USF Libraries
USF Digital Collections

Crystal engineering of molecular and ionic cocrystals

MISSING IMAGE

Material Information

Title:
Crystal engineering of molecular and ionic cocrystals
Physical Description:
Book
Language:
English
Creator:
Ong, Tien Teng
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Amino Acids
Aqueous Solubility
Ellagic Acid
Lithium Diamondoid Metal-organic Materials
Lithium Salts
Dissertations, Academic -- Chemistry, Pharmaceutical -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Solubility enhancement of poorly-soluble active pharmaceutical ingredients (APIs) remains a scientific challenge and poses a practical issue in the pharmaceutical industry. The emergence of pharmaceutical cocrystals has contributed another dimension to the diversity of crystal forms available at the disposal of the pharmaceutical scientist. That pharmaceutical cocrystals are amenable to the design principles of crystal engineering means that the number of crystal forms offered by pharmaceutical cocrystals is potentially greater than the combined numbers of polymorphs, salts, solvates and hydrates for an API. The current spotlight and early-onset dissolution profile ("spring-and-parachute" effect) exhibited by certain pharmaceutical cocrystals draw attention to an immediate question: How big is the impact of cocrystals on aqueous solubility? The scientific literature and in-house data on pharmaceutical cocrystals that are thermodynamically stable in water are reviewed and analyzed for trends in aqueous solubility and melting point between the cocrystal and the cocrystal formers. There is poor correlation between the aqueous solubility of cocrystal and cocrystal former with respect to the API. The log of the aqueous solubility ratio between cocrystal and API has a poor correlation with the melting point difference between cocrystal and API. Structure-property relationships between the cocrystal and the cocrystal formers remain elusive and the actual experiments are still necessary to investigate the desired physicochemical properties. Crystal form (cocrystals, polymorphs, salts, hydrates and solvates) diversity is and will continue to be a contentious issue for the pharmaceutical industry. That the crystal form of an API dramatically impacts its aqueous solubility (a fixed thermodynamic property) is illustrated by the histamine H2-receptor antagonist ranitidine hydrochloride and HIV protease inhibitor ritonavir. For more than a century, the dissolution rate of a solid has been shown to be directly dependent on its solubility, cçterîs paribus. A century later, it remains impossible to predict the properties of a solid, given its molecular structure. If delivery or absorption of an API are limited by its aqueous solubility, aqueous solubility then becomes a critical parameter linking bioavailability and pharmacokinetics of an API. Since the majority of APIs are Biopharmaceutical Classification System (BCS) Class II (low solubility and high permeability) compounds, crystal form screening, optimization and selection have thus received more efforts, attention and investment. Given that the dissolution rate, aqueous solubility and crystal form of an API are intricately linked, it remains a scientific challenge to understand the nature of crystal packing forces and their impact upon physicochemical properties of different crystal forms. Indeed, the selection of an optimal crystal form of an API is an indispensable part of the drug development program. The impact of cocrystals on crystal form diversity is addressed with molecular and ionic targets in ellagic acid and lithium salts. A supramolecular heterosynthon approach was adopted for crystal form screening. Crystal form screening of ellagic acid yields molecular cocrystals, cocrystal solvates/hydrates and solvates. Crystal form screening of lithium salts (chloride, bromide and nitrate salts) afforded ionic cocrystals and cocrystal hydrates.
Thesis:
Disseration (Ph.D.)--University of South Florida, 2011.
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 Tien Teng Ong.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 212 pages.
General Note:
Includes vita.

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-SFE0004876
usfldc handle - e14.4876
System ID:
SFS0028139:00001


This item is only available as the following downloads:


Full Text
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 s2011 flu ob 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004876
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Ong, Tien Teng
0 245
Crystal engineering of molecular and ionic cocrystals
h [electronic resource] /
by Tien Teng Ong.
260
[Tampa, Fla] :
b University of South Florida,
2011.
500
Title from PDF of title page.
Document formatted into pages; contains 212 pages.
Includes vita.
502
Disseration
(Ph.D.)--University of South Florida, 2011.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
520
ABSTRACT: Solubility enhancement of poorly-soluble active pharmaceutical ingredients (APIs) remains a scientific challenge and poses a practical issue in the pharmaceutical industry. The emergence of pharmaceutical cocrystals has contributed another dimension to the diversity of crystal forms available at the disposal of the pharmaceutical scientist. That pharmaceutical cocrystals are amenable to the design principles of crystal engineering means that the number of crystal forms offered by pharmaceutical cocrystals is potentially greater than the combined numbers of polymorphs, salts, solvates and hydrates for an API. The current spotlight and early-onset dissolution profile ("spring-and-parachute" effect) exhibited by certain pharmaceutical cocrystals draw attention to an immediate question: How big is the impact of cocrystals on aqueous solubility? The scientific literature and in-house data on pharmaceutical cocrystals that are thermodynamically stable in water are reviewed and analyzed for trends in aqueous solubility and melting point between the cocrystal and the cocrystal formers. There is poor correlation between the aqueous solubility of cocrystal and cocrystal former with respect to the API. The log of the aqueous solubility ratio between cocrystal and API has a poor correlation with the melting point difference between cocrystal and API. Structure-property relationships between the cocrystal and the cocrystal formers remain elusive and the actual experiments are still necessary to investigate the desired physicochemical properties. Crystal form (cocrystals, polymorphs, salts, hydrates and solvates) diversity is and will continue to be a contentious issue for the pharmaceutical industry. That the crystal form of an API dramatically impacts its aqueous solubility (a fixed thermodynamic property) is illustrated by the histamine H2-receptor antagonist ranitidine hydrochloride and HIV protease inhibitor ritonavir. For more than a century, the dissolution rate of a solid has been shown to be directly dependent on its solubility, cters paribus. A century later, it remains impossible to predict the properties of a solid, given its molecular structure. If delivery or absorption of an API are limited by its aqueous solubility, aqueous solubility then becomes a critical parameter linking bioavailability and pharmacokinetics of an API. Since the majority of APIs are Biopharmaceutical Classification System (BCS) Class II (low solubility and high permeability) compounds, crystal form screening, optimization and selection have thus received more efforts, attention and investment. Given that the dissolution rate, aqueous solubility and crystal form of an API are intricately linked, it remains a scientific challenge to understand the nature of crystal packing forces and their impact upon physicochemical properties of different crystal forms. Indeed, the selection of an optimal crystal form of an API is an indispensable part of the drug development program. The impact of cocrystals on crystal form diversity is addressed with molecular and ionic targets in ellagic acid and lithium salts. A supramolecular heterosynthon approach was adopted for crystal form screening. Crystal form screening of ellagic acid yields molecular cocrystals, cocrystal solvates/hydrates and solvates. Crystal form screening of lithium salts (chloride, bromide and nitrate salts) afforded ionic cocrystals and cocrystal hydrates.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor:
Zaworotko, Michael J.
653
Amino Acids
Aqueous Solubility
Ellagic Acid
Lithium Diamondoid Metal-organic Materials
Lithium Salts
690
Dissertations, Academic
z USF
x Chemistry, Pharmaceutical
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4876



PAGE 1

Crystal Engineering of Molecular and I onic Cocrystals by Tien Teng Ong 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: Michael J. Zaworotko, Ph.D. Ning Shan, Ph.D. rn Almarsson, Ph.D. Shengqian Ma, Ph.D. Date of Approval: March 25, 2011 Keywords: Aqueous Solubility, Ellagic Acid, Lithium Salts, Amino Acids, Lithium Diamondoid Meta l-Organic Materials Copyright 2011, Tien Teng Ong

PAGE 2

Dedication To my dearest parents and esteemed mentor who is apt to j udge and appraise this artefact

PAGE 3

Acknowledgements First and foremost, I will seize this opportunity to express my deepest gratitude to my Ph.D. mentor, Dr. Michael J. Zaworot ko, for providing me with the physical and mental space to nucleate and crystallize into who I am today. I must thank the members of my gra duate committee: Dr. Ning Shan, Dr. rn Almarsson, Dr. Shengqian Ma, for their scientific inputs, guidance and support throughout my doctoral research, and Dr. R. D ouglas Shytle for kindly agreeing to serve as the chairperson of my examination committee. I thank all past and presen t members in the research groups of Dr. Zaworotko and Dr. Eddaoudi for their informative discussion s, especially the cocrystal team whose earlier work highlighted the trials that lie ahead of me. I especia lly like to acknowledge and thank Dr. Kapildev K. Arora, Dr. ukasz Wojtas and Jason A. Perman for imparting their knowledge in the prac tice of single crystal X-ra y crystallography as an indispensable tool in my docto ral research. I than k Dr. Wenge Qiu for lending his ear and more during a testing period of my candidatu re. And I thank Alexander Schoedel for his help in many ways during the prep aration of this dissertation. Thank you all.

PAGE 4

i Table of Contents List of Tables …………………………………………………………………………....vii List of Figures ………………………………………………………………………...….ix Abstract ………………………………………………………………………………….xv Chapter 1. Molecular crystal forms .............................................................................. 1 1.1 Preamble ..................................................................................................... 1 1.2 From molecules to crystals ......................................................................... 2 1.3 Allotropism ................................................................................................. 3 1.4 Polymorphism of molecular crystals .......................................................... 4 1.5 Concomitant polymorphism........................................................................ 7 1.6 Enantiotropy and Monotropy ...................................................................... 9 1.7 Remarks .................................................................................................... 11 Chapter 2. The Impact of Cocrystal on Aqueous Solubility ....................................... 12 2.1 Preamble ................................................................................................... 12 2.2 Biopharmaceutical Classification System (BCS) ..................................... 13

PAGE 5

ii 2.3 Noyes-Whitney Equation .......................................................................... 14 2.4 Thermodynamic versus apparent solubility .............................................. 15 2.5 Solubility and Melting Point ..................................................................... 16 2.6 Trends in Aqueous Solubility of Cocrystals ............................................. 17 2.7 Conclusion ................................................................................................ 36 Chapter 3. Molecular Co crystals: Crystal Forms of Ellagic Acid .............................. 38 3.1 Preamble ................................................................................................... 38 3.1.1 CSD Survey of Ellagic Acid ..................................................................... 44 3.1.2 Cocrystals of Polyphenols with Group 1 Basic Cocrystal Formers (Pyridine, Imidazole, Pyrazole) ................................................................ 45 3.1.3 Cocrystals of Polyphenols with Group 2 Zwitterionic Cocrystal ................ Formers ..................................................................................................... 48 3.1.4 Cocrystals of Polyphenols with Group 3 Neutral Cocrystal Formers (Carbonyl) ................................................................................................. 49 3.2 Results and Discussion ............................................................................. 50 3.2.1 Crystal Forms of Ellagic Acid with Group 1 Basic Cocrystal ..................... Formers (Pyridine, Imidazole, Pyrazole) .................................................. 50 3.2.2 Crystal Forms of Ellagic Acid with Group 2 Zwitterionic Cocrystal Formers ..................................................................................................... 58

PAGE 6

iii 3.2.3 Crystal Forms of Ellagic Acid with Group 3 Neutral Cocrystal .................. Formers (Carbonyl) ................................................................................... 60 3.2.4 Solvates of Ellagic Acid ............................................................................ 62 3.3 Conclusion ................................................................................................ 67 3.4 Materials and Methods .............................................................................. 70 3.4.1 Synthesis of Crystal Forms of Ellagic Acid .............................................. 71 3.4.2 Crystal Form Characterization .................................................................. 74 Chapter 4. Ionic Cocrystals: Cr ystal Forms of Lithium Salts..................................... 82 4.1 Preamble ................................................................................................... 82 4.1.1 Amino acids as the ideal cocrystal formers .............................................. 83 4.1.2 CSD survey of ionic cocrystals of LiX (X = Cl-, Brand NO3 -) and ............ amino acids ............................................................................................... 84 4.1.3 1:2 ionic cocrystals of LiX (X = Cl-, Brand NO3 -) and amino acids ......... as structural analogues of zeolites and silica ............................................ 86 4.2 Results and Discussion ............................................................................. 93 4.2.1 1:1 ionic cocrystals of LiX and amino acids: A discrete ............................. supermolecule and linear chains ............................................................... 94 4.2.2 1:2 ionic cocrystals of LiX a nd amino acids: Square grids, ........................ Diamondoids and a Zeolitic ABW net .................................................... 103 4.3 Conclusion .............................................................................................. 112

PAGE 7

iv 4.4 Materials and Methods ............................................................................ 114 4.4.1 Synthesis of Crystal Forms of Lithium Salts .......................................... 114 4.4.2 Crystal Form Characterization ................................................................ 118 Chapter 5. Conclusions a nd Future Directions ......................................................... 126 5.1 Conclusions ............................................................................................. 126 5.2 Future Directons ...................................................................................... 128 5.3 Crystal Engineering and Crystallization ................................................. 130 References ..................................................................................................... 132 Appendices ..................................................................................................... 159 Appendix 1. Solid state characte rization data for ELANAM .......................... 160 Appendix 2. Solid state characte rization data for ELAINM•1.7H2O .............. 161 Appendix 3. Solid state characteri zation data for ELAINM•4NMP ............... 162 Appendix 4. Solid state characteri zation data for ELAINM•4DMA ............... 163 Appendix 5. Solid state charac terization data for ELACAF•H2O ................... 164 Appendix 6. Solid state characte rization data for ELATPH•0.13H2O ............ 165 Appendix 7. Solid state characteri zation data for ELATPH•2DMA ............... 166

PAGE 8

v Appendix 8. Solid state characterization data for ELADMP .......................... 167 Appendix 9. Solid state characte rization data for ELASAR ........................... 168 Appendix 10. Solid state characterization data for ELADMG .......................... 169 Appendix 11. Solid state characte rization data for ELACAP ........................... 170 Appendix 12. Solid state characteri zation data for ELAURE•2NMP ............... 171 Appendix 13. Solid state characte rization data for ELADMA4 ........................ 172 Appendix 14. Solid state characte rization data for ELADMA2 ........................ 173 Appendix 15. Solid state characte rization data for ELADMSO2 ...................... 174 Appendix 16. Solid state characte rization data for ELANMP2 ........................ 175 Appendix 17. Solid state characte rization data for ELAPG2 ............................ 176 Appendix 18. Solid state characte rization data for LICSAR ............................. 177 Appendix 19. Solid state characte rization data for LINSAR ............................ 178 Appendix 20. Solid state characte rization data for LICDMG ........................... 179 Appendix 21. Solid state characte rization data for LINDMG ........................... 180 Appendix 22. Solid state characteriza tion data for LICBAL-anhydrate ........... 181 Appendix 23. Solid state characte rization data for LICBAL ............................ 182

PAGE 9

vi Appendix 24. Solid state characte rization data for LICABA ............................ 183 Appendix 25. Solid state characte rization data for LICPRO ............................. 184 Appendix 26. Solid state characte rization data for LIBPRO ............................. 185 Appendix 27. Solid state characte rization data for LINPRO ............................ 186 Appendix 28. Solid state characte rization data for LICSAR2 ........................... 187 Appendix 29. Solid state characte rization data for LINBTN2 .......................... 188 Appendix 30. Solid state characte rization data for LICDMG2 ......................... 189 Appendix 31. Solid state characte rization data for LIBDMG2 ......................... 190 Appendix 32. Solid state characte rization data for LICPRO2 ........................... 191 Appendix 33. Solid state characte rization data for LIBPRO2 ........................... 192 Appendix 34. Solid state characte rization data for LINPRO2 ( dia ) .................. 193

PAGE 10

vii List of Tables Table 1-1. Solubility of ritonavir polymorphs at 5C. ................................................. 7 Table 1-2. Melting points and tors ion angles of ROY polymorphs. ............................ 8 Table 2-1. Solubility data for itraconazole cocrystals.b .............................................. 18 Table 2-2. Solubility data for fluox etine hydrochloride cocrystals. ........................... 19 Table 2-3. Solubility data for Pfizer 1 succinic acid cocrystal.b ................................ 20 Table 2-4. Solubility data for norfl oxacin isonicotinamide cocrystal solvate.a ......... 20 Table 2-5. Solubility data for AMG517 cocrystals. ................................................... 21 Table 2-6. Solubility data for et henzamide gentisic acid cocrystal.b ......................... 22 Table 2-7. Solubility data for cocrystals of hexamethylenebisacetamide derivative (A).a .......................................................................................... 23 Table 2-8. Solubility data for lamotrigine cocrystals. ................................................ 24 Table 2-9. Solubility data for cytosine cocrystals.b .................................................... 24 Table 2-10. Solubility data for pterostilbene cocrystals. .............................................. 25 Table 2-11. Molecular structures of pheno lic acids used in this study and their three letter code. ........................................................................................ 26 Table 2-12. Polyphenols used in this study and their three letter code. ....................... 27 Table 2-13. Pyridines, Purines a nd their three letter code. ........................................... 27 Table 2-14. Zwitterions and th eir three letter code. ..................................................... 28 Table 2-15. Molecules with carbonyl groups and their three letter code. .................... 28 Table 2-16. Aqueous solubility and me lting point for target molecules. ..................... 28

