here between short and long notation is the latt er computes the numb er of cycles at each of the respective angles. The use of vertex symbol s is currently the most informative and commonly used notation to cla ssify MOMs because the extende d Schlfli notation can be misleading. Because a cycle is not always th e shortest distance between the starting and ending point and can lead to inaccurate results. The vertex symbol computed for each node often provides enough information to make a decisive topological distinction. Fo r example, Wells’ (10,3) -a and (10,3)-b nets can be uniquely identified through the use of vertex symbols: 105.105.105 and 102.104.104, respectively. Yet there are excepti ons, as in cubic (crystabolite; dia ) and hexagonal (longstalite; lon ) diamond which both are uninodal 4-c onnected nets possessing the same vertex symbol of, 62.62.62.62.62.62.117 This notation implies that each of the 4-connected nodes are surrounded by 6 angles and each angl e contains two 6-membered rings that meet at the vertex but does not provide e nough information to uniquely identify the net.

PAGE 72

35 1.6.3. Coordination Sequence In order to topologically distinguish between the dia and lon nets, one must further determine the coordination sequence (CS) of each node. According to O’Keeffe et al. a CS consists of a sequence of numbers ( n1, n2, n3, …, nk …) in which the k th term is the number of nodes in the shell k that are connected to nodes in shell k -1. In short it refers to the number of ne ighbors surrounding each node as th e structure grows. The CS of dia up to k = 10 is 4, 12, 24, 42, 64, 92, 124, 162, 204, 252,..etc (Table 1.2.). In this case, the [CS(0)] refers to the 4-connected node which has a set of 4 nearest neighbors, this subsequently generates the first shell [C S(1)]. These nodes are then linked to set of 12 nearest neighbors, further extended to 24, a nd so on. There is no specific limit for the k th term but typically computer softwa re programs determine this value up to k = 10. The sum of the CS is regarded as the topological density (Td10) and reveals the total number of vertices in the k th subshell. The computed coordination sequence for dia and lon up to k = 10 reveals a difference at the third ne ighbor, thereby illust rating the topological difference between these nets. The topological density determined for each net, reveals lon is denser than dia having values of 1027 and 981, respectively. Table 1.2. Coordination sequence computed up to k = 10 for the cubic and hexagonal diamond nets.117 Symbol VS cs1 cs2 cs3 cs4 cs5 cs6 cs7 cs8 cs9 cs10 td10 dia lon 62.62.62.62.62.62 62.62.62.62.62.62 4 4 12 12 24 25 42 44 64 67 92 96 124 130 162 170 204 215 252 264 981 1027 Difference 0 0 1 2 3 4 6 8 10 12

PAGE 73

36 Likewise, in 2003, a similar observation was reported by Zaworotko and coworkers between two fundamentally di fferent supramolecular isomers of [M2(bdc)2(L)2]n. Both nets possess the same novel point symbol (65.8) and vertex symbol (6.6.6.6.6.*) but were determined to have a different CS. Note th at an asterisk is present in the VS because no ring is contained at that a ngle but there is always a cycl e, hence the point symbol of (65.8). An in depth topological study confirmed that one of the nets had a similar connectivity to the 4-connected CdSO4 net and was classified as having cds topology. Interestingly, the other net was identifi ed as adopting an unprecedented network topology, coined usf1.118 Vertex symbols are more informative topol ogical descriptors th an point symbols. Nevertheless, for nets that possess higher c oordination numbers it is much more practical to report the point symbol as oppose to the vertex symbol. For example, Eddaoudi and co-workers were the first to report a (3,24)-connected MOM having rht topology and there are 276 angles around the 24-connectcted node!!! As you can imagine, reporting the vertex symbol for this node would not only be time consuming but extremely complex. It is therefore more appropriate to report the point symbol, (43)8(472.6132.872). Alternatively, similar nets may have the same coordination sequence but different vertex symbols. This is attested in two zeolite nets, lta (Linda A) and rho The CS for both nets up to k = 10 is 4, 9, 17, 28, 42, 60, 81, 105, 132, 162, and 641; theref ore both nets have a topological density of 641. However, the nets are indeed unique as they possess different vertex symbols of 4.6.4.6.4.8 and 4.4.4.6.8.8 for rho and lta respectively.

PAGE 74

37 1.6.4. Tiling and Transitivity A powerful tool implemented for the system atic enumeration of structures is to describe the net in terms of the tiling that it carries because each tiling can encode valuable information ( e.g. transitivity, natural tiling, duals, etc ), all of which contribute to the understanding of existing nets and development of future target structures. Tiling representations for 2-peri odic lattices appear in the form of edge-to-edge tiles that cover a plane, for instance covering a plane with e quivalent hexagons would reveal a regular tiling for the (6,3) honeycomb lattice. Tilings for 3-periodic nets on the other hand consist of an assortment of face-to-f ace tiles that fill space through the formation of generalized polyhedra or cages (Figure 1.22.). A visualizatio n of the tiling that carries at net can provide a clearer picture of the open channels and cages present in the structure; however, in complex networks the tiling may be less appropriate a nd look like nothing more than a collage of colors and shapes. Figure 1.22. Tiling representation and face symbols for two zeolite nets: (left) lta 3[46] + [46.68] + [412.68.86]; and (right) rho 3[48.82] + [412.68.86]. Most nets adopt a so-called natural tiling which consists of the smallest possible tiles and thus it is uniquely suited to desc ribe a particular net. Computer software

PAGE 75

38 packages are available to determine (TOPOS)119 and visualize (3dt)120 the natural tiling by applying the following restrictions in the calculation: (a) each tile must carry the maximum symmetry for that net and (b) all f aces must represent st rong, essential rings. Consequently, a face symbol is generated for each tiling which reveals the ratio and composition of the tile in the form of [Aa.Bb…. etc ]. One may recall that this notation looks quite similar to that used for point symbols. Nonetheless, the meaning is quite different and refers to a faces that are A-membered rings (MRs), b faces that are B-rings, etc For example, the face symbol for the lta net could be broken down as follows: 3[46] + [46.68] + [412.68.86] the tiling contains three different tiles in a ratio of 3:1:1. The largest tile, an -cage, is assembled from twelve 4MRs, eight 6-MRs, and six 8-MRs. Each of the 6-MRs are shared faces with a -cage comprised of six 4-MRs and eight 6MRs. Likewise, the 4-MRs of the -cage share all faces with the cube tile comprised of six 4-MRs. A valuable property obtained from tiling is a measure of its regularity which is categorized in terms of transitivity, pqrs This notation specifies the number of kinds of vertices ( p ), edges ( q ), faces ( r ), and tiles ( s ) in a 3-periodic netw ork. Alternatively, since 2-periodic nets do not contai n tiles, it is expressed as pqr MOMs that possess lower numerical transitivities have a higher degree of regularity a nd are therefore regarded as highly feasible targets in crystal chemistry and are easier to acce ss synthetically. Thus, the assembly of flexible building blocks, in the absence of structure directing agents, frequently leads to the formation of a so called, default net.121 Each vertex in a net can be simplif ied into a geometrical building unit by replacing a single vertex with a group of ve rtices (augmentation) by connecting the points

PAGE 76

39 of extension to reveal a vertex figure/coor dination figure. The most regular MOMs have transitivities of 1111 which means they are ve rtex-, edge-, face-, and tile-transitive. The vertex figures of these nets correspond to regular polygons or polyhe dron, specifically a triangle, tetrahedron, square, octahedron, or cube. There are exac tly five regular 3periodic default nets: srs (triangle), dia (tetrahedron), nbo (square), pcu (octahedron), and bcu (cube). Alternatively, the 12-coordinated fcu net has a transitivity of 1112 because it is comprised of a quasiregular polyhedron (cuboctahedron) which generates two kinds of holes in the stru cture and leads to the forma tion of two different tiles – tetrahedron and octahedron. Semi-regular nets are vertexand edge-transitive, represented by a transitivity of 11rs with r > 1. Edge-transitiv e nets are attractive targets in crystal chemistry. In f act, many MOMs are assembled from two different vertex figures linked together through a common edge.121 1.6.5. Reticular Chemistry Structure Resource (RCSR) Database The RCSR117 is a comprehensive database developed by O’Keeffe and coworkers. It is dedicated to facilitating the rapid development of reticular chemistry and is comprised of over 1600 entries from known ch emical compounds and crystal structure predictions ranging from 0to 3-periodic architectures. RCSR is modeled after the zeolite database and therefore a th ree-letter nomenclature system was adopted. The key distinction between these code s is that all RCSR symbols are lower case bold as oppose to the upper case symbols used in the zeolite database.122 The symbols help to alleviate confusion between nets because nets ar e associated with different names ( e.g. Laves net, (10,3)-a, Y*, etc ). The RCSR letter code assigned to this uninodal 3-connected net is

PAGE 77

40 therefore srs because it is the Si net in SrSi2. When applicable, the nets are assigned a symbol based on their relation to zeolites, for instance LTA (Zeolite-A) is given the RCSR symbol lta Many of the symbols reported in th e database are in fact derived from their relation to minerals such as diamond ( dia ) Niobium Oxide ( nbo ) Platinum Sulfide ( pts ) Quartz ( qtz ) etc In special circumstances how ever nets with unprecedented topologies maybe assigned a symbol that is named after a person, such as the med net, named after Professor Mohamed Eddaoudi. The database is user friendly whereby a known crystal net can be located either by searching for the symbol, name, keyword(s), modifiers, and/or bounds ( i.e. density, td10, kinds of vertex, edge, face, etc ). Perhaps most importantly, each entry is unique and reveals valuable crystallographic information ( i.e. unit cell parameters) and topological details including the vertex symbol and coordination sequence for each crystallographically independent node. Furtherm ore, information pertaining to the natural tiling and transitivity of a net is available fo r 3-periodics structures accompanied with a file in a coordinates of verti ces in faces (cdg) that can be visualized using a useful tiling and analysis program (3dt) desi gn by Olaf Delagado-Friedrichs.120 1.7. Properties and Potential Applications of MOMs The unique structural and chemical feat ures that are char acteristic of MOMs separate them from other so lid state materials and so the field has witnessed an exponential growth concerning in terest in their ph ysical and chemical properties. The attractive features include but are not limite d to, (1) hybrid inorganic-organic components allow for a high degree of modularity where by the choice of metal ion, as well as, the

PAGE 78

41 length and functionality of the organic linker can often be mo dified while maintaining the geometry of the MBBs; (2) varying degrees of dimensionality are possible, e.g. discrete, 2-, and 3-periodic; (3) decora tion and expansion of traditi onal zeolite nets topologies using the single-metal-ion-based MBB appro ach has proven to yield ZMOFs with finetunable pores and enlarged cav ities; (4) many open structures are known to exhibit high surface areas, pore volumes, and low framewor k densities; (5) open metal centers have shown to enhance H2-MOF binding interactions and thus lead to higher Q st (at low loading); (6) typically facile and inexpensive synthetic protocols with high yields; and (7) many structures are thermally and chemically stable. Academic and industrial researchers from across the world have therefore placed a particular focus on investigating the potenti al utility of these materials in desired applications. The national meeting of the Am erican Chemical Society in fact held a special symposium in 2008 comprised of invited speakers devoted to th is topic, entitled: “Metal-Organic Frameworks: What are they good for?” A themed issue of Chemical Society Reviews (Title: MetalOrganic Frameworks) also emphasized this topic whereby 18 peer-reviewed articles were published concerning design principles, molecular modeling, and various applications of which MOMs are well suited.123 Examples of some applications included, gas storage ( e.g. H2, CH4, etc), selective gas adsorption and separation, drug delivery, catalysis, post-syn thetic modifications, sensing, magnetism, and thin films. Studies concerning the stor age and binding affinity of hydrogen, methane, and carbon dioxide gases in select ed MOFs is the focus of th is dissertation. Accordingly, details concerning other applic ations are beyond the scope of this discussion and can be found in the cited themed issue on MOFs and references therein.

PAGE 79

42 1.8. Potential Clean Energy Alternative: Bridging the Interface between Hydrogen Storage and On-Board Applications The rapid expenditure of non-renewabl e petroleum-based fossil fuels coupled with escalating environmental concerns ( e.g. global warming and air pollution) has drawn considerable interest worldwide on a both scien tific and societal level. The industrial and transportation sectors are the two primary c onsumers of fossil fuel s and therefore this demand is not expected to subside in the fu ture as the global p opulation continues to increase. The United States currently relies heavily on foreign countries to fulfill this quota and from an economical perspective this is dangerous because it can lead to price fluctuations and supply disruptions.124 This dilemma has prompted researchers from a range of disciplines to work together towa rds the rational development of alternative renewable energy technologies that will alle viate the harmful side effects brought by the consumption of fossil fuels, yet preserve or even exceed their optimal energy output. Vehicles powered by battery, hybrid, and fu el-cell technologies are currently used by many individuals to reduce their carbon f ootprint. This technology certainly offers improvements over traditional methods but it is not an ultimate solution to the problem because fuel-cells partially rely on the use of fossil fuels to power the internal combustion engine. Nissan however plans to launch the LEA F this year which is a 100 % electric car and therefore has zero emissions. It has an estimated driving range of 100 miles before recharging and is thus well suited for city commuters.125 An attractive energy alternative that has gained much momentum in recent years is the use of hydrogen as an energy carrier for pot ential on-board applications.126-128 Although hydrogen is the simplest and most abundant chemical element on earth, less than 1 % is fact available in the form of a pure gas.126, 129 It is therefore often extracted

PAGE 80

43 from domestic resources ( e.g. water, biomass, hydrocar bons ) by means of chemical processes.130 The use of hydrogen as an energy ca rrier offers many advantages over gasoline and diesel fuel, perhaps most notab le that it is 100 % car bon-free which means it is clean burning, with only water as a byproduct.131 It also has an outstanding gravimetric mass energy density which is almost three times that of liquid hydrocarbons: 142 MJ kg-1 versus 44.5 MJ kg-1.132, 133 Significant drawbacks however hinder its utility especially with regards to identifying cost-effective and safe stor age technologies. Note that molecular hydrogen exists as an extremely volatile gas at ambient temperatures and therefore it has a very low volumetric energy de nsity, which is much too low for practical applications ( i.e. dH2 = 0.08988 kg m-3 at STP versus gasoline ~700 kg m-3). To put this number into perspective, 5 kg of hydr ogen gas occupies a volume of 60 m3 at room temperature and atmosphere pressures! In order to facilitate the potential wide spread commerc ialization of hydrogenfueled light weight vehicles the Bush ad ministration implemented the Hydrogen Fuel Initiative in 2003. The US Department of En ergy (DoE) introduced the Energy Efficiency and Renewable Energy (EERE) and Fuel Cells Technologies (FCT) program to provide funding to universities, as well as, industrial and government agencies in order to accelerate the research and development in this area.134 The program focuses on many factors including hydrogen production, deliver y, storage, infrastructure, and safety concerns. The DoE has therefore developed a sp ecific set of technical targets (Table 1.3.) that span across numerous storage parameters.

PAGE 81

44 Table 1.3. Select “revised” technical targets (Septemb er 2009) as set forth by the US Department of Energy (DoE) for on-board hydrogen storage systems in light-duty vehicles.135 Storage parameter Units 2010 2015 Ultimate System Gravimetric Capacity: Usable, specific-energy from H2 (net useful energy/max system mass) kWh / kg (kg H2 /kg system) 1.5 (0.045) 1.8 (0.055) 2.5 (0.075) System Volumetric Capacity: Usable energy density from H2 (net useful energy/max system volume) kWh/L (kg H2/L system) 0.9 (0.028) 1.3 (0.040) 2.3 (0.070) Storage system cost and fuel cost $/kWh net $/kg H2 $/gge at pump TBDa TBD 3-7 TBD TBD 2-6 TBD TBD 2-3 Durability/Operability • Operating ambient temperature • Min/max delivery temperature • Cycle life (1/4 tank to full) Cycle life variation • Maximum delivery pressure from a tank • Refueling rate for 5-kg of H2 C C Cycles bar (abs) min (kg H2/min) -30/50 -40/85 1000 12 4.2 (1.2) -40/60 -40/85 1500 12 3.3 (1.5) -40/60 -40/85 1500 12 2.5 (2.0) a To be determined. These targets serve as criteria for candidate mate rials if they are to be used as alternative fueling systems for on-board hydrogen storage systems in light-duty vehicles. Every five years the targets are re-examine d and adjusted according to th e progress made in terms of the development. If a material was to meet or exceed all of these parameters it would essentially provide the same effici ency as current gasoline tanks.

PAGE 82

45 1.8.1. Conventional Hydrogen Storage Tec hnologies: Advantages and Limitations Current hydrogen storage technologies involve storing large quantities of H2 gas in compressed steel cylinders at elevat ed pressures (5000 – 10000psi), as well as, sufficiently cooling the gas below its critical temperature so that it becomes liquefied, i.e. cryogenic storage. Both methods have lim itations concerning on-board applications. Compressed tanks store hydroge n at elevated pressures and therefore pose neither obvious safety concerns, while liquefaction of hydrogen is not cost-effective nor energy efficient. Note that the latter is still used in space technology and re lated applications but it must be contained in well insulated and pressurized vessels to avoid loss through evaporation.136 In summary, these methods are pla gued by limited storage densities as a result of the added weight ( e.g. insulation) and this will in turn prevent them from reaching the technical targets. Ongoing studi es are however being conducted to develop hybrid compressed/liquefied storage systems ( i.e. lower pressures and warmer temperatures) because this approach would be deemed safer and reduce the amount of H2 lost from constant boil off. 1.8.2. Potential Hydrogen Storage Materials In an effort to overcome the limitations imposed by compressed and cryogenic hydrogen storage technologies researchers ar e exploring the potential of solid-state materials to remedy this problem by storing hyd rogen in or on the surface of wide range of materials via physisorption (r eversible), chemisorption (irrev ersible), and/or spill-over processes.128, 133, 137-139 The relative strength of the binding affinity of dihydrogen with the adsorbent will ultimately dictate which mechanism is observed and play a decisive role in

PAGE 83

46 determining the uptake capacity and revers ibility of the system. A tank containing a porous material (adsorbent) loaded with hydrog en would in theory store the same amount of gas in less volume and at lower pressures. Materials that are currently being expl ored and developed as viable hydrogen storage materials (HSMs) include metal hydrides, complex hydrides, carbon-based compounds, zeolites, and MOFs. Each of the proposed platforms offer a unique set of advantages but in many cases are plagued by severe limitations that impede their development for practical on-board applications. The revised DoE targets for 2010 stipulate that a viable HSM storing 5 kg of H2 must have a gravimetric and volumetric storage capacity of 4.5 wt% of H2 and 0.028 kg L-1, respectively and should reversibly adsorb/desorb H2 under moderate pressures in the temperature range of -30 to 50oC, and permit a driving range of greater than 300 mile s (480 km), as highlighted above in Table 1.3.135 It is important to bear in mind howev er that these targets include all system components such as the tank, regulators, va lves, piping, mounting br ackets, insulation, cooling capacity, etc. The actual storage capac ity of the adsorbent must therefore exceed these values to compensate for the loss due to infrastructure. Not to mention that the tank must be durable and the re-fueling process must be safe, fast, and as efficient as currently gasoline tanks. These targets become even more demanding for 2015. 1.8.2.1. Chemical versus Physical Adsorption Adsorption can be subdivided into two cate gories based upon the relative strength of interaction between the adsorbate and the host ( e.g. adsorbent); that is, chemical (irreversible) and physical (reversible) ad sorption. The former, commonly referred to as

PAGE 84

47 chemisorption is associated with strong adsorbate-adsorbent interactions ( e.g. > 50 kJ mol-1). The estimated binding energy is therefore similar in magnitude to the strength of a chemical bond that is ionic or covalent in nature.140 The adsorption process often takes place at elevated temperatures ( i.e. above the critical temperat ure of the adsorbate) and frequently leads to very high isosteric heats of adsorption ( Q st) values ( i.e. too high for vehicular applications). This is because the mechanism involves the making (adsorption) and subsequent breaking (desorption) of chemical bonds. Unfavorable temperature, pressure, and/or chemical reaction conditions may therefore be required to remove the adsorbate and could compromise the stru ctural integrity of the adsorbent. The other adsorption category encomp asses physisorbent materials. This particular class of compounds is well-known to exhibit facile reversibility because the adsorption process is governed by rather weak intermolecular interactions ( e.g. ~ 4 to 8 kJ mol-1). The so-called van der Waals for ces, i.e. dispersion and London forces141 are the most prevalent types of forces in physisorption but others ca n also be present and include ion-dipole, ion-induced dipole, dipol e-dipole, and quadr upole interactions.140, 142 This technique is well suited for surface area, pore volume, and pore size distribution determinations because the adsorbed phase is free to migrate on the surface and completely fill the pores. 1.8.2.2. Proposed Chemisorption-Based Adsorbents Metallic and chemical hydrides fall into the chemisorption cat egory and therefore are often associated with very high binding affi nities which lead to issues with refueling due to irreversibility 143-146 Fully reversible adsorption of hydrogen accompanied by high

PAGE 85

48 volumetric capacities has indeed been observed in some metallic hydrides but their high framework densities often lead to gravimetri c storage capacities that are far too low to meet technical targets. Complex hydrides have demonstrated good gravimetric energy densities but are plagued by slow kinetics and unfavorable desorption temperatures. Reports have shown that hybrid materials comprised of both metallic and complex hydrides offer superior performance. Full reversibility is achieved along with good kinetics and gravimetric uptake capacities; how ever, desorption at elevated temperatures is still a reoccurring problem. 1.8.2.3. Hydrogen Storage in Physisorption-Based Materials The fastest way to charge and discharge hydr ogen from an adsorbent ideally is to keep it in its molecular form, i.e. using physis orption-based materials. The most extensive amount of research in this area has focuse d on the storage of dihydrogen in carbon-based materials, zeolites, and permanently porous MOFs.72, 133, 147-151 The latter class of solidstate crystalline materials has received cons iderable interest since the first reported H2 uptake by Yaghi and co-workers in 2003 ( i.e. MOF-5).152 In order to achieve high uptake capacities that will satisfy the DoE technica l targets the optimal binding affinity is estimated to be in the range of 15 to 20 kJ mol-1. A compromise must therefore be made with respect to the binding affinity observed in physiand chemisorption-based materials. A promising hydrogen storage candidate must exhibit lower H2-MOF binding energies than those observed in dissociative chemisorp tion in order to sustain the reversibility and fast kinetics but exhibit a significant increase in binding en ergy must be sustained over

PAGE 86

49 what is classically observed for physisorption-ba sed materials in order to adequately store large amounts of hydrogen. 1.9. Key Factors which Govern H2 Binding in MOFs: Literature Summary and Future Outlook Permanently porous MOFs are well suite d to potentially circumvent the H2 storage challenge. This is due to their extens ive list of attractive features, which include modularity, high surface areas and H2 uptake capacities (at 77 K), low framework densities, and moderate chemical/thermal stabilities. Several promising avenues have been pursed in recent years to increase the H2-MOF binding interactions, while maintaining high surface areas and low fram ework densities. These include, most notably: (1) Embedding coordinatively unsatur ated metal sites into the MOF via the framework backbone ( e.g. using inorganic or organic MBBs);153 (2) Optimizing the pore dimensions ( e.g. < 1 nm);98 (3) Incorporation of highly polarizable moieties into the framework;154 and (4) Inclusion of an electrostatic field by ha ving a charged framework ( e.g. extraframework cations/anions).155-157 Each pathway has at tracted considerable attention because the Q st determined for these systems is estimated to be in the range of 7 to 13 kJ mol-1. This is an improvement over the 5 kJ mol-1 benchmark ( e.g. MOF-5). Few reports however embark on this chal lenge in a systemic fashion and this makes it exceedingly difficult to independen tly evaluate the contribution of each parameter towards the overall H2-MOF binding affinity. The design of potential made-toorder hydrogen storage candidate s is therefore critically de pendent on gaining a detailed chemical understanding of the preferential binding sites of molecular hydrogen in these frameworks in a systematic fashion (see Chapter 2).

PAGE 87

50 1.9.1. Importance of High Surface Area The rational construction of porous MOFs with high anticipated apparent surface areas is a crucial parameter to maintain thr oughout the design proce ss. A large accessible surface facilitates more H2-MOF interactions and in doing affords an increased number of binding affinities. This will ultimately contri bute to improving the overall performance of the MOF. The abundance of sorption data ava ilable for MOFs in th e literature reveals a direct correlation between surface area and por e volume; that is, MOFs with high surface areas often have large corre sponding pore volumes (Figur e 1.23.). For example, NOTT116 published be Schroder and co-workers with rht topology has an apparent BET surface of 4664 m2/g and a pore volume that is estimated to be 2.17 cm3 g-1.158 MOFs with large pore volumes however rarely reach saturation at lower pressures ( e.g. 1 atm) but often reveal outstanding gravimetric up take capacities at higher pressures ( i.e. compressibility factor). The highest surface area reported for a MOF ( i.e. as of May 2010) is UMCM-2 synthesized by Matzger and co-workers.159 Apparent BET and Langmuir surface areas were determined to be 5200 and 6060 m2/g, respectively. Note that since 2004 this field has witnessed unpr ecedented breakthroughs in terms of reported surface areas ( e.g. MOF-177, MIL-101, NOTT-116) which implies the limits have not yet been realized for these materials.86, 158, 160

PAGE 88

51 Figure 1.23. A plot of pore volume versus surface area for select MOFs to illustrate the correlation between these factors. Select examples of MOFs having appa rent surface areas greater than 3000 m2/g is provided in Table 1.4. Note that the calcula ted framework density for MOFs is often inversely proportional with respect to the su rface area and pore volume whereby higher surface area materials predominantly have lo wer densities. The data presented below clearly reveals that MOFs with high surf ace areas are dominated by neutral frameworks having low densities. Yet, those with smalle r pores and potential open metal sites reveal enhanced H2-MOF interactions as evidenced by the la rger isosteric heat of adsorption (at low loading), as will be discussed below.

PAGE 89

52 Table 1.4. Select examples of MOFs with high surface areas. refMOF MBBs Ca P. S. ( ) S. A. (m2/g) H2 Uptakeb Q st (kJ/mol) Inorganic Organic BET Lang. 159UMCM-2c Zn4O(CO2)6 BTB; T2DC N ~14-30 5200 6060 1.3 6.4 – 4.2 160MOF-177 Zn4O(CO2)6 BTB N ~10-15 4746 5640 1.25 4.4 161UMCM-1 Zn4O(CO2)6 BTB; p -BDC N ~14-32 4730 4160 N.R.d N.R. 158NOTT-116e Cu(CO2)4L2 ATPB N ~13-24 4664 N.R. 9.2 f 6.7 162MIL-101g Cr3OF(CO2)6 p -BDC N ~29-34 4230 5500 2.5 10-9.3 163, 164IRMOF-20 Zn4O(CO2)6 T2DC N ~15-20 4024 4593 1.35 N.R. 165PCN-66 h Cu(CO2)4L2 NTEI N ~13-20 4000 4600 N.R. N.R. 166, 167IRMOF-1 Zn4O(CO2)6 p -BDC N ~10-15 3800 4400 1.32 4.8 168NOTT-112 Cu(CO2)4L2 TPB N ~13-20 3800 N.R. 2.30 5.6 82PMOF-2 Cu(CO2)4L2 TEPB N ~13-20 3730 4180 2.21 9.2 97rht-1 Cu(CO2)4L2 Cu3O(N4CR)3 TZI C ~13-20 2847 3223 2.4 9.5 169PCN-6 Cu(CO2)4L2 TATB N ~10-15 N.R. 3800 1.90 N.R. 163IRMOF-6 Zn4O(CO2)6 o -DCB N ~10-15 2476 3263 1.48 N.R. aOverall charge on the framework, where N refers to neutral, A is anionic, and P is cationic; bH2 uptake measured at 77 K and 1 atm; cUniversity of Michigan Crystalline Material; dNot reported; eNottingham; f35 bar and 77 K; gMaterials Institut Lavisoier; fIsoreticular Metal-Organic Framework; gPorous Coordination Network. Abbreviations: BTB = 1,3,5-benzenetribenzoate; T2DC = thieno[3,2-b]thiophene-2,5dicarboxylate ; p -BDC = 1,4-benzenedicarboxylate; ATPB = 1,3,5-tris(3,5dicarboxylbiphenylethynyl)benzene; NTEI = 5,5',5''-(4,4',4''-nitrilotris(benzene-4,1-diyl)tris(ethylene-2,1diyl)triisophthalte; TPB = 1,3, 5-tris(p-phenyldicarboxyl)ben zene; TEPB = 1,3,5-tris(3,5dicarboxylphenylethynyl)benzene; TZI = 5-tetrazolylate; TATB = 4,4'.4''s -triazine-2,4,6-triyltribenzoate; o -DCB = 1,2-dihydrocyclobutab enzene-3,6-dicarboxylate. 1.9.2. Design Strategies to Incorporate Po tential Open Metal sites into MOFs Experimental and computational studies have shown that a high concentration of potentially accessible open-metal binding si tes, i.e. multinuclear clusters, enhances Q st in MOFs at low loading.153 This is therefore a sought af ter feature because it facilitates higher gravimetric H2 uptake capacities at lower pr essures (see Table 1.5.). The impregnation of coordinatively unsaturated meta l centers into MOFs, whether in the form of intraor extra-framework metal cations ofte n leads to stronger bind ing sites due in part to favorable metal-H2 orbital interactions and enhanced electrostatic interactions. The organic components on the other hand are believed to have less of an effect, with binding energies in the range of ~4 – 5 kJ mol-1.128 Note that the contemporary challenge is to

PAGE 90

53 synthesize a MOF that also maintains these favorable binding interactions at higher loadings (see Chapter 2). Table 1.5. Select examples of MOFs to highlight the relationship between Q st and accessible to potential open-metal centers. refMaterial Inorganic MBB Surface Area (m2/g) H2 Uptakeb Q stb (kJ/mol) BET Langmuir 170Ni2(DHTP) rod-like Ni clusters N.R. N.R. 2.0 13.5 171[Cd5(TZ)9]NO3 Cd5(TZ)6 310 338 0.75 13.3 172Co3[(Mn4Cl)3(BTT)81.7CoCl2 Co3[Mn4Cl(N4CR)8] 2096 2268 2.12 10.5 173NaNi3(OH)(SIP)2 Na2Ni6O34 700 N.R. 0.94 10.4 172Fe3[(Mn4Cl)3(BTT)8]2 FeCl2 Fe3[Mn4Cl(N4CR)8] 2033 2201 2.21c 10.2 172Mn3[(Mn4Cl)3(BTT)8(CH3OH)10]2 Mn3[Mn4Cl(N4CR)8] 2057 2230 2.20c 10.1 MIL-101 [Cr3OF(CO2)6] 4230 5500 2.5 10.0 97[Cu6O(TZI)3]NO3 [Cu2(CO2)4]; Cu3O(N4CR)3 2847 3223 2.4 9.5 172Ni2.75Mn0.25[(Mn4Cl)3(BTT)8]2 Ni2.75Mn0.25 [Mn4Cl(N4CR)8] 2110 2282 2.29a 9.1 71Zn3(BDT)3 Zn3(BDT)3 640 N.R. 1.46 8.7 71Mn3(BDT)3 Mn3(BDT)3 290 N.R. 0.97 8.4 163Zn2(DHTP) rod-like Zn clusters 783 1132 1.77 8.3 163Cu3(BTC)2 [Cu2(CO2)4] 1507 2175 2.5 6.8 98[In3O(ABTC)1.5]NO3 In-TMBB N.R. 1417 2.6a 6.5 aPore size; bHydrogen uptake capacity at 77 K and 1 atm; cIsosteric heat of adsorption at low loading; cat 1.2 bar. Abbreviations: DHTP = 2,5-dihydroxyterephthlate; TZ = tetrazolate; BTT = 1,3,5benzenetristetrazolate; SIP = 5-sulfoisophthalate; TZI = 5-tetrazolylate; BDT = 1, 4,-benzeneditetrazol-5-yl; BTC = 1,3,5-benzenetricarboxylate; ABTC = 3,3',5,5'-azobenzenetetracarboxylate. An urgent need exists for the developmen t of various platforms so that the overall effect of the metal ion can be truly assesse d in a systematic fashion. In many cases a MOF that exhibits a high Q st may also be contain other key parameters that promote stronger interactions ( i.e. narrow pores, electrostatic field, etc ). This is exemplified in the case of soc -MOF which exhibits an exceptional H2 uptake capacity of 2.61 wt % at 1.2 atm and 77 K. The high uptake is in agreement with the constancy observed in Q st and most importantly it is maintained at 6.5 kJ mol-1 even at higher lo adings. Eddaoudi and

PAGE 91

54 co-workers attribute to th e potential open In-TMBB metal centers, narrow pores, and higher localized charge density in the channels (see Chapter 2).98, 174 To date, three different strategies are actively being pursued by researchers to incorporate accessible open-metal binding sites into MOFs. These include, (1) the synthesis of solvent-containing MOFs wher eby the metal-bound solvent molecules on the inorganic MBB ( e.g. H2O, DMF, MeOH, etc ) can be removed post-synthetically via thermal and/or solvent exchange techni ques to produce unsaturated metal centers,87, 153 (2) introducing metal cations into the framework backbone by utilizing metallated organic-based bridging ligands ( e.g. metalloporphyrins, crown ethers, salen-type ligands, etc ),175-177 and (3) the synthesis of charged frameworks in which case the extraframework cations can be substituted for metal ions or metallated complexes ( i.e. anionic MOFs).72, 155 The first approach is certainly th e most exploited strategy used by experimentalists, especially since the latter have only been recently developed. In addition to selecting a suitable strategy, the type of metal cation is very important because the binding affinity of molecular hyd rogen can vary significantly from one metal ion to the next. Specific characteristics that must be taken into consideration include the charge, atomic radius, and the classification of the metal ( e.g. s -, d -, p -, or f -block). Studies suggest that alkali-based meta ls are unlikely to satisfy the optimal Q st demand at room temperature, yet monovalent transition me tals may lead to bindi ng energies that are in fact too strong ( i.e irreversible) for on-board applica tions. The most explored group of metal ions are the dicationic first row transition metal ions and select s and p -block metals ( e.g. Na+, Li+, Mg2+, Al3+). Note that the highest H2 uptake recorded for a MOF is

PAGE 92

55 comprised of dicationic copper centers ( i.e. paddlewheel). Accordingly, PCN-12 has an apparent BET and Langmuir surface area that is estimated to be 1943 and 2425 m2/g, respectively and a corres ponding pore volume of 0.94 cm3 g-1.178 The neutral MOF adsorbs 3.05 wt % of H2 at 77 K and 1 atm. The Q st was not reported for the this material but the authors attri bute the high uptake to presence of open copper metal sites and the efficient packing of the spheroid-like building blocks. 1.9.2.1. Extra-framework Approaches: Ion-Exchange and Metal Doping The relative binding strength of me tal-hydrogen interactions has been systematically evaluated in ion-exchange d zeolites for many years. For example, temperature-dependent infrared spectr oscopy measurements carried out on Li+, Na+, and K+ ion exchanged zeolites ( e.g. ZSM-5, Ferrierite) confirm that the interaction energy is strongest for Na+ than the others.179-181 High framework densities associated with zeolite frameworks coupled with their lack of modularity render them unsuitable for hydrogen storage applications. Eddaoudi and co-workers however conducted similar studies using anionic ZMOFs as platforms. The charge ba lancing extra-framewor k organic cations ( e.g. DMA+) present in the enlarged cavitie s were readily exchanged for Li+ and Mg2+ ions, which resulted in improved Q st over the parent compound ( e.g. Li > Mg) and more importantly as much as an 50% increase compared to neutral MOFs. Unlike their inorganic counterparts, ZMOFs are not thermally stable at elevated temperatures ( i.e. > 300oC) and therefore the metal ions are still hydrated. Accordingly, the authors associated the enhanced isosteric heat of adsorption to th e presence of an elec trostatic field in the cavities resultant from the metal ions.155

PAGE 93

56 A comprehensive ion-exchange study wa s conducted by Long and co-workers to investigate the effect of metal cation on H2 binding affinities in a sodalite-type platform formulated as, [(Mn4Cl)3(BTT)8(CH3OH)10]2. Exchange with metal chloride salts resulted in the formation of an isostruc tural series of mix-metal MOFs, M3[(Mn4Cl)3(BTT)8(CH3OH)10]2xMCl2 where M = Co2+, Ni2+, Cu2+, Zn2+, Fe2+. The isosteric heats of adsorption for these MOFs we re estimated to be in the range of 8.5 kJ mol-1 to 10.5 kJ mol-1 with the Cu analog being the lowest and Co the highest.172 Hupp and co-workers on the hand introduced a nove l strategy by doping a neutral pillared MOF with Li+ cations. The results were encouraging whereby the H2 uptake was increased from 0.93 wt % to 1.63 w% (77 K; 1 atm). This was found to be in agreement with the isosteric heat of adsorption, which revealed an increased over th e full pressure range.182 1.9.2.2. Experimental Technique s: Detection of Metal-H2 Binding in MOFs Many strategies have been develope d in recent years to incorporate potential open metal binding sites into MOFs. Only a handful of these studies however use experimental techniques to unequivocally prove that the metal centers are ind eed open binding sites and thus accessible to molecular hydrogen. Th e possibility that hydrogen can in fact bind to a metal center in molecular form wa s discovered in 1984 by Kubas in the complex [W(CO)3(P i Pr3)(H2)].183 The formation of such a side-on 2, -H2 complex was confirmed with various experimental and theo retical techniques, including for example, single-crystal neutron diffracti on studies. Following this extremely significant discovery many analogous transition metal -H2 complexes have been reported, where it must be

PAGE 94

57 noted that oxidative addition of the H2 to the metal, and hence binding as hydrides, is still far more common.184, 185 1.9.2.2.1. SCD, FT-IR, and Low-temperature Po wder Neutron Diffraction Studies Some experimental techniques that have been employed to provide evidence of an open metal center include single-crystal Xray diffraction (SCD), infrared (IR) spectroscopy, low-temperature powder neutron diffraction studies, a nd inelastic neutron scattering (INS) experiments. Yaghi and co-w orker were the first to investigate the presence of accessible open metal sites by us ing SCD studies. The authors collected SCD data on the hydrated and dehydrated forms of the porous MOF-11 with pts topology.186 The framework is built-up from copper te tracarboxylate MBBs and therefore has two potential open metal sites per MBB. Indeed, the crystal structure of the dehydrated form revealed no evidence of coordinated axial aqua ligands and the structure was more relaxed as compared to the hydrated form. Bordiga and co-workers were however the first to confirm binding of molecular hydroge n to open metal sites in a MOF, i.e. HKUST-1 formulated as [Cu3(BTC)2], which is also assembled from copper paddlewheel MBBs.187 This was carried using IR spectrosc opy whereby a stretching band at 4100 cm-1in the IR spectrum confirmed the presence of a metal-H2 interaction ( i.e. physisorbed). Note however that this type of interaction is nothing like that observed in the Kubas complexes and is nothing mo re than a physisorbed interaction. More extensive studies in recent years have focused on the use of lowtemperature neutron powder diffracti on experiments by determining the Mx+-D2 distances. Results on HKUST-1 and revealed 6 different binding sites, with the accessible

PAGE 95

58 square pyramidal copper site be ing the most favorable, i.e. Cu2+-D2 distance of 2.39 Long and co-workers carried out a systematic study of metal on H2-MOF interactions in two isostructural MOFs with sodalitetype topology and formulated as, HM[(M4Cl)3(BTT)8]3.5HCl, where M = Mn or Cu. M2+-D2 distance of 2.27 and 2.47 for the Mn2+ and Cu2+ analogs, respectively, were found in neutron powder diffraction data.172 This evidence suggests that hydrogen will bind more st rongly in the Mn analog and is therefore anticipated to have a higher Q st for H2 at low loadingsw. Indeed, sorption experiments confirmed that the Q st was higher for Mn analog (10.1 kJ mol-1) compared to the copper-based framework (9.5 kJ mol-1). It is noteworthy to mention that Q st for the copper analog was more constant over the enti re coverage because of the presence of a higher concentration of open metal sites. This is because the manganese analog cannot be fully dehydrated owing to the presen ce of residual methanol ligands. A neutron scattering study of the zinc analog of MOF-74 ( i.e. CPO-27) found a rather long Zn2+-D2 distances of 2.6 which is in accord with the lower Q st of 8.8 kJ mol-1 as compared to complexes with shorter Mx+-D2 bond lengths. D2-D2 distance between neighboring adsorbed molecules were determined to be 2.85 which is rather short in view of the fact that the distance between H2 molecules in the solid state is 3.6 This finding therefore does s uggest that open metal centers le ad to more dense packing of hydrogen in these systems.188 Cheetham and co-workers explored the Ni2+-D2 interactions via INS studies in two microporous phosphateand sulfoisophthalate-based frameworks. Both TPD studes and INS spect ra an unusually strong interaction of hydrogen with Ni in accord with the high value of Q st of 10.4 kJ mol-1.173

PAGE 96

59 1.10. Inelastic Neutron Scattering (INS) Neutrons can be used in a wide variety of scattering t echniques to investigate the bulk properties and surrounding chemical environm ent of materials in condensed phases. Neutrons are neutral subatomic particles and in teract directly with atomic nuclei rather than the electrons. Scattering of neutrons is a function of th is neutron-nuclear interaction, and it has an irregular variation through the periodic tabl e, whereas the intensity of the scattering in photon-based techniques de pends upon the number of electrons ( i.e. in X-ray diffraction) or their polari zability or the resulting dipole moment (in optical spectroscopy). For example, it can be difficult to obtain information for light atoms such as hydrogen, which has the largest neutron scat tering cross section of any element (Figure 1.24.). Neutrons are unique scattering probes in the sense that both of their energies and wavelengths match, respectiv ely, excitations and intera tomic dimensions found in condensed matter. They can, therefore be used in diffraction experiments, i.e. elastic scattering (as X-rays) and in a variety of spectrocopies, i.e. inelastic scattering, to study dynamics of molecules over a very large ra nge of timescales. In the latter case the neutron can either lose energy by exciting a vibrational, or ro tational mode in the solid in the scattering process, or gain energy from de-excitation of such a mode. INS maybe regarded as the most sensitive spectroscopic tool for determining the affinity of adsorbed H2 at various binding sites in porous materials. This is based on the fact that the lowest tran sition between the rotational energy levels of the bound H2 molecule is extremely sensitive to the surround ings which give rise to the barrier to rotation. Neutrons can observe the rotational tr ansitions associated with a change in the total nuclear spin of the hydrogen molecule, wh ich is not normally possible using optical

PAGE 97

60 spectroscopy. A change in the nuclear spin ( I) coincides with a change in the molecular rotational quantum number ( J ) and is forbidden in optical spectroscopy because photons are unable to couple to nuclear spins.189 Figure 1.24. Schematic illustration of select elements and the deuterium isotope to emphasize the cross sections due to neutron scattering (coherent = blue; incoherent = purple) and absorption (green).190 The rotational transitions observed in an INS spectrum can tentatively be assigned using a phenomenological model for the rotational energy levels. We use the model previously described by Eckert and co-workers for rotati on in a simple double-minimum potential with two angular degrees of freedom which has just one parameter, namely the barrier height, V2, to be determined from the observed transitions.98, 191 In the absence of a barrier to rotation the energy levels of the hydrogen molecule are quantized and given by B J ( J +1) whereby B is the rotation al constant (7.35 meV or 59.6 cm-1) and J the rotational quantum number. The lowest transition for the free rotor is the ortho to para H2 transition which occurs at 14.7 meV (119 cm-1 = 2B). Introduction of a barrier to rotation from guest-host interactions results in a hindered rotation, which lifts some or all

PAGE 98

61 of the degeneracy of the higher levels (Figure 1.25.).192 We have chosen to simply label the energy levels sequentially (0,1,2, etc .) as the exact attribution (mJ sublevels) is not trivial, and depends on the form of the actual rotational potenti al energy surface. Figure 1.25. Schematic of the rotational energy leve l diagram for a dihydrogen molecule with two rotational degrees of freedom, under the influence of a hindering potential (Courtesy of Dr. Eckert at UCSB). The 0-1 transition decreases very strongly with increasing barrier height as it may be regarded as a rotational tunneling tran sition. The tunneling probability is always smaller for larger barriers, whereas the highe r transitions (torsional oscillator) increase with barrier height. A smaller value for th e 0-1 transition therefore suggests a stronger interaction of H2 with the host. We can associate di fferent sets of observed transitions with particular binding sites when we carry out a series of experi ments as a function of H2 loading, related to know st ructural features in the ma terial, and thereby assess the relative strengths of different binding sites. Th is type of INS experi ment has the potential to provide critical insights in terms of rationalizing which structural and chemical

PAGE 99

62 features contribute to favorab le hydrogen interacti ons with the host material. Efforts are underway to obtain rotational potential en ergy surfaces directly from ab-initio calculations for a given binding site, and then to calculate the expected rotational and translational transition directly using quantum dynamics. This will make it possible to use the INS data to give direct and quantitativ e information of the interactions of H2 at particular sites, and thereby direct effo rts in synthesis in very specific ways. 1.11. Other Applications: MOFs as Platform s for the Capture, Sequestration and Separation of Gases 1.11.1. Methane Storage The movement towards a potential hydr ogen-based economy is certainly the most practical solution from both an enviro nmental and economic standpoint. Methane however is also regarded as attractive substitu te for gasoline and diesel fuels because of its high volatility ( e.g. dissociates quickly in the event of a spill) and the byproducts from the combustion process contain less carbon dioxide emissions than conventional hydrocarbon fuel sources. Methane is very abundant, non-toxic, and the principle component in natural gas (87 % by volume). The reoccurring problem however is that it also exists as a gas at ambient temperatures and standard pressures. This feature therefore hinders its utility in on-boar d storage applications. Accord ingly, the development of a safe, efficient, and cost-effective storage system is of the utmost importance. Note that a vehicle powered by adsorbed natu ral gas must have an energy density that is at least equivalent to the commercially available compressed natural gas powered vehicles.193, 194 The US DoE methane storage target is 180v( STP)/v, which represents the equivalent volume (cm3) of methane stored per volume of the adsorbent at near ambient temperature

PAGE 100

63 and pressures under 35 bar. A variety of porous adsorbents have been explored for their potential to serve as suitable methane storage media ( e.g. zeolites, activated carbons, single-walled carbon nanotubes, and MOFs.45, 84, 195-198 The results are promising, yet an adsorbent that adequately stores significant amounts of both methane and hydrogen has not been isolated and therefore remains a contemporary challenge. The highest absolute methane adsorption capacity is exhibited by a MOF (PCN-14) which indeed surpasses the DoE target by 28 % by storing 230v(STP)/v at 290K and 35 bar.199 It is assembled from two types of nanoscopic cages (octahedral and cuboctahe dral) comprised of copper paddlewheel inorganic building blocks whic h are linked together through an anthracene backbone. The assembly of the two 4-connect ed nodes results in the formation of a neutral 3-periodic MOF having nbo topology. The presence of coordinately unsaturated metal sites and enlarged pores having sm all window apertures contributes to an unprecedented isosteric heat of adsorption, i.e. Q st 30 to 35 kJ mol-1. It is noteworthy to mention that certain carbon-based adsorbents have also exceeded the DoE target; however, they exhibit very weak adsorbate-ad sorbent interactions and it is difficult to enhance these interactions. 1.11.2. Storage and/or Sequestration of Carbon Dioxide The combustion of fossils fuels generate s the most abundant greenhouse gas; that is, carbon dioxide (CO2). In 2005 alone the United Stat es independently released 11 billion metric tonnes of CO2 into the atmosphere, 1/3 of which was from power plants and other large point sources.200 Amid escalating environmental concerns ( e.g. global warming) the DoE introduced the Car bon Sequestration Program (CSP) in 1997 to

PAGE 101

64 promote the research and development of high scale cost-effective CO2 capture technologies. The ultimate goal of this program is to significantly reduce large scale CO2 gas emissions (by 2012) through the development of a fossil fuel conversion system that efficiently captures at least 90 % of the gas (in the pure form).201 Once isolated and captured the gas would be compressed and th en transported to an underground storage site for disposal or recycled for use in chem ical processes, enhan cement of oil recovery, and/or enhancement of coal bed methane producti on. Note that it is v ital for all of these steps to accomplished whilst simultaneously bein g efficient and cost effective; that is, with a maximum of 10-20% incr ease in production costs. The mo st expensive step of the entire process is estimated to be the capture process, which repres ents three-fourths the cost of the entire system.202 Methods have been proposed and explored to capture CO2 via pre-combustion, post-combustion, or via oxy-combustion; each of which are characterized by specific adva ntages and limitations. The most prevalent industrial adsorbents currently used to sequester CO2 via postcombustion from flue gas streams are so-called chemical ( e.g. monoethanolamine, MEA) and physical solvents ( e.g. Selexol) which are utilized at low and high pressures, respectively.203, 204 These solvent methods are expensive and pose an environmental challenge for large scale CO2 sequestration due to potential solvent degradation and loss during the operation process. Some large point sources also dilute CO2 with N2 using air fired combustors but this method is not ideal and be tter separation methods must be devised. Scientists are theref ore exploring alternative CO2 sequestration systems that are highly efficient, practical, and cost-effective. Some of the proposed concepts include lowtemperature distillation, gas separation membra nes functionalized with mobile or fixed

PAGE 102

65 amine carriers, mineralization, biological -based capture systems, absorption and adsorption (physical and chemical).205, 206 Porous MOFs with high surface areas re present an attractive class of solid adsorbents that are well suited to potentially remedy this problem. An optimal adsorbent must therefore be designed to possess the following traits: (1) high gas selectivity for CO2 in a mixed gas stream, (2) high adsorption capacity, (3) high stability ( e.g. in the presence of contaminants), (4) favorable equilibrium and kinetics, (5) ease of regeneration, and (6) of course be cost effective. The underlying mechanisms that govern adsorption-based gas separation processes are not fully underst ood. Numerous studies however have proven that the molecular sieving eff ect, thermodynamic equilibrium effect, kinetic effect, and/or quantum sieving effect can independently or collectively play a dominant role in governing the overall se lectivity process.207 An increasing number of rigid and flexible MOFs have been evaluated in recent years for their CO2 storage capabilities.59, 208-213 The results of these studies are encouraging, as illustrated in Table 1.6., but still of hundreds if not thousands of MOFs ha ve yet to be tested.

PAGE 103

66 Table 1.6. CO2 adsorption capacities ( e.g. low and high) for selected MOFs. refMaterial S. A. (m2/g)a Framework Features P. S. ( )b Max. CO2 Uptake Exp. Conditions (T K, P bar) Q stc (kJ/mol) BET Lang. 214MOF-177 4750 5640 high porosity; large cavities 11 x 17 33.5 mmol g-1 40.0 mmol g-1 298, 42 298, 50 N.R. 215MIL-100 1900 N.R. OMSd; large cages; small window apertures C: 25 x 29; W: 5.5 x 8.6 18.0 mmol g-1 304, 50 62 215MIL-101c 4100 5900 mesoporous cages with small window apertures C: 29 x 34; W: 12 x 16 40.0 mmol g-1 304, 50 44 216MIL-102 N. R. 42.1 1-D circular channels decorated with Fanions ~4.4 3.4 mmol g-1 304, 50 N.R. HKUST-1 1154 1958 OMSd; high porosity 5 and 15 10.7 mmol g-1 298, 42 35 IRMOF-1 3800 4400 high porosity 12 x 15 21.7 mmol g-1 298, 35 N.R. IRMOF-3 1568 N. R. amino functionalized pores 10 x 15 18.7 mmol g-1 298, 35 N.R. IRMOF-6 2516 N. R. alkyl functionalized pores 10 x 15 21.7 mmol g-1 298, 35 N.R. IRMOF-11 2096 N. R. catenated 7 x 12 14.7 mmol g-1 298, 35 N.R. 217Bio-MOF-11 1040 N. R. narrow pores with lewis basic sites ~5.2 6.0 mmol g-1 298, 35 45 174In soc-MOF N.R. 1417 narrow pores; high localized charge density < 10 9.0 mmol g-1 298, 25 28.5 213ZIF-78 620 N. R. polar functional group; small windows C: ~ 7.1; W:~3.8 6.0 mmol g-1 298, 1 N.R. 213ZIF-69 950 N. R. Large pores and large windows C: ~12.4; W:~5.9 37.6 cm3 g-1 Or 82.6 L/L 298, 1 N.R. aApparent surface area; bAverage pore size; cIsosteric heat of adsorp tion at low loading; and dOpen metal sites. Similar to hydrogen storage there is a con tinuous quest to find the optimal balance between a maximum adsorption capacity and fa vorable isosteric heat of adsorption. The highest amount of CO2 adsorbed in a MOF to date at room temperature is 33.5 mmol g-1 and 40.0 mmol g-1 at 42 bar and 50 bar, respectively for MOF-177.214 To put these numbers into perspective the highest gravimetric CO2 adsorption capacity for a zeolite

PAGE 104

67 material is 7.4 mmol g-1 at 32 bar and room temperature (zeolite 13X),218 while for a carbon-based material (MAXSORB) it is 25.0 mmol g-1 at 35 bar and room temperature.219 Several MOFs exhibit high ad sorption capacity for pure CO2 as shown above. The adsorption behavior however can be significantly altered when the material is exposed to a mixture of gases and thus th e adequate purification and separation of specific gases is of paramount importance for their potential u tility in industrial applications. Dynamic adsorption measurements in the form of kinetic breakthrough curves are commonly employed to quantify an d assess the adsorption behavior of solid adsorbents with respect to multi-component gas mixtures. The methodology behind this type of experiment separates the gases based on their affinity towards the adsorbent; that is, the gas mixture is introduced to a column that is packed with the adsorbent and the mixture is circulated until equilibrium is es tablished. Those gases with a strong affinity for the adsorbent will be retained in the column, while weakly bound gases will be permitted to pass through. The adsorbed phase can be readily released in a controlled fashion via mild heating. A numerical value fo r selectivity (separation factor) can be then be extrapolated from these curves and correla ted to the pore structure and composition of the adsorbent. The current state of the art industrial separation technologies utilize generalpurpose filters composed of activated carbon be ds that are impregnated with metal salts ( e.g. Cu, Zn, Ag, Mo). This class of adsorbents is efficient for some systems but it is not transferrable to all systems and their lack of modularity is impedes future optimization. The unique attributes afforded by MOFs ci rcumvent these limitations and systematic breakthrough studies carried out by Yaghi and co-workers, amongst others, demonstrates

PAGE 105

68 the importance of pore structur e and functionality for adequa te gas separation using this class of adsorbents.207 For example, it was shown that IRMOF-3 exhibits an 18 fold increase in ammonia dynamic capacity over MO F-5 upon exposure to a mixture of eight challenge cases (ammonia, ethylene oxide, chlorine, carbon monoxide, sulfur dioxide, dichloromethane, and tetrahydrothiophene).220 The vast number of MOFs synthesized versus those that have been tested for their selective gas adsorption and separation capabilities is not comparable. We therefore an ticipate that future studies in this area will be fruitful and lead to unprecedented results. 1.12. Characterization of Porous Solid s via Physical Adsorption Methods Porous compounds are delimited by an inte rconnected network of pores or voids. The vast majority of solid-state materials can therefore be described as having some degree of porosity because many contain ope n channels and/or cavities. Permanently porous solids however encompass a unique subset of porous materials whereby the structural integrity of the framework is maintained even after the complete removal of all guest molecules. Due to their inherent porosit y, they are uniquely suited for a plethora of potential applications ranging from gas storage and separa tion, catalysis, drug delivery, etc In order to optimize the performance of thes e materials in practical applications it is imperative to perform a comprehensive te xtual characterizatio n, in addition to, a systematic study to evaluate their adsorption properties ( e.g. surface area, pore volume, pore size). Numerous experiment al techniques are available to assess these features ( e.g. small angle x-ray and neutron scatteri ng, NMR-methods, mercury porosimetry, gas adsorption, etc ) but the applicability of each met hod is highly dependent on a length

PAGE 106

69 factor. Each method is therefore not univ ersal for the entire pore size regime. Gas adsorption is widely recognized as a reliab le, convenient, and co st-effective technique that is commonly employed to characterize porous materials because it permits a full evaluation of pore sizes ranging from micr oto macropores (from 0.35 nm up to 100 nm).140 Note that the amount of gas adsorbed is critically de pendent upon several experimental and structural parameters which include the absolute temperature, pressure, and the interaction potential ( E ) between the vapor (adsorbate) and the surface (adsorbent). 1.12.1. Exploiting the Porosity of MOFs : Protocols for Sample Activation An ongoing challenge for experimentalists is to identify a suitable protocol to fully activate porous MOFs, so that the experimental values agree well with computational and/or SCD studies. This is a pr erequisite step in or der to take advantage of the porosity whilst avoiding collapse of the framework. Se veral activation methods are currently available but each system often be haves differently and therefore what works for one material may not work for the next. MOFs that display permanent porosity are predominately activated via solvent exchange; that is, the higher boiling point gue st molecules residing in the channels/cages are exchanged for volatile solvents ( e.g. DCM, EtOH, (CH3)2CO, CH3CN, etc ). In a typical experiment, 30 – 100 mg of the as-syn thesized material is washed and then soaked in the chosen solvent for an unspecified amount of time ( e.g. 1 – 7 days). This step is predominately carried out at room te mperature but mild heating has been used to speed up the equilibrium process. The low boi ling point guest molecules can then be

PAGE 107

70 removed from the MOF using relatively mild evacuation procedures ( i.e. heating under vacuum). The success of this approach is evidenced by the larger number of highly porous MOFs reported that were activated us ing solvent exchange but still occasionally fails for many compounds.153 Promising alternatives have been propos ed to circumvent this challenge in solvent-containing MOFs and demonstrate gr eat promise. In one method, Hupp and coworkers introduced a hybrid technique c onsisting of liquid activation followed by supercritical drying (ScD) using CO2 activation.221 The authors observe up to 1200% increase over traditional liquid and thermally assisted evacuation protocols. Note that ScD is often used in the pr eparation of silica aerogels, organosilicates, and polymer synthesis because it is an effective way to remove liquids in a controlled fashion without comprising the structural integrity of the material.222 This method is therefore uniquely suited to enhance access to the internal surf ace area of solvent-containing MOFs by because the fluid can be removed by reducing th e surface tension in the material. In this context, the preferred gas of choice is CO2 because it is cheap and more importantly it becomes supercritical under mild conditions (Tc = 31oC; P = 73 atm). A similar concept, yet less explored, was introduced by Lin a nd co-workers (2009) whereby the higher boiling point solvents were exchanged with solvents that can be removed via freezedrying ( i.e. benzene).223 In this sense the liquid phase is precluded owing to the direct gas to solid transition. In summa ry, the enormous internal surface areas offer by this emerging classes of crystalline materials is critically dependent on development and identification of proper activation procedures that are unique for a particular material.

PAGE 108

71 1.12.2. Classification of Adsorption Isotherms An adsorption isotherm provides a re lationship between the amount of gas adsorbed per gram of solid as a function of pressure at a c onstant temperature ( i.e. boiling point of the adsorbate). Volumetric and gr avimetric sorption measurements are the two most popular techniques currently used to de termine the amount of gas adsorbed in a porous material. The shape of the isotherm is dependent upon the relati ve strength of the adsorbate-adsorbent and adsorbate-adsorbat e interactions and therefore it reveals important structural information. This is b ecause these interactions are governed by the size, shape, and chemical composition of th e framework. The sorption behavior is also affected by the experimental conditions, i.e. the temperature relative to the boiling point of the adsorbate. This is due to the fact that differences in temperature ( i.e. suband supercritical) will invoke changes in the th ermodynamic states of the pore and bulk fluid phases (see Chapter 2). In 1985, the International Union of Pure and Applied Chemistry (IUPAC) recognized the need for an efficient and uni versal classification system regarding the terminology, classification, and inte rpretation of sorption isotherms.224 As a subjective guide to researchers IUPAC therefore propos ed to classify pore sizes based on the diameter of the pore. The limits were divided into three categories: (1) Micropores (2 nm or less); (2) Mesopores (2 nm to 50 nm); a nd (3) Macropores (> 50 nm). With regards to sorption behavior, porous materials fall into one of the six types of adsorption isotherms (Figure 1.26.). A fully reversible Type I isotherm is characteristic of microporous materials whereby the adsorption process is re stricted to only a few molecular layers and is exemplified in the case of many MOFs, zeolites, porous ox ides, and activated carbons.