PAGE 11

viii Table 2-17. Solubility data for in-house cocrystals. ..................................................... 29 Table 3-1. Molecular Structures of E llagic Acid and Group 1 Basic Cocrystal Formers with their three letter code. ......................................................... 42 Table 3-2. Group 2 Zwitterionic Cocrys tal Formers with their three letter code. .......................................................................................................... 42 Table 3-3. Group 3 Neutral Cocrystal Form ers with their thr ee letter code. ............. 42 Table 3-4. Solvates of Ellagic Acid. ........................................................................... 43 Table 3-5. Selected hydrogen bond parame ters for ellagic acid cocrystals. .............. 76 Table 3-6. Selected hydrogen bond para meters for ellagic acid cocrystal solvates/hydrates. ...................................................................................... 77 Table 3-7. Selected hydrogen bond parame ters for ellagic acid solvates. ................. 78 Table 3-8. Crystallographic data for ellagic acid cocrystals. ..................................... 79 Table 3-9. Crystallographic data for ella gic acid cocrystal solvates/hydrates. .......... 80 Table 3-10. Crystallographic data for ellagic acid solvates. ........................................ 81 Table 4-1. Ionic cocrystals of lithium salts from the CSD. ........................................ 85 Table 4-2. Table of zwitterionic cocrystal formers. ................................................... 93 Table 4-3. Analysis of Li-C-Li angles forming 4-, 6and 8-membered ring motifs. ..................................................................................................... 112 Table 4-4. Selected hydrogen bond para meters for 1:1 ionic cocrystals. ................. 120 Table 4-5. Selected hydrogen bond para meters for 1:2 ionic cocrystals. ................. 122 Table 4-6. Crystallographic data for 1:1 ionic cocrystals. ....................................... 123 Table 4-7. Crystallographic data for 1:1 ionic cocrystals (continued). .................... 124 Table 4-8. Crystallographic data of 1:2 ionic cocrystals. ......................................... 125

PAGE 12

ix List of Figures Figure 1-1. Crystal Packing of Benzamid e (Form I, top left, Form II, bottom left, Form III, bottom right) and the amide dimer supramolecular homosynthon present in all three polymorphs (top right). .......................... 5 Figure 1-2. The conformations of ritonavi r Form I (top) and Form II (bottom). .......... 6 Figure 1-3. Molecular st ructure of ROY. ...................................................................... 8 Figure 2-1. Significance of quadrants Q1, Q2, Q3 and Q4 in log plot of solubility ratios. CCF1 denotes API. CCF2 denotes cocrystal former. ....................................................................................................... 30 Figure 2-2. Solubility plot of in-house cocrystals. ...................................................... 31 Figure 2-3. Solubility plot of literature cocrystals. ...................................................... 32 Figure 2-4. Relationship between cocrystal solubility and cocrystal former solubility with respect to the API.............................................................. 33 Figure 2-5. The relationship between ch ange in melting point (CC mp CCF1 mp) and the log of solubility ra tio of cocrystal and API. CCF1 denote API. ............................................................................................... 35 Figure 3-1. The phenol-carbonyl supramol ecular heterosynthon is featured in the cocrystals of (caffeine) •(pterostilbene) (left) and (carbamazepine)•(pterostilbene) (right). ................................................... 40 Figure 3-2. Crystal forms of ellagic ac id: anhydrate (top left), dihydrate (top right) and tetrakis(pyridin e) solvate (bottom). .......................................... 44

PAGE 13

x Figure 3-3. The phenol-pyridine supram olecular heterosynthon aligns two olefins into a discrete ph otoreactive supermolecule105 (top). A projection of the discrete photoprod uct on the ac plan e (bottom). ........... 46 Figure 3-4. A chain of alternating ca ffeine molecules and methyl gallate molecules in (caffeine)(methyl ga llate) cocrystal (top), alternating pairs of methyl gallate and theophy lline dimer forms a flat sheet in the (theophylline)(methyl gallate) cocrystal (bottom). ............................ 47 Figure 3-5. A 1:1 cocrystal of L-asco rbic acid and sarcosine sustained by carboxylate-hydroxyl supramol ecular heterosynthon. .............................. 49 Figure 3-6. -caprolactam molecules exist as dimers and bridge resorcinol molecules. ................................................................................................. 49 Figure 3-7. The undulating ELANAM ta pes are crosslinked by sandwiching the amide dimer with a pair of ellagic acid molecules. ............................. 51 Figure 3-8. Crystal packing of ELAINM1.7H2O reveals isonicotinamide dimers with lateral hydrogen bonding to phenolic and carbonyl moieties of ellagic acid. Water molecules (shown as large vdw spheres) occupy channels parallel to the crystallogr aphic a axis. ............. 52 Figure 3-9. The supramolecular ta pe in ELAINM has excess hydrogen bond donors from phenolic moieties of ellagic acid and amide dimer of isonicotinamide, attracting the NMP solvent molecules (top) and DMA solvent molecules (bottom). ........................................................... 53 Figure 3-10. Crystal Packing in ELACAFH2O reveals hydrogen bonding between ELA and CAF molecules (top). Water molecules

PAGE 14

xi reinforce the hydrogen bonding in teraction between adjacent sheets (bottom). ......................................................................................... 55 Figure 3-11. A projection of ELATPH0.13H2O on the ac plane (top), a chain is constructed in ELATPH2DMA (bottom). ............................................... 56 Figure 3-12. A discrete assembly is established in ELADMP. ..................................... 57 Figure 3-13. The phenol-carboxylate s upramolecular hete rosynthon (top), stacking interactions in the crysta l packing of ELASAR (bottom) .......... 59 Figure 3-14. The phenol-carboxylate s upramolecular hete rosynthon (left), stacking interactions reinforce the crystal packing forces in ELADMG (right) ...................................................................................... 60 Figure 3-15. A supramolecular heterocat emer is assembled in ELACAP. ................... 61 Figure 3-16. A chain of hydrogen bonded urea molecules occupy channels down the crystallographic a axis in the crystal packing of ELAURE2NMP. ...................................................................................... 62 Figure 3-17. The similarity in crysta l packing between the ELADMA4 solvate and the tetrakis(pyridine) solvate (inset). .................................................. 63 Figure 3-18. The similarity in crysta l packing between the ELADMA2 solvate and the dihydrate (inset)............................................................................ 64 Figure 3-19. DMSO molecules tether ELA molecu les into a straight chain. ................ 65 Figure 3-20. A projection of ELANMP2 on the ac plane. ............................................. 66 Figure 3-21. ()PG molecules align the ellagic acid molecules. ................................... 67 Figure 4-1. A chain of fused six-atom rings in HEFWUK (top) and alternating eight-atom and four-atom rings in ALUNEA (bottom). ........................... 85

PAGE 15

xii Figure 4-2. Secondary building units (SBUs) in zeolites. ........................................... 87 Figure 4-3. Synthetic zeolite based on imidazolates, LTA (top left), RHO, (top right), RHO (BIF-9-Li, bottom left), SOD (bottom right). Yellow spheres highlight the em pty polyhedral cages. ......................................... 89 Figure 4-4. Polymorphs of silica, -cristobalite (top left), -tridymite (top right), -quartz, (bottom left), fa ujasite (bottom right). ............................ 90 Figure 4-5. Distribution of the Li-C-Li angle in crysta l structures that contain lithium cations bridged by carboxylate moieties (data obtained using CSD version 5.31). .......................................................................... 93 Figure 4-6. A chain of fused six-atom rings is observed in LICSAR.......................... 94 Figure 4-7. The nitrate anions coordinate to the lithium cations in LINSAR. ............ 95 Figure 4-8. Li2(DMG)2 clusters are hydrogen bo nded into a 1-periodic structure..................................................................................................... 96 Figure 4-9. The nitrate anions are co ordinated to the lithium cations in LINDMG................................................................................................... 97 Figure 4-10. An undulating chain is constructed by bridging discrete (LiCl)2(H2O)2 clusters with betaine zwitterions. ..................................... 97 Figure 4-11. A chain of fused alternati ng eight-atom rings and four-atom rings in LICBAL-anhydrate. .............................................................................. 98 Figure 4-12. Water molecules coordinate to the lithium cations in LICBAL. .............. 99 Figure 4-13. The chloride anions are c oordinated to the lithium cations in LICABA. ................................................................................................. 100

PAGE 16

xiii Figure 4-14. A chain of fused six-atom rings in LICPRO (top) and LIBPRO (bottom)................................................................................................... 101 Figure 4-15. An adamantane cage (left) is constructed with L-proline zwitterions and nitrate anions as bridging lig ands, hexagonal channels in the dia net (right) are occupied by the pyrrolidinium ring of the zwitterions to mitigate interpenetration. ................................................. 102 Figure 4-16. Structural dive rsity in the networks form ed by 1:2 cocrystals of lithium salts with amino acids: (a) square grids; (b) diamondoid LiDMOM nets; (c) zeolitic ABW topology in the first LiZMOM. Hydrogen atoms and counteranions are omitted for clarity. ................... 104 Figure 4-17. Undulating square grid (lef t) and bilayered p acking arrangement (right) in LICSAR2. ................................................................................ 104 Figure 4-18. Undulating square grid (lef t) and bilayered p acking arrangement (right) in LINBTN2. ................................................................................ 105 Figure 4-19. Hexagonal channels in dia nets of LICDMG2 (left) and LIBDMG2 (right). ..................................................................................................... 106 Figure 4-20. Hexagonal channels in dia nets of LICPRO2 (left) and LIBPRO2 (right). ..................................................................................................... 108 Figure 4-21. Hexagonal channels in LINP RO2 (dia, top) and octagonal channels in LINPRO2 (ABW, bottom). ................................................................. 109 Figure 4-22. Significant expansion of dime nsions occurs because of the ditopic carboxylate linker in the ABW-t ype LiZMOM LINPRO2 when compared to prototypal ABW zeolite. .................................................... 111

PAGE 17

xiv Figure 5-1. Triclinic modi fication of quinhydrone. ................................................... 128 Figure 5-2. Carbonyl as the hydrogen bond acceptor in the cocrystal of diphenylamine and benzophenone. ......................................................... 129

PAGE 18

xv Abstract Solubility enhancement of poorly-sol uble active pharmaceutical ingredients (APIs) remains a scientific challenge and poses a practical issue in the pharmaceutical industry. The emergence of pharmaceutical cocr ystals has contributed another dimension to the diversity of crystal forms available at the disposal of the pharmaceutical scientist. That pharmaceutical cocrystals are amenable to the design principles of crystal engineering means that the number of crysta l forms offered by pharmaceutical cocrystals is potentially greater than the combined num bers of polymorphs, salts, solvates and hydrates for an API. The current spotlight and early-onset dissolution profile (“springand-parachute” effect) exhibited by certain pharm aceutical cocrystals draw attention to an immediate question: How big is the impact of cocrystals on aque ous solubility? The scientific literatu re and in-house data on pharmace utical cocrystals that are thermodynamically stable in water are review ed and analyzed for trends in aqueous solubility and melting point between the cocr ystal and the cocrystal formers. There is poor correlation between the a queous solubility of cocrysta l and cocrystal former with respect to the API. The log of the aqueous solubility ratio between cocrystal and API has a poor correlation with the melting point difference between cocrystal and API. Structure-property relationships between the cocrystal and th e cocrystal formers remain elusive and the actual experiments are stil l necessary to investigate the desired physicochemical properties. Crystal form (cocrystals, polymorphs, salts, hydrates and solvates) diversity is and will continue to be a contentious issue for the pharmaceutical industry. That the crystal

PAGE 19

xvi form of an API dramatically impacts its aqueous solubility (a fixed thermodynamic property) is illustrated by the histamine H2-receptor antagonist ranitidine hydrochloride and HIV protease inhibitor ritonavir. For more than a century, the dissolution rate of a solid has been shown to be directly dependent on its solubility, c ter s paribus A century later, it remains impossible to predict the properties of a solid, given its molecular structure. If delivery or absorption of an API are limited by its aqueous solubility, aqueous solubility then becomes a criti cal parameter linking bioavailability and pharmacokinetics of an API. Since the ma jority of APIs are Biopharmaceutical Classification System (BCS) Class II (low solubility and high permeability) compounds, crystal form screening, optimization and sel ection have thus received more efforts, attention and investment. Give n that the dissolution rate, aqueous solubility and crystal form of an API are intricatel y linked, it remains a scientific challenge to understand the nature of crystal packing forces and thei r impact upon physicoche mical properties of different crystal forms. Indeed, the selection of an optimal crystal form of an API is an indispensable part of the drug development pr ogram. The impact of co crystals on crystal form diversity is addressed with molecular a nd ionic targets in ella gic acid and lithium salts. A supramolecular heterosynthon approach was adopted for crystal form screening. Crystal form screening of ellagic acid yields molecular cocrystals, cocrystal solvates/hydrates and solvates. Crystal form sc reening of lithium salts (chloride, bromide and nitrate salts) afforded ionic co crystals and cocr ystal hydrates.

PAGE 20

1 Chapter 1. Molecular crystal forms “…, crystals should not be rega rded as chemical graveyards.”1 Michael J. Zaworotko 1.1 Preamble Scientists have a natural inclination to correlate structure with properties. It is owing to the curiosity (and I must add pers istence) of the indi vidual to make new discoveries and relate them to what is known. I was trying to recall my first chemistry lesson ever but strangely enough, I could not. It must have been a lesson on the periodic table. Afterall, the periodic table provide s a systematic classi fication of all known elements into different horizontal rows (per iods) and vertical colu mns (groups) to help chemists to rationalize the properties of these elements as atoms. In short, the periodic table is useful in correlating the electronic configur ation (electronic structure) of atoms with their properties. Dmitri Ivanovich Me ndeleev is credited for creating the first version of the periodic table.2 With the table, he was su ccessful in pr edicting the properties of elements that have yet to be discovered and was honoured with the radioactive “mendelevium”, name d after him. Early scientists have demonstrated success in establishing structure-prope rty relationships in atoms. Linking atoms together, we can have ions or molecules. According to the Oxford English Dictionary,3 a molecule is ( 1 ) one of the minute discre te particles of which material substances were thought to be composed; ( 2 ) the smallest unit of a chemical compound that can take part in the react ions characteristic of that compound; ( 3 ) a group

PAGE 21

2 of atoms chemically bonded together and actin g as a unit. One may think of a molecule as an assembly of atoms held together by an attractive force. Following the introduction to acids, bases and general organic chemis try, the floodgates were opened to a massive amount of information on the states of matte r. The kinetic theory of gases assumes negligible interactions between gaseous molecules.4 Liquids and solids enter the picture when the interactions between molecules ar e significant and cannot be ignored. Moving along a homologous series of alkanes, gase s, liquids and solids are encountered, characterized by their boili ng points or melting points.5-6 In a way, this collection of data enabled us to associate structural features (o f the molecules) with properties. There is an urge to analyze for trends in the physical properties across the same class of compounds and correlate the molecular stru cture with their properties. When the intermolecular forces of a ttraction are sufficien tly strong, molecules aggregate into a solid. The controlled aggreg ation in an organized mode results in a crystalline solid. How much do we know about the solid as a crystalline material? 1.2 From molecules to crystals “The crystal is … the supermolecule par excellence.”7 A crystal is a material made up of atoms, molecules or ions that are organized in an orderly periodic arrangement that repeats in three spatial directions. “While analyzing crystal structures in terms of intermolecular interactions, one recognizes certain repetitive structural units These are typically small in size and are com posed of specific mol ecular functionalities and their resulting interactions arranged in specific ways The identification of these structural units is somewhat subjective but they are important in that they may be

PAGE 22

3 associated with particular cr ystal packing characteristics.”8 Thus molecules recognize themselves in the solid state and organize themselves through molecular recognition events so that one can think of the smallest repeating unit as the unit cell The propagation of the unit cell along the three cr ystallographic axes illustrates a crystal. However, the same molecules can organize themselves with different molecular recognition events to achieve another crystal. This gives rise to polymorphism. McCrone describes a polymorph9 as “a solid crystalline phase of a given compound resulting from the possibility of at least two different ar rangements of the molecules of that compound in the solid state.” 1.3 Allotropism The phenomenon and consequences of at l east two different arrangements of the constituents of the same compound in the soli d state, is perhaps best illustrated by the element carbon.10 Carbon exists in the form of diamond, graphite, lonsdaleite, C60 11 (buckminsterfullerene or buckyball), C540, C70, single-walled carbon nanotube and amorphous carbon. In diamond, each carbon atom is bonded to four other carbon atoms covalently in a tetrahedral arrangement. The overall structure is a 3-periodic diamondoid ( dia ) network of carbon atoms featuring six-me mbered carbon rings in the stable chair conformation that results in the hardness of diamond.12 This hardness makes diamond an excellent abrasive and used industrially in cutting and drilling ope rations. In graphite, each carbon atom is covalently bonded to th ree other carbon atoms in a plane, forming a (6,3) honeycomb net. Each carbon atom has f our valence electrons and contributes one electron to a deloca lized system of -electron density above and below the plane,

PAGE 23

4 allowing these delocalized electrons to move freely throughout the plane. Hence graphite becomes a conductor of electri city along the planes of the carbon atoms. The layer structure of graphite13 enables it as a lubricant. Using elemental carbon as an example, diamond and graphite bring forth the contrast in properties genera ted by their distinct spatial arrangements (crystal packing) of the carbon atoms. 1.4 Polymorphism of molecular crystals The first example of polymorphism of molecular compounds was reported by Liebig and Whler14 as early as 1832 when they described the slow cooling of a “boiling hot” aqueous solution of benzamide which produ ced white silky needles that transformed to rhombic crystals. They were alerted by th e changes in the morphology of the crystal. The single crystal X-ray structure of benzamide (Form I)15 was reported in 1959. David16 and Blagden17 pursued the structure solution of benzamide (Form II) using powder X-ray diffraction techinques in 2005 until Form III18 was reported four years later. The polymorphism of benzamide was identified by the keen eye for detail of Liebig and Whler but the lack of contro l in polymorphic selectivity from a crystallization process had been a barrier in the solid stat e characterization of the polymorphs.