PAGE 109

72 These isotherms are predominantly collected using nitrogen and ar gon as adsorbates at their corresponding boiling temperature of 77 K and 87 K, respectively. With respect to shape, the curve is concave with respect to the P/P0 axis and reaches a plateau as P/P0 approaches 1. This therefore allows for th e determination of the surface area and pore volume of the material. Note that the st eepness and high uptakes (micropore filling) observed in the isotherm at low pressures is attributed to strong adsorbate-adsorbent interactions. Figure 1.26. IUPAC classification of sorption isotherms.224 A reversible Type II isotherm denotes a non-porous or macroporous material. The overall shape is concave with respect to the P/P0 axis at low pressure s but it exhibits an inflection point or knee ( e.g. point B) at higher pressures. This point is believed to indicate the transition from monoto multila yer coverage. The amount adsorbed does not

PAGE 110

73 reach a plateau as P/P0 approach 1 and therefore materi als that fall into this category possess such large pores that monolayer-m ultilayer coverage is not restricted. The reversible Type III isotherm adopts a shape that is completely convex to the P/P0 axis and is rarely observed in the literature. The adsorption behavior is primarily influenced by adsorbate-adsorbate intera ctions and less on the attractive adsorbateadsorbent interactions ( i.e. as observed in Type I). This is evidenced by the horizontal slope with respect to the P/P0 axis at low pressure, which si gnifies rather weak fluid-wall interactions. At higher pressures the amount adso rbed increases at an exponential rate and thus reflects the importance of strong fluid-fluid interactions. Mesoporous materials commonly exhibit the Type IV isotherm, which is accompanied by a distinct hysteresis loop thereb y indicating that the adsorption process is not fully reversible. This l oop results from pore (capillary ) condensation and is observed when the adsorbate reaches a liquid-like state at pressures below the saturation pressure (P0) of the bulk fluid. The general shape re sembles a Type II isotherm (at lower pressures) due to the presence of a knee but following micropore f illing the mesopores become saturated at higher pr essures. This is indicated by the observed plateau in the isotherm as P/P0 approach 1. The Type V isotherm can be rationalized as a combination of the Types III and IV isotherms. It resembles the former with respect to the weak fluid-wall interactions, yet it is similar to the la tter due to the presence of pore c ondensation. It is noteworthy to mention this type is rarely observed for porous materials, i.e. as for Type III. Lastly, the reversible Type VI isotherm is unique as compared to the aforementioned isotherms because each adsorbed layered is reflected in the form of stepwise multilayer adsorption.

PAGE 111

74 The relative height and defin ition of each step is dependent upon the surface chemistry of the adsorbent, adsorptive, a nd the temperature. Each step represents the monolayer capacity for the respective adsorbed layer and typically remains constant for two to three layers. The type of isotherm is characteri stic of adsorption on a uniform non-porous adsorbent, as exemplified by the adsorption of argon or krypton on gr aphitized carbon at a temperature of 77 K ( i.e. below the boiling point of the adsorptive). 1.12.3. Surface Area Analysis 1.12.3.1. Adsorbate Selection The standard adsorbate used for surface area analysis is nitrogen at its boiling temperature of 77 K. Alternative probe mol ecules have however been proposed and are often more reliable such as Ar at 87 K. The cross-sectional area of the adsorptive is also an important parameter to consider because its size can be affected by the surrounding environment. For example, the customary cro ss-sectional value for nitrogen at 77.35 K is 0.162 nm2 but when it is exposed to a hydroxlated surface this values is significantly reduced to 0.135 nm2 .140 This discrepancy arises from the fact that nitrogen has a permanent quadruple moment and has the potential to interact with specific groups on the surface of the adsorbent. To overcome this limitation alternative probe molecules have been investigated with similar kinetic diam eters to nitrogen and most notably include, argon (0.34nm), carbon dioxide (0.33nm), and methane (0.38nm) (see Chapter 2). Note that argon adsorption at 87.3 K is a more re liable probe molecule for pore size and surface area analysis because it does not have a quadruple. It therefore does not interact with specific sites on surface and high reso lution data can be collected in a timely

PAGE 112

75 fashion. The pores are then filled at higher re lative pressures in the case of argon (Figure 1.27.). Figure 1.27. Semi-logarithmic plot of the low pressure sorption data collected for indium soc MOF using nitrogen (red) and argon (blue) as adsorptives at 77.3 K and 87.3 K, respectively shown to highlight the differences in pore filling ranges. ( i.e. 10-7 < P/P0 <10-5 for N2 and 10-5 < P/P0 <10-3 for Ar).174 1.12.3.2. Langmuir and BET Theories The two most recognized methods for determining the apparent surface area (m2/g) is determined by applying the Langmuir225 and/or BET theories226 ( i.e. within the appropriate pressure region). The former theory assumes that the adsorbate covers the surface of the adsorbent with only one layer ( i.e. monolayer coverage ) and therefore the surface area can be obtained by experimentally determining the total number of adsorbed molecules that are required to completely cover the surface. The shortcomings of this approach are encountered when one consider s the mechanism or sequence of adsorption on the surface. At low pressures the str onger binding sites on the surface will favor adsorbate coverage over the lower energetic s ites and therefore subsequent layers will

PAGE 113

76 surely form before monolayer coverage is co mplete. This has the potential to lead to uncertainties with respect to the number of adsorbed molecules and thus the surface area. The experimentally determined surface ar ea will therefore be underestimated as compared to the theoretical values. The La ngmuir equation can be written as follows:140 where P represents the equilibrium pressure of the adsorbate, K is the Langmuir constant, and W and Wm refer to the adsorbed weight and monolayer weight, re spectively. A plot of versus P should ideally reveal a straight line. The slope and intercept correspond to and respectively. Once a solution is dete rmined for the monolayer weight, the total surface area, St, can be computed using the equation given below, where Ax is the cross-sectional area of the adsorbate, is Avogadro’s number, and is the molecular weight of the adsorbate. Three scientists by the names of Br aunauer, Emmett, and Teller modified Langmuir’s theory by accounting for multilayer adsorption, referred to as the BET theory (1938). This method can be used for surf ace area analysis of nonporous and mesoporous materials but strictly speaking it is not a ppropriate for microporous materials. The reason for this is that it is difficult to distin guish between monolayer coverage and micropore filling and therefore is more of an apparent or estimated surface area when reporting on microporous materials. This theory does however account for surface area analysis up to three layers of adsorption, which is not possi ble with the Langmuir model. Surface area

PAGE 114

77 analysis using the BET method involves two steps. The weight of the monolayer (Wm) must first be derived by transforming the adsorption isotherm in to a linear BET plot, i.e. consisting of a slope and intercept. The si mplified form of the BET equation can be written as follows:140 where W is the adsorbed weight, Wm is the weight of the monolayer, and C is an empirical constant. The latter value reflects the relative strength of the attractive adsorbate-adsorbent interactions and must always be positive ( i.e. negative value is not physical). A negative result implies that the B ET plot is outside the reasonable pressure range. A plot of versus typically yields a straight line; that is, if it is applied in the appropriate pressure regi on. The classical BET range of is between 0.05 and 0.35 atm and therfore the equations for the slope (s) and intercept (i) are represented as and respectively. Combining these equations and solving for the weight of the monolayer gene rate the following expression: while a solution for C is expressed as The second and last step in volves determining the total BET surface area using the expression below ( i.e. as described for the Langmuir theory). It is noteworthy to mention the specific su rface area is predominan tly reported in the literature, meaning the total surface area divided by the weight of the sample.

PAGE 115

78 To reiterate, the succe ss of the BET theory for microporous materials lies in choosing the appropriate relativ e pressure range to generate a linear BET plot. It must therefore display a positive value for C and the term must continuously increase with respect to Failure to do so will certainl y lead to inaccurate BET surface area analysis. To illustrate the importance of selecting the optimal pressure range, two BET plots were generated for indium soc -MOF using the argon adsorption data collected at 87 K (Figure 1.28.). Data points selected within the classical BET range (0.05 and 0.35 atm) do not yield a straight line. The appa rent surface area was estimated to be 891.5 m2/g) and the value for C is negative. Moreove r, the points lie in the decreasing portion with respect to Deleting the points above 0.04 atm however generate a linear plot with a positive value fo r C and a surface area of 1148.3 m2/g. Note that this value is 13% higher than if the classi cal BET range is applied.

PAGE 116

79 Figure 1.28. BET plots for indium soc -MOF derived from the argon adsorption data collected at 87 K: (a) a plot of versus where n is the amount adsorbed; (b) BET plot in the classical relative pressure range; and (c) a linear B ET plot generated by deleting all point above 0.04 atm.174 1.12.4. Pore volume and Pore size Analysis Experimental pore volume and pore si zes of porous materials are routinely obtained from high resolution adsorption isotherms i.e. using nitrogen (77 K) or argon (87 K). It is assumed that the pores are filled with a liquidlike fluid after micropore filling and therefore the easiest way to dete rmine the pore volume is to correlate the (c) ( b ) ( a )

PAGE 117

80 amount adsorbed at higher relative pressures ( e.g. in the plateau region) by applying the following equation:140 where represents the volume of the liquid adsorptive contained in the pores, and T are the ambient temperature and pressure, R is the gas constant, is the volume of adsorbate adsorbed, and is the molar volume of the liquid adsorbate ( e.g. 34.7 cm3 g-1 for nitrogen). Insights into pore size analysis ( e.g. using a pore size distribution) can be conducted by applying a variety of classica l macroscopic and modern approaches. The Dubinin-Raduschkevitch (D-R) method is based on Polanyi’s potential theory and is one of the classical approaches for pore volume and pore size analysis.227-230 This equation is not reliable for heterogeneous microporous adso rbents and therefore a linear fit for the adsorption data cannot be obtained. This is be cause the points are taken within a limited pressure range. To overcome this limitati on the Dubinin-Asthakov (D – A) equation was developed because to be applicable over a much wider pressure range. 231 It does not however account for adsorbate-ad sorbent interactions which are particularly important to consider in the case of microporous material s. This therefore paved the way for the development of the Horvath-Kawazoe (H-K ) method because it recognizes that the relative size and shape of the pore has a di rect effect on the adsorption properties.232 A limitation which impedes the success of this appr oach is that it is only applies to pores with a slit-like shape and it still assumes the pr operties of the fluid are identical to that of the bulk liquid. This assumption can lead to uncertainties in th e pore size and volume analysis because it does not rationall y describe the micropore filling process.140 The

PAGE 118

81 Saito-Foley method was therefore developed to account for cylindrical-like pores ( e.g. zeolites).233 The most advanced methods currently av ailable for pore size analysis of both microand mesoporous materials are based on statistical mechanics; that is, Density Functional Theory (DFT), Monte Carlo (M C) simulation, and Molecular Dynamics (MD).234, 235 A more accurate pore size can be obt ained using these methods because they recognize that the thermodyanamic state of the liquid is affected by the size and shape of the pore. Accordingly, the NLDFT method can be applied to extrapolate this data but it is imperative to choose the appropriate kernel in the sense that chosen kernel must be in agreement with the experimental data. A suitable kernel is evaluated based on the following parameters: (1) type of adsorptive and temperature ( e.g. nitrogen at 77 K); (2) Pore geometry ( e.g. cylindrical, spherical, slit-like pores); and lastly (3) Type of adsorbent ( e.g. oxidic/zeolite, car bonaceous, silica). The importance of choosing the appropria te NLDFT kernel is shown in Figure 1.29. whereby the semi-logarithmic plot of th e experimental argon ( 87.3 K) isotherm for indium soc -MOF (with cylindricallike pores) is compared to the NLDFT-Fit using both the oxidic/zeolite and carbonaceous models. No te that the surface chemistry of MOFs is more comparable to zeolites as oppose to carbonaceous materi als. The semi-logarithmic plot of the experimental data therefore appears to be in better agreement with the oxidic/zeolite method. This is confirmed fr om the corresponding por e size distribution plot which reveals a pore size of 6.1 and 12 for the oxidic/zeolite and carbon models, respectively. The estimated diameter of the channels in soc -MOF as calculated from the crystal structure is approximately 10 (point to point and not considering van der Waals

PAGE 119

82 radii). This specific NLDFT method therefore provides a very accurate assessment of the experimental pore size. Figure 1.29. (a) Semi-logarithmic plot of the experimental argon (87.3 K) isotherm for indium soc -MOF with cylindrical-like pores compared to the NLDFT-Fit using both the oxidic/zeolite and carbonaceous models and (b) Corresponding pore size distribution.174 1.13. References (1) Feynman, R. P. Eng. Sci. 1960 22. (2) Baughman, R. H.; Zakhidov, A. A.; de Heer, W. A. Science 2002 297 787-792. (3) Hoskins, B. F.; Robson, R. J. Am. Chem. Soc. 1989 111 5962-5964. (4) Clegg, W.; Blake, A. J.; Gould, R. O.; Main, P., Crystal Structure Analysis Principles and Practice Oxford University Press: Oxford, 2001; Vol. 6, p 1-265. (5) Wood, E. A., Crystal Orientation Manual Dover Publications: NY, 1977. (6) Bloss, F. D., Crystallography and Crystal Chemistry: An Introduction Holt, Rinehart and Wilson, Inc.: New York, 1971 (7) Ewald, P. P., International Union of Crystallography Utrecht: Netherlands, 1962 (8) Bragg, L., The Crystalline State Cornell University Press: Ithaca, New York, 1975 ; Vol. 1. (9) Guinier, A., X-Ray Diffraction In Crystals, Imperfect Crystals, and Amorophous Bodies 2 ed.; Unabridged Dover: San Francisco, 1994; p 1-375. (10) Steed, J. W.; Atwood, J. L., Supramolecular Chemistry John Wiley & Sons, Ltd: West Sussex, 2002 ; p 19-30. (11) Allen, F. H. Acta Crystallogr., Sect. B: Struct. Sci 2002 58 380-388. (12) Cambridge Structural Database (CSD). http://www.ccdc.cam.ac.uk/ products/csd /statistics/ (a) ( b )

PAGE 120

83 (13) Bernstein, F. C.; Koetzle, T. F.; Williams, G. J. B.; Meyer, E. F. J.; Brice, M. D.; Rodgers, J. R.; Kennard, O.; Shimanouchi, T.; Tasumi, M. J. Mol. Biol. 1977 112 535-542. (14) RCSB: Protein Data Bank. http://www.pdb.org/pdb/home/home.do (15) Inorganic Crystal St ructure Database (ICSD). http://www.fiz-karlsruhe.de/icsd.html (16) CRYSTMET Database. http://www.tothcanada.com/about.htm (17) Cotton, F. A.; Wilkinson, G., Basic Inorganic Chemistry John Wiley & Sons: New York 1976 (18) Keggin, J. F.; Miles, F. D. Nature 1936 137 (19) Ludi, A.; Guedel, H. U.; Ruegg, M. Inorg. Chem. 1970 9 2224-2227. (20) Bartoll, J.; Jackisch, B.; Most, M. ; Wenders de Calisse, E.; Vogtherr, C. M., Early Prussian Blue. Blue and green pigments in the paintings by Watteau, Lancret and Pater in the collection of Frederick II of Prussia In: TECHNE 2007 ; Vol. 25, p 3949. (21) Dunbar, K. R.; Heintz, R. A. Prog. Inorg. Chem. 1997 45 283-391. (22) Ferlay, S.; Mallah, T.; Ouahes, R.; Veillet, P.; Verdaguer, M. Nature 1995 378 701-703. (23) Sato, O.; Iyoda, T.; Fujishima, A.; Hashimoto, K. Science 1996 272 704-705. (24) Verdaguer, M.; Girolami, G. S., Magnetism: Molecules to Materials V John Wiley & Sons: 2005 ; p 283-346. (25) Yan, B.; Zhou, H.; Lachgar, A. Inorg. Chem. 2003 42 8818-8822. (26) Hofmann, K. A.; Kuspert, F. Z. Anorg. Chem. 1897 15 204-207. (27) Iwamoto, T.; Miyoshi, T.; Miya moto, T.; Sasaki, Y.; Fujiwara, S. Bull. Chem. Soc. Jpn. 1967 40 1174-1178. (28) Iwamoto, T., In Inclusion Compounds: Structural Aspects of Inclusion Compounds Formed by Inorganic and Organometallic Host Lattices Academic Press: London, 1984 ; Vol. 1, p 29-57. (29) Nishikiori, S.; Iwamoto, T. J. Struct. Chem. 1999 40 726-749. (30) Iwamoto, T. J. Inclusion Phenom. Mol. Recognit. Chem. 1996 24 61-132. (31) Nishikiori, S.; Takahashi, A.; Ratcliffe, C. I.; Ripmeester, J. A. J. Supramol. Chem. 2002 2 483-496. (32) Bowman-James, K. Acc. Chem. Res. 2005 38 671-678. (33) Nassimbeni, L. R.; Papanicolaou, S.; Moore, M. H. J. Inclusion Phenom. 1986 4 31-42. (34) Xu, H.; Li, J. Y.; Wu, Z. Z.; Z ou, J. H.; Xu, Z.; You, X. Z.; Dong, Z. C. Polyhedron 1993 12 2261-2264. (35) Nassimbeni, L. R.; Papanicolaou, S.; Moore, M. H. J. Inclusion Phenom. Mol. Recognit. Chem. 1986 4 31-42. (36) Li, B.-L.; Xu, Y.; Liu, Y.-J.; Xu, J. Chem. Res. Chin. Univ. 2000 17 237-239. (37) Abrahams, B. F.; Hoskins, B. F.; Michail, D. M.; Robson, R. Nature 1994 369 727-729. (38) Hoskins, B. F.; Robson, R. J. Am. Chem. Soc. 1989 111 5962-5964. (39) Hoskins, B. F.; Robson, R. J. Am. Chem. Soc. 1990 112 1546-1554. (40) Wells, A. F., Three-Dimensional Nets and Polyhedra Wiley-Interscience: New York, 1977 (41) Ferey, G. J. Solid State Chem. 2000 152 37-48.

PAGE 121

84 (42) Kitagawa, S.; Kitaura, R.; Noro, S. Angew. Chem. Int. Ed. 2004 43 2334-2375. (43) Eddaoudi, M.; Moler, D. B.; Li, H. L.; Chen, B. L.; Reineke, T. M.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2001 34 319-330. (44) Moulton, B.; Zaworotko, M. J. Chem. Rev. 2001 101 1629-1658. (45) Noro, S.; Kitagawa, S.; Kondo, M.; Seki, K. Angew. Chem. Int. Ed. 2000 39 20822084. (46) Macgillivray, L. R.; Subramanian, S.; Zaworotko, M. J. Chem. Commun. 1994 1325-1326. (47) Fujita, M.; Kwon, Y. J.; Washizu, S.; Ogura, K. J. Am. Chem. Soc. 1994 116 11511152. (48) Hennigar, T. L.; MacQuarrie, D. C.; Losier, P.; Rogers, R. D.; Zaworotko, M. J. Angew. Chem. Int. Ed. 1997 36 972-973. (49) Subramanian, S.; Zaworotko, M. J. Angew. Chem. Int. Ed. 1995 34 2127-2129. (50) Fujita, M.; Yazaki, J.; Ogura, K. J. Am. Chem. Soc. 1990 112 5645-5647. (51) Stang, P. J.; Zhdankin, V. V. J. Am. Chem. Soc. 1993 115 9808-9809. (52) Stang, P. J.; Cao, D. H. J. Am. Chem. Soc. 1994 116 4981-4982. (53) Stang, P. J.; Cao, D. H.; Saito, S.; Arif, A. M. J. Am. Chem. Soc. 1995 117 62736283. (54) Fujita, M.; Sasaki, O.; Mitsuhashi, T.; Fujita, T.; Yazaki, J.; Yamaguchi, K.; Ogura, K. Chem. Commun. 1996 1535-1536. (55) Uehara, K.; Kasai, K.; Mizuno, N. Inorg. Chem. 2007 46 2563-2570. (56) Schnebeck, R. D.; Randaccio, L.; Zangrando, E.; Lippert, B. Angew. Chem. Int. Ed. 1998 37 119-121. (57) Delgado-Friedrichs, O.; O'Keeffe, M.; Yaghi, O. M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2006 62 350-355. (58) Masciocchi, N.; Bruni, S.; Cariati, E.; Cariati, F.; Galli, S.; Sironi, A. Inorg. Chem. 2001 40 5897-5905. (59) Banerjee, R.; Phan, A.; Wang, B.; Knobl er, C.; Furukawa, H.; O'Keeffe, M.; Yaghi, O. M. Science 2008 319 939-943. (60) Park, K. S.; Ni, Z.; Cote, A. P.; Choi, J. Y.; Huang, R. D.; Uribe-Romo, F. J.; Chae, H. K.; O'Keeffe, M.; Yaghi, O. M. Proc. Nat. Acad. Sci. U.S.A. 2006 103 1018610191. (61) Huang, X. C.; Lin, Y. Y.; Zhang, J. P.; Chen, X. M. Angew. Chem. Int. Ed. 2006 45 1557-1559. (62) Hayashi, H.; Cote, A. P.; Furukawa, H.; O'Keeffe, M.; Yaghi, O. M. Nat. Mater. 2007 6 501-506. (63) Fujita, M.; Oguro, D.; Miyazawa, M. ; Oka, H.; Yamaguchi, K.; Ogura, K. Nature 1995 378 469-471. (64) Liu, Y. L.; Kravtsov, V. C.; Larsen, R.; Eddaoudi, M. Chem. Commun. 2006 14881490. (65) Sava, D. F.; Kravtsov, V. C.; Nouar, F.; Wojtas, L.; Eubank, J. F.; Eddaoudi, M. J. Am. Chem. Soc. 2008 130 3768-3770. (66) Alkordi, M. H.; Brant, J. A.; Wojtas, L.; Kravtsov, V. C.; Cairns, A. J.; Eddaoudi, M. J. Am. Chem. Soc. 2009 131 17753-17755. (67) Liu, Y. L.; Kravtsov, V. C.; Eddaoudi, M. Angew. Chem. Int. Ed. 2008 47 84468449.

PAGE 122

85 (68) Zaworotko, M. J. Chem. Soc. Rev. 1994 23 283-288. (69) Uemura, T.; Hiramatsu, D.; Kubota, Y.; Takata, M.; Kitagawa, S. Angew. Chem. Int. Ed. 2007 46 4987-4990. (70) He, J.; Yin, Y. G.; Wu, T.; Li, D.; Huang, X. C. Chem. Commun. 2006 2845-2847. (71) Dinca, M.; Yu, A. F.; Long, J. R. J. Am. Chem. Soc. 2006 128 8904-8913. (72) Dinca, M.; Dailly, A.; Liu, Y.; Br own, C. M.; Neumann, D. A.; Long, J. R. J. Am. Chem. Soc. 2006 128 16876-16883. (73) Ardizzoia, G. A.; LaMonica, G.; Maspero, A.; Moret, M.; Masciocchi, N. Eur. J. Inorg. Chem. 2000 181-187. (74) Zhai, Q.-G.; Lu, C.-Z.; Ch en, S.-M.; Xu, X.-J.; Yang, W.-B. Cryst. Growth Des. 2006 6 1393-1398. (75) Ardizzoia, G. A.; Cenini, S.; La Moni ca, G.; Masciocchi, N.; Maspero, A.; Moret, M. Inorg. Chem. 1998 37 4284-4292. (76) Dinca, M.; Dailly, A.; Tsay, C.; Long, J. R. Inorg. Chem. 2007 47 11-13. (77) Lin, X.; Telepeni, I.; Blake, A. J.; Dail ly, A.; Brown, C. M.; Simmons, J. M.; Zoppi, M.; Walker, G. S.; Thomas, K. M.; Mays, T. J.; Hubberstey, P.; Champness, N. R.; Schroder, M. J. Am. Chem. Soc. 2009 131 2159-2171. (78) Moulton, B.; Lu, J. J.; Hajndl, R.; Hariharan, S.; Zaworotko, M. J. Angew. Chem. Int. Ed. 2002 41 2821-2823. (79) Bourne, S. A.; Lu, J. J.; Mondal, A.; Moulton, B.; Zaworotko, M. J. Angew. Chem. Int. Ed. 2001 40 2111-2113. (80) Moulton, B.; Lu, J. J.; Mondal, A.; Zaworotko, M. J. Chem. Commun. 2001 863864. (81) Eddaoudi, M.; Kim, J.; Vodak, D.; Sudi k, A.; Wachter, J.; O'Keeffe, M.; Yaghi, O. M. Proc. Nat. Acad. Sci. U.S.A. 2002 99 4900-4904. (82) Hong, S.; Oh, M.; Park, M.; Yoon, J. W.; Chang, J. S.; Lah, M. S. Chem. Commun. 2009 5397-5399. (83) Li, H.; Eddaoudi, M.; O'Keeffe, M.; Yaghi, O. M. Nature 1999 402 276-279. (84) Eddaoudi, M.; Kim, J.; Rosi, N.; Voda k, D.; Wachter, J.; O'Keeffe, M.; Yaghi, O. M. Science 2002 295 469-472. (85) Ferey, G.; Serre, C.; Mellot-Draznieks C.; Millange, F.; Su rble, S.; Dutour, J.; Margiolaki, I. Angew. Chem. Int. Ed. 2004 43 6296-6301. (86) Ferey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F.; Dutour, J.; Surble, S.; Margiolaki, I. Science 2005 309 2040-2042. (87) Chui, S. S. Y.; Lo, S. M. F.; Charmant, J. P. H.; Orpen, A. G.; Williams, I. D. Science, 1999 283 1148-1150. (88) Kondo, M.; Yoshitomi, T.; Seki, K.; Matsuzaka, H.; Kitagawa, S. Angew. Chem. Int. Ed. 1997 36 1725-1727. (89) Li, H.; Eddaoudi, M.; Groy, T. L.; Yaghi, O. M. J. Am. Chem. Soc. 1998 120 8571-8572. (90) Cotton, F. A.; A., W. R., Multiple Bonds Between Metal Atoms Oxford University Press: Oxford, 1982 (91) Oconnor, B. H.; Maslen, E. N. Acta Cryst. 1966 20 824-&. (92) Eddaoudi, M.; Kim, J.; Wachter, J. B.; Chae, H. K.; O'Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2001 123 4368-4369.

PAGE 123

86 (93) Abourahma, H.; Coleman, A. W.; M oulton, B.; Rather, B.; Shahgaldian, P.; Zaworotko, M. J. Chem. Commun., 2001 2380-2381. (94) McManus, G. J.; Wang, Z.; Zaworotko, M. J. Cryst. Growth Des., 2003 4 11-13. (95) Furukawa, H.; Kim, J.; Plass, K. E.; Yaghi, O. M. J. Am. Chem. Soc., 2006 128 8398-8399. (96) Perry, J. J.; Kravtsov, V. C.; McManus, G. J.; Zaworotko, M. J. J. Am. Chem. Soc. 2007 129 10076-10077. (97) Nouar, F.; Eubank, J. F.; Bousquet, T. ; Wojtas, L.; Zaworotko, M. J.; Eddaoudi, M. J. Am. Chem. Soc. 2008 130 1833-1835. (98) Liu, Y. L.; Eubank, J. F.; Cairns, A. J.; Eckert, J.; Kravtsov, V. C.; Luebke, R.; Eddaoudi, M. Angew. Chem. Int. Ed. 2007 46 3278-3283. (99) Brant, J. A.; Liu, Y. L.; Sava, D. F.; Beauchamp, D.; Eddaoudi, M. J. Mol. Struct. 2006 796 160-164. (100) Sava, D. F.; Kravtsov, V. C.; Eckert J.; Eubank, J. F.; Nouar, F.; Eddaoudi, M. J. Am. Chem. Soc. 2009 131 10394-10396. (101) Jia, J. H.; Lin, X.; Wilson, C.; Blake, A. J.; Champness, N. R.; Hubberstey, P.; Walker, G.; Cussen, E. J.; Schroder, M. Chem. Commun. 2007 840-842. (102) Okubo, T.; Kondo, M.; Kitagawa, S. Synth. Met. 1997 85 1661-1662. (103) Evans, O. R.; Lin, W. B. Acc. Chem. Res. 2002 35 511-522. (104) O'Connor, C. J.; Klein, C. L.; Majeste, R. J.; Trefonas, L. M. Inorg. Chem. 1982 21 64-67. (105) Liu, Y. L.; Kravtsov, V.; Walsh, R. D.; Poddar, P.; Srikanth, H.; Eddaoudi, M. Chem. Commun. 2004 2806-2807. (106) Liu, Y. L.; Kravtsov, V. C.; Beauch amp, D. A.; Eubank, J. F.; Eddaoudi, M. J. Am. Chem. Soc. 2005 127 7266-7267. (107) Natarajan, S.; Mahata, P. Chem. Soc. Rev. 2009 38 2304-2318. (108) Delgado-Friedrichs, O.; O'Keeffe, M. J. Solid State Chem. 2005 178 2480-2485. (109) Wells, A. F. Acta Crystallogr. 1954 7 535-544. (110) Wells, A. F. Acta Crystallogr. 1954 7 545-554. (111) Wells, A. F. Acta Crystallogr. 1954 7 842-848. (112) Wells, A. F. Acta Crystallogr. 1954 7 849-853. (113) Wells, A. F. Acta Crystallographica 1955 8 32-36. (114) Wells, A. F. Acta Crystallogr. 1956 9 23-28. (115) Friedrichs, O. D.; O'Keeffe, M.; Yaghi, O. M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2003 59 22-27. (116) Blatov, V. A.; O'Keeffe, M.; Proserpio, D. M. CrystEngComm 2010 12 44-48. (117) O'Keeffe, M. Reticular Chemistry Structure Resource. http://rcsr.anu.edu.au/ (118) Moulton, B.; Abourahma, H.; Bradne r, M. W.; Lu, J. J.; McManus, G. J.; Zaworotko, M. J. Chem. Commun. 2003 1342-1343. (119) Blatov, V. A.; Carlucci, L. ; Ciani, G.; Prosperpio, D. M. Cryst. Eng. Comm. 2004 6 378-395. (120) Delgado-Friedrichs, O. 3dt The Gavrog Project http://gavrog.sourceforge.net/ (121) Delgado-Friedrichs, O.; O'Keeffe, M.; Yaghi, O. M. Phys. Chem. Chem. Phys. 2007 9 1035-1043. (122) O'Keeffe, M.; Peskov, M. A.; Ramsden, S. J.; Yaghi, O. M. Acc. Chem. Res. 2008 41 1782-1789.

PAGE 124

87 (123) Guest Editors: Long, J. R.; Yaghi, O. M. Chem. Soc. Rev. 2009 38 1201-1508. (124) Davis, S. C.; Diegel, S. W.; Boundy, R. G., Transportation Energy Data Book US Department of Energy: Washington, 2009 (125) Nissan. http://www.nissanusa.com/leaf-electriccar/index?dcp=ppn.39666654.&dcc=0.216878497#/l eaf-electric-car/index (126) Schlapbach, L.; Zuttel, A. Nature 2001 414 353-358. (127) Coontz, R.; Hanson, B. Science 2004 305 957-957. (128) Lochan, R. C.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2006 8 1357-1370. (129) McMuray, F., Chemistry 4 ed.; Pearson Prentice Hall: Upper Saddle River, 2004 (130) Gratzel, M. Nature 2001 414 338-344. (131) DOE Office of Energy Efficiency and Renewable Energy and the FreedomCAR and Fuel Partnership: Targets for Onboa rd Hydrogen Storage Systems for LightDuty Vehicles (September 2009). http://www1.eere.energy.gov/ hydrogenandfuelcells/stora ge/pdfs/targets_onboard_hy dro_storage_explanation.pdf (132) van den Berg, A. W. C.; Arean, C. O. Chem. Commun. 2008 668-681. (133) Zhao, D.; Yuan, D. Q.; Zhou, H. C. Energy Environ. Sci. 2008 1 222-235. (134) DOE Office of Energy and Efficien cy and Renewable Energy. Hydrogen Storage Technical Plan, April 2009 http://www1.eere.energy.gov/ hydrogenandfuelcells/ (135) DOE Office of Energy and Efficien cy and Renewable Energy. The FreedomCAR and Fuel Partnership, September 2009 http://www1.eere.energy .gov/hydrogenandfuelcells/ (136) Sherif, S. A.; Zeytinoglu, N.; Veziroglu, T. N. Int. J. Hydrogen Energy 1997 22 683-688. (137) Graetz, J. Chem. Soc. Rev. 2009 38 73-82. (138) Li, Y. W.; Yang, R. T. J. Am. Chem. Soc. 2006 128 726-727. (139) Li, Y. W.; Yang, R. T. J. Am. Chem. Soc. 2006 128 8136-8137. (140) Lowell, S.; Shields, J. E.; Thomas, M. A.; Thommes, M., Characterization of Porous Solids and Powders: Surface Area, Pore Size, and Density Kluwer Academic Publishers: Do rdecht / Boston / London, 2004 (141) London, F. Z. Phys. 1930 11 222-251. (142) Israelachvili, J. N., Intermolecular and Surface Forces with Applications to Colloidal and Biological Systems Academic Press: London, 1985 (143) Zhou, L. Renew. Sust. Energ. Rev. 2005 9 395-408. (144) Bogdanovic, B.; Brand, R. A.; Marj anovic, A.; Schwickardi, M.; Tolle, J. J. Alloys Compd. 2000 302 36-58. (145) Grochala, W.; Edwards, P. P. Chem. Rev. 2004 104 1283-1315. (146) Vajo, J. J.; Skeith, S. L.; Mertens, F. J. Phys. Chem. B 2005 109 3719-3722. (147) Eckert, J.; Trouw, F. R.; Mojet, B.; Forster, P.; Lobo, R. J. Nanosci. Nanotechnol. 10 49-59. (148) Kazansky, V. B.; Borokov, V. Y.; Serykh, A. I.; van Santen, R. A.; Stobbelaar, P. J. Phys. Chem. Chem. Phys. 1999 1 2881-2886. (149) Brown, C. M.; Yildirim, T.; Neumann, D. A.; Heben, M. J.; Gennett, T.; Dillon, A. C.; Alleman, J. L.; Fischer, J. E. CHem. Phys. Lett. 2000 329 311-316. (150) Ren, Y.; Price, D. L. Appl. Phys. Lett. 2001 79 3684-3686.

PAGE 125

88 (151) Dillon, A. C.; Jones, K. M.; Bekke dahl, T. A.; Kiang, C. H.; Bethune, D. S.; Heben, M. J. Nature 1997 386 377-379. (152) Rosi, N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim, J.; O'Keeffe, M.; Yaghi, O. M. Science 2003 300 1127-1129. (153) Dinca, M.; Long, J. R. Angew. Chem. Int. Ed. 2008 47 6766-6779. (154) Trewin, A.; Darling, G. R.; Cooper, A. I. New J. Chem. 2008 32 17-20. (155) Nouar, F.; Eckert, J.; Eubank, J. F.; Forster, P.; Eddaoudi, M. J. Am. Chem. Soc. 2009 131 2864-2870. (156) An, J.; Rosi, N. L. J. Am. Chem. Soc. 2010 132 5578-5579. (157) Chen, S.; Zhang, J.; Wu, T.; Feng, P.; Bu, X. J. Am. Chem. Soc. 2009 131 1602716029. (158) Yan, Y.; Telepeni, I.; Yang, S. H.; Li n, X.; Kockelmann, W.; Dailly, A.; Blake, A. J.; Lewis, W.; Walker, G. S.; Allan, D. R.; Barnett, S. A.; Champness, N. R.; Schroder, M. J. Am. Chem. Soc. 2010 132 4092-4094. (159) Koh, K.; Wong-Foy, A. G.; Matzger, A. J. J. Am. Chem. Soc. 2009 131 41844185. (160) Furukawa, H.; Miller, M. A.; Yaghi, O. M. J. Mater. Chem. 2007 17 3197-3204. (161) Koh, K.; Wong-Foy, A. G.; Matzger, A. J. Angew. Chem. Int. Ed. 2008 47 677680. (162) Latroche, M.; Surble, S.; Serre, C.; Mell ot-Draznieks, C.; Llewellyn, P. L.; Lee, J. H.; Chang, J. S.; Jhung, S. H.; Ferey, G. Angew. Chem. Int. Ed. 2006 45 8227-8231. (163) Rowsell, J. L. C.; Yaghi, O. M. J. Am. Chem. Soc. 2006 128 1304-1315. (164) Wong-Foy, A. G.; Matzger, A. J.; Yaghi, O. M. J. Am. Chem. Soc. 2006 128 3494-3495. (165) Zhao, D.; Yuan, D. Q.; Sun, D. F.; Zhou, H. C. J. Am. Chem. Soc. 2009, 131 9186-9188. (166) Kaye, S. S.; Dailly, A.; Yaghi, O. M.; Long, J. R. J. Am. Chem. Soc. 2007 129 14176-14177. (167) Rowsell, J. L. C.; Millward, A. R.; Park, K. S.; Yaghi, O. M. J. Am. Chem. Soc. 2004 126 5666-5667. (168) Yan, Y.; Lin, X.; Yang, S. H.; Blake, A. J.; Dailly, A.; Champness, N. R.; Hubberstey, P.; Schroder, M. Chem. Commun. 2009 1025-1027. (169) Ma, S. Q.; Sun, D. F.; Ambrogio, M.; Fillinger, J. A.; Parkin, S.; Zhou, H. C. J. Am. Chem. Soc. 2007 129 1858-1859. (170) Zhou, W.; Wu, H.; Yildirim, T. J. Am. Chem. Soc. 2008 130 15268-15269. (171) Zhong, D.-C.; Lin, J.-B.; Lu, W.-G.; Jiang, L.; Lu, T.-B. Inorg. Chem. 2009 48 8656-8658. (172) Dinca, M.; Long, J. R. J. Am. Chem. Soc. 2007 129 11172-11176. (173) Forster, P. M.; Eckert, J.; Heiken, B. D.; Parise, J. B.; Yoon, J. W.; Jhung, S. H.; Chang, J. S.; Cheetham, A. K. J. Am. Chem. Soc. 2006 128 16846-16850. (174) Moellmer, J.; Celer, E. B.; Luebke, R.; Cairns, A. J.; Staudt, R.; Eddaoudi, M.; Thommes, M. Microporous Mesoporous Mater. 2010 129 345-353. (175) Kosal, M. E.; Chou, J. H.; Wilson, S. R.; Suslick, K. S. Nat. Mater. 2002 1 118121. (176) Kitaura, R.; Onoyama, G.; Sakamot o, H.; Matsuda, R.; Noro, S.; Kitagawa, S. Angew. Chem. Int. Ed. 2004 43 2684-2687.

PAGE 126

89 (177) Cho, S. H.; Ma, B. Q.; Nguyen, S. T.; Hupp, J. T.; Albrecht-Schmitt, T. E. Chem. Commun. 2006 2563-2565. (178) Wang, X. S.; Ma, S. Q.; Forster, P. M.; Yuan, D. Q.; Eckert, J.; Lopez, J. J.; Murphy, B. J.; Parise, J. B.; Zhou, H. C. Angew. Chem. Int. Ed. 2008 47 7263-7266. (179) Arean, C. O.; Manoilova, O. V.; Bone lli, B.; Delgado, M. R.; Palomino, G. T.; Garrone, E. Chem. Phys. Lett. 2003 370 631-635. (180) Arean, C. O.; Nachtiga llova, D.; Nachtigall, P.; Garrone, E.; Delgado, M. R. Phys. Chem. Chem. Phys. 2007 9 1421-1436. (181) Bushnell, J. E.; Kemper, P. R.; Bowers, M. T. J. Phys. Chem. 1994 98 20442049. (182) Mulfort, K. L.; Hupp, J. T. J. Am. Chem. Soc. 2007 129 9604-9605. (183) Kubas, G. J.; Ryan, R. R.; Swanson, B. I.; Vergamini, P. J.; Wasserman, H. J. J. Am. Chem. Soc. 1984 106 451-452. (184) Kubas, G. J. Chem. Rev. 2007 107 4152-4205. (185) Kubas, G. J., Metal Dihydrogen and -Bond Complexes, Structure, Theory, and Reactivity Kluwer-Academic/Plenum Publishers: New York, 2001 (186) Chen, B. L.; Eddaoudi, M.; Reineke, T. M.; Kampf, J. W.; O'Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2000 122 11559-11560. (187) Prestipino, C.; Regli, L.; Vitillo, J. G.; Bonino, F.; Damin, A.; Lamberti, C.; Zecchina, A.; Solari, P. L.; Kongshaug, K. O.; Bordiga, S. Chem. Mater. 2006 18 1337-1346. (188) Liu, Y.; Kabbour, H.; Brown, C. M.; Neumann, D. A.; Ahn, C. C. Langmuir 2008 24 4772-4777. (189) Georgiev, P. A.; Albinati, A.; Mo jet, B. L.; Ollivier, J.; Eckert, J. J. Am. Chem. Soc. 2007 129 8086-8087. (190) Eckert, J. Spectrochim. Acta, Part A 1992 48, 271-283. (191) Curl, R. F.; Hopkins, H. P.; Pitzer, K. S. J. Chem. Phys. 1968 48 4064-4070. (192) Spencer, E. C.; Howard, J. A. K.; McIntyre, G. J.; Rowsell, J. L. C.; Yaghi, O. M. Chem. Commun. 2006 278-280. (193) US Energy Policy Act of 1992 (EPAct). (194) Burchell, T.; Rogers, M., SAE Tech. Pap. Ser. 2000 2000-01-2205. (195) Muris, M.; Dupont-Pavlovsky, N.; Bienfait, M.; Zeppenfeld, P. Surf. Sci. 2001 492 67-74. (196) Duren, T.; Sarkisov, L.; Yaghi, O. M.; Snurr, R. Q. Langmuir 2004, 20 26832689. (197) Bourrelly, S.; Llewellyn, P. L.; Serre C.; Millange, F.; Lo iseau, T.; Ferey, G. J. Am. Chem. Soc. 2005 127 13519-13521. (198) Kondo, M.; Shimamura, M.; Noro, S.; Minakoshi, S.; Asami, A.; Seki, K.; Kitagawa, S. Chem. Mater. 2000 12 1288-1299. (199) Ma, S.; Sun, D.; Simmons, J. M.; Collier, C. D.; Yuan, D.; Zhou, H.-C. J. Am. Chem. Soc. 2007 130 1012-1016. (200) Energy Information Administration, Official Energy Statistics from the U.S. Government, International Energy Annual 2005. http://www.eia.doe.gov/aer/ (201) U. S. Department of Ener gy Carbon Dioxide Sequestration Program. http://fossil.energy.gov/progr ams/sequestration/index.html

PAGE 127

90 (202) Yu, K. M. K.; Curcic, I.; Gabriel, J.; Tsang, S. C. E. ChemSusChem 2008 1 893899. (203) Leci, C. L. Energy Convers. Manage. 1996 37 915-921. (204) Figueroa, J. D.; Fout, T.; Plasyns ki, S.; McIlvried, H.; Srivastava, R. D. Int. J. Greenhouse Gas Control 2008 2 9-20. (205) Huang, J.; Zou, J.; Ho, W. S. W. Ind. Eng. Chem. Res. 2008 47 1261-1267. (206) Rao, A. B.; Rubin, E. S. Environ. Sci. Technol. 2002 36 4467-4475. (207) Li, J. R.; Kuppler, R. J.; Zhou, H. C. Chem. Soc. Rev. 2009 38 1477-1504. (208) Walton, K. S.; Millward, A. R.; Dubbeldam, D.; Frost, H.; Low, J. J.; Yaghi, O. M.; Snurr, R. Q. J. Am. Chem. Soc. 2008 130 406-407. (209) Keskin, S.; Sholl, D. S. J. Phys. Chem. C 2007 111 14055-14059. (210) Bae, Y. S.; Mulfort, K. L.; Frost, H.; Ryan, P.; Punnathanam, S.; Broadbelt, L. J.; Hupp, J. T.; Snurr, R. Q. Langmuir 2008 24 8592-8598. (211) Wang, B.; Cote, A. P.; Furukawa, H.; O'Keeffe, M.; Yaghi, O. M. Nature 2008 453 207-211. (212) Bastin, L.; Barcia, P. S.; Hurtado, E. J.; Silva, J. A. C.; Rodrigues, A. E.; Chen, B. J. Phys. Chem. C. 2008 112 1575-1581. (213) Phan, A.; Doonan, C. J.; Uribe-Romo, F. J.; Knobler, C. B.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2010 43 58-67. (214) Millward, A. R.; Yaghi, O. M. J. Am. Chem. Soc. 2005 127 17998-17999. (215) Llewellyn, P. L.; Bourrelly, S.; Serre C.; Vimont, A.; Daturi, M.; Hamon, L.; De Weireld, G.; Chang, J. S.; Hong, D. Y.; Hwang, Y. K.; Jhung, S. H.; Ferey, G. Langmuir 2008 24 7245-7250. (216) Surble, S.; Millange, F.; Serre, C. ; Duren, T.; Latroche, M.; Bourrelly, S.; Llewellyn, P. L.; Ferey, G. J. Am. Chem. Soc. 2006, 128 14889-14896. (217) An, J. Y.; Fiorella, R. P.; Geib, S. J.; Rosi, N. L. J. Am. Chem. Soc. 2009 131 8401-8403. (218) Cavenati, S.; Grande, C. A.; Rodrigues, A. E. J. Chem. Eng. Data 2004 49 10951101. (219) Himeno, S.; Komatsu, T.; Fujita, S. J. Chem. Eng. Data 2005 50 369-376. (220) Britt, D.; Tranchemontagne, D.; Yaghi, O. M. Proc. Nat. Acad. Sci. U.S.A. 2008 105 11623-11627. (221) Nelson, A. P.; Farha, O. K.; Mulfort, K. L.; Hupp, J. T. J. Am. Chem. Soc., 2009 131 458-460. (222) Cooper, A.; Howdle, S. Mater. World 2000, 8 10-12. (223) Ma, L. Q.; Jin, A.; Xie, Z. G.; Lin, W. B. Angew. Chem. Int. Ed. 2009 48 99059908. (224) Sing, K. S. W.; Everett, D. H.; Ha ul, R. A. W.; Moscou, L.; Pierotti, R. A.; Rouquerol, J.; Siemieniewska, T. Pure Appl. Chem. 1985 57 603-619. (225) Langmuir, I. J. Am. Chem. Soc. 1918 40 1368. (226) Brunauer, S.; Emmett, P. H.; Teller, E. J. Am. Chem. Soc. 1938 60 309-319. (227) Polyanyi, M. Verh. Dtsch. Phys. Ges. 1914 16 1012. (228) Dubinin, M. M.; TImofeev, D. P. Zh. Fis. Khim. 1948 22 133. (229) Dubinin, M. M.; Radushkevich, L. V. Dokl. Akad. Nauk. SSSR 1947 55 331. (230) Dubinin, M. M. Chem. Rev. 1960 60 235-241. (231) Dubinin, M. M.; Astakhov, V. A. Adv. Chem. Series 1970 102 69.

PAGE 128

91 (232) Horvath, G.; Kawazoe, K. J. Chem. Eng. Jpn. 1983 16 470-475. (233) Saito, A.; Foley, H. C. Microporous Mater. 1995 3 531-542. (234) Evans, R.; Marconi, U. M. B.; Tarazona, P. J. Chem. Soc., Faraday Trans. 1986 82 1763. (235) Tarazona, P.; Marconi, U. M. B.; Evans, R. Mol. Phys. 1987 60 573-595.

PAGE 129

92 Chapter 2. Design and Synthesis of Porous Isostructural MOMs with “soc” Topology: Insights into Adsorption Characterization and Adsorbate-MOF Interactions 2.1. Introduction Self –assembly of molecular building bl ocks (MBBs) with prefabricated shapes, geometries, and functionality that are amenab le to design from first principles, have proven to facilitate the cons truction of discrete and ex tended functional solid-state crystalline materials with specific functionalities.1-8 The success of this approach is exemplified by the burgeoning number of me tal-organic polyhedra (MOPs), metalorganic frameworks (MOFs), covalent orga nic frameworks (COFs), and coordination polymers, reported in the literature on a daily basis.9-15 The MBB approach is a rational top-down design, bottom-up synthetic strategy and thus is inherently modular and predictable (to a certain degree). Poten tial made-to-order MOMs can therefore be targeted for their ability to exhibit desired properties by employing the appropriate MBBs that possess the warranted structur al and chemical attributes. The unique inorganic-organic hybrid nature a characteristic feature of MOMs, together with effective design strategies, i. e. reticular chemistry, and facile synthesis makes it possible to construct rigid and di rectional pre-designed MBBs of varying connectivity by exploiting the diverse coor dination modes and geometries provided by metal ions and organic molecules.From a design perspective, inorganic and organic MBBs can be translated into building units (BUs) that are analogous to geometrical

PAGE 130

93 shapes. The BU is defined by connecting th e points of extensio n of the respective building block via the carboxylato carbon at om. The vertex figure (or coordination figure) of the original n -connected node(s) is thereby rev ealed and this representation corresponds to the augmente d conformation of the net.16 Inorganic MBBs ( i.e. metal clusters) which are typically formed in situ and organic MBBs are th erefore targeted prior to the assembly process for the desired shape, geometry, and directi onality necessary to augment a specific predicted net. Highly symmetrical building blocks with 3-, 4-, and/or 6-connectivity are commonly expl oited in solid-state chemistr y, as their vertex figures correspond to regular polygons and/or polyhedr a such as those depicted in Figure 2.1.17-22 Figure 2.1. Commonly employed inorganic and organic MBBs and their corresponding augmented forms: (a) BTEC orga nic MBB, square BU; (b) [Zn4O(RCO2)6] MBB, octahedral SBU; (c) BTC organic MBB; triangular BU; and (d) [M3O(RCO2)6(H2O)3] trimer MBB (TMBB), trigonal prismatic SBU. Porous MOFs have gained much momentum in recent years as a promising class of solid-state materials to be utilized as potential hydrogen storage mediums (host) for onboard vehicular applications.11, 23-26 MOMs are ideal physisorption-based candidates (a) (b) (c) (d)

PAGE 131

94 because by their very nature they exhi bit fast kinetics and fully reversible adsorption/desorption isotherms.27-29 Molecular hydrogen can therefore be readily released under mild conditions (in contrast to chemisorp tion methods). Several MOFs have certainly exceeded expectations whereby unprecedented surface areas, uptake capacities, and Q st values have been reported.19, 30-35 Many key obstacles however must be overcome before a hydrogen-based econom y can become an efficient and safe alternative fuel to fulfill future energy demands. A significant limitation which impedes the development of MOFs as potential solid-state hydrogen storage materials is dominated by weak binding affinities between molecular hydrogen and the constituents of the framework, i.e. typically in the range of ~ 5 – 8 kJ mol-1.36 These fundamental interactions pr edominately arise from a collective group of forces; that is, van der Waals, charge-quadrupole, a nd induction forces.37, 38 To date no porous physisorbent based material has in fact concomitantly sustained the optimal binding energy (15 – 20 kJ mol-1)39 and uptake capacity required for on-board applications.40 In their present state it is therefore impossible for these materials to store adequate amounts of hydrogen at operable temperatures and pr essures. It is this very reason why it is imperative to devi se pathways to increase the H2 binding energies in a systematic fashion to afford stronger bindi ng affinities via stru ctural and chemical modifications (see Chapter 1).41 We sought to gain a better understandi ng of the preferential binding of hydrogen in porous MOFs by conducting a systematic adsorption study by synthesizing an isostructural series of MOFs having soc topology (= s quare oc tahedron).20, 26, 42 The MOFs described herein, 1 – 5 were constructed from the se lf-assembly of predesigned 6-

PAGE 132

95 connected oxygen-centered me tal trimer molecular building blocks (TMBBs) having trigonal prismatic geometry and a 4-c onnected rectangular-planar tetracarboxylate organic ligand (3,3',5,5'-azobe nzenetetracarboxylic acid, H4-ABTC). To the best of our knowledge, these compounds represent the only examples of MOFs having soc topology. The unique features offered by the prototypical platform include: (1) Narrow pores (< 1 nm); (2) Higher localized charge density; a nd (3) Accessibility to potential vacant metal sites. All of these parameters are recognized in literature as a means to improve the binding energy of hydrogen in MOFs; however, systematic studies that independently evaluate the effect of each parameter remain scarce.20, 43, 44 It therefore seems highly plausible that the rational construction of a MOF having a large surface area and bearing the aforementioned characteristics would repr esent an ideal candidate to permit the attainment of the DoE systems targets for H2 uptake at operable temperatures and pressures. An increasing number of MOFs are being constructed in recent years using rigid and flexible tetracarboxylate organic ligands, with the nbo pts and lvt nets being predominately observed.45-48 A dominant factor which prevents the formation of the soc net using other tetracarboxylat e ligands is the relative scal e and geometry of the ligand. To generate a MOF with soc topology using a longer ligand would require expansion in both directions ( x and y ). This is because the faces of the cuboidal cages are delimited by the ligands, which are oriented perpendicula r with respect to the neighboring face. The H4-ABTC ligand was therefore deliberately chosen for this project to link the TMBBs because the distance between th e azo moieties is precisely th e right size to connect the opposing m -BDC units of neighboring trimers. It is important to note however that the

PAGE 133

96 soc net is the most symmetrical outcome for the assembly of octahedral and square BUs. In our case, H4-ABTC is not a perfect square and th e inorganic MBB exhibits trigonal prismatic geometry which is certainly not octahedral. Accordi ngly, the rectangular geometry of the ligand causes elongation al ong one direction which thereby permits the formation of trigonal prismatic BUs as oppos e to octahedral BUs but the coordination sequence is still a match to soc topology. The structural description provided in this chapter c oupled with the sorption and INS data will reveal that th e unique structural features o ffered by this platform provide valuable insights for enhancing H2-MOF interactions. Acco rdingly, fine-tuning these features to further enhance these interactions plays an important role in governing the maximum uptake capacity and enthalpy of adsorption. The results will independently assess the effect of the following parameters on H2-MOF interactions using the cationic soc -MOF as a platform: (1) Effect of meta l cation whereby the TMBB has the potential to be replaced with different metal cations ( e.g. In, Al, Cr, Mn, Fe, Ni, etc ) to generate accessible open-metal centers and lower framew ork densities; (2) Fine-tuning the extraframework counter ions which are confined in the cuboidal cage ( e.g. NO3 -, Cl-, Br-, etc ); and (3) Effect of polarizabil ity via the incorporation of decorated TMBBs; that is, one of the axial sites is occupied by a halide which thereby leaves the cuboidal cage anion-free. These modifications were accomplished in situ while preserving the st ructural integrity of the frameworks.

PAGE 134

97 2.2. Results and Discussion The first MOF possessing soc topology was isolated from the solvothermal reaction of H4-ABTC with In(NO3)32H2O in a DMF/CH3CN solution in the presence of piperazine.20 This led to the formation of orange polyhedral crystals characterized and formulated by elemental microanalysis and SCD studies as {[In3O(ABTC)1.5(H2O)3](NO3)(H2O)3}, ( 1 ). The purity of 1 was confirmed by similarities between the experimental and calculated powder X-ra y diffraction (PXRD) patterns (Appendix B). Compound 1 crystallizes in the cubic space group P43 n with a = 22.4567(11) and has a unit cell volume of 11325.0(10) 3. The 3-periodic framework is built up from 3-oxygen-centered indium TMBBs whereby each indium cation adopts an octahedral coordination environment (Fi gure 2.2.). The equatorial plane of the TMBB is occupied by four deprot onated carboxylate oxygen atoms (dIn-O = 2.162(2) ) from four independent ABTC4ligands, each of which coordinate in a bis-monodentate fashion. A terminal aqua ligand (dIn-O = 2.173(6) ) and a central 3-oxo anion (dIn-O = 2.040(3) ) are bound in the axial positions and complete the coordination sphere of the metal cation, [InO5(H2O)]. The central oxo anion (O2-) is located on a th reefold axis and unites the three indium octahedra, which leads to three planar In-( 3-O)-In angles of 120o.

PAGE 135

98 Molecular Building Blocks Building Units Figure 2.2. Select fragments from the X-ray crystal structure of 1 : (a) In-TMBB, generated in situ as [In3O(RCO2)6(H2O)3] and the organic ligand, H4-ABTC; (b) Inorganic and organic MBBs can be viewed as a 6-connected node having trigonal pr ismatic geometry and a 4-connected node with rectangular-planar geometry, respectively; (c) Ball-and-stick representation of the cage that houses the partially occupied [NO3]ions; and (d) Polyhedral bui lding unit representation of the cage. (center) Optical image of 1 to show the crystal morphology of the as-synthesized compound. Color code: In = green; C = gray; N = blue; O = red. Hydrogen atoms and solvent molecules are omitted for clarity. Each pair of indium centers are br idged together via two independent ABTC4ligands that coordinate to each indium in a bis-monodentate fashion through the carboxylato oxygen atoms. Each TMBB is linked t ogether by six separa te organic linkers to result in the formation of a novel 3-periodic structure. In 1 each indium atom is trivalent and the 3-oxo anion contributes a negative two charge and therefore this yields an overall cationic framework (+1 per formula unit). The charge balance is provided by disordered extra-framework nitrate, [NO3]-, anions which statistically occupy two (a) (b) (c) (d)

PAGE 136

99 positions on the three fold axis with an equa l probability of being located at either position. The nanometer –scale car cerand-like cage therefore en capsulates a total of four [NO3]ions. The cage is cube-like in shape and is delimited by six ABTC4ligands and eight TMBBs, which occupy the faces and vertic es of the cage, respectively. Each cage is shared between eight neighboring cages via the TMBBs and thereby reveals a periodic array of connected cages and infinite intersecting channels resembling a bcc arrangement. The cage is comprised of small window aper tures that measure approximately 4.612 x 2.431 , taking into considering van der Waals radii. Consequently, the [NO3]anions are trapped due to steric hindrance and so the cage remains inaccessible to guest molecules in the absence of external forces, i.e. pressure. As eluted to in the introduction and th e title of this ch apter, all compounds described herein possess the same c onnectivity; that is, they all have soc topology.42 The TMBB can be rationalized as a 6-connected trigonal prismatic building unit and can be regarded as a pseudo-octahedron. The 4-conne cted node is represented by the organic MBB, H4-ABTC, which is rationalized as rect angular-planar node, or pseudo-square. Accordingly, the assembly of such 4and 6-connected nodes results in the formation of an edge-transitive 3-periodic network with soc topology. Compound 1 exhibits an interesting pore system because it contains two well-defined types of infinite channels, namely an intersecting hydrophobic and hydr ophilic channel system (Figure 2.3.).

PAGE 137

100 Figure 2.3. Select fragment from the crystal structure of 1 to emphasize the packing of the cages and highlight the two types of intersecting cha nnels. The green tubular rods run through the hydrophobic channels, while the pink tubular r ods run through the hydro philic channels. Color code: In = green; C = gray; N = blue; O = red. Hydrogen atoms, solvent molecules, and [NO3]anions are omitted for clarity. The relative span of the cage is estimated to be 10 , while the channels measure approximately 6.5 . The hydrophilic channel syst em sustains a polar environment due to the presence of aqua ligands bound in the ax ially positions of the indium centers. The water molecules are directed inward towards these channels and participate in hydrogen bonding with other guest water molecules located in these channels. The hydrophobic channels however were determined to be completely guest-free in the as-synthesized material. 2.2.1. Low Pressure Adsorption Measurements for 1. The interesting structural and chemical features of 1, that is, narrow pores with higher localized charge density and potent ially accessible open-metal binding sites,

PAGE 138

101 prompted us to investigate the gas storage properties of 1 The total solvent accessible volume was obtained using the PLATON softwa re by summing the voxels that are more than 1.2 away from the framework.49 For 1 it was determined to be approximately 57.2 % of the unit cell volume. Gas-sorpti on studies were conducted on the fully evacuated sample following exchange in CH3CN for a period of 36 h. The exchanged sample was loaded into a 6-mm sample cell (dry) and evacuated at room temperature for 24 h and then gradually heated to 135oC for 8 h to afford the desolvated form of 1 The permanent porosity was confirmed by the nitroge n sorption isotherm collected at 77 K. It revealed a fully reversible Type I isothe rm, which is characteristic of microporous materials. The apparent Langmuir surface area was estimated to be 1417 m2/g with a corresponding experimental pore volume of 0.50 cm3g-1. This value agrees very well with the theoretical value of 0.51 cm3g-1. The hydrogen capacity was assessed at atmospheric pressures and temperatures of 77 K and 87 K (Figure 2.4a.). Regardle ss of the fact that 1 has a moderate free volume and exhibits a lo wer surface area as compared to other MOFs ( i.e > 3000 m2/g) it was still able to store a significant amount of H2. It was found to store up to 2.61 wt% at 77 K and 895 Torr (1.2 atm), as shown in Figure 2.4b.

PAGE 139

102 Figure 2.4. Sorption data collected on 1 : (a) Nitrogen sorption isotherm at 77 K; (b) Hydrogen sorption isotherm at 77 K and 87 K; and (c) Isosteric heat of adsorption for H2. The ability of 1 to adsorb such high amounts of H2 indicates a higher density of H2 in the pores ( i.e. 0.05 g cm-3) and therefore the bulk properties of the H2 adsorbed in 1 changed from that of a gas to resemble that of a liquid-like state (0.0708 g cm-3 at 20 K). The fact that 1 is able to achieve this density at 77 K and atmosphere pressures is interesting and most likely dr iven by the overlap of poten tial energy fields from the narrow pore dimensions, which facilitate st rong adsorbate-adsorben t interactions. In (a) (b) (c)

PAGE 140

103 addition to the collective contributions from the residual charge density from the extraframework nitrates anions a nd accessibility to potentially vacant In(III) metal binding sites. The corresponding isosteric h eat of adsorption for H2 was calculated from the 77 K and 87 K H2 sorption data, up to 1.8% loading of H2 per sorbent weight and determined to be 6.5 kJ mol-1 (Figure 2.4c.) The Q st is higher than values reported for porous carbon materials, yet similar to those reported for an assortment of MOFs. This may not seem all that significant at low coverage but it is remarkable that it is maintained even at higher loadings. This observation can be attributed to a significant averaging of the binding sites which is a unique feat ure that is rarely observed in MOFs.26, 44 In most reported cases Q st is significantly lower at higher c overage as compared to low loadings because the structural and chemical featur es of the MOF become less accessible to the adsorbate ( i.e. open-metal sites, organic functionality, counter ions, etc ). In other words, the stronger H2-MOF binding sites become fully occupi ed at lower loadings and thus are less accessible at higher loadings which cause Q st to decrease. 2.2.2. Inelastic Neutron Scatteri ng (INS) Studies for 1. Owing to the fact 1 can adequately store such high amounts of H2 and sustain the H2-MOF interactions even at higher loadings illustrate s the importance of potential vacant metal sites, higher localized charge density, and pore dimensions on the energetics of sorbed H2 molecules in MOFs. INS experime nts were therefore carried out on 1 in order to gain a better under standing of the preferential H2-MOF sorption sites in this material. This technique is well suited to probe H2 interactions because it is sensitive to

PAGE 141

104 the chemical environment of the host mate rial that causes th e hindered rotational transitions of the adsorbed hydr ogen molecules (see Chapter 1).50-55 The INS experiments conducted on 1 were carried out on the QENS spectrometer (neutron energy loss) at th e Institute Pulsed Neutr on Source at Argonne National Laboratory. Approximately 2g of the fully exchanged and evacuated sample was transferred to an aluminum sample holder, sealed under a He atmosphere, and connected to an external gas dosing system. A “b lank” spectrum was first collected on 1 in the absence of hydrogen. It was then warmed to approximately 70 K and hydrogen was adsorbed in-situ at various loadings e quivalent to one, two, three, five, and seven H2 molecules per indium atom. The sample was th en cooled to 15 K for data collection to observe the respective interactions of molecular hydrogen with the framework. The INS spectra for H2 in 1 reveals reasonably well-defined peaks even at the lowest loading (1 H2/In), i.e. after subtracting the blank spectrum. Multiple peaks are observed in the spectra, even at the initial loading of 1 H2/In. This is indicative of multiple binding sites which are occupied simultaneously and is evidenced by the observable peaks in the range of 12 to 16 meV and another at 25.3 meV (Figure 2.5.). Note that the In-TMBB, [In3O(RCO6)], contains three potentia lly vacant metal sites and therefore one would predict to observe only one peak at th e lowest loading since the interaction of H2 with a metal center ought to be the strongest. Conversely, the occupancy of multiple sites in MOFs having no vacant metal sites is expected as the binding energies are typically within a close range. Ne vertheless, the presence of multiple sites at the lowest loading is in accord with the overa ll consistency observed in the isosteric heat of adsorption for 1 which is maintained up to 1.8 wt% of H2.

PAGE 142

105 Figure 2.5. INS spectra of 1 obtained at 15 K with loadings of 1, 3, and 5 H2 per indium. The relative peak positions (Table 2.1) in the INS spectra of 1 were assigned on the basis of the phenomenological model (see section 2.3.2.). It is important to keep in mind that a hindered transition observed at lo wer energy is caused by a higher barrier to rotation with respect to H2. This indicates a stronger interaction which means that particular binding site(s) should be occupied first (at lower loadings). In the case of 1 the most intense bands are labeled as the 0-1 a nd 0-2 transitions which correspond to weakly bound (physisorbed) H2 at different sites in the backbone of the framework. In lieu of the previous discussion concerning th e barrier to rotation, the pote ntially vacant indium site was tentatively assigned (in 2007) as the peak located at 25.3 meV. This corresponds to the 0 – 2 transition for weakly bound H2. It is noteworthy to mention that due to computational advances specu lation currently surrounds the origin of this peak. The original phenomenological model may not have been appropriate for this system ( i.e. over simplified). The presence of a 0 – 2 transiti on stipulates that the 0 – 1 transition must be present at 4.8 meV. The relatively low va lue of this observed rotational transition ( i.e.

PAGE 143

106 high barrier to rotation) suggests a stronger overa ll interaction as compared to the other 01 transitions observe d at 12.7 and 14.1 meV. Table 2.1. Tentative rotational transitions (meV) assi gned for the hydrogen adsorbed at different sites on 1 Transition 0-1 0-2 1-2 (calcd) Barrier (V/B) Sites 4.8 25.3 20.5 6.8 12.7 15.9 3.2 1.05 13.3 15.4 2.1 0.7 14.1 15.0 1.1 0.3 Additional Sites 11.8 16.4 4.6 1.55 11.2 16.9 5.7 1.9 ~9.5 18.3 8.7 3.0 Accordingly, we speculate that the latte r transitions arise from interactions associated with the weaker binding sites near the cationic TMBB; while the former peak arises from interactions with the organic components of the framework ( i.e. carboxylates, azo bridge, phenyl moieties). This assumpti on was supported following the addition of 3 H2/In because the intensity of the non-indium sites increased owing to subsequent filling of H2. There was little change howev er in the relative intensity of the peak assigned to the vacant indium site. This indicates the stronge r binding site is predominately occupied after the initial loading. When the dosage of H2 is subsequently increased to 5 H2/In the aforementioned non-indium binding sites did not gain intensity but instead they became overshadowed by new peaks at even lower energies than t hose observed after loadings of 1 and 3 H2/In, as highlighted in the difference spectra in Figur e 2.6. The fact that non-indium sites become populated at higher loadings is not surp rising and is often observed in MOFs.41, 44, 50 The

PAGE 144

107 astonishing observation is th at these binding sites corres pond to stronger non-indium interactions (higher barrier to rotation) than those observed in lowe r loadings whereby the H2-MOF interactions are typically weak in nature. It is important to mention that in situ loading of 5 H2/In and higher surpass 1.8 wt% of H2 and approaches the maximum uptake capacity of 2.61 wt% at 77 K and 1.2 atm. This is relevant to the discussion because we hypothesize that these additio nal binding sites become available only at higher loading due to pressure induced interactions within the isolated na nometer-scale cuboidal cages which ho use the nitrate ions. Figure 2.6. Difference spectra calculated for 1 between 3H2/In – 1H2/In and 5H2/In and 3H2/In. At lower loadings ( i.e. 1 and 3 H2/In) we speculate that the H2 is unable to overcome the sorption kinetic-energy diffusion ba rrier to facilitate access into the cage owing to the steric constraints imposed by the window apertures. Thus, at lower loadings H2 is preferentially occupied at the vacant indium centers and throughout the intersecting channel system. At higher loadings however th e framework is subject to elevated levels

PAGE 145

108 of H2 (compressibility factor) and the isolated cages become accessible near the entrances of the cage. The increased local pressure of H2 allows the kinetic ba rrier to be overcome and permits access to the once inaccessible cage. The confined dimensions of the cavity coupled with the higher localized charge dens ity induced from the nitrate ions and the proximal distance to the cationic TMBBs pr omote stronger sorption sites and thus correspond to increased H2-MOF interactions. This evid enced by the strong binding sites in the INS at higher loadings (5 H2/In). These findings support the elevated density of H2 (0.05 g cm-3) observed in the pores at 77 K and atmospheric pressures. 2.2.3. High Resolution Physical Adsorpti on Measurements: A Systematic Benchmark Study on 1. In order to explore th e vast potential of 1 as well as, other MOFs for existing and novel applications ( e.g. gas storage and separation) it is imperative to fully characterize the materials using a combination of experi mental techniques such as X-ray diffraction and physical adsorption. The data provides esse ntial structural details that compliment the quantitative physisorption measurements c oncerning the surface area, porosity, relative pore size, as well as, the re lative strength of the adsorb ate-MOF interactions. This therefore allows for a reliable assessment concerning the adsorption capacity and surface energetics with respect to different ad sorptives for gas storage and separation applications. The preliminary low pre ssure gas sorption data collected on 1 using nitrogen and hydrogen as adsorptives revealed promising results, as discussed above. This prompted us to conduct a comprehens ive and systematic phys ical adsorption study to evaluate the uptake capacity and Q st of 1 using a variety of adsorptives ( i.e. H2, CH4, and CO2) over a wide range of temperatures and pressures. High resolution hydrogen

PAGE 146

109 isotherms were collected for 1 at three closely spaced temp eratures of 87 K, 97 K, and 107 K.26 These measurements were performed in order to validate the previously reported Q st of 6.5 kJ mol-1. The reason for choosing the aforementione d adsorptives was twofold: (1) We were interested in assessing their reliability as alternative probe molecules for pore size and surface area analysis as compar ed to traditional adsorptives ( i.e. nitrogen and argon); and (2) the uptake capacity can be eval uated as a function of temperature ( e.g. subto supercritical range) whereby the impact of c hoosing alternative ther modyanamic states of the pore and bulk fluid states can be identified. 2.2.3.1. Adsorption Characterization using H2 and CH4 as Adsorptives The Clausius-Clapeyron expression is qu ite effective and r outinely used to calculate Q st in MOFs and other porous materials. Values derived from this expression however using just two experimental temperat ures are associated w ith a high degree of uncertainty. To provide the most accurate result this method requires at least three adsorption isotherms collected in a closely spaced temperature range.56-60 High resolution low pressure volumetric hydrogen adsorption isotherms were therefore collected on 1 at 87 K, 97 K, and 107 K (Figure 2.7.). All three is otherms are indeed fully reversible with the highest capacity observed at 87 K. The heats of adsorption for H2, derived from these isotherms show a slight deviat ion in the region of low loadin g but are very close to the expected value of 6.5 kJ mol-1. The consistency is maintained across this temperature regime, even at high loadings and theref ore validates our previous measurements.

PAGE 147

110 Figure 2.7. (a) Low pressure volumetric hydrogen adsorption isotherms for 1 collected at 87, 97, and 107 K and (b) Corresponding isosteric heats of adsorption for H2. The overall shape of an adsorption isotherm is influenced by many factors, one of which experimental conditions, such as th e temperature and pressure at which the measurement is performed. This is because thes e parameters effect th e observed states of the pore and bulk fluid phases of the adsorpti ve. For example, a classical Type I argon isotherm collected at 87 K [Tc(Ar) = 151 K] will fill the pore s of a framework with a liquid-like state at lower pre ssures, which is indicative of the observed plateau following saturation. A hydrogen isotherm collected at 87 K [Tc(H2) = 33.3 K] however will not fill the pores in a liquid-like state at this te mperature until higher pressures because you are above the critical temperature of H2. An excellent example to illustrate the dependence of thermodynamic states of pore and bulk fluid pha ses on the overall shape of an isotherm is shown in Figure 2.8. for methane and hydrogen ad sorption isotherms collected at 107 K. (a) (b)

PAGE 148

111 Figure 2.8. Adsorption isotherms for 1 collected at 107 K using methane and hydrogen as adsorptives illustrated in the linear display and (left) semi-logarithmic plot to highlight the adsorption behavior at low pressures (right). The critical temperature of methane is 190 K and therefore at this experimental temperature it is considered to be subcritical, while hydrogen is in a supercritical state because it is above its critical temperature. Th is explains why methane fills the pores in a liquid-like state as evidenced by the observe d plateau (typical of a Type I behavior), while the H2 adsorption isotherm exhibits a linear fit as a function of pressure. The overall shape of the methane adsorption isothe rm is analogous to that observed for a Type I nitrogen or argon adsorption isotherm coll ected at 77 K and 87 K, respectively and therefore it is possible to ex trapolate surface area and pore vol ume information. Note that methane is a non polar molecule and ther efore has no dipole or quadrupole moment. Accordingly, methane should in theory be a reliable probe molecule for textural characterization because it will not give rise to specific interactions with functional groups on the surface of the adsorbent ( i.e. similar to Ar). The pore volume calculated from the methane adsorption at 0.95 atm (0.45 cm3 g-1) is close to that calculated from the argon isotherm at 87 K (0.47 cm3 g-1). This result confirms that a methane adsorption

PAGE 149

112 isotherm collected at 107 K can also be employed to gain meaningful information pertaining to pore volume and porosity. 2.2.3.2. Polar and Non-polar Adsorptives: CO2 versus CH4 Adsorption Studies The CO2 and CH4 sorption capabilities of 1 were assessed in the temperature range of 273 K to 323 K and pressures up to 3 MPa (30 atm). This was achieved by using two instruments, namely a high resolution vol umetric apparatus in combination with a gravimetric instrument equipped with a ma gnetic suspension balance. The data points collected at relative pressures ranging from 10-5 to 10-2 were measured on the volumetric instrument, while data collection at highe r pressures was obtained using the magnetic suspension balance. Since the is otherms were collected at high pressures it is necessary to distinguish between surface excess adsorption (n) and absolute adsorption (nA) which are related by the following expression: Nex = Nads – bulkVads. The aforementioned values can differ substantially at higher temperatures and pressures, which is in contrast to sorption at lower temperatures and pressures. The quantity of gas that is experiment ally measured using standard physical adsorption measurements ( i.e. volumetric or gravimetric te chniques) provides the amount of surface excess gas adsorbed (not the absolute value). For simplicity, at very low temperatures ( i.e. the boiling point of the adsorptive) the amount of gas adsorbed can be rationalized as a two phase model. This mean s that the gas adsorbed at the surface does not experience the same type of environment or interactions as the gas located further away from the surface. An adsorbed phase is considered to coexist with the bulk phase at lower pressures because the gas molecules will initially accumulate (adsorb) on the

PAGE 150

113 surface as a dense monolayer. At higher pres sures however multilaye r coverage occurs and a liquid-like density builds up on the surf ace owing to strong interactions with the surface of the adsorbent as P0 is approached -a gas and liquidlike phase thereby coexist.37, 56, 61 At very low temperatures the density of the gas is insignificant compared to the density of the bulk phase near the su rface. The surface excess is therefore roughly equivalent to the adsorbed amount and there is no need to differen tiate betwee n these two values. At higher pressures however this anal ogy is false because there is a significant difference in the density be tween the bulk gas phase and the adsorbed phase at the surface. The absolute adsorbed amount of CO2 was calculated for 1 from the surface excess using two different approaches. The fi rst method is more applicable to systems with narrow pores, as in the case of soc -MOF. It relies on the a ssumption that the volume adsorbed in the system is fixed. The volume ad sorbed therefore is correlated to the pore volume as determined from argon or nitrog en adsorption isotherms at 77 K and 87 K, respectively. The equation used to calculate the absolute ad sorbed amount is based on the following equality below, where g refers to the molar density of the bulk gas at the experimental temperature and pressure, Vp is the corresponding pore volume, and n represents the experimentally dete rmined surface excess adsorbed amount. nA = n + gVp If one assumes however that the volume of the adsorbed phase is not fixed and varies as a function of the density of the adsorbed phase then the following expression is revealed whereby the density of the adsorbed and gas phase is considered.37 Nex = Nads – bulkVads Nex = Nads/[1-( bulk/bulk)].

PAGE 151

114 The absolute adsorbed amount of CO2 for 1 was determined by employing both methods using a pore volume of 0.50 cm3/g (N2 isotherm) and assuming the adsorbed CO2 was in a liquid-like phase owing to the f act that the experiment al conditions (273 K) were conducted below the cr itical temperature of CO2 (Tc = 304.1 K). Both methods in fact are in good agreement. The pore volum e obtained from the absolute adsorption isotherm at 273 K was determined to be 0.44 cm3 g-1, which agrees with the aforementioned pore volumes obtained using nitrogen (0.50 cm3 g-1), argon (0.47 cm3 g-1), and methane (0.45 cm3 g-1). The work described herein therefore demonstrates that a variety of adsorptive s can be employed as probe molecules to extrapolate meaning porosity information if the experiment is conducted at the appropriate temperature. However, since the values are in such good agreement across all adsorptives it only stands to r eason that Ar adsorption at 87 K is an adequate probe to be employed over the others. This is due to c onvenience, reliability, and because it does not have a quadrupole moment and thus is not expected to give ri se to specific interactions with most surfaces and ions (contrary to N2 and CO2). The uptake capacity of 1 for CO2 and CH4, plotted as surface excess and absolute amounts adsorbed, is shown in Figure 2.9. Th ese plots clearly demonstrate how the excess and absolute adsorbed values deviate as a function of pressure We predicted that 1 would exhibit stronger adso rbate-adsorbent interactions for the adsorption of CO2 and thus have a higher uptake capacity as compared to CH4. The reason being that CO2 has a quadruple moment and 1 exhibits narrow pores, higher localized charge density, and vacant In(III) centers. This should lead to stronger binding sites and enhance the H2-MOF interactions. Methane on th e other hand is a non-polar molecule with no dipole or

PAGE 152

115 quadruple moment. Indeed, 1 does adsorb more CO2 than CH4 at temperatures ranging from 273 K to 323 K up to pressures of 30 atm. The steep rise observed at low loadings in the CO2 adsorption isotherm, as compared to CH4, indicates that CO2 has a strong affinity for the surface of the MOF. Figure 2.9. (a) Surface excess and absolute amounts of CO2 adsorbed for 1 at 273, 298, and 323 K; and (b) Surface excess and absolute amounts of CH4 adsorbed for 1 over the same temperature range. The isosteric heats of adsorption calculated from the CO2 and CH4 adsorption isotherms at three closely spaced temperatures confirms that CO2 has a stronger binding affinity for the MOF adsorption sites in 1 as compared to CH4. For CO2 it is 28.5 kJ mol-1 versus 18.8 kJ mol-1 in the case of CH4 (Figure 2.10.). In contrast to the Q st obtained from the hydrogen adsorption isotherm at lo w pressures, these values do not remain constant at higher loadings but instead slightly increase. At low loading the interaction energy is primarily influenced by adsorbate-ad sorbent interactions but at higher pressures adsorbate-adsorbate interactions becomes dominant and is facilitated by the narrow pores. ( a ) (b)

PAGE 153

116 Figure 2.10. (a) A comparison between the absolute CO2 and CH4 adsorption isotherms for 1 as a function of pressure to illustrate the differences in uptake capacity; and (b) Isosteric heat of adsorption for CO2 and CH4. The maximum uptake capacity observed for CO2 and CH4 in 1 is higher than those values reported for zeolites but still lower than those reported for other MOFs.62-64 The current record holder fo r the maximum amount of CO2 stored in a MOF is MOF-177 which adsorbs 33.5 mmol g-1 and 40.0 mmol g-1 of CO2 at 42 bar and 50 bar, respectively.62 While in the case of methane storage, Zhou and co-workers reported the synthesis of a compound that surpasses the DoE targets and stores 230v(STP)/v at 290 K and 35 bar.65 The ability of 1 to efficiently separate CO2 from CH4 was investigated as a function of temperature (273 K to 323 K) and pr essure (0 to 3 MPa), as shown in Figure 2.11. Since the critical temperature of CO2 is 304.1 K, it is highly plausible that it fills the pores in a liquid-like state. Altern atively, the critical temperature of CH4 is 190 K therefore it will remain in the supercritical state across the temperature range and is not (a) (b)

PAGE 154

117 expected to fill the pores in this fashion. The loading ratio of CO2/CH4 was predicted to be the highest at 273 K and the lowest at 323 K and equivalent pressures whereby more CO2 should be adsorbed when it is in a liquid-li ke state. Indeed, the loading ratio is the highest at 273K in the lo w pressure region (below 0.3 MPa) and decreases with increasing pressure. This finding implies that in the pressure range of <0.3 MPa and 273 K compound 1 is more efficient at CO2/CH4 separation. Above this pressure there is a crossing of the loading curves whereby the loading ratio at 323 K is the highest. The reason for this stems from the fact that at high pressures the supercritical phase CO2 is more compressible than for the liquid-like state whereas the thermodynamic state of CH4 remains constant throughout the entire temperature range. Figure 2.11. Evaluation of 1 to separate CO2 and CH4 as a function of temperature and pressure: (a) Loading ratio of CO2/CH4 up to pressures of 3 MPa (30 atm); and (b) Zoomed in plot up to 0.3 atm to emphasize the behavior of the loading ratio at low loadings. In summary, the data presented herein suggests that efficient CO2/CH4 separation at high pressures is more favorable at temp eratures above the bulk critical temperature while at low pressures (< 3 MPa) the temp erature should be maintained well below the (a) (b)

PAGE 155

118 critical temperature of CO2. This observation further va lidates our hypothesis that different thermodynamic states of adsorbed phases can drastically affect the amount adsorbed at the same pressure. In conclusi on, the high resolution physical adsorption data discussed in this section provides a systema tic evaluation of the temperature dependency of H2, CH4, and CO2 adsorption over a wide range of temperatures and pressures in a robust MOF, 1 The results are both informative a nd encouraging but more importantly emphasize the importance of identifying differe nt thermodynamic states of pore and bulk fluid phases because this can substantiall y affect the adsorption behavior, uptake capacity, and surface energetics of the material. 2.2.4. Isostructural Series of soc-MOF Anal ogs: Prototypical Platform to Conduct Systematic H2-MOF Interaction Studies The structural analysis of 1 coupled with the high resolution physical adsorption measurements and INS data reveals that high uptakes of hydrogen along with favorable H2-MOF interactions are attainable when the ap propriate structural and chemical features are incorporated into the MOF. Thus, th e importance of potential open-metal binding sites and narrow pores (< 1nm) with higher lo calized charged density was highlighted in the context of gas storage a nd separation applications. In the ongoing quest towards the development of an optimal hydrogen storage material, we sought to gain a better understanding of the preferential binding of hydrogen in porous MOFs by conducting a systematic adsorption study using soc -MOF as a prototypal platform.

PAGE 156

119 2.2.4.1. Iron-Based TMBBs Herein, we report on the design and s ynthesis of two isostructural MOFs constructed from Fe-TMBB and H4-ABTC. The reaction of H4-ABTC with FeSO42H2O under solvothermal conditions in a mildly acidic solution containing DMF/H2O/chlorobenzene affords a orange cube -shape microcrystalline material, formulated as {[Fe3(ABTC)1.5(H2O)3](NO3)(H2O)0.67}, ( 2 ) determined using SCD studies. The as-synthesized ma terial is not homogenous; it is accompanied by an unidentified brown powder that we speculate to be iron oxide. The impurity can be readily removed to afford a homogenous phase by repeatedly washing the as-synthesized sample with DMF, as confirmed by the similarities between the calculated and experimental PXRD patterns (Appendix B) SCD studies, as well as, XPRD patterns confirmed that compound 2 is isostructural to 1 The 3-periodic framework is built up from the same organic MBBs ( i.e. ABTC4-) as used in 1 but each octahedral metal center of the TMBB is replaced by a lighter Fe(III) metal cation, [FeO5(H2O)]. The trimers are linked together by six independent ABTC4ligands (dFe-O = 2.005 to 2.018 ), each of which are comprised of three Fe oc tahedra that share one central 3-oxo anion. The trimer is located on the three fold axis and yields Fe-(3-O)-Fe units at angles of 120o (dFe-O = 1.923 ). Iron has an open-shell electron configurat ion and therefore it can adopt different spin states, which means the metal ion can adopt different oxidation states. In the crystal structure of 2 each Fe atom is trivalent, the 3-oxo anion and deprotonated ABTC ligand contribute a negative two and four charge, re spectively. The crystallographic assignment of 3-O and Fe(III) was based on literature precedents. This yields an overall cationic

PAGE 157

120 framework (+1 per formula unit) whereby di sordered extra-framework nitrate, [NO3]-, anions are anchored in the corners of the carcerand-like cage in a tetrahedral arrangement and provide charge balance. The anions st atistically occupy two positions on the three fold axis with an equal probability of being located at either position (as observed in 1 ). The anions are unable to escape this cavity owing to electrostatics and steric hindrance (window dimensions: 5.139 x 7.639 point to poi nt). The relative scal e of the cage in 2 is comparable to that of 1 and thus has an effective di ameter of approximately 10.437 and the intersecting channels have an estimated diameter of 6.004 . During the course of these studies, H4-ABTC was reacted with an assortment of iron salts in an attempt to establish a series of Fe-based socMOF analogs. This platform would allow us to systematically evaluate the contribution of char ge balancing anions ( i.e. NO3 -, Cl-, Br-, OH-, etc ) on H2-MOF interactions by fine-t uning the localized charge density. A mixed valence Fe soc -MOF analog would also afford a neutral MOF and thus the overall contribution from the charge dens ity could be investigated. The following iron salts were employed in this study: FeSO46H2O, FeCl2 (anhydrous), FeCl24H2O, FeBr2, Fe(NO3)39H2O, as well as, ferrocene. A brief list of the extensive experimental parameters which were adjusted during the co urse of this project include the metal-toligand ratio, solvent polarit y, structure directing agen ts (SDAs), concentration, temperature, and pH. Indeed, reaction between H4-ABTC and FeCl2 in a DMF/H2O/chlorobenzene solution in the presence of warm diluted HC l yields orange homogenous microcrystalline material with cube-shape morphology. The pur ity of the as-synthesized material was confirmed by similarities between the e xperimental and calculated PXRD patterns

PAGE 158

121 (Appendix B). The as-synthesized mate rial, stable and insoluble in H2O and common organic solvents, was characterized by elem ental microanalysis and SCD studies and formulated as {[Fe3(ABTC)1.5(H2O)3](Cl)}, ( 3 ). In the crystal structure of 3 each Febased TMBB adopts the same coordination en vironment as that observed in compounds 1 and 2 [Fe3O(CO2)6(H2O)3]. Each Fe cation is trivalent and thus charge balance for the cationic framework is provi ded by chloride ions (Cl-), as oppose to nitrate ions. Note that in this case it is impossible for the charge to be balanced by nitrate ions because there is no nitrate source in the starting materials. Th e chloride assignment was also supported by the halide elemental microanalysis studies wh ich detected chloride ions in the sample. The disordered chloride ions are located in the cavities but statistically occupy three positions about the three fold axis with equa l probability. In other words the disordered nitrate ions in 2 are substituted for chloride ions in 3 (Figure 2.12.). As observed in the previous analogs, the chloride ions are una ble to escape owing to steric hindrance (window dimensions: 7.599 x 5.133 ; point to point and excluding van der Waal radii). The two types of narrow channels; that is, the hydrophobic and hydrophilic have similar diameters, which are estimated to be 6.024 .

PAGE 159

122 Figure 2.12. (a) Ball-and-stick representation of the nitrate ions encapsulated in the cages of 2 ; and (b) Ball-and-stick representation of the ch loride ions encapsulated in the cages of 3 The nitrate and chloride ions are highlighted in sp ace-filling representation. Color code: Fe = green; C = gray; N = blue; O = red; Cl = light green. Hydrogen atoms and solvent molecules are omitted for clarity. 2.2.4.2. Decorated TMBBs: Incorporation of Coordinated Halide Anions The solvothermal reaction of H4-ABTC and InCl3 or InBr3 in separate vials containing solutions of DMF/EtOH/H2O and DMF/CH3CN/H2O, respectively yields orange homogenous microcrystalline materials. In both cases, the morphology of the assynthesized samples is cube-shape. The as -synthesized compounds we re characterized by SCD studies and formulated as {[In3(ABTC)1.5(Cl)(H2O)2](H2O)5.35} ( 4 ) and {[In3(ABTC)1.5(Br)(H2O)2](H2O)6.25} ( 5 ), respectively. Indeed, 4 and 5 are isostructural compounds and crystallize in the cubic P-43 n space group with a = 22.4570(3) and a = 22.4530(4) , respectively. The location of the ch arge balancing counter ions is different in this set of soc -MOF analogs as compared to compounds 1 – 3 In the previous soc MOF structures the nitrate and chloride ani ons were encapsulated in the cages but in 4 and 5 the halide ions are not locate d in the cages. The anions are coordinated in the apical (a) (b)

PAGE 160

123 position on the In-TMBB (1 halide pe r trimer) to reveal a so-called decorated TMBB. Note that oxo-bridged indium trimers are scarce, with only three structures in the Cambridge Structural Database (CSD: February, 2010).20, 66 Two of which are reported by Eddaoudi and co-workers. Simila rly, the 3-periodic structure of 4 is built-up from indium TMBBs which contain three indium-cen tered octahedra that share one central 3oxo anion, In-(3-O)-In angles of 120o (dIn-O = 2.045(2) ). Crystall ographic analysis shows that one of the three apical pos itions is occupied by a chloride ion (Cl-), which is disordered over the three apical sites with an equal probability of be ing at either position (dIn-Cl = 2.460(4) ). The two remaining apical positions are saturated by aqua ligands (dIn-O = 2.164(2) ) to afford a [In3O(CO2)6(Cl)(H2O)2] TMBB. Note that the presence of a halide ion coordinatively bound in the axial pos ition is unprecedented for indium-based TMBBs. The crystal structure of 5 reveals a similar coordination environment, except the apical chloride ion is replaced by a bromide ion (Br-), which is also disordered over the three apical sites with an equal probability of being located at either of the positions (dInBr = 2.619 ). Thus, the assembly of eight [In3O(RCO2)6(Br)(H2O) 2] TMBBs and twelve ABTC4organic ligands delimit the anion-free cu boidal cage, as depicted in Figure 2.13. The relative size of the channels in 4 and 5 must be reported within a range owing to the fact the halide ions are deloca lized over the three apical site s. Accordingly, in the crystal structure of 4 and 5 the estimated diameter of the cha nnels can range in size from 3.333 to 5.952 and 2.867 to 5.966 , respectively depending on the location of the halide ions. The approximate dimensions of the window apertures of the cuboidal cages in 4 and

PAGE 161

124 5 measure 4.562 x 7.622 and 4.590 x 7.626 , respectively (point to point and excluding van der Waals radii). Figure 2.13. Select fragments from the crystal structure of 5 : (a) Ball-and-stick representation of the indium-based TMBB, [In3O(RCO2)6(Br)(H2O)2]; (b) Ball-and-stick representation of the anion-free cuboidal cage. Color code: In = green; C = gray; N = blue; O = red; Br = purple. Hydrogen atoms and solvent molecules are omitted for clarity. The yellow sphere represents the largest sphere that can fit inside the cage (~10.137 in diameter), considering van der Waals radii. 2.2.4.3. Low Pressure Gas Adsorption Measurements The total solvent-accessible free volume for compounds 2 – 5 was calculated to be 55.5 %, 58.9 %, 59.2 %, and 58.5 %. The set of isostructural soc -MOFs described herein, 1 – 5 made it possible for us to independently evaluate the effect of each of the following structural parameters on H2-MOF binding interactions: (1) Accessibility to potential open-metal binding sites (In versus Fe); (2) The impact of localized charge (a) (b)

PAGE 162

125 density by utilizing different extra-framework anions (NO3 versus Cl-); and (3) The effect of polarizi ng substituents (Clversus Br-). 2.2.4.3.1. Evaluating the Effect of Open Metal Binding Sites on H2-MOF Interactions In order to assess the sorpti on properties of the Fe-based soc -MOF with extraframework nitrates, 2 the guest molecules were exchanged with CH3CN. The orange microcrystalline sample was air dried and loaded into a 6-mm glass sample cell, where it was initially outgassed at room temperature for 12 h and then gradually heated to 135oC for 8 h. TGA spectra for the as-synthesized and exchanged samples indicate that 2 is indeed thermally stable above 135oC (Appendix C). We decided however to proceed with a full sorption study at 135oC in order to gain insights into the nature of the axial aqua ligands. Most notable the quest ion was if they were bound at this temperature or removed to yield open Fe binding sites. Since iron inte racts more strongly with water than indium (dFe-O = 2.060 ; dIn-O = 2.173 ), we anticipated that el evated temperatures would be required to ensure complete removal of the axial ligands. The argon sorption isotherm collected at 87 K is of Type I isotherm, which is characteristic of permanen tly porous microporous materi als (Figure 2.14a.). BET and Langmuir surface areas were estimated to be 1381 m2/g and 1678 m2/g and the total pore volume was found to be 0.563 cm3/g (P.V.theo = 0.570 cm3g-1). The internal pore diameter, 6.42 , was obtained from a cy lindrical NLDFT pore model assuming an oxidic (zeolite) surface (Figure 2.14b.). This value agrees reasonably well with the accessible pore diameter of the cylindrical-like channels in 2 which have an estimated diameter of 6.085 .

PAGE 163

126 Figure 2.14. (a) Argon sorption isotherm at 87 K for 2 {Fe(NO3)3 soc -MOF}, after evacuation at 135oC; and (b) Pore size distribution in 2 calculated from the Ar isotherm (~6.42 ). The hydrogen sorption capacity for 2 measured at 77 K and 87 K and atmospheric pressures, was determined to be as high as 2.17 wt % of H2 (Figure 2.15a.) and the Q st was determined to reach a maximum of 6.50 kJ mol-1 at low loading. This value decreased slightly to 5.80 kJ mol-1 at higher loading ( e.g. 1.5 % per weight). As illustrated in Figure 2.15b., Q st at low loading is similar for 1 and 2 The lower than expected Q st (at low loading) and the gradual decline observed in 2 may indicate that some of the coordinated water molecules are still bound to the axial positions of the FeTMBB. The cationic cluster would be less po larizable and contribute less towards the ionic character of the MOF because the iron si tes are not open and thus less accessible to H2. (a) (b)

PAGE 164

127 Figure 2.15. Sorption studies for 2 {Fe(NO3)3 soc -MOF}, after evacuation at 135oC: (a) Hydrogen sorption isotherm at 77 K and 87 K; an d (b) Isosteric heat of adsorption comparison between 1 (red) and 2 (green). Heating the exchanged sample, 2 to elevated temperatures (175oC / 8h) did indeed result in a significant change in the appearance of the crystalline sample; that is, 2 changed color from orange to dark brown when heated from 135oC to 175oC. This observation coupled with the TGA da ta leads us to believe that 2 is desolvated and thus the water molecules should be removed from the axial positions in the Fe-TMBB. Argon sorption studies on 2 at 175oC at 87 K revealed BET and La ngmuir surface areas were estimated to be 1701 m2/g and 2061 m2/g, respectively. The corresponding pore volume and pore size were determined to be 0.687 cm3g-1 and 6.42 , respectively. As illustrated in Figure 2.16., compound 2 certainly adsorbs more argon after evacuation at 175oC as compared to 135oC and a significant increase in th e pore volume is observed. This data coupled with the observable color change upon heating helps to support the argument that higher temperatures are needed to rem ove the coordinated water molecules in 2 (a) (b)

PAGE 165

128 Figure 2.16. A comparison of the Argon sorption isotherms at 87K for 2, {Fe(NO3)3 soc -MOF}, after evacuation at 135oC (blue) and 175oC (green). An evaluation of the H2 sorption properties at 175oC revealed that 2 can store higher amounts of H2, up to 2.62 wt % at 77 K and atmos pheric pressure and approaches saturation (Figure 2.17a.). This is indicative of the near li quid density of the sorbed hydrogen in the pores (0.04 g cm-3). Interestingly, the isoste ric heat of adsorption was calculated to be 7.84 kJ mol-1 at low loading, which is noticeably higher as compared to the 135oC data for both 1 and 2 (Figure 2.17b.). In contrast to 1 the isosteric heat of adsorption is not constant at higher loadings and gradually decreases to an estimated value of 5.82 kJ mol-1, which is comparable to the 135oC Q st data calculated for 2 This result implies that the binding sites in 2 are not homogenous (averaged) up to the experimental loading capacity. This is not overly surprising because it is known that vacant Fe binding sites facilitate stronger H2-MOF interactions than open In binding sites and thus the binding sites ar e not expected to be hom ogenous. Note that at higher loadings the heat of adsorption is not all that different between 1 and 2 and therefore if multiple hydrogen sorption isotherms were collected on 2 at three or more closely related temperatures ( i.e. 77, 87, and 97 K) it may reveal little difference.

PAGE 166

129 Figure 2.17. (a) Hydrogen sorption isotherm at 77 K and 87 K for 2 {Fe(NO3)3 soc -MOF}, after evacuation at 175oC; and (b) Isosteric heat of adsorption for 2 after evacuation at 135oC and 175oC compared to the Qst for 1, {In(NO3)3 soc -MOF}, after evacuation at 135oC. The enhanced H2-MOF interactions at lower loadi ngs may therefore be attributed to: (1) the electric field created by the posit ively charged and highly polarizable open Fe coordination sites on the TMBBs; and (2) A ccessibility to the windows of the cage, which is geometrically proximal to all of the dominantly charged components. 2.2.4.3.2. Contributions from Different Extra-framework Anions: Clversus NO3 The importance of open coordination meta l sites and the choice of metal cation employed can have a significant effect on the H2 sorption properties of a material. The former was exemplified in the previous section by comparing the Q st of 2 after evacuation at 135oC and 175oC, while the significance of th e latter was demonstrated by reporting higher Q st values for the iron versus the indium analog at low loading. Unfortunately, devising pathways to satisfy the DoE technical targets is not as simple as just fine-tuning the ch oice of metal center. In this context, compounds 2 and 3 are ideally (a) ( b )

PAGE 167

130 suited to evaluate the overall effect regardi ng the type of extra-fr amework counter ions on H2-MOF interactions in the charged soc -MOFs. Both compounds are constructed from the same Fe-based TMBBs and ABTC4organic MBBs and therefore contain the same number of potential open Fe binding sites. The two MOFs differ however with respect to the type of tetrahedrally positio ned counter ions in nanometer-scale cavity; that is, nitrate anions in 2 and chloride anions in 3 In order to assess the sorption properties 3 the guest molecules were exchanged with CH3CN for 3 days. The orange microcrystalline sample was air dried and loaded into a 6-mm glass sample cell, where it was initially outgassed at room temperature for 28 h. Subsequently, the hydrogen sorption propertie s were investigated so that a similar comparison could be made with respect to th e accessibility of the potential open Fe sites ( i.e. RT evacuation vs. elevated temperatures). The estimated BET and Langmuir surface areas were determined to be 1006 m2/g and 1117 m2/g, respectively (N2 isotherm) with a corresponding pore volume of 0.418 cm3 g-1 (P.V.theo = 0.619 cm3 g-1). It was evidenced that 3 stores up to 1.75 wt % of H2 at 77 K and atmospheric pressures. The Q st was calculated to be in th e range of 6.50 kJ mol-1 and 5.80 kJ mol-1, which is identical to 2 after heating to 135oC (Figure 2.18.). The lower than expect ed values can be attributed to improper sample activation due to the presen ce of axial water ligands on the Fe TMBBs, as previously demonstrated in the case of 2

PAGE 168

131 Figure 2.18. Sorption studies for 3 {FeCl3 soc -MOF}, after evacuation at room temperature: (a) Hydrogen sorption isotherm at 77 K and 87 K; and (b) Isosteric heat of adsorption for H2. These results raise an important questi on – Do the counter ions provide any contribution towards enhancing the polarizab ilty of the ionic framework below the optimal evacuation temperature? Computational studies provide valuable insights to help answer this question. Indeed, Space and co-wor kers have demonstrated that a high dipole population is localized inside the carcerandlike capsule near the window aperture. The authors ascribe this high dipol e population to the relative position of the window with respect to the dominantly charged components (ABTC4-, In-TMBB, and nitrate ions). With respect to H2-MOF interactions, a hydrated TMBB would not only reduce the contributions from the metal sites but restri ct accessibility to the windows of the cage. Thus, we speculate the overall contribution fr om the counter ions would have a reduced effect if the TMBB is hydrated. In order to fully assess the porosity of 3 the sample was heated to 165oC for 8h. During the evacuation process the microcrystalline sample changed color from orange to dark brown, similar to that observed in the case of 2 The apparent BET and Langmuir (a) (b)

PAGE 169

132 surface areas were calculated from the Ar so rption isotherm at 87 K and estimated to be 1648 m2/g and 1969 m2/g, respectively. The corresponding pore volume and pore size were estimated to be 0.65 cm3 g-1 and 6.42 , respectively (Figure 2.19.). Figure 2.19. Sorption studies for 3 ,{FeCl3 soc -MOF}, after evacuation at 165oC: (a) Argon sorption isotherm at 87 K; and (b) Pore size distribution calculated from the Ar isotherm (~6.42 ) by applying a cylindrical NLDFT por e model assuming an oxidic/zeolite surface. Hydrogen sorption studies revealed 3 adsorbs up to 2.53 wt% at 77 K and atmospheric pressures, while the associated is osteric heat of adsorption was estimated to reach a maximum of q = 7.10 kJ mol-1 at lower loadings (Figure 2.20.). The heat of adsorption for 3 after evacuation at 165oC shows increased H2-MOF interactions at low loading as compared to the Q st after room temperature evacuation. This observation suggests that the axial ligands on the Fe-T MBB were removed. The total number of potential open Fe sites is the same in 2 and 3 and thus a direct comparison can be made with respect to the extra-fram ework counter ions. Also, the re lative size of the channels and cages is nearly identical in both stru ctures and the density of the desolvated frameworks is similar having calculated values of 0.975 g cm-3 and 0.952 g cm-3 for 2 and (a) (b)

PAGE 170

133 3 respectively. The chloride ion would be expect ed to be more polarizable than a nitrate ion on account of the fact that the negative charge is loca lized on one atom as oppose to being distributed over four atom s. We therefore expected the presence of chloride ions in 3 would lead to a higher dipole popula tion and thereby promote stronger H2-MOF interactions. However, this is not the cas e and the heat of ad sorption was higher in 2 at low loadings. Figure 2.20. (a) Hydrogen sorption isotherm at 77 K and 87 K for 3 ,{FeCl3 soc -MOF}, after evacuation at 175oC; and (b) Isosteric heat of adsorption for 3 after evacuation at 165oC compared to the Qst for 3 after evacuation at 175oC. These findings can be rationalized by noting the relative size of the counter ions. A nitrate ion is by far larger than a chloride ion so that nitrates occupy more space in the cage than the chloride ions. The nitrate ions should therefore be more accessible to H2 via the windows than the chloride i ons in the desolvated MOFs. (a) (b)

PAGE 171

134 2.2.4.3.3. Polarizability Effects via D ecorated Indium-based TMBBs The two MOFs described herein are constructed from decorated In-TMBBs whereby one of the axial positions of each TMBB in 4 and 5 contain a disordered chloride ion and bromide ion, respectively. The aforementioned halides have the same formal charge of negative one but differ in size (RCl= 181 pm ; RBr= 196 pm) and thus polarizability (Br> Cl-). The desolvated forms of 4 and 5 will possess one less potential metal binding site as compared to compounds 1 – 3 and for that reason the cavity is anion-free. They therefore represent an idea l pair of compounds to determine if minor structural and chemical modifications in por e structure, i.e. size and localized charge density, influence the H2-MOF preferential binding sites. The solvent activation protocol carried out on 4 and 5 was similar to that used for the previous compounds whereby the as-synt hesized samples were exchanged in CH3CN for 4 to 7 days. The activated microcryst alline samples were outgassed at room temperature for 8 h and then gradually heated to 135oC and 145oC for 8 h in the case of 4 and 5 respectively. The apparent BET and Langmui r surface areas of the evacuated form of 4 were estimated to be 1244 m2/g and 1494 m2/g, respectively (Figure 2.21a.). The specific pore volume of 0.50 cm3 g-1 agrees reasonably well w ith the theoretical pore volume that is estimated to be 0.53 cm3 g-1. The internal pore diameter, calculated from the Argon adsorption isotherm at 87 K, shows that the majority of the accessible pores exhibit a diameter of 6.12 (Figure 2.21b.), which is comparable to that determined for 1 At first glance is may seem surprising that 1 and 4 have similar experimental pore sizes owing to the fact that the hydrophilic pores of 4 are decorated with chloride ions. However, it is important to consider that each TMBB only contains one chloride per

PAGE 172

135 trimer and therefore on average the pores are not e xpected to be substantially different in size. Figure 2.21. Sorption studies for 4 {InCl3 soc -MOF}, after evacuation at 135oC: (a) Argon sorption isotherm at 87K; (b) Pore size distribution obtained from a cylindrical NLDFT pore model assuming an oxidic/zeolite surface (~6.12 and 7.15 ). In the case of 4 the pore size distribution does rev eal an additional peak, albeit very small, at approximately 7.2 . The relative peak height indicates that the majority of the pores do not adhere to this diameter. There is uncertainty surrounding the origin of this peak because it is neither in agreem ent with the dimensions of the channels calculated from the crystal structure nor the di mensions of the cage (if Ar could enter). Alternatively, it would be more understanda ble if the peak was located below 6.1 owing to the reduced pore dimensions imposed by the disordered chloride ions. The H2 sorption uptake of 4 was explored and it was found to adsorb 2.04 wt % at 77 K and atmospheric pressures (Figure 2.22a.), which is significantly lower than the capacities reported for compounds 1 – 3 The isosteric heat of adsorption for 4 was calculated to be in the range of 6.79 kJ mol-1 to 5.93 kJ mol-1 at low and high loadings, respectively (a) ( b )

PAGE 173

136 (Figure 2.22b.). To further validate this resu lt, the sample was subsequently heated to 200oC to ensure complete removal of the axial water molecules (2 per trimer). Indeed, the sample adsorbed similar amounts of H2 (2.03 wt %) at 200oC and the isosteric heat of adsorption was in good agreement with the data at 135oC having a range of 6.53 kJ mol-1 to 6.04 kJ mol-1 (Figure 2.22b.). Figure 2.22. (a) Hydrogen sorption isotherm at 77 K and 87 K for 4 {InCl3 soc -MOF}, after evacuation at 135oC; and (b) Isosteric heat of adsorption of H2 for 4 after evacuation at 135oC and 200oC. The heat of adsorption is comparable to 2 and 3 at higher loadings; meaning it is not completely flat as observed in the case of 1 A plausible explanation to account for this result could be associated with the ani on-free cage. The bound ch loride ions that are delocalized on the cationic cluste r may lead to a lower dipolar population as compared to the extra-framework anions in a confined space, 1 – 3 Strictly speaking it is difficult to make a direct comparison between the Q st of 1 and 4 because more than one parameter has been modified: (1) compound 4 has less potential open metal sites than 1 (2 versus 3); and (2) the carcerand cage has is anion-free in 4 (a) (b)

PAGE 174

137 Gas adsorption studies conducted on 5 using argon and hydrogen as adsorbates revealed similar behavior as that reported for 4 The desolvated form of 5 was found to exhibit apparent BET and La ngmuir surface areas of 1237 m2/g and 1486 m2/g, respectively. The experimentally determined pore volume of 0.4986 cm3 g-1 agrees very well with the theoretical value that is estimated to be 0.50 cm3 g-1. The pore size distribution was comparable to 4 whereby the internal diameter of the majority of the pores measure 6.12 while a small per centage exhibits a width of 7.16 . Figure 2.23. Sorption studies for 5 ,{InBr3 soc -MOF}, after evacuation at 145oC: (a) Argon sorption isotherm at 87 K; (b) Pore size dist ribution obtained from a cylindrical NLDFT pore model assuming an oxidic/zeolite surface (~6.12 and 7.16 ). The framework densities for the desolvated forms of 4 and 5 are very similar having values of 1.09 g cm-3 and 1.14 g cm-3, respectively. Therefore, it was not overly surprising that 5 was found to adsorb 1.95 wt % of H2 at 77 K and atmospheric pressures (Figure 2.24). The isosteri c heat of adsorption for 5 was determined from the 77 K and 87 K H2 adsorption isotherms after evacuation at 145oC and 200oC. In both cases a maximum of q = 7.00 kJ mol-1 and 6.72 kJ mol-1 was recorded at low loadings for the (a) (b)

PAGE 175

138 aforementioned temperatures, respectively. These results demons trate that having a bromide versus a chloride ion coordinated in the apical position of the TMBB has little if any effect on the H2-MOF interactions at low loadi ng, as illustrated in Figure 2.24c. Figure 2.24. (a) Hydrogen sorption isotherm at 77 K and 87 K for 5 ,{InBr3 soc -MOF}, after evacuation at 145oC; (b) Isosteric heat of adsorption for 5 after evacuation at 145oC and 200oC; and (c) Isosteric heat of adsorption of H2 for 5 compared with 4 {InCl3 soc -MOF}. The fact that the Q st curves for 4 and 5 is similar to compounds 1 – 3 is interesting because the former contain one less poten tial open metal site per TMBB and the carcerand cage contains no ani ons. We may conclude that the added polarizability and (a) (b) (c)

PAGE 176

139 modest pore size reduction from the halide ions is large enough to compensate for the loss of one open In-TMBB binding site and loca lized charge density associated with the cuboidal cage of 1 2.2.5. Validity of the Isosteric Heats of Adsorption via Computational Studies Space and co-workers assisted us in th is study by fitting the hydrogen sorption isotherms collected at 77 and 87 K, i.e. for compounds 2 – 5 to a function in order to determine if the experimentally determined Q st was appreciably highe r at low loadings as we observed. It is important to r ecall that all data points below P/P0 equal to 10-3 were deleted for the Q st curves shown above because at such low pressures there is a high degree of uncertainty between the two temper atures. Accordingly, in trying to fit the experimental data to their parameterized model Space et al. also observed significant discrepancies at very low cove rage, i.e. it was not a linear fit, and therefore many points had to be deleted. The resultant curves s hown in Figure 2.25. do not reveal a significant difference between the analogs. It does howev er show the same trend as we observed experimentally; that is, Q st decreases according to the following trend Fe(NO3)3 ( 2 ) > FeCl3( 3 ) > InBr3( 4 ) > InCl3( 3 ).

PAGE 177

140 Figure 2.25. Isosteric heats of adsorp tion curves for compound 2 – 5 : (left) unzoomed and (right) zoomed computed by fitting the experimental hyd rogen sorption data at 77 and 87 K to a function. 2.2.6. INS Studies: Insights into the H2-MOF Preferential Binding Sites in 2 and 4 INS spectra were collected for compounds 2 and 4 on the direct geometry IN5 spectrometer at the Institut Laue-Langevin in Grenoble, France. For these specific experiments an incident wavelength of 2 was used in order to access a significantly large energy transfer range in neutron energy loss. In doing so however this resulted in a rather modest energy resolution for the spect ra. Prior to conducting the measurements, both samples were activated (CH3CN) on site and evacuated using the same protocol as noted in the sorption experime nts above. The INS spectra for 2 and 4 collected after the first in situ loading of H2, is shown in Figure 2.26. for 1 H2 / trimer, [Fe3O(RCO2)6] ( 2 ) and [In3O(RCO2)Cl] ( 4 ). This loading corresponds to 1/3 H2 per Fe and 1/2 H2 per In, respectively.

PAGE 178

141 Figure 2.26. INS spectra (T = 4.3 K) corresponding to 1 H2 per trimer of 2 {Fe(NO3)3 soc -MOF}, (red) and 4 {InCl3 soc -MOF} (green) in neutron energy loss over the range: (a) From 1 – 17 meV; and (b) enlarged range below 12 meV. An appreciable difference is observed between the spectra at this loading, which is precisely the range whereby a noticeable difference is obser ved in the experimentally determined isosteric heat of adsorption. That is, the Q st at low loading is appreciably greater for 2 than the indium analogs ( 1 4 and 5 ). The low-energy region of the (a) (b)

PAGE 179

142 spectrum provides useful insights to understand the underlying interactio ns at this loading because those peaks located at lower energy in dicate a higher barrier to rotation and thus correspond to stronger binding sites (lower fre quency transitions). A careful inspection of the low-energy region of the spectrum for 2 reveals two well-defined peaks at 7.7 meV and 10 meV. Compound 4 however does not have clearly vi sible peaks in this region. As eluted to previously, the lower rotational tr ansition frequencies are associated with high barrier (rotational tunneling) a nd therefore a stronger intera ction is observed in the case of H2 with Fe than compound 4 The second set of loadings corresponding to 3 and 2/3 H2 per formula unit in the case of 2 and 4 respectively (Figure 2.27.) correspond to filling of the metal binding sites. The pair of peaks observed in 2 (7.7 meV and 10 meV) show a commensurate increase in intensity, while a weak, and ra ther broad band appears in the spectrum of 4 at approximately 8 meV. The aforementioned peaks observed in 2 show a parallel increase in intensity that is comparable to the H2 adsorbed at open –metal binding sites in the Ni-, Co-, and Mg-CPO-27 analogs of MOF-74.67

PAGE 180

143 Figure 2.27. INS spectra (T = 4.3 K) of 3 H2 per trimer in 2 {Fe(NO3)3 soc -MOF} (black) and 3/2 H2 per trimer in 4 {InCl3 soc -MOF} (yellow) in neutron en ergy loss over the range from 1 to 12 meV. Accordingly, these peaks can be assigned as two related rotational transitions68 for H2 at the Fe binding site. The broad peak observed in the spectrum of 4 can therefore be tentatively assigned to H2 binding sities near the chloride ion located in the apical position of the indium-based TMBB. This assign ment is in accord with a very similar band observed in ZIF-68 whereby the organi c ligand was functionalized by replacing a hydrogen atom for a chloro substituent.69 Accordingly, the INS da ta presented herein explicitly demonstrate that it is the Febinding sites in 2 which are responsible for the enhancement of Q st at low loadings relative to th at of the indium analogs. At higher loading, the INS spectra for 2 and 4 are similar and reveal strong albeit broad bands at subsequent loadings in the re gion above 10 meV, which may be resolved into three peaks (Figure 2.28.).

PAGE 181

144 Figure 2.28. INS spectra (T = 4.3 K) to emphasize the differences at higher loading in neutron energy loss over the range from 1 to 12 meV: (a) Compound 2 {Fe(NO3)3 soc -MOF}, shown after loadings of 2, 6, 18, and 30 mmol whic h are represented in red, green, blue, and yellow respectively; (b) Compound 4 {InCl3 soc -MOF} at loadings corresponding 1.3, 2.7, 12, and 20 mmol of H2 represented in red, blue, gree n, and yellow, respectively. It is important to take into consider ation that the mode st energy resolution available from this type of instrument when used with a short incident wavelength, 2 , will not resolve fine structure in these bands as was possible on the QENS instrument at (a) (b)

PAGE 182

145 IPNS used in previous measurements. In pr evious studies, we assigned peaks in this frequency region to weaker binding sites between the TMBB and the organic components of the framework; that is, carboxylate, azo, a nd phenyl moieties. At low loading the INS spectra appear to be fairly similar in this region for both compounds, i.e. one may discern three peaks at 13.3, 14.4 and 15.6 meV for compound 2 and 12.6, 14.5 and 15.6 meV for 4 The two highest frequency transitions (highe r energy region of th e spectra) are likely to be related, and thus arise from similar sites. Note that at higher loadings in 2 the lower frequency transition (13.3 meV) shifts to even lower frequency (12.1 meV) when the loading is increased to the equivalent of 2 H2 per Fe. A similar, yet less pronounced, effect was observed in the parent indium-based soc -MOF compound, 1 A plausible explanation to account for th is observation can be expl ained by noting that in both 1 and 2 the carerand-like cage encapsula tes four nitrate anions, whic h are statically occupy two positions on the three fold axis with an equa l probability of being located in either position (see section 2.2.4.1.), whereas this cage is anion-free in 4 It is thereby apparent, that hydrogen can be forced into the partially nitrate filled cage at high loadings, whereas this cage simply fills gradually (peak at 12.6 meV, shifts slightly to 12.0 meV at high loadings) in 4 2.3. Experimental Section 2.3.1. Materials and Methods Unless otherwise noted, all MOMs discusse d in the following chapters were synthesized and characterized by Amy J. Cair ns in Prof. Mohamed Eddaoudi’s research group in the Department of Chemistry at the University of South Fl orida (USF) according

PAGE 183

146 to the outlined experimental methods a nd procedures. Please note that compound 1 was first synthesized by Professor Yunling Liu wh en he was a post doctoral fellow in Prof. Mohamed Eddaoudi’s research group at USF. My contribution focused on preparing 1 for gas sorption measurements, collecting sorpti on data, and preparing bulk quantities of 1 for INS experiments. 2.3.2. Instrumentation and Software Single-crystal X-ray diffraction (SCD) data were collected on a Bruker AXS SMART-APEX CCD diffractometer using MoK radiation ( = 0.71073 ) or CuK radiation ( = 1.54178 ) operated at 2000 W power ( 50 kV, 40 mA). The frames were integrated using the SAINT software package70 with a narrow frame algorithm. The structures were solved using direct met hods and refined by full-matrix least-squares on| F |2. The SHELXTL 5.1 program package71 was used for all crystallographic calculations. The crystallographic data reported in the following chapters were primarily and solved by Dr. Lukasz Wo jtas. Assistance was also graciously Dr. Gregory J. McManus, Dr. Derek Beau champ, and Mr. Mohamed Alkordi in the Department of Chemistry at the University of South Florida and Dr. Victor Kravtsov at the Institute of Applied Physics of Academy of Science of Moldova. The crystallographic data for compound 2 was collected at the Small Mo lecule Crystallography Beamline (11.3.1) at the Advanced Light Source in Berk ley, California by Dr. Paul Forster from the University of Las Vegas in Nevada. Complete lists of crystallographi c tables are included in Appendix A.

PAGE 184

147 X-ray Powder Diffraction (XRPD) meas urements were carried out at room temperature on a Bruker AXS D8 Advance 50kV instrument using a 40mA for CuK ( = 1.5418 ), with a scan speed of 1/min and a step size of 0.02 in 2 Calculated XRPD patterns were produced using the PowderCell 2.4 software72 and/or Materials Studio MS Modeling version 4.0.73 See Appendix B for a comparison between calculated and experimental XRPD patterns. Conventional thermogravimetric anal yses (TGA) were performed under N2 and recorded on a Perkin Elmer Precisely STA6000 thermogravimetric analyzer. TGA profiles are included in Appendix C. Volumetric gas sorption studies performe d at the University of South Florida (USF) were conducted on a fully automate d micropore gas analyzer Autosorb-1 MP (Quantachrome Instruments) at relativ e pressures up to 1 atm. The cryogenic temperatures were controlled using liquid nitr ogen and liquid argon at temperatures of 77 K and 87 K, respectively. Gravimetric gas so rption studies were performed by Mr. Ryan Luebke in the Department of Chemistry at the University of Sout h Florida on a VTI MB300 GHP (gravimetric high pressu re analyzer) gas adsorption in strument up to a pressure of 19,000 Torr at temperatures ranging from 273 to 298 K. High resolution physical adsorption experiments were performed on 1 using argon, hydrogen, methane and carbon dioxide as adsorptives. The low pressure m easurements were collected on a volumetric adsorption analyzer equipped with a novel cryos tat (Quantachrome Instr. / Oxford Instr.) from 0 to 0.1 MPa over a wide range of temp eratures (77 – 323 K). While, high pressure gravimetric adsorption measurements were measured on 1 from 273 K to 323 K on a Rubotherm GmbH instrument equipped w ith a magnetic suspension balance up to

PAGE 185

148 pressures of 50 MPa. The high resolution sorp tion data was collected and analyzed by our collaborators at Quantachrome Instrument s (M. Thommes and E.B. Celer) in Boynton Beach, FL and our collaborators at the Institu t fr Nichklassische Chemie (J. Moellmer and R. Staudt) in Leipzig, Germany. INS spectra of hydrogen adsorbed on 1 were collected by Drs. Juergen Eckert and Jarrod Eubank on the quasielast ic neutron spectrometer (QENS) spectrometer at the Intense Pulsed Neutron Source (IPNS) at Ar gonne National Laboratory (ANL). High resolution INS spectra of hydrogen adsorbed on compounds 2 and 4 were obtained on the cold-neutron time-of-flight IN5 spectromete r at the high-flux reactor located at the Institut Laue-Langevin (ILL) in Grenoble, France. The experiments were carried out by Dr. Juergen Eckert, Peter Geor giev, and Amy Cairns. All spec tra were analyzed using the LAMP software.74 The observed binding sites in th e INS spectra were tentatively assigned using a two-dimensional phenomenol ogical model previously described by Eckert and co-workers. That is, for simplicity, the model assumes H2 to be a hindered rotor that is subjected to a barrier to rotation with 2 angul ar degrees of freedom in a simple double-minimum potential. The transiti ons for the hindered rotor occur between different energy levels ( i.e. 0, 1, 2, etc ) with increasing energy a nd are therefore labeled as (0 – 1), (0 – 2), etc transitions. Accordingly, in the ab sence of a barrier to rotation the H2 molecule is permitted to rotate freely and the lowest transition for H2 occurs at 14.7 meV (or 119 cm-1 = 2B) and is observed for the para and ortho -H2 Alternatively, when H2 is introduced to a host material ( i.e MOF) that exhibits various structural and functional features, the H2 molecules are subjected to a ba rrier that hinders the rotation. This partially lifts the degeneracy of the J = 1 level and the lowe st transition frequency

PAGE 186

149 for the hindered rotor between J = 0 and J =1, regarded as the 0 – 1 transition, decreases exponentially as the barrier hei ght is increased. Thus, a 0-1 tr ansition at lower energy is the direct result of a higher ba rrier to rotation and is indicativ e of a stronger interaction at that particular binding site. It is imperative to point out th at these energy values are not equivalent to the rotation quantum number J because it is not applicable to the hindered rotor, only the free rotor. Atomic Absorption (AA) experiments we re carried out on a Varian Spectra AA 100 instrument, with the help of Mr. Farid N ouar in the Department of Chemistry at the University of South Florida. Total solvent-accessible volumes were determined using PLATON49 software by summing voxels more than 1.2 away from the framework. Materials Studio MS Modeling version 4.0 was used for the graphical structural analysis. A topological evaluation for each compound was performed using Topos75 software. Subsequently, the topological term s were compared to those present in the literature and the RCSR database.42 For known topologies, the three-letter symbols adopted by Prof. Michael O’ Keeffe are used (e.g. pts represents the topology for the Platinum Sulfide net). Tiling representati ons were evaluated using 3dt software.76 2.3.3. Synthesis and Characterization All chemicals and solvents used in the preparation of compounds 1 5 were of reagent grade and used w ithout further purification.

PAGE 187

150 Figure 2.29. Synthetic strategy followed for the preparation of 3,3',5,5'azobenzenetetracarboxylic acid (H4-ABTC). Reagents and Conditions: (i) H2O / NaOH / stir for 30 min; (ii) slow addition of hot glucose ( aq ) and bubble air through the solution for 24 h; (iii) filter the precipitate; dissolve in a minimal amount of H2O; acidify to pH = 1 using 12 M HCl. Preparation of 3,3',5,5'-azoben zenetetracarboxylic acid, H4-ABTC: The product was synthesized according to a m odified procedure from the literature.77 In a typical reaction, (9.5g, 0.045mol) of 5-nitroisophthalic acid was gradually added to a round bottom flask (r.b.f) containing 120 mL of H2O and (25g, 0.63mol) of NaOH pellets. Concomitantly, (50g, 0.28mol) of D(+ )-Glucose was dissolved in 75mL of H2O, the dissolving process was aided by heati ng the solution. The hot aqueous glucose solution was added dropwise to the mixt ure over a 1 h period. During the addition process, many color changes are observed; that is, the mixtur e is initially yellow but as more glucose is added the solution changes color to orange, then to a deep red, and finally upon full addition the solution appears almost black with a hint of orange. To facilitate the azo coupling reaction, a constant flow of air is bubbled through the mixture for approximately 24 h. This causes the sodium salt of the ligand to crash out of solution in the form of a yellow precipitate. The precipi tate was dissolved in a minimal amount of H2O to yield a transparent deep orange soluti on whereby the final product is isolated by acidification to pH = 1 using 12M HCl. The orange product was filtered, washed thoroughly with ice cold water, and air-dried. Note: No further purification is required.