PAGE 24

5 Figure 1-1. Crystal Packing of Benzamide (Form I, top left, Form II, bottom left, Form III, bottom right) and the amide di mer supramolecular homosynthon present in all three polymorphs (top right). In the crystal structures of benzamide Fo rms I, II and III, a supramolecular tape established by the lateral stacking of the amide dimers, is present in all three polymorphs. However, the relative orientation of these supramolecular tapes is different amongst the three polymorphs and hence the cause of the polymorphic behaviour of benzamide. The meticulous efforts to unravel the polymorphs of a simple molecule in benzamide, highlights the alertness of the individua l and the challenges in understanding and controlling the crystal lization conditions.

PAGE 25

6 Figure 1-2. The conformations of riton avir Form I (top) and Form II (bottom). 166 years later, the lack of contro l in polymorphic selectivity from a crystallization process was thrust onto centr e stage. Ritonavir (HIV-1 protease inhibitor) is a Biopharmaceutical Classification System19 (BCS) Class 4 drug with low solubility and low permeability. One crystal form of ritonavir20 (Form I) was known since its development and production of 240 lots of No rvir capsules without the problem of phase transformation. Alarm was raised when severa l lots of the capsules failed a series of

PAGE 26

7 dissolution experiments, powder X-ray diffr action revealed the appearance of a polymorph (Form II). The crystal structures21 of Forms I and II reveal the compound exists in different conformations around the carbamate in the solid state. The subtle change in the conformation results in crystals with a decrease in aqueous solubility. In addition, the presence of Form II in the or iginal capsule formulation decreases the bioavailability. The emergence of undesire d Form II had an instant impact on the manufacturing and clinical fronts. Form II appeared according to Ostwald’s Law of Stages but there was no crystallization strate gy and lack of knowledge at that time to prevent its formation. Table 1-1. Solubility of rito navir polymorphs at 5C. EtOH/H2O 99/1 75/25 Form I (mg/mL) 90 170 Form II (mg/mL) 19 30 In hindsight, polymorphism is not the pr oblem, the issue lies in the lack of 1) control of the crystallization cond itions and 2) a proactive approach22 to explore the crystal form landscape of the API in the marketed product. 1.5 Concomitant polymorphism To understand the relationship between structure and properties of crystals, polymorphic systems are ideal since any changes in the properties of the crystal form is invariably due to its structure (crystal packing), c ter s paribus Indeed, the intrigue of polymorphism is best captured in 5-methyl-2-[(2-nitrophenyl)amino]-3thiophenecarbonitrile,23 also known as ROY for its red, orange and yellow crystals. To

PAGE 27

8 date, it has ten polymorphs (1. R, 2. Y, 3. ON, 4. OP, 5. YN, 6. ORP, 7. RPL, 8. Y04, 9. YT04 and 10. R05, numbered in order of disc overy and named in terms of colour and morphology). Of these, seven polymorphs have been characterized by single crystal Xray crystallography at ambient conditions (20C to 23C) to elucidat e the origins of its polymorphic behaviour and difference in co lour. X-ray analysis of ROY polymorphs highlighted the torsion angle as an important variable, pr oducing distinct conformations of ROY in the crystal packing. -conjugation increases between the thiophene and benzene rings as the torsion angle nears zero, resulting in a red shift of the visible absorption spectrum. Figure 1-3. Molecular structure of ROY. Table 1-2. Melting points and torsion angles of ROY polymorphs. Form R Y ON OP YN ORP YT04 mp (C) 106.2 109.8 114.8 112.7 99 97 106.9 () 21.7 104.7 52.6 46.1 104.1 39.4 112.8 The pure melt of ROY yields several polymorphs simultaneously upon cooling to near room temperature. Mixtures of polymorphs are also obtained from crystallization in methanol. All known polymorphs of ROY ha ve been crystallized near ambient N N O O H S CN

PAGE 28

9 conditions. The concomitant polymorphism24 of ROY makes it a challenging model system to understand the relationship be tween nucleation and crystal growth of polymorphs in general. Furthermore, their ki netic stability (at ambient conditions) allows them to be isolated and studied for their st ability relationships. Th e relative Gibbs free energy of seven ROY polymorphs were calcula ted using data collected from differential scanning calorimetry experiments25-26 to plot an energy versus temperature diagram and reveal the stability order of ROY polymor phs and their monotropi c or enantiotropic relationships.27-28 The goal is to dictate polymorphic selectivity through a crystallization process and ultimately gain the control of selective crystallization of the desired polymorph in a complicated system. 1.6 Enantiotropy and Monotropy The relative stability of two polymorphs depends on their Gibbs free energies. G = H – TS (G = Gibbs free energy, H = enthalp y, T = temperature in K, S = entropy) When the stability order of polymorphs is invariant of temperature, the polymorphs are monotropic. When the stab ility order of polymorphs changes with temperature, they become enantiotropic pol ymorphs. The knowledge of the enantiotropic or monotropic relationship between different polymorphs requires the generation of energy versus temperature diagram of the pol ymorphic system. This diagram can be used

PAGE 29

10 as a map to direct the crystallization pr ocess to achieve a de sired polymorph at the expense of the undesired polymorph. For a dimo rphic system, there ar e four possibilities: 1) The desired thermodynamically stable form is in a monotropic system. Phase transformation will not occur to generate another form hence precautions are not necessary to guard agains t such a transformation. 2) The desired thermodynamically stable form is in an enantiotropic system. Precautions must be exercised to maintain the thermodynamic conditions at which the Gibbs free energy of the desired polymorph is lower than that of the other. 3) The desired thermodynamically metastable form is in a monotropic system. A kinetically controlled phase transformati on will occur to generate the undesired thermodynamically stable form. Precautions must be exercised to retard the kinetics of the transformation by employing extreme conditions such as very low temperatures, very dry conditi ons and storage in the dark. 4) The desired thermodynamically metastable form is in an enantiotropic system. The generation of the energy versus temperatur e diagram of the polymorphic system will reveal the temperature range at which the Gibbs free energy of the desired form is lower than that of the other, to ob tain and maintain this desired form.

PAGE 30

11 1.7 Remarks “There are many mysteries of nature that we have not solved. Hurricanes, for example, continue to occur and often cause massive devastation. Meteorologists cannot predict months in advance when and with what velocity a hurricane will strike a specific community. Polymorphism is a parallel phenomenon. We know that it will probably happen. But not why or when. Unfortunately, there is nothing that we can do today to prevent a hurricane from st riking any community or poly morphism from striking any drug.”29 Dr. Eugene Sun This underscores the necessity for due diligence in solid form screening.30

PAGE 31

12 Chapter 2. The Impact of Cocrystal on Aqueous Solubility 2.1 Preamble Solubility enhancement31-32 of poorly-soluble APIs33-34 remains a scientific challenge and poses a practical issue in the pharmaceutical industry. The emergence35 of pharmaceutical cocrystals has contributed another dimension to the diversity of crystal forms available for pharmaceutical development. The long-standing reliance of the pharmaceutical industry and regulatory authorit ies on crystal forms and the “spring-andparachute”36-type dissolution profile uniquely e xhibited by certain pharmaceutical cocrystals draw attention to an immediate question: How big is the impact of cocrystals on aqueous solubility? The scientific litera ture and in-house data on pharmaceutical cocrystals that are thermodynamically stable in water are reviewed and analyzed for trends in aqueous solubility and melting po int between the cocrystal and the cocrystal formers. There is poor correlation between the aqueous solubility of cocrystal and cocrystal former with respect to the API. Th e log of the aqueous solubility ratio between cocrystal and API has a poor correlation with the melting point difference between cocrystal and API. Structure-property re lationships between the cocrystal and the cocrystal formers remain elusive and the actual experiments are still necessary to investigate the desired physicochemical properties.

PAGE 32

13 2.2 Biopharmaceutical Classification System (BCS) A critical parameter deciding the fate of any API is its aqueous solubility. Hence to increase drug solubility while maintaining a st able solid form is often an integral part of a pharmaceutical development process. This goal is imperative since solubility and permeability are the two criteria used to illust rate oral absorption of a drug according to the BCS.19 BCS class II (low solubility, high permeability) drugs are limited in oral absorption by their solubility. This class of drugs is currently estimated to account for about 30% of commercial and developmental drugs37. BCS class I (high solubility, high permeability) is the ideal scenario for any API to be in. Increased solubility can result in a significant improvement of oral absorption, leading to higher bioavailability, pushing a drug from BCS class II to BCS cl ass I. This can be achieved via pharmaceutical cocrystals38, which have emerged and establishe d itself as a promising methodology to modify important physicochemical properties of poorly water soluble drugs, such as solubility, dissolution rate and bioavailabil ity, to the extent that the pharmaceutical industry cannot afford not to explore. Ind eed crystal form scre ening of APIs have traditionally been restricted to polymorphs salts and solvates (including hydrates). Pharmaceutical cocrystals are also possible for nonionizable39, weakly acidic40 and weakly basic41 drugs which prevent and restrict the opportunity of a salt screen respectively, and present a viable option for solubility enhancement without changing the molecular structur e of the drug.

PAGE 33

14 2.3 Noyes-Whitney Equation The Noyes-Whitney equation42-43 dictates the dissolution pr ofile of all substances, at least until the dissolution of cocrystals was reported. It states “the rate at which a solid substance dissolves in its own solution is proportional to the difference between the concentration of that solution and the concen tration of the saturated solution”. This has practical relevance for the dissolution profile. A solid with high solubility will boast a high initial rate of dissolution. A solid with low solubility will exhibi t a low initial rate of dissolution and take a longer time to reach ma ximum concentration. There will not be a solid with low solubility but demonstrat e a high initial rate of dissolution or vice versa With cocrystals, there can be a sudden spik e in concentration (“ spring-and-pa rachute” effect) of the API followed by a precipitous dip to a level sim ilar to that of the pure API. This is achieved by a poorly wate r soluble antidepressant Prozac.36 The powder dissolution profiles of cocrys tals of fluoxetine hydrochlorid e were measured and each cocrystal has its unique dissolu tion properties. Obviously, ther e are several controls such as temperature, stirring rate, particle size distribution (PSD), volume and type of dissolution media. These factors are meant to mitigate the effects of diffusion phenomenon and surface area. The solubility of cocrystals had been reported in a variety of dissolution media, including 0.1 N HCl, fasted simulated intestinal fluids (FaSIF) and water. The dissolution media and context us ed for performing dissolution experiments can also change the outcome of the experi ment. This was exemp lified by the celecoxib44 nicotinamide cocrystal. Most studies reporte d the dissolution profile over time. Particle size control via sieving of samples was reported o ccasionally. There was also no mention of controls in some reports. This highlights the wide range of experimental parameters

PAGE 34

15 that are adopted in solubility experiments which can be adju sted to achieve the necessary information. 2.4 Thermodynamic versus apparent solubility There are several pointers to consid er when consulting solubility data45. The first is equilibrium versus apparent solubility measur ements. Apparent solubility measurements are approximate data points usually based on one measurement at one time point. It is unknown if the solid material is thermodynamically stable in the system or if the system has reached equilibrium in the time frame of the experiment unless additional experiments are conducted to verify the phase pu rity of the material before and after each dissolution experiment. For equilibrium solubility, multiple data points are recorded at different time points to ensure the system has attained equilibrium, by observation of a plateau in the concentration-time plot, which is also referred to as a powder dissolution profile. Powder dissolution rates can be infl uenced by PSD since a smaller PSD affords a greater surface area in contac t with the dissolution medium, given the same mass of material. The next factor is phase change during the experiment, as mentioned earlier about thermodynamic stability. When there is a change in the phase purity of the solid material, the resultant solubility data may be irrelevant to the starting material in the experiment. Phase transformation can be de tected by powder X-ray diffraction (PXRD) and differential scanning calor imetry (DSC). They are t ypically indicated by the crystallization of a less soluble material afte r a dramatic plunge in concentration during the collection of solubility measurements at various time points. Hence the maximum

PAGE 35

16 solubility observed over a certain time pe riod is reported along with the time of occurrence but this is not an equilibrium sol ubility measurement. An analysis of the solid material remaining after the experiment is required to confirm any phase transformation. Solubility studies had been conducted to correlate the aqueous solubilities of cocrystals and their respective cocrystal form ers. Success in these studies will pave the way to a solubility-guided strategy46 towards cocrystal screen ing. Alternatively, the origin of pharmaceutical cocrystals47 is the product of an approach based on the understanding and implementatio n of supramolecular synthons.48 A solubility-guided strategy towards cocrystal screening for solubili ty enhancement is indeed very attractive. Should aqueous solubility become a transferab le property, all cocrystal former libraries will only include those with high aqueous so lubility and significantly reduce the number of cocrystal screen ing experiments. 2.5 Solubility and Melting Point Solubility is a fixed thermodynamic property of a crystal form. From a thermodynamic standpoint, the Gibbs free ener gy of solution is the summation of the lattice energy and solvation energy in an aqueous system. Going from a pure crystalline API to a cocrystal, the Gibbs free energy of solution is obviously going to change since the cocrystal is a new composition of matter and not a physical mixture of two solids. The hydration energy of both components (as a cocrystal or a physical mixture) is assumed to be a constant as they retain their molecular integrity in solution. But the lattice energy changes since the cocrystal is a new crystalline material. A stable crystal

PAGE 36

17 form with high lattice energy is unlikely to disintegrate and dissolve in solution easily. Likewise, a less stable crystal form with low lattice energy is more likely to disintegrate and dissolve in solution readily. Since melting point is a direct manifestation of the lattice energy and also an experimentally verifiable physical property of a crystalline solid that is easily obtainable, melting point may have a better correlation with the experimentally determined solubility measurements. In fact if the inverse relationship between lattice energy and thermodynamic solubility is a simple one, solubility enhancement of an API upon cocrystal formation should be accompan ied by melting point depression, observable by the difference in melting point be tween the cocrystal and the API. To search for trends in solubilities of cocrystals, we undertook the endeavor to investigate the aqueous solubilities of cocr ystals, eliminating systems where a solventmediated phase transformation occurred. 2.6 Trends in Aqueous Solubility of Cocrystals Dissolution experiments of crystalline itraconazole (95% of all crystalline particles < 10 m), commercial Sporanox beads (amorphous itra conazole) and three itraconazole cocrystals (succinic acid, L-malic acid and L-tartaric acid) were conducted in solutions of 0.1 N HCl at 25C. All the cocrystals de monstrated higher solubility than crystalline itraconazole. The solubility of the succinic acid cocrystal was about 2 10-4 M, while the solubilities of the L-malic acid and L-tartaric acid cocrystals were about 7 10-4 M which rivals the commercial amorphous Sporanox formulation. These cocrystals recorded a sustained increase in concentration ranging from 4to 20-fo ld higher than that

PAGE 37

18 of crystalline itraconazole49. This provides an early example of the solubility improvements that cocrystals can deliver. Table 2-1. Solubility data for itraconazole cocrystals.b Cocrystal [CCF] (mg/mL) [API] (M) [CC] (M) Itraconazole Succinic acid 76.9a ~(1/4)(210-4) ~2 10-4 Itraconazole L-malic acid 363.5a ~(1/20)(710-4) ~7 10-4 Itraconazole L-tartaric acid 1390a ~(1/20)(710-4) ~7 10-4 aValues are obtained from the Merck Index. bPhase transformation during the dissolution experiment is not reported. The dissolution profiles of fluoxetine HCl36 and three fluoxetine HCl cocrystals (benzoic acid, succinic acid and fumaric acid) were monitored in water at 20C for 120 minutes using ultraviolet spectroscopy (UV) at 227 nm. The samples were sieved to achieve a PSD of 53-150 m. About 800 mg of sample we re introduced to 30 mL of water at a stirring rate of 144 rpm. Any rema ining solids after the dissolution experiment were filtered, dried in vacuum and analy zed by powder x-ray diffraction (PXRD). The solubility of fluoxetine HCl was 11.4 mg/mL. Th e solubility of the benzoic acid cocrystal was lower at 5.6 mg/mL while the fumaric aci d cocrystal was higher at 14.8 mg/mL. The succinic acid cocrystal recorded a peak so lubility of 20.2 mg/mL after about 1 minute before decreasing to a concentration which ne ars that of fluoxetine HCl. The validity of these results is confirmed by PXRD analysis of the undissolved solids matching those of the starting phases for fluoxetine HCl, benz oic acid and fumaric acid cocrystals. PXRD also confirmed the identity of the undissolved solids after dissolution of the succinic acid cocrystal, to be fluoxetine HCl, highlight ing that the cocrys tal dissociated and

PAGE 38

19 recrystallized to fluoxetine HCl under the expe rimental conditions. This series of three cocrystals is unique in providing an example each of increasing (fumaric acid cocrystal) and decreasing (benzoic acid cocrystal) the aqueous solubility as well as being the first example (succinic acid cocrystal) of the “spring-and-parachute” effect. Table 2-2. Solubility data for fluoxetine hydrochloride cocrystals. Cocrystal [CCF] (mg/mL) [CC] (mg/mL) Conversion during experiment CC mp (C) Fluoxetine HCl Benzoic acid 3.4a 5.6 No 131.46 Fluoxetine HCl Succinic acid 76.9a 20.2 Yes 134.08 Fluoxetine HCl Fumaric acid 6.3a 14.8 No 161.06 aValues are obtained from the Merck Index. [API]=11.4 mg/mL. API mp=157.18C. A developmental cancer candidate compound50 (Pfizer 1), a potent and selective ErbB2 inhibitor, is a weak base with an aqueous solubility of 0.0008 mol/mL. 20 crystal forms of acid-base pairs were obtained after a salt screen. Three of these; a sesquisuccinate (neutral, 0.79 mol/mL), a dimalonate (mixed ionic and zwitterionic, 3.83 mol/mL) and a dimaleate (salt, 10.4 mol/mL), were selected for development. The location of the protons is probed by X-ray structural analysis and confirmed by cross polarization magic angle spinning (CPMAS) 15N nuclear magnetic resonance (NMR) spectroscopy. There was no di scussion on the experimental conditions involving solubility measurements. However, this is the first example of a cocrystal that exhibited an aqueous solubility approximating 3 orders of magnitude higher than that of the pure drug.