PAGE 188

151 Synthesis of {[In3(ABTC)1.5(H2O)3](NO3)(H2O)3}n, (1). A solution of In(NO3)3 .2H2O (22.0mg, 0.065mmol) and H4-ABTC (16.6mg, 0.044mmol) in 1 mL DMF, 1 mL CH3CN, 0.100 mL piperizi ne (0.4 M in DMF), and 0.300 mL warm HNO3 (3.5M in DMF) was prepared in a 20 mL sc intillation vial. The so lution was heated to 85oC for 16h, after which orange polyhedra-shaped crystals were obt ained. Crystals of 1 were harvested and air-dried. The as-synt hesized material is insoluble in H2O and common organic solvents. (Yie ld: 28.0 mg, 95.6 %). Elementa l analysis calcd (%) for 1 C24H21N4O22In3: C 27.15, H 1.99, N 5.28; found: C 27.32, H 2.51, N 5.18. Synthesis of {[Fe3(ABTC)1.5(H2O)3](NO3)(H2O)0.67}, (2). A solution of FeSO4 .6H2O (30.0mg, 0.11mmol) and H4-ABTC (12.9mg, 0.036mmol) 1 mL DMF, 1 mL H2O, 0.5 mL chlorobenzene, and 0.300 mL HNO3 (3.5M in DMF) was prepared in a 20 mL scintillation vial. Th e solution was heated at a constant rate of 1.5oC/min to 85oC and held for 12h, then cooled at a constant rate of 1.0oC/min to room temperature. The as-synthesized sample was purified through repeated washings with DMF solvent, revealing small cube-shaped crystals that are insoluble in wate r and common organic solvents. Crystals of 2 were harvested and air-drie d. (Yield: 10.2 mg, 33.0 %). Synthesis of {[Fe3(ABTC)1.5(H2O)3](Cl)}n, (3). A solution of FeCl2 (13.7mg, 0.1 mmol) and H4-ABTC (12.9mg, 0.034mmol), 1 mL DMF, 1 mL H2O, 0.5 mL chlorobenzene, and 0.450 mL of warm HCl ( 3.5 M in DMF) was prepared in a 20 mL scintillation vial. The solution was heated at a constant rate of 1.5oC/min to 85oC and held for 12h, then cooled at a constant rate of 1.0oC/min to room temperature resulting in orange cube-shaped crystals. The as-synthesi zed crystals are insoluble in water and

PAGE 189

152 common organic solvents. Crystals of 3 were harvested and air-dried. (Yield: 14.2 mg, 52.9 %). Synthesis of {[In3(ABTC)1.5(Cl)(H2O)2](H2O)5.35}n, (4). A solution of InCl3 (30.0mg, 0.14mmol) and H4-ABTC (24.3mg, 0.068mmol) in 1 mL DMF, 1 mL ethanol, and 0.5 mL H2O was prepared in a 20 mL scintilla tion vial. The solution was heated to 85oC for 12h, and pure orange cube-shaped cr ystals were obtained. Crystals of 4 were harvested and air-dried. The crystals were found to be insoluble in H2O and common organic solvents. (Y ield: 26.2 mg, 52.9 %). Synthesis of {[In3(ABTC)1.5(Br)(H2O)2](H2O)6.25}n, (5). A solution of InBr3 (49.6mg, 0.14mmol) and H4-ABTC (24.1mg, 0.067mmol) in 1 mL DMF, 1 mL CH3CN, and 0.5 mL H2O was prepared in a 20 mL scintilla tion vial. The solution was heated to 85oC for 12h and pure orange cube-shape cr ystals were obtained. Crystals of 5 were harvested and air-dried. (Y ield: 23.4 mg, 44.7 %) and found to be insoluble in H2O and common organic solvents. 2.4. Summary and Conclusions In summary, the rational design and synthesi s of an isostructural series of MOFs with soc topology, 1 – 5 permitted a methodical hydrogen sorption study to be conducted in which case some factors that govern H2 binding were isolat ed. To gain a better understanding of the sorption sites within select compounds ( 1 2 and 4 ), the sorption data was complimented by INS studies. Colle ctively, the data obtained from these measurements provide valuable in sights into the preferential H2-MOF interactions as a direct consequence of varying specific stru ctural and chemical components. This study

PAGE 190

153 independently evaluated the effect of th ree key parameters: (1) Metal cation (In versus Fe) in two MOFs whose structure and compos ition differ only with respect to the choice of metal center; (2) Cont ributions from extra-fram ework counter ions (NO3 versus Cl-) housed in a confined space in two Fe-based MOFs; a nd (3) Conbributions from polarizability via decorated In-TMBBs (Clversus Br-). Table 2.2. Selected sorption data for compounds 1 – 5 Cpd Evac.a (oC) ddesolv (g/cm3) BET/Langmuirb (m2/g) P.V.exp (cm3/g) P.V.theo d (cm3/g) P.Sexp () % H2 e (77 K) H2 (g/L) # H2 / Metalf Q st g (kJ/mol) 1 135 1.12 N.R.c; 1417 0.50 0.51 (P) 6.12 2.46 27.5 3.88 6.50 2 175 0.975 1701 / 2061 0.68 0.54 (P) 0.57 (M.S) 6.42 2.62 25.5 3.36 7.84 3 165 0.952 1648 / 1969 0.66 0.58 (P) 0.62 (M.S) 6.42 2.53 24.1 3.14 7.10 4 135 1.09 1244 / 1494 0.50 ~0.55 (P) ~0.53 (M.S) 6.12 2.04 22.2 3.13 6.66 5 145 1.14 1237 / 1486 0.50 ~0.51 (P) ~0.50 (M.S) 6.12 1.95 22.2 3.13 6.86 aEvacuation temperature of the exchanged samples; bApparent BET and Langmuir surface areas obtained from the Ar sorption isotherms at 87 K, except for 1 which was determined from the N2 isotherm at 77 K; cBET surface area was not reported for 1 ; dTheoretical pore volume calculat ed using PLATON (P) and Materi als Studio (M.S.) using argon probe; eHydrogen uptake (wt %) at 77 K and 724 Torr; fThe unit cell of each compound contains 24 metal cations; gIsosteric heat of adsorption for H2 at low loading. With regards to the choice of metal cation, it was demons trated that potentially open Fe cationic TMBBs facilitate stronger H2-MOF interactions at low loading as compared to the indium analogs. This was evidenced by the notable increase in the isosteric heats of adsorption; that is, 6.50 kJ mol-1 and 7.84 kJ mol-1 for 1 and 2 respectively. Since compound 2 is constructed from Fe metal cations (MWFe = 55.85 g

PAGE 191

154 mol-1; MWIn = 114.82 g mol-1) this affords a lighter framew ork density and thus a larger pore volume and surface area. These structural features coupled with the enhanced interaction of Fe with H2 led to a higher uptake of H2 at 77 K and 723 Torr as compared to 1 and is supported by INS data. The types of extra-framework anions were found to play a role as demonstrated in two Fe-based MOFs, 2 and 3 The materials have similar framework densities, pore sizes and volumes but differ with respect to the t ype of anion located in the cage. Compound 2 having four disordered nitrate i ons was found to exhibit a higher Q st at low loading as compared to 3 having disorder chloride ions. Indeed, compound 2 was found to have a higher Q st and adsorbs slightly more H2 than observed in the case of 3 The effect polarizability and to a less er extent pore size was assessed via the incorporation of decorated In-TMBBs. This st udy revealed an insi gnificant difference by having one bromide ion coordinated in the apic al position of each trimer versus a chloride ion. This conclusion, based on the sorption da ta, is further supported by the INS data collected on 4 which did not reveal well-defined p eaks at lower transitional frequencies (below 10 meV). Compound 5 exhibits a slightly higher framework density than 4 due to the increased molecular weight from th e bromide and thus it adsorbs less H2 at 77 K and atmospheric pressures. However, its larger atomic radius and higher polarizability afforded a slightly higher isosteric heat of adsorption as compared to the chloride analogue. In conclusion, we have shown that the soc -MOF platform is well suited to achieve high uptake capacities in which case the H2-MOF interactions are maintained even at higher loadings. Nevertheless, the chemical changes described herein afforded minor

PAGE 192

155 changes in Q st and therefore we are still far away from satisfying the DoE targets. We are currently exploring alternative avenue s to increase the uptake capacity and Q st by designing isoreticular soc -MOF analogs using expanded square-planar tetracarboxylate organic ligands. The relative scale imposed by such ligands would result in the formation of larger cages and would therefore allow for encapsulation studies to be conducted, i.e. metal complexes. This could be interesting for gas storage and/or sensing applications. 2.5. References (1) Eddaoudi, M.; Kim, J.; Vodak, D.; Sudik, A.; Wachter, J.; O'Keeffe, M.; Yaghi, O. M. Proc. Nat. Acad. Sci. U.S.A 2002 99 4900-4904. (2) Moulton, B.; Zaworotko, M. J. Chem. Rev. 2001 101 1629-1658. (3) Ferey, G. J. Solid State Chem 2000 152 37-48. (4) Kitagawa, S.; Kitaura, R.; Noro, S. Angew. Chem. Int. Ed. 2004 43 2334-2375. (5) Tranchemontagne, D. J.; Mendoza-Cort es, J. L.; O'Keeffe, M.; Yaghi, O. M. Chem. Soc. Rev. 2009 38 1257-1283. (6) Furukawa, H.; Kim, J.; Ockwig, N. W.; O'Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2008 130 11650-11661. (7) Eddaoudi, M.; Moler, D. B.; Li, H. L.; Chen, B. L.; Reineke, T. M.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res 2001 34 319-330. (8) Seidel, S. R.; Stang, P. J. Acc. Chem. Res. 2002 35 972-983. (9) Tranchemontagne, D. J. L.; Ni, Z.; O'Keeffe, M.; Yaghi, O. M. Angew. Chem. Int. Ed. 2008 47 5136-5147. (10) El-Kaderi, H. M.; Hunt, J. R.; MendozaCortes, J. L.; Cote, A. P.; Taylor, R. E.; O'Keeffe, M.; Yaghi, O. M. Science 2007 316 268-272. (11) Nouar, F.; Eubank, J. F.; Bousquet, T. ; Wojtas, L.; Zaworotko, M. J.; Eddaoudi, M. J. Am. Chem. Soc. 2008 130 1833-1835. (12) Ferey, G. Chem. Soc. Rev. 2008 37 191-214. (13) Phan, A.; Doonan, C. J.; Uribe-Romo, F. J.; Knobler, C. B.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2010 43 58-67. (14) Sumida, K.; Hill, M. R.; Horike, S.; Dailly, A.; Long, J. R. J. Am. Chem. Soc. 2009 131 15120-15121. (15) Perry, J. J.; Perman, J. A.; Zaworotko, M. J. Chem. Soc. Rev. 2009 38 1400-1417. (16) Ockwig, N. W.; Delgado-Friedr ichs, O.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2005 38 176-182. (17) Eddaoudi, M.; Kim, J.; Rosi, N.; Voda k, D.; Wachter, J.; O'Keeffe, M.; Yaghi, O. M. Science 2002 295 469-472. (18) Chui, S. S. Y.; Lo, S. M. F.; Charmant, J. P. H.; Orpen, A. G.; Williams, I. D. Science 1999 283 1148-1150.

PAGE 193

156 (19) Latroche, M.; Surble, S.; Serre, C.; Me llot-Draznieks, C.; Llewellyn, P. L.; Lee, J. H.; Chang, J. S.; Jhung, S. H.; Ferey, G. Angew. Chem. Int. Ed. 2006 45 8227-8231. (20) Liu, Y. L.; Eubank, J. F.; Cairns, A. J.; Eckert, J.; Kravtsov, V. C.; Luebke, R.; Eddaoudi, M. Angew. Chem. Int. Ed. 2007 46 3278-3283. (21) Sudik, A. C.; Cote, A. P.; Yaghi, O. M. Inorg. Chem. 2005 44 2998-3000. (22) Wang, Z. Q.; Kravtsov, V. C.; Zaworotko, M. J. Angew. Chem. Int. Ed. 2005 44 2877-2880. (23) Murray, L. J.; Dinca, M.; Long, J. R. Chem. Soc. Rev. 2009 38 1294-1314. (24) Morris, R. E.; Wheatley, P. S. Angew. Chem. Int. Ed. 2008 47 4966-4981. (25) Zhao, D.; Yuan, D. Q.; Zhou, H. C. Energy Environ. Sci. 2008 1 222-235. (26) Moellmer, J.; Celer, E. B.; Luebke, R.; Cairns, A. J.; Staudt, R.; Eddaoudi, M.; Thommes, M. Microporous Mesoporous Mater. 2010 129 345-353. (27) Rosi, N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim, J.; O'Keeffe, M.; Yaghi, O. M. Science 2003 300 1127-1129. (28) Rowsell, J. L. C.; Yaghi, O. M. J. Am. Chem. Soc. 2006 128 1304-1315. (29) Sagara, T.; Ortony, J.; Ganz, E. J. Chem. Phys. 2005 123 014701/1 014701/4. (30) Koh, K.; Wong-Foy, A. G.; Matzger, A. J. J. Am. Chem. Soc. 2009 131 41844185. (31) Furukawa, H.; Miller, M. A.; Yaghi, O. M. J. Mater. Chem. 2007 17 3197-3204. (32) Zhao, D.; Yuan, D.; Sun, D.; Zhou, H.-C. J. Am. Chem. Soc. 2009 131 9186-9188. (33) Yan, Y.; Telepeni, I.; Yang, S.; Lin, X. ; Kockelmann, W.; Dailly, A.; Blake, A. J.; Lewis, W.; Walker, G. S.; Allan, D. R.; Barnett, S. A.; Champness, N. R.; Schroder, M. J. Am. Chem. Soc. 2010 132 4092-4094. (34) Zhong, D.-C.; Lin, J.-B.; Lu, W.-G.; Jiang, L.; Lu, T.-B. Inorg. Chem. 2009 48 8656-8658. (35) Chen, B.; Zhao, X.; Putkham, A.; Hong, K.; Lobkovsky, E. B.; Hurtado, E. J.; Fletcher, A. J.; Thomas, K. M. J. Am. Chem. Soc. 2008 130 6411-6423. (36) Lochan, R. C.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2006 8 1357-1370. (37) Lowell, S.; Shields, J. E.; Thomas, M. A.; Thommes, M., Characterization of Porous Solids and Powders: Surface Area, Pore Size, and Density Kluwer Academic Publishers: Do rdecht / Boston / London, 2004 (38) Israelachvili, J. N., Intermolecular and Surface Forces with Applications to Colloidal and Biological Systems Academic Press: London, 1985 (39) Myers, A. L. Langmuir 2006 22 1688-1700. (40) DOE Office of Energy and Efficien cy and Renewable Energy. Hydrogen Storage Technical Plan, September 2009. http://www1.eere.energy .gov/hydrogenandfuelcells/ (41) Nouar, F.; Eckert, J.; Euba nk, J. F.; Forster, P.; Eddaoudi, M. J. Am. Chem. Soc. 2009 131 2864-2870. (42) O'Keeffe, M. Reticular Chemistry Structure Resource. http://rcsr.anu.edu.au/ (43) Belof, J. L.; Stern, A. C.; Eddaoudi, M.; Space, B. J. Am. Chem. Soc. 2007 129 15202-15210. (44) Sava, D. F.; Kravtsov, V. C.; Eckert J.; Eubank, J. F.; Nouar, F.; Eddaoudi, M. J. Am. Chem. Soc. 2009 131 10394-10396.

PAGE 194

157 (45) Lin, X.; Telepeni, I.; Blake, A. J.; Dail ly, A.; Brown, C. M.; Simmons, J. M.; Zoppi, M.; Walker, G. S.; Thomas, K. M.; Mays, T. J.; Hubberstey, P.; Champness, N. R.; Schroder, M. J. Am. Chem. Soc. 2009 131 2159-2171. (46) Wang, X. S.; Ma, S. Q.; Rauch, K.; Simmons, J. M.; Yuan, D. Q.; Wang, X. P.; Yildirim, T.; Cole, W. C.; Lopez, J. J.; de Meijere, A.; Zhou, H. C. Chem. Mater. 2008 20 3145-3152. (47) Chen, B. L.; Ockwig, N. W.; Millward, A. R.; Contreras, D. S.; Yaghi, O. M. Angew. Chem. Int. Ed. 2005 44 4745-4749. (48) Yang, S.; Lin, X.; Blake, A. J.; Thom as, K. M.; Hubberstey, P.; Champness, N. R.; Schroder, M. Chem. Commun. 2008 6108-6110. (49) Spek, A. L. Acta Crystallogr. 2009 D65 148-155. (50) Rowsell, J. L. C.; Eckert, J.; Yaghi, O. M. J. Am. Chem. Soc. 2005 127 1490414910. (51) Kubas, G. J. Chem. Rev. 2007 107 4152-4205. (52) Ma, S. Q.; Eckert, J.; Forster, P. M.; Yoon, J. W.; Hwang, Y. K.; Chang, J. S.; Collier, C. D.; Parise, J. B.; Zhou, H. C. J. Am. Chem. Soc. 2008 130 15896-15902. (53) Eckert, J.; Trouw, F. R.; Mojet, B.; Forster, P.; Lobo, R. J. Nanosci. Nanotechnol 2010 10 49-59. (54) Forster, P. M.; Eckert, J.; Heiken, B. D.; Parise, J. B.; Yoon, J. W.; Jhung, S. H.; Chang, J. S.; Cheetham, A. K. J. Am. Chem. Soc. 2006 128 16846-16850. (55) Georgiev, P. A.; Albinati, A.; Mo jet, B. L.; Ollivier, J.; Eckert, J. J. Am. Chem. Soc. 2007 129 8086-8087. (56) Keller, J. U.; Staudt, R., Gas Adsorption Equilibria, Experimental Methods and Adsorption Isotherms Springer: New York 2004 (57) Gregg, S. J.; Sing, K. S. W., Adsorption, Surface Area and Porosity Academic Press: London, 1982 (58) Rouquerol, F.; Rouquerol, J.; Sing, K., Adsorption by Powders and Porous Solids Academic Press: London, 1999 (59) Bhatia, S. K.; Myers, A. L. Langmuir 2006 22 1688-1700. (60) Vuong, T.; Monson, P. A. Langmuir 1996 12 5425-5432. (61) Findenegg, G. H.; Thommes, M. ; in: Fraissard, J.; Conner, W. C., Physical Adsorption: Experiment Theory and Application Kluwer Dordrecht, 1997 (62) Millward, A. R.; Yaghi, O. M. J. Am. Chem. Soc. 2005 127 17998-17999. (63) Phan, A.; Doonan, C. J.; Uribe-Romo, F. J.; Knobler, C. B.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2010 43 58-67. (64) Llewellyn, P. L.; Bourrelly, S.; Serre C.; Vimont, A.; Daturi, M.; Hamon, L.; De Weireld, G.; Chang, J. S.; Hong, D. Y.; Hwang, Y. K.; Jhung, S. H.; Ferey, G. Langmuir 2008 24 7245-7250. (65) Ma, S.; Sun, D.; Simmons, J. M.; Collier, C. D.; Yuan, D.; Zhou, H.-C. J. Am. Chem. Soc. 2007 130 1012-1016. (66) Volkringer, C.; Loiseau, T. Mater. Res. Bull. 2006 41 948-954. (67) Dietzel, P. D. C.; Georgiev, P. A.; Ec kert, J.; Blom, R.; Straessle, T.; Unruch, T. Chem. Commun. 2010 in press (68) Kong, L.; Roman-Perez, G.; Soler, J. M.; Langreth, D. C. Phys. Rev. Let. 2009 113 096103.

PAGE 195

158 (69) Eckert, J.; Bangerjee, R.; Georgiev, P. A.; Matonovic, I.; Orcajo, G.; Izaola, Z.; Russina, M.; Albinati, A.; Eddaoudi, M.; Yaghi, O. M. Phys. Chem. Chem. Phys. 2010 submitted (70) Bernstein, F. C.; Koetzle, T. F.; Williams, G. J. B.; Meyer, E. F. J.; Brice, M. D.; Rodgers, J. R.; Kennard, O.; Shimanouchi, T.; Tasumi, M. J. Mol. Biol. 1977 112 535-542. (71) Sheldrich, G. M. SHELXTL, v. 5.10; Bruker Analytical X-ray, Madison, WI, 1997 (72) Kraus, W.; Nolze, G. Powder Cell for Windows v2.4; Berlin, 2000. (73) CRYSTMET Database. http://www.tothcanada.com/about.htm (74) Richard, D.; Ferrand, M.; Kearley, G. J. J. Neutron Research 1996 4 33-39. (75) Blatov, V. A.; Shevchenko, A. P.; Serezhkin, V. N. TOPOS (v. 4.0 Professional) http://www.topos.ssu.samara.ru/versions.html (76) Delgado-Friedrichs, O. 3dt The Gavrog Project http://gavrog.sourceforge.net/ (77) Wang, S.; Wang, X.; Li, R.; Advincula, R. C. J. Org. Chem. 2004 69 9073-9084.

PAGE 196

159 Chapter 3. Structural Diversit y of 4-Connected Nodes: Serendipity versus Predictability in Crystal Chemistry 3.1. Synthesis of 2-Periodic Metal-Organic Materials 3.1.1. Introduction Metal-organic materials (MOMs) are wi dely recognized as a unique class of solid-state crystalline materials. They e xhibit superior fine-t unable structural and chemical features which make them ideally suited for many desired applications ( e.g. gas storage and separation, magnetism, drug delivery, etc ).1-8 The molecular building block (MBB) approach has proven to be remark ably successful at producing modular MOMs with a certain degree of predictability that range from di screte (0-periodic), 1-, 2-, to 3periodic, as described in Chapter 1.9-13 In this chapter we will address the significance and versatility of such MBBs for the construction of 2-periodic MOMs,14-23 with an emphasis on structures having kgm and sql topology. Layered networks are naturally more comp lex than 1-periodic structures because the latter are delineated by chain-like struct ures having at most 2-connected nodes and thus the geometry (angularity) around each node governs the formation of the net. The geometry around each node is equally important in the realm of 2-periodic networks but the existence of two spatial dir ections supports the formation of n -connected nodes where n is greater than 2. This modification facili tates the formation of geometrical shapes ( e.g. building units; BUs) through the self-assembly of the n -connected nodes. In the context

PAGE 197

160of MOMs, the overall shape and pa cking motif of the layers can lead to the formation of channels and/or cavities, which will have a profound effect on the properties of the material ( e.g. porosity, magnetism). Layered structures have a ri ch history with clay-like ma terials and graphite being two well-known examples.24, 25 Some of the earliest report s of metal-organic analogues were inclusion compounds base d on metal-cyanide compounds,26 more specifically Prussian blue (PB) compounds and Ho fmann clathrates (see Chapter 1).27-29 One of the fascinating features of this broad class of mate rials is their ability to trap guest molecules between neighboring layers through the expansio n and/or contraction of the layers. The intercalated guest molecules can be incorpor ated into the host lattice as a means to control the overall packing of the layers ( e.g. AAA, ABAB, ABC, etc ) and the relative distance between neighboring layers. Hence the chemical composition and functionality of the framework, as well as, the packing motif provides a useful stra tegy for controlling and optimizing the porosity of these materials. 3.1.1.1. Topological Perspective: Regular and Semi-Regular Plane Tilings Two-periodic metal-organic architectures are commonly re ferred to as layered or sheet structures. This analogy stems from the f act that the underlyi ng topology of the net is described by planar networks. In order to evaluate the topology of a framework, it must be reduced into its simplest com ponents; that is, the assembly of n -connected nodes connected through edges (spacers). In this context the plane can be enclosed (covered) by convex polygons to reveal a tiling representation that is unique for that particular net. In most cases the product of the self-assembl y process yields the simplest and most

PAGE 198

161symmetric nets.30 There exists three exclusive regular plane nets or Platonic tilings. Their transitivity is described by [ 111] which denotes one type of vertex-, edge-, and tile. As illustrated in Figure 3.1., these tilings incl ude: the (3,6)-hexagona l lattice which is assembled from 6-connected nodes to reveal tr iangular tiles, the (4,4) -square grid lattice composed of 4-connected nodes which assemble to afford square tiles, and the (6,3)honeycomb lattice comprised of 3-co nnected nodes and hexagonal tiles.31 Figure 3.1. Tiling representation of the three exclusive regular [111] plane nets: (left) (3,6) hexagonal lattice ( hxl ); (middle) (4,4) square grid lattice ( sql ); and (right) (6,3) honeycomb lattice ( hcb ).32 The regularity of a given net is reduced for systems composed of more than one type of convex polygon whose vertices inte rsect whereby an increased level of topological complexity is in troduced. Accordingly, a sub-cl ass of plane tilings which warrant special attention is the semi-regular (Archimedean) nets and there are exactly eight tilings in this group (F igure 3.2.). A diverse collection of 2-peroidic nets exist, a few of which have been highlighted here. In particular, the edge-transitive nets must be emphasized because this group of nets is salie nt to crystal chemistry. There are precisely five 2-periodic edge-transitive nets: the three regular nets, the quasi-regular kgm net, and the Kagom dual ( kgd ; isohedral tiling). The results de scribed below will highlight the

PAGE 199

162versatility and significance of the sql and kgm nets by reporting the design and synthesis of several decorated metal-organic arch itectures having the noted topologies. Figure 3.2. Tiling representation of the semiregular plane nets constructed from two or more regular polygons. Top (left to right): (4.82) fes or sql-a ; (3.122) hca ; (32.4.3.4) tts ; (33.42) cem Bottom (left to right): (3.4.6.4) htb ; (4.6.12) fxt ; (34.6) fsz ; (3.6.3.6) kgm .32 3.1.1.2. Design Principles and Applicatio ns: Nanoscale Kagom and Square Grid Lattices Metal-organic materials with kgm topology are attractive ta rgets because of their properties, which include potential magnetis m applications in resulting from spin frustration.22, 33-38 The Kagom lattice is a uninodal ne t (vertex transitive) comprised of 4connected square planar nodes and thus the se lf-assembly of suitable MBBs can lead to the formation of this layered network with ( 3.6.3.6) topology. This par ticular sequence of numbers indicates that the kgm net is comprised of two sets of triangular and hexagonal convex polygons, which meet at each 4-connected node.39 The triangular polygons are vertex-linked and thus delimit the hexagona l polygon. The tiling representations of the kgm and sql nets provide a blueprint for the ratio nal design of nanoscale metal-organic

PAGE 200

163analogues. Thus, the self-assembly of pre-fabr icated MBBs with the necessary geometric attributes have the potential to generate analogous nets under the appropriate reaction conditions. A clever design st rategy was introduced by Zawo rotko and co-workers to target such nets by utilizing nanoscale SBUs ( n SBUs).21, 22, 40-42 The authors demonstrated that square and triangular n SBUs can be formed via the self-assembly of vertex-linked regular molecu lar squares (Figure 3.3.). Figure 3.3. Schematic illustration of the two nanoscale SB Us that can be generated by linking molecular squares through m -BDC type ligands: (a) Square n SBU comprised of four square SBUs; and (b) Triangular n SBUs comprised of three square SBUs. The dimetal tetracarboxylate, [M2(RCO2)4L2], MBB is commonly employed to construct a diverse collection of MOMs, wh ich require a 4-connected square planar building unit.43-46 This analogy stems from the fact that connecting the points of extension of the paddlewheel MBB reveals a molecular square, when viewed down the 4fold axis (Figure 3.4.). In a similar vein, a linear spacer and octahedral building unit are attainable if only the axial positions represent the points of extension or if all six positions are extension points, respectively. This particular MBB can be readily synthesized and thus is a ubiqu itous metal cluster in solid-state chemistry as exemplified by the extensive number of crystal struct ures reported in the CSD (over 1600!). Transition metal ions predominately form this cluster, specifically Cu, Rh, Ru, and Mo. The inherent modularity and ri gidity afforded by the paddlewheel MBB renders it an (a) (b)

PAGE 201

164ideal building block in crystal design: (1) the metal cations can be readily substituted; (2) the carboxylato ligands can be varied; and (3) the axial positions can be easily altered. Figure 3.4. (left) Dimetal tetracarboxylate, [M2(RCO2)4L2]n, “paddlewheel” building block, which can be regarded as (right) a square bu ilding unit when viewed down the 4-fold axis. Zaworotko and co-workers in fact de monstrated that triangular and square n SBUs can be generated by linking these molecular squa res at their vertices using rigid ditopic organic ligands ( e.g. m -BDC). The natural angularity im posed by the relative position of the carboxylate groups dictates that such ligan ds are predisposed to link the squares at a 120o angle. The versatility and predictability of linking square paddlewheel MBBs with m -BDC and its derivativ es is exemplified by the large collection of discrete nanoballs (triangular and square nSBUs ),41, 47, 48 Kagom lattices (triangular n SBUs),22, 49 and 2periodic tetragonal (square n SBUs) lattices.21, 42 Hence, the aforementioned MOMs are excellent examples of supramolecular isomeris m whereby more than one type of network is generated from the same MBBs. In the context of this discussion, the Kago m lattice is assembled from alternating bowl-shaped triangular n SBUs. This arrangement of atom s leads to the formation of small triangular and large hexagonal cavit ies having van der Waals dimensions of approximately 7.804 and 15.751 , respectively (Figure 3.5a,c.). The amplitude between the layers and their overall packing can be influe nced by ligand functionality,

PAGE 202

165non-covalent interactions, templa te-effects, steric hindrance, etc A higher degree of diversity is possible in the tetr agonal square grid structure wi th respect to the orientation of the molecular squares in the n SBUs (Figure 3.5b,d.) whereby the square grid is constructed from both the cone and 1,3-altern ate forms. This leads to the formation of uniform rectangular-shaped channels in wh ich each bowl has an outer diameter of approximately 9.024 and a depth of 8.412 . Such topological differences have the potential to lead to rather different porosities and magnetic properties. Figure 3.5. Examples of Kagom and Square Grid MOMs constructed from m -BDC and [Cu2(RCO2)4L2] MBBs: (a-b) Space-filling representation of a single 2-periodic layer in a Kagom and tetragonal square grid stru cture, respectively, viewed down the z -axis; (c) Schematic illustration of the Kagom lattice which is comp rised of 6and 3-membered rings; and (d) Schematic illustration of the tetragonal square grid lattice comprised of solely 4-memebered rings. Color Code: C = gray; O = red; hydrogen = white; and Cu = green. The 5-position of the BDC moiety is highlighted in purple. (a) (c) (b) (d)

PAGE 203

1663.1.1.3. Taxonomy of Calixarene-like MOMs The resultant bowl-shape of the square n SBU bears a striking resemblance to the conformation adopted by calixarene compounds ( e.g. calix[4]arene). It was therefore shown that tetragonal square grid structur es, amongst others, can form analogous metalorganic atropisomers in the solid state.42 Calixarenes are macrocyclic compounds derived from the assembly of aromatic and alkyl components.50, 51 They are routinely synthesized via a condensation reaction by combining N equivalents of phenol, resorcinol, or pyrogallol with N equivalents of an aldehyde. Ty pically, phenol is reacted with formaldehyde while the substituted phenyl components require larger aldehydes ( e.g. acetaldehyde).52 The resultant bowl-shape cavities f acilitate the encapsulation of guest molecules and thus this class of compounds represent an ever growing field in supramolecular Host:Guest chemistry.53-57 Atropisomerism is a form of isomerism that can occur in systems when free rotation about the sp3 carbon-carbon bond is si gnificantly hindered.58 This can invoke the formation of different stereoisomers, i. e. compounds having the sample chemical composition but differs only in their arrangemen t in space. This form of isomerism is predominately observed in biaryls, porphyrins, and calixarene-based materials. The four possible atropisomerisms of a calix [4]arene identified by Gutsche54 (Figure 3.6.).59-63

PAGE 204

167 Figure 3.6. The four types of atropisomers of calix[4 ]arene: (a) Cone; (b) Partial cone; (c) 1,2alternate; and (d) 1,3-alternate. The basis for this nomenclature arises fr om the orientation of neighboring arene units. In the cone arrangement, all arene units are directed up ward, while having one arene directed downward affords the partial cone. The 1,2and 1,3-alternates correspond to cis -and alternating arenes directed downward, respectively. The organic MBBs used to construct the 2and 3-periodic structures pr esented in this chapter are all derived from 1,3-BDC ligands. Accordingly, this type of terminology will be utilized throughout this chapter to described networks that are inde pendently assembled from clusters of four square and tetragonal building units, as well as, triangular n SBUs. 3.1.2. Results and Discussion 3.1.2.1. Decorated Kagom Lattice Herein, we contribute to the collection of decorated metal-organic 2-periodic Kagom lattices by reporting the serendipitous discovery of compound 6 {[Cu2(5NH2BDC)2(H2O)2]}, with kgm topology. The goal of this project was in fact to synthesize an isoreticular series of porous 3-periodic MOFs having nbo topology, whose (a) (b) (c) (d)

PAGE 205

168cross-linked layers mimic the connectivity of the kgm lattice. A so-called ligand-toligand pillaring design strategy was employed to connect the adjacent la yers in an attempt to systematically control th e resultant pore size (along the z -direction) by using an assortment of m -BDC derived tetracarboxylate orga nic ligands. The linkers were functionalized in the m -position to precisely fuse the neighboring layers. A detailed discussion of this design strategy will be pr ovided in the following section. The rationale behind this methodology stems from the fact that planar networ ks exhibit modular interlayer distances and the packi ng arrangement of these layers ( e.g. AAA, ABAB, ABC, etc ) in the solid state is predominately controlled by non-covalent intermolecular interactions between guest molecules. Thus covalent cross-linking of neighboring layers through pre-designed tetracarboxylate orga nic ligands of various lengths and functionality is expected to facilitate th e formation of highly porous 3-periodic MOFs with tunable pore dimensions. The tetracarboxylate organic ligand empl oyed in this study, 5,5'-{benzene-1,4diylbis[( E )methylylidenenitrilo]}benzen e-1,3-dicarboxylic acid (H4-ImTC) is shown in Figure 3.7. Mild solvothermal reaction between H4-ImTC and Cu(NO3)2 .2.5H2O however yields a 2-periodic kgm net instead of the expected 3-periodic nbo structure. There are several possible explanations to account fo r this outcome: (1) the aqueous reaction conditions may have promoted the imine f unctionality to undergo a hydrolysis reaction which liberated 5-NH2BDC as a product, and/or (2) th e final product was contaminated with unreacted starting mate rial in the form of 5-NH2BDC.

PAGE 206

169 Figure 3.7. 5,5'-{benzene-1,4-diylbis[( E )methylylidenenitrilo]}benzene-1,3-dicarboxylic acid, H4-ImTC. We were unable to reproduce these crystals using H4-ImTC and therefore the next logical step was to focu s on the reaction of 5-NH2BDC with copper. Indeed, mild solvothermal reaction of 5-NH2BDC with Cu(NO3)2 .2.5H2O in a DMF/CH3CN solution in the presence of triethylamine (TEA) yields homogenous green hexagonal crystals, designed as compound 6 The crystal structure revealed a 2-periodic network similar to that of the parent Kagom structure.22 Accordingly, the network is built up from copper paddlewheel building blocks, whereby each c opper ion exhibits the expected square pyramidal geometry (dCu-Cu = 2.630 ). The equatorial plane is occupied by four independent carboxylato oxygen atoms that brid ge the dinuclear copper ions in a bismonodentate fashion (dCu-O = 1.950 , 1.971 ), while aqua ligands lie in the axial positions (dCu-O = 2.110 ) to yield the CuO5 primary building block. The self-assembly of the triangular n SBUs affords an infinite nanoscale Kagom lattice sustained by large hexagonal cavities and small triangular ca vities (Figure 3.8.). The estimated window dimensions of the aforementioned cavitie s are 15.751 (6.138 vdw sphere) and 5.174 , respectively. The approximate depth th e hexagonal and triangul ar cavities is 9.072 and 10.600 , respectively which roughly corr esponds to the interlayer separation.

PAGE 207

170 Figure 3.8. X-ray crystal structure of 6 : (a) Ball-and-stick representation of the Kagom lattice viewed down the z -axis; (b) Schematic illustration to show the assembly of triangular n SBUs; (c) Ball-and-stick representation of th e AAA layers in viewed down the x -axis; and (d) Schematic illustration of the arrangement of the undulating triangular n SBUs (highlighted in green). Color Code: C = gray; O = red; N = blue; hydrogen = white; and Cu = green. The natural curvature imparted by th e angularity in the functionalized m -BDC ligand promotes efficient packing of the undulating layers (AAA) with an effective interlayer separation of 10.600 . Each of the layers interact through N-H…O hydrogen bonding interactions (dis tance is 1.749 ); that is, aqua liqands are pos itioned between the layers which serve as a bridge. Thus, each aqua ligand participates in N-H…O hydrogen bonding with two N-H groups located on different phenyl ri ngs in neighboring layers (Figure 3.9.). (a) (b) (c) (d)

PAGE 208

171 Figure 3.9. N-H…O hydrogen bonding interactions in 6 : (a) Viewed down the z -axis; and (b) The y -axis. The independent layers are represented in purple and yellow to highlight each layer. 3.1.2.2. Synthesis of an Anionic Te tragonal Square Grid Lattice from Tetrahedral MBBs The natural angularity imparted by m -BDC is not a prerequisite for the synthesis of tetragonal square grid structures. These networks have also been constructed using linear ditopic ligands, such as 1,4-BDC19 and others others.42, 64 In a similar vein, the vertex figure of the inorganic MBB does not have to be a perfect square to facilitate the formation of a square grid lattice. Mild solvothermal reaction between 5-NH2BDC and In(NO3)6H2O in a DMF/H2O solution indeed yields a tetragonal square grid lattice, formulated by SCD studies as, {[In(5-AmBDC)2](DMA)(DMF)}n, ( 7 ). It must be noted that the reaction conditions employed in the synthesis of 7 promote a condensation reaction, which converts the amin e functionality into the am ide derivative. Amides are routinely synthesized in orga nic synthesis via the reaction of a carboxylic acid with an amine. The general reaction mechanism is outlined in Figure 3.10. Under appropriate reaction conditions, DMF frequently decompos es to liberate DMA cations and formic acid as bi-products. We therefore hypot hesize that 5-AmBDC was formed via a (a) (b)

PAGE 209

172condensation reaction between formic acid and the amine group on 5-NH2BDC following the decomposition of DMF. To the best of our knowledge, compound 7 represents the first example of a MOM constructed usi ng this ligand (CSD update: February 2010). Figure 3.10. A general reaction mechanism for the direct synthesis of amides via a condensation reaction between a carboxylic acid and an amine.65 Each indium metal ion in the crystal structure of 7 is surrounded by four independent 5-AmBDC ligands, which coordi nate to each In(III) cation through the carboxylato groups in a bidentate fashion. A 4-connected In(RCO2)4 MBB is thereby revealed having In…O distances in the range of 2. 141 to 2.562 (Figure 3.11a.). The O…In…O angles are in the range of 83.084 to 165.769o, which significantly deviate from an ideal tetrahedral arrangement. Thus, each In (III) metal ion is said to adopt a distorted tetrahedral configuration. Each independent angular 5-AmBDC ligand bridges two In(III) centers and thus the assembly of the 4-c onnected nodes results in the generation of 4membered windows to give an overall sql topology. A square n SBU consisting of four vertex-linked tetrahedral buildi ng units is thereby obtained.

PAGE 210

173 Figure 3.11. X-ray crystal structure of 7 : (a) In(RCO2)4 MBB which can be viewed as a 4connected node; (b) Ball-and-stick representation of the tetragonal sheets which pack in an ABAB fashion down the z -axis; and (c) The layers interact via offset face-to-face interactions through the phenyl rings of 5-AmBDC. Color Code: C = gray; O = red; N = blue; hydrogen = white; and In = green. The channels are re presented by the undulating yellow columns. The 1,3-alternate conformation is the only isomer observed in 7 and therefore the four dicarboxylate ligands orient in an alternating fashion. The anionic layers propagate in the xy -plane and stack in an ABAB fashion along the z -axis with an interlayer separation of approximately 6.660 . The rectan gular-shaped channels have an estimated diameter of 6.110 . Non-covalent intermolecu lar interactions govern the overall packing arrangement of the layers. These include offset face-to-face interactions and hydrogen bonding interactions. In the case of the former, the two opposing metal carboxylate groups in the inorganic MBB are facing downward while these groups in the subsequent layers are directed upward. An offset face-to-face interaction occurs (a) (b) (c)

PAGE 211

174between neighboring phenyl moieties and an N-H…O hydrogen bonding interaction (2.006 ) is observed between the amide f unctionality of one layer and a carbonyl oxygen atom of the next layer. Disordered DMF solvent molecules reside between the layers, as well as, DMA+ for charge balance. 3.1.3. Experimental Section 3.1.3.1. Materials and Methods All materials and methods are described in Chapter 2, unless otherwise noted. The H4ImTC ligand was prepared by Mohamed Alkordi. 3.1.3.2. Synthesis and Characterization Synthesis of {[Cu2(5-NH2BDC)2(H2O)2]}n, (6) A solution of Cu(NO3)2 .2.5H2O (23.3mg, 0.10mmol), 5-aminoisophthalic acid (5-NH2BDC) (36.2mg, 0.20mmol), DMF (0.5mL), CH3CN (1mL), triethylamine (0.2mL, 1M in DMF), and HNO3 (0.2mL, 3.5M in DMF) was prepared in a 20 mL scintill ation vial. The solution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours to afford homogenous green hexagonal crystals. Hom ogenous green crystals of 6 were harvested and air-dried. The as-synthesized material was determined to be insoluble in H2O and common organic solvents. Synthesis of {[In(5-AmBDC)2](DMA)(DMF)}n, (7) A solution of In(NO3)2 .6H2O (69.1mg, 0.23mmol), 5-NH2BDC (41.7mg, 0.23mmol), DMF (1mL), H2O (1mL) and HNO3 (0.3mL, 3.5M in DMF) was prepared in a 20 mL scintillation vial. The solution was heated to 85oC for 12 hours followed by additional heating to 105oC for

PAGE 212

17523 hours and 115oC for 23h hours, after which colorless block shape crystals were obtained. Crystals of 7 were harvested and air-dried. The as-synthesized material was determined to be insoluble in H2O and common organic solvents; however, the crystals do degrade upon exposure to aqueous media. 3.2. From 2-Periodic Layers to 3-Periodic Frameworks: Pillaring as a Design Strategy 3.2.1. Introduction The previous section highlighted the significance, complexity, and structural diversity of 2-perioidic arch itectures. This information pr ovides a solid foundation for the transition into and development of a speci fic class of 3-periodic MOMs regarded as pillared frameworks. It is important to note that layered stru ctures possess interesting properties and are useful for many applications as note above; however their inability to enclose space often limits their developmen t in a wide range of porosity related applications. A logical approach to circumvent this shortcoming is to merge the attractive features of 2and 3-peroidic frameworks into one class of material; that is, to bridge the layers through judiciously chosen organic MBBs ( e.g. pillars) to form fine-tunable cavities and/or channel systems that have th e potential to adsorb gases and/or organic molecules. The diverse topologies adopted by layered structures lead to fundamental changes in pore size and shape, as exemplified by supramolecular isomerism whereby more than one superstructure can be isolated from the same MBBs ( e.g. nanoball, kgm and sql structures).10 Moreover, each topological class repr esents a modular platform where the metal cation or organic component(s) can be re adily varied to give charged or neutral

PAGE 213

176isostructural or isoreticular analogs, respect ively. These types of platforms are urgently needed in order to systematically evaluate the effect of each parameter on gas-MOF interactions. Thousands of layered structures have been documented as a result of both serendipitous discoveries and rational design strategies ( e.g. MBB approach). We have therefore instituted a plan which specifically targets 3-periodic MOMs whose layers are based upon the sql and kgm topologies. The reason was twofold: (1) the fundamental differences in pore shape and size and (2) predic tability because these nets are proven to be readily constructed from a wide range of different chemical components. The pillaring approach provides a unique pathway to incorpor ate these desired features into 3-periodic frameworks. Gas sorption studies could then be conducted in a systematic fashion using a series of modular microporous MOM platform s to independently discern the impact of pore size/shape, chemical functionality, and charge. 3.2.1.1. Pillaring Strategy: Design Principles Layered structures can in principle be pillared into 3-pe riodic MOMs via the inorganic or organic building blocks by choos ing pre-designed spacer/pillar moieties that can fulfill geometrical ( e.g. size, angles, etc ) criteria and at the same time adhere to the coordination environment around the metal center. There are at least three ways to pillar layers, which are categorized based on the locati on of the pillar in the layer: (1) axial-toaxial; (2) axial-to-ligand; and (3) li gand-to-ligand (Figure 3.12.). The dimetal tetracarboxylate “paddlewheel” MBB is well suited for all three types of pillaring

PAGE 214

177because it commonly forms laye red structures when reacte d with ditopic carboxylatebased organic molecules. Figure 3.12. Schematic representation to illustrate the th ree possible design strategies to pillar 2D layers into 3D frameworks: the purple, blue, a nd yellow pillars independently represent the axialto-axial, axial-to-ligand, and ligandto-ligand strategies, respectively. Linear bipyridyl-based organic moieties are ideally suited to connect collimated layers via the axial-to-a xial positions on the paddl ewheel MBB (Figure 3.13.).66-74 This form of pillaring can take place in situ by combining all necessary chemical components in a one-pot reaction or by using a post-synth etic approach whereby the layers are first synthesized and subsequently pillared.

PAGE 215

178 Figure 3.13. Select examples of bipyridyl-based organi c ligands (left to right): pyrazine; 1,4diazabicyclo[2.2.2.]octane; 4,4'-bipyridine; 1,2di(pyridine-4-yl)ethane1,2-diol; and N,N'-di(4pyridyl)1,4,5,8-naphthalenetetracarboxyldiimide. Axial-to-ligand pillaring on the hand is regarded as a hybrid method because the eclipsed layers are covalently cross-linked by judiciously chosen bifunctional organic moieties. One type of functional group forms the necessary metal clus ter and thus bridges the clusters to form the infinite layer, such as a decorated isophtalate unit. The role of the other functionality ( e.g. pyridyl-type) is to pillar the layers via the 5-position of the isopthalate unit to the axial position of the metal cluster in the ne ighboring layer (Figure 3.14.). Eddaoudi and co-workers have demonstrat ed the versatility and predictability of this approach by pillaring sql and kgm lattices using 5-(4-pyrid inylmethoxy)-isophthalate (PMOI). To the best of our knowledge, this design strategy has not previously been observed in the literature.75

PAGE 216

179 Figure 3.14. Bifunctional organic MBB, 5-(4-pyridinylmethoxy)-isophthalate (PMOI2-) surrounded by three [Cu2(RCO2)4N2] MBBs. Color Code: C = gray; O = red; N = blue; and Cu = green. All hydrogen atoms are omitted for clarity. The pillared structures described he rein are of the ligand-to-ligand type.46, 76-81 The layers are therefore fused together in an eclipsed fashion through a functionalized tetracarboxylate organic bridge. There are several examples of highly porous MOFs constructed using this approach. The type of tetracarboxylate pill ar employed, whether it be rigid, flexible, or bulky in nature can l ead to structural di versity as a direct consequence of ligand design. 3.2.2. Results and Discussion The ligand-to-ligand pillaring method was us ed to design and synthesize a diverse collection of open 3-periodic architectu res, whose layers are based upon the kgm and sql lattices. The tetracarboxylate organic ligands ( e.g. the pillars) employed in these studies are shown in Figure 3.15.

PAGE 217

180 Figure 3.15. Tetracarboxylate organic ligands utilized fo r the ligand-to-ligand pillaring strategy: (left) H4-ABTC; (middle) H4-BIPATC; and (right) H4-BAYTC. The rational assembly of open 3-peri odic MOMs based upon the aforementioned lattices generates several platforms that po ssess various pore sizes and shapes, due to their connectivity and relative scale and f unctionality of the pillars. These platforms therefore (in theory) provide a systematic pa thway to attain high surface area materials whereby the effect of metal cation, pore size and shape, and ligand functionality can be evaluated with respect to th e gas adsorption properties ( e.g. uptake capacity and Q st). A major obstacle however which impeded these studies was improper sample activation and therefore obtaining the expected results pr oved to be very challenging, as will be discussed below. 3.2.2.1. Isoreticular MOFs with nbo To pology: Pillared Kagom Layers The solvothermal reaction of H4-ABTC and Cu(NO3)22.5H2O, Zn(NO3)26H2O, or MnCl2 in separate vials cont aining solutions of DMF/CH3CN/IMI, DEF/CH3CN, and DMF/CH3CN/HMTA, respectively generates crystal line material in high yield. The assynthesized compounds were characterized by SCD studies and determined to be

PAGE 218

181isostructural frameworks (Appendix A). The following structural description will only pertain to the copper analogue which was formulated as {[Cu2(ABTC)(H2O)2]}n, ( 8 ). The crystal structure of 8 is built-up from dicoppe rtetracarboxylate MBBs by bridging two Cu(II) cations with four deprot onated carboxylate moieties from the ABTC ligand to reveal a [Cu2(RCO2)4(H2O)2] paddlewheel configuration (dCu1-Cu1=2.675 ). Each Cu(II) cation adopts a square pyramidal coordination environment. The equatorial plane is defined four oxygen atoms from four independent ABTC4ligands, which coordinate through the carboxylato gr oups in a bis-monodentate fashion (dCu-O= 1.927, 1.927, 1.915, 1.915 ). The axial position is o ccupied by an aqua ligand (dCu-O= 2.155 ). Each ABTC4ligand is linked to four paddlewheel MBBs and vice versa ; therefore both the inorganic and organic MBBs can be redu ced to 4-connected nodes. Augmentation of the aforementioned nodes reveals square and rectangular-pla nar building units, respectively. A topological assessment wa s performed using the TOPOS software82 and the vertex symbol and coordinati on sequence were calculated to be [6(2).6(2).6(2).6(2).8(2).8(2)] and 4, 12, 28, 50, 76, 110, 148, 194, 244, and 302, respectively. This confirms that the 4connected 3-periodic framework possesses nbo topology. The topology of 8 can be delineated as a neut ral 3-periodic pillared MOM constructed from 2-periodic layers that are connected by covale ntly cross-linked ABTC4pillars. A space-filling representation of 8 viewed along the x -axis illustrates this concept (Figure 3.16.). The undulating layers propagate along the xy -plane and are pillared diagonally to neighboring layers through ABTC4via the 5-position of the diisophthalate units. A select fragment of the layer reveals it has the same connec tivity as the (3.6.3.6)

PAGE 219

182Kagom lattice. The pillared Kagom network, 8 is constructed from the self-assembly of alternating bowl-shap ed triangular cavities having a “cone” confirmation, [Cu2(ABTC)(H2O)2]3. Hexagonal channels having a 1,3,5-alternate conformation are thereby revealed (Figure 3.16c.). The relativ e diameters of the metalla[3]calix and metalla[6]calix are predetermined by the ge ometry of the paddlewheel MBB and the distance between the carboxylato moieties on the BDC unit and therefore have an estimated diameter of 3.387 and 7.606 , respectively. Figure 3.16. Space-filled views from the x-ray crystal structure of 8 : (a) Viewed along the x -axis to highlight the undulating 2-D layers (shown in black); (b) Select fragment of one Kagom layer which propagates in the xy -plane (the triangular and hexagonal cavities are highlighted with black circles); (c) Layers are pillared along the z -axis; (d) Two metal-organic calixarene-like conformations observed in 1 : cone and 1,3,5-alternate. Color Code: C = gray; O = red; N = blue; hydrogen = white; and Cu = green. All solvent mo lecules and axial aqua ligands are removed for clarity, as well as, hydrogen atoms in figures (a) and (c). (a) (b) (c) (d)

PAGE 220

183The centroids of the diisophthalate benzen e ring were chosen as the points of extension to reveal a square SBU. Triangular n SBUs are obtained by linking three square SBUs at their vertices. These nanosized triangl es then self-assemble at their vertices to create a 2D Kagom lattice with in finite undulating layers (Figure 3.17.). Figure 3.17. Schematic representation of the arrangement of the n SBUs in the 2D Kagom layer: (top) Triangle bowl-shaped n SBU are formed by the assembly of three vertex-linked molecular squares, which are formed by linking together the centroids of the BDC units; (bottom) the natural curvature of the BDC ligand leads to the formation of undulating layers. The Kagom layers can be covalently pill ared into 3-periodic frameworks in two ways: (1) pillaring a triangle to a triangle and a hexagon to a hexagon ( e.g. AAA face-toface packing); or (2) pillari ng a triangle to a hexagon ( e.g. ABAB offset face-to-face packing). The former arrangement requires that the ligand have some degree of flexibility because the BDC units would be directed out of plane. Ligand flexibility is not required in the latter case and therefore a planar ri gid ligand is well suited to facilitate this geometrical conformation. In the crystal structure of 8 the layers are connected through a triangle-to-hexagon conformation. Layer A is covalently cross-linke d to layer B through three ABTC4ligands in the following way: three BDC moieties of the cone triangular

PAGE 221

184building block in layer A are directed downw ard to three BDC moieties directed upward in the 1,3,5-alternate hexagon in layer B. La yer A and B are thus connected via the 5position of the BDC moiety through the azo moiety. The pillared layers generate two types of cages with stoichiometries of [Cu12(ABTC)12] and [Cu24(ABTC)6], denoted as cage I and II, respectively (Figure 3.18.). The dimensions of each cage are pro portional to the relative size ABTC4-. Cage I is sphere-like in shape and is assembled from six paddlewheel MBBs and twelve ligands, while cage II is elongated and comprises tw elve paddlewheel M BBs and six ligands. Figure 3.18. Two types of cages whose dimensions are predetermined by the relative size of the ABTC4ligand: (a) Spheroid-like cage, [Cu12(ABTC)12]; (b) Elongated cylindrical-like cage, [Cu24(ABTC)6]; (c) Schematic n SBU representation of the spheroid cavity; (d) The elongated cavity (highlighted in blue) results from the p acking arrangement of the spheroid-like cavity. Color Code: C = gray; O = red; N = blue; hydrogen = white; and Cu = green. (a) (b) (c) (d)

PAGE 222

185The cages are delimited by two types of in tersecting windows defined as A and B. The overall shape can be corre lated with the connections between adjacent paddlewheel MBBs. This thereby reveals an overall triangular arrangement The estimated size of the internal cavity in cage I is 9.522 and it has 8 windows which lead to its interior, two of which are TYPE A and 6 are TYPE B. Th e former window is defined by the cone arrangement (equilateral triangle) and thus the dimensions (~3.387 ) are not dependent on the length of ABTC4-. On the other hand, type B is defined by isosceles triangles whereby the two equivalent edges represent th e ligand and the third edge is occupied by one BDC moiety. Thus, the window dimensi on is predetermined by the length of ABTC4. The Cu…Cu separation between adjacent padd lewheel MBBs is 6.457 , while the estimated diameter of the window is 7.397 . The eight triangular windows described herein are shared between cag e B, which passes through three layers. The internal cavity of this cage is estimated to be 11.891 and the vertical distance of measures approximately 22.633 . Isoreticular MOMs can be generated in a similar manner using elongated pillars to yield non-interpenetrated netw orks with nanoscopic cages. It is noteworthy to mention that nbo is a dual net ( i.e. bcu ) and thus interpenetration is possible. The structure described above however, 8 does not favor interpenetration because the effective diameter of the regular triangular window is too small to allow the penetration of a pillar(s). It is possible however for an orga nic spacer to penetrate the isosceles windows but the ligand would have to be quite long and this could be avoided by functionalizing the pillar with bulky substituents.

PAGE 223

1863.2.2.2. Enroute to Larger Functionalized Cavities and Higher Surface Area: Pillared Kagom Lattices from Expanded Tetracarboxylate Ligands The organic pillars described in this section are two polyphe nyl tetracarboxylate ligands of varying length and functionality, i.e. H4-BAYTC and H4-BIPATC. The resultant MOMs therefore has the same number of potential unsaturated metal centers but the size and functionality of the pillar would directly a ffect the pore size and shape. A systematic gas sorption study could then, in theo ry, be conducted to eval uate the effect of pore structure and metal cation on the uptak e capacity and gas-MOF interactions. Solvothermal reaction between H4-BAYTC and Cu(NO3)22.5H2O in a mildly acidic solution consisting of DMF and chlorobenzene in the presence of tetrabutylammonium bromide (TBABr) yields blue-green microcrystalline material with parallelepiped morphology. The as-synthesi zed compound, which is insoluble in H2O and common organic solvents, was characterized and formulated by single-crystal X-ray diffraction studies as {[Cu2(BAYTC)(H2O)2]}n, ( 11 ). Compound 11 crystallizes in the rhombohedral R-3 m space group with a = 15.527(3) , b = 18.527(3) , c = 54.221(1) and a unit cell volume of 16117(5) 3. The crystal structure has a framework comprised of the same inorganic MBBs as compound 8 i.e. the paddlewheel MBB (dCu1-Cu1=2.659 ). A topological evalua tion confirmed that 11 exhibits the same nbo network topology as compounds 8-10 but the structures differ with resp ect to the pore sizes because of the length of the pillars. Similarly, the un dulating Kagom layer propagates in the xy -plane and its windows are unaffected by the le ngth of the organic ligand (Figure 3.19.).

PAGE 224

187 Figure 3.19. X-ray crystal structure of 11 : (left) Space-filling view down the y -axis to illustrate the covalent cross-linking between undulating layers; and (right) Space-filling view down the z axis. Color Code: C = gray; O = red; and Cu = green. Hydrogen atoms, axial ligands, and solvent molecules are omitted for clarity. The two cages generated from pillaring the layers with BAYTC4have the same metal-to-ligand ratio as observed in 8 [Cu12(BAYTC)12] and [Cu24(BAYTC)6] for cage I and II, respectively (Figure 3.20.). The relative span of BAYTC4(16.430 ) versus ABTC4(9.283 ) leads to the generati on of much larger cavities. Figure 3.20. Two types of cages in 11 : (left) Spheroid-like cage I, [Cu12(BAYTC)12]; and (right) Elongated cylindrical-like cage II, [Cu24(BAYTC)6]. Color Code: C = gray; O = red; and Cu = green.

PAGE 225

188Cage I in 11 can accommodate a van der Waals sphere with an estimated diameter of 12.974 . In both structures the size of the intersecting triangular windows (TYPE A) are roughly the same in both structures. Th e isosceles window (TYPE B), which can be viewed along the x or y -axis, has an estimated edge a nd length that span 6.697 and 13.030 , respectively. The increased length of th e pillar makes cage II to be much more elongated with a distance from the top laye r to the third layer of ~38.312 . It can however only accommodate a van der Waals sphe re with a diameter of ~9.833 because of the bottle-neck shape. The inherent modularity of the MBB appr oach also facilitates fine-tuning of the interior cavities and windows th rough the use of judiciously chosen pillars with desired built-in functionality. Thes e parameters can then be modified by employing functionalized organic ligands whose substitu ents range from bulky to small, electrondonating to electron-wit hdrawing, etc. The H4-BIPATC pillar is well suited for this purpose, i.e. to link the Kagom because it is comprised of a naphthalene moiety that is decorated with four oxyge n atoms. Reaction of H4-BIPATC with Cu(NO3)32.5H2O in a mildly acidic solution consisting of DMF a nd chlorobenzene indeed results in the formation of blue-green microcrystalline material having a polyhedral morphology. The as-synthesized material was characterized us ing single-crystal X-ra y diffraction studies and formulated as {[Cu2(BIPATC)(H2O)2]}, ( 12 ). The crystal structure of 12 features a 3periodic pillared Kagom network consisting of infinite Kagom layers {[Cu2(RBDC)2(L)2]3 }n that are covalently cr oss-linked through BIPATC4-. The coordination environment is identical to that observed in 8-11 but one paddlewheel MBBs has a terminal aqua ligand located in the axial positions (dCu-O =

PAGE 226

1892.165, 2.274 ), while the apical sites on the second MBB are occupied by a DMF molecule (dCu-O = 2.171 ). There are two crystall ographically independent BIPATC4ligands: one type coordinates to four aqua paddlewheel M BBs and the other coordinates to both an aqua and DMF paddlewheel MBB in a trans fashion. The 1,3-BDC moieties remain coplanar but steric hindrance be tween neighboring hydrogen atoms causes the naphthalene-based moiety to twist perpendicular to the plane having an estimated dihedral angle of 82.2o and 72.9o (Figure 3.21.). Figure 3.21. Crystal structure of 12 : (top) Each BIPATC4coordinates to four paddlewheel MBBs and the naphthalene-based moiety twists out of plane; (bottom) Space-filling view down the y -axis. Color Code: C = gray; O = red; N = blue; and Cu = green. The dimensions of the cages and wi ndows are directly affected by the conformation of BIPATC4-. Accordingly, the spheroid-like cage has a larger interior cavity as compared to the previous com pounds and thus can accommodate a van der Waals sphere with an estimated diameter of 15.164 . Access to the interior is only permitted through the isosceles windows and the dimensions are significantly reduced as

PAGE 227

190compared to compound 8 and 11 because the naphthalene moiety is in the plane of the window. The length and width of the sma ll window apertures are 12.250 and 3.94 , respectively. The elongated cavity has a ve rtical distance of ~36.432 and it largest diameter can fit a van der Waals sphere of ~11.838 (Figure 3.22.). Figure 3.22. Two types of cages in 12 : (left) Spheroid-like cage, [Cu12(BIPATC)12]; and (right) Elongated cylindrical-like cage, [Cu24(BIPATC)6]. Color Code: C = gray; O = red; and Cu = green. 3.2.2.3. Pillared Tetragonal Square Grid Networks The reaction of H4-BIPATC with Cu(NO3)22.5H2O in a DMA/H2O solution with a minimal amount of pyridine generates homoge nous microcrystalline material in high yield. The parallelepiped shaped crystals we re characterized and formulated by singlecrystal X-ray diffrac tion studies as {[Cu2(BIPATC)(DMA)2]}n9H2O x solvent ( 13 ). The purity of 13 was confirmed by similarities between the calculated and experimental X-ray powder diffraction (XRP D) patterns (Appendix B). The crystal structure of 13 reveals a 3-periodic framework consisting of pi llared tetragonal square grid layers. The layers are built-up from dinuclear paddle wheel MBBs consisting of two copper ions bridged through four deprotona ted carboxylato groups of the diisophthalate units in the

PAGE 228

191BIPATC ligand, [Cu2(RCO2)4(DMA)2]n (dCu-Cu = 2.665 ). The BDC-linked moieties adopt the “1,2-D” configur ation whereby two adjacent meta -carboxyl groups are directed downward and thus the square grid laye r adopts the 1,2-alternate conformation. The SBUs do not assemble to form a cone-shaped square n SBU due to pronounced twisting of the adjacent moieties (Figur e 3.23.). Two opposing molecular squares lie in the same plane while the adjacent molecular squa res are directed upward and downward, respectively to form [Cu2( m -BDC)4]4. Figure 3.23. (left) “1,2-D” configuration associated with the BDC-linked SBU in 13 which can be interpreted as a square building unit by choosin g the centroids of the benzene ring as the points of extension; (right) A cluster of four squares vertex-linked at 120o through 1,3-BDC reveals nanoscale square n SBUs. Color Code: C = gray; O = red; and Cu = green. The nanosized square n SBUs in turn self-assemble at their vertices into a corrugated square grid latti ce that propagates in the xz -plane and is pillared along the y axis through BIPATC4to give a 3-periodic framework with intersection channels (Figure 3.24.). The geometrical conformation of the n SBUs facilitates the formation of an infinite intersecting channel system. These channels are decorated with naphthalene moieties (dihedral angle: 86.13o) and thus the dimensions are redu ced as compared to a coplanar polyphenyl backbone. When viewed down the y -axis it resembles a zigzag channel because the “1,2-D” conformation in one la yer is connected with the opposing layer which is directed upward and vice versa

PAGE 229

192 Figure 3.24. (a) Schematic representation of the corrugated square grid layers in 13 ; (b) Naphthalene moiety in BIPATC4reduced to a linear spacer; and (c) Schematic representation of the pillared square grid layers. Color Code: C = gray; O = red; N = blue. The effective dimensions of the grid s have an edge length of 6.574 and diagonal length of 12.407 . The grids are covalently cross-linked by BIPATC4with an interlayer separation which corresponds to the relative scale of the pillar, ~15.537 . From a topological perspective, the coordi nation sequence and vertex symbol of the 4connected node was determined to be 4, 10, 24, 44, 72, 104, 144, 188, 240, 296 and [4.4.(8)4.8(8).8(8)], respectivel y which corresponds to the lvt network topology. Spacefilling views of 13 (Figure 3.25.) highlight th e intersecting channel sy stem. To the best of our knowledge, no tetracarboxylate pillared structures having lvt have been reported in the literature.

PAGE 230

193 Figure 3.25. Space-filling views from the crystal structure of 13 : (a) Viewed along the x -axis; (b) Tetragonal square grid layer assembled from nanos ized building blocks having the 1,2-alternate conformation; (c) Layers are pillared along the y -axis; (d) Select fragment viewed down the y axis to highlight the conformation (twisting) of the BIPATC ligand. Color Code: C = gray; O = red; N = blue; and Cu = green. Aqua ligands and guest molecules are omitted for clarity. 3.2.2.4. Potential Gas Storage Applications Compounds 8 13 represent three modular platforms which, in principle, can be used to systematically assess adsorbate-MOF interactions by varyi ng specific structural components. They can be categorized and used to evaluate the effect of the following parameters: (1) Metal cation (Cu, Zn and Mn) in an isostructural ( nbo ) series of MOMs constructed from ABTC pillar s; (2) Pore size and functi onality by varying the pillar (ABTC vs. BAYTC vs. BIPATC) in an isoreticular series of MOMs having nbo topology; (3) Pore shape and size between two MOMs having nbo (triangular windows and cages) and lvt (rectangular-shaped channels) topology. Unfortunately, sample activation proved to be very challenging for all compounds due either to framework collapse or improper solvent system.

PAGE 231

194 Indeed, the effect of metal cati on could not be evaluated because 9 and 10 collapse upon removal of the solvent molecu les. This was confirmed by PXRD patterns because no peaks were observed, as is ch aracteristic of an amorphous material. Framework collapse is comm only observed for porous MOMs constructed from paddlewheel building blocks comprised with transition metals other than copper. A possible explanation for this observation is instability of the inorganic MBB upon removal of the axial ligands for metals which do not unde rgo a Jahn-Teller distortion. The copper analogue, 8 was exchanged in acetone for 36 hours and evacuated at 100oC for 12 hours. The solvent-accessible volume for 8 was determined to be 70.9% and the apparent BET and Langmuir surface areas were for 8 determined to be 882 m2/g and 1239 m2/g, respectively (Figure 3.26a.). These values are significantly lower than anticipated because the calculated su rface area is estimated to be 3004 m2/g. The H2 uptake capacity at 723 torr and 77 K was found to be 1.41 wt% (Figure 3.27b.). The isosteric heat of adsorption at low load ing begins at appr oximately 6.8 kJ mol-1 and decreases slightly at high er loading to ~5 kJ mol-1.

PAGE 232

195 Figure 3.26. Sorption data for 8 : (a) Nitrogen sorption isotherm at 77 K; and (b) Hydrogen sorption isotherm at 77 K. Two other research groups independ ently published the same structure, 8 at about the same time. They used different solvent ac tivation procedures which resulted in higher surface area values. Zhou and co-workers re ported apparent BET and Langmuir surface area values of 1407 m2/g and 1779 m2/g using a combined methanol and DCM exchange protocol.80 Suh and co-workers heated the hydrated sample of 8 to 170oC to remove the axial aqua ligands and obtained an a pparent Langmuir su rface area of 2850 m2/g.78 Suh et al. noted that they could not reproduce the sorption data published by Zhou et al. These reports further support the importance of proper sample activation and reproducibility in order to obtain reliable sorption data. The effect of pore size and functionality could in principle be addressed in compounds 8 11 and 12 which are constructed from th e dicopper paddlewheel MBBs but with different pillars – ABTC4-, BAYTC4-, and BIPATC4-, respectively. Compound 12 unfortunately could not be activated and consequently a tentative comparison can only be made between 8 and 11 Table 3.1., provides a summary of the cage dimensions and (a) (b)

PAGE 233

196window apertures for compounds 8 12 The incorporation of a longer organic ligand is a logical design strategy to obtain MOMs w ith higher surface areas and larger pore volumes. Table 3.1. Comparison of the cages and window dimensions between compounds 8 11 and 12 having nbo topology. Compound Cage Dimensions () Window Dimensionsd () Type Ia Type IIb,c 8 ; [Cu2(ABTC)] 9.522 11.891; 22.633 5.883 x 6.885 11 ; [Cu2(BAYTC)] 12.974 9.833; 38.312 13.030 x 6.697 12 ; [Cu2(BIPATC)] 15.164 11.838; 36.432 12.250 x 3.94 a The largest van der Waals sphere that can fit inside the spheriod-shape cage; b The largest van der Waals sphere that can fit inside the elongated cage; c The vertical length of the Type II cage; d Aperture dimensions of the isosceles windows shared between cage types I and II. The cage and window dimensions of 8 11 and 12 correlate with the length and functionality of the organic ligands. As expected, 8 has the smallest cages and windows and 11 and 12 are comparable in size. The window dimensions in 12 are significantly reduced due to twisting of the naphthale ne moiety. The steric conformation of 12 accommodates a larger van der Waals sphere in ca ge I and decorates the interior cavity of cage II which can be advantageous in te rms of increasing gas-MOF interactions. The large solvent-accessible volume for 11 (78.5%) was determined by summing voxels more than 1.2 away from the framework using PLATON software. The estimated BET and Langmuir surfac e areas determined from the N2 sorption were 2055 m2/g and 2499 m2/g, respectively. The H2 uptake capacity was determ ined to be 1.71 wt % at 722 torr and 77 K, Figure 3.27. Compound 11 has a larger pore volume than 8 ; therefore, it is expected to adsorb more H2, particularly at high pressure.

PAGE 234

197 Figure 3.27. Sorption data for 11 : (a) Nitrogen sorption isotherm at 77 K; and (b) Hydrogen sorption isotherm at 77 K. A summary of the sorption data collected on 8 and 11 is provided in Table 3.2. A comparison between the experimental and ca lculated surface area values implies further studies are needed to optimize the porosity of these materials. Ongoing studies are exploring alternative solvent activation methods such as the use of supercritical CO2 activation. Table 3.2. Sorption data for compounds 8 11 and 12 Cpd F. V. a (%) Simulated S.A.b (m2/g) Experimental S. A. (m2/g) H2 Uptake (%) c BET Lang 8 70.9 3004 882 1239 1.41 11 78.5 4991 2055 2499 1.71 12 69.2 3124 N.A. N.A. N.A. a Estimated solvent-accessible free volume calculated using PLATON software; b calculated accessible surface area determined using the Cygwin software package; c H2 uptake at 722 torr; N.A. refers to not applicable due to sample activation problems. Sample activation issues encountered with compound 12 impeded the studies of the effect of pore shape ( lvt versus nbo topologies) on gas-MOF interactions. However, (b) (a)

PAGE 235

198preliminary sorption studies were carried out on the pillared squa re grid structure, 13 with a solvent-accessible volum e of 64.4 %. The as-synthesized sample was exchanged in a variety of volatile solvents such as CH3CN, CH3Cl, DCM, THF, and (CH3)3CO but none were effective, as shown by the low surface area values ( e.g. below 500 m2/g). We decided to take a different approach and exch ange the crystals in a bulkier solvent having a moderate boiling point. The as-synthesized sample was exchanged in chlorobenzene for 4 days and evacuated at 110oC for 12h. The resultant sorp tion data did show an improvement over the other solvent syst ems. The argon sorption isotherm for 13 was a fully reversible type I isotherm, characteris tic of microporous materials. The estimated BET and Langmuir surface areas are 1052 m2/g and 1308 m2/g, respectively, which is appreciably lower than the cal culated surface area of 2621 m2/g. The H2 uptake capacity of 13 was found to be as much as 1.73 wt % at 77 K and 722 Torr. The isosteric heat of adsorption is very similar to that observed in 8 At low loading the enthalpy of adsorption is 6.60 kJ mol-1 and it decreases slightly to 5.76 kJ mol-1 at higher loadings of H2 (Figure 3.28.).

PAGE 236

199 Figure 3.28. Sorption data for 13 : (a) Argon isotherm at 87 K; (b) Hydrogen isotherms at 77 K and 87 K; and (c) Isosteric heat of adsorption for hydrogen; and (d) CO2 isotherm measured at 0oC and 25oC. While compound 13 possesses larger cages than 8 ; however, the increased functionality of the BIPATC ligand facil itates stronger gas-MOF interactions which results in a trade off with respect to Qs t. The enthalpy of adsorption was thereby maintained despite the fact the dimensi ons of the cages were increased. Figure 3.29d. depicts the CO2 sorption isotherms measured at 0oC and 25oC, respectively. The full reversibility of these isotherms further s upports the microporosity of the material, as (a) (c) (b) (d)

PAGE 237

200outlined by the Type I isotherm. The saturati on capacity is reached at relatively low pressures (~ 4 atm) which can be attri buted to the narrow pore size. The maximum amount of CO2 stored in 13 is 31.3 wt% and 27.1 wt% at 0oC and 25oC, respectively. These values are comparable to so me of the best performing zeolites,83 yet lower than some of the recorder holder MOMs.84 The shape of the sorption isotherm is influenced by many factors, one of which is functional gr oups. Accordingly, we may speculate that the interesting structural features in 13 namely narrow pores and carbonyl moieties promote favorable CO2-MOF interactions. This hypothesis is supported by the relatively steep rise in the isotherm at low loading. 3.2.2.5. Lanthanide-Based lvt-MOMs The 15 elements from lanthanum to lutetium are termed lanthanides and constitute the f -block group of metals. Ra re earth metals have rece ived less attention than the transition and main group elements for th e design of MOMs. The main reason for this is their cost and the lack of control over the coordination number and geometry about the metal center. Lanthanides possess unique spectroscopic properties and thus MOMs constructed from such metals have the potenti al to be used in a variety of sensing and related applications.85-89 MOMs are commonly constructed from transition metal cations with coordination numbers ranging from 2 to 6 and in some cases up to 8 ( e.g. Cadmium) has been reported. Rare earth metals have la rger atomic radii and so higher coordination numbers are possible. They are therefore idea lly suited to target nets where the vertex figure presents the need for higher connectivity.

PAGE 238

201Lanthanides ions are hard acids and have a high affin ity for hard donor ligands containing oxygen or hybrid oxygen-nitrogen atoms. It is therefore important for the ligand to have an adequate number of donor atoms. Otherwise the coordination sphere will become saturated with solvent molecules. This could be considered as an advantage or a limitation: the coordinated solvent mol ecules could potentially be removed to expose open metal centers but the inorganic MBB may become unstable upon removal of the ancillary ligands and cause the framework to collapse. H4-ABTC was reacted as part of this project with an assortme nt of lanthanides metal salts such as erbium, ytterbium, cerium, europium, and terbium using a divers e set of reaction condi tions. The objective was to isolate robust 3-periodic MOFs having a high degree of potential coordinately unsaturated metal sites for gas storage studies. Solvothermal reaction between H4-ABTC and Yb(NO3)35H2O in a mildly acidic solution of DMF/CH3CN with a minimal amount of pi perazine (PIP) generates yellow homogenous microcrystalline material. Microcry stalline material was also isolated under similar reaction conditions from the reaction of H4-ABTC with Er(NO3)35H2O from a solution containing DMF/CH3Cl/PIP. The as-synthesized materials, which were determined to be insoluble in H2O and common organic solvents, and were characterized and formulated by SCD studies as {[Yb2(ABTC)(NO3)2(DMF)4]}n ( 14 ) and {[Er2(ABTC)(NO3)2(DMF)4]}n ( 15 ), respectively. Compounds 14 and 15 were determined to be isostructural framewor ks based on similar unit cell parameters (Appendix A). The crystal structure of 14 reveals a 3-periodic framework built-up from dinuclear ytterbium MBBs consisting of tw o crystallographically equivalent ytterbium

PAGE 239

202ions bridged through two deprotona ted carboxylato ABTC moieties (dYb-Yb = 5.180 ), see Figure 3.29. Figure 3.29. Select fragments from the X-ray crystal structure of 14 : (left) 4-connected inorganic MBB, [Yb2(ABTC)(NO3)2(DMF)4]n; and (right) Each ABTC4ligand is connected to four inorganic MBBs. C = gray; O = red; N = blue; and Yb = green. All hydrogen atoms are omitted for clarity. Each trivalent ytterbium ion is 8-coor dinate and is bound to four carboxylato atoms: one ABTC4ligand coordinates in a bidentate fashion (dYb-O = 2.383 , 2.326 ) and two ABTC4ligands are bridged between the me tals in a bis-monodentate fashion (dYb-O = 2.181 , 2.215 ). The remainder of the c oordination sphere is occupied by two oxygen atoms from two DMF molecules (dYb-O = 2.259 , 2.249 ) and two bidentate oxygen atoms from a nitrate ion (dYb-O = 2.417 , 2.374 ). The inorganic metal cluster is linked to four ABTC4ligands and vice versa The assembly of these two 4-connected nodes results in the generation of a cationic MOM with lvt topology. The charge balanced is provided by th e coordinated nitrate (NO3 -) ions. The inorganic MBBs form a chain-like arrangement and propagate in the xzplane. This unit is pillared along the y axis through ABTC4to afford a 3-periodic MOM w ith rectangular-shaped intersecting channels (Figure 3.30.). (a) (b)

PAGE 240

203 Figure 3.30. Select fragments from the X-ray crystal structure of 14 : (left) 4-connected inorganic MBB, [Yb2(ABTC)(NO3)2(DMF)4]n; and (right) Each ABTC4ligand is connected to four inorganic MBBs and vice versa C = gray; O = red; N = blue; and Yb = green. All hydrogen atoms and coordinated DMF molecules are omitted for clarity. The coordinated DMF molecules and nitr ate ions are directed towards these channels. The largest ch annel, viewed along the z -axis, has an estimated diameter of 16.792 by 5.465 . The total solvent-accessible free volumes for the desolvated forms of 14 and 15 are estimated to be 62.7 % and 62.1 %, respectively. The interesting structural features of these charged framew orks are potential open metal sites (4 per metal cluster), open channels, and moderate free volumes. In an attempt to exploit the porous features of 14 and 15 the as-synthesized samples were exchanged in a variety of volatile solvents. Unfortunately, the B ET surface area values were always 0 m2/g. A plausible explanation to account for this observation is that upon removal of the coordinated DMF molecules the metal cluster becomes flexible which leads to the (a) (b)

PAGE 241

204collapse of the framework or that a suitabl e solvent exchange protocol has not been discovered. 3.2.2.6. Default versus Non-Default Structures in Cr ystal Chemistry: MOMs Assembled from Tetrahedral and Rectangular Building Units Inorganic and organic chemistries offer a vast repertoire of MBBs with varying connectivities, which can be utilized to constr uct target networks with desired properties. Despite the large number of structural possibili ties, it is a general thesis that only a small fraction of simple high-symmetry nets will be generated via the self-assembly of simple building blocks. These nets are ther efore often regarded as so-called default nets and are particularly salient to crystal chemistry. In the context of this discussion, binodal edgetransitive nets (Table 3.3.) re present an interesting class of networks; that is, networks assembled from two types of ve rtices and one type of edge. Table 3.3. Select examples of bi nodal edge-transitive nets.90 Coord. Number Vertex Figure Symbol Tiles Transitivity 3,4 triangle; square pto 3[84]+[86] 2122 3,4 triangle; rectangle tbo 2[64]+[86]+[68.86] 2123 3,4 triangle; tetrahedron bor [64]+[64.86] 2122 3,4 triangle; tetrahedron ctn 2[83]+3[83] 2122 3,6 triangle; octahedron pyr [66]+2[63] 2112 3,24 triangle; rhombicuboctahedron rht 6[44]+2[46]+[412] 2123 4,4 square; tetrahedron pts [84]+[42.82] 2132 4,4 square; tetrahedron pth [4.82]+[83] 2132 4,6 square; octahedron soc [412]+3[44.84] 2122 4,8 tetrahedron; cube flu [412] 2111 6,6 octahedron; trigonal prism nia [43]+[49] 2122 The default net for the assembly of sole ly 4-connected square planar nodes is nbo (vertex-transitive); however, the simplest a nd most symmetrical linkage of 4-connected tetrahedral and square planar nodes (in a 1:1 ratio) affords the binodal pts (cooperite) net,

PAGE 242

205which is derived as the edge-net of the CdSO4 ( cds ) uninodal net. Altern atively, the less symmetrical variant of pts ; that is, the hexagonal pth network is also a plausible outcome for the assembly of such MBBs and is derived from the 4-connected quartz ( qzt ) uninodal net. The pts net is topologically equivalent to Platinum Sulfide whereby the planar node ( i.e. Pt) forms PtS4 rectangles and the tetrahedral node ( i.e. S) forms SPt4 tetrahedra. Significant advances in crystal chemistry have made it possible to attain decorated networks by repl acing these nodes with suitabl e MBBs, as exemplified by several reports of MOFs having the pts topology.91, 92 Prefabricated MBBs are thus chosen prior to the assembly process for thei r desired shape, geomet ry, and directionality which are prerequisites to augment an anticipated net. The overall objective of this project was to design rigid archit ectures that support permanent porosity to be utilized in potent ial gas storage applicat ions, as well as, ionexchange studies. The pts net is an ideal synthetic target for such purposes because of its open channel system. Furthermore, this net is not self-dual and ther efore interpenetration is less likely when using long pillars. These structural characteris tics coupled with the inherent modularity afforded by MOFs permits the designer to fine-tune both the inorganic and organic MBBs to favor the formation of decorated pts nets with diverse pore dimensions, functionality, and charge. 3.2.2.6.1. MOFs Constructed from p -block Metal Centers The reaction of H4-ABTC with InCl3 under mild solvothermal conditions indeed yields orange homogeneous crysta ls having a polyhedral morphology ( 16 ). The purity of the as-synthesized compound was confirmed by comparison of the experimental and

PAGE 243

206calculated PXRD patterns. The microcryst alline material, which was found to be insoluble in H2O and common organic solvents, was characterized and formulated by single-crystal X-ray diffraction studies as {[In(ABTC)](DMA)}n ( 16 ). Compound 16 was not initially isolated in a homogenous phase, however. Prior to optimizing the reaction conditions the vial contained a mixtur e of two phases, i.e. crystals with both polyhedral and plate-like mo rphologies were observed. SCD studies confirmed that the plate-like phase was in fact a supramolecular isomer of 16 which was characterized and formulated as {[In(ABTC)](DMA)}n, ( 17 ). Many experimental pr otocols were pursed in an attempt to obtain a pure phase of 17 but were unsuccessful. The single-crystal structure of 16 is a 3-periodic pts -MOF constructed from [In(RCO2)4] inorganic MBBs bridged by the tetracarboxylate organic MBB, ABTC4(Figure 3.31.) Each In(III) is surrounded by four indepe ndent carboxylate groups which coordinate to the metal ion in a bidentate fa shion. This results in a distorted tetrahedral coordination environment with In…O distances ranging from 2.068 to 2.314 and O…In…O angles ranging from 82.088o to 94.595o. Each deprotonated ABTC ligand coordinates to four individua l indium ions and thus acts as a rectangular-planar 4connected node. Since each indium center is trivalent, this yields an overall anionic framework. Charge balance is provided by dimethylammonium cations (DMA+) located in the channels.

PAGE 244

207 Figure 3.31. Schematic representation of 16 and 17 : (center) In(RCO2)4 and H4-ABTC MBBs which can be viewed as tetrahedral and rectangul ar building units, respectively; (left) ball-andstick and augmented representation of 16 ; and (right) ball-and-stick and augmented representation of 17 Color Code: C = gray; O = red; N = blue; and In = green. The infinite channels are represented by the green and purple columns. The crystal structure of 17 is built up from two crys tallographically independent In(III) metal centers. Both are surrounded by four independent carboxylate groups. Each ABTC4ligand coordinates to In(III) in a bide ntate fashion to generate 4-connected [In(RCO2)4] MBBs having distorted tetrahedral geometry. It is worth noting that compound 17 is comprised of two types of inorganic MBBs which are differentiated based on the spacial orientation of the BDCmoiety around the metal centers. One type of [In(O2CR)4] MBB and thus adopts a so-called “1,3-D” conformation whereby two opposite meta -carboxyl groups are directed downward (same as that observed in 16 ). The second MBB resembles a “3-D” confor mation, i.e. three of the four meta -carboxyl groups are facing downward. Both inorganic MBBs a dopt a distorted tetrahedral geometry with

PAGE 245

208In-O distances ranging from 2.136 to 2.483 and 2.136 to 2.386 , respectively. Two crystallographically independent ABTC ligands are present in 17 and both coordinate to four individual indium ions : three “1,3-D” confor mations and one “3-D” surround each ligand. The propagation of the 4-connected tetrahedral and rectangularplanar nodes leads to the fo rmation of an anionic 3-periodic framework with an unprecedented network topology. DMA+ cations in the channels balance the charge. A topological distinction was made usi ng the TOPOS software package whereby the coordination sequence was determined up to k = 10 (Table 3.4). The vertex symbols of the two independent nodes in 16 are 4.4.8(2).8(2).8(8).8(8) and 4.4.8(7).8(7).8(7).8(7). Compound 17 on the other hand is more complex and contains four independent vertices with corresponding VSs of 4.4.8(2).8(2) .8(8).8(8), 4.6(2).6(2) .8(8).8(2).8(2), 4.4.6.8(7).6.8(7), 4.6(2).6.8(7).6.8(7). The degree of regularity can be described based on the transitivity of the ne t which is 2132 and 4896 for 16 and 17 respectively. The network topology of compound 17 is therefore less regular than 16 and not classified as an edge-transitive net. Table 3.4. Coordination sequence up to k = 10 determined for 16 and 17 net vertex cs1 cs2 cs3 cs4 cs5 cs6 cs7 cs8 cs9 cs10 cum10 pts V1 4 10 24 42 64 92 124 162 204 252 979 V2 4 10 24 42 64 90 124 162 204 250 975 novel V1 4 10 24 46 70 100 136 177 230 280 1078 V2 4 11 26 45 71 102 140 182 228 283 1093 V3 4 10 25 47 72 101 137 180 230 284 1091 V4 4 11 25 43 69 101 139 181 228 282 1084

PAGE 246

209The structure of 17 can, for a better understanding of the framework topology, be rationalized as a pillared 3-pe riodic MOM constructed from 2-periodic tetragonal square grid layers (Figure 3.32.). Figure 3.32. Select fragments from the crystal structure of 17 : (a-b) Illustration of the two 4connected indium MBBs, [In(RCO2)4]; (c) Ball-and-stick presentation viewed along the x -axis to show the three types of channels as highlighted by the green, purple, blue, and yellow column; (d) CPK presentation of the 2-D undulating tetragonal layer viewed down the z -axis. Color Code: C = gray; O = red; N = blue; and In = green. Hydrogen atoms and guest molecules are omitted for clarity. The existence of the “3-D” conformation in 17 is the structure directing feature which facilitates the formation of the novel net, as oppose to the pts network topology, 16 The enhanced curvature (be cause of position 4) of the “3-D” conformation cause the cones to alternate and thus in finite undulating layers are obtained. The neighboring layers are pillared by ABTC4(along the z -direction) via a ligand-to-lig and approach to generate (a) (b) (c) (d)

PAGE 247

210the pillared 3-periodic MOF. This arrangement reveals an interesting channel system consisting of four types of infinite narrow pores. The curvature resulting from the “3-D” MBB conformation gives rise to an undulating channel system along the z -direction with a diameter of approximately 6.526 . Thes e channels are inte rsected along the x -direction by the 5-membered ring channels that are hi ghlighted in blue (~6.661 ). Moreover, 1D 4-membered rectangular and square-like channe ls also run along this axis and measure 6.135 x 10.093 and 6.902 x 4.206 , respectively. Compound 16 on the other hand, is solely comprised of the “1,3-D” conformation (Figure 3.33.). For this reason the tetragonal s quare grid layers are not undulating and the bridged layers make this a 3-periodic MOF with infinite and regular 3D channel system with regular channels. Figure 3.33. Select fragments from the crystal structure of 16 : (a) Illustration of the indium MBB, [In(RCO2)4] whereby two ABTC ligands are direct ed upward and two downward; (b) Balland-stick presentation to show the three types of channels, as highlighted by the green, pink, and yellow columns; (c) CPK view of the 2-D tetragonal layer viewed down the z -axis. Color Code: C = gray; O = red; N = blue; and In = green. All hydrogen atoms and guest molecules are omitted for clarity. (a) (b) (c)

PAGE 248

211The 4-membered ring channels directed down the z -axis measure approximately 6.468 in diameter and are intersected by the 4-memb ered ring channel which is rectangular in shape (7.088 x 6.468 ) (purple column in Figure 3.33.). The aforementioned channel is further intersected by another small 4-memb ered ring channel whic h is square-like in shape (4.473 x 4.147 ) (yellow column in Figure 3.33.). 3.2.2.6.2 Compound 16: Prospective Properties The interesting structural features of 16 namely a large solvent-accessible free volume (70.2 %) and the anionic character and infinite channel syst em of intersecting narrow pores, prompted us to assess the gas so rption and ion-exchange properties of this material. We were equally as interested in investigating these properties in 17 but were unable to do so because isolating a pure phase was problematic. Future studies will therefore be devoted to isol ating a homogenous phase of 17 We attempted to investigate the gas sorption properties of 16 by conducting initial studies on the as-synthesized sample, as well as the as-synthesized crystals soaked in a variety of low boiling point organic solvents The list of solvents included acetonitrile, dichloromethane, chloroform, ethanol, benzene, ether, tetrahydrofu ran. Unfortunately, we were unable to obtain a suitable solvent activatio n protocol for this system as the apparent BET surface areas were only in the range of 120 to 280 m2/g, which is much lower than we anticipated. The unexpected porosity data can be attrib uted to: (1) improper sample activation and/or (2) framewor k collapse under vacuum. The latter explanation is less likely because we anticipate the [In(RCO2)4] MBB to be rigid and thus enhance the robustness of the framework. This has been observed in permanently porous pts and pth

PAGE 249

212MOFs synthesized by Eddaoudi and co-workers using 1,2,3,4-BTEC93, as well as, a indium-based MOF synthesized by Schrder and co-workers with pts topology.92 We also explored the potential of exchanging the DMA+ organic cations for a variety of mono-, di-, and triv alent metal cations. The purpose of this study was to assess the H2-MOF interactions in this system as a function of varying the extraframework metal cation, if suitable solvent activation methods could be found. The data would then be compared with the parent co mpound comprised of solely DMA+ cations. A sample of 16 (activated in ethanol) was therefore exchange d in a variety of meta l stock solutions, as summarized in Table 3.5. Table 3.5. Summary of the metal cation exchange studies for 16 {[In(ABTC)]DMA}. Metal Concentrationa Durationb Stability Color Change Ratio S.A.c Na+ 0.5; 0.1 24 N yellow powder N.M. N.M. Li+ 0.5 24 Y no change 2.05:1 129 Mg2+ 0.5 24 Y no change 1.73:1 0 Cu2+ 0.5 24 Y orange to green 1.25:1 351 Co2+ 1.0 24 Y deep orange 1.75:1 257 Ni2+ 1.0 24 Y no change 1.45:1 398 Cr3+ 0.5; 0.1 5 min N xtals dissolve N.M. N.M. Fe2+ 0.5; 0.1 10 min N xtals dissolve N.M. N.M. aConcentration (Molarity); bDuration in hours (unless otherwise noted); cBET surface area (m2/g) measured on the NOVA instrument; N. R. refers to no t measured due instability after metal exchange. The trivalent metal salts immediately a ttacked the framework and dissolved the crystals within minutes. Resu lts obtained for those sample s exchanged with monoand divalent metals were more encouraging, excep t for the sodium sample which rapidly lost crystallinity. The atomic abso rption data collected on the Cu2+, Co2+, and Ni2+ exchanged samples revealed the DMA+ cations were almost completely replaced by the aforementioned metal salts. The esti mated BET surface areas for the Cu2+, Co2+, and Ni2+

PAGE 250

213exchanged samples were found to be 351, 257, and 398 m2/g, respectively. These results are lower than that obtained with the pa rent compound. Future st udies are therefore needed to explore alternative activation methods for this material, such as the use of supercritical CO2 activation. 3.2.2.7. Strategy to Incorporate Unsaturate d Metal Centers into MOFs via Organic Building Bl ocks: Porphyrin-Based Metalloligands Computational and experimental studies have shown that the incorporation of unsaturated metal centers into porous MOFs enhances H2-MOF interactions, as demonstrated by an increase in the isosteric heat of adsorption (often observed at low loadings).94 A considerable effort is therefore bei ng made to incorporate this particular property into MOFs in conjunction with other important characteristics ( e.g. high surface area and localized charge density, optim al pore sizes, low framework density, etc ).1 The most common approach is to create a high concentration of exposed metal sites into the framework of MOFs by th e design and synthesis of solvated MOFs constructed from inorganic MBBs that cont ain coordinated solven t molecule(s) bound to the metal center. This strategy is often most effective for porous MOFs constructed from rigid MBBs ( e.g. metal clusters) because single-metalion based MBBs are frequently too labile to support permanent porosity, once the solvent molecules are removed. Another powerful strategy for creating additional meta l centers, which may however, be hydrated is derived from examples in inorganic zeol ites. Eddaoudi and co-workers utilized the anionic rho -ZMOF as a platform to evaluate the H2-MOF interactions as a function of extra-framework metal cationic complexes. The presence of an electrostatic field induced from the hydrated metal cations was indeed found to enhance the sorption energetics and

PAGE 251

214led to an increase in the is osteric heat of adsorption.95 Incorporation of metal centers in MOMs now includes the organic components whereby additional binding sites are created within the ligand in the form of metal fragments and/or chelated metal centers, as for example in MOFs assembled from sa len-type and porphyrin-based ligands.66, 68, 96-99 We are interested in using tetracarbo xylate porphyrin-based organic ligands (Figure 3.34.) as a pathway to create metal binding centers within the framework backbone of 3-periodic MOFs with pts topology. This approach could generate a platform with a high concentration of exposed metal sites whereby the effect of metal cations on H2-MOF interactions could be evaluated in a systematic fashion if successful. The metal sites could be incorporated into the MOF via the framework backbone and/or extra-framework cations. In the latter case, the porphyin functionality could be metallated with an assortment of me tal cations by using preor post-synthetic modification techniques and in the form the extra-framew ork organic cations could be exchanged for different metal cations. Figure 3.34. Porphyrin-based linkers used in this study: (left) 5,15-bis(2,6-dibromophenyl)10,20-bis(3,5-dicarboxyphenyl)porphyin, (H4-TCBrPP); (right) 5,10,15,20-tetrakis(4-carboxy)-2 1 H ,23 H -porphine (H4-TCPP).

PAGE 252

215 Reaction of H4-TCBrP with In(NO3)36H2O in a mildly acidic solution containing DMF/H2O/2,6-lutidine yields purple crystals with cube-shape morphology. The assynthesized compound was charac terized by single-crystal X -ray diffraction studies and formulated to be {[In(TCBrPP)(H2O)DMA]}n, 18 The crystal structure of 18 reveals a 3-periodic (4,4)-connected MOF c onstructed from mononuclear [In(RCO2)4] nodes that are bridged by the tetracarboxylate TCBrP4porphyrin-based ligand. Each In(III) metal ion is seven-coordinate and binds to oxygen centers from four carboxylato moieties and one terminal aqua ligand to form a squa re pyramidal coordination geometry. Four independent deprotonated TCBrP ligands occupy the equatorial plan e: opposing ligands coordinate to In(III) through the ca rboxylato groups in a bidentate (dIn-O = 2.194 to 2.322 ) and monodentate fashion (dIn-O = 2.020 to 2.109 ), respectively, and an aqua ligand (dIn-O = 2.234 ) completes the coordination sphere in the axial position. The conformation around the inorganic MBB can be rationalized as “3-D ” because three of the four meta -carboxyl groups are facing downward (F igure 3.36.); however, in order to make the pts net the “1,3-D” conformation is desire d. When only the points of extension are considered, the geometry around the inorganic MBB resembles a distorted tetrahedron. As a general theme presented throughout this section, the MOF can be dissected as a 3-periodic fram ework comprised of pillared 2-D tetragonal square grid layers. The layers are assembled from alte rnating cones which measure approximately 7.461 in diameter, as illustrated in Figur e 3.35b. The distance betw een each layer is relative to the length of the pi llar; that is, approximately 9.971 as measured from the 5position of the BDC unit on neighboring layers.

PAGE 253

216 Figure 3.35. Select fragments from the crystal structure of 18 : (a) “3-D” conformation; and (b) Upon deleting the pillars, TCBrP, 2periodic tetragonal square grid layers are revealed, as viewed down the z -axis. Color Code: C = gray; O = red; N = blue; Br = brown; and In = green. The purple atom located in the 5-position of the BDC unit represents the point of extension of the pillar. All hydrogen atoms and guest molecules are omitted for clarity. Each TCBrP4ligand binds to four separate In (III) metal ions which is thus regarded as a 4-connected node. The assemb ly of the two types of 4-connected nodes leads to the formation of a 3-periodic MOF with nia-4,4-Pbca topology. The framework is anionic because the porphyrin unit is not metallated in th e as-synthesized material and each indium metal ion is trivalent. The charge is balanced by disordered DMA+ cations located in the interconnected i rregular shaped channels. The re lative size of the channels is significantly reduced because the porphyrin is oriented perpendicular to the BDC unit with a dihedral angle of 53.92o. The observed conformation results from steric hindrance which takes place between the neighbori ng hydrogen atoms on the BDC and pyridyl units, similar to that observed with BIPA TC. The bromo-functiona lized phenyl moieties are also twisted with respected to th e porphyrin with a dihedral angle of 82.50o. The infinite channels are deco rated by the porphyrin-based group and have approximate diameters of 7.268 and 5.356 (green and pu rple column, respectively; Figure 3.36). The total solvent-accessible free volume for 18 corresponds to approximately 35.4 % of the unit cell volume.

PAGE 254

217 Figure 3.36. X-ray crystal structure of 18 : (a) TCBrPP4organic MBB which coordinates to four [In(RCO2)4(H2O)] inorganic MBBs and thus can be viewed as 4-connected rectangular and distorted tetrahedral building units, respectively; (b) Ball-and-stick and schematic representation viewed along the y-axis to highlight the two typ es of infinite channels shown with green and purple columns; (c) Ball-and-stick and schematic representation viewed along the x-axis to show the undulating conformation of the 2-D layers. Color Code: C = gray; O = red; N = blue; Br = brown; and In = green. All hydrogen atoms and guest molecules are omitted for clarity. From a topological perspective the assemb ly of the two 4-connected nodes did not yield the pts net. Instead the framework adopts a t opology of a new subnet that is derived from a net with a large net relation gr aph (NRG). The binodal network can be topologically described as nia-4,4-Pbca having a transitivity of [2474] and thus is not an edge-transitive net. The 4-c onnected inorganic and organi c nodes have a coordination sequence up to k = 10 of 4, 11, 27, 52, 81, 117, 161, 210, 266, 332 and 4, 11, 27, 51, 81, 116, 161, 209, 266, 329, respectively. These sequen ces correspond to vertex symbols of [4.6.6.8(6).6.8(6)] and [4.6.6.8.6.8(7)], respectively. This net is rarely observed in MOMs and to the best of our knowledge, only two ot her MOFs exist with this network topology. ( a ) ( b ) ( c )

PAGE 255

218This is in addition to one hydrogen b onded network (CSD ref codes: NEHMET, WINJOS, CEJRUE). While our objective was to isolate MOFs with pts topology using the porphyrinbased ligands presented, we have shown in this section that this was not the case. This unexpected result raises important questions such as: (1) why do some nets occur more frequently? Or (2) why is it that some nets rarely or never occur in nature? A general answer is that nature prefers to form the simplest nets with the highest symmetry. The connectivity observed in compound 18 is an excellent example of a rare net but this raises the question of why the pts net did not form? A plausible explanation may be the steric hindrance brought about by the twisted conformation of the porphyrin. The relative size of the 4-membered ring channels along the z -direction (square grid tetragonal layers) in a pts net constructed from 1,3-BDC organic li nkers is somewhat predetermined by the BDC units. Formation of the target net might be blocked if bulky substituents, such as the porphyrin unit, are directed inward towards the channel, as opposed to lying in the plane of the BDC unit, this co uld prevent the formatio n of the target net. We report a 3-periodic porphyrin-bas ed MOF having an unprecedented topology. Despite many attempts no crystalline mate rial could however be isolated with H4-TCPP. The anionic character of 18 offers the potential for post-synthetic modification experiments whereby the extraframework meta l cations are exchanged for different metal ions. In a similar vein, post-s ynthetic modificatio n is feasible with the porphyrin unit because it is not metallated in the as-synthe sized compound. This material may therefore be of interest for potential catalysis applications.

PAGE 256

2193.2.2.8. MOFs with pts Topology Construc ted from Monoand Bimetallic Transition Metal Centers Solvothermal reaction between H4-ABTC and Co(NO3)26H2O or Cd(NO3)24H2O in a mildly acidic mixture containing DMF/CH3CN/H2O and DMF/NMP yields crystalline samples with a polyhedral morphology. The as-synthesized compounds were characterized by single-crystal X-ray diffr action studies and formulated to be {[(Co4(ABTC)2(DMF)2(EtOH)3](EtOH)4H2O}n, ( 19 ), and {[(Cd2(ABTC)(DMF)3](DMF)3(H2O)5}n, ( 20 ), respectively. Compound 19 crystallizes in the P2(1)2(1)2(1) space group and belongs to the orthorhombic crystal system with a = 13.514(3) , b = 13.766(3) , c = 45.290(8) . Compound 19 was determined to be isostructural to 20 with a = 23.516(0) , b = 15.120(7) , c = 12.859(7) . The crystal structure of 19 is that of a 3-periodic network built-up from bimetallic Co2(RCO2)4(DMF)(EtOH) and Co2(RCO2)4(DMF)(EtOH)2 building blocks, which are linked through ABTC4units to exhibit a “1,3-D” conformation. The only difference between the aforementioned inorganic MBBs is the number of coordinated solvent molecules; otherwise, the geometry around the Co(II) cations and thus the points of extension are the same. Four deprotonated AB TC ligands coordinate to Co1 and Co2, in each cluster, by means of the car boxylato oxygen atoms (Figure 3.37.).

PAGE 257

220 Figure 3.37. X-ray crystal structure of 19 : (top-center) Ball-and-stick representation of the two “1,3-D” inorganic MBBs: [M(ABTC)4(DMF)(EtOH)] and [M(ABTC)4(DMF)(EtOH)2], where M = Co or Cd. The MBBs can be simplified into 4connected distorted tetrahedral and rectangularplanar nodes; (left) Ball-and stick and schematic representation viewed down the z -direction; and (right) ball-and stick and schematic representation viewed down the x -direction. All hydrogen atoms and solvent molecules are omitted for clarity. Color Code: C = gray; O = red; N = blue; and Co = green. The coordination environment around Co1 consists of six oxygen atoms: one ligand coordinates in a bidentate fashion (dCo O = 2.097 and 2.217 ), two coordinate in a bis-monodentate fashion (dCo O = 2.026 and 2.085 ; dCo O = 2.008 and 1.936 ) and serve as a bridge between Co1 and Co2, and the coordination sphere is completed by one ABTC ligand which coordinate s in a bidentate fashion (dCo O = 2.243 ) and simultaneously is bridged to Co2 via the deprotonated hydroxyl moiety (dCo O = 2.058 ). Each Co2 metal cation adopts a square pyram idal geometry. The equatorial plane is occupied by two carboxylato oxygen atoms from two independent ligands which coordinate in a bis-monodentate (dCo O = 2.085 ) and bifurcated bis-monodendate

PAGE 258

221fashion (dCo O = 2.058 ), respectively. The remaini ng equatorial sites contain oxygen atoms from coordinated DMF a nd ethanol solvent molecules (dCo O = 2.100 and dCo O = 1.969 ). This axial sites is occupied by one deprotonated ligand which coordinates to Co2 in a bis-monodentate fashion (dCo O = 1.936 ). The bimetallic Co(II) clusters can be viewed as 4-connected nodes having distorted tetrahedral geometry, if all the c oordinated solvent mol ecules are disregarded. Each tetrahedral building unit is linked by four separate re ctangular-planar building units to form a decorated 3-periodic MOF with pts topology. The overall charge is neutral on account of the Co(II) bimetallic cluster in contrast to the indium-based pts net (compound 16 ). The infinite channel system is an alogous to that desc ribed for compound 16 ; however, the rectangular (4.985 x 7.526 ) and rhombic-like (5.480 x 10.193 ) channels are filled with coordinated and guest DMF and ethanol molecules that are directed towards the interior of the channe ls. The total solvent-accessible free volumes calculated for 19 and 20 is estimated to be 70.9 and 71.4 %, respectively (coordinated and guest solvent molecules were omitted for this calculation). We were unable to study the gas-sorpti on properties of these materials because the as-synthesized and solvent exchanged samp les did not exhibit permanently porosity, as confirmed by a B ET surface area of 0 m2/g. The XPRD pattern following sorption analysis did not reveal any p eaks which is indicative of fr amework collapse. These results are not surprising, as many MOFs constructed from inorganic MBBs with a high degree of coordinated solvent molecules becomes less rigid upon solvent removal. It was our goal to circumvent this shortcoming th rough the design and synthesis of compound 16

PAGE 259

222which is constructed from rigid single -metal-ion based inorganic MBBs (section 3.2.3.3.1.). The primary objective of th is project was to react H4-ABTC with variety of transition metal cations ( e.g. Cr, Mn, Fe, Co, Ni, Cu, Zn, Cd) which are capable of maintaining an octahedral c oordination environment. The target was to generate an isoreticular series of MOFs with soc (= s quare oc tahedron) topology – see Chapter 2 for a detailed explanation. A serious challenge in the synthesis of such compounds is to uncover the optimal reaction conditions that can always lead to the formation of a target MBB, such as the reactions conditions which consistently lead to the formation of the oxo-centered metal trimer clus ter. During the course of this work, the following experimental parameters were adjusted in an attempt to favor the formation of this trimeric MBB: M:L ratio, concentration, pol arity, acidity, temperature, alternative counter ions, and various SDAs. The TMBB was only isolated for In(III) and Fe(III) metal cations, while Co, Ni, Cu, Zn, Cd co mplexes predominately formed 4-connected nodes with square-planar or te trahedral geometries having nbo and pts topologies, respectively. It was therefore not very surprising that the solvothermal reaction of H4-ABTC with Co(NO3)26H2O or Mn(OAc)32H2O under mildly acidic conditions lead to the formation of two isostructural MOFs with pts topology. The as-synthesized compounds were characterized by single-c rystal X-ray diffraction studi es and formulated to be {[(Co(H-ABTC)(H2O)2]}n, 21 and {[(Mn(H-ABTC)(H2O)2]}n, 22 respectively. The former compound was determined to crystallize in the P21/c space group with a = 11.024(3) , b = 22.005(4) , c = 9.691(9) . Similarly, compound 22 crystallizes in the

PAGE 260

223C2/c space group with a = 11.117(8) , b = 22.102(5) , c = 9.722(6) . A crystallographic analysis of 21 revealed that its structur e is built up from 4-connected single-metal ion [Co(RCO2)4] building blocks, which adopt the “1,3-D” conformation. Four independent BDC moieties of the ABTC ligands are positioned in the same plane, which is in contrast to other MOFs in which the opposing BDC groups are oriented perpendicular to the plane. Four deprotona ted ABTC ligands thus coordinate to each Co(II) metal center through the carboxylato oxygen atoms: the equatorial plane is occupied by two independent ligands coordinated in a cis -monodentate fashion (dCo-O = 2.064 and 2.093 ), and two disordered DMF molecules (dCo-O = 2.082 and 2.101 ), while the axial sites are occupied by two additi onal ligands which are also coordinated in a monodentate fashion (dCo-O = 2.062 and 2.104 ), as shown in Figure 3.38a. The inorganic MBB can be interpreted as a 4connected MBB, if the DMF molecules are neglected, with a see-saw or a highly distorted tetrahedral ge ometry. The self-assembly of the two 4-connected MBBs results in an anionic 3-periodic MOF with pts topology. Charge is balance is provided by DMA+ cations, which are located in the infinite rectangular channels that measure approximately 4.6 x 7.3 in diameter (Figure 3.38bc.).

PAGE 261

224 Figure 3.38. Select fragments from the crystal structure of 21 : (a) [Co(ABTC)4(DMF)2], which can be viewed as a 4-connected node with a see-saw or highly distorted tetrahedral geometry; (b) Ball-and-stick representation of the infinite rectangular channels that run along the z -axis; (c) Schematic representation of the 4-connected build ing units down the z-axis. All hydrogen atoms and solvent molecules are omitted for clarity. Color Code: C = gray; O = red; N = blue; and Mn = green. The open channels are highlig hted with the green columns. The underlying connectivity observed in 21 and 22 is the same but the presence of a monoversus bimetallic MBB influences th e packing of the pillared 2-D layers along the y -axis. The conformation of the BDC units in 21 which are all being parallel to the plane, thus prevents the formation of 2-D te tragonal square grid layers. Each layer is comprised of 4-membered rings, constr ucted from two opposing BDC units which interact with neighboring layers in a zi g-zag fashion. Gas-so rption properties of 21 and 22 could not be investigated since neither of the MOFs retained their crystallinity upon the removal of coordinated solvent and guest molecules. (a) (b) (c)

PAGE 262

2253.2.3. Experimental Section 3.2.3.1. Materials and Methods All materials and methods are described in Chapter 2, unless otherwise noted. A detailed synthesis of H4-ABTC is described in Chapter 2. The tetracarboxylate porphryin-based organic ligand, H4-TCBrP, was synthesized by Chungsik Kim in Dr. Peter Zhang’s lab at the University of South Florida and H4-BAYTC was synthesized by Bao Zhong in Dr. Graham Bodwell’s lab at Memorial University in St. John’s, Newfoundland. 3.2.3.2. Synthesis and Characterization Figure 3.39. A schematic to illustrate the cocrysta l controlled solid-state synthesis (C3S3) of benzoimidephenanthroline tetracarboxylic acid (H4-BIPATC). Reagents and Conditions: (i) DMF, (ii) Grind the solids together (uniform purple color), and (iii) = 180oC for 1 hr. Preparation of benzoimidephenant hroline tetracarboxylic acid, H4-BIPATC: 5-Aminoisophthalic acid (1.81g, 10mmol) and 1,4,5,8-naphthalenetetracarboxylic dianhydride (1.34g, 5mmol) were ground together using a mortar and pestle in a minimal amount of DMF (~0.150mL). The homogenous purple solid was heated at 180oC for approximately 1 hr and cooled to room te mperature to yield a yellow powder. The assynthesized product does not requ ire additional purification step s; however, if desired the solid can be refluxed in DMF overnight if a dditional purification is deemed necessary.

PAGE 263

226Synthesis of {[Cu2(ABTC)(H2O)2]}n, (8) A mixture of Cu(NO3)22.5H2O (8.8mg, 0.070mmol), H4-ABTC (12.1mg, 0.034mmol), DMF (0.5mL), CH3CN (1mL), H2O (0.5mL), and HMTA (0.15mL 1M in DMF) was prepared in a 20 mL scintillation vial. The solution was heated to 85oC for 12 hours and cooled to room temperature. Orange cube-shaped crystals of 8 which were found to be insoluble in H2O and common organic solvents, were ha rvested and air-dried. Synthesis of {[Zn2(ABTC)(H2O)2]}n, (9) A mixture of Zn(NO3)26H2O (8.8mg, 0.070mmol), H4-ABTC (12.1mg, 0.034mmol), DEF (1 mL) was prepared in a 20 mL scintillation vial. The so lution was heated to 85oC for 12 hours and cooled to room temperature. Orange cube-shaped crystals of 9 which were found to be insoluble in H2O and common organic solvents, we re harvested and air-dried. Synthesis of {[Mn2(ABTC)(H2O)2]}n, (10) A mixture of MnCl2 (8.8mg, 0.070mmol), H4-ABTC (12.1mg, 0.034mmol ), DMF (0.5mL), CH3CN (1mL) and HMTA (0.15mL, 1M in DMF) was prepared in a 20 mL scintillation vial. The solution was heated to 85oC for 12 hours and cooled to room temperature. Orange cube-shaped crystals of 10 which were found to be insoluble in H2O and common organic solvents, were harvested and air-dried. Synthesis of {[Cu2(BAYTC)(H2O)2]}n, (11) A mixture of Cu(NO3)2 .2.5H2O (20.0mg, 0.086mmol), H4-BAYTC (15.4mg, 0.035mmol), DEF (1mL), chlorobenzene (1mL), tetrabutylammonium bromide (0.2mL, 1M in H2O), and HNO3 (0.mL, 3.5M in DMF) was prepared in a 20 mL scintillati on vial and subsequently heated to 85oC for 12 hours and cooled to room temperature to a fford green parallelepi ped crystals. The as-

PAGE 264

227synthesized crystals of 11 were harvested, air-dried, and determined to be insoluble in H2O and common organic solvents. Synthesis of {[Cu2(BIPATC)(H2O)2]}n, (12) A mixture of Cu(NO3)2 .2.5H2O (35.8mg, 0.20mmol), H4-BIPATC (31.2mg, 0.20mmol), DMF (1mL), CH3CN (1mL) and guanidine nitrate (0.2mL, 1M in DMF) was prepared in a 20 mL scintillation vial. The solution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours and 115oC for 23h hours. Each of the heating cy cles were increa sed at rate of 1.5oC/minute and cooled to room temperature at a rate of 1.0oC/minute. Green hexagonal crystals of 12 were harvested and air-dried. The as -synthesized material was determined to be insoluble in H2O and common orga nic solvents. Synthesis of {[Cu2(BIPATC)(H2O)2]}n, (13) A mixture of Cu(NO3)2 .2.5H2O (11.6mg, 0.050mmol), H4-BIPATC (28.5mg, 0.052mmol), DMA (1mL), H2O (1mL) and pyridine (0.1mL, 1M in DMF) was prepared in a 20 mL scintillati on vial. The solution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours and 115oC for 23h hours. Green para llelepiped crystals of 13 were harvested and air-dried. The as-synthesized material was determined to be insoluble in H2O and common organic solvents. Synthesis of {[Yb2(ABTC)(NO3)2(DMF)4]}n, (14) A mixture of Yb(NO3)3 .5H2O (50.0mg, 0.11mmol), H4-ABTC (7.97mg, 0.022mmol), DMF (1mL), CH3CN (1mL), PIP (0.15mL, 1M in DMF), and HNO3 (0.15mL, 3.5M in DMF) wa s prepared in a 20 mL scintillation vial. The so lution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours and cooled to room te mperature. Orange block-shaped

PAGE 265

228crystals of 14 which were found to be insoluble in H2O and common organic solvents, were harvested and air-dried. Synthesis of {[Eb2(ABTC)(NO3)2(DMF)4]}n, (15) A mixture of Er(NO3)3 .5H2O (50.0mg, 0.11mmol), H4-ABTC (8.08mg, 0.022mmol), DMF (1mL), CH3Cl (0.5mL), PIP (0.2mL, 1M in DMF), and HNO3 (0.15mL, 3.5M in DMF) was prepared in a 20 mL scintillation vial. The so lution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours and cooled to room te mperature. Orange block-shaped crystals of 15 which were found to be insoluble in H2O and common organic solvents, were harvested and air-dried. Synthesis of {[In(ABTC)](DMA)}n, (16) A mixture of InCl3 (22.1mg, 0.10mmol), H4-ABTC (35.8mg, 0.10mmol), DMF (1mL), NMP (0.5mL), H2O (0.5mL) and HNO3 (0.45mL, 3.5M in DMF) was prepared in a 20 mL scintillation vial and heated to 85oC for 12 hours and cooled to room temper ature. Orange polyhedral crystals of 16, which were found to be insoluble in H2O and common organic solvents, were harvested and air-dried. Synthesis of {[In(ABTC)](DMA)}n, (17) A mixture of InCl3 (30.0mg, 0.14mmol), H4-ABTC (24.4mg, 0.07mmol), DMF (1mL), ethanol (0.5mL), H2O (0.5mL) and HNO3 (0.45mL, 3.5M in DMF) was prepared in a 20 mL scintillation vial and heated to 85oC for 12 hours and cooled to room temperat ure to afford a heterogeneous mixture of 16 and 17 Despite many trials, a homogenous phase of 17 could not be isolated. Synthesis of {[In(TCBrPP)(H2O)2]DMA}n, (18) A mixture of In(NO3)2 .6H2O (3.24mg, 0.011mmol), H4-TCBrPP (3.00mg, 0.003mm ol), DMF (1mL), H2O (1mL), 2,6lutidine (0.1mL in 1M DMF), and HNO3 (0.25mL, 3.5M in DMF) was prepared in a 20

PAGE 266

229mL scintillation vial. The solution was heated to 85oC for 12 hours and cooled to room temperature whereby red cube-shaped crystals were obtained. The as-synthesized crystals of 18 were harvested, air-dried, and determined to be insoluble in H2O and common organic solvents. Synthesis of {[Co4(ABTC)2(DMF)2(EtOH)3](EtOH)4(H2O)}n, (19) A mixture of Co(NO3)2 .6H2O (15.0mg, 0.052mmol), H4-ABTC (9.23mg, 0.03mmol), DMF (1mL), CH3CN (1.5mL), H2O (0.25mL), and HNO3 (0.15mL, 3.5M in DMF) was prepared in a 20 mL scintillation vial. The solution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours and cooled to room temperature. Purple polyhedral crystals of 19 were harvested and air-dried. The as-synthesized material was determined to be insoluble in H2O and common organic solvents. Synthesis of {[Cd4(ABTC)2(DMF)2(EtOH)3]}n, (20) A mixture of Cd(NO3)2 .4H2O (30.8mg, 0.1mmol), H4-ABTC (11.6mg, 0.03mmol), DMF (1mL), NMP (1mL), and HNO3 (0.15mL, 3.5M in DMF) was prepar ed in a 20 mL scintillation vial. The solution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours and cooled to room temperatur e. Orange polyhedral crystals of 20 were harvested and air-dried. The as -synthesized material was determined to be insoluble in H2O and common organic solvents. Synthesis of {[Co(ABTC)(H2O)2]}n, (21) A mixture of Co(NO3)2 .6H2O (29.1mg, 0.10mmol), H4-ABTC (9.23mg, 0.03mmol), DMF (1mL), CH3CN (1mL), and H2O (0.25mL) was prepared in a 20 mL scintil lation vial and subsequently heated to 85oC for 12 hours and cooled to room temperate to afford pink plate-shaped crystals. The

PAGE 267

230as-synthesized crystals of 21 were harvested, air-dried, and determined to be insoluble in H2O and common organic solvents. Synthesis of {[Mn(ABTC)(H2O)2}n, (22) A mixture of Mn(OAc)3 .2H2O (20.0mg, 0.074mmol), H4-ABTC (17.8mg, 0.05mmol), DMF (1mL), H2O (1mL) and HNO3 (0.3mL, 3.5M in DMF) was prepared in a 20 mL scintillation vial. The solution was heated to 85oC for 12 hours and cooled to room te mperate to afford pale orange plate-shaped crystals. The as-synthesized crystals of 22 were harvested, air-dried, and determined to be insoluble in H2O and common organic solvents. 3.3. Summary and Conclusions A series of MOMs built up from solely 4-connected nodes have been described. The majority of the structures presented in this chapter were in fact targeted and therefore exemplify the power of the MBB approach. Seve ral of the structures did, however, form by accident but are also interest ing in their own regard. Three nets in particular were the focus of this study; that is, the nbo lvt and pts nets. The resultant MOMs were isolated with a ligand-to-ligand pillar ing strategy by employing suitab le tetracarboxylate organic spacers ( e.g. H4-ABTC, H4-BIPATC, and H4-BAYTC) in conjunction with inorganic MBBs having the appropriate geometry needed to yield the target net. We have thereby shown that covalently cross-linking kgm and sql layers with a rigid ligand results in 3periodic MOMs with nbo and lvt topology, respectively. Although both of these nets are constructed from 4-connected nodes their channel sy stems are significantly different in that the form is comprised of two types of cages while the latter is exhibits regular channels. A platform has thus been establis hed on the basis of which the effects from

PAGE 268

231metal cation, pore size, and pore shape can be evaluated with respect to H2-MOF interactions. Sample activation problems did, however, impede thes e studies but efforts are underway to explore altern ative activation protocols ( e.g. ScD CO2 activation). Gas sorption properties of the anionic indium-based pts -MOF were investigated but the values were much lower than anticipated. It will, therefore be necessary to study this problem in more detail to determine if the MO F collapses under vacuum or if it is just a matter of identifying an optimal activation protocol. Future work will work will also focus upon synthesizing 4-connected MO Ms based upon flexible tetracarboxylate ligands. This would allow for an elevated degree of diversity via ligand design which may otherwise be unattainable w ith the use of rigid spacers. 3.4. References (1) Murray, L. J.; Dinca, M.; Long, J. R. Chem. Soc. Rev. 2009 38 1294-1314. (2) Li, J. R.; Kuppler, R. J.; Zhou, H. C. Chem. Soc. Rev. 2009 38 1477-1504. (3) Horcajada, P.; Chalati, T.; Serre, C.; G illet, B.; Sebrie, C.; Baati, T.; Eubank, J. F.; Heurtaux, D.; Clayette, P.; Kreuz, C.; Chang, J. S.; Hwang, Y. K.; Marsaud, V.; Bories, P. N.; Cynober, L.; Gil, S. ; Ferey, G.; Couvreur, P.; Gref, R. Nat. Mater. 2010 9 172-178. (4) Lee, J.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen, S. T.; Hupp, J. T. Chem. Soc. Rev. 2009 38 1450-1459. (5) Ma, L. Q.; Abney, C.; Lin, W. B. Chem. Soc. Rev. 2009 38 1248-1256. (6) Wang, Z. Q.; Cohen, S. M. Chem. Soc. Rev. 2009 38 1315-1329. (7) Allendorf, M. D.; Bauer, C. A.; Bhakta, R. K.; Houk, R. J. T. Chem. Soc. Rev. 2009 38 1330-1352. (8) Zacher, D.; Shekhah, O.; Woll, C.; Fischer, R. A. Chem. Soc. Rev. 2009 38 14181429. (9) Eddaoudi, M.; Moler, D. B.; Li, H. L.; Chen, B. L.; Reineke, T. M.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2001 34 319-330. (10) Moulton, B.; Zaworotko, M. J. Chem. Rev. 2001 101 1629-1658. (11) Ferey, G. J. Solid State Chem. 2000 152 37-48. (12) Robson, R. J. Chem. Soc., Dalton Trans. 2000 3735-3744. (13) Tranchemontagne, D. J.; Mendoza-Cort es, J. L.; O'Keeffe, M.; Yaghi, O. M. Chem. Soc. Rev. 2009 38 1257-1283.