PAGE 39

20 Table 2-3. Solubility data for Pf izer 1 succinic acid cocrystal.b Cocrystal [CCF] (mg/mL) [CC] ( mol/mL) CC mp (C) Pfizer 1 Succinic acid 76.9a 0.79 143 aValue is obtained from the Merck Index. bPhase transformation during the dissolution experiment is not reported. [API]=0.0008 mol/mL. API mp=167C. Norfloxacin51 is a poorly water soluble fluor oquinolone antibacterial drug. An isonicotinamide cocrystal and three salts with succinic acid, malonic acid and maleic acid were prepared and structurally characteri zed. Solubility studies were performed by suspending about 21 mg of sample in 2.5 mL of water, maintained at 25C (1) in a laboratory oven for 72 hours at a stirring ra te of 300 rpm using absorbance measured at 276 nm. The succinic acid, malonic acid and maleic acid salts have an apparent solubility of 6.60, 3.90 and 9.80 mg/mL. The apparent so lubility of norfloxacin is 0.21 mg/mL while that of the isonicotinamide cocrystal is 0.59 mg/mL. This cocr ystal resulted in a modest 3-fold increase in solubility. Table 2-4. Solubility data for norfloxac in isonicotinamide cocrystal solvate.a Cocrystal [CCF] (mg/mL) [CC] (mg/mL) CC mp (C) Norfloxacin Isonicotinamide Chloroform solvate 192 0.59 180 aPhase transformation during the dissolu tion experiment is not reported. [API]=0.21 mg/mL. API mp=221C. AMG 51752, a transient receptor potential vanilloid 1 antagonist (TRPV1) developed for the treatment of chronic pa in, possesses an aqueous solubility of 5 g/mL as the free base. A total of 21 cocrystals were reported on two occasions. Solubility

PAGE 40

21 measurements were taken in fasted simulated in testinal fluid (FaSIF) at 25C within a 24 hour timeframe. There was no reported control of particle size. Of th ese cocrystals, there are 6 cocrystals with trans -cinnamic acid, 2,5-dihydroxybenz oic acid, 2-hydroxycaproic acid, benzoic acid, benzamide and cinnami de, that did not undergo any phase transformation during the duration of the experi ment. The solubilities of the cocrystals are 1, 2, 3, 21, 12.6 and 6.3 g/mL respectively53-54. The benzoic acid cocrystal reported a four-fold increase in solubility. Table 2-5. Solubility data for AMG517 cocrystals. Cocrystal [CCF] (mg/mL) [CC] ( g/mL) Conversion during experiment CC mp (C) AMG 517 trans -cinnamic acid 0.23a 1 No 204 AMG 517 2,5dihydroxybenzoic acid 22b 2 No 229 AMG 517 2-hydroxycaproic acid 32c 3 No 130 AMG 517 Benzoic acid 3.4d 21 No 146 AMG 517 Benzamide 13.5d 12.6 No 164 AMG517 Cinnamide 1.37 6.3 No 184 aLiterature value. bValue is obtained from the Handbook of Aqueous Solubility Data. cValue is calculated from Scifinder. dValues are obtained from the Merck Index. [API]=5 g/mL. API mp=230C. Ethenzamide, a nonsteroidal anti-inflamma tory drug (NSAID) for the treatment of moderate pain, forms polymorphic cocrystals55 with gentisic acid with three distinct crystalline modifications. Dissolution expe riments were carried out at 25C for ethenzamide and the three polymorphs. The eq uilibrium solubilities for ethenzamide and

PAGE 41

22 the cocrystal (Form 1) were estimated to be 0.0344 and 0.0365 mg/mL. There was no solubility enhancement in this example. Table 2-6. Solubility data for ethenzamide gentisic acid cocrystal.b Cocrystal [CCF] (mg/mL) [CC] (mg/mL) CC mp (C) Ethenzamide Gentisic acid Form I 22a 0.0365 100.65 aValue obtained from the Handbook of Aqueous Solubility Data. bPhase transformation during the dissoluti on experiment is not reported. [Ethenzamide]=0.0344 mg/mL. Ethenzamide mp=132C. Hexamethylenebisacetamide is used in the treatment of myelodysplastic syndrome and resistant acute myelogenous leukemia. A pyridine derivative56 (A) was prepared and forms five cocrystals with su ccinic acid, adipic acid, suberic acid, sebacic acid and dodecanedioic acid. Crystal structur e analysis reveals that the linear diacids serve as spacer molecules such that the increa se in length of the crys tallographic c axis in space group P is a reflection of the increasing nu mber of carbon atoms from succinic acid to dodecanedioic acid. As the melting point of the diacid decreases from succinic acid to dodecanedioic acid, the cocrystals follow the same trend. The aqueous solubilities of the cocrystals also decrease as the aqueous solubilities of the diacids decreases. However, this is not a common occurrence and has been attributed to the crystal packing of the cocrystals. Dissolution experiments were conducted with A and its five cocrystals by stirring excess cocrystal so lid phase in water (25C). The concentration of A was monitored by UV spectroscopy ( max = 236.4 nm).

PAGE 42

23 Table 2-7. Solubility data for cocrystals of hexamethylenebisacetamide derivative (A).a Cocrystal [CCF] (mg/mL) [CC] (mg/mL) CC mp (C) Asuccinic acid 76.9 167.18 186 Aadipic acid 14.4 117.3 165 Asuberic acid 1.6 92.042 158 Asebacic acid 1 12.109 148 Adodecanedioic acid 0.03 7.2459 146 aPhase transformation during the dissolution experiment is not reported. [A]=70.24 mg/mL. A mp=181C. Lamotrigine is marketed as Lamictal by GlaxoSmithKline for oral administration as a conventional or chewable tablet for the treatment of epilepsy and psychiatric disorders such as bipolar disord er. Ten crystal forms of lamotrigine were reported: lamotrigine methylpara ben cocrystal form I (1:1) ( 1 ), lamotrigine methylparaben cocrystal form II (1:1) ( 2 ), lamotrigine nicotinamide cocrystal (1:1) ( 3 ), lamotrigine nicotinamide co crystal monohydrate (1:1:1) ( 4 ), lamotrigine saccharin salt (1:1) ( 5 ), lamotrigine adipate salt (2:1) ( 6 ), lamotrigine malate salt (2:1) ( 7 ), lamotrigine nicotinate dimethanol solvate (1:1:2) ( 8 ), lamotrigine dimethanol solvate (1:2) ( 9 ) and lamotrigine ethanol monohydrate (1:1:1) ( 10 ). Dissolution experiments were performed on 2 3 4 and 5 in deionized water (25C) and 0.1 M HCl (37C). The samples were sieved to achieve a PSD between 53 m and 75 m. 100 mg of samples were introduced into 100 mL of water while 500 mg of sa mples were introduced into 50 mL of 0.1 M HCl. The slurries were stirred at a rate about 200-300 rpm. Samp les were filtered with a 0.45 m syringe filter and the concentration was determined using UV spectroscopy. The maximum concentration of 2, 3, 4, 5 and pure lamotrigine were 0.21, 0.30, 0.23, 0.45 and 0.28 mg/mL respectively. PXRD analysis of the remaining solid after dissolution

PAGE 43

24 confirmed 2 and 5 were the same solid phase while 3 and 4 converted to a hydrated form of lamotrigine57. Table 2-8. Solubility data for lamotrigine cocrystals. Cocrystal [CCF] (mg/mL) [CC] (mg/mL) Conversion during experiment CC mp (C) Lamotrigine Methylparaben Form II 2.5a 0.21b No 162.40 Lamotrigine Nicotinamide 1000a 0.30b Yes 167.23 Lamotrigine Nicotinamide Monohydrate 1000a 0.23b Yes 174.81 aValues are obtained from the Merck Index. Maximum [API] observed=0.28 mg/mL. API mp=221C. The solubility of cytosine cocrystals58 with oxalic acid, malonic acid and succinic acid were measured by gravimetric met hod at 15, 25, 40 and 60 C. The solubility of cytosine and its oxalic acid, malonic acid and su ccinic acid cocrystals were about 6, 4, 20 and 7 mg/mL respectively. Table 2-9. Solubility data for cytosine cocrystals.b Cocrystal [CCF] (mg/mL) [CC] (mg/mL) CC mp (C) Cytosine Oxalic acid 98.1a ~4 270.2 Cytosine Malonic acid 1538a ~20 219.2 Cytosine Succinic acid 76.9a ~7 249.5 aValues are obtained from the Merck Index. [API]=7.69 mg/mL.a bPhase transformation during the dissolutio n experiment is not reported. API mp=320C(dec.).

PAGE 44

25 Pterostilbene is a dimethyl ated analogue of resveratrol59 that forms cocrystals with caffeine and carbamazepine. Three cocrystals60 were obtained: caffeine pterostilbene cocrystal (Form I and Form II) and carbamazepine pterostilbene cocrystal. Dissolution experiments were performed for caffeine pterostilbene cocrystal (Form I) and the carbamazepine pterostilbene cocrystal by slurrying in water for 72 hours. Samples were filtered with a 0.2 m nylon filter and analyzed via a 96 well quartz plate using UV spectroscopy at 275 nm for carbamazepine. Table 2-10. Solubility data for pterostilbene cocrystals. Cocrystal [CCF] (mg/mL) [CC] (mg/mL) Conversion during experiment CC mp (C) CBZ Pterostilbene 0.021 0.0085 No 134 Caffeine Pterostilbene Form I 0.021 0.56 Yes 114 [CBZ]=0.056 mg/mL. CBZ mp=190C.

PAGE 45

26 In-house studies Table 2-11. Molecular structures of phenolic acids used in this study and their three letter code. OH O OH SAL O OH HO HO PCA OH O HO HO HO GAL OH O H3CO HO H3CO SYR OH O OH 1HY OH O HO COU OH O HO HO CFA OH O HO H3CO FER OH OH O O HO OH OH HO O CGA Tables 2-11 to 2-15 represent different categories of molecules based on different functional groups to facilitate co crystal screening and preparation via a supramolecular heterosynthon approach. In-house dissolution e xperiments were performed for cocrystals of epigallocatechin-gallate, protocatechuic ac id, coumaric acid, caffeic acid, gallic acid, ferulic acid, caffeine and ()baclofen. All samp les were prepared and sieved with ASTM standard sieve to achieve a PSD between 53 m and 75 m. The dissolution experiments were done in water at room temperature with the exception of ()bac lofen cocrystals at

PAGE 46

27 36.9C. Residual solids at the end of the di ssolution experiments were analyzed with powder X-ray diffraction to ensure the in tegrity of the cocrystal solid phase. Table 2-12. Polyphenols used in this study and their three letter code. O O HO OH OH OH OH O O H O H OH EGCG O OH HO OH OH OH O QUE O O HO HO O OH OH O ELA O O HO HO HO ETG Table 2-13. Pyridines, Purines and their three letter code. N NH2 O NAM N NH2 O INM N N H O NH2 INZ N N N N O O CAF N H N N N O O TPH N N HN N O O TBR

PAGE 47

28 Table 2-14. Zwitterions and their three letter code. +HN OO NAC +HN OO INA Cl OO NH3 + BAC Table 2-15. Molecules with carbonyl groups and their three letter code. NH O CAP H2N NH2 O URE NH HN O O GAH HN N H NH O O O CYA Many of these cocrystals routinely contain two components that can be considered as APIs in their own right based on their pharmacological properties. Thus the description of API and cocrystal former beco mes interchangeable. For the correlation of the properties of the cocrystals, CCF1 will be used to denote the API while CCF2 will be denoted as the cocrystal former. Table 2-16. Aqueous solubility and me lting point for ta rget molecules. CCF1 [CCF1] (mg/mL) CCF1 mp (C) Epigallocatechin gallate (EGCG) 23.6 140 Protocatechuic acid (PCA) 10.67 200 Coumaric acid (COU) 0.789 210 Caffeic acid (CFA) 0.485 223 Gallic acid (GAL) 11.49 258 Ferulic acid (FER) 0.6 168 Caffeine (CAF) 21.7 238 ()-baclofen (BAC) 5 206

PAGE 48

29 Table 2-17. Solubility data for in-house cocrystals. Cocrystal Refcode (CCF1CCF2) [CCF2] (mg/mL) [CC] (mg/mL) CC mp (C) EGCGNAM 1000 2.9 90 EGCGINM 192 1.32 155 EGCGINA 5.2 1.23 190 PCACAP 4560 17.21 131 PCAINM 192 7.74 195 PCATPH 8.3 1.02 254 PCANAM 1000 2.94 201 COUNAM 1000 1.078 159 COUINM (I) 192 1.15 172 COUINM (II) 192 1.12 165.5 COUCAF 22 0.314 184 COUTHP 8 1.157 223 COUTHB.H2O 0.5 0.294 228 COUURE 1200 1.52 123.5 COUINZ 12.5 1.096 178 COUTHB 0.5 0.2812 228 CFAINM 192 1.9415 151 CFAGAH 142 1.2613 221 CFAINZ 12.5 1.0616 185 GALINM 192 2.4 202 GALTBR 0.5 0.5 271 GALNAC 17 8 211 GALGAH 142 4.7 260 FERINM 192 1.8 152 FERNAM 1000 1.4 126 FERTBR 0.5 0.6 192 CAFCYA 2.7 9.69 228 CAFELA 0.0038 0.07 300 CAFCOU 0.789 1.1 178 CAFFER 0.6 2.85 146 CAFSAL 2 3.5 146 CAF1HY 0.5 0.19 190 CAFETG 4 0.54 144 CAFCGA 25 11.9 131 CAFGAL 11 5.7 244 CAFCFA 0.485 0.671 199 CAFSYR 5.7 1.17 180 CAFQUR 0.0026 8.55 244 QURCAF 22 13.3 244 BACGAL(36.9C) 26.4 9.37 178 BACFER(36.9C) 1.76 1.943 154 CCF1 denotes API. CCF2 denotes cocrystal former.

PAGE 49

30 Figure 2-1. Significance of quadrants Q1, Q2 Q3 and Q4 in log plot of solubility ratios. CCF1 denotes API. CCF2 denotes cocrystal former. Plotting the solubility ratio of cocrystal to API against the solubility ratio of cocrystal former to API on a logarithmic scal e, any negative value implies a solubility ratio of less than 1. When a more soluble co crystal former (compared to API) leads to a more soluble cocrystal, these data points are captured in quadrant Q1 When the solubility of the cocrystal decreases with a less soluble cocrystal former (compared to API), they are represented in quadrant Q3 Of particular concern are the data points in quadrants Q2 and Q4 Quadrant Q2 depicts cocrystals that re veal a decrease in aqueous solubility of the cocrystal despite a cocrysta l former with an aqueous solubility higher

PAGE 50

31 than that of the API. Albeit with a less soluble cocrystal former, if the solubility of the cocrystal records an increase, they are found in quadrant Q4 Figure 2-2. Solubility plot of in-house cocrystals. First and foremost, without a logarithmic scale, most data points are within an order of magnitude. A logarithmic plot enables a spread of the indivi dual data points. For the quercetin caffeine cocrystal (QUECAF), the dissolution experiment was performed with caffeine as the analyte of interest. Since the cocrystal is thermodynamically stable in water, therefore the molar concentration of both components (caffeine and quercetin) has to be the same to arrive at 13.3 mg/mL. Datapoints in Q1 and Q3 suggest that solubility

PAGE 51

32 is a transferable property from the cocrysta l former to the cocrystal. However, the presence of several data points in Q2 suggests this is not the case. Figure 2-3. Solubility plot of literature cocrystals. The solubility plot of cocrystals from the scientific literature tells a similar story. While the examples of the Pfizer cance r drug succinic acid cocrystal and the cis-itraconazole cocrystals are encouragi ng, the AMG517 cocrystals undoubtedly refutes the notion that a more soluble cocrystal form er (compared to API) translates to a more soluble cocrystal. Figure 2-3 reveals data points in every quadrant. Clearly, the solubility of the cocrystal is not predictable based on the solubility of the cocrystal former.

PAGE 52

33 Figure 2-4. Relationship between cocrystal solubility and cocrystal former solubility with respect to the API. The resultant plot in Figure 2-4 features a cluster of datapoi nts near the origin with a few notable exceptions such as QUECAF, the Pfizer cancer drug succinic acid cocrystal and the cis-itraconazole cocrys tals. Solubility is not a transferable physicochemical property. No longer can one be contented with a restricted cocrystal former library that favours high aqueous so lubility of cocrystal former to boost the aqueous solubility of cocrystals. Furthe rmore, the opportunity cost of missing the cocrystal with desired aqueous solubility after substantial rese arch and development

PAGE 53

34 efforts, is too high. That aque ous solubility is no t predictable means that every cocrystal potentially features utility in advantageous solubility and expands the searchable space61 of APIs but the solubility data has to be av ailable. While a solubility-guided approach to cocrystal screening is not advisable, the form ation of a cocrystal, using a supramolecular synthon approach, is more reliable by employing persistent supramolecular heterosynthons62-64. Although the desired aqueous solubil ities of the cocr ystals are not predictable, they are by all mean s achievable after due diligence. As mentioned earlier in this chapter, a crystal form with high lattice energy is unlikely to disintegrate and dissolve in solution readily, manifested by a high melting point which is an easily obtai nable, experimentally verifi able physical property of a crystalline solid. Hence solubility enha ncement upon cocrystal formation should be accompanied by melting point depression ( mp = CC mp – API mp) reflected in the melting point of the cocrystal relative to the API, i.e. mp < 0 for solubility enhancement and vice versa

PAGE 54

35 Figure 2-5. The relationship between chan ge in melting point (CC mp CCF1 mp) and the log of solubility ratio of cocrystal and API. CCF1 denote API. Indeed, if the inverse relationship between lattice energy and thermodynamic solubility upon cocrystal formation is a simp le one, all data points in figure 2-5 should only occupy the top left and bottom right qua drants. That most of the data points in figure 2-5 fall in the lower half suggests th at melting point depression may be a common phenomenon among the cocrystals discussed here in. There may be a hi gher probability of obtaining a cocrystal with a lower melting poi nt than a higher melting point. Hence if a crystal form of a solid with a lower me lting point is necessary without covalent modifications to the API, cocrystals represent viable options.

PAGE 55

36 2.7 Conclusion The “spring-and-parachute” effect has been a unique feature of pharmaceutical cocrystals that are not stable in an aqueous environment. As the dissolution profile of more cocrystal systems become publicly avai lable, we can expect to witness more examples of such systems. That the “springand-parachute” effect features a precipitous plunge to a concentration level similar to that of the pure AP I, is not nece ssarily a bad one. Similar dissolution profiles are presented by amorphous solids sin ce the initial boost in solubility (due to the metastable amor phous form) will decline once the conversion of the amorphous solid (into a crystalline form) commences. However, that the pharmaceutical cocrystal is a crystalline solid implies it possesses all the advantages that a crystal form has to offer. Compared to am orphous solids, crystallin e solids are preferred from a manufacturing and regulatory perspect ive due to a number of factors such as purity, stability and reproducibility. Furthermor e, formulation strategies such as the judicious choice of excipients can control the kinetics of th e conversion process, in the same way that amorphous solids are stabilized by excipients. Is solubility a transferable physicoche mical property of a cr ystalline solid? The results presented herein suggest that this is not the case. Several cocrystals are observed to be less water soluble upon cocrystal formation with co crystal formers possessing a favourable solubility ratio (>1) with respect to the API. 20.3% (13/64) of the data points in this study fall into this category. Given the lack of publicly available dissolution and solubility data of cocrystals calculation of this percenta ge may not be statistically

PAGE 56

37 representative. The lack of predictability of the desired physicochemical properties points to a need for a cocrystal former library to be exhaustive and incl ude cocrystal formers with a wide range of aqueous solubilities. In the same mann er that the cocrystals were overlooked after initial salt scr eens, due diligence should be ta ken not to miss the targeted cocrystals.

PAGE 57

38 Chapter 3. Molecular Cocrystals: Crystal Forms of Ellagic Acid 3.1 Preamble Crystal form (cocrystals, polymorphs, salt s, hydrates and solvates) diversity is and will continue to be a contentious issue for the pharmaceutical industry.65-66 That the crystal form of an API dramatically im pacts its aqueous solubility (a fixed thermodynamic property) is illustrated by the histamine H2-receptor antagonist ranitidine hydrochloride and HIV protease inhibitor r itonavir. For more than a century, the dissolution rate of a solid has been shown to be directly dependent on its solubility,42-43 c ter s paribus A century later, it remains impossible to predict the properties of a solid, given its molecular structure.67 If delivery and absorption of an API are limited by its aqueous solubility, aqueous solubility then becomes a critical parameter linking bioavailability and pharmacokinetics of an AP I. Since the majority of APIs are BCS Class II (low solubility and high permeability) compounds, crystal form screening, optimization and selection have thus receive d more efforts, attention and investment. Given that the dissolution rate, aqueous solu bility and crystal form of an API are intricately linked, it remains a scientific cha llenge to understand the nature of crystal packing forces and their impact upon physicoc hemical properties of different crystal forms. Indeed, the selection68 of an optimal crystal form of an API is an indispensable part of the drug development program. Developing a crystalline form of an API has other advantages. Crystallization has been employed routinely in the pharmaceutical industry to separate, isolate and purify

PAGE 58

39 drug molecules to yield a high purity product with reliable reproducibility and is a scalable process for manufacturing purposes. Compared to amorphous solids, a stable crystal form is highly desirable from manufacturing, handling and regulatory perspectives. In addition, the recent introduction69 of pharmaceutical cocrystals has expanded the patentable space of an API.61 Their novelty and utility is illustrated by the fact that they profoundly modify the physic ochemical properties of the parent API. Pharmaceutical cocrystals can be defined as multiple component crystals in which at least one component is molecular and a solid at room temperature (the pharmaceutically acceptable cocrystal former) and forms a s upramolecular synthon with a molecular or ionic API. They have added another dimensi on to the diversity of possible crystal forms (traditionally restricted to polymorphs, salts, solvates a nd hydrates) and augmented the drug discovery funnel. With a myriad of possi ble crystal forms waiting to be discovered, it is not surprising that pharmaceutical scie ntists have resorted to high throughput screening (HTS) techniques to as sist them in their endeavor. Screening techniques such as solv ent drop grinding, SonicSlurry™, HTS evaporation experiments, and reaction crystall ization are some of the routine screening tools. Using carbamazepine, these four screen ing techniques identified a total of 27 solid phases with 18 carboxylic acids as cocrystal formers.70 50 piroxicam cocrystals71 were identified in screening experiments using 23 carboxylic acids. 18 carboxylic acids were combined with meloxicam to produce 19 cocrystals.72 Other examples of API carboxylic acid cocrystals include itraconazole,49 fluoxetine hydrochloride,36 developmental sodium channel blocker,73 fluconazole,74 caffeine,75 theophylline,75 gabapentin,76 AMG517,52 minoxidil77 and ethenzamide.55, 78-79 Many carboxylic acids are included in a Generally

PAGE 59

40 Recognised As Safe80 (GRAS) list and are thus routin ely engaged as pharmaceutically acceptable cocrystal formers. However, pha rmaceutically acceptable cocrystal formers must go beyond carboxylic acids to truly dive rsify the crystal forms available to APIs. For instance, polyphenols possess potent hydrogen bonding functional groups capable of molecular recognition. Furthermore, they are present in our diet81 in the fruits, tea82 and wine83 that are commonly consumed and often exhibit pharmacological properties as well. Obviously, they have to be pha rmaceutically acceptable even though the toxicology84 profile may not be thoroughly inve stigated. Indeed, this has been demonstrated with pterostilbene as a cocrystal former to prepare cocrystals with caffeine and carbamazepine.60 Figure 3-1. The phenol-carbonyl supramolec ular heterosynthon is featured in the cocrystals of (caffeine)•(pterostilbene) (l eft) and (carbamazepine)•(pterostilbene) (right). Ellagic acid85 is a naturally occurring polyphenolic nutraceutical86-87 derived from the hydrolysis of ellagitannins88 and it is abunda nt in pomegranates and raspberries. According to DeFelice, a nutraceutical is defi ned as “a food (or part of a food) that

PAGE 60

41 provides medical or health be nefits, including the prevention and/or treatment of a disease. It has been widely studied for its pharmacological properties.89-90 However, it is reported to possess an aque ous solubility of only 3.8 g/ml91 at 25C. This has restricted recent nutritional research to focus on the pha rmacokinetics of pomegranate juice/extract rather than pure ellagic acid in healthy subjects,92-96 since its poor aq ueous solubility limits bioavailability. In two separate pharmacokinetic studies administrating pomegranate extract and pomegranate juice in healthy human subjects, ellagic acid was detected in human plasma with a maximum concentration (Cmax) of about 33 ng/mL97-98 at a time of maximum concentration (Tmax) of 1 hour after ingestion. The poor aqueous solubility of ellagic acid has impeded its pot ential as a therapeutic molecule. There is a need to improve the aqueous solubility of ellagic acid from a clinical perspective. Although a number of pharmacokinetic studies have been focused on ellagic acid, little information on its solid state chemistry has been reported in the open literature. That ellagic acid possesses extremely low aqueous solubility makes it a worthy candidate for cocrystal formation since aqueous solubility is a critical physicoc hemical property that can decide the fate of an API.

PAGE 61

42 Table 3-1. Molecular Structures of Ell agic Acid and Group 1 Basic Cocrystal Formers with their three letter code. O O HO HO O OH OH O ELA N NH2 O NAM N NH2 O INM N N N N O O CAF N H N N N O O TPH H NN DMP Table 3-2. Group 2 Zwitterionic Cocrystal Formers with their three letter code. H2 +N OO SAR NH+ OO DMG Table 3-3. Group 3 Neutral Cocrystal Formers with their three letter code. NH O CAP H2N NH2 O URE

PAGE 62

43 Table 3-4. Solvates of Ellagic Acid. N O NMP N O DMA S O DMSO HO OH ()PG With this in mind, we explore the cr ystal form landscape of ellagic acid. Amenable to crystal engineering design prin ciples, ellagic acid has an imbalance of hydrogen bond donors over acceptors, which s uggests that perhaps it may be prone towards cocrystal formation and/or solvat ion by molecules with superior hydrogen bond acceptor capabilities. Its molecular planarity and rigidity also ma kes it an attractive crystal engineering target. To date, three crystal structur es of ellagic acid (anhydrate, dihydrate and tetrakis(pyridine) solvate) were reported in the Cambridge Structural Database99 (CSD). Cocrystals of ellagic acid had addressed supramolecular heterosynthons64 (between phenols and carboxylat es) and structure-stability relationships100 (thermal stability of cocrystal hydr ates). A grouped library of successful cocrystal formers include ( 1 ) nicotinamide, isonicotinamide, caffeine, theophylline, 3,5dimethylpyrazole, ( 2 ) sarcosine, N,N -dimethylglycine, ( 3 ) -caprolactam and urea. Group 1 involves cocrystal formers with basic pyri dine, imidazole and pyrazole moieties. Group 2 contains zwitterionic cocrystal form ers having negatively-charged carboxylate moieties. Group 3 features the neutral carbonyl group. Ot her crystal forms of ellagic acid include solvates from N,N -dimethylacetamide (DMA), N -Methyl-2-pyrrolidone (NMP), dimethylsulfoxide (DMSO) and ()propylen e glycol (PG). Although the discovery of

PAGE 63

44 solvates was serendipitous, the supramolecu lar synthons observed in these structures nevertheless render themselves usef ul in the design of cocrystals. 3.1.1 CSD Survey of Ellagic Acid Figure 3-2. Crystal forms of ellagic acid: anhydrate (top left), dihydrate (top right) and tetrakis(pyridine) solvate (bottom). Analysis of existing crystal structures represents the first step in a crystal engineering experiment.101 A survey of the CSD for crystal forms of ellagic acid returned three hits, an anhydrate101 (Refcode: ELLAGC), a dihydrate102 (KUVBEI) and tetrakis(pyridine) solvate103 (VASYEU). In the crystal st ructure of anhydrous ellagic acid, ellagic acid molecules are arranged in to a chain by phenol-phenol supramolecular homosynthon. Adjacent chains are cro ss-linked by phenol-carbonyl supramolecular ELLAGC KUVBEI VASYEU

PAGE 64

45 heterosynthon to generate a fl at sheet. Adjacent sheets are -stacked to complete the crystal packing. In the dihydrat e, water molecules insert themselves between ellagic acid molecules so that the chains are sustai ned by OH•••O(water) interactions. Adjacent chains are linked by OH•••O( water) and OH(water)•••O=C interactions. Similarly, adjacent sheets are reinforced by -stacking104 interactions. The crystal structure of the tetrakis(pyridine) solvate of ellagic acid reveals a discrete supermolecule105 assembled from one ellagic acid molecule and four pyridine solvent molecules linked via phenolpyridine supramolecular hetero synthon. The presence of the te trakis(pyridine) solvate suggests the phenol-pyridine supramolecular heterosynthon is more robust than the phenol-phenol supramolecular homosynthon a nd should be exploited for cocrystal formation. 3.1.2 Cocrystals of Polyphenols with Group 1 Basic Cocrystal Formers (Pyridine, Imidazole, Pyrazole) The supramolecular heterosynthons respons ible for the formation of cocrystals between phenols and pyridine/imidazole/pyr azole has been documented. The dominant phenol-pyridine supramolecular heterosyntho n has been employed by the groups of MacGillivray, Nangia and Zaworotko in c onstructing cocrystals of polyphenols. MacGillivray106 et al. used resorcinol as a template to align trans -1,2-bis(4pyridyl)ethylene into a photor eactive supermolecule for the [2+2] photodimerization of trans -1,2-bis(4-pyridyl)ethylene (which is otherwise photostable) to yield rctt -tetrakis(4pyridyl)cyclobutane.

PAGE 65

46 Figure 3-3. The phenol-pyridine supramolec ular heterosynthon aligns two olefins into a discrete photoreactive supermolecule105 (top). A projection of the discrete photoproduct on the ac plane (bottom). Peddy et al. prepared a series of cocrysta ls of phenols with isonicotinamide107 that exhibit the phenol-pyridine supramolecular hete rosynthon. To demonstrate the preference of phenols for pyridines over cyano groups in a competitive hydrogen bonding environment, Bis et al. prepared and characterized a series of cocrystals with persistent phenol-pyridine supramolecular heterosynthon63 in the presence of cyano groups. That the phenol-pyridine supramolecular hetero synthon is superior to the phenol-cyano supramolecular heterosynthon is confirmed wh en mixtures of starting materials were obtained from experiments to obtain cocrystals of 3-hydroxypyridine or 5hydroxyisoquinoline with cyano derivatives.

PAGE 66

47 Figure 3-4. A chain of alternating caffeine molecules and methyl gallate molecules in (caffeine)(methyl gallate) cocrystal (top), alternating pairs of methyl gallate and theophylline dimer forms a flat sheet in th e (theophylline)(methyl gallate) cocrystal (bottom). Substituted xanthines (such as caffeine and theophylline) are biologically active molecules that feature basic imidazoles, and are potent corystal formers for acidic molecules. Trask et al. reported a series of cocrystals with caffeine108 and theophylline109 using a homologous series of dicarboxylic acids (oxalic acid, maloni c acid, maleic acid, succinic acid110 and glutaric acid). Coincidently, both caffeine/theophylline oxalic acid cocrystals were stable to hydration at 98% RH for a period of seven weeks, confirmed by powder X-ray diffraction. Weakly acidic phenol s are equally adept in interacting with substituted xanthines111, demonstrated by their methyl gallate cocrystals.40, 112 The crystal

PAGE 67

48 packing of the cocrystal helps to improve the mechanical properties compared to caffeine due to the introduction of a slip system. The ab ility of pyrazoles to form cocrystals with polyphenols was illustrated through 3,5dimethypyrazole and phloroglucinol.113 3.1.3 Cocrystals of Polyphenols with Gro up 2 Zwitterionic Cocrystal Formers Zwitterions such as the proteinogenic ami no acids are perhaps the best cocrystal formers in terms of their toxicology profile, cost and ready availability. In general, a zwitterion is a dipolar ion that carri es both a positive and negative charge114 and has a total net charge of zero.115 They had been used in the resolution of optical isomers involving cocrystals of ho mochiral mandelic acid. ( S )-alanine116 and ( R )-cysteine117 can be used to resolve racemic mandelic acid. ( S )-mandelic acid is able to resolve racemic 2aminobutyric acid and phenylalanine through fractional crystalli zation. Importantly, ( S )mandelic acid is used to resolve racemic pregabalin118 (active ingredient in LYRICA that is used to treat neuropathic pain) on an industrial scale. Examples of zwitterionic cocrystals exist in the CSD.119-127 The utility of carboxylates in cocrystal formation has expanded the scope of zwitterionic cocrystals beyond carboxylic acids to include molecules with weakly acidic hydroxyl moieties100, as exemplified by polyphenols and Lascorbic acid.