PAGE 269

232(14) Biradha, K.; Domasevitch, K. V.; M oulton, B.; Seward, C.; Zaworotko, M. J. Chem. Commun. 1999 1327-1328. (15) Zaworotko, M. J. Chem. Commun. 2001 1-9. (16) Biradha, K.; Sarkar, M.; Rajput, L. Chem. Commun. 2006 4169-4179. (17) Gable, R. W.; Hoskins, B. F.; Robson, R. J. Chem. Soc., Chem. Commun. 1990 1677-1678. (18) Kitaura, R.; Seki, K.; Akiyama, G.; Kitagawa, S. Angew. Chem. Int. Ed. 2003 42 428-431. (19) Li, H.; Eddaoudi, M.; Groy, T. L.; Yaghi, O. M. J. Am. Chem. Soc. 1998 120 8571-8572. (20) Biradha, K.; Hongo, Y.; Fujita, M. Angew. Chem. Int. Ed. 2002 41 3395-3398. (21) Bourne, S. A.; Lu, J. J.; Mondal, A.; Moulton, B.; Zaworotko, M. J. Angew. Chem. Int. Ed. 2001 40 2111-2113. (22) Moulton, B.; Lu, J. J.; Hajndl, R.; Hariharan, S.; Zaworotko, M. J. Angew. Chem. Int. Ed. 2002 41 2821-2823. (23) Furukawa, H.; Kim, J.; Ockwi g, N. W.; O'Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2008 130 11650-11661. (24) Oostinga, J. B.; Heersche, H. B.; Li u, X. L.; Morpurgo, A. F.; Vandersypen, L. M. K. Nat. Mater. 2008 7 151-157. (25) Choy, J. H.; Choi, S. J.; Oh, J. M.; Park, T. Appl. Clay Sci. 2007 36 122-132. (26) Dunbar, K. R.; Heintz, R. A. Prog. Inorg. Chem. 1997 45 283-391. (27) Iwamoto, T.; Nakano, T.; Morita, M.; Miyoshi, T.; Miyamoto, T.; Sasaki, Y. Inorg. Chim. Acta. 1968 2 313-316. (28) Iwamoto, T., Inclusion Compounds: Structural Aspects of Inclusion Compounds Formed by Inorganic and Organometallic Host Lattices Academic Press: London, 1984; Vol. 1, p 29-57. (29) Lipowski, J., Inclusion Compounds: Structural Aspects of Inclusion Compounds Formed by Inorganic and Organometallic Host Lattices Academic Press: London, 1984; Vol. 1, p 59-103. (30) Ockwig, N. W.; Delgado-Friedr ichs, O.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2005 38 176-182. (31) Delgado-Friedrichs, O.; O'Keeffe, M.; Yaghi, O. M. Phys. Chem. Chem. Phys. 2007 9 1035-1043. (32) O'Keeffe, M. Reticular Chemistry Structure Resource. http://rcsr.anu.edu.au/ (33) Nytko, E. A.; Helton, J. S.; Muller, P.; Nocera, D. G. J. Am. Chem. Soc. 2008 130 2922-2923. (34) Mahata, P.; Sen, D.; Natarajan, S. Chem. Commun. 2008 1278-1280. (35) Murugesu, M.; Clerac, R.; Anson, C. E.; Powell, A. K. J. Phys. Chem. Solids 2004 65 667-676. (36) Volkringer, C.; Meddouri, M.; Loiseau, T.; Guillou, N.; Marrot, J. r. m.; Ferey, G. r.; Haouas, M.; Taulelle, F.; Audebrand, N.; Latroche, M. Inorg. Chem. 2008 47 11892-11901. (37) Horike, S.; Hasegawa, S.; Tanaka, D.; Higuchi, M.; Kitagawa, S. Chem. Commun. 2008 4436-4438. (38) Syozi, I. Theor. Phys. 1951 6 306-308.

PAGE 270

233(39) Wells, A., Three-dimensional Nets and Polyhedra John Wiley & Sons: Hoboken, 1977 (40) Lu, J. J.; Mondal, A.; Moulton, B.; Zaworotko, M. J. Angew. Chem. Int. Ed. 2001 40 2113-2116. (41) Moulton, B.; Lu, J. J.; Mondal, A.; Zaworotko, M. J. Chem. Commun. 2001 863864. (42) Abourahma, H.; Bodwell, G. J.; Lu, J. J.; Moulton, B.; Pottie, I. R.; Walsh, R. B.; Zaworotko, M. J. Cryst. Growth Des. 2003 3 513-519. (43) Chui, S. S. Y.; Lo, S. M. F.; Charmant, J. P. H.; Orpen, A. G.; Williams, I. D. Science, 1999 283 1148-1150. (44) Cotton, F. A.; Lin, C.; Murillo, C. A. Acc. Chem. Res. 2001 34 759-771. (45) Nouar, F.; Eubank, J. F.; Bousquet, T. ; Wojtas, L.; Zaworotko, M. J.; Eddaoudi, M. J. Am. Chem. Soc. 2008 130 1833-1835. (46) Lin, X.; Telepeni, I.; Blake, A. J.; Dail ly, A.; Brown, C. M.; Simmons, J. M.; Zoppi, M.; Walker, G. S.; Thomas, K. M.; Mays, T. J.; Hubberstey, P.; Champness, N. R.; Schroder, M. J. Am. Chem. Soc. 2009 131 2159-2171. (47) Abourahma, H.; Coleman, A. W.; M oulton, B.; Rather, B.; Shahgaldian, P.; Zaworotko, M. J. Chem. Commun., 2001 2380-2381. (48) Furukawa, H.; Kim, J.; Plass, K. E.; Yaghi, O. M. J. Am. Chem. Soc. 2006 128 8398-8399. (49) Perry, J. J.; McManus, G. J.; Zaworotko, M. J. Chem. Commun. 2004 2534-2535. (50) Bott, S. G.; Coleman, A. W.; Atwood, J. L. J. Am. Chem. Soc. 1986 108 17091710. (51) Da Silva, E.; Lazar, A. N.; Coleman, A. W. J. Drug Delivery Sci. Technol. 2004 14 3-20. (52) Gutsche, C. D.; Dhawan, B.; No, K. H.; Muthukrishnan, R. J. Am. Chem. Soc. 1981 103 3782-3792. (53) Rebek, J. Chem. Soc. Rev. 1996 25 255-264. (54) Gutsche, C. D. Acc. Chem. Res. 1983 16 161-170. (55) Wieser, C.; Dieleman, C. B.; Matt, D. Coord. Chem. Rev. 1997 165 93-161. (56) Atwood, J. L.; Barbour, L. J.; Hardie, M. J.; Raston, C. L. Coord. Chem. Rev. 2001 222 3-32. (57) Kim, J. S.; Quang, D. T. Chem. Rev. 2007 107 3780-3799. (58) Lloyd-Williams, P.; Giralt, E. Chem. Soc. Rev. 2001 30 145-157. (59) Gutsche, C. D.; Dhawan, B.; Levine, J. A.; No, K. H.; Bauer, L. J. Tetrahedron 1983 39 409-426. (60) Araki, K.; Iwamoto, K.; Shinkai, S.; Matsuda, T. Chem. Lett. 1989 1747-1750. (61) Pappalardo, S.; Ferguson, G.; Neri, P.; Rocco, C. J. Org. Chem. 1995 60 45764584. (62) Iwamoto, K.; Shinkai, S. J. Org. Chem. 1992 57 7066-7073. (63) Groenen, L. C.; Vanloon, J. D.; Verboom W.; Harkema, S.; Casnati, A.; Ungaro, R.; Pochini, A.; Ugozzoli, F.; Reinhoudt, D. N. J. Am. Chem. Soc. 1991 113 23852392. (64) Eddaoudi, M.; Kim, J.; Vodak, D.; Sudi k, A.; Wachter, J.; O'Keeffe, M.; Yaghi, O. M. Proc. Nat. Acad. Sci. U.S.A. 2002 99 4900-4904.

PAGE 271

234(65) Wade, L. G., Organic Chemistry 6 ed.; Pearson Prentice Hall: Upper Saddle River, New Jersey 2006 (66) Choi, E.-Y.; Barron, P. M.; Novot ny, R. W.; Son, H.-T.; Hu, C.; Choe, W. Inorg. Chem. 2008 48 426-428. (67) Farha, O. K.; Malliakas, C. D.; Kanatzidis, M. G.; Hupp, J. T. J. Am. Chem. Soc. 2010 132 950-952. (68) Shultz, A. M.; Farha, O. K.; Hupp, J. T.; Nguyen, S. T. J. Am. Chem. Soc. 2009 131 4204-4205. (69) Chung, H.; Barron, P. M.; Novotny, R. W.; Son, H. T.; Hu, C.; Choe, W. Cryst. Growth Des. 2009 9 3327-3332. (70) Chen, S.; Zhang, J.; Bu, X. Inorg. Chem. 2008 47 5567-5569. (71) Chun, H.; Moon, J. Inorg. Chem. 2007 46 4371-4373. (72) Ma, B.-Q.; Mulfort, K. L.; Hupp, J. T. Inorg. Chem. 2005 44 4912-4914. (73) Chen, B.; Ma, S.; Zapata, F.; Fr onczek, F. R.; Lobkovsky, E. B.; Zhou, H.-C. Inorg. Chem. 2007 46 1233-1236. (74) Kondo, M.; Okubo, T.; Asami, A.; Noro, S. ; Yoshitomi, T.; Kitagawa, S.; Ishii, T.; Matsuzaka, H.; Seki, K. Angew. Chem. Int. Ed. 1999 38 140-143. (75) Eubank, J. F., Rational Synthesis Toward the De sign of Functional Metal-Organic Materials. ProQuest / University of South Florida: Tampa, 2008 (76) Chen, B. L.; Ockwig, N. W.; Millward, A. R.; Contreras, D. S.; Yaghi, O. M. Angew. Chem. Int. Ed. 2005 44 4745-4749. (77) Xue, M.; Zhu, G.; Li, Y.; Zhao, X.; Jin, Z.; Kang, E.; Qiu, S. Cryst. Growth Des. 2008 8 2478-2483. (78) Lee, Y. G.; Moon, H. R.; Cheon, Y. E.; Suh, M. P. Angew. Chem. Int. Ed. 2008 47 7741-7745. (79) Lin, X.; Jia, J. H.; Zhao, X. B.; Th omas, K. M.; Blake, A. J.; Walker, G. S.; Champness, N. R.; Hubberstey, P.; Schroder, M. Angew. Chem. Int. Ed. 2006 45 7358-7364. (80) Wang, X. S.; Ma, S. Q.; Rauch, K.; Simmons, J. M.; Yuan, D. Q.; Wang, X. P.; Yildirim, T.; Cole, W. C.; Lopez, J. J.; de Meijere, A.; Zhou, H. C. Chem. Mater. 2008 20 3145-3152. (81) Ma, S.; Sun, D.; Simmons, J. M.; Collier, C. D.; Yuan, D.; Zhou, H.-C. J. Am. Chem. Soc. 2007 130 1012-1016. (82) Blatov, V. A.; Shevchenko, A. P.; Serezhkin, V. N. TOPOS (v. 4.0 Professional) http://www.topos.ssu.samara.ru/versions.html (83) Cavenati, S.; Grande, C. A.; Rodrigues, A. E. E. J. Chem. Eng. Data 2004 49 1095-1101. (84) Llewellyn, P. L.; Bourrelly, S.; Serre C.; Vimont, A.; Daturi, M.; Hamon, L.; De Weireld, G.; Chang, J. S.; Hong, D. Y.; Hwang, Y. K.; Jhung, S. H.; Ferey, G. Langmuir 2008 24 7245-7250. (85) Devic, T.; Serre, C.; Audebran d, N.; Marrot, J. r. m.; Ferey, G. J. Am. Chem. Soc. 2005 127 12788-12789. (86) Chandler, B. D.; Cramb, D. T.; Shimizu, G. K. H. J. Am. Chem. Soc. 2006 128 10403-10412. (87) Pereira, G. A.; Peters, J. A.; Almeida Paz, F. A.; Rocha, J.; Geraldes, C. F. G. C. Inorg. Chem. 2010 49 2969-2974.

PAGE 272

235(88) Black, C. A.; Costa, J. S. n.; Fu, W. T.; Massera, C.; Roubeau, O.; Teat, S. J.; Aromi, G.; Gamez, P.; Reedijk, J. Inorg. Chem. 2009 48 1062-1068. (89) Ma, S.; Yuan, D.; Wang, X.-S.; Zhou, H.-C. Inorg. Chem. 2009 48 2072-2077. (90) Delgado-Friedrichs, O.; O'Keeffe, M.; Yaghi, O. M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2006 62 350-355. (91) Chen, B. L.; Ockwig, N. W.; Froncze k, F. R.; Contreras, D. S.; Yaghi, O. M. Inorg. Chem. 2005 44 181-183. (92) Yang, S.; Lin, X.; Blake, A. J.; Thom as, K. M.; Hubberstey, P.; Champness, N. R.; Schroder, M. Chem. Commun. 2008 6108-6110. (93) Liu, Y.; Eddaoudi, M. unpublished results (94) Dinca, M.; Long, J. R. Angew. Chem. Int. Ed. 2008 47 6766-6779. (95) Nouar, F.; Eckert, J.; Euba nk, J. F.; Forster, P.; Eddaoudi, M. J. Am. Chem. Soc. 2009 131 2864-2870. (96) Ohmura, T.; Usuki, A.; Fukumori K.; Ohta, T.; Ito, M.; Tatsumi, K. Inorganic Chemistry 2006 45 7988-7990. (97) Kitaura, R.; Onoyama, G.; Sakamot o, H.; Matsuda, R.; Noro, S.; Kitagawa, S. Angew. Chem. Int. Ed. 2004 43 2684-2687. (98) Goldberg, I. Chem. Commun. 2005 1243-1254. (99) Kaye, S. S.; Long, J. R. J. Am. Chem. Soc. 2008 130 806-807.

PAGE 273

236 Chapter 4. From Molecular Building Bloc ks to Supermolecular Building Blocks: Highly Connected MOMs that Possess a Sup erior Level of Hierarchical Complexity 4.1. Covalent Cross-Linking of Nanoscale Faceted Polyhedra 4.1.1. Introduction The evolvement of the node-and-spacer approach into the MBB approach is widely recognized as a powerful design st rategy for the ratio nal construction of functional solid-state MOMs ( e.g. metal-organic polyhedra, coordination polymers, metal-organic frameworks, etc ). This is exemplified by the burgeoning academic and industrial interest in th is class of materials.1 An exceptional feature offered by this strategy is that the desired functionality and directionality can be introduced into the inorganic and organic building blocks at th e molecular level, prior to the assembly process, to generate modular MOMs that are uni quely suited for a wide range of pertinent applications (see Chapter 1). The transition from single metal MBBs to enlarged rigid metal clusters with 3-, 4-, and/or 6-connec tivity therefore provided a heightened degree of predictability and more importantly an attractive pathway to target highly porous MOFs with increased surface areas and larger pore volumes. The relative size of metal clusters versus single metal ions has undoubtedly led to unprecedented breakthroughs in terms of sc ale and porosity. It remains however a challenge to absolutely predict which net will form once the MBBs self-assemble. This is because more than one net can often be isol ated from a combination of building blocks.

PAGE 274

237 For example, the self-assembly of tetrahedra l and triangular SBUs has the potential to form augmented networks with ctn or bor topology but other nets are also possible (see Chapter 5).2 Furthermore, metal ions often adhere to more than one type of coordination geometry and thus the corresponding building units can possess different shapes. This is exemplified in the case of indium where by 4-connected single metal ions MBBs or 6connected metal clusters have been observed having tetrahedral and trigonal prismatic building units, respectively a nd therefore MOMs with different topologies are generated (see Chapters 2 and 3).3, 4 Due to the numerous topologi cal possibilitie s for a given combination of building blocks one once must not take the term “desi gn” too literally at this stage. A review article published by Yaghi and co-workers in 2005 addresses the issue of ambiguity by st ating the following: “ We believe that a structure can be truly designed when a specific building block leads to a specific predetermined structure, which is only possible for that building block. Such structures fall into the highest level, which in essence conceives the construction of a net from a building block that codes specifically and only for that net. ”5 Accordingly, nets whose vertex figure s designate the need for high connectivity are ideal targets to serve as nodes for the c onstruction of extended frameworks. This is due, in part, because fewer nets are likely to form based on the combination of complex building blocks and the enhanced structural a nd directional informati on is already built-in to the MBBs to facilitate the formation of a specific network. This higher level of complexity therefore offers greater potentia l toward prediction, de sign, and synthesis of prospective made-to-order MOMs with unpreced ented pore metrics. Such structural features are required for many important app lications, as will be discussed below.

PAGE 275

238 In crystal chemistry the deliberate c onstruction of MOMs fr om prefabricated nodes with high connectivity is less prevalent than basic MBBs. It is an ongoing synthetic challenge to target such nets using traditional MBBs, i.e. with connectivity 8, because the necessary MBBs are often far more comp lex than the connectivity offered by organic ligands and multinuclear clusters. Zaworotko an d co-workers therefor e initiated a clever design strategy in 2004 to overcome the limita tions of scale and connectivity imposed by traditional MBBs and SBUs by engaging in a form of suprasupermolecular chemistry using metal-organic polyhedra (MOP) as nodes. The authors demonstrated the feasibility of employing highly symmetric nano scale metal-organic nanoballs, i.e. small rhombihexahedra ([Cu2( m -BDC)2L2]12), as nodes for the c onstruction of extended networks with intended topologies. The di screte spheroid-lik e nanoballs exhibit Oh symmetry and are assembled from 12 c opper paddlewheel building blocks ( i.e. closed faces) which in turn reveal 8 triangula r and 6 square accessible windows. The prototypical nanoball is theref ore uniquely suited to act as a rigid and directional node because the peripheral sites can be decorated through the axial metal sites and/or the meta position of the m -BDC ligand to yield extended arch itectures. The underlying topology of these nets is therefore governed by whic h face/window facilitates the connection to adjacent nanoballs. The 4-connected diamondoid nets, 6-connected octahedral or tetragonal nets, or 8-connected bcc nets are therefore suitable targets. Substitution of the meta-hydrogen atom on the BDC ligand for a sulfonate group indeed resulted in the formation of decorated anionic nanoballs having 24 sulfonate moieties located at the vertices. The crystal packing of the nanoballs resembles a bcc arrangement sustained by double cross-linking whereby 16 sulfonate moie ties of one nanoball are bridged to an

PAGE 276

239 adjacent nanoball through the axial positions of 16[Cu(methoxypyridine)4]2+ cations.6 This connectivity therefore yields a 3-period ic MOM built-up from discrete MOP (Figure 4.1a-b.). Shortly thereafter, Za worotko et al. generated a pcu arrangement of nanoballs sustained by quadruple covalent cross-linki ng of each of the 24-connected nodes. This was accomplished using a flexible tetraca rboxylate organic ligand, 1,3-bis(5-methoxy1,3-benzene dicarboxylic acid)be nzene (Figure 4.1c-d.).7 Figure 4.1. (a) Ball-and-stick representation of th e sulfonated nanoball; (b) Schematic representation of the bcc packing of nanoballs, which is f acilitated by double cross-linking; (c) Select fragment from pcu net of nanoballs viewed along the b c axes, which is sustained by quadrupule covalent cross-linking to six nanoballs; (d) Schematic to illustrate the pcu packing of nanoballs. The realization that MOP can be employe d as nanoscale nodes to target specific nets represents the next generation of bu ilding blocks, coined supermolecular building (a) (b) (c) (d)

PAGE 277

240 blocks (SBBs) by Eddaoudi and Zaworot ko et al. in 2008. Recall that SBUs are assembled from MBBs and therefore connecti ng the points of extension of SBBs reveals a tertiary building unit (TeBU), which correspond s to the vertex figure of that particular net. Pre-designed SBBs therefore accolade existing design strategies and offer an exceptional degree of hierarchic al structural complexity to construct higher connected and functional 3-perioidic MOMs that are ot herwise unattainable via simple MBBs and SBUs. The SBB approach offers many advantag es over traditional design strategies because the remarkable scale attainable by S BBs guarantees that the relative size of the resultant material will be enhanced since S BBs are considerably larger than their SBU counterparts. Prior to the assembly process the relative cavity dimensions of the SBBs can be controlled and functionaliz ed to ensure that the desire d features are present in the resultant material. This will be, in part, dete rmined by the cross-linker positioned at each of the vertices. The SBBs of ten have narrow accessible windo w apertures that permit the passage of small molecules into the hollow interior micropore cavity, which simultaneously hinders the forma tion of interpen etrated nets. 4.1.1.1. Classification of Metal-Organic Polyhedra Many examples of MOPs having 8, 12, 20, a nd 24 vertices have been reported in the literature. They therefore represent ideal synthetic targets to se rve as nanoscale nodes in the construction of extended nets that indicate the need for higher connectivity, as exemplified above.8, 9 The design and synthesis of a pa rticular subset of MOPs are graciously inspired by a specific class of solids called the Platonic solids The convex

PAGE 278

241 polyhedra are assembled from edge-shari ng regular polygons wherein all faces are congruent polygons that meet at the same angles with the same number of polygons at each vertex (Figure 4.2.).10 They are therefore described as being vertex-, edge-, and face-transitive regular polyhe dra and have the lowest possi ble transitivity of [1111] due to the highest degree of regularity. Figure 4.2. Schematic representation of the five regular Platonic Solids (left to right): Tetrahedron, hexahedron (cube), octahe dron, dodecahedron, and icosahedron. The implementation of design strategies that consistently form supramolecular polygons provided the foundations for the develo pment of more complex structures akin to specific classes of polyhe dra (see Chapter 1). In a sim ilar manner, early reports of MOPs were constructed by adopting a molecular paneling approach and this gave rise to a diverse collection of cage-like compounds ( e.g. molecular tetrahedra, octahedra, etc ). An immense number of metal-organic nanostr uctures have also been constructed from carboxylate-based and heterofunctional orga nic ligands which exhibit similar cagelike characteristics (Figure 4.3.). The windows and accessible interior cavities have the potential to offer unique hostguest chemistry applications and the restricted cavity dimensions of these nanoscale polyhedra have shown to promote chem ical reactions that are otherwise not observed in less confined environments. MOPs possessing peripheral functionalities can be further employed as SBBs to construct unprecedented extended MOMs, as will be discussed below.

PAGE 279

242 Figure 4.3. Examples of MOPs: (a) molecular tetrahedron, [(Fe3O)4(SO4)12(BPDC)6(Py)12]8(IRMOP-51);11 (b) molecular cube, [Ni8(HImDC)12]8(MOC-1);4 and (c) molecular octahedron, [Pd6(2,4,6-tri(4-pyrid yl)-1,3,5-triazene)4(2,2'-BIPY)4]12+.12 Hydrogen atoms and guest molecules are omitted for clarity. Color code: M = green; C = gray; O = red; N = blue; S = yellow. A large number of MOPs on the other hand can be classified as Archimedean in nature because they are assembled from two or more types of polygons that meet at identical vertices as oppose to just one type of face. They are therefore regarded as semiregular vertex-transitive polyhedra (Figure 4.4.).The simplest and most widely reported Archimedean polyhedron is the truncated tetrahedr on, also known as the augmented form of the regular tetrahedron. Many examples of MOPs rela ted to truncated octahedron, cuboctahedron, truncated cuboctahedron and others have also been reported. (a) (b) (c)

PAGE 280

243 Figure 4.4. Schematic representations of the 13 semi-regular Archimedean Solids. Top (left to right): Truncated tetrahedron, truncated octahedron, truncated cube, cuboctahedron, rhombicuboctahedron (small r hombicuboctahedron). Middle (left to right): Truncated cuboctahedron (great rhombi cuboctahedron), snub cube, icosidodecahedron, truncated dodecahedron. Bottom (left to right): Truncated icosahedron (bucky ball), rhombicosidodecahedron (small rhombicosidodecahed ron), truncated icosidodecahedron (great rhombicosidodecahedron, and snub dodecahedron.13, 14 Lastly, a class of polyhedra that ar e derived from both the Platonic and Archimedean solids must be highlighted as they are well-represented in the literature and are particularly salient to the results descri bed in this chapter. These uniform polyhedra are classified as faceted polyhedra because they are sustained by vertex-linking of one or more types of regular polygons which m eet at identical vertices. The polygons comprising this group assemble in such a way so that none of their edges are shared. This is in contrast to the aforementioned pur ely convex polyhedra. Accordingly, polyhedra that are comprised of both closed faces (convex polygons) and open (concave) windows are thereby revealed. There are nine faceted polyhedra ,15 all generated from regular polygonal faces ( e.g. square, triangle, pentagon) and open windows (Figure 4.5.).

PAGE 281

244 Figure 4.5. Nine faceted polyhedra. Top (left to right): Tetrahemihexahedron, cubohemioctahedron, octa hemioctahedron, small rhombihexa hedron, small cubicuboctahedron. Bottom (left to right): Small dodecahemidodecahedron, small icosihemidodecahedron, small dodecaicosidodecahedron, small rhombidodecahedron.14, 16 4.1.2. Results and Discussion A particular group of faceted polyhedra that are assemb led from solely square polygons will be the focus of this section. Th ey are suitable synt hetic targets to be exploited as SBBs because the ubi quitous dimetal tetracarboxylate, [M2(RCO2)4] paddlewheel MBB, defined as a square SBU is readily accessible in crystal chemistry.1724 This is exemplified by the over 1600 crysta l structures reported in the Cambridge Structural Database (CSD) that are built up from this particular MBB. The assembly of vertex-linked square polygons oriented at 90o, 120o, and 144o angles has the potential to afford the following faceted polyhedra, the cubohemioctahedron (12 vertices), small rhombihexahedron (24 vertices), and small rhombidodecaheron (60 vertcies), respectively (Figure 4.6.).15 Employing a suitable spacer moiety to link each of the aforementioned vertices will, in theory, be a governing factor for determining the formation of one polyhedron over the other.

PAGE 282

245 Figure 4.6. Three uniform faceted polyhedra that can be generated via linking of molecular squares only: (left) cubohemioctahedron (middle) small rhombihexahedron (right) small rhombidodecahedron The first examples of MOMs containi ng only vertex-linked square SBUs was based on a molecular small rhombihexahedron also known as the nanoball, truncated cuboctahedron or rhombicuboctahedron (MOP-1). The crystal structure is built up from copper paddlewheel MBBs that link at 120o through the carboxyl ate moieties of m -BDC (see Chapter 1).22, 23 Further studies revealed the outer surface of the prototypical MOP could be fine tuned via decoration of the m -BDC unit and/or the axial metal sites 6, 25-28 to yield multiple derivatives that exhibit unique properties29-32 and molecular weights of at least 6 kDa. To be the best of our knowledge, a molecular form of the small rhombidodecahedron has not yet been synthesized as a discrete or extended entity. We report herein the first example of a MOF rationalized as the assembly of metal-organic cubohemioctahedral SBBs, generated in situ as cuboctahedral TeBUs to construct an edge-transitive 12-connected MOF, as oppose to discrete molecules or ions. The m -BDC derived tetracarboxylate organic ligand, i.e. H4-ABTC, is precisely located at the vertices of the vertex figure and there by permits covalent cros s-linking to adjacent SBBs to generate a 12-connected MOF having fcu topology, i.e. the only quasi-regular net. A (3,24)-connected rht MOF was also targeted usi ng the SBB approach by linking metal-organic small rhombihexahedral SBBs through tritopic hexacarboxylate organic

PAGE 283

246 ligands. This particular SBB was targeted be cause the TeBU corresponds to the requisiste 24-connected rhombicuboctaheral vertex fi gure. The trigonal building units are positioned exactly at the 24 vertices to facilita te the construction of this anticipated net (Figure 4.7.). Figure 4.7. (left) Tiling representation of the (3,24)-connected rht net; and (right) augmentation of the 24and 3-connected nodes reveals rhom bicuboctahedron and tria ngular vertex figures. Note that the hexacarboxylate organic ligands, containing three m -BDC moieties, were judiciously chosen to be analogous to the trigonal m -BDC units generated from the inorganic [Cu3O(N4CR)3] trimer MBBs in the original (3,24)-connected rht net ( i.e. rht 1) as previously reported by our group (Figure 4.8.).33 The data presented in this chapter will attempt to convey that molecular analogs of the fcu and rht nets represent unique prototypical platforms. The cav ities and bulk properties of th e materials can be fine tuned via judicious substitution of the metal ions and organic ligands to yield an isoreticular series of functional MOFs. They are therefor e ideal systems to analyze and correlate a large number of structureproperty relationships ( e.g. gas storage).

PAGE 284

247 Figure 4.8. (left) rht -1 was constructed using a heterof unctional ligand, TZI, to promote the formation of two differe nt inorganic MBBs, [Cu2(RCO2)4] and [Cu3O(N4CR)3]; (right) Isoreticular analogues can also be generate d by substituting the inorganic trimer for hexacarboxylate organic ligands that contains three m -BDC moieties. 4.1.2.1. Structural Analysis: 12-connected fcu nets Reaction of H4-ABTC with Ni(NO3)2 .6H2O in a DMF/HNO3 solution, in the presence of HMTA, yields green homogene ous crystals with a cube-shape morphology. The purity of the as-synthesized compound was confirmed by similarities between the experimental and calculated PXRD patterns (A ppendix B). The microcrystalline material was found to be insoluble in H2O and common organic solvents It was characterized and formulated by single-crystal Xray diffraction studies as [Ni2(ABTC)(H2O)3]n, ( 23 ). The crystal structure is built up from centrosymmetric [Ni6(ABTC)12]12anions and therefore a novel anionic metal-organic cubohemioctahedral SBB is revealed (Figure 4.9.). Each SBB is covalently cross-linked to twelve adjacent SBBs through the meta position of the rigid diisophthalate te tracarboxylate ligand, ABTC4which thereby leads to the formation of a 3-periodic network with fcu topology. A pure phase of the cobalt analogue of 23 24

PAGE 285

248 can also be also be prepar ed under mild solvothermal c onditions via the reaction of H4ABTC with Co(NO3)26H2O to yield homogenous red crys tals with a truncated cubeshape morphology. Figure 4.9. (left) Schematic representation of the cubohemioctahedron assembled from square polygons sustained by a 90o vertex connection; Prototypical metal-organic cubohemioctahedron [M6(bdc)12]12SBB in ball and stick (middle) and space-filling modes (right). Color code: M (Ni, Co) = green; C = gray; O = red. All hydrogen atoms and the decorated metal site have been omitted for clarity. The inorganic MBB in 23 assembled from two octahe drally coordinated Ni(II) cations, adopts a pseudo-paddlewheel conformation. Four indepentent ABTC4ligands coordinate to Ni1 in a monodentate fashi on through the carboxyl ato oxygen atoms along the equatorial plane (dNi – O = 2.02 to 2.12 ; O…Ni…O angle of 89.9o), while the axial positions are occupied by aqua ligands (dNi – O = 2.00 to 2.08 ). Charge balance for the anionic metal-organic cubohemioctahedron is provided by a crystallograpically disordered Ni(II) cation, Ni2, which bridges one of the axial aqua lig ands and two of the four carboxylato moieties (Ni…2-O distance of 1.98 ; dNi – O = 1.48 ) (Figure 4.10.). From a topological perspective, Ni2 is not a point of extension and therefore plays no decisive role in governin g the underlying topology of the network. Since the decorated metal site is symmetrically disordered across the four carboxylato moieties this appears to

PAGE 286

249 facilitate crystallization in the F m -3 space group. Taking into c onsideration the decorated metal site, a cubohemioctahedron SBB of formula M12L6 is thereby revealed. Figure 4.10. Ball and stick representation of the inorganic MBB in 23 : (left) octahedral coordination environments about Ni1 and the symmetrically disordered Ni2 cation; (right) Ni2 metal center shown without disorder. The cubohemioctahedral SBB has inner cav ity dimensions of ca. 8.7 and the relative size of this cavity is essentiall y predetermined by the distance between the carboxylate groups in ABTC4-. Much larger tetrahedral and octahedral cavities are unveiled as a direct consequence of linki ng the nanoscale SBBs (Figure 4.11.). The truncated octahedral cage cons ists of six inorganic MBBs which occupy the vertices of the octahedron, viewed as square fa ces, linked together by twelve ABTC4ligands. The ligand therefore constitutes th e edges of the octahedron and reveals idealized triangular windows proportional to the dimensions of th e ligand. The octahedral cavity is crossshaped and exhibits an internuclear dist ance of 23.3 . Suitable guest molecules can enter this cavity via the acces sible triangular windows that measure approximately 7.4 . The tetrahedral cavity is comprised of smalle r window apertures, 4.9 , which lead into a cavity with an internuclear dist ance of approximately 18.9 .

PAGE 287

250 Figure 4.11. (a) Fragment of the single-crystal structure of 23 comprised of (b) cubohemioctahedral; (c) tetrahedra l; and (d) truncated octahedral cavities. Hydrogen atoms, decorated metal site, and solvent molecules are omitted for clarity. Color code: Ni = green; C = gray; O = red. MOMs are commonly simplified by dissecting the networks into an assembly of n -connected nodes. The frameworks are s ubsequently classified based upon the underlying topology adopted by that particular net. This assessment is critically dependent upon the MBB assignments because this will ultimately determine the points of extension that govern the growth of the ne t. In lieu of this anomaly, most structures can be interpreted in more than one yet equa lly acceptable ways. This often leads to more than one topological possibility, as exemplified by in the case of 23 and 24 One way of rationalizing the topology of 23 is to view the spheroid SBB as a 12-connected node. (a) (b) (b) (c)

PAGE 288

251 Augmentation reveals a cuboctahedron vertex figure which corresponds to the vertex figure of the fcu net.34 Each SBB is linked to 12 adjacent SBBs to generate the observed face-centered cubic, fcu network (Figure 4.12a-b.). Figure 4.12. A topological evaluation of 23 and 24 reveals the structure can be interpreted in two ways: (a) Augmented 12-connected fcu topology; (b) Tiling representation of the fcu net comprised of two types of tiles; (c) Schematic of the augmented nbo network; and (d) Tiling view of the nbo net consisting of one type of tile. An nbo type network on the other hand is reveal ed if the carbon bearing the carboxylato groups are deemed the points of extension. In this case, the inorga nic and organic MBBs are represented as two distinct 4-conn ected nodes which correspond to square and rectangular building units, re spectively. The assembly of M12L6 moieties would therefore (a) (c) (b) (d)

PAGE 289

252 be regarded as the assembly of squares, rectangles, and hexagons, generating cages resembling those in the zeolite structure sodalite, SOD (Figure 4.12c-d.). 4.1.2.2.1. Properties Both the enlarged cavities and pot ential open metal sites exhibited by 23 and 24 encouraged us to investigate their H2 uptake capacity. We found that 23 can store 0.6 wt% of H2 at 77 K and 1 atm, whereas, 24 could not be activated despite trying a variety of protocols ( e.g. solvent exchange and supercritical CO2 activation). We also carried out preliminary spectroscopic studies to inves tigate the feasibility of exchanging the decorated SBB metal site for different metal cations. The UV-Vis spectra however did not show any evidence of exchange as no sh ift in the maxima wa s observed between the exchanged samples as compared to the parent compound. It is noteworthy to mention that Zaworotko and co-workers reported the synthesis of an isoreticular fcu framework builtup from cubohemioctahedral SBBs that were covalently cross-linked using an expanded tetracarboxylato ligand, BIPATC4-.35 The relative span of the ligand facilitated the formation of enlarged cavities (> 3nm) comprised of a similar inorganic MBB (M = Co or Ni) as 23 which did readily undergo full metal exchange for Cu. 4.1.2.2. Structural and Topological Analysis : (3,24)-connected rht nets from Hexatopic Organic Ligands Mild solvothermal reaction of H6-PTMOI with Cu(NO3)22.5H2O in a DMF/NMP mixture yields a blue homogenous microcrystal line material in 86% yield with truncated octahedron morphology. The purity of the ma terial was confirmed by similarities between calculated and experimental PXRD patterns (Appendix B). The as-synthesized

PAGE 290

253 compound, which is insoluble in H2O and common organic solvents, was characterized and formulated by single-crystal X-ray diffraction studies as {[Cu3(PTMOI)(H2O)3]}n, ( 25 ). Zinc, cobalt, and manganese analogs of 25 26 – 28 were also successfully isolated in crystalline forms under similar reaction condi tions and therefore support the versatility of the approach as a function of varying the metal cation. Compounds 26 – 28 are insoluble in common organic solvents; however the crystals rapidly degrade in air and aqueous environments. For that reason, th e purity of the bulk samples could not be confirmed. SCD data collected on several cr ystals harvested from independent vials confirmed compounds 25 – 28 are indeed isostructural frameworks with rht topology. The only structural difference between the af orementioned networks is the metal ion and therefore the following structural description will pertain to 25 Compound 25 crystallizes in the cubic F m -3 m space group with a = 41.4786(3) and a unit cell volume of 71,362.9(9) 3. The crystal structure exhibits a 3-periodic network built up from copper paddlewheel M BBs, whereby each copper ion exhibits the expected square pyramidal geometry. The equatorial plane is occupied by four independent carboxylate oxygen atoms that brid ge the dinuclear copper ions in a bismonodentate fashion, while an aqua lig and lies in the axial position, CuO5. The natural 120o angle subtended by the three angular m -BDC units of the C3-symmetric hexatopic ligand, PTMOI6-, facilitates the assembly of vert ex-linked molecular squares (Figure 4.13top.).

PAGE 291

254 Figure 4.13. Select fragments from the crystal structure and schematic representation of 25 : Top. m -BDC naturally accommodates the 120o angle needed to generate metal-organic truncated cuboctahedral SBBs and rhombic uboctahedral TeBUs. Bottom. Cr oss-linking of the nanoscale polyehdra is facilitated via decoration at the 5-position of the m -BDC unit by employing PTMOI6to afford a (3,24)-connected MOF with rht topology. Hydrogen atoms and solvent molecules are omitted for clarity. Color code: Cu = green; C = gray; O = red; the 5-position is highlighted in orange. Twelve [Cu2(RCO2)4] paddlewheel MBBs assemble in a cis fashion to generate the anticipated neutral metal-organic SBBs with 24-vertices, which are generated in situ as rhombicuboctahedral tertiary bu ilding units (TeBUs). Note th at since the SBB is indeed comprised of both open windows and closed fa ces, it can be represented as a nanoscale small rhombihexahedra consisting of 12 vert ex-linked, not edge shared, square SBUs (Figure 4.13middle.). Each SBB is covalently cross-linked to twelve adjacent SBBs

PAGE 292

255 through the triconnected coplanar isophtha late BDC units of the hexatopic ligand, regarded as a triangular build ing unit (Figure 4.13bottom.). This results in an edgetransitive (3,24)-connected MOF with rht topology. This topology was in fact only recently reported by Delgado-Friedrichs and O’Keeffe to be the sole edge transitive net for the assembly of rhombic uboctahedral and trigonal BUs.36 The overall neutral framework is compri sed of three distinct types of open polyhedral cages (Figure 4.14.). Covalent cross-linking of the SBBs (cage A) through PTMOI6permits the formation of tetrahedral and oc tahedral cavities referred to as cages B and C, respectively. The largest cage, C, is defined by 8 triangular organic MBBs and 24 [Cu2(RCO2)4] inorganic MBBs. The inner cavity di mensions of this cage are large enough to accommodate a van der Waals s phere measuring approximately 16.624 and it is surrounded by 6 truncated cubo ctahedra cages and 8 tetrahedrallike cages. The truncated cuboctahedral cag e, comprised of 12 [Cu(RCO2)4] MBBs, exhibits an internuclear distance of a pproximately 13.069 and is surr ounded by six large cages and eight tetrahedrallike cages. While the tetrahedral-like cage can accommodate a van der Waals sphere measuring approximately 10.460 and is assembled from 12 [Cu(RCO2)4] MBBs and 4 PTMOI6ligands; consequently, each of these cages are bordered by four truncated cuboctahedra and f our of the largest cages.

PAGE 293

256 Figure 4.14. Top. (left to right): Ball-and-stick and tili ng representation of the cages C, A, and B, respectively in 25 Bottom (left to right): Tiling representation showing the assembly of the smaller polyhedral cages around the largest cage in the rht net. Color code: Cu = green; C = gray; O = red. Hydrogen atoms and solvent molecules are omitted for clarity. The results reported so far in this se ction have addressed the modularity and predictability of the SBB appro ach by showing that isostructural rht analogues can indeed be readily obtained under similar reaction conditions by merely replacing the copper ions with different dicationic transition metals such as zinc, cobalt, and manganese. MOFs with this particular connec tivity have demonstrated great promise as potential hydrogen storage systems because of the unique structural features of the rht network, which include (1) high surface area (2) large free volume, (3) low framework density, (4) potential open metal sites and (5) large open windows and cavities. Once proper activation procedures ha ve been developed, compounds 25 – 28 are expected to be an ideal series of materials to analyze and correlate the contribution of potential open metals sites on framework dihydrogen interactions.

PAGE 294

257 Isoreticular rht MOFs assembled from expanded tritopic organic ligands can therefore be targeted to achieve even hi gher surface areas and larger pore volumes. An elongated C3-symmetric hexacarboxylate organic ligand, H6-ABPTMOI, was therefore designed to mimic the connec tivity and geometry of H6-PTMOI. We predicted that each of the 4-connected square planar [M2(RCO2)4] nodes would assemble with each of the m BDC units whereby one ABPTMOI6ligand is shared be tween three truncated cuboctahedra to generate a trigonal 3-connected node with an estimated edge length of 24.6 (Figure 4.15.). Figure 4.15. Tritopic hexacarboxylate organic ligands (L6-) employed in this study: (left) PTMOI6spans ~13.6 , these dimensions can be substantially increased to ~24.6 as demonstrated in ABPTMOI6(right). MOFs constructed from flexible elon gated organic ligands commonly form interpenetrated networks but porous MOFs cons tructed from SBBs are in fact less likely to exhibit interpenetration when using such ligands due the relative window dimensions of the SBB. Accordingly, solvothermal reaction of H6-ABPTMOI and Zn(NO3)26H2O in a DMF/NMP solution yields a orange homoge nous microcrystalline material in 81% yield with a truncated octahedron morphology, ( 29 ). The connectivity of 29 is the same as that observed in 25 28 but ligand expansion permits th e formation of much larger

PAGE 295

258 tetrahedral and octahedral cages (Figur e 4.16.). The diameter and windows of the truncated cuboctahedral cage are relativel y consistent for all the compounds ( i.e. ~13 ) because the dimensions are delimited by the scale of the inorganic MBBs and the 120o angle between the m -BDC units. Figure 4.16. Ball-and-stick representation of two of the three cages in 29 : (a) Tetrahedrallike cage can accommodate a van der Waals sphere th at measures approximately 17.138 ; and (b) Largest cage is estimated to 27.199 in diameter Color code: Zn = green; C = gray; O = red; N = blue. Hydrogen atoms and solvent molecules are omitted for clarity. As exemplified in the previous section of this chapter, the topological assessment of a net can come in many flavors. Likewi se, the present isoreticular frameworks, 25 – 29 can also be interpreted as novel (3,3,4)-connected ternary net based on the assembly of three different building units to afford a trinodal n -connected net (Figure 4.17.).

PAGE 296

259 Figure 4.17. (left) Novel topological assessment of 25-29 rationalized as a trinodal ternary net and its corresponding augmented conformation (ri ght). Topological Terms for Node 1: VS = 6.6.8.8.8(2).8(2); CS = 4,8,18, 29, 52, 61, 106, 120, 170, 187; TD10 = 755. Topological Terms for Node 2: VS = 6.8(3).8(3); CS = 3, 8, 15, 29, 40, 69, 81, 131, 146, 206; TD10 = 728. Topological Terms for Node 2: VS = 8(3).8(3).8(3); CS = 3, 6, 18, 24, 42, 63, 84, 93, 183, 175; TD10 = 691. The 4-connected paddlewheel MBBs ( i.e. square vertex figure) are linked to four trigonal nodes ( i.e. three m -BDC units) which constitute one of the 3-connected nodes. Three of these nodes surround the central benzene ring of the ligand at 120o, which delimits the second 3-connected vertex. The connectivity of the decorated nets of 25-29 can also be rationalized as related to Zeolite A with lta topology; that is, assembled from and -cages and d4r or the reo e net comprised of rhombicuboctahedron, cuboctahedron, and d4Rs.34, 37

PAGE 297

260 4.1.2.2.1. Gas Sorption Measurements for Compound 25 The total solvent-accessible volume calculated for 25 is estimated to be 74 % of the unit cell volume38 and the calculated density for the desolvated framework was found to be 0.615 g cm-3. The high free volume and low density paired with the unique structural features of this framework, i. e. the large accessible windows leading to the large open cavities with potential open meta l sites therefore prompted us to investigate the gas sorption properties of this MOF. The sorption data is compared to the copper TZI ( rht -1) MOF which exhibits a hi gher framework density, larger cavities, and potentially a higher concentration of open metal sites. This was done as a comparative study to evaluate the effect of thes e structural parameters on H2 uptake and isosteric heat of adsorption. Gas sorption investigations were carried out on the fully evacuated sample after exchanging the crystals in methanol for a pe riod of 72 h. The activated blue crystalline material was loaded into a 6mm sample cell (slightly wet), initially evacuated at room temperature for a period of 6 h, then gradually heated to 95oC for 12 h and lastly heated to 115oC and held for 6 h. The permanent porosity of 25 was confirmed by the argon sorption isotherm collected at 87 K which re veals a pseudotype I isotherm (Figure 4.18.). The overall shape of the isotherm is ch aracteristic of a mate rial containing both micropores and larger pores that approach the mesoporous range. This is in accord with the calculated dimensions of cages A – C.

PAGE 298

261 Figure 4.18. Argon sorption isotherm for 25 collected at 87 K. The smaller cages undergo micropore filling at lower pressures, i.e. below 0.01, while a second slope appears in the range of 0.01 to 0.1 atm. This is due to gradual mesopore filling of the largest cage but note that saturation is still not achieved at 1 atm. The small hysteresis loop that is apparent in the desorption isotherm implies that the adsorbate cannot be removed from the framew ork as readily as it enters and therefore stronger adsorbate-MOF binding sites are likely responsible. This type of behavior is not observed in rht -1 and hexatopic organic-based rht analogs reported previously by other groups.39-41 It is unlikely that the hysteresis is due to the copper metal sites because the aforementioned analogues are also comprised of the same [Cu2(RCO2)4] inorganic MBBs. It is plausible however that u nder vacuum the seemingly rigid PTMOI6hexatopic ligand undergoes a structural change around the methoxy gr oups whereby the relative size/shape of the windows and cav ities is altered and as a resu lt hinders the desorption of the adsorbate. Note that this is purely specu lation and further experi ments are required to validate this hypothesis.

PAGE 299

262 The apparent BET and Langmuir surface areas of 25 were estimated to be 1688 and 1105 m2/g, respectively. The corresponding total pore volume was estimated to be 0.572 cm3 g-1. The experimental data is not however in agreement with the calculated values, i.e. half of what we anticipate. This implies that either the sample was not fully activated or the framework partially collapses under vacuum. The calculated pore volume of 25 is estimated to be 1.20 cm3 g-1, while the anticipated BET surface area computed for this compound was found to be 3601 m2/g. The CO2 sorption capabilities of 25 were tested on the methanol exchanged sample following evacuation at 115oC for 6 h. The permanent porosity of the material was once again confirmed by the pseudo-t ype I adsorption isotherms which were collected at 0oC and 25oC (Figure 4.19.). The overall shap es of the isotherms reveal a steep rise in the lower pressure regio n. This is due to stronger adsorbate-MOF interactions which may be associated with the potential open copper sites. The maximum CO2 uptake capacity of 25 at 25 atm is 42 and 79 wt% at 25oC and 0oC, respectively. The large accessible cavities and acces sible windows permit saturation at higher pressures in the 25oC isotherm but this is not observed at 0oC. A hysteresis is observed upon desorption in both isotherms, and is particular ly apparent in the isotherm collected at 0oC. This could be attributed to framework flexib ility but as previously stated this is only speculation. A second slope is observed for pr essures between 7 to 10 atm and 10 to 16 atm in the 0oC and 25oC isotherm, respectively which is similar to that observed in the argon isotherm at 87 K ( i.e. mesopore filling).

PAGE 300

263 Figure 4.19. (a) CO2 sorption isotherms measured at 0oC and 25oC on compound 25 after exchange in methanol; and (b) CO2 sorption isotherm measured at 25oC to emphasize the steep rise at low loading and filling in the largest cag es that approach the mesoporous range at higher pressures. Interestingly, the second slope does not appe ar in the same pressure range in the two isotherms. In order to provide an accurate assessment of this behavior, a more informative full sorption study must be comp leted. Meanwhile, I speculate that at 0oC and higher pressures the framework becomes flexible in nature which would further support the presence of the third slope obser ved for pressures between 20 to 25 atm. The pronounced hysteresis loop remains constant during desorption from 25 atm to approximately 13 atm and decreases steadily un til it closes at approximately 2 atm. The sorption studies report ed herein concerning the CO2 storage capacity of 25 shows great promise for the rht platform as a potential CO2 storage medium. This compound is however subjected to an unidentif ied pressure and/or temperature dependent structural change which we have not observed in similar materials (at 25oC). This highlights the importance of generating a nd identifying highly favorable sites of interaction between the gas and material to establish pertinent structure-function (a) (b)

PAGE 301

264 relationships. Ongoing studies are directed towards invest igating the reason for the unexpected CO2 sorption behavior. The hydrogen uptake capacity for 25 evaluated at 77 K a nd 87 K at atmospheric pressures, revealed that it can adsorb 1.9 wt% of H2 (Figure 4.20a.). The isosteric heat of adsorption, up to 1.4 % loading of H2 per sorbent weight, has an estimated value of 9.7 kJ mol-1 at low loading (Figur e 4.20b.). Note that the Q st for 25 is appreciably larger than that observed for porous carbon materials and higher than most neutral MOFs. Nevertheless, the heat of adsorption is not ma intained at higher loadings, evidenced by the decrease to 6.0 kJ mol-1. This is in accord with pore f illing of the larger cavities at higher loading. Figure 4.20. Sorption measurements for 25 following activation in methanol: (a) Hydrogen sorption isotherms at 77 K and 87 K; and (b) Isosteric heat of adsorption for H2. We proceeded to try alternative ac tivation protocols to fully activate 25 since the sorption data was much lower than we an ticipated. Accordingly, the as-synthesized crystals were exchanged in a variety of other solvents such as: EtOH, CH3CN, CHCl3, CH2Cl2, THF, and acetone. Despite repeated at tempts, the estimated BET surface areas (a) (b)

PAGE 302

265 did not exceed the original values ( i.e. MeOH exchange). It is a well known fact that open MOFs are often plagued by activation probl ems. Confronted with these challenges, I turned to supercritical dr ying (ScD) in the form of CO2 activation as an alternative approach (see Chapter 1). This form of samp le activation causes less surface tension than traditional thermal evacuation techniques and th erefore is more likely to preserve the structural integrity of the MOF.42 The as-synthesized crystals were excha nged in anhydrous methanol for 7 days, transferred to a plastic holder and placed in the CO2 activation chamber. The blue microcrystalline material was then repeatedly washed with liquid CO2, taken to its critical point (T = 31oC, P = 73 atm), held for 45 min, and finally set to vent/bleed out slowly over a 12 h period. The deep purpl e crystals (dry) we re loaded into a 6 mm sample cell, evacuated at room temperature for 12 h and heated to 115oC for 6 h. The argon isotherm exhibits similar behavior as determined for the MeOH exchanged sample and thus the apparent BET and Langmuir su rface areas are 1676 and 1088 m2/g, respectively with a corresponding total pore volume of only 0.56 cm3 g-1 (Figure 4.21a.).

PAGE 303

266 Figure 4.21. Sorption data collected on 25 after supercritical CO2 activation: (a) Argon sorption isotherm at 87 K; (b) Hydrogen sorption isot herms at 77 K and 87 K; and (c) Comparison between the isosteric heats of adsorption for 25 and rht -1. The maximum H2 uptake for 25 at 77 K and atmospheric pressures was determined to be 1.87 wt% with an estimated Q st of 9.5 kJ mol-1 at low loadings (Figure 4.21b-c.). The ScD CO2 activation method there did not enhance the sorption properties of this material. In fact, it was virtually identical to that of the methanol exchanged sample. A comparison between the isos teric heats of adsorption between 25 and the (a) (b) (c)

PAGE 304

267 original Cu-TZI rht MOF ( rht -1) reveals similar behavior at low loading, i.e. approximately 9.5 kJ mol-1 for both compounds. At highe r loadings however the Q st decreases more rapidly in rht -1. This is perhaps driven by the larger dimensions of the tetrahedrallike and largest cavity. It is noteworthy to mention that this is somewhat counterintuitive because rht -1 is comprised of two inorga nic MBBs and therefore has a higher concentration of poten tial open-copper binding centers A possible explanation to account for this unexpected result may be that some of th e copper sites in rht -1 are still saturated and thus higher evacuation temperat ures are needed to fully desolvate the copper centers. Infrared spectroscopy and/or low-temperature powder neutron diffraction studies should be conducted on both compounds in order to prove or disprove the evidence of H2 binding to the copper centers, as wa s first evidenced in the case of HKUST-1, soon followed by other MOFs (see Chapter 1).A more accurate comparison concerning the expected binding affinities obs erved in these and related materials could then be conducted. It is impor tant to mention that compounds 26 – 29 were subjected to gas sorption measurements but all exhibited poor sorption behavior; this is the surface area values never exceeded 80 m2/g for any of the compounds. This could be attributed to inadequate sample activation or the di metal paddlewheel MBBs comprised of Mn2+, Co2+, and Zn2+ are unstable following the removal of the axial ligands; whereas, dicopper complexes are more readily stabilized by the Jahn-Teller effect. In summary, this section has a ddressed the unique traits of the rht network and highlighted the potential benef its this platform offers to reticular chemistry and gas storage applications. That is, larger surface areas and free volumes can be easily achieved via ligand expansion while avoiding interpen etrated nets. Optimal activation of these

PAGE 305

268 materials however poses a significant challe nge and must be overcome but in principle such values are attainable. 4.1.3. Experimental Section 4.1.3.1. Materials and Methods All materials and methods are described in Chapter 2, unless otherwise noted. The azo-derived hexatopic orga nic ligand, 1,3,5-tris(5'-[( E )-(p-phenyloxy)diazenyl]benzene1,3-dicarboxylic acid) benzene (H6-ABPMTOI) was prepared by Mohamed Alkordi. For clarification purposes, the orig inal synthesis of compounds 25 and 26 was carried out by an undergraduate student in the Eddaoudi gr oup (Matthew Hight) but pure phases could not be isolated. Following his graduation, I started working on this project and purified these compounds and later synthesized 27 and 28 under similar reaction conditions. 4.1.3.2. Synthesis and Characterization Figure 4.22. Synthetic strategy followed for the preparation of 1,3,5-tris(5-methoxy-1,3-benzene dicarboxylic acid)benzene (H6-PTMOI). Reagents and Conditions: (i) DMF / KI / K2CO3 / = 100oC for 1h, (ii) add H2O and filter, (iii) NaOH / H2O / MeOH / = 50oC for 12h / HCl, pH = 1.

PAGE 306

269 Preparation of 1,3,5-tris(5-methoxy-1,3benzene dicarboxylic acid)benzene methyl ester: The product was synthesized according to a modified procedure from the literature.43 In a typical reaction, (12.7g, 0.06mol) of dimethyl-5-hydroxyisophthalate was dissolved in 250mL of DMF. A catalytic amount of KI (74mg, 4.4mol) was added to the solution followed by (26g, 0.26mol) K2CO3. The solution was then heated to 100oC for 1 hour. Subsequently, (2.88g, 0.008mol) of 1,3,5-tr is(bromomethyl)benzene was dissolved in 10mL of DMF and added dropwise to th e mixture and allowed to heat at 100oC for an additional hour and then cool ed to room temperature. Approximately 800mL of water was added to the solution and the resulting white precipitate was filtered, washed thoroughly with ice cold water, and air-dried. Note: No further purification is required. Preparation of 1,3,5-tris(5-methoxy-1,3benzene dicarboxylic acid)benzene, H6-PTMOI: In a typical hydrolysis reaction, the me thyl ester product isolated from the previous step was added to a round botto m flask containing 250mL of methanol. Subsequently, an aqueous Na OH solution (6g in 60mL H2O) was added to this mixture and then refluxed at 50oC for 12 hours. The solution was aci dified to pH = 1 using 15M HCl. The precipitate was separated by filtration, washed with cold H2O, and air-dried. Synthesis of {[Ni2(ABTC)(H2O)3]}n, (23). A solution of Ni(NO3)2 .6H2O (15mg, 0.05mmol), H4-ABTC (9.24mg, 0.025mmol), DMF (1 mL), HMTA (0.1mL, 1M in H2O), and HNO3 (0.4mL, 3.5M in DMF) wa s prepared in a 20 mL scintillation vial. The solution was heated to 85oC increasing at a rate of 1.5oC/minute, held for 12 hours, and cooled to room temper ature at a rate of 1.0oC. Green cube-shaped crystal of 23 which were found to be insoluble in H2O and common organic solvents were harvested and airdried.