PAGE 68

49 Figure 3-5. A 1:1 cocrystal of L-ascorbic acid and sarcosine sustained by carboxylate-hydroxyl supramolecular heterosynthon. 3.1.4 Cocrystals of Polyphenols with Group 3 Neutral Cocrystal Formers (Carbonyl) Figure 3-6. -caprolactam molecules exist as dimers and bridge resorcinol molecules. The carbonyl group is a component of many organic functional groups such as the carboxylic acid, primary, seconda ry and tertiary amides, alde hyde, ketone, ester, lactone and lactam. While the supramolecular chemistry of carboxylic acids, primary and

PAGE 69

50 secondary amides have been explored, the sa me cannot be said of the other functional groups. Using -caprolactam and urea, we probe the propensity of the carbonyl group to engage in hydrogen bonding in teraction with polyphenols. Urea is a small organic molecule that is prone to forming urea inclusion128-129 compounds. Supramolecular heterocatemers130 have been constructed from pol yphenols (such as 4,4’-biphenol and resorcinol) and -caprolactam to highlight thei r role in cocrystal design. 3.2 Results and Discussion Cocrystals, cocrystal solvates/hydrates a nd solvates of ellagic acid were prepared and fully characterized. The synthesis of thes e cocrystals demonstrated the utility of polyphenols as cocrystal formers for a wide variety of functi onal groups, such as, pyridine, imidazole, pyrazole, zw itterions and carbonyl groups. 3.2.1 Crystal Forms of Ellagic Acid with Group 1 Basic Cocrystal Formers (Pyridine, Imidazole, Pyrazole) ELANAM: The crystal structure of ELANAM reveals it is a 1:2 cocrystal of ellagic acid and nicotinamide. ELANAM exhibits hydrogen bonding between the pyridyl group of nicotinamide and the phenolic moieties of ellagic acid [OH•••N: 2.630(5) ]. The neutral nature of nicotinamide is s upported by the C-N-C angle of 116.35 in the pyridyl group. Two nicotinamide molecules a dopt the amide dimer. The phenol-pyridine supramolecular heterosynthon63, 107, 131 and the amide dimer supramolecular

PAGE 70

51 homosynthon132-134 [NH•••O: 3.040(7) ] generate an undulating tape. Two ellagic acid molecules sandwich the amide dimer with the carbonyl and phenolic moieties to form a (10) trimer [OH•••O: 2.694(6) and NH•••O: 2.953(7) ]. The trimeric interaction (figure 2-6) links the undulat ing tapes together to yield the overall crystal packing. Figure 3-7. The undulating ELANAM tapes are crosslinked by sandwiching the amide dimer with a pair of ellagic acid molecules. ELAINM: Three crystal form s of ELAINM were obtained: a variable hydrate, a NMP solvate and DMA solvate. Crystal structure analysis of ELAINM•1.7H2O reveals one ellagic acid molecule, two isonicotinamide molecules and disordered water molecules in the unit cell. ELAINM•1.7H2O exhibits hydrogen bonding be tween the pyridyl group of isonicotinamide and the phenolic moieties of ellagic acid [OH•••N: 2.611(2) ]. The

PAGE 71

52 neutral nature of isonicotinamide is suppor ted by the C-N-C angle of 118.07 in the pyridyl moiety. Two isonicotinamide molecules exist as dimers through amide dimer supramolecular homosynthons [NH•••O: 2.988(3) ]. The phenol-pyridine supramolecular heterosynthon and the amide dimer supramolecular homosynthon generate a tape (figure 2-7). The amid e dimer also acts as both a hydrogen bond donor and acceptor towards the carbonyl [NH•••O: 2.930( 3) ] and phenolic mo ieties of ellagic acid [OH•••O: 2.749(2) ] to form a (10) trimer that links adjacent chains. Disordered water molecules occupy channels para llel to the crysta llographic a axis. Figure 3-8. Crystal packing of ELAINM1.7H2O reveals isonicotinamide dimers with lateral hydrogen bonding to phenolic and carbonyl moieties of ellagic acid. Water molecules (shown as large vdw sp heres) occupy channels parallel to the crystallographic a axis. The crystal structure of ELAINM•4NMP reveals it is a tetrakis(NMP) solvate of the 1:2 cocrystal of ellagic acid and isonico tinamide. Isonicotinamide molecules exist as

PAGE 72

53 dimers via the amide dimer supramolecular homosynthon [NH•••O: 2.923(4) ]. The pyridyl group interacts with the phenolic moiety [OH•••N distance of 2. 647(4) ] to form a linear chain. The crystal structure is su stained by the phenol-pyridine and amide-amide supramolecular synthons. Along the periphery of the chain, four NMP molecules are positioned to engage the two remaining hydrogen bond donors (from ellagic acid) via phenol-carbonyl supramolecular heterosynt hon [OH•••O: 2.615(4) ] and another two dangling N-H donors [from the amide dimer, NH•••O: 2.827(5) ]. Figure 3-9. The supramolecular tape in ELAINM has excess hydrogen bond donors from phenolic moieties of ellagic acid and amide dimer of isonicotinamide, attracting the NMP solvent molecules (top) and DMA solvent molecules (bottom).

PAGE 73

54 The crystal structure of ELAINM•4D MA reveals a 1:2: 4 ellagic acid, isonicotinamide and DMA cocrystal solvate in which two of the four DMA molecules are crystallographically disorder ed. ELAINM•4DMA is isostr uctural with ELAINM•4NMP. Isonicotinamide molecules are dimers via the amide dimer supramolecular homosynthon [NH•••O: 2.895(3) ]. The pyridyl group inter acts with the phenolic moiety [OH•••N: 2.647(3) ] to form a linear chain. Along the periphery of the chain, four DMA molecules are positioned to engage th e two remaining hydrogen bond donors (from ellagic acid) via phenol-carbonyl supramolecular he terosynthon [OH•••O: 2.608(3) ], and another two dangling N-H donors [from the amide dimer, NH•••O: 2.792(7) ].

PAGE 74

55 Figure 3-10. Crystal Packing in ELACAFH2O reveals hydrogen bonding between ELA and CAF molecules (top). Water molecules reinforce the hydrogen bonding interaction between adjacent sheets (bottom). ELACAF•H2O: Analysis of the crystal structure reveals that ELACAF•H2O contains two ellagic acid molecules, two caffeine molecules and two water molecules in the unit cell. The phenolic moieties of ella gic acid act as hydrogen bond donors to the basic imidazole [OH•••N: 2.702(4) ] a nd carbonyl moieties of caffeine molecules [OH•••O: 2.708(3) ] so as to form straight chains of alte rnating ellagic acid and caffeine molecules. Adjacent chains are linked via OH(ELA)•••O=C(CAF) hydrogen bonds [2.708(3) ] and water molecules [2.608(3) and 2.854(4) ] th at hydrogen bond to two

PAGE 75

56 ELA molecules, thereby forming a sheet [Figure 3-10(top)]. Adjacent sheets are crosslinked by hydrogen bonds between water molecules and carbonyl oxygen atoms of ellagic acid [2.920(4) ]. The water mo lecules are in isolated environments. ELATPH: Two crystal forms of ELATPH were obtained: a partial hydrate and a DMA solvate. Figure 3-11. A projection of ELATPH0.13H2O on the ac plane (top), a chain is constructed in ELATPH2DMA (bottom). The X-ray structure of ELATPH•0.13H2O reveals a 1:2 cocrys tal of ellagic acid and theophylline. Theophylline molecules exist as (10) dimers [N-H•••O: 2.714(2) ], allowing the remaining carbonyls from the dimer to engage the phenolic moieties of ellagic acid [OH•••O: 2.757(2) ] to form a straight chain. This straight chain is crosslinked at an angle with adjacent straight chains via phenol-imidazole supramolecular heterosynthon [N-H•••O: 2.711(2) ] to genera te stacks of crosses. Traces of water are present in the X-ray crystal structure repres ented by the isolated red spheres in figure

PAGE 76

57 3-11(top). The water molecules hydrogen bond with the carbonyls [OH•••O: 2.611 ] of ellagic acid and theophylline [OH•••O: 2.785]. In ELATPH•2DMA, theophylline mol ecules exist as dimers [N-H•••O: 2.804(3) ], allowing the imidazole nitrogen of theophylline to interact with the phenolic moieties [OH•••N: 2.849(3) ] and establish a linear chain. Similarly, the crystal structure is sustained by the phenol-imid azole supramolecular heterosynthon and the theophylline dimer. Two disordered DMA mol ecules are located al ong the periphery of the chain to satisfy the two remaini ng hydrogen bond donors (from ellagic acid) via phenol-carbonyl supramolecular hete rosynthon [OH•••O: 2.548(4) ]. Figure 3-12. A discrete assembly is established in ELADMP. ELADMP: X-ray structural analysis reve als a 1:2 cocrystal of ellagic acid and 3,5-dimethylpyrazole. Centrosymmetric 3,5dimethylpyrazole molecules sandwich an ellagic acid molecule via OH(phenol)•••N(pyrazole) interaction [OH•••N: 2.588(4) ]

PAGE 77

58 and NH(pyrazole) •••O=C(lactone) interact ion [NH•••O: 2.973(4) ] to form a supermolecule. The remaining phenolic moieties engage in T-shaped OH••• [OH••• : 3.263(5) ] interactions135 with adjacent 3,5-dimethylpyrazo le molecules to generate a herringbone arrangement. 3.2.2 Crystal Forms of Ellagic Acid with Gr oup 2 Zwitterionic Cocrystal Formers ELASAR: The crystal structure of ELASAR reveals that the 1:1 cocrystal of ellagic acid and sarcosine is sustained by charge-assisted O-H•••O supramolecular heterosynthon [2.594(2) and 2.630(2) ] to form (18) tetramers [figure 3-13(top)]. Each carboxylate moiety of sarcosine interact s with four ellagic acid molecules whereas each ammonium moiety forms a charge-assist ed N-H•••O supramolecular heterosynthon [2.905(3) and 3.025(3) ] with the phenolic moie ties of two ellagic acid molecules. The ellagic acid molecules are -stacked in a slipped orientation [figure 3-13 (bottom)].

PAGE 78

59 Figure 3-13. The phenol-carboxylate supramolecular heterosynthon (top), stacking interactions in the cry stal packing of ELASAR (bottom) ELADMG: The 1:1 cocrystal of ellagic acid and N,N-dimethylgycine also exhibits the (18) tetramer [OH•••O: 2.649(1) a nd 2.903(2) ] [figure 3-14(left)]. The ammonium moiety of N,N-dimethylglycine interacts with one phenolic moiety [NH•••O: 2.887(2) ] whereas th e ellagic acid molecules are -stacked in a staggered orientation [figure 3-14(right)].

PAGE 79

60 Figure 3-14. The phenol-carboxylate supramolecular heterosynthon (left), stacking interactions re inforce the crystal packing forces in ELADMG (right) 3.2.3 Crystal Forms of Ellagic Acid with Group 3 Neutral Cocrystal Formers (Carbonyl) ELACAP: The X-ray crystal structure reveal s it is a 1:2 cocrys tal of ellagic acid and -caprolactam. The phenolic moieties act as hydrogen bond donors towards the carbonyls of -caprolactam [OH•••O: 2.696(9) and 2.724(9) ]. The carbonyl groups of ellagic acid act as hydrogen bond acceptors towards the NH of -caprolactam [NH•••O: 2.903(10) ]. The two supramolecula r heterosynthons generate a 3-periodic network of supramolecular heterocatemers.

PAGE 80

61 Figure 3-15. A supramolecular hetero catemer is assembled in ELACAP. ELAURE•2NMP: Analysis of the crystal st ructure reveals a bis(NMP) solvate of a 1:2 cocrystal of ellagic acid and urea. The phenol-carbonyl supramolecular heterosynthon is prevalent with all phenolic moieties intera cting with the carbonyl of two crystallographically independent urea [OH •••O: 2.543(5) and 2.684(4) ] and NMP molecules [OH•••O: 2.581(4) and 2.657(5) ]. Centrosymm etric urea dimers [NH•••O: 2.899(6) ] bridge four ellagic acid mo lecules [NH•••O: 2.983(5) and 3.058(5) ] from different staggered layers. Channels ar e generated along the crystallographic a axis to accommodate chains of hydrogen bonded urea molecule s [NH•••O: 2.922(5) and

PAGE 81

62 2.959(6) ]. NMP molecules are present to interact with the remaining hydrogen bond donors from ellagic acid. Figure 3-16. A chain of hydrogen bonded urea molecules occupy channels down the crystallographic a axis in the crystal packing of ELAURE2NMP. 3.2.4 Solvates of Ellagic Acid Additional crystal forms of ellagic acid were obtained in the form of solvates using DMA, NMP, DMSO and ()-PG. The crystal packing of some of these solvates resembles that of the dihydrate and tetr akis(pyridine) solvate of ellagic acid.

PAGE 82

63 Figure 3-17. The similarity in crystal packing between the ELADMA4 solvate and the tetrakis(pyridine) solvate (inset). ELADMA4: The tetrakis(DMA) solv ate is sustained by phenol-carbonyl supramolecular heterosynthon between the pheno lic moieties of one ellagic acid molecule with four DMA molecules [OH•••O: 2.6148( 16) and 2.5770(16) ]. This crystal structure resembles that of the tetrakis(pyridine) solvate (Refcode: VASYEU) with slightly further hydrogen bonding distances [OH•••N: 2. 622(3) and 2.795(4) ] compared to the tetrakis(DMA) solvate. A di screte assembly is formed and the overall crystal packing is a herringbone arrangement.

PAGE 83

64 Figure 3-18. The similarity in crystal packing between the ELADMA2 solvate and the dihydrate (inset). ELADMA2: The bis(DMA) solvate of ellagic acid is su stained by phenolcarbonyl supramolecular heterosynthon, with the phenolic moieties of two adjacent ellagic acid molecules bridged by two DMA molecules to form a (14) tetramer [OH•••O: 2.6696(14) and 2.8393(14) ]. This unit propagates into a linear chain and resembles the crystal packing of ellagi c acid dihydrate (Refcode: KUVBEI). In the crystal structure of ellagic ac id dihydrate, the crystal packing forces were maximized with water molecules bridging adjacent chains via OH(water)•••O=C(lact one) interaction, reinforced by -stacking interaction. The same is not possible for this DMA solvate since DMA does not possess labile protons av ailable for hydrogen bonding and the comparatively bigger DMA molecules (relative to H2O molecules) caused adjacent chains to be displaced such that -stacking interactions canno t be made efficiently.

PAGE 84

65 Figure 3-19. DMSO molecules tether ELA molecules into a straight chain. ELADMSO2: The bis(DMSO) solvate is sustained by phenol-sulfoxide supramolecular heterosynthon [OH•••O=S: 2. 674(3) and 2.695(3) ]. Two adjacent ellagic acid molecules are tethered by tw o bridging DMSO molecules to form a (14) tetramer. This unit propagates linearly into a chain and the crystal packing resembles that of both the bis(DMA) solvate a nd ellagic acid dihydrate. Here, -stacking interactions reinforced the crystal packing forces.

PAGE 85

66 Figure 3-20. A projection of ELANMP2 on the ac plane. ELANMP2: The bis(NMP) solvate is sustained by phenol-carbonyl supramolecular heterosynthon where the phenolic moieties of two ellagic acid molecules are bridged by one NMP mol ecule [OH•••O: 2.697(4) and 2.745(4) ]. Unlike the bis(DMA) solvate and bis(DMSO) solvate, the phenolic moieties do not hydrogen bond in a concerted manner to the same two NMP mo lecules to avoid the formation of another chain structure. Instead, the NMP molecules are bridging ellagic acid molecules to generate a supramolecular heterocatemer packed in a herringbone arrangement.

PAGE 86

67 Figure 3-21. ()PG mol ecules align the ellagic acid molecules. ELAPG2: The bis()PG solvate is sustained by OH(phenol)•••OH(aliphatic alcohol) and OH(aliphatic alcohol)•••O=C(l actone) interactions at 2.660(3) and 2.734(3) respectively. PG molecules of opposite chirality bridge ad jacent ellagic acid molecules into -stacks, with the shortest distance of two ellagic acid planes at 3.280 . 3.3 Conclusion 17 crystal forms (5 cocrystals, 7 cocrys tal solvates/hydrates and 5 solvates) of ellagic acid have been presented herein. It seems ellagic acid is particularly prone to solvation. Although they are usually discovered serendipitously, the supramolecular synthons observed in these structures can sti ll render themselves usef ul in the design of cocrystals. Unlike calixarenes136 and cyclodextrins137 that act as hosts to accommodate

PAGE 87

68 guest molecules in their cavities, ellagic ac id does not possess any molecular feature to suggest a tendency towards small molecule uptake such as solvent molecules. The solvated/hydrated crystal struct ures reveal they are neithe r inclusion compounds, i.e. the crystal packing does not generate pockets of empty spaces to trap molecules of appropriate volume. As mentione d at the start of this chapter, the imbalance of hydrogen bond donors over acceptors implies that given the opportunity, ellagic acid will prefer to accommodate a hydrogen bond acceptor in its lo cal environment to satisfy its hydrogen bonding requirements. Furthermore, the lact one functional group is not known to be a target in crystal engineering studies, i.e ., a stronger hydrogen bond acceptor will engage in hydrogen bonding interaction with the phenolic moieties. Hence, it is not surprising that the phenol-carbonyl(lactone) interaction is only observed in the crystal structure of anhydrous ellagic acid. However, in the pres ence of molecules w ith superior hydrogen bonding acceptor capabilities, th e phenolic moieties participate in hydrogen bonding readily, disrupting the crystal p acking of anhydrous ellagic aci d. The crystal structures of the five cocrystal solvates/hydrates are a furt her testament of this imbalance of hydrogen bond donors over acceptors. Referring to the isonicotinamide and theophylline cocrystal solvates, there remains an excess of hydr ogen bond donors after the formation of the phenol-pyridine supramolecular heterosynthon and the phenol-imidazole supramolecular heterosynthon. Hence the solvent molecules ar e included in the crys tal lattice around the periphery of the established unit. Examination of the crystal form landscape of ellagic acid has demonstrated that it has a preference for molecules capable of hydrogen bond acceptor properties in its local environment. Pyridines, imidazoles, pyrazole zwitterions and carbonyl groups have been

PAGE 88

69 manipulated to yield the cocrystals. Th e results presented herein further our understanding of the supramolecu lar chemistry of polyphenols, confirm the feasibility of polyphenols to serve as cocrystal formers for a diverse variety of functional groups and demonstrate the phenol-carbonyl supramolecular heterosynthon to be reliable and useful in a predictable manner. The crystal structur e of anhydrous ellagic acid reveals linear chains of ellagic acid molecules (establishe d by forming a catechol dimer), crosslinked via OHO=C(lactone) interaction and reinforced by -stacking interactions. The DMA and NMP solvates of ellagic acid suggest that tertiary amides are perhaps better hydrogen bond acceptors than lactones. The crystal structur es of the cocrystal solvates highlight the synergistic cooperation exhibited among the phenol-pyridine supramolecular heterosynthon, phenol-imidazole supramolecu lar heterosynthon and the phenol-carbonyl supramolecular heterosynthon. That the solvents are included in the crystal lattice, despite the phenol-pyridine supramolecular heterosynthon and phenol-imidazole supramolecular heterosynthon playin g a critical role in the se lf-assembly of the cocrystal solvates, implies that the phenol-carbonyl supramolecula r heterosynthon is robust, especially when there is an excess of hydrogen bond donors such as in the pr esence of a polyphenol. The synergism demonstrated be tween the phenol-pyridine supramolecular heterosynthon and phenol-imidazole supram olecular heterosynthon with the phenolcarbonyl supramolecular heterosynthon (in the cocr ystal solvates) sugges ts the possibility of a ternary cocrystal and adds another level of complexity to the possible structures that can be constructed. In conclusion, the crystal forms (cocry stals, cocrystal solvates/hydrates and solvates) of ellagic acid are achieved by a variety of supramolecular heterosynthons

PAGE 89

70 formed between polyphenols with pyridines imidazoles, pyrazole, carboxylates and carbonyl moieties. The propensity of polypheno lic molecules to engage in molecular recognition events, leading to the formation of supramolecular heterosynthons, has been the subject of intense scrutiny. The delin eation of supramolecular synthons into a hierarchy has undoubtedly unleashed the power of the supramolecular heterosynthon approach to the crystal e ngineering of cocrystals. Th e synergism exhibited among supramolecular heterosynthons is a coveted attribute in a hydroge n bonding cooperative environment. That APIs routinely contai n multiple functional groups enables polyphenols to play a role as cocrystal formers in the crystal form screening of APIs. This information is relevant to the pharmaceutical, food and nutrition industry in their crystal form screening since polyphenols are f ound in a variety of food sources (tea, wine, fruits) with known toxicology profile. Indeed, the range of pharmaceutically ac ceptable cocrystal formers available for crystal form screen ing would be considerably broadened if polyphenols were to be included in cocrystal former libraries. 3.4 Materials and Methods Ellagic acid dihydrate was purchased fr om TCI America (Lot. 6LHXG) and used as received. Ellagic acid di hydrate from other suppliers may not reproduce these results if they are not easily soluble in pyridine, DMA and NMP.