PAGE 307

270 Synthesis of {[Co2(ABTC)(H2O)3]}n, (24). A solution of Co(NO3)2 .6H2O (29.1mg, 0.1mmol), H4-ABTC (9.23mg, 0.025mmol), DMF (1mL), ethanol (1.5mL), H2O (0.25mL), and HNO3 (0.3mL, 3.5M in DMF) was prep ared in a 20 mL scintillation vial. The solution was heated to 85oC increasing at a rate of 1.5oC/minute, held for 12 hours, and cooled to room te mperature at a rate of 1.0oC. Red cube-shaped crystal of 24 which were found to be insoluble in H2O and common organic solvents, were harvested and air-dried. Synthesis of {[Cu3(PTMOI)(H2O)3]}n, (25). A solution of Cu(NO3)3 .2.5H2O (19.5mg, 0.084mmol), H6-PTMOI (37mg, 0.056mmol), DMF (1mL), and H2O (1mL) was prepared in a 20 mL scintillatio n vial. The solution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours and 115oC for 23h hours. Each of the heating cycles were increased at rate of 1.5oC/minute and cooled to room temperature at a rate of 1.0oC/minute. Green polyhedral crystals of 25 were harvested and air-dried. The as-synthesized material was determined to be insoluble in H2O and common organic solvents. Synthesis of {[Zn3(PTMOI)(H2O)3]}n, (26). A solution of Zn(NO3)3 .6H2O (12.5mg, 0.042mmol), H6-PTMOI (18.5mg, 0.028mmol), DMF (1mL), and NMP (1mL) was prepared in a 20 mL scintillatio n vial. The solution was heated to 85oC for 12 hours and cooled to room temperature at a rate of 1.0oC/minute. Colorless polyhedral crystals of 26 were harvested and air-dried. The as-synthesized material is insoluble in H2O and common organic solvents. Synthesis of {[Mn3(PTMOI)(H2O)3]}n, (27). A solution of Mn(NO3)2 .x H2O (35.8mg, 0.2mmol), H6-PTMOI (37mg, 0.056mmol), DMF (1mL) and NMP (1mL) were

PAGE 308

271 combined in a 20 mL scintillation vial and heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours. Both heating cycles were increased at rate of 1.5oC/minute and cooled to room temperature at a rate of 1.0oC/minute. Colorless polyhedral crystals of 27 were harvested and air-dried. The as-synthesized material is insoluble in H2O and common organic solvents. Synthesis of {[Co3(PTMOI)(H2O)3]}n, (28). A solution of Co(NO3)2 .6H2O (24.4mg, 0.084mmol), H6-PTMOI (37mg, 0.056mmol), DMF (1mL), and NMP (0.5mL) was prepared in a 20 mL scintillatio n vial. The solution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours and subsequently cooled to room temperature at a rate of 1.0oC/minute. Red cube-shaped crystal of 28 which were found to be insoluble in H2O and common organic solvents, were harvested and air-dried. Synthesis of {[Zn3(ABPTMOI)(H2O)3]n, (29). A solution of Zn(NO3)2 .6H2O (12.5mg, 0.042mmol), H6-ABPTMOI (27mg, 0.028mmol), DMF (1mL) and NMP (1mL) was prepared in a 20 mL scintillatio n vial. The solution was heated to 85oC for 12 hours followed by heating to 105oC for 23 hours heating increasing at rate of 1.5oC/minute and cooled to room temper ature at a rate of 1.0oC/minute. Orange cube-shaped crystal of 29 which were found to be insoluble in H2O and common organic solvents, were harvested and air-dried. 4.2. Rational Directed Assembly of Finite Metal-Organic Cubes (MOCs): A Viable Pathway to Target Zeolitelike Metal-Organic Frameworks (ZMOFs) 4.2.1. Introduction Zeolites are typically defined as pur ely inorganic aluminosilicate microporous crystalline materials. They are co nstructed from corner-sharing [AlO4]5and [SiO4]4

PAGE 309

272 tetrahedra bridged at an a pproximate T-O-T angle of 145o (Figure 4.23.). The incorporation of a trivalent metal ion ( e.g. Al3+) into the silicate network generates anionic frameworks whereby additional meta l ions and/or cationic organic molecules reside within the cavities/channels for charge balance. The assembly of simple tetrahedral building units (TBUs) generates a diverse ra nge of open and robust 3-periodic structures. They may possess large cavities and unior multi-dimensional pore systems or channels, as exemplified by the extensiv e number of different framewor ks (over 170) recognized by the IZA.37 Figure 4.23. Ball and stick representation of corner-sharing [AlO4]5and [SiO4]4tetrahedra bridged via O2at a T-O-T angle of ~145o. Color Code: Al = green; Si = yellow; O = red. The frameworks are differentiated based on a number of factors such as cage dimensions, channel system, ring sizes, framework and topological densities, etc Many synthetic protocols, therefore, rely on the use of an assortment of structure directing agents ( e.g. alkyl ammonium salts) to provide charge balance as well as a certain degree of control over the resultan t pore dimensions. The cationi c templates are often readily exchanged through post-synthetic techniques for an assortment of guest molecules and/or metal ions. Zeolites occur in a hydrated form and exhibit varying degr ees of reversible dehydration owing to their exceptional thermal stability, as indicated by their name which translates into “boiling stone”. This unique class of compounds comes in two flavors

PAGE 310

273 namely naturally occurring and synthetic st ructures. Over 40 natu rally occurring zeolite frameworks are known;37 however, their use in commercial and industrial applications is often limited because they are rarely isol ated in a pure phase and obtaining bulk quantities can be challenging. An alternative strategy implemented to overcome some of these limitations is focused upon the design a nd synthesis of topologically equivalent synthetic zeolites that possess unique properties akin to the natural an alogues. In addition, this approach offers feasible synthetic path ways to construct a pl ethora of novel zeolite structures and/or hypothe tical zeolite structures44 with intrinsic features that have the potential to address pertinent scient ific and societal needs (Figure 4.24.). Figure 4.24. Schematic illustration of some naturally occurring and synthetic zeolite frameworks which exhibit versatile pore structures that rende r them suitable materials for many commercial and industrial applications: (left) Faujasite, FAU (middle) Linde Type A, LTA ; and (right) ZSM5, MFI Zeolites are the largest class of commercially av ailable functional porous materials. Specific applications include, but are not limited to, ion-exchange ( e.g. water softening and purification), shape-selective catalysis ( e.g. petrochemical cracking), drug delivery and separation and removal of gases and solvents.45-55 A severe limitation, which impedes the development and applicability of purely inorganic zeolite s, is the inherent lack of organic functionality. Generally speak ing, this restricts th e pore dimensions and

PAGE 311

274 makes it difficult to extend th e size beyond the “1 nm prison”.56 Because of this shortcoming, zeolites have proven to be uns uitable for some applications, such as molecular magnetism, hydrogen storage ( due to high framework density), and separation/catalysis/encapsulation of large molecules. In recent years, our group and othe rs have strived to develop rational methodologies for the deliberate construction of functional MOMs that possess zeolitic topologies with fine tunable properties that exceed the boundar ies of traditional inorganic zeolites.57-65 Zeolitelike metal-organic frameworks (ZMO Fs) represent a perfect merger between two classes of materials, namely MO Fs and zeolites. The increased tunability and functionality, along with the porosity a nd facile preparation methods of MOFs are combined with the typical an ionic character exhibited by th eir zeolite analogues. Like their counterparts ZMOFs lack interpenetrati on but do contain extra-large cavities which renders them ideal platforms for cationic ex change and encapsulation of large organic molecules and molecular complexes. ZMOFs ar e an emerging class of materials that are attractive for a much wider range of potential applications, such as but not limited to, sensing, ion-exchange, catalysis, magne tism, and gas storage/sequestration. Rational synthesis of MOMs with zeoli tic topologies remains a significant scientific challenge because the assembly of non-directional and flexible TBUs in combination with flexible ditopic organic ligands predominantly yields MOMs having the default cubic diamond topology ( dia ).34 Our group has demonstrated that non-default structures, such as zeolites, can be targ eted by employing the si ngle-metal-ion-based MBB approach.57, 58, 65 Accordingly, this strategy perm its the formation of the necessary MBBs ( e.g. MN4O4, MN4O2) which are regarded as rigid and directional TBUs via the

PAGE 312

275 combination of angular hetero functional organic ligands ( e.g. H3ImDC) with 6 or 8coordinated metal ions. In the construction of ZMOFs, the angular ligands are judiciously chosen so that the nitrogen atom is part of the aromatic ring to direct the framework topology and the carboxylate mo ieties are located in the -position relative to the nitrogen atom to sustain the rigidity of the resultant framework by means of the formation of rigid heterochelated five -membered rings (Figure 4.25.). Figure 4.25. An example of a ZMOF constructed us ing the single-metal-ion-based MBB approach: (right) Fragment of rho -ZMOF shown in ball and stick and node-and-spacer representations; composed of (left) InN4O4 MBBs which can be viewed as InN4 TBUs. Color code: In = green; C = gray; N = blue; O = red.57 A particular subset of zeolite nets possessing LTA AST ACO and ASV topologies share a common composite building unit comprised of eight tetrahedra linked together in a cubelike arrangement referred to as a double 4-ring, d4R.37 A d4R is analogous to a specific type of Platonic solid, namely a hexahedron (cube) and thus gives rise to a highly symmetric metal-organic cube (MOC). As previously reported by us, finite and rigid MOP possessing appropriately positioned peripheral functionalities can be further employed as SBBs to construct ex tended MOFs. The ability to control the connectivity between SBBs offers the potential for structural diversity in the sense that

PAGE 313

276 MOCs can be connected via the edge and/or vertex in a lin ear or tetrahedral manner to generate an assortment of structures. Furt hermore, the modular nature of MOMs permits functionalization of the SBBs ( e.g. interior and/or exterio r) whereby topologically identical or novel architect ures can be constructed. The single-metal-ion-based MBB approach has been shown by our group to be a successful design strategy to control the coor dination number and thus geometry to afford discrete MOCs via the self-assembly of fac -MN3(CO2)3 MBBs with a ditopic heterofunctional organic ligand (Figure 4.26a-b.).4, 66 The peripheral carboxylate oxygen atoms embedded into the MOC, via the liga nd, have the potential to participate in hydrogen bonding interactions with or wit hout metal-coordination with neighboring cubes.67, 68 Thus, MOCs can be employed as SBBs for the directed assembly and deliberate construction of MOMs based on edge -transitive 8-connected nets and having zeolitelike topologies (Figure 4.26c-d.). MOCs linke d at the vertex in a linear fashion could afford ZMOFs having LTA and ASV like topologies; while a vertex-connection via tetrahedral nodes has the potential to form ACO and AST like topologies.

PAGE 314

277 Figure 4.26. Examples of ZMOFs closely related to edge -transitive 8-connected nets synthesized using H3ImDC as an organic linker: (a) Single metal-ion-based fac -MN3(CO2)3 MBBs facilitate the assembly of (b) Finite MOCs which can be u tilized as 8-connected SBBs to generate (c) Left to right: LTA -, ACO -, and AST like topologies which are closely related to the reo -, bcu -, and flu like edge-transitive nets, respectively. The aforementioned zeolite nets, i.e. LTA ASV ACO and AST are particularly interesting in reticular chemistry as their nets correspond to the augmented conformation of the edge-transitive 8-connected nets reo scu bcu and flu respectively. To be more specific, reo is a semi-regular binodal net; bcu is a regular net and the default net for the (a) (b) (c)

PAGE 315

278 assembly of cubelike vertex figures; scu and flu are both (4,8)-connected nets, as outlined in the introductory chapter of this dissertation.69 The construction of functional solid state crystalline materials with just one type of edge makes life much easier for the synthetic chemist as there are fewer experiment al parameters to take into consideration. This subclass of nets are highly feasible targets, since existing or predicted edgetransitive nets can serve as blueprints fo r the generation of a plethora of compounds having the same topology but with a different chemical composition. Herein, I will demonstrate the util ization of MOP, MOCs, as SBBs by reporting the deliberate de sign and construction of an astZMOF sustained by vertexand edge-connected SBBs re lated to the (4,8)-connected edge-transitive flu like net. 4.2.2. Results and Discussion Solvothermal reaction of H3ImDC with Mn(NO3)2 .6H2O in a DMF / CH3CN solution, in the presence of guanidine nitrate, yields homogeneous red polyhedral crystals. The purity of the microcrystalline material was confirmed by similarities between the calculated and experimental PXRD patterns (Appendix B). The assynthesized compound, which was determined to be insoluble in H2O and common organic solvents, was characterized a nd formulated using SCD studies as {[Mn8(ImDC)8(HImDC)4]Mn4 (DMF)8(H2O)3 (guanidinium)8}, ( 30 ). The crystal structure of 30 reveals a 3-perioidc network constr ucted from edgeand vertex-connected MOCs, generated in situ wherein each anionic MOC formulated as [Mn8(ImDC)8(HImDC)4]16is comprised of four doubly and eight triply deprotonated imidazoledicarboxylate exo -ligands that coor dinate to eight Mn2+ ions located at the

PAGE 316

279 vertices of each MOC. Three inde pendent ditopic he terofunctional HnImDC ligands coordinate to each of the Mn2+ ions in a bis(bidentat e) fashion through N-, Oheterocoordination, fo rming the essential fac -MN3(CO2)3 MBB needed to construct a MOC. Each ligand constitutes the edges of the cube and bridges two individual Mn2+ ions by forming two rigid five-membered chelate ri ngs coplanar with the imidazole ring which reinforce the rigidity and directionality of the assembly by locking the metal into position. The regularity of the cube is akin to that of an ideal cube with Mn…Mn distances along the edges of the cube of 6.525 and Mn…Mn…Mn angles of 90o (Figure 4.27.). The presence of excess manganese and guani dinium ions not only provides charge balance to the framework but permits vertex -to-vertex and edge-t o-edge connection to twelve adjacent cubes through the peripheral oxygen atoms of the HnImDC ligand, to reveal an extended zeolitelike framework possessing ast topology. Each of the twelve edge connections occur though a Mn2+ ion that adopts an octahedral coordination environment whereby four carboxylate oxygen atoms (dMn – O = 2.13 ) from two ImDC ligands of two independent MOCs lie along th e equatorial plane while disordered DMF molecules in the axial positions saturate the coordination sphere of the metal (dMn – O = 2.16 ).

PAGE 317

280 Figure 4.27. (a) Ball and stick representation of the single-crystal structure of 30 which reveals edge-to-edge connections throug h octahedrally coordinated Mn2+ ions; (b) vertex-to-vertex connections occur through char ge-assisted H-bonding be tween guanidinium ions which link four MOCs through (c) a supramolecular tetrahedron represented in yellow; (d) six MOCs (red tile) assemble to generate the AST -cage (green tile). Hydrogen at oms and solvent molecules are omitted for clarity Color code: Mn, or ange; C, gray; N, blue; O, red. The vertex-to-vertex interm olecular connections however are sustained by chargeassisted hydrogen bonds that occur between four guanidinium ca tions and carboxylate oxygen atoms located at the vertices (N-H…O distances of 2.92 to 3.00 ) of neighboring cubes. The cations are uniquely situated to represent a supramolecular tetrahedron and play an influential role in governing the overall outcome of the assembly. Hence, each tetrahedron and can be simplified into a 4-conne cted tetrahedral node th at extends to four

PAGE 318

281 neighboring MOCs via the vertices to gene rate a 3-periodic network possessing the zeolitelike AST topology. As previously mentioned, a similar appro ach can be used to target binodal edgetransitive nets whereby the MOC is regarded as an 8-connected node instead of a d4R. The 8-connected nodes are therefore linked t ogether through the 4-connected tetrahedral nodes, to reveal a binodal edge-transitiv e (4,8)-connected net represented by the flu (Figure 4.28.). It is not eworthy to mention, the flu net is the dual of the only quasi-regular net, a uninodal 12-connected fcu net, described in the firs t section of this chapter. Figure 4.28. (a) MOC and tetrahedron represented as 8and 4-connected nodes, respectively which self-assembly to form (b) a binodal edge-transitive flu -like net. The total solvent-accessible volume for 30 was obtained using the PLATON software38 by summing the voxels that are more than 1.2 away from the framework which revealed an estimated free-volume of 48.3%. The approximate diameter of the AST -cage is 15 ; however, the 6-membered ri ng windows which lead to the interior of the cavity are severely obstructed by the pr esence of the guanidinium ions which form supramolecular H-bonded panels that span mu ch of the window aperture. Nevertheless,

PAGE 319

282 we proceeded to investigate the potential gas sorption properties of this material. Despite numerous attempts to activate the compound, BET surface area values were only found to be in the range of 10 to 25 m2/g. The PXRD pattern collect ed on these samples after sorption did not show any peaks, indicating co llapse of the framework. It is conceivable that the edge-directed Mn2+ ions become unstable upon evacua tion if the axial ligands are removed and contributes to the framework collapse. The same design strategy as detailed above was used to construct similar nets, lead by Mohamed Alkordi ( ast -ZMOFs) and Jacilynn Brant ( lta -ZMOF), whereby the octahedral metal ions were substituted for an assortment of metal ions which include Co, Cd, In, and Zn. Likewise, the supramolecular tetrahedron formed via the assembly of guanidinium ions can readily be replaced by analogous monovalent metal clusters ( i.e. K+, Na+, Cs+) to generate isoreticular edgea nd vertex-connected nets. The results described herein, therefore highli ght the versatile nature and a ccessibility of this approach toward the construction of ZMOFs based on different octa hedrally coordinated metal ions. Work is in progress to investigate th e potential of these ma terials for host-guest applications for catalysis and/or small mo lecule sensing and to expand the structure library to include novel ZMOFs based on the di rected assembly of the readily accessible MOCs using an assortment of functional ditopic organic ligands and metal ions. 4.2.3. Experimental Section 4.2.3.1. Materials and Methods All materials and methods are describe d in Chapter 2, unless otherwise noted.

PAGE 320

283 4.2.3.2. Synthesis and Characterization Synthesis of {[Mn12(guanidinium)8(ImDC)8(HImDC)4](DMF)8(H2O)3}n, (30). A solution of Mn(NO3)2 .6H2O (35.8mg, 0.20mmol), H3ImDC (31.2mg, 0.20mmol), DMF (1mL), CH3CN (1mL) and guanidine ni trate (0.2mL, 1M in DMF) was prepared in a 20 mL scintillation vial. The solution was heated to 85oC for 12 hours followed by additional heating to 105oC for 23 hours and 115oC for 23h hours. Each of the heating cycles were increased at rate of 1.5oC/minute and cooled to room temperature at a rate of 1.0oC/minute. Red polyhedral crystals of 30 were harvested and air-dried (Yield: 13.6 mg, 22.7 %). The as-synthesized material was determined to be insoluble in H2O and common organic solvents. 4.3. Summary and Conclusions The work reported in this chapter furt her supports the evolution of the MBB approach by introducing the next generation of MOMs that are delib erately constructed from externally functionaliz ed rigid and directional MOP, termed as SBBs. They are ideal nodes to be exploited in the constructi on of target nets where the vertex figures allocate the need for high connectivity. Fu rthermore, expanded non-interpenetrated nets are readily obtained due to the narrow window dimensions of the SBB, which in turn provides enhanced stability to the overall framework. This chapter outlined several examples of 3-periodic MOMs built-up from discrete MOP. They can be divided into tw o classifications based on the SBBs employed; namely two prototypical plat forms based upon linking nanoscale faceted polyhedra and another platform sustained by linking meta l-organic cubes which are related to

PAGE 321

284 hexahedra. The first stu dy provided two examples of 12-connected MOFs with fcu topology. The frameworks are sustained by ri gid cross-linking of cubohemioctahedral SBBs, {[M2(ABTC)](H2O)3}n where M = Ni or Co, thr ough a tetracarboxylate organic ligand, ABTC4-. Work is in progress to synthesize isostructural and isoreticular anionic fcu nets with large cavities in which case the de corated metal site is to be substituted for extra-framework organic cations ( e.g. alkylammoniun ions) or charged complexes ( e.g. porphyrin-based molecules). The incorporati on of extra-framework cations of varying size and functionality is a practical strategy to tune the pore size of these materials. If successful, these materials could be useful in gas storage and/or cat alytic applications. The second study focused on the synthesis of (3,24)-connected MOFs with rht topology. Small rhombihexadral SBBs where covalently cross-li nked through a tritopic hexacarboxylate organic ligand, PTMOI6-. This strategy therefore afforded a series of isostructural compounds formulated as {[M3(PTMOI)(L)3]}n where M = Cu2+, Zn2+, Co2+, Mn2+. The uniqueness of the rht network to the practice of reticular chemistry was exemplified by reporting the deliberate constr uction of an expanded isoreticular analogue formulated as {[Zn3(ABPTMOI)(H2O)3]}n. Future studies for this neutral platform will be devoted to finding the appr opriate conditions to make th ese materials porous. This would permit a systematic analysis to be conducted in order to correlate the dihydrogen interactions as a function of both varyi ng the metal cation and size of the cages. Concurrent work will also focus on impre gnating these MOFs with enlarged metal complexes, as well as, exploring the possibi lity of generating an ionic organic-based rht MOFs by coordinating halide ions in the ap ical position of the paddlewheel MBB.

PAGE 322

285 In the last study, predesigned finite MOCs were employed as rigid and directional SBBs. This provides a reliable pathway to target to specific zeolite nets comprised of the composite d4R building unit, while simultaneously generating nets based on edgetransitive nets. Work related to this projec t is in progress to explore the potential of constructed ZMOFs as hosts for molecules with applications in cat alysis and/or small molecule sensing. Additionally, we are extending this appro ach further to construct novel ZMOFs based on the directed assembly of the readily accessible MOCs.68 In conclusion, the developed SBB me thodologies delineated throughout this chapter provide a unique and modular pathway to access complex structures. This approach offers new horizons for solid state ma terials in terms of ch emical functionality, scale, and structural diversity that are ot herwise unattainable using basic MBBs and SBUs. 4.4. References (1) Guest Editors: Long, J. R.; Yaghi, O. M. Chem. Soc. Rev. 2009 38 1201-1508. (2) El-Kaderi, H. M.; Hunt, J. R.; Mendoza-Cortes, J. L.; Cote, A. P.; Taylor, R. E.; O'Keeffe, M.; Yaghi, O. M. Science 2007 316 268-272. (3) Liu, Y. L.; Eubank, J. F.; Cairns, A. J. ; Eckert, J.; Kravtsov, V. C.; Luebke, R.; Eddaoudi, M. Angew. Chem. Int. Ed. 2007 46 3278-3283. (4) Liu, Y. L.; Kravtsov, V.; Walsh, R. D.; Poddar, P.; Srikanth, H.; Eddaoudi, M. Chem. Commun., 2004 2806-2807. (5) Ockwig, N. W.; Delgado-Friedric hs, O.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2005 38 176-182. (6) McManus, G. J.; Wang, Z.; Zaworotko, M. J. Cryst. Growth Des., 2003, 4 11-13. (7) Perry, J. J.; Kravtsov, V. C.; McManus, G. J.; Zaworotko, M. J. J. Am. Chem. Soc. 2007 129 10076-10077. (8) Tranchemontagne, D. J. L.; Ni, Z.; O'Keeffe, M.; Yaghi, O. M. Angew. Chem. Int. Ed. 2008 47 5136-5147. (9) Perry, J. J.; Perman, J. A.; Zaworotko, M. J. Chem. Soc. Rev. 2009 38 1400-1417. (10) Coxeter, H. S. M., Regular Polytopes MacMillian: New York, 1963. (11) Sudik, A. C.; Cote, A. P.; Wong -Foy, A. G.; O'Keeffe, M.; Yaghi, O. M. Angew. Chem. Int. Ed. 2006 45 2528-2533.

PAGE 323

286 (12) Kusukawa, T.; Fujita, M. J. Am. Chem. Soc. 2002 124 13576-13582. (13) Webb, R. Stella4D v4.4; Melbourne, Australia, 2008 http://www.software3d.com/Stella.php (14) Webb, R. Symmetry: Culture and Science 2000 11 231-268. (15) Holden, A., Shapes, Space, and Symmetry Columbia University Press: New York, 1971 ; p 94. (16) Moulton, B.; Abourahma, H.; Bradne r, M. W.; Lu, J. J.; McManus, G. J.; Zaworotko, M. J. Chem. Commun. 2003 1342-1343. (17) Cotton, F. A., Walton, R. A., Multiple Bonds Between Metal Atoms Oxford University Press: Oxford, 1982 (18) Batten, S. R.; Hoskins, B. F.; Moubaraki, B.; Murray, K. S.; Robson, R. Chem. Commun. 2000 1095-1096. (19) Li, H.; Eddaoudi, M.; Groy, T. L.; Yaghi, O. M. J. Am. Chem. Soc. 1998 120 8571-8572. (20) Eddaoudi, M.; Kim, J.; Vodak, D.; Sudi k, A.; Wachter, J.; O'Keeffe, M.; Yaghi, O. M. Proc. Nat. Acad. Sci. U.S.A. 2002 99 4900-4904. (21) Lin, X.; Telepeni, I.; Blake, A. J.; Dail ly, A.; Brown, C. M.; Simmons, J. M.; Zoppi, M.; Walker, G. S.; Thomas, K. M.; Mays, T. J.; Hubberstey, P.; Champness, N. R.; Schroder, M. J. Am. Chem. Soc. 2009 131 2159-2171. (22) Eddaoudi, M.; Kim, J.; Wachter, J. B.; Chae, H. K.; O'Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc., 2001 123 4368-4369. (23) Moulton, B.; Lu, J. J.; Mondal, A.; Zaworotko, M. J. Chem. Commun. 2001 863864. (24) Chen, B. L.; Eddaoudi, M.; Reineke, T. M.; Kampf, J. W.; O'Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2000 122 11559-11560. (25) Abourahma, H.; Coleman, A. W.; M oulton, B.; Rather, B.; Shahgaldian, P.; Zaworotko, M. J. Chem. Commun. 2001 2380-2381. (26) Furukawa, H.; Kim, J.; Plass, K. E.; Yaghi, O. M. J. Am. Chem. Soc. 2006 128 8398-8399. (27) Furukawa, H.; Kim, J.; Ockwi g, N. W.; O'Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2008 130 11650-11661. (28) Ke, Y.; Collins, D. J.; Zhou, H.-C. Inorg. Chem., 2005 44 4154-4156. (29) Larsen, R. W.; McManus, G. J.; Perry, J. J.; Rivera-Otero, E.; Zaworotko, M. J. Inorg. Chem. 2007 46 5904-5910. (30) Mohomed, K.; Abourahma, H.; Zaworotko, M. J.; Harmon, J. P. Chem. Commun. 2005 3277-3279. (31) Mohomed, K.; Gerasimov, T. G.; Abour ahma, H.; Zaworotko, M. J.; Harmon, J. P. Mater. Sci. Eng. A, 2005, 409 227-233. (32) Jung, M.; Kim, H.; Baek, K.; Kim, K. Angew. Chem. Int. Ed. 2008 47 5755-5757. (33) Nouar, F.; Eubank, J. F.; Bousquet, T. ; Wojtas, L.; Zaworotko, M. J.; Eddaoudi, M. J. Am. Chem. Soc. 2008 130 1833-1835. (34) O'Keeffe, M. Reticular Chemistry Structure Resource. http://rcsr.anu.edu.au/ (35) Cairns, A. J.; Perman, J. A.; Wojtas, L.; Kravtsov, V. C.; Alkordi, M. H.; Eddaoudi, M.; Zaworotko, M. J. J. Am. Chem. Soc. 2008 130 1560-1561. (36) Delgado-Friedrichs, O.; O'Keeffe, M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2007 63 344-347.

PAGE 324

287 (37) Li, H.; Eddaoudi, M.; O'Keeffe, M.; Yaghi, O. M. Nature 1999 402 276-279. (38) Spek, A. L. Acta Crystallogr. 1990 A46 c34. (39) Hong, S.; Oh, M.; Park, M.; Yoon, J. W.; Chang, J. S.; Lah, M. S. Chem. Commun. 2009 5397-5399. (40) Zhao, D.; Yuan, D. Q.; Sun, D. F.; Zhou, H. C. J. Am. Chem. Soc. 2009 131 91869188. (41) Yan, Y.; Lin, X.; Yang, S. H.; Blake, A. J.; Dailly, A.; Champness, N. R.; Hubberstey, P.; Schroder, M. Chem. Commun. 2009 1025-1027. (42) Nelson, A. P.; Farha, O. K.; Mulfort, K. L.; Hupp, J. T. J. Am. Chem. Soc. 2009 131 458-460. (43) Shao, X. B.; Jiang, X. K.; Wang, X. Z.; Li, Z. T.; Zhu, S. Z. Tetrahedron 2003 59 4881-4889. (44) Foster, M. D. a. T., M. M. J. A Database of Hypothetical Zeolite Structures. http://www.hypotheticalzeolites.net/ (45) Davis, M. E. Nature 2002 417 813-821. (46) Hourdin, G.; Germain, A.; Moreau, C.; Fajula, F. Catal. Lett. 2000 69 241-244. (47) Yang, X. Gongye Cuihua 2003 11 19-24. (48) Maxwell, I. E.; Stork, W. H. J. Stud. Surf. Sci. Catal. 2001 137 747-819. (49) Breck, D. W., Zeolite Molecular Sieves: Structure, Chemistry, and Use John Wiley & Sons, Inc.: New York, 1974 (50) Glaeser, R.; Weitkamp, J. Springer Ser. Chem. Phys. 2004, 75 161-212. (51) Corma, A. NATO ASI Ser., Ser. C 1992 352 373-436. (52) Le Van Mao, R.; Vu, N. T.; Xiao, S.; Ramsaran, A. J. Mater. Chem. 1994 11431147. (53) Claridge, R. P.; Lancaster, N. L.; Mi llar, R. W.; B, M. R.; Sandall, J. P. B. J. Chem. Soc., Perkin Trans. 2 2001 197-200. (54) Kuznicki, S. M.; Langner, T. W.; Curran, J. S.; Bell, V. A. U. S. Pat. Appl. US 20020077245, 2002. (55) Kuznicki, S. M.; Bell, V. A.; Langner, T. W.; Curran, J. S. U. S. Pat. Appl. US20020074293, 2002. (56) Paillaud, J. L.; Harbu zaru, B.; Patarin, J.; Bats, N. Science 2004 304 990-992. (57) Liu, Y. L.; Kravtsov, V. C.; Larsen, R.; Eddaoudi, M. Chem. Commun. 2006 14881490. (58) Liu, Y. L.; Kravtsov, V. C.; Eddaoudi, M. Angew. Chem. Int. Ed. 2008 47 84468449. (59) Banerjee, R.; Phan, A.; Wang, B.; Knobl er, C.; Furukawa, H.; O'Keeffe, M.; Yaghi, O. M. Science 2008 319 939-943. (60) Hayashi, H.; Cote, A. P.; Furukawa, H.; O'Keeffe, M.; Yaghi, O. M. Nat. Mater. 2007 6 501-506. (61) Park, K. S.; Ni, Z.; Cote, A. P.; Choi, J. Y.; Huang, R. D.; Uribe-Romo, F. J.; Chae, H. K.; O'Keeffe, M.; Yaghi, O. M. Proc. Nat. Acad. Sci. U.S.A. 2006 103 1018610191. (62) Ferey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F.; Dutour, J.; Surble, S.; Margiolaki, I. Science 2005 309 2040-2042. (63) Fang, Q. R.; Zhu, G. S.; Xue, M.; Sun, J. Y.; Wei, Y.; Qiu, S. L.; Xu, R. R. Angew. Chem. Int. Ed. 2005 44 3845-3848.

PAGE 325

288 (64) Huang, X. C.; Lin, Y. Y.; Zhang, J. P.; Chen, X. M. Angew. Chem. Int. Ed. 2006 45 1557-1559. (65) Sava, D. F.; Kravtsov, V. C.; Nouar, F.; Wojtas, L.; Eubank, J. F.; Eddaoudi, M. J. Am. Chem. Soc. 2008 130 3768-3770. (66) Brant, J. A.; Liu, Y. L.; Sava, D. F.; Beauchamp, D.; Eddaoudi, M. J. Mol. Struct. 2006 796 160-164. (67) Sava, D. F.; Kravtsov, V. C.; Eckert J.; Eubank, J. F.; Nouar, F.; Eddaoudi, M. J. Am. Chem. Soc. 2009 131 10394-10396. (68) Alkordi, M. H.; Brant, J. A.; Wojtas, L.; Kravtsov, V. C.; Cairns, A. J.; Eddaoudi, M. J. Am. Chem. Soc. 2009 131 17753-17755. (69) Delgado-Friedrichs, O.; O'Keeffe, M.; Yaghi, O. M. Phys. Chem. Chem. Phys. 2007 9 1035-1043.

PAGE 326

289 Chapter 5. Tuning Pore Size via Charged Metal-Organic Frameworks: Pathway to Enhance Adsorbate-MOF Interactions 5.1. Introduction Porous materials, specifically MOFs, are widely recognized as an interesting class of solid-state adsorbents for use in applica tions such as gas stor age and sequestration, as evidenced throughout this dissertation.1-3 A contemporary challenge in the context of hydrogen storage is to significantly increase the adsorbate-MOF interactions, i.e. by factors of 3 – 5 over traditional physisorbents, while maintaining a high free volume and low framework density. These tr aits are imperative in order to facilitate high uptake capacities that will ultimately satisfy the Do E technical targets at working pressure and temperatures. Several key parameters have shown great promise towards increasing the sorption energetics in MOFs (see chapter 1 a nd 2); two of which are in fact the relative pore size and the presence of a strong elec trostatic field. Synthe tic strategies that therefore allow for control over these features in a systematic fashion are particularly salient to reticular chemistry. One method that has been explored to study the effects of por e size is so-called interpenetration, also known as framework catenation, whereby one or more nets are entangled within a single network.4-6 Note that early reports ofte n regarded this feature as an undesirable outcome for gas storage appli cations due to mitigation in the free volume and an increase in the framework density. In recent years however partitioning the empty

PAGE 327

290 space using this approach has endured success in that higher surface areas and isosteric heats of adsorption have been attained.7-12 A limitation to this strategy however lies in the fact that it is often difficult to control th e degree of interpenetration and factors which influence its formation remain generally unclear. Furthermore, for some nets it is strictly forbidden all together such as in the case of ZMOFs.13 Reports on anionic ZMOFs and other charged MOFs have therefore show n that occupying the large open space with extra-framework cations ( e.g. hydrated metal ions or or ganic cations) leads to an electrostatic field with charged-induced forces and therefore enhances Q st.1, 14-22 The objective of this project was th erefore to design a charged and robust structural prototype that w ould permit facile modifications to the pore metrics and functionality through th e use of different extra-framewor k organic cations (Figure 5.1.). This would serve as a dual feature whereby th e organic cations could also be substituted for an array of metal cations. Note that depending on the window dimensions of the resultant MOF, the organic cations may al so be introduced into the MOF using postsynthetic exchange techniques. Figure 5.1. Examples of targeted extra-framework alkyla mmonium organic cations (left to right): Dimethylammonium (DMA+), diethylammonium (DEA+), triethylammonium (TEA+), and tetrabutylammonium (Bu4N+). In this sense, charged frameworks of fer a significant advantage over neutral MOFs because it is often eas ier to fine-tune the pore f eatures using post-synthetic

PAGE 328

291 exchange techniques as oppose to finding th e appropriate reacti on conditions to synthesize an isoreticular or isostructural MO F. In planning the synt hesis we chose to use tetrahedral and triangular building units (BUs) because they are less likely deform in the assembly process due to their rigidity and so we anticipated the resultant MOF would also be robust.23 The self-assembly of tetrahedral and planar triangular BUs has the potential to form an infinite number of nets. In the absence of structure directing effects however nets with ctn and bor topology are most likely to form by linking the nodes through one kind of edge. The reason being that they are the most symmetric structures for the assembly of the aforementioned MBBs,24 i.e. edge-transitive. They are in fact ideal targets for this study b ecause they possess large cavities.25 In terms of finding the suitable MBBs, Eddaoudi and co-workers have already confirmed the feasibility of generati ng rigid single-metal ion TBUs with p -block metal cations ( i.e. In3+)26 and therefore we chose to target th is group of metals to represent the tetrahedral MBB. A trivalent node in conjunc tion with a trivalent organic linker would afford a charged framework and so we chos e to bridge the tetr ahedral MBBs with 1,3,5benzenetricarboxylic acid (H3BTC) because it is a rigid MBB and its vertex figure corresponds to a planar triangul ar BU. The diverse coordina tion geometries sustained by BTC (Figure 5.2.) has afforded a wide range of MOMs with intr iguing topologies and fascinated properties, as evidenced by the hundreds of reported structures in the literature.27-34 In order to form the targeted (3,4)-c onnected nets it is vital that the three carboxylate moieties in H3BTC coordinate to the metal ion in either a bidentate fashion or mondentate fashion. This is di fficult to control however as BTC has many permutations

PAGE 329

292 (Figure 5.2.) and slight variations in the r eaction conditions can lead to an unprecedented results ( i.e. quaternary net) Figure 5.2. Examples of common coordination modes generated by BTC. 5.2. Results and Discussion 5.2.1. Porous Anionic MOFs with ctn To pology: Structural Description Reaction of H3-BTC with InBr3 in a DMF/CH3CN solution in the presence of tetraethylammonium chloride (TEACl) indeed yi elds colorless tetrahedral-shape crystals characterized and formulated using single-crystal diffr action studies as, {[In(BTC)1.33]DMA}n ( 31 ). The purity of the as-synthesi zed material was confirmed by similarities between the calculated and experi mental PXRD patterns (Appendix B). In the crystal structure of 31 each indium ion is coordinate d to eight carboxylate oxygen atoms

PAGE 330

293 from four independent deprotonated BTC ligands. This configuration affords a rigid single metal ion MBB, In(RCO2)4 (dIn – O = 2.114 to 2.374 ) which can be portrayed into the expected TBU. Each indepe ndent trivalent BTC anion is coordinated to three In(III) ions in a bidentate fashion and is therefore regarded as a 3-connect ed node (Figure 5.3.). The assembly of these 3and 4-connect ed nodes generates a microporous anionic 3periodic MOF with ctn (C3N4) topology.25 Charge balance for 31 however can potentially come from two different extra-framework or ganic cations; that is dimethylammonium cations (DMA+) due to thermal decomposition of DMF or tetraethylammonium cations (TEA+) from the SDA. Figure 5.3. Ball-and-stick and schematic representation of the ctn analogs, 31 and 32 Top: the 3and 4-connected organic and inorganic MBBs can be rationalized as triangular and tetrahedral BUs, respectively. Bottom: MBBs self-assemble to yield anionic 3-periodic MOFs with a complex intersecting channel system. Color Code: Ga = yellow; In = green; C = gray; O = red. All hydrogen atoms and guest molecules are omitted for clarity. Gallium ctn-MOF Indium ctn-MOF

PAGE 331

294 The crystal structure of the indium-based ctnMOF, 31 revealed too much disorder in the channels and thus it was not possible to de finitively confirm the presence of one cation over the other. Preliminary meta l exchange studies nevertheless lead me to believe that DMA+ cations are present in 31 due to the rapid rate of the exchange, as will be discussed below. Note that other solvent systems were pursued in an attempt to avoid the presence of DMA+ cations ( e.g. DMSO). This was carried out in conjunction with an assortment of larger alkylammonium-bas ed SDAs, such as but not limited to diethylamine (DEA), triethylam ine (TEA), tetrapropylamine (Pr4N), tetraethylamine (Et4N), and tetrabutylamine (Bu4N). Attempts to rationally fi ne-tune the pore dimensions in 31 were not successful but studies carried out on another p -block metal cation ( i.e. Ga3+) offered encouraging results. Accordingly, reaction of H3BTC with Ga(NO3)3 x H2O in a DMA/butanol solution in the presence of TEACl generates colorl ess polyhedral crystals. The as-synthesized material was characterized and formulated using single-crystal di ffraction studies as, {[Ga(BTC)1.33]DMA}n ( 32 ). The crystal structure of 32 therefore reveals the same network topology as 31 ( i.e. ctn ) but in this case each Ga(III) ion is 4-coordinate due its smaller atomic radius and thus each BTC3ligand is coordinated to three gallium centers in a monodentate fashion (Figure 5.3.). The same dilemma concerning the organic source for the charge balance is encountered in 32 because DMA can also decompose into DMA+, while TEA+ could also be present. Reaction between H3BTC and Ga(NO3)3 x H2O in a DMF/butanol solution also generates co lorless microcrystalline material, which was determined to be another ctn analog ( 33 ). This was confirmed by comparing similarities between the experimental PXRD patterns (Append ix B). Charge balance in this case must

PAGE 332

295 therefore be provided by extra-framework DMA+ cations because there is no other source. Two more ctn analogs were isolated using a similar approach and contain DEA+ ( 34 ) and Bu4N+ ( 35 ). They were generated by reacting H3BTC with Ga(NO3)3 x H2O in a DEF/butanol solution and DMSO/buta nol solution in the presence of tetrabutylammonium bromid e (TBABr), respectively. Despite the high symmetry of the Inand Ga-based ctn -MOFs ( i.e. 31 and 32 ), the pore system is quite complex. It is po ssible to simplify the structure into four independent intersecting channels, wh ich form a repeating array of [M9(BTC)11] 6cagelike units where M refers to indium or gallium (Figure 5.4.). The apical positions of the cage are defined by two parallel BTC3moieties oriented in an eclipsed fashion with an interlayer separation that is estimated to be 5.472 and 5.255 in 31 and 32 respectively. The BTC3ligands coordinate to six independe nt M(III) centers, three on the top and bottom and are br idged through three BTC3ligands located in the equatorial plane. The relative dimensions of the cage ar e significantly enhanced due to the presence of three additional equatorial M(III) ions. These are connect ed in a bidentate mode via two BTC3units. A large triangular-shape cage is thereby revealed having a 3-fold rotation axis ( i.e. C3 symmetry). The relative dimensions of this large cage are reflected in the edge length which is estimated to measure 16.236 and 15.927 for 31 and 32 respectively with accessible wi ndows that range from approximately 6.74 to 7.55 . The total accessible free volume to solvent molecule s and charge balancing organic cations in 31 and 32 was calculated to be 66.7 % and 63.1 %, respectively. Note that gallium has a smaller atomic radius than indium and ther efore exhibits slightly smaller dimensions.

PAGE 333

296 Figure 5.4. Select fragments from the crystal structure of 31 : (a) Triangular cage viewed down x axis and formulated as [In9(BTC)11]6-; and (b) Alternate view of the cage to highlight the four type of intersecting channels with window dimensions in the range of 5.473 to 7.880 . Color Code: In = green; C = gray; O = red. Hydrogen atoms and solvent molecules are omitted for clarity. In summary, the direct synt hesis of five porous anionic ctn -MOFs housing sizetunable extra-framework organic cations in th e channels was reported. It is important to mention that during the course of these studies two other groups reported on this work ( i.e. indium analogs only).15, 30 This therefore demonstrates the facile experimental protocols and modularity to which the pore metr ics can be modified in these materials by simply employing cations of varying size and functionality. All of the compounds described above were isolated using direct synthesis; however it is entirely possible that extra-framework cations of suitable size can be introduced via post-synthetic methods. In principle, these MOFs therefore represent prototypical platforms that provide a unique opportunity to investigate th e impact of pore size on the gas adsorption properties ( e.g. adsorbent-adsorbent interactions). 5.2.1.1. Properties: Preliminary Gas Sorpti on and Metal Ion Exchange Studies The anionic nature, large accessible ca ges, and high thermal and chemical stability ( i.e. > 300oC) exhibited by these ctnMOFs permits the incorporation of a wide (a) (b)

PAGE 334

297 range of fine-tunable extra-framework cations ( i.e. organic or metal ions). Preliminary gas adsorption studies carried out on the indium and galliu m analogs having either DMA+ or TEA+ extra-framework cations residing in the cages are discussed below. Note that ongoing investigations are being conducted with similar analogs. Due to time constraints and the high demand for the gas adsorption anal yzer they could not be included in this dissertation. In order to assess the sorpti on properties of the indium ctn -MOF, 31 the guest molecules were exchanged with ethanol for 3 days. The colorless microcrystalline sample was then air dried and loaded into a 6-mm sample cell, initially outgassed at room temperature for 8 h and then gradually heated to 280oC for 3 h. PXRD patterns confirmed that the framework does not change after evacuation (Appendix B). TGA spectra for the as-synthesized and exchanged samples indicate that 31 is indeed stable up to 400oC (Appendix C). The exceptional thermal stability displayed by this compound is a feature rarely observed in MOFs and lik ely attributed to the rigid MBBs in conjunction with the intricate pore structure. This is interesting because it offers the potential for dehydrating the extra-framework metal cations to genera te accessible open-metal binding sites, i.e. analogous to zeolites. This has proven to be problematic in fact for many reported anionic MOFs owing to the lower thermal stability of the framework, i.e. at temperatures required to fully dehydrate the metal ions. The nitrogen and argon adsorption isot herms collected at 77 K and 87 K, respectively reveal Type I behavior but satu ration is not reached at 1 atm (Figure 5.5.). BET and Langmuir surface areas were estimated from the Ar isotherm to be 515 m2/g and 851 m2/g and the total pore volu me was found to be 0.23 cm3/g (at P/P0 = 0.60). The

PAGE 335

298 internal pore diameter of 5.52 and 7.93 was obtained from both a cylindrical and spherical NLDFT pore model, respectively assuming an oxidic (zeolite) surface. The lower than expected pore volume may be attr ibuted to the location of the alkylammonium cations in the cage and/or improper sample activ ation. Note that in bo th cases a hysteresis loop is observed in the desorption isotherms a nd therefore the adsorbate is irreversibly sorbed in 31 The loop closes at lower re lative pressures in the N2 isotherm as compared to Ar, which is likely associated with the fact that N2 has a larger kinetic diameter and dipole moment, i.e. can lead to interactions with th e surface of the MOF. This is in agreement with the slightly lower values obtained from the N2 isotherm. Figure 5.5. N2 and Ar sorption isotherm for 31 collected at 77 K and 87 K, respectively. The hydrogen sorption capacity for 31 measured at 77 K and 87 K and atmospheric pressures revealed a moderate uptake of 1.2 wt % of H2 (Figure 5.6a.). The Q st was calculated to be 7.39 kJ mol-1 and found to be nearly c onstant at higher loadings, which is indicative of averaging of th e binding sites (Figure 5.6b.). Since compound 31 contains no open-indium binding site s the consistency observed in the Q st can be

PAGE 336

299 attributed to the presence of an electrostatic field and the reduced pore size due to the presence of the alkylammonium cations ( i.e. DMA+ or TEA+). Ongoing studies are therefore focused upon identifying which cation is present and synthe sizing other analogs with larger cations, i.e. using direct and postsynthetic techniques. Figure 5.6. Gas sorption studies for the indium ctn -MOF, 31 : (a) N2 and Ar sorption isotherm at 77 K and 87 K, respectively; (b) H2 sorption data at 77 K and 87 K; and (c) isosteric heat of adsorption for H2. Gas sorption studies carried out on the gallium ctn -MOF after exchange in ethanol and evacuated at 280oC revealed different behavi or with regards to the N2 and Ar isotherms (Figure 5.7a.), i.e. the MO F selectively adsorbs Ar over N2. It is important to note that the identity of the extra-framwork organic cation is also under speculation in 32 i.e. either DMA+ or TEA+. The cage and window dimensions in compounds 31 (In) and 32 (Ga) are very similar and therefore I believe that it is possible that DMA+ and TEA+ cations are present in the cages, respectively. The larger size of the latter cation could hinder accessibility into th e cage, depending on its location, and help to explain the hysteresis observed in the desorption for both adsorbates. Since N2 has a larger kinetic (a) (b)

PAGE 337

300 diameter than Ar and has a dipole moment is not surprising that it adsorbed less, that is, if the windows are indeed blocked. The BET and Langmuir surfaces areas estimated from the Ar isotherm at 87 K are 256 and 368 m2/g, which are lower than expected. The H2 isotherm however is fully reversible a nd has a maximum uptake of 1.5 wt % at 77 K (Figure 5.7b.). This is in accord with what we expect because the gallium analog has a lower framework density. Figure 5.7. Gas sorption studies for a gallium ctn -MOF, 32 : (a) N2 and Ar sorption isotherm at 77 K and 87 K, respectively; (b) H2 sorption data at 77 K and 87 K; and (c) Isosteric heats of adsorption for H2 in 32 compared with the indium analog ( 31 ). (a) (b) (c)

PAGE 338

301 The isosteric heats of adsorption for 31 and 32 are in fact quite comparable but the latter deviates initially at higher loadings (Figure 5.7c .). This result was unexpected because if 32 does indeed have smaller pores due to the presence of the extra-framework organic cations then one would expect the Q st to be higher across the entire range. A possible explanation to account for this may in fact have nothing to do with cations, meaning the non-coordinated carb oxylate oxygen atoms from BTC3may interact with N2 and hinder is mobility in the MOF. Future studies will be carried to confirm the identity of the extra-framework cation because wit hout this key bit of information a true comparison between 31 and 32 cannot be conducted. Post-synthetic metal ion exchange studies were carried out on 31 to assess the feasibility of substituting the organic cations for various mono-, di-, and trivalent metal ions (Table 5.1.). Note that sim ilar studies were also done with 32 but the exchange process could not be quantified because a gallium lamp is needed for the Atomic Absorption (AA) instrument to deduce the meta l ratios. Exchange studi es carried out with trivalent metal ions, i.e. Cr3+ or Fe3+, were not successful for 31 and 32 because the ions instantly attack the ethanol exchanged MOF. In an attempt to alleviate this issue the concentration of the stock solution was reduce d, as well as, the durat ion of the exchange but the crystals still degraded to powder. The divalent metal ion exchange studies however revealed promising results in whic h case the colorless crystals changed color almost instantly and remained crystalline even after exchange for 15 h in the stock solution. The AA experiments for the Cu2+ and Co2+ exchanged samples of 31 revealed full exchanges, while the Ni2+ and Mg2+ analogs were determined to be partially exchanged. The BET surface areas were m easured for select samples on the NOVA

PAGE 339

302 surface area analyzer but the results are not that significant at this stage and further studies are required to make m eaningful conclusions. In short, we have demonstrated that full metal ion exchange is feasible using the robust ctn platform but the surface areas are lower than expected and therefore much e ffort is needed to optimize the activation conditions. The PXRD patterns are in good ag reement with the as-synthesized and calculated patterns and therefore the unexpected results are not attributed to framework collapse (Appendix B). Accordingly, it would be beneficial to this study to collect SCD data of the exchanged samples to confirm their location in the MOF. Table 5.1. Summary of the metal cation exchange studies carried out on 31 which is tentatively formulated as [In12(BTC)16]-1212DMA+ Metal Conc.a (M) Duration Stabilityb Color Change Ratio In3+:Mx + # of Cations Surface Areac (m2/g) Na+ 0.1 1215 h Y colorless N.R. N.R. N.R. Li+ 0.5 1215 h Y colorless N.R. N.R. 20 – 32 Mg2+ 0.5 1215 h Y colorless 5.5:1 2 282 Cu2+ 0.5 1215 h Y light blue 1.88:1 6 76 – 118 Ni2+ 0.5 1215 h Y light green 2.75:1 4 410 Co2+ 0.5 1215 h Y light red 1.76:1 6 12 Cr3+ 0.1; 0.5 5 min N N/A N/A N/A N/A Fe3+ 0.1; 0.5 15 min N N/A N/A N/A N/A aConcentration of the metal stock solution; bYes (Y) or no(N) refers to the stability of the crystals after exchange; cBET surface area measured on the NOVA instrument after evacuation at 200oC. 5.2.2. Novel Quaternary Net formed by Self-Asse mbly of Triangles (2), Tetrahedral, and Trigonal Prismatic Building Units A vast number of 3-periodic MOMs ca n be categorized as uninodal nets ( i.e. vertex transitive).35, 36 The best examples are the five re gular nets with a transitivity of [1111] and include the srs dia nbo pcu and bcu nets.37 They therefore correspond to

PAGE 340

303 the default nets for the assembly of triangles tetrahedra, squares, octahedra, and cubes, respectively and are thus constructed from one type of polygon or polyhedra. Binodal nets on the other hand offer more diversity since they are built-up from two different nodes.24, 38 The (3,4)-connected nets are excellen t examples to illustrate this concept because employing the simplest of MBBs ( e.g. regular polygon/polyhedra) can afford very different structures. For example, the as sembly of squares that are exclusively linked to triangles and vice versa can afford nets with Pt3O4 ( pto )39 and twisted boracite ( tbo )27, 40 topology, while linking tetrah edra to only triangles and vi ce versa can lead to the formation of the boracite ( bor )41 and C3N4 ( ctn )23, 30, 42 nets, as demonstrated above for the latter case. The MBB approach can also be used to ta rget tertiary nets that consist of three different vertex-linked polygons or polyhedra. The rational design and synthesis of nets having more than two nodes is a considerable synthetic challenge be cause the reaction of unifunctional ligands with metal ions predominantly leads to th e formation of one type of metal chromophore in given structure. To overcome this limitation one must therefore either (1) use a metal cation that has the potential to adopt multiple coordination modes ( e.g. Zn), (2) employ a bifunctional organic ligan d that will facilitate the formation of MBBs with different geometries ( e.g. azolates), and/or (3) intr oduce two different ligands into the reaction scheme. Zaworotko and co-worke rs were in fact the first to demonstrate the feasibility of isolating tertiary nets from simple MBBs by synthesizing USF-3 and USF-4, both of which are built-up from tria ngular, square, and tetrahedral MBBs and exhibit unprecedented network topologies.29 The novel nets are therefore represented by the topological acronyms USF which stand for U niversity of S outh F lorida. The

PAGE 341

304 triangular MBB is represented by a BTC ani on, while the square and tetrahedral MBBs are based on different zinc chromophores. An interesting feature is that the exclusive formation of one net over the other can be pr ecisely controlled using template effects. The authors later extended this simple stra tegy by reporting on the synthesis of USF-5, a 3-periodic copper-based MO M also built-up from tria ngles, squares, and pseudotetrahedral MBBs.43 Note that in this case the au thors used a bifunctional ligand ( i.e. 5aminoisophthatlic acid) to promote the forma tion of geometrically and chemically diverse metal coordination environments as oppos e to using a metal ion with many chromophores. The success of this one-pot MBB approach has witnessed success in as evidenced by reports on othe r novel (3,4)-connected nets,44, 45 as well as, ternary MOFs constructed from a range of geometrically different MBBs.31, 46 5.2.2.1. Structural Description of 36 Herein, we extend this simple strategy to the next hierarchical level by reporting on the serendipitous discovery a 3-periodic MO F comprised of four types of MBBs. This framework therefore represents a quaternary MOF and was in fact isolated while trying to synthesize an indium-based MOF with bor topology. To the best of our knowledge, this represents the first example of a MOF built-up from four different polygons or polyhedra; that is, trigonal prismatic, tetrahedral, a nd two topologically differe nt triangles. Reaction of H3BTC with In(NO3)3 x H2O in a NMP/DMF solution in the presence of TEABr affords colorless polyhedral crystals formulated by SCD diffraction studies as, {[In30(O4)(BTC)24(HCOO)4Br6]}n ( 36 ). The structure is complex as shown in ball-andstick and polyhedral representations (F igure 5.8.). It is comprised of three crystallographically independent 3-connected BTC3anions which display the following

PAGE 342

305 coordination environments: (1) adjacent carboxyl ate moieties bridge to two In(III) ions in a bidentate fashion and the third is coordina ted to another In in a bis(monodentate) fashion, referred to as BTC1 coordination mode; (2) adjacent carboxylate moieties bridge to two In in a monodentate fashion and the third also bridges to another In(III) in a bis(monodentate) fashion, BTC2; and (3) all ca rboxylate moieties coordinate to six In(III) ions in a bis(mondentat e fashion) in BTC3. Figure 5.8. Optical image and select fragments from the crystal structure of 36 including the corresponding vertex figures when augmented. (Left) The four MBBs include BTC, [In(RCO2)4], [In3O(RCO2)6(Br)], and [In3O(RCO2)3(HCOO)6(Br)]. (Right) SBUs for each of the inorganic and organic MBBs. (Middle) the assembly of the n -connected nodes (where n = 3,4, and 6) yields a porous 3-periodic MOF comprised of 3 types of cag es. Color Code: In = green, C = gray, O = red, Br = pink. In3+ + MBBs S B U s

PAGE 343

306 Three independent indium ions are present in the crystal structure of 36 In1 is coordinated to four oxygen atoms, that is, two from BTC1and BTC2, respectively. This reveals a [In(RCO2)4] MBB and reveals a TBU. In2 on the other hand displays an octahedral coordination geometry in which case three octahedra, [InO5], are bridged through a central 3-oxo anion to generate a trimer MBB. Each In is linked to six BTC3 ligands and thereby reveals a trigonal prismatic building unit (TPBU). The coordination mode exhibited by In3 on the other hand is unpr ecedented and so is regarded as a pseudoTMBB. Each indium in latter case c oordinates to two oxygen atoms from BTC3bound in the cis position of the equatorial plane, while the opposing positions are occupied by oxygen atoms from two formic acid ligands. Each octahedral MBB is however still linked through a central 3-oxo anion and thus when augm ented reveals a triangular BU. The four vertices of the supe rtetrahedral building block ( i.e. ST1) are occupied by the 6-connected TMBBs while the ligand ( i.e. BTC3) is located at the four faces of the SBB. Each corner is thereby bridged to six In(III) carboxylate-based tetrahedral MBB, [In(RCO2)4] through three BTC1 ligands to reveal f our cages that are hexagonal in shape, referred to as cage C. Accordingly, this cag e is delimited by 6 tetrahedral MBBs which are positioned in the equatorial plane bri dged through 6 deprotonated BTC2 ligands to the 6-connected pseudo-trimer MBBs located in the axial positions The four hexagonal cages are united through a sec ond supertetrahedral MBB ( i.e. ST2) which links the pseudo-TMBBs through formic acid. The ST2 cage is therefore delimited by 4 pseudoTMBBs located at the vertices and 12 deprot onated formic acid ligands (FA) lie in the faces of the cage ( i.e. two per face). A summary of the selected features of 36 in terms of the pores sizes and window dimensions is pr ovided in Table 5.2. Note that according to

PAGE 344

307 the crystallographic data, the indi um ions that make up cage B ( i.e. ST2) are partially occupied and therefore the pseudo-TMBB is diso rdered over the four positions with an equal probability of being lo cated at either position. Table 5.2. Selected structural features of compound 36 Feature Description Topology novel Crystallographic data Cubic; Fd-3m; a = 50.5457 ; V = 129138 3 # Independent In(III) ions 3 # Independent BTC3anions 3 Density 1.003 g cm-3 Supertetrahedron (ST1) – Cage A Edge length 9.902 Internal diameter 7.918 Window dimensions 4.707 x 7.538 Supertetrahedron (ST2) – Cage B Edge length 7.015 Internal diameter 7.314 Window dimensions 4.167 x 6.369 Hexagonal-shape Cage C Edge length 8.627 Internal diameter 13.295 Window dimensions 7.422 x 9.134 Cage D Internal diameter 7.790 Window dimensions 7.780 x 11.765 Note: all measurements are point-to-point and do not include van der Waals radii between adjacent atoms. Each indium ion is trivalent in 36 and thus yields an overall cationic MOF. The charge is broken down as follows with re spect to each In carboxylate-based MBB: [In(BTC)1.3]-1, [In3O(BTC)2]+1, and [In3O(BTC)(FA)1.5]+1 for the tetrahedral, TMBB, and pseudo-TMBB, respectively. Charge balance for the cationic MOF is provided a coordinated bromide ion that is coordinated in the ap ical position of the TMBB (dIn – Br = 2.558 ). The ion is disordered over the three po sitions with an equal probability of being

PAGE 345

308 located at their site. From a topological perspective compound 36 can be simplified as consisting of 3,4,12,12-connected nodes, which corresponds to BTC, tetrahedral MBB, ST1, and ST2 respectively. Topological anal ysis performed using TOPOS software confirms the frameworks unprecedented topol ogy, as defined by the unique coordination sequences and vertex symbols (Table 5.3. and 5.4.). Table 5.3. Coordination sequence for the independent nodes in 36 Node cs1 cs2 cs3 cs4 cs5 cs6 cs7 cs8 cs9 cs10 In(RCO2)4 4 8 48 70 138 102 268 246 482 342 BTC 3 17 28 79 80 191 162 341 277 567 BTC 3 17 28 78 81 182 155 353 282 565 ST1 12 24 48 38 132 156 300 206 468 360 ST2 12 24 48 38 144 156 288 206 456 360 Table 5.4. Vertex symbols for the independent nodes in 36 Node Vertex Symbol In(RCO2)4 [6.6.6.6.6(2).6(2)] BTC [6(2).6(2).6(2)] BTC [6(2).6(2).6(2)] ST1 [6.6.6.6.6.6.6.6.6.6.6.6.6.6.6.6.6.6.6.6.6.6 .6.6.8.8.8.8.8.8.8.8.8 .8.8.8.*.*.*.*.*.*.*. *.*.*.*.*.*.*.*.*.*.*.*.*.*.*.*.*.*.*.*.*.*.*] Point Symbol: {6^24;8^12;10^24;12^6} ST2 [6.6.6.6.6.6.6.6.6.6.6.6.6(2).6(2).6(2).6(2) .6(2).6(2).8.8.8.8.8.8.8.8.8.8.8.8.8.8.8.8. 8.8.8.8.8.8.8.8.8.8.8.8.8.8.8.8.8.8.8.8.*.*.*.*.*.*.*.*.*.*.*.*] Point Symbol: {6^18;8^36;12^12} The transitivity of this net is [5444] a nd therefore it is comprised of 5 types of nodes, 4 edges, 4 faces, and 4 types of tiles. In many cases the natural tiling that carries a net provides a clearer view of the structure in terms of understanding the connectivity of the cages and/or channels by having a polyhedral illustration. This is not the case however for 36 (Figure 5.9.). Note that TOPOS r ecognizes the BTC ligands as being 2

PAGE 346

309 independent nodes but from a geometrical pers pective they are both triangular BUs and therefore in our assessment we re gard them as being identical ( i.e. quaternary net) with respect to the types of BUs present in 36 Figure 5.9. Tiling representation of 36 5[6^6] + 2[6^9]. 5.2.2.2. Properties: Preliminary Gas Sorption Studies Although compound 36 was not predicted, it cert ainly possesses desirable structural features which offe r much potential from a gas storage point of view. This MOF in fact contains many of the key pa rameters that are suggested to increase Q st; that is, a significant number of accessible open metal sites (indium-based TMBBs), the framework is charged and there by induces an electrostatic fiel d into the cavities, charge balance is provided by polarizable bromide i ons, has a moderate solvent-accessible free volume of 63 %, and the pore size is rather narrow (s ee Table 5.1.). Compound 36 was exchanged in CH3CN for 5 days and its sorption properties assesse d after evacuation at 115oC for 9 h. The Ar sorption studies carried out at 87 K on 36 did not reveal a typical

PAGE 347

310 Type I isotherm, in the sense that it does not reach saturation a nd it displays a small hysteresis in the desorption isotherm (Figure 5.10a.). Figure 5.10. Gas sorption studies for compound 36 : (a) Ar sorption isotherm at 87 K; (b) H2 sorption data at 77 K and 87 K; and (c) Isosteric heats of adsorption for H2. This may be explained by improper sample activation or a heterogeneous sample and current studies are focused on addressing th is issue. The apparent BET and Langmuir surface areas are estimated to be 540 and 748 m2/g with a corresponding pore volume of 0.26 cm3 g-1. Note that obvious structural and chemical differences exist between 36 and (a) (b) (c)

PAGE 348

311 the ctn -MOFs discussed above. Nevertheless, the H2 uptake capacity at 77 K for was found to be similar, 1.2 wt % (Figure 5.10b.). The Q st was also comparable within the range of 7.4 to 5.7 kJ mol-1 at low and high loadings, respectively (Figure 5.10c.). These findings therefore point towards the importa nce of pore size and charged frameworks towards achieving enhanced binding affinities in these materials but more importantly the materials presented herein demonstrate that su ch interactions can be maintained even at higher loadings. 5.3. Experimental Section 5.3.1. Materials and Methods All chemicals and solvents used in the preparation of compounds 31 – 36 were of reagent grade and used w ithout further purification. 5.3.2. Synthesis and Characterization Synthesis of {[(In(BTC)1.33]DMA]}n, (31). A solution of InBr3 (17.7mg, 0.05mmol) and BTC (21.0mg, 0.1 mmol), DMF (1.0mL), CH3CN (0.5mL), and tetraethylammonium chloride (TEACl) (0.1mL, 1M in H2O) were placed in a 20 mL scintillation vial which was sealed, heated to 85oC for 12h, and then cooled to room temperature. Colorless tetrahedral-shape crystals of 31 were harvested and air-dried. The as-synthesized material was dete rmined to be insoluble in H2O and common organic solvents. Synthesis of {[(Ga(BTC)1.33]DMA]} n, (32). A solution of Ga(NO3)3 x H2O (12.8mg, 0.05mmol), BTC (21.0m g, 0.05mmol), DMA (0.5 mL), butanol (1.0 mL), and

PAGE 349

312 TEACl (0.1 mL, 1M in H2O) was prepared in a 20 mL scintillation vial, which was sealed and heated to 85oC for 12h and then to105oC for24h and cooled to room temperature. Colorless tetrahedral-shape crystals of 32 which were found to be insoluble in H2O and common organic solvents, we re harvested and air-dried. Synthesis of {[(Ga(BTC)1.33]DMA]}n, (33). A solution of Ga(NO3)3 x H2O (12.8mg, 0.05mmol), BTC (21.0 mg, 0.05mmol), DMF (0.5 mL ) and butanol (1.0 mL) was prepared in a 20 mL scintillation vial. The solution was then sealed and heated to 85oC for 12h, 105oC for 24h, and 115oC for 24h and cooled to room temperature. Colorless tetrahedral-shape crystals of 33 which were found to be insoluble in H2O and common organic solvents, were harvested and air-dried. Synthesis of {[(Ga(BTC)1.33]DEA]} n, (34). A solution of Ga(NO3)3 x H2O (12.8mg, 0.05mmol), BTC (21.0m g, 0.05mmol), DEF (0.5 mL) and butanol (1.0 mL) was prepared in a 20 mL scintillation vial. The solution was then sealed and heated to 85oC for 12h followed by additional heating to 105oC for 24h and 115oC for 23h. Colorless tetrahedral-shape crystals of 34 which were found to be insoluble in H2O and common organic solvents, were harvested and air-dried. Synthesis of {[(Ga(BTC)1.33]Bu4N]}n, (35). A solution of Ga(NO3)3 x H2O (12.8mg, 0.05mmol), BTC (21.0m g, 0.05mmol), DMSO (0.5 mL ) and butanol (1.0 mL), and tetrabutylammonium bromide (TBABr; 1M in H2O) was prepared in a 20 mL scintillation vial. The solution wa s then sealed and heated to 85oC for 12h, 105oC for 24h, and 115oC for 23h and cooled to room temperature. Colorless tetrahedral-shape crystals of 35 were harvested and air-dried. The as-synth esized material was found to be insoluble in H2O and common organic solvents.