PAGE 90

71 3.4.1 Synthesis of Crystal Forms of Ellagic Acid ELANAM: Ellagic acid di hydrate (20.0 mg, 0.06 mmol) was dissolved in 0.5 mL of NMP with heat. Nicotinamide (144.0 mg, 11.8 mmol) was dissolved in 0.5 mL of deionised water with heat. The contents we re mixed and left for slow evaporation. Yellow plate-like crystals (15.6 mg, 288C) were harvested after a week. ELAINM•1.7H2O: Ellagic acid dihydrate (20.0 mg, 0.06 mmol) was dissolved in 2 mL of DMA. Isonicotinamide (290 mg, 2.37 mmo l) was dissolved in 1 mL of deionised water. Both contents were mixed to give a light brown solution and left for slow evaporation. Yellow plate-like crystals ( 18.1 mg, mp 300C) were harvested after one month. ELAINM•4NMP: Ellagic acid dihydrate (20.0 mg, 0.06 mmol) and isonicotinamide (80.0 mg, 0.66 mmol) were disso lved in 1 mL of NMP with heat. The yellow solution was left for slow evaporat ion. Yellow crystals (16.6 mg, 290C) were harvested overnight. ELACAF•H2O: Ellagic acid dihy drate (10.0 mg, 0.03 mmol) and caffeine (29.0 mg, 2.37 mmol) were added to 5 mL of ()1,2-propanediol/ethanol (6:4 volume ratio) and heated on a hotpla te until boiling occurred. The cont ents were cooled to room temperature and filtered to obtain a yellow solution. The solution was left to stand. Yellow oval plate-like crysta ls (3.9 mg, mp, 300C) were harvested after a week. ELATPH•0.13H2O: Ellagic acid dihydrate ( 10.0 mg, 0.03 mmol ) and anhydrous theophylline (27.0 mg, 0.15 mmol) were added to 5 mL of ethanol with heat until boiling

PAGE 91

72 occurred. The contents were cooled to room temperature and filtered to obtain a yellow solution. The solution was left for slow evapor ation. Yellow plate-like crystals (11.1 mg, 326C) were harvested the next day. ELATPH•2DMA: Ellagic acid dihydrate (50.0 mg, 0.15 mmol) and and anhydrous theophylline (134.0 m g, 0.75 mmol) were dissolved in 2 mL of DMA with heat. The solution was left for slow evapor ation. Yellow plate-like crystals (63.7 mg, 326C) were harvested after a week. ELADMP: Ellagic acid dihydrate (5.0 mg, 0.015 mmol) was dissolved in 5 mL of hot ()PG. 3,5-dimethylpyrazole (143.0 mg, 15 mmol) were added. The solution was left for slow evaporation. Yellow needle-like crys tals (2.9 mg, 237C) were harvested after one month. ELASAR: Ellagic acid dihydrate (10.0 m g, 0.03 mmol) was dissolved in 5 mL of hot methanol. Sarcosine (10.0 mg, 0.11 mmol) was dissolved in 5 mL of hot methanol. Both contents were mixed to give a yellow solution and left for slow evaporation. Yellow plate-like crystals (6.5 mg, 320oC) were harvested after two days. ELADMG: Ellagic acid dihydrate (10.0 mg, 0.03 mmol) was dissolved in 5 mL of hot methanol. N,N-dimethylglycine (10.0 mg, 0.11 mmol) was dissolved in 5 mL of hot methanol. Both contents were mixed to give a yellow solution and left for slow evaporation. Yellow plate-like crystals (5.6 mg, 293C) were harvested after two days. ELACAP: Ellagic acid dihydr ate (10.0 mg, 0.03 mmol) and -caprolactam (268.0 mg, 2.37 mmol) were added to 5 mL of isopropanol with heat. The content was

PAGE 92

73 filtered to obtain a yellow solution. It was allowed to slowly evaporate at room temperature. Yellow triangular plate-like crystals (6.2 mg, 2 76C) were harvested after a month. ELAURE•2NMP: Ellagic acid dihydrate (10.0 m g, 0.03 mmol) and urea (27.0 mg, 0.45 mmol) were di ssolved in 2 mL NMP/H2O (1:1 volume ratio) with heat. The solution was left to stand in air. Yellow block crystals (9.3 mg, 200C) were harvested after a week. ELADMA4: Ellagic acid dihydrate (50.0 m g, 0.15 mmol) was dissolved in 1 mL of DMA. The brown solution was left for sl ow evaporation. Colourless block crystals (45.0 mg) were harvested after a week. ELADMA2: Ellagic acid dihydrate ( 20.0 mg, 0.06 mmol) and oxalic acid (107.0 mg, 1.18 mmol) was dissolved in 1 mL of DMA with heat. The brown solution was left to stand in air. Yellow block crys tals (10.2 mg) were harvested after a week. ELADMSO2: Ellagic acid dihydrate (10.0 mg, 0.03 mmol) was dissolved in 1 mL of DMSO with heat. The brow n solution was left for slow evaporation. Yellow plate-like crystals (9.9 mg) were ha rvested after three days. ELANMP2: Ellagic acid dihydrate (50.0 m g, 0.15 mmol) was dissolved in 1 mL of NMP. The brown solution was left for sl ow evaporation. Yellow plate-like crystals (42.0 mg) were harvested after a week.

PAGE 93

74 ELAPG2: Ellagic acid dihydr ate (10.0 mg, 0.15 mmol) was dissolved in 2 mL of ()PG with heat. The yellow solution was filte red and left for slow evaporation. Yellow block crystals (5.0 mg) were harvested after a week. 3.4.2 Crystal Form Characterization Single crystals suitable for X-ray diffr action were selected using an optical microscope. The X-ray diffraction data were collected using Bruker-AXS SMART-APEX/APEXII CCD diffractomet er with monochromatized Cu K radiation ( = 1.54178) for all crystal sa mples, except for ELATPH•0.13H2O with monochromatized Mo K radiation ( = 0.71073). Indexing was performed using SMART v5.62537a or using APEX 2008v1-0.37b Frames were integrated with SaintPlus 7.5138 software package. Absorption correct ion was performed by multiscan method implemented in SADABS. The structures were solved using SHELXS-97 (Direct methods) and refined using SH ELXL-97 (full matrix nonlinear least-squares) contained in SHELXTL v6.1040 and WinGX v1.70.0141-43 pr ogram packages. All non-hydrogen atoms were refined anisotropically. A Bruker AXS D8 X-ray powder diffract ometer was used for all PXRD measurements with experimental parameters as follows: Cu K radiation ( = 1.54056 ); 40 kV and 30 mA; detector t ype: scintillation type; scanning interval: 3 40 2 ; time per step: 0.5 s. The experimental PXRD pa tterns and calculated PXRD patterns from single crystal structures were compared to confirm the composition of bulk materials.

PAGE 94

75 Thermal analysis was performed on a TA Instruments DSC 2920 differential scanning calorimeter. Aluminum pans were us ed for all samples and the instrument was calibrated using an indium standard. For refere nce, an empty pan sealed in the same way as the sample was used. Using inert nitrogen conditions, the samples were heated in the DSC cell from 30C to 350C at a rate of 10C/min. A Perkin-Elmer STA 6000 Simultaneous Th ermal Analyzer was used to conduct thermogravimetric analysis. Open alumina crucibles were used to heat the samples from 30C to 350C at a rate of 10 C/min under a nitrogen stream. Characterization of the co crystals by fourier transf orm infrared spectroscopy (FTIR) was accomplished with a Nicolet Avat ar 320 FT-IR instrument. Sample amounts of 1-2 mg were used, a nd spectra were measured over the range of 4000-400 cm-1 and analyzed using EZ Omnic software.

PAGE 95

76 Table 3-5. Selected hydrogen bond parame ters for ellagic acid cocrystals. Compound Hydrogen bond d (HA) / D (DA) / / ELANAM N(7)-H(71)•••O(7)#2 2.17 3.044 (7) 176.5 N(7)-H(72)•••O(15) 2.07 2.939 (7) 169.1 O(17)-H(17)•••N(1)#3 1.59 (6) 2.634 (5) 174 (5) O(11)-H(11)•••O(7)#2 1.85 2.694 (6) 179.3 #1 -x,-y,-z #2 -x+1,-y,-z+ 2 #3 x-1/2,-y+1/2,z-3/2 ELADMP O(2)-H(2)•••N(1)#3 1.58 (5) 2.588 (4) 163 (4) N(2)-H(1)•••O(4)#2 1.90 (5) 2.973 (4) 173 (4) #1 -x+2,-y+1,-z+1 #2 x,-y+1/ 2,z-1/2 #3 x,-y+1/2,z+1/2 ELASAR N(1)-H(1N)•••O(7) 1.99 (2) 2.905 (3) 163 (3) N(1)-H(2N)•••O(8)#1 2.24 (2) 3.006 (3) 142 (3) O(6)-H(6)•••O(10)#2 1.79 2.595 (3) 160.6 O(7)-H(7)•••O(9)#3 1.80 2.594 (2) 156.7 O(8)-H(8)•••O(9)#4 1.93 2.714 (2) 154.5 #1 x,y,z+1 #2 x,y-1,z-1 #3 -x,-y+1,-z+1 #4 -x,-y+1,-z ELADMG O(5)-H(3)•••O(10)#1 1.72 (2) 2.649 (1) 174 (2) O(6)-H(4)•••O(9)#2 1.86 (2) 2.682 (1) 157 (2) O(1)-H(1)•••O(9)#3 2.17 (2) 2.903 (2) 149 (2) O(2)-H(2)•••O(10)#4 1.63 (2) 2.587 (1) 170 (2) N(1)-H(5)•••O(5)#5 2.09 (2) 2.887 (2) 144 (2) #1 -x+1/2,-y+1/2,-z #2 x,y+1,z #3 x+1/2,y+1/2,z #4 -x+1,-y+1,-z #5 -x+1/2,y-1/2,-z+1/2 ELACAP O(11)-H(11)•••O(1) 1.96 (11) 2.724 (9) 160 (11) N(1)-H(1N)•••O(17)#2 2.15 (11) 2.903 (10) 130 (8) O(12)-H(12)•••O(1)#4 1.98 (8) 2.696 (9) 171(9) #1 -x+1,-y+1,-z+1 #2 x-1/2,-y-1/2, z-1/2 #3 -x+1/2,y-1/2,-z+1/2 #4 -x+1/2,y+1/2,-z+1/2

PAGE 96

77 Table 3-6. Selected hydrogen bond para meters for ellagic acid cocrystal solvates/hydrates. Compound Hydrogen bond d (HA) / D (DA) / / ELAINM•1.7H2O O(3)-H(3)•••O(11)#1 1.91 2.749 (2) 177.1 O(4)-H(4)•••N(11) 1.86 2.611 (2) 147.8 N(12)-H(12N)•••O(1)#2 2.03 (3) 2.930 (3) 174 (3) N(12)-H(13N)•••O(11)#3 2.06 (3) 2.988 (3) 173 (3) #1 -x-1,-y+1,-z #2 -x,-y+ 1,-z+1 #3 -x-2,-y+2,-z ELAINM•4NMP N(4)-H(16)•••O(7)#4 2.01 (5) 2.923 (4) 176 (5) O(4)-H(17)•••N(3)#3 1.78 (6) 2.647 (4) 147 (5) N(4)-H(10)•••O(1)#2 1.97 (5) 2.827 (5) 166 (4) O(3)-H(15)•••O(2)#3 1.72 (6) 2.615 (4) 166 (5) #1 -x+2,-y+1,-z+1 #2 x+1,y,z+1 #3 x-1,y,z #4 1-x,2-y,2-z ELAINM•4DMA N(2)-H(2A)•••O(1)#2 2.04 2.895 (3) 172.4 N(2)-H(2B)•••O(51) 2.01 2.792 (7) 150.2 O(3)-H(3)•••N(1) 1.916 (2) 2.647 (3) 148.1 (1) O(4)-H(4)•••O(6) 1.790 (2) 2.608 (3) 177.3 (1) #1 -x+1,-y+1,-z+2 #2 -x,-y+2,-z+1 ELACAF•H2O O(2)-H(2)•••N(4)#1 1.93 2.702 (4) 151.7 O(4)-H(4)•••O(6)#2 1.90 2.708 (3) 161.3 O(7)-H(7)•••O(1W)#3 1.78 2.608 (3) 167.2 O(8)-H(8)•••O(5)#4 1.93 2.701 (3) 152.4 O(1W)-H(1W)•••O(3)#3 2.08 (2) 2.920 (4) 164 (4) O(1W)-H(2W)•••O(9) 1.99 (2) 2.854 (4) 162 (4) #1 x+1,y,z-1 #2 x,y,z-1 #3 x-1,y,z #4 x,y+1,z ELATPH•0.13H2O N(4)-H(4)•••O(4)#2 1.80 (2) 2.714 (2) 169 (2) O(11)-H(111)•••O(2)#3 1.86 (3) 2.757 (2) 171 (2) O(12)-H(121)•••N(3)#4 1.82 (4) 2.711 (2) 166 (3) #1 -x+1,-y,-z+1 #2 -x+2,-y,-z+ 2 #3 -x+1/2,y-1/2,-z+3/2 #4 -x+1,-y,-z+2 ELATPH•2DMA O(3)-H(3)•••N(13)#2 2.09 2.849 (3) 154.7 O(4)-H(4)•••O(21)#3 1.75 2.548 (4) 165.6 N(14)-H(14A)•••O(11)#4 1.95 2.804 (3) 170.0 #1 -x+2,-y+1,-z+1 #2 x-1,y,z #3 x,y-1,z #4 -x+1,-y,-z

PAGE 97

78 ELAURE•2NMP O(2)-H(2)•••O(11)#1 1.81 (6) 2.543 (5) 169 (7) O(6)-H(11)•••O(12)#2 1.99 (7) 2.684 (4) 160 (7) O(5)-H(10)•••O(10)#3 1.39 (8) 2.581 (4) 166 (7) O(1)-H(1)•••O(9)#4 2.05 (6) 2.657 (5) 164 (7) N(3)-H(3A)•••O(11)#5 2.061 (3) 2.899 (6) 164.6 (3) N(4)-H(4B)•••O(6)#5 2.348 (3) 2.983 (5) 130.9 (3) N(3)-H(3B)•••O(8)#5 2.256 (3) 3.058 (5) 155.4 (3) N(2)-H(2A)•••O(12) 2.064 (3) 2.922 (5) 175.9 (3) N(1)-H(1A)•••O(12)#3 2.248 (3) 2.959 (6) 139.9 (3) #1 x,1+y,z #2 x,-1+y,z #3 -1+x,y,z #4 1-x,1-y,-z #5 1-x,-y,-z Table 3-7. Selected hydrogen bond para meters for ellagi c acid solvates. Compound Hydrogen bond d (HA) / D (DA) / / ELADMA4 O(3)-H(20)•••O(2)#2 1.70 2.5770 (16) 170.4 O(4)-H(21)•••O(1)#2 2.01 2.6148 (16) 129.9 #1 -x,-y+2,-z #2 -x+1,y+1/2,-z+1/2 ELADMA2 O(1)-H(1)•••O(5)#2 1.83 2.6696 (14) 172.8 O(2)-H(2)•••O(5) 2.08 2.8393 (14) 149.9 #1 -x,-y+2,-z+2 #2 -x+1,-y+1,-z+2 ELADMSO2 O(11)-H(11)•••O(1)#2 1.87 (5) 2.674 (3) 169 (5) O(12)-H(12)•••O(1)#3 1.97 (5) 2.695 (3) 153 (5) #1 -x+1,-y+2,-z+1 #2 x+1,y,z #3 -x+2,-y+1,-z+1 ELANMP2 O(1)-H(1)•••O(5)#2 2.09 (6) 2.745 (4) 150 (6) O(2)-H(2)•••O(5)#3 2.01 (6) 2.697 (4) 172 (8) #1 -x+1,-y,-z #2 x,y+1,z #3 -x+1/2,y+1/2,-z+1/2 ELAPG2 O(21)-H(21)•••O(14)#2 1.99 (5) 2.734 (3) 166 (5) O(12)-H(12)•••O(22)#3 1.84 (5) 2.660 (3) 164 (4) O(11)-H(11)•••O(21)#4 1.75 (5) 2.602 (3) 171 (5) #1 -x+1,-y+1,-z #2 x,y+1,z #3 x+1,y-1,z #4 x+1,y,z