PAGE 350

313 Synthesis of {In30(O)4(BTC)24(HCOO)4Br6}n, (36). A solution of In(NO3)3 x H2O (19.5mg, 0.05mmol), BTC (10.5mg, 0.05mmol), DMF (0.5mL), NMP (0.5mL), tetraethylammonium brom ide (TEABr; 0.3mL, 1M in H2O), and warm HNO3 (0.3mL, 3.5M in DMF) was prepared in a 20 mL scintillation vial. The solution was heated to 85oC for 12h and cooled to room temperature to yield colorless polyhedral crystals. The as-synthesized crystals, 36 were isolated and air-dr ied and determined to be insoluble in H2O and common organic solvents. 5.4. Summary and Conclusions The work conducted in the first part of this chapter focused on the rational synthesis of anionic (3,4)-connect ed nets. Structures possessing ctn and bor topologies were specifically targeted for this study because we anticipated that the frameworks would be robust and contain larg e cavities. The objective of this study was to incorporate extra-framework organic cations of varying si ze and functionality into the cavities as a means to control the pore size and study the effects on adsorbate-MOF interactions. We deliberately selected rigid and directional pr efabricated tetrahedra l and triangular MBBs in the form of [M(RCO2)4] and H3BTC, respectively. Studies aimed at isolating charged MOFs with bor topology were not successful. The latter is a viable synthetic target but I believe that was not successful at finding the appropriate reaction conditions to generate this structure. A series of anionic indiumand ga llium-based 3-periodic MOFs with ctn topology were isolated housing different extr a-framework alkylammonium cations in the cavities, i.e. DMA+, DEA+, TEA+, and Bu4N+. Preliminary metal ion exchange studies

PAGE 351

314 confirmed the feasibility of substituting the organic cations of suitable size for an array of divalent metal ions. A significant challeng e however was encountered when trying to determine the identity of the cation using SC D studies. A high degree of disorder in the cavities made it impossible to confirm wh ich alkylammonium cation was there in instances where more than one cation could be present, i.e. DMA+ or TEA+. To alleviate this issue DMF solvent was avoided from th e reaction conditions and in doing so charged balance could only be provided by the chosen template. Preliminary sorption studies were carried out on the In and Ga ctn -MOFs having either DMA+ or TEA+ cations in the cavities. Note that at this st age it is difficult to make any reliable conclusi ons since it is unknown which counter ion is located in the cavity. The isosteric heats of adsorption nevertheless were similar for both compounds and an interesting feature observed in both was the consistency in Q st at low and high loadings, i.e. ~7 kJ mol-1. This plateau is indicative of a significant averaging of the binding sites in these MOFs which therefore can be attributed to the presence of an el ectrostatic field and the relative pore size. The second part of this chapter revealed a serendipitous discovery that was in fact isolated while trying to synthe size an indium-based MOF with bor topology. The porous MOF has a novel topology and is built-up from four di fferent types of MBBs. Accordingly, we have classified it as a qua ternary MOF and to the best of our knowledge it is unprecedented for MOFs. It is therefore an excellent proof of principle in that complex MOFs can indeed be obtained from simple MBBs and we anticipate that this strategy will lead to other modular structures. In summary, the ctn -MOFs presented in this chapter represent ideal molecular platforms for which systematic adsorbate-MO F interactions can be conducted. Future

PAGE 352

315 studies for this project will focus on comp leting full sorption studies on each of the alkylammonium analogs to assess the contributions of pore size on Q st. Sorption studies can be concurrently carried out on the metal exchanged samples to assess differences in Q st as a function of varying the metal. This ha s the potential to be afford very interesting results owing to the high stability of the ctn -MOFs ( i.e. > 300oC). Moreover, studies can be also be done to investigate the feasibility of isolating isoreticular ctn or borMOFs with enlarged cavities by using H3BTB. This would therefore allow for the encapsulation of even large alkylammonium cati ons and even metal complexes. 5.5. References (1) Nouar, F.; Eckert, J.; Eubank, J. F.; Forster, P.; Eddaoudi, M. J. Am. Chem. Soc. 2009 131 2864-2870. (2) Zhao, D.; Yuan, D. Q.; Zhou, H. C. Energy Environ. Sci. 2008 1 222-235. (3) Banerjee, R.; Phan, A.; Wang, B.; Knobler C.; Furukawa, H.; O'Keeffe, M.; Yaghi, O. M. Science 2008 319 939-943. (4) Zhang, J. J.; Wojtas, L.; Larsen, R. W.; Eddaoudi, M.; Zaworotko, M. J. J. Am. Chem. Soc. 2009 131 17040-17041. (5) Batten, S. R. CrystEngComm 2001 18 1-7. (6) Batten, S. R.; Robson, R. Angew. Chem. Int. Ed. 1998 37 1460-1494. (7) Chen, B. L.; Ma, S. Q.; Hurtado, E. J.; Lobkovsky, E. B.; Zhou, H. C. Inorg. Chem. 2007 46 8490-8492. (8) Ma, S. Q.; Sun, D. F.; Ambrogio, M.; Fi llinger, J. A.; Parkin, S.; Zhou, H. C. J. Am. Chem. Soc. 2007 129 1858-1859. (9) Choi, E. Y.; Park, K.; Yang, C. M.; Kim, H.; Son, J. H.; Lee, S. W.; Lee, Y. H.; Min, D.; Kwon, Y. U. Chem. -Eur. J. 2004 10 5535-5540. (10) Cho, S. H.; Ma, B. Q.; Nguyen, S. T.; Hupp, J. T.; Albrecht-Schmitt, T. E. Chem. Commun. 2006 2563-2565. (11) Kesanli, B.; Cui, Y.; Smith, M. R.; Bittner, E. W.; Bockrath, B. C.; Lin, W. B. Angew. Chem. Int. Ed. 2005 44 72-75. (12) Ma, S. Q.; Eckert, J.; Forster, P. M.; Yoon, J. W.; Hwang, Y. K.; Chang, J. S.; Collier, C. D.; Parise, J. B.; Zhou, H. C. J. Am. Chem. Soc. 2008 130 15896-15902. (13) Liu, Y. L.; Kravtsov, V. C.; Larsen, R.; Eddaoudi, M. Chem. Commun. 2006, 14881490. (14) Dinca, M.; Dailly, A.; Liu, Y.; Br own, C. M.; Neumann, D. A.; Long, J. R. J. Am. Chem. Soc. 2006 128 16876-16883.

PAGE 353

316 (15) Chen, S.; Zhang, J.; Wu, T.; Feng, P.; Bu, X. J. Am. Chem. Soc. 2009 131 1602716029. (16) An, J.; Rosi, N. L. J. Am. Chem. Soc. 2010 132 5578-5579. (17) Dinca, M.; Long, J. R. J. Am. Chem. Soc. 2007 129 11172-11176. (18) Sava, D. F.; Kravtsov, V. C.; Nouar, F.; Wojtas, L.; Eubank, J. F.; Eddaoudi, M. J. Am. Chem. Soc. 2008 130 3768-3770. (19) Mulfort, K. L.; Farha, O. K.; Stern, C. L.; Sarjeant, A. A.; Hupp, J. T. J. Am. Chem. Soc. 2009 131 3866-3867. (20) Himsl, D.; Wallacher, D.; Hartmann, M. Angew. Chem. Int. Ed. 2009 48 46394642. (21) Yang, S.; Lin, X.; Blake, A. J.; Thom as, K. M.; Hubberstey, P.; Champness, N. R.; Schroder, M. Chem. Commun. 2008 6108-6110. (22) Yang, S. H.; Lin, X.; Blake, A. J.; Walk er, G. S.; Hubberstey, P.; Champness, N. R.; Schroder, M. Nat. Chem. 2009 1 487-493. (23) El-Kaderi, H. M.; Hunt, J. R.; MendozaCortes, J. L.; Cote, A. P.; Taylor, R. E.; O'Keeffe, M.; Yaghi, O. M. Science 2007 316 268-272. (24) Delgado-Friedrichs, O.; O'Keeffe, M.; Yaghi, O. M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2006 62 350-355. (25) O'Keeffe, M. Reticular Chemistry Structure Resource. http://rcsr.anu.edu.au/ (26) Liu, Y.; Eddaoudi, M. unpublished results (27) Chui, S. S. Y.; Lo, S. M. F.; Charmant, J. P. H.; Orpen, A. G.; Williams, I. D. Science 1999 283 1148-1150. (28) Ferey, G.; Mellot-Draznieks, C.; Serre, C.; Millange, F.; Dutour, J.; Surble, S.; Margiolaki, I. Science 2005 309 2040-2042. (29) Wang, Z. Q.; Kravtsov, V. C.; Zaworotko, M. J. Angew. Chem. Int. Ed. 2005 44 2877-2880. (30) Lin, Z.-Z.; Jiang, F.-L.; Chen, L.; Yue, C.-Y.; Yuan, D.-Q.; Lan, A.-J.; Hong, M.-C. Cryst. Growth Des. 2007 7 1712-1715. (31) Luo, F.; Che, Y.-x.; Zheng, J.-m. Cryst. Growth Des. 2008 8 176-178. (32) Lu, J. J.; Mondal, A.; Moulton, B.; Zaworotko, M. J. Angew. Chem. Int. Ed. 2001 40 2113-2116. (33) Bradshaw, D.; Prior, T. J.; Cussen, E. J.; Claridge, J. B.; Rosseinsky, M. J. J. Am. Chem. Soc. 2004 126 6106-6114. (34) Forster, P. M.; Kirn, D. S.; Cheetham, A. K. Solid State Sci. 2005 7 594-602. (35) Ockwig, N. W.; Delgado-Friedr ichs, O.; O'Keeffe, M.; Yaghi, O. M. Acc. Chem. Res. 2005 38 176-182. (36) Delgado-Friedrichs, O.; O'Keeffe, M.; Yaghi, O. M. Phys. Chem. Chem. Phys. 2007 9 1035-1043. (37) Friedrichs, O. D.; O'Keeffe, M.; Yaghi, O. M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2003 59 22-27. (38) Delgado-Friedrichs, O.; O'Keeffe, M. Acta Crystallogr., Sect. A: Found. Crystallogr. 2007 63, 344-347. (39) Chen, B. L.; Eddaoudi, M.; Hyde, S. T.; O'Keeffe, M.; Yaghi, O. M. Science 2001 291 1021-1023. (40) Sun, D. F.; Ma, S. Q.; Ke, Y. X.; Collins, D. J.; Zhou, H. C. J. Am. Chem. Soc. 2006 128 3896-3897.

PAGE 354

317 (41) Wang, X. S.; Ma, S. Q.; Yuan, D. Q.; Yoon, J. W.; Hwang, Y. K.; Chang, J. S.; Wang, X. P.; Jorgensen, M. R.; Chen, Y. S.; Zhou, H. C. Inorg. Chem. 2009 48 7519-7521. (42) Dybtsev, D. N.; Chun, H.; Kim, K. Chem. Commun. 2004 1594-1595. (43) McManus, G. J.; Wang, Z.; B eauchamp, D. A.; Zaworotko, M. J. Chem. Commun. 2007 5212-5213. ( 44) Abrahams, B. F.; Batten, S. R.; Hamit, H.; Hoskins, B. F.; Robson, R. Angew. Chem. Int. Ed. 1996 35 1690-1692. (45) Hong, S.; Zou, Y.; Moon, D.; Lah, M. S. Chem. Commun. 2007 1707-1709. (46) Nouar, F.; Eubank, J. F.; Bousquet, T. ; Wojtas, L.; Zaworotko, M. J.; Eddaoudi, M. J. Am. Chem. Soc. 2008 130 1833-1835.

PAGE 355

318 Chapter 6. Conclusions and Prospects for Future Studies 6.1. Summary The research presented in this dissert ation focused upon the rational construction of 3-periodic porous MOMs with a particul ar emphasis on gaining insights into the structure-property relati onships that govern H2-MOF interactions. It is imperative to the advancement of this field to have a deta iled chemical understanding of these binding affinities at the molecular level in order to design potential made-to-order MOMs bearing the desired features that would meet future DoE targets for hydrogen storage in mobile applications. Herein, we have shown that highly porous MOFs may be capable of reaching these targets; however; these capacities are currently only achievable at low temperatures ( i.e. 77 K). Nevertheless, pertinent info rmation can be extrapolated from these studies and several of the necessary comp onents have been identified that lead to improvements in the sorption energetics in these materials. This is certainly a step in the right direction and therefore to achieve the optimal binding energies at working temperatures and pressure it is a general th esis that an ideal physisorbent-based MOM must strike a comprise between having a high surface area, large pore volume, narrow pore dimensions, and low framework density. The MOMs described here were isolat ed by employing a top down design bottom up synthesis approach. Prior the assembly process a collection of pre-designed MBBs was therefore specifically targeted for their re quisite shape, geometry and directionality

PAGE 356

319necessary to augment a given anticipated net. In this context, the notion of design is not an exact science in the sense that a combin ation of building blocks often encodes for more than one net. Accordingl y, these studies also led to se rendipitous discoveries, which are nevertheless interes ting in their own right. Note that functional molecular analogs of such nets were deliberately targeted for their anticipated properties, i.e. exhibit large/small cages or channels, etc The evolution of the MBB approach wa s evidenced whereby the transition from MOFs constructed from single metal ion node s to highly symmetric multinuclear metal clusters was provided. The ability to targets ne ts which are exclusive to a combination of MBBs was a significant achievement and thereb y represents the next hierarchical level ( i.e. supermolecular building blocks) of design because it offers great potential towards prediction and synthesis of resu ltant structures. Briefly, this dissertation has contributed to reticular chemistry in the following ways: (i) An isostructural series of cationic 3-periodic MOFs with soc topology have been deliberately prepared from the self-a ssembly of 6-connected Inand Fe-based 3oxygen centered trigonal prismatic MBBs with a 4-connected recta ngular planar ligand, H4-ABTC. This established a modular platform bearing unique stru ctural and chemical features, i.e. accessible open metal centers a nd narrow pores with higher localized charge density. A systematic gas sorption study complimented by INS studies on selected compounds was conducted to gain in sights into the preferential H2-MOF interactions as a function of varying the metal center ( e.g. In versus Fe), extraframework counter ions ( e.g. NO3 versus Cl-), and polarizability ( e.g. Brversus Cl-). Exceptional H2 uptake capacities at 77 K and 1 atm were reported and the Q st values at low loading revealed a

PAGE 357

320stronger interaction in the case of the Fe analogs as oppose to the In structures. The differences however were not all that substantial (~ 1 kJ mol-1) but it is important to emphasize that these values were even maintained at high loadings for all analogs. This feature is rarely observed in ot her materials in the sense that Q st decreases significantly at higher loadings. (ii) A platform of 4-c onnected 3-periodic neutral MOMs has been generated by employing a pillaring design strategy. R eadily accessible layered structures ( e.g. Kagom and tetragonal lattices) whic h exhibit different pore shapes and sizes were covalently cross-linked using rigid diisophtha late tetracarboxylate pillars ( e.g. H4-ABTC, H4BIPATC, and H4-BAYTC). Accordingly, the pillared Kagom and tetragonal lattices afforded MOMs with nbo and lvt topology, respectively while the assembly of tetrahedral and square BUs led to the formation of the pts net. The former platform, in theory, permits a systematic so rption studies to be conducted whereby the effect of metal cation, pore size and shape on H2-MOF interactions could be independently assessed. The frameworks however were plagued by activati on issues and therefor e further studies are needed to address this issue. (iii) A higher level of structural complexity was delineated through the introduction of SBBs in the form of MOP ( i.e. small rhombihexahedron and cubohemioctahedron), in which case the corres ponding vertex figure re presents a tertiary building unit. The relationship between the SBB and a specific net illustrates the potential to derive pathways for the construction of 3D MOFs from SBBs that when rigidly crosslinked are uniquely suited for a particular net. That such SBBs can be exploited to form MOFs via a crystal engineering approach imp lies that polyhedral SBBs are amenable to

PAGE 358

321crystal design approaches that facilitate th e generation of nets that cannot be readily obtained using traditional MBBs, that the limit of scale has not yet been realized and that a high degree of fine-tunability will be feasible for such materials. Future work will focus upon the properties of the fcc and rht platforms with regards to their gas storage, catalytic and sensing capabilities. (iv) The direct synthesis of a series of porous anionic Inand Ga-based MOFs with ctn topology was reported. Charge balance is provided in the form of various-sized extra-framework organic cations, which were sh own to be readily exchanged for divalent metal ions. Accordingly, this system provide s a pathway to fine-tune the pore metrics of these materials as to influence the gas sorp tion behavior. Unfortunately, due to lack of instrument time we were unable to comple te this study and no correlation between the sorption properties and the size of the or ganic cation was provided. Nevertheless, preliminary gas sorption studi es did reveal constant Q st values estimated to be 7.0 kJ mol-1, at both low and high loadings for the In and Ga-based analogs. These tentative results therefore support the effects and pore size and electrostatic on H2-MOF interactions, as these MOFs do not contain any open metal sites. We anticipate that larger organic cations will lead to enhanced bindi ng affinities. Finally, to the best of our knowledge we reported the first example of a 3-periodic MOF constructed from four different types of MBBs and therefore is rationalized as a quaternary MOF. 6.2. Future Prospects This dissertation has highlighted a handful of extended structures but in the big picture they represent only a diminutive percen tage of the many thousands of structures

PAGE 359

322that have been reported to date. Without a doubt, this field has witnessed tremendous success in a short period of time concerning ad vancements in design, as evidenced by the transition towards targeted functional MOMs with intended properties. If the field continues to expand at this alarming pace and sc ientists continue to be creative then it surely is a conceivable thought that made-t o-order MOMs will be a way of the future. In my personal opinion, future research studies should place a primary focus on conducting a comprehensive evaluation of exis ting structures to understand their potential properties and surface chemistry ( e.g. in-depth adsorption characterization studies), particularly with regards to sample activa tion. Significant advancements in recent years in the computational area would be a nice compliment to aid in these studies. This dissertation is an example of this concept wh ereby a pool of structures was reported but it contains many incomplete studies. In hindsight it would be have much more informative to focus on a particular set of platforms because this would ultimately provide more meaningful information and provide a cleare r vision as to which direction to move forward.

PAGE 360

323Appendix A. Crystal data and structu re refinement for Select Compounds Table A.1. Compound 1 {In(NO3)3 soc -MOF}. Identification Code 1 Empirical formula C24 H21 In3 N4 O22 Formula weight 1061.91 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Cubic, P-43 n Unit cell dimensions a = 22.4567(11) = 90o Volume 11325.0(10) 3 Z, Calculated density 8, 1.246 Mg/m3 Absorption coefficient 1.270 mm-1 Theta range for data collection 2.03o to 25.01o Limiting indices -15 h 15, 0 k 18, 1 l 26 Reflections collected / unique / observed 3352 / 1819 [R(int) = 0.0572] Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 3334 / 6 / 210 Goodness-of-fit on F^2 1.039 Final R indices [ I >2 ( I )] R1 = 0.0680, w R2 = 0.1760 R indices (all data) R1 = 0.0855, w R2 = 0.1844

PAGE 361

324Appendix A (continued) Table A.2. Compound 2 {Fe(NO3)3 soc -MOF}. Identification Code 2 Empirical formula C24 H16.33 Fe3 N4 O19.67 Formula weight 842.95 Temperature 100(2) K Wavelength 0.77090 Crystal system, space group Cubic, P-43 n Unit cell dimensions a = 21.9524(6) = 90o Volume 10579.0(5) 3 Z, Calculated density 8, 1.059 Mg/m3 Absorption coefficient 1.078 mm-1 F(000) 3389 Crystal size 0.01 x 0.01 x 0.01 mm Theta range for data collection 3.77o to 27.55o Limiting indices -6 h 26, -24 k 10, -7 l 25 Reflections collected / unique / observed 12785 / 2945 [R(int) = 0.0700] Completeness to theta = 27.55 99.3 % Max. and min. transmission 0.9893 and 0.9893 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 2945 / 4 / 149 Goodness-of-fit on F^2 1.064 Final R indices [ I >2 ( I )] R1 = 0.0471, w R2 = 0.1265 R indices (all data) R1 = 0.0534, w R2 = 0.1293 Absolute structure parameter 0.08(3) Largest diff. peak and hole 0.532 and -0.608 e.A-3

PAGE 362

325Appendix A (continued) Table A.3. Compound 3 {FeCl3 soc -MOF}. Identification Code 3 Empirical formula C24 H15 Cl Fe3 N3 O16 Formula weight 804.39 Temperature 225(2) K Wavelength 1.54178 Crystal system, space group Cubic, P-43 n Unit cell dimensions a = 21.8720(9) = 90o Volume 10463.2(7) 3 Z, Calculated density 8, 1.021 Mg/m3 Absorption coefficient 7.477 mm-1 F(000) 3224 Crystal size 0.04 x 0.04 x 0.04 mm Theta range for data collection 2.86o to 64.98o Limiting indices -17 h 25, -17 k 7, -24 l 10 Reflections collected / unique / observed 12512 / 2672 [R(int) = 0.0980] Completeness to theta = 64.98 98.0 % Max. and min. transmission 0.7541 and 0.7541 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 2672 / 0 / 147 Goodness-of-fit on F^2 1.016 Final R indices [ I >2 ( I )] R1 = 0.0621, w R2 = 0.1665 R indices (all data) R1 = 0.0917, w R2 = 0.1804 Absolute structure parameter 0.025(14) Largest diff. peak and hole 0.399 and -0.329 e.A-3

PAGE 363

326Appendix A (continued) Table A.4. Compound 4 {InCl3 soc -MOF}. Identification Code 4 Empirical formula C24 H23.70 Cl In3 N3 O20.35 Formula weight 1059.69 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Cubic, P-43 n Unit cell dimensions a = 22.4570(3) = 90o Volume 11325.4(3) 3 Z, Calculated density 8, 1.243 Mg/m3 Absorption coefficient 10.603 mm-1 F(000) 4124 Crystal size 0.10 x 0.10 x 0.10 mm Theta range for data collection 5.57o to 68.14o Limiting indices -24 h 26, -24 k 26, -16 l 23 Reflections collected / unique / observed 27765 / 3412 [R(int) = 0.0688] Completeness to theta = 68.14 99.2 % Absorption correction Semi -empirical from equivalents Max. and min. transmission 0.4169 and 0.4169 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 3412 / 0 / 175 Goodness-of-fit on F^2 1.053 Final R indices [ I >2 ( I )] R1 = 0.0385, w R2 = 0.1116 R indices (all data) R1 = 0.0417, w R2 = 0.1139 Absolute structure parameter 0.000(15) Largest diff. peak and hole 0.906 and -0.637 e.A-3

PAGE 364

327Appendix A (continued) Table A.5. Compound 5 {InBr3 soc -MOF}. Identification Code 5 Empirical formula C24 H25.50 Br In3 N3 O21.25 Formula weight 1120.34 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Cubic, P-43 n Unit cell dimensions a = 22.4530(4) = 90o Volume 11319.4(3) 3 Z, Calculated density 8, 1.315 Mg/m3 Absorption coefficient 11.021 mm-1 F(000) 4340 Crystal size 0.10 x 0.10 x 0.10 mm Theta range for data collection 4.82o to 67.40o Limiting indices -25 h 24, -25 k 18, -15 l 26 Reflections collected / unique / observed 24026 / 3334 [R(int) = 0.0771] Completeness to theta = 67.40 98.5 % Max. and min. transmission 0.4053 and 0.4053 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 3334 / 6 / 182 Goodness-of-fit on F^2 1.076 Final R indices indices [ I >2 ( I )] R1 = 0.0391, w R2 = 0.1121 R indices (all data) R1 = 0.0426, w R2 = 0.1145 Absolute structure parameter 0.002(14) Largest diff. peak and hole 0.816 and -0.523 e.A-3

PAGE 365

328Appendix A (continued) Table A.6. Compound 6 { kgm lattice}. Identification Code 6 Empirical formula C16 H14 Cu2 N2 O10 Formula weight 521.38 Temperature 293(2) K Wavelength 0.71073 Crystal system, space group Hexagonal, P-3 Unit cell dimensions a = 18.521(3) = 90o b = 18.521(3) = 90o c = 10.594(2) = 120o Volume 3147.2(9) 3 Z, Calculated density 3, 0.901 Mg/m3 Absorption coefficient 1.049 mm-1 F(000) 846 Crystal size 0.10 x 0.05 x 0.03 mm Theta range for data collection 1.27o to 25.02o Limiting indices -11 h 18, -16 k 21, -8 l 12 Reflections collected / unique / observed 6115 / 3698 [R(int) = 0.0742] Completeness to theta = 25.02o 99.2 % Max. and min. transmission 0.9692 and 0.9024 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 3698 / 0 / 149 Goodness-of-fit on F^2 0.786 Final R indices [ I >2 ( I )] R1 = 0.0728, wR2 = 0.1550 R indices (all data) R1 = 0.1247, wR2 = 0.1676 Largest diff. peak and hole 0.810 and -0.517 e.A-3

PAGE 366

329Appendix A (continued) Table A.7. Compound 7 { sql lattice}. Identification Code 7 Empirical formula C23 H25 In N4 O11 Formula weight 648.29 Temperature 100(2) K Wavelength 0.71073 Crystal system, space group Monoclinic, C2/c Unit cell dimensions a = 15.533(3) = 90o b = 12.515(2) = 90.254(3)o c = 13.320(3) = 90o Volume 2588.9(9) 3 Z, Calculated density 4, 1.663 Mg/m3 Absorption coefficient 0.981 mm-1 F(000) 1312 Crystal size 0.10 x 0.10 x 0.10 mm Theta range for data collection 2.09o to 25.25o Limiting indices -18 h 17, -5 k 15, -15 l 15 Reflections collected / unique / observed 5129 / 2216 [R(int) = 0.0322] Completeness to theta = 25.25o 95.1 % Max. and min. transmission 0.9083 and 0.9083 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 2216 / 0 / 195 Goodness-of-fit on F^2 1.016 Final R indices [ I >2 ( I )] R1 = 0.0325, w R2 = 0.0834 R indices (all data) R1 = 0.0378, w R2 = 0.0863 Largest diff. peak and hole 0.604 and -0.512 e.A-3

PAGE 367

330Appendix A (continued) Table A.8. Compound 8 {Cu(ABTC) nbo -MOF}. Identification Code 8 Empirical formula C24 H9 Cu3 N3 O15 Formula weight 769.96 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Rhombohedral, R-3 m Unit cell dimensions a = 22.9196(8) = 90o b = 22.9196(8) = 90o c = 19.9060(7) = 120o Volume 9055.8(5) 3 Z, Calculated density 6, 0.847 Mg/m3 Absorption coefficient 1.567 mm-1 F(000) 2286 Crystal size 0.10 x 0.10 x 0.10 mm

PAGE 368

331Appendix A (continued) Table A.9. Compound 9 {Zn(ABTC) nbo }. Identification Code 9 Empirical formula C27.33 H6 N4.33 O10.67 Zn2 Formula weight 696.44 Temperature 293(2) K Wavelength 1.54178 Unit cell dimensions a = 21.8708(2) = 90o b = 19.4600(2) = 99.1640(10)o c = 29.3141(3) = 90o Volume 12317.0(2) 3 Z, Calculated density 12, 1.127 Mg/m3 Absorption coefficient 1.854 mm-1 F(000) 4148 Crystal size 0.10 x 0.08 x 0.08 mm Theta range for data collection 2.35o to 67.41o Limiting indices -25 h 25, -22 k 22, -35 l 33 Reflections collected / unique / observed 106455 / 21700 [R(int) = 0.0715] Completeness to theta = 67.41o 97.9 % Max. and min. transmission 0.8658 and 0.8363 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 21700 / 0 / 847 Goodness-of-fit on F^2 1.047 Final R indices [ I >2 ( I )] R1 = 0.0906, w R2 = 0.2639 R indices (all data) R1 = 0.1045, w R2 = 0.2768 Largest diff. peak and hole 1.298 and -1.630 e.A-3

PAGE 369

332Appendix A (continued) Table A.10. Compound 11 {Cu(BAYTC) nbo }. Identification Code 11 Empirical formula C26 H10 Cu2 O10 Formula weight 609.44 Temperature 293(2) K Wavelength 0.71073 Crystal system, space group Rhombohedral, R-3 m Unit cell dimensions a = 15.527(3) = 90o b = 18.527(3) = 90o c = 54.221(1) = 120o Volume 16117(5) 3 Z, Calculated density 9, 0.565 Mg/m3 Absorption coefficient 0.614 mm-1 F(000) 2736 Crystal size 0.25 x 0.18 x 0.15 mm Theta range for data collection 3.85o to 25.02o Limiting indices -10 h 15, -10 k 21, -62 l 64 Reflections collected / unique / observed 12739 / 3439 [R(int) = 0.1141] Completeness to theta = 25.02o 98.6 % Max. and min. transmission 1.00 and 0.474 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 3439 / 0 / 97 Goodness-of-fit on F^2 1.458 Final R indices [ I >2 ( I )] R1 = 0.1304, w R2 = 0.4094 R indices (all data) R1 = 0.1889, w R2 = 0.4416 Largest diff. peak and hole 1.407 and -0.517 e.A-3

PAGE 370

333Appendix A (continued) Table A.11. Compound 12 {Cu(BIPATC) nbo }. Identification Code 12 Empirical formula C30 H10 Cu2 N2 O14 Formula weight 749.48 Temperature 293(2) K Wavelength 0.71073 Crystal system, space group Monoclinic, C2/m Unit cell dimensions a = 32.461(12) = 90o b = 18.024(6) = 119.964(9)o c = 19.748(7) = 90 o Volume 10010(6) 3 Z, Calculated density 6, 0.746 Mg/m3 Absorption coefficient 0.672 mm-1 F(000) 2244 Crystal size 0.05 x 0.04 x 0.03 mm Theta range for data collection 1.19o to 18.53o Limiting indices -28 h 28, -10 k 16, -17 l 17 Reflections collected / unique / observed 11846 / 3886 [R(int) = 0.1828] Completeness to theta = 18.53 99.1 % Max. and min. transmission 0.9801 and 0.9672 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 3886 / 0 / 149 Goodness-of-fit on F^2 1.028 Final R indices [ I >2 ( I )] R1 = 0.1111, w R2 = 0.2458 R indices (all data) R1 = 0.1703, w R2 = 0.2651 Largest diff. peak and hole 1.036 and -1.285 e.A-3

PAGE 371

334Appendix A (continued) Table A.12. Compound 13 {Cu(BIPATC) lvtMOF}. Identification Code 13 Emperical formula C38 H28 Cu2 N4 O23 Formula weight 1035.72 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Orthorhombic, Imma Unit cell dimensions a = 15.209(1) = 90o b = 36.441(3) = 90o c = 10.324(1) = 90 o Volume 5721.8(9) 3 Z, Calculated density 4, 1.202 Mg/m3 Absorption coefficient 1.542 mm-1 F(000) 2104 Crystal size 0.10 x 0.10 x 0.10 mm Theta range for data collection 4.85o to 67.64o Limiting indices -17 h 13, -41 k 43, -10 l 12 Reflections collected / unique / observed 12397 / 2685 [R(int) = 0.0479] Completeness to theta = 67.64 97.7% Max. and min. transmission 0.8611 and 0.8611 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 2685 / 0 / 213 Goodness-of-fit on F^2 1.102 Final R indices [ I >2 ( I )] R1 = 0.0574, w R2 = 0.1603 R indices (all data) R1 = 0.0705, w R2 = 0.1722 Largest diff. peak and hole 0.631 and -0.424 e.A-3

PAGE 372

335Appendix A (continued) Table A.13. Compound 14 {Yb(ABTC) lvtMOF}. Identification Code 14 Empirical formula C28 H34 Yb2 N8 O18 Formula weight 1116.69 Temperature 298(2) K Wavelength 0.71073 Crystal system, space group Monoclinic, C2/c Unit cell dimensions a = 27.703(1) = 90o b = 9.325(4) = 121.379(2)o c = 21.117(1) = 90o Volume 4657(4) 3 Z, Calculated density 4, 1.678 Mg/m3 Absorption coefficient 4.072 mm-1 F(000) 2280 Crystal size 0.07 x 0.06 x 0.05 mm Theta range for data collection 1.72o to 25.23o Limiting indices -32 h 32, -11 k 11, -13 l 25 Reflections collected / unique / observed 11745 / 4150 [R(int) = 0.1267] Completeness to theta = 25.23 o 98.4 % Max. and min. transmission 1.000 and 0.074 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 4150 / 0 / 272 Goodness-of-fit on F^2 0.905 Final R indices [ I >2 ( I )] R1 = 0.0609, w R2 = 0.1389 R indices (all data) R1 = 0.1331, w R2 = 0.1742 Largest diff. peak and hole 1.496 and -1.876 e.A-3

PAGE 373

336Appendix A (continued) Table A.14. Compound 15 {Er(ABTC) lvtMOF}. Identification Code 15 Empirical formula C31 H34 Er2 N9 O22 Formula weight 1219.19 Temperature 100(2) K Wavelength 0.71073 Crystal system, space group Monoclinic, C2/c Unit cell dimensions a = 27.953(3) = 90o b = 9.299(1) = 121.923(2)o c = 21.288(2) = 90o Volume 4697.0(9) 3 Z, Calculated density 4, 1.724 Mg/m3 Absorption coefficient 3.634 mm-1 F(000) 2380 Crystal size 0.40 x 0.30 x 0.20 mm Theta range for data collection 2.25o to 28.31o Limiting indices -35 h 17, -12 k 12, -23 l 27 Reflections collected / unique / observed 11877 / 4881 [R(int) = 0.0235] Completeness to theta = 28.31o 83.3 % Max. and min. transmission 1.000 and 0.464 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 4881 / 0 / 274 Goodness-of-fit on F^2 1.171 Final R indices [ I >2 ( I )] R1 = 0.0466, w R2 = 0.1354 R indices (all data) R1 = 0.0566, w R2 = 0.1791 Largest diff. peak and hole 2.123 and -1.984 e.A-3

PAGE 374

337Appendix A (continued) Table A.15. Compound 16 {In(ABTC) ptsMOF}. Identification Code 16 Empirical formula C18 H8 In N3 O8 Formula weight 509.09 Temperature 293(2) K Wavelength 1.54178 Crystal system, space group Monoclinic, C2 Unit cell dimensions a = 14.778(3) = 90o b = 13.081(3) = 102.37(3)o c = 21.597(4) = 90o Volume 4078.0(14) 3 Z, Calculated density 4, 0.826 Mg/m3 Absorption coefficient 4.863 mm-1 F(000) 992 Crystal size 0.20 x 0.20 x 0.20 mm Theta range for data collection 2.09o to 49.99o Limiting indices -13 h 14, -10 k 12, -21 l 19 Reflections collected / unique / observed 6487 / 3318 [R(int) = 0.0619] Completeness to theta = 49.99o 94.9 % Max. and min. transmission 0.4430 and 0.4430 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 3318 / 1 / 28 Goodness-of-fit on F^2 1.217 Final R indices [ I >2 ( I )] R1 = 0.1122, w R2 = 0.2762 R indices (all data) R1 = 0.1194, w R2 = 0.2878

PAGE 375

338Appendix A (continued) Table A.16. Compound 17 {In(ABTC) novelMOF}. Identification Code 17 Empirical formula C18 H8 In N3 O8 Formula weight 509.09 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Monoclinic, P21/c Unit cell dimensions a = 9.926(2) = 90o b = 35.737(7) = 98.088(4)o c = 23.401(5) = 90o Volume 8218(3) 3 Z, Calculated density 8, 0.820 Mg/m3 Absorption coefficient 4.826 mm-1 F(000) 1984 Crystal size 0.10 x 0.10 x 0.10 mm Theta range for data collection 3.82o to 67.81o Limiting indices -10 h 11, -41 k 42, -27 l 27 Reflections collected / unique / observed 62551 / 14220 [R(int) = 0.1266] Completeness to theta = 67.81o 95.5 % Max. and min. transmission 0.6440 and 0.6440 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 14220 / 0 / 494 Goodness-of-fit on F^2 0.956 Final R indices [ I >2 ( I )] R1 = 0.0750, w R2 = 0.1976 R indices (all data) R1 = 0.1324, w R2 = 0.2229 Largest diff. peak and hole 1.304 and -1.560 e.A-3

PAGE 376

339Appendix A (continued) Table A.17. Compound 18 {porphyrin-based MOF}. Identification Code 18 Empirical formula C48 H20 Br4 In N4 O9 Formula weight 1231.14 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Orthorhombic, Pbca Unit cell dimensions a = 26.734(17) = 90o b = 13.358(9) = 98.088(4)o c = 33.07(2) = 90o Volume 11810(13) 3 Z, Calculated density 8, 1.385 Mg/m3 Absorption coefficient 6.771 mm-1 F(000) 4776 Crystal size 0.04 x 0.02 x 0.02 mm Theta range for data collection 2.67o to 33.53o Limiting indices -19 h 18, -9 k 9, -23 l 21 Reflections collected / unique / observed 8766 / 2238 [R(int) = 0.1783] Completeness to theta = 33.53o 98.4 % Max. and min. transmission 0.8765 and 0.7734 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 2238 / 0 / 290 Goodness-of-fit on F^2 0.802 Final R indices [ I >2 ( I )] R1 = 0.0545, w R2 = 0.0939 R indices (all data) R1 = 0.1252, w R2 = 0.1103 Largest diff. peak and hole 0.378 and -0.302 e.A-3

PAGE 377

340Appendix A (continued) Table A.18. Compound 19 {Co2(ABTC) ptsMOF}. Identification Code 19 Empirical formula C52 H70 Co4 N6 O26 Formula weight 1430.87 Temperature 100(2) K Wavelength 0.71073 Crystal system, space group Orthorhombic, P212121 Unit cell dimensions a = 13.514(3) = 90o b = 13.766(3) = 90o c = 45.290(8) = 90o Volume 8426(3) 3 Z, Calculated density 4, 1.353 Mg/m3 Crystal size 0.10 x 0.10 x 0.10 mm Refinement method Full-matrix least-squares on F2 Goodness-of-fit on F^2 1.053 Final R indices [ I >2 ( I )] R1 = 0.0365, w R2 = 0.0997

PAGE 378

341Appendix A (continued) Table A.19. Compound 21 {Co(ABTC) ptsMOF}. Identification Code 21 Empirical formula C22 H22 Co N4 O10 Formula weight 561.37 Temperature 293(2) K Wavelength 1.54178 Crystal system, space group Monoclinic, P21/c Unit cell dimensions a = 11.024(4) = 90o b = 22.005(7) = 105.094(6)o c = 9.692(3) = 90o Volume 2270.1(13) 3 Z, Calculated density 4, 1.643Mg/m3 Absorption coefficient 6.541 mm-1 F(000) 1156 Crystal size 0.15 x 0.10 x 0.06 mm Theta range for data collection 4.02o to 38.77o Limiting indices -8 h 8, -17 k 17, -7 l 7 Reflections collected / unique / observed 5015 / 1240 [R(int) = 0.0907] Completeness to theta = 38.77o 97.8 % Max. and min. transmission 0.6949 and 0.4403 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 1240 / 0 / 154 Goodness-of-fit on F^2 1.086 Final R indices [ I >2 ( I )] R1 = 0.0679, w R2 = 0.1596 R indices (all data) R1 = 0.1242, w R2 = 0.1804 Largest diff. peak and hole 0.521 and -0.381 e.A-3

PAGE 379

342Appendix A (continued) Table A.20. Compound 22 {Mn(ABTC) ptsMOF}. Identification Code 22 Empirical formula C19 H13 Mn N4 O9 Formula weight 497.01 Temperature 100(2) K Wavelength 0.71073 Crystal system, space group Monoclinic, C2/c Unit cell dimensions a = 11.178(4) = 90o b = 22.103(7) = 104.284(6)o c = 9.723(3) = 90o Volume 2327.9(13) 3 Z, Calculated density 4, 1.544 Mg/m3 Absorption coefficient 0.632 mm-1 F(000) 1084 Crystal size 0.1 x 0.1 x 0.1 mm

PAGE 380

343Appendix A (continued) Table A.21. Compound 23 {Ni(ABTC) fccMOF}. Identification Code 23 Empirical formula C18 H12 Ni2 N2 O11 Formula weight 549.68 Temperature 100(2) K Wavelength 0.71073 Crystal system, space group Cubic, F m -3 Unit cell dimensions a = 22.340(7) = 90o Volume 30782.0(12) 3 Z, Calculated density 96, 0.660 Mg/m3 Absorption coefficient 0.757 mm-1 F(000) 6125 Crystal size 0.10 x 0.10 x 0.10 mm Theta range for data collection 1.30o to 19.98o Limiting indices -30 h 18, -30 k 30, -30 l 29 Reflections collected / unique / observed 19812 / 1153 [R(int) = 0.1212] Completeness to theta = 19.98 88.5 % Max. and min. transmission 0.9631 and 0.9631 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 1153 / 5 / 62 Goodness-of-fit on F^2 1.111 Final R indices [ I >2 ( I )] R1 = 0.1432, w R2 = 0.3462 R indices (all data) R1 = 0.1575, w R2 = 0.3575 Largest diff. peak and hole 0.887 and -0.851 e.A-3

PAGE 381

344Appendix A (continued) Table A.22. Compound 25 {Cu(PTMOI) rhtMOF}. Identification Code 25 Empirical formula C33 H18 Cu3 O18 Formula weight 893.09 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Cubic, F m -3 m Unit cell dimensions a = 41.4786(3) = 90o Volume 71362.9(9) 3 Z, Calculated density 32, 0.665 Mg/m3 Absorption coefficient 1.115 mm-1 F(000) 14304 Crystal size 0.10 x 0.10 x 0.10 mm Theta range for data collection 1.84o to 58.55o Limiting indices -45 h 33, -45 k 42, -37 l 45 Reflections collected / unique / observed 40041 / 2529 [R(int) = 0.1454] Completeness to theta = 58.55 99.9 % Max. and min. transmission 0.8967 and 0.8967 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 2529 / 0 / 45 Goodness-of-fit on F^2 1.032 Final R indices [ I >2 ( I )] R1 = 0.1172, w R2 = 0.3004 R indices (all data) R1 = 0.1390, w R2 = 0.3213 Absolute structure parameter 0.002(13) Largest diff. peak and hole 0.807 and -0.972 e.A-3

PAGE 382

345Appendix A (continued) Table A.23. Compound 26 {Zn(PTMOI) rhtMOF}. Identification Code 26 Empirical formula C40.50 H18 N2.50 O18 Zn3 Formula weight 1023.68 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Tetragonal, I4/m Unit cell dimensions a = 29.271(2) = 90o b = 29.271(2) = 105.094(6)o c = 41.637(3) = 90o Volume 35674(4) 3 Z, Calculated density 16, 0.762 Mg/m3 Absorption coefficient 1.275 mm-1 F(000) 8200 Crystal size 0.05 x 0.05 x 0.05 mm Theta range for data collection 1.84o to 39.08o Limiting indices -12 h 23, -22 k 23, -22 l 31 Reflections collected / unique / observed 23763 / 4903 [R(int) = 0.0470] Completeness to theta = 39.08o 93.8 % Max. and min. transmission 0.9390 and 0.9390 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 4903 / 0 / 262 Goodness-of-fit on F^2 1.147 Final R indices [ I >2 ( I )] R1 = 0.0877, w R2 = 0.2579 R indices (all data) R1 = 0.0979, w R2 = 0.2721 Largest diff. peak and hole 0.813 and -0.634 e.A-3

PAGE 383

346Appendix A (continued) Table A.24. Compound 29 {Cu(ABPTMOI) rhtMOF}. Identification Code 29 Empirical formula C51 H36 Cu3 N6 O18 Formula weight 1211.5 Temperature 293(2) K Wavelength 1.54178 Crystal system, space group Cubic, F m -3 Unit cell dimensions a = 56.798(9) = 90o Volume 183232(50) 3 Z, Calculated density 32, 1.349 Mg/m3 Absorption coefficient 2.031 mm-1 F(000) 75264 Crystal size 0.10 x 0.10 x 0.10 mm Theta range for data collection 1.56o to 24.69o Limiting indices -40 h 38, -40 k 14, -30 l 28 Reflections collected / unique / observed 24564 / 3128 [R(int) = 0.1550] Completeness to theta = 33.53o 99.9 % Max. and min. transmission 0.8227 and 0.8227 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 3128 / 0 / 116 Goodness-of-fit on F^2 0.876 Final R indices [ I >2 ( I )] R1 = 0.0790, w R2 = 0.1817 R indices (all data) R1 = 0.1614, w R2 = 0.2130 Largest diff. peak and hole 0.144 and -0.174 e.A-3

PAGE 384

347Appendix A (continued) Table A.25. Compound 30 {Mn ast -ZMOF}. Identification Code 30 Empirical formula C86 H120 Mn12 N56 O60 Formula weight 3664.56 Temperature 100(2) K Wavelength 0.71073 Crystal system, space group Cubic, F m -3 Unit cell dimensions a = 26.291(6) = 90o Volume 18172(7) 3 Z, Calculated density 4, 1.313 Mg/m3 Absorption coefficient 0.888 mm-1 F(000) 6999 Crystal size 0.10 x 0.10 x 0.10 mm Theta range for data collection 2.19o to 20.43o Limiting indices -25 h 6, -21 k 13, -3 l 25 Reflections collected / unique / observed 3669/817 [R(int) = 0.04101] Completeness to theta = 20.43o 98.0 % Max. and min. transmission 0.9165 and 0.9165 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 816 / 4 / 103 Goodness-of-fit on F^2 1.245 Final R indices [ I >2 ( I )] R1 = 0.0841, w R2 = 0.2320 R indices (all data) R1 = 0.0952, w R2 = 0.2416 Largest diff. peak and hole 0.339 and -0.464 e.A-3

PAGE 385

348Appendix A (continued) Table A.26. Compound 31 {In(BTC) ctnMOF}. Identification Code 31 Empirical formula C14 H10 In N O8 Formula weight 435.05 Temperature 100(2) K Wavelength 1.54178 Crystal system, space group Cubic, I-43d Unit cell dimensions a = 20.417(2) = 90o Volume 8511.2(17) 3 Z, Calculated density 12, 1.019 Mg/m3 Absorption coefficient 0.857 mm-1 F(000) 2568 Crystal size 0.20 x 0.20 x 0.20 mm Theta range for data collection 3.16o to 24.69o Limiting indices -22 h 7, -16 k 16, -1 l 24 Reflections collected / unique / observed 3692 / 1207 [R(int) = 0.0425] Completeness to theta = 49.99o 99.5 % Max. and min. transmission 0.9192 and 0.9192 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 1207 / 0 / 49 Goodness-of-fit on F^2 1.054 Final R indices [ I >2 ( I )] R1 = 0.0489, w R2 = 0.1255 R indices (all data) R1 = 0.0654, w R2 = 0.1322 Absolute structure parameter 0.03(9) Largest diff. peak and hole 0.421 and -0.289 e.A-3

PAGE 386

349Appendix A (continued) Table A.27. Compound 32 {Ga(BTC) ctnMOF}. Identification Code 32 Empirical formula C14 H12 Ga1 N1 O11 Formula weight 439.97 Temperature 173(2) K Wavelength 1.54178 Crystal system, space group Cubic, I-43d Unit cell dimensions a = 19.982(6) = 90o Volume 7978(4) 3 Z, Calculated density 12, 1.116 Mg/m3 Absorption coefficient 1.077 mm-1 F(000) 2704 Crystal size 0.20 x 0.20 x 0.20 mm Theta range for data collection 2.88o to 20.78o Limiting indices -12 h 3, -8 k 18, -3 l 19 Reflections collected / unique / observed 1403 / 519 [R(int) = 0.1293] Completeness to theta = 20.78o 96.1 % Max. and min. transmission 0.9000 and 0.9000 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 519 / 0 / 61 Goodness-of-fit on F^2 1.160 Final R indices [ I >2 ( I )] R1 = 0.0996, w R2 = 0.1760 R indices (all data) R1 = 0.1430, w R2 = 0.1995 Absolute structure parameter 0.02(10) Largest diff. peak and hole 0.248 and -0.228 e.A-3

PAGE 387

350Appendix A (continued) Table A.28. Compound 36 {In(BTC) novelMOF}. Identification Code 36 Empirical formula C2040 H672 Br48 In240 O1536 Formula weight 81146.26 Temperature 293(2) K Wavelength 0.71073 Crystal system, space group Cubic, Fd-3 m Unit cell dimensions a = 50.546(5) = 90o Volume 129138(22) 3 Z, Calculated density 1, 1.043Mg/m3 Absorption coefficient 1.475 mm-1 F(000) 38640 Crystal size 0.04 x 0.04 x 0.04 mm Theta range for data collection 2.09o to 20.84o Limiting indices -18 h 50, -43 k 40, -27 l 50 Reflections collected / unique / observed 40013 / 3168 [R(int) = 0.1504] Completeness to theta = 20.84o 99.2 % Max. and min. transmission 0.9433 and 0.9433 Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 3168 / 18 / 213 Goodness-of-fit on F^2 1.096 Final R indices [ I >2 ( I )] R1 = 0.0946, w R2 = 0.2226 R indices (all data) R1 = 0.1186, w R2 = 0.2392 Largest diff. peak and hole 1.237 and -1.047 e.A-3

PAGE 388

351Appendix B. Powder X-ray Diffraction (PXRD) Patterns for Select Compounds Figure B.1. PXRD pattern for compound 1 {In(NO3)3 soc -MOF}. Figure B.2. PXRD pattern for compound 2 {Fe(NO3)3 soc -MOF}.

PAGE 389

352Appendix B (Continued) Figure B.3. PXRD pattern for compound 4 {InCl3 soc -MOF}. Figure B.4. PXRD pattern for compound 5 {InBr3 soc -MOF}.

PAGE 390

353Appendix B (Continued) Figure B.5. PXRD pattern for compound 6 {Cu kgm -MOM}. Figure B.6. PXRD pattern for compound 8 {Cu(ABTC) nboMOF}.

PAGE 391

354Appendix B (Continued) Figure B.7. PXRD pattern for compound 9 {Zn(ABTC) nboMOF}. Figure B.8. PXRD pattern for compound 13 {Cu(BIPATC) lvtMOF}.

PAGE 392

355Appendix B (Continued) Figure B.9. PXRD pattern for compound 14 {Yb(ABTC) lvtMOF}. Figure B.10. PXRD pattern for compound 15 {Er(ABTC) lvtMOF}.

PAGE 393

356Appendix B (Continued) Figure B.11. PXRD pattern for compound 16 {In(ABTC) ptsMOF}. Figure B.12. PXRD pattern for compound 18 {porphyrin based MOF}.

PAGE 394

357Appendix B (Continued) Figure B.13. PXRD pattern for compound 23 {Ni(ABTC) fcu -MOF}. Figure B.14. PXRD pattern for compound 24 {Co(ABTC) fcu -MOF}.

PAGE 395

358Appendix B (Continued) Figure B.15. PXRD pattern for compound 25 {Cu(PTMOI) rht -MOF}. Figure B.16. PXRD pattern for compound 26 {Zn(PTMOI) rht -MOF}.

PAGE 396

359Appendix B (Continued) Figure B.17. PXRD pattern for compound 29 {Mn ast -ZMOF}. Figure B.18. PXRD pattern for compound 31 {In ctn -MOF}.

PAGE 397

360Appendix B (Continued) Figure B.19. PXRD pattern for compound 32 {Ga ctn -MOF}.

PAGE 398

361Appendix C. TGA for Select Compounds Figure C.1. Compound 1 {In(NO3)3 soc -MOF}, as-synthesized in DMF. Figure C.2. Compound 1 {In(NO3)3 soc -MOF}, exchanged in CH3CN.

PAGE 399

362Appendix C (Continued) Figure C.3. Compound 2 {Fe(NO3)3 soc -MOF}, as-synthesized in DMF. Figure C.4. Compound 2 {Fe(NO3)3 soc -MOF}, exchanged in CH3CN.

PAGE 400

363Appendix C (Continued) Figure C.5. Compound 3 {FeCl3 soc -MOF}, as-synthesized in DMF. Figure C.6. Compound 3 {FeCl3 soc -MOF}, exchanged in CH3CN.

PAGE 401

364Appendix C (Continued) Figure C.7. Compound 4 {InCl3 soc -MOF}, as-synthesized in DMF. Figure C.8. Compound 4 {InCl3 soc -MOF}, exchanged in CH3CN.

PAGE 402

365Appendix C (Continued) Figure C.9. Compound 5 {InBr3 soc -MOF}, as-synthesized in DMF. Figure C.10. Compound 5 {InBr3 soc -MOF}, exchanged in CH3CN.

PAGE 403

366Appendix C (Continued) Figure C.11. Compound 6 {Cu kgm -MOF}, as-synthesized in DMF. Figure C.12. Compound 7 {In sql -MOF}, as-synthesized in DMF.

PAGE 404

367Appendix C (Continued) Figure C.13. Compound 8 {Cu(ABTC) nbo -MOF}, as-synthesized in DMF. Figure C.14. Compound 9 {Zn(ABTC) nbo -MOF}, as-synthesized in DEF.

PAGE 405

368Appendix C (Continued) Figure C.15. Compound 13 {Cu(BIPATC) lvt -MOF}, as-synthesized in DMA. Figure C.16. Compound 13 {Cu(BIPATC) lvt -MOF}, exchanged in chlorobenzene.

PAGE 406

369Appendix C (Continued) Figure C.17. Compound 14 {Yb(ABTC) lvt -MOF}, as-synthesized in DMF. Figure C.18. Compound 15 {Er(ABTC) lvt -MOF}, as-synthesized in DMF.

PAGE 407

370Appendix C (Continued) Figure C.19. Compound 16 {In(ABTC) pts -MOF}, as-synthesized in DMF. Figure C.20. Compound 18 {porphyrinbased MOF}, as-synthesized in DMF.

PAGE 408

371Appendix C (Continued) Figure C.21. Compound 19 {Co2(ABTC) pts -MOF}, as-synthesized in DMF. Figure C.22. Compound 20 {Cd2(ABTC) pts -MOF}, as-synthesized in DMF.

PAGE 409

372Appendix C (Continued) Figure C.23. Compound 21 {Co(ABTC) pts -MOF} in DMF. Figure C.24. Compound 22 {Mn(ABTC) pts -MOF} in DMF.

PAGE 410

373Appendix C (Continued) Figure C.25. Compound 23 {Ni(ABTC) fcc -MOF}, as-synthesized in DMF. Figure C.26. Compound 24 {Co(ABTC) fcc -MOF}, as-synthesized in DMF.

PAGE 411

374Appendix C (Continued) Figure C.27. Compound 25 {Cu(PTMOI) rht -MOF}, as-synthesized in DMF. Figure C.28. Compound 25 {Cu(PTMOI) rht -MOF}, exchanged in methanol.

PAGE 412

375Appendix C (Continued) Figure C.29. Compound 26 {Zn(PTMOI) rht -MOF}, as-synthesized in DMF. Figure C.30. Compound 27 {Mn(PTMOI) rht -MOF}, as-synthesized in DMF.

PAGE 413

376Appendix C (Continued) Figure C.31. Compound 28 {Co(PTMOI) rht -MOF}, as-synthesized in DMF. Figure C.32. Compound 30 {Mn(ImDC) ast -MOF}, as-synthesized in DMF.

PAGE 414

377Appendix C (Continued) Figure C.33. Compound 31 {In(BTC) ctn -MOF}, as-synthesized in DMF. Figure C.34. Compound 31 {In(BTC) ctn -MOF}, exchanged in CH3CN.

PAGE 415

378Appendix C (Continued) Figure C.35. Compound 32 {Ga(BTC) ctn -MOF}, as-synthesized in DMF. Figure C.36. Compound 32 {Ga(BTC) ctn -MOF}, exchanged in ethanol.

PAGE 416

379 Figure C.37. Compound 36 {In(BTC) novel -MOF}, as-synthesized in DMF.

PAGE 417

About the Author Amy J. Cairns was born in Sydney, Nova Scotia in 1981 and received her BSc in chemistry (with honors) from Saint Mary’s Un iversity in 2004. As undergraduate student she was granted the opportunity to be a resear ch assistant in the la boratory of Dr. Hilary A. Jenkins. From 2001 – 2004 she worked on nume rous projects related to the synthesis and characterization of nitr ogen-based metal-organic ma terials (MOMs). In August 2004, Amy was admitted to the Ph.D. program at USF and joined Dr. Mohamed Eddaoudi’s research group. Her research interests pertain to the desi gn, synthesis, and characterization of functional MOMs with an emphasis on in vestigating their potential utility in pertinent applications ( e.g. gas storage and separation). She is a member of the American Chemical Society (ACS) and as a result of her work and collaborations has co-authored 5 public ations in peer-reviewed journals and has several other manuscripts in preparati on. During her studies at USF she had the opportunity to travel to local, regional, a nd national conferences for which she made a total of 8 oral or poster pr esentations. In 2006, she receiv ed a travel award from the International Center for Materials Research (ICMR) to attend a two week summer school on porous materials at UC Santa Barbara. Sh e also had the chance in 2009 to visit the Institut Laue-Langevin (ILL) in Grenoble, France to conduct inelastic neutron scattering (INS) experiments on two porous metal-or ganic frameworks (MOFs) that she synthesized.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004537
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Cairns, Amy.
0 245
Structural diversity in crystal chemistry :
b rational design strategies toward the synthesis of functional metal-organic materials
h [electronic resource] /
by Amy Cairns.
260
[Tampa, Fla] :
University of South Florida,
2010.
500
Title from PDF of title page.
Document formatted into pages; contains X pages.
502
Dissertation (PHD)--University of South Florida, 2010.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: Metal-Organic Materials (MOMs) represent an important class of solid-state crystalline materials. Their countless attractive attributes make them uniquely suited to potentially resolve many present and future utilitarian societal challenges ranging from energy and the environment, all the way to include biology and medicine. Since the birth of coordination chemistry, the self-assembly of organic molecules with metal ions has produced a plethora of simple and complex architectures, many of which possess diverse pore and channel systems in a periodic array. In its infancy however this field was primarily fueled by burgeoning serendipitous discoveries, with no regard to a rational design approach to synthesis. In the late 1980s, the field was transformed when the potential for design was introduced through the seminal studies conducted by Hoskins and Robson who transcended the pivotal works of Wells into the experimental regime. The construction of MOMs using metal-ligand directed assembly is often regarded as the origin of the molecular building block (MBB) approach, a rational design strategy that focuses on the self-assembly of pre-designed MBBs having desired shapes and geometries to generate structures with intended topologies by exploiting the diverse coordination modes and geometries afforded by metal ions and organic molecules. The evolution of the MBB approach has witnessed tremendous breakthroughs in terms of scale and porosity by simply replacing single metal ions with more rigid inorganic metal clusters whilst preserving the inherent modularity and essential geometrical attributes needed to construct target networks for desired applications. The work presented in this dissertation focuses upon the rational design and synthesis of a diverse collection of open frameworks constructed from pre-fabricated rigid inorganic MBBs (i.e. [M(CO2)4], [M2(RCO2)4], [M3O(RCO2)6], MN3O3, etc), supermolecular building blocks (SBBs) and 3-, 4- and 6-connected organic MBBs. A systematic evaluation concerning the effect of various structural parameters (i.e. pore size and shape, metal ion, charge, etc) on hydrogen uptake and the relative binding affinity of H2-MOF interactions for selected systems is provided.
590
Advisor: Mohamed Eddaoudi, Ph.D.
653
Metal-organic frameworks
Gas storage
Isosteric heats of adsorption
Supermolecular building blocks
690
Dissertations, Academic
z USF
x Chemistry
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4537