PAGE 98

79 Table 3-8. Crystallographic data for ellagic acid cocrystals. ELANAM ELADMP ELASAR ELADMG ELACAP Formula C26H18N4O10 C24H22N4O8 C17H13NO10 C18H15NO10 C26H28N2O10 MW 546.44 494.46 391.28 405.31 528.51 Crystal system Monoclinic Monoclinic Tricli nic Monoclinic Monoclinic Space group P 21/n P 21/c P C 2 /c P 21/n a () 7.035 (3) 4.9393 (6) 7.6115 (2) 20.7781 (4) 12.452 (3) b () 22.963 (6) 18.155 (2) 9.4592 (2) 12.4012 (2) 6.0627 (15) c () 7.290 (2) 12.1795 (12)11.5037 (3) 13.0557 (2) 15.624 (4) (deg) 90 90 105.294 (2) 90 90 (deg) 106.50 (2) 93.451 (7) 109.068 (2) 107.333 (1) 91.562 (10) (deg) 90 90 91.848 (2) 90 90 V /3 1129.2 (6) 1090.2 (2) 748.69 (3) 3211.34(9) 1179.1 (5) Dc/mg m-3 1.607 1.506 1.736 1.677 1.489 Z 2 2 2 8 2 2 range 3.85 to 39.04 4.38 to 68.14 4.25 to 63.65 4.20 to 65.08 4.48 to 42.87 Nref./Npara. 623 / 172 1916 / 178 2369 / 263 2695 / 285 753 / 212 T /K 100 (2) 100 (2) 100 (2) 100 (2) 100 (2) R1 [I>2sigma(I)] 0.0375 0.0636 0.0431 0.0313 0.0515 w R2 0.0771 0.1404 0.1032 0.0844 0.1346 GOF 1.011 1.002 1.001 1.030 1.065 Abs coef. 1.077 0.972 1.269 1.206 0.974

PAGE 99

80 Table 3-9. Crystallographic data for ella gic acid cocrystal solvates/hydrates. ELAINM •1.7H2O ELAINM •4NMP ELAINM •4DMA ELACAF •H2O Formula C26H21.4N4O11.7 C46H54N8O14 C42H54N8O14 C22H18N4O11 MW 574.65 942.97 894.93 514.40 Crystal system Triclinic Triclinic Triclinic Triclinic Space group P P P P a () 3.6207 (2) 6.6365 (2) 6.6402 (4) 9.705(3) b () 12.4908 (6) 10.6281 (3) 10.4995 (6) 10.926(5) c () 13.5227 (6) 16.1445 (4) 16.2469 (8) 11.469(7) (deg) 72.641 (3) 87.391 (2) 86.469 (3) 104.74(3) (deg) 89.258 (3) 87.722 (2) 86.959 (3) 95.68(3) (deg) 83.979 (3) 78.243 (2) 78.012 (3) 113.80(3) V /3 580.39 (5) 1113.14 (5) 580.39 (5) 1047.9 (9) Dc/mg m-3 1.644 1.407 1.345 1.630 Z 1 1 1 2 2 range 3.42 to 65.36 2.74 to 67.72 2.73 to 68.33 4.09 to 65.36 Nref./Npara. 1916 / 210 3735 / 395 3710 / 354 3301 / 349 T /K 100(2) 100(2) 100(2) 100(2) R1 [I>2sigma(I)] 0.0439 0.0798 0.0593 0.0549 w R2 0.1093 0.2067 0.1638 0.1380 GOF 1.037 1.052 1.052 1.013 Abs coef. 1.136 0.881 0.855 1.150 ELATPH •0.07H2O ELATPH •2DMA ELAURE •2NMP Formula C28H2N8O12.13 C36H40N10O14 C52H64N12O24 MW 664.62 836.75 620.57 Crystal system Monoclinic Triclinic Triclinic Space group P 21/n P P a () 11.636 (3) 4.543 (3) 7.1468 (7) b () 6.8183 (19) 12.373 (6) 13.1766 (11) c () 16.868 (5) 17.486 (8) 14.7365 (12) (deg) 90 104.923 (8) 88.038 (5) (deg) 95.607 (6) 94.003 (13) 79.457 (6) (deg) 90 93.217 (12) 86.578 (6) V /3 1331.9 (6) 944.6 (8) 1361.5 (2) Dc/mg m-3 1.657 1.471 1.514 Z 2 1 2 2 range 2.04 to 25.03 2.63 to 65.49 3.05 to 67.71 Nref./Npara. 2339 / 271 3207 / 296 4548 / 416 T /K 100(2) 100(2) 100 (2) R1 [I>2sigma(I)] 0.0474 0.0576 0.0916 w R2 0.1108 0.1438 0.1966 GOF 1.099 1.013 1.459 Abs coef. 0.133 0.979 1.035

PAGE 100

81 Table 3-10. Crystallographic data for ellagic acid solvates. ELADMA4 ELADMA2 ELADMSO2 ELANMP2 ELAPG2 Formula C30H42N4O12 C22H24N2O10 C18H18O10S2 C24H24N2O12 C20H22O12 MW 650.68 476.43 458.44 500.46 454.38 Crystal system Monoclinic Triclinic Triclinic Monoclinic Triclinic Space group P 21/c P P P 21/n P a () 11.2531 (1) 7.0277 (1) 5.040 (2) 12.7369 (6) 5.0201 (4) b () 7.9074 (1) 7.8128 (1) 10.095 (5) 6.1858 (3) 7.2345 (5) c () 18. 9832 (3) 10.8493 (2) 10.304 (4) 14.1230 (8) 13.8053 (10) (deg) 90 90.697 (1) 71.61 (3) 90 74.893 (5) (deg) 109.448 (1) 107.440 (1) 77.13 (3) 96.182 (4) 87.069 (5) (deg) 90 113.140 (1) 76.07 (3) 90 77.273 (6) V /3 1592.80 (4) 516.822 (14) 476.7 (3) 1106.25 (10) 472.15 (6) Dc/mg m-3 1.357 1.531 1.597 1.502 1.598 Z 2 1 1 2 1 2 range 4.17 to 68.23 4.32 to 66.26 4.58 to 65.03 4.44 to 65.63 3.32 to 63.64 Nref./Npara. 2843 / 214 1716 / 159 1520 / 146 1878 / 172 1459 / 162 T /K 100 (2) 100 (2) 100 (2) 100 (2) 100 (2) R1 [I>2sigma(I)] 0.0460 0.0374 0.0440 0.0687 0.0464 w R2 0.1279 0.1077 0.1066 0.2169 0.1058 GOF 1.088 1.106 1.025 1.389 1.060 Abs coef. 0.887 1.041 3.065 1.005 1.157

PAGE 101

82 Chapter 4. Ionic Cocrystals: Crystal Forms of Lithium Salts 4.1 Preamble Bipolar disorder138 affects more than one percent of the population, is a disabling mental illness characterized by alternating “moo d swings” between periods of elevated or irritable mood (mania) and periods of depressi on. These mood swings can be very abrupt with occasional suicidal tendencies. There are three types of bipolar disorder (type I, II and cyclothymic), differing only in the severity of the m ood swings. Type I involves at least one manic or mixed episode (mania a nd depression occurring simultaneously) and one or more depressive episodes, that lasts for at least 7 days in a regular alternating pattern. Untreated patients average 4 episodes of mood swings each year with untreated mania lasting at least one week to months while depressive episodes can last 6 to 12 months. Type II involves major depressive ep isodes with occasiona l hypomania episodes, lasting at least 4 days. Patients suffer from significantly more depressive episodes with shorter periods of well-being between episodes compared to type I patients, leading to a higher risk for suicidal tendencies. Cyclothymic disorder is less severe compared to types I and II but is more chronic. It can deteriorate to t ype I or II in some patients or remain a mild chronic condition. In addition, there is bipolar disorder not otherwise specified (NOS) that does not fit into the three categor ies and bipolar disorder with rapid cycling. Bipolar disorder with rapid cycling is a temporary condition with at least 4 rapid alternating manic, hypomania or depressive episodes in a year.139

PAGE 102

83 Lithium (especially carbonate and citrat e salts) remains the “gold standard” for the treatment of bipolar disorder and th e prevention of relapses in patients.140 Further, it is the only medication that consis tently reduces suicidality in recurrent unipolar major depressive disorder an d in bipolar disorder.141 Lithium is readily absorbed from the gastrointestinal tract and attains its maximum concentr ation between to 3 hours. However, lithium treatment suffers from a na rrow therapeutic index with a therapeutic serum level between 0.5 mM to 1.5 mM. Lithium intoxication occurs with a serum level of 2.0 mM. Animal studies with adult male Spra gue-Dawley rats indicate lithium takes about 2 hours (Tmax) to achieve maximum serum concentration (Cmax) but 24 hours for maximum brain concentration.142 That lithium experiences difficulty in crossing the blood brain barrier coupled with the narrow therapeutic index, means any patient on lithium treatment for bipolar disorder must be monitored constantly to guard against lithium intoxication. From a clinical perspective, the Cmax in blood serum should best be quickly followed by the Cmax in brain, i.e., the time difference between the Tmax for blood serum and brain should be minimal. 4.1.1 Amino acids as the ideal cocrystal formers A pharmaceutical cocrystal143 is a cocrystal in which a pharmaceutically acceptable cocrystal former forms a supramolecular synthon48 with an API. That there are active transporters to facil itate the movement of ami no acids across the blood brain barrier144 implies that intuitively, amino acids may be the appropriate cocrystal formers to target ionic cocrystals145 of lithium salts. Furthermore, am ino acids as dipolar zwitterions

PAGE 103

84 are ideally suited to engage the lithium cation via the negatively-charged carboxylate while the positively-charged ammonium inter act with the countera nion, thus segregating the lithium cation from its counteranion by acting as a spac er. In an aqueous environment, the carboxylic acid moiety of the amino acid undergoes deprotonation readily with simultaneous protonation of the am ine moiety. In essence, the amino acid is a zwitterion that is dipolar with a positiv e-charged ammonium group and a negativecharged carboxylate moiety and has a total net charge of zero. The carboxylate moiety will bind strongly to the lithium cation due to its oxophilic character, leaving the ammonium group to hydrogen bond with the counteranion. Lithium chloride is the primary objective since th e chloride anion is abundant in stomach acid. In addition, other pharmaceutically acceptable anions such as br omide and nitrate will also be studied. 4.1.2 CSD survey of ionic co crystals of LiX (X = Cl-, Brand NO3 -) and amino acids The coordination chemistry of the lithium146 cation has been re viewed to provide in-depth analysis of the bond length, geom etry, coordination number and solvent of crystallization of lithium complexes ba sed on X-ray crystallographic data. The coordination numbers of lithium complexes ra nge from 4 to 8. But it prefers the 4coordinate tetrahedral coordi nation mode which is the most frequently encountered. The lithium cation is defined as a hard acid.147-149 Thus lithium cations are expected to bond to the oxygen atoms of ligands. A survey of the Cambridge Structural Database (CSD) on lithium chloride/bromide/nitrate complexes with amino acids and peptides reveals the

PAGE 104

85 coordination of the bridging carboxylates onto lithium cations while hydrogen bonding interaction between the anions and the ammoni um groups serves to reinforce the crystal packing forces. Table 4-1. Ionic cocrystals of lithium salts from the CSD. Lithium Salt Amino Acid Lithium: Amino Acid Stoichiometry CSD Refcode LiCl Glycine 1:1 HEFWUK150 GLY-GLY 1:1 HEFXEV150 L-proline 1:1 YOXBET151 LiBr ALA-GLY 1:1 ALGLYL152 GLY-GLY 1:1 GLYGLB GLY-GLY-GLY 1:1 GLYLIB LiNO3 Glycine 1:1 ALUNEA153 Glycine 1:2 ROZTUW154 Figure 4-1. A chain of fused six-atom rings in HEFWUK (top) and alternating eight-atom and four-atom ri ngs in ALUNEA (bottom).

PAGE 105

86 Analysis of these crystal structures rev eals the lithium cations in the 4-coordinate tetrahedral coordination environment. The carbox ylates bridge adjacent lithium cations to form either a chain of fused six-atom rings or a chain of alternating eight-atom and fouratom rings. This is a dominant feature in th e 1:1 ionic cocrystals of lithium salts and amino acids discussed in this chapter. 4.1.3 1:2 ionic cocrystals of LiX (X = Cl-, Brand NO3 -) and amino acids as structural analogues of zeolites and silica The interest in developing ionic cocrysta ls of lithium salts is two-fold. From a materials design perspective, lithium is the lig htest metal in the periodic table that prefers to adopt a 4-coordinate tetrahedral envir onment. The attention is drawn to zeolites (aluminosilicates) and silica (SiO2) as manifestations of tetrah edral nodes to create a rich diversity of 3-periodic architec tures. Synthetic inorganic ze olites typically consist of oxide anions that link tetrah edral aluminum or silicon ca tions (nodes) in a 2:1 ratio.155 The key to the existence of microporosity in zeo lites is that the oxide linkers are angular (M-O-M angles typically range from 140 to 165), thereby facilitating the generation of a wide range of topologies that are based upon rings, fused rings and polyhedral cages.

PAGE 106

87 Figure 4-2. Secondary building units (SBUs) in zeolites. Which particular topology exists for a given chemical composition is typically controlled by reaction conditions, counterio ns and/or structure directing agents156 (SDAs). The absence of counterions, SDAs or the use of a linear linker more typically manifests the tetrahedral node in the form of a diamondoid (dia) net157-158 that, unlike most zeolitic topologies, can interpenetrate to mitigate the creation of free space. The ground rules for generating zeolitic and/or dia networks are therefore selfevident and they have been validated across a remarkably diverse range of tetrahedral nodes (e.g. phosphates159, transition metal cations160-162, metal clusters163) and linkers (including purely organic ligands that form coordination bonds164 or hydrogen bonds165167). Coordination polymers that exploit the di versity of tetrahedral moieties and angular Frequency of occurrence given in parentheses

PAGE 107

88 or linear organic ligands have recently affo rded new levels of scale that includes new classes of zeolitic structures with hitherto unattainable levels of porosity. Such zeolitic metal-organic materials are exemplif ied by zeolitic imidazolate frameworks168-170 (ZIFs), boron imidazolate frameworks171 (BIFs), zeolite-like me tal-organic frameworks172 (ZMOFs) and the zeolite NPO su stained by metal-carboxylate cl usters only (denoted as CPM = crystalline porous material)173. ZIFs are based upon imidazolate ligands that subtend an angle of ca. 145 whereas the prototypal ZMOFs use 4,5-imidazoledicarboxylate174 and pyrimidine-based ligands175-176 in the presence of SDAs to coordinate to 8-coordinate metals such as In and Cd. BIFs are inherently of low density because they are based upon tetrahedral boron atoms.

PAGE 108

89 Figure 4-3. Synthetic zeolite based on imid azolates, LTA (top left ), RHO, (top right), RHO (BIF-9-Li, bottom left), SOD (botto m right). Yellow spheres highlight the empty polyhedral cages. Tetrahedral nodes are also found in se veral crystalline forms (polymorphs) of silica (SiO2). Among them, quartz is the only stable form under ambient conditions. Under normal pressure conditions, -quartz (trigonal polymo rph) converts to -quartz (hexagonal polymorph) at 573C which further converts to -tridymite (hexagonal polymorph) at 870C. -tridymite exhibits the rarely enc ountered lonsdaleite177-178 (lon)

PAGE 109

90 net. At 1470C, -tridymite converts to -cristobalite (cubic polymorph) which exhibits the dia net and finally melts at 1705C.179 The multitude of 3-periodic architectures from zeolites and silica has been constructed from h eavy atoms. That low density is a desirable property means that lithium, the lightest metal in the periodic table, is a particularly attractive target to serve as a tetr ahedral node in either zeolitic or dia networks. Figure 4-4. Polymorphs of silica, -cristobalite (top left), -tridymite (top right), quartz, (bottom left), fa ujasite (bottom right).

PAGE 110

91 Lithium forms many air and water stable coordination environments and not all existing zeolitic metal-organic materials are water stable. In this context, a prototypal structure was reported by Pinkerton et al. who isolated a lithium-based zeolitic ABW network with hexachlorotantalum anion embedded in what was described as a threedimensional Li-Cl-dioxane network180. However, this compound is extremely moisture sensitive. Bu and coworkers addressed the ch allenge elegantly in BIFs by employing both lithium and boron with imidazolates in BIF-9-Li, a compound with RHO topology181. Other approaches to low density porous materials based upon lithium include the following: Robson et al. reported a microporous lithium isonicotinate with square channels182; Henderson and coworkers isolated a pillared bilayer and a diamondoid net with solvated lithium aryloxides183; Parise et al. reported a MOF based on lithium and 2,5-pyridinedicarboxylic aci d that loses porosity upon solvent removal184. A new and general strategy to build lithiu m based zeolitic metal-organic materials (LiZMOMs) and lithium based diamondoid me tal-organic materials (LiDMOMs) by exploiting the Li-carboxylate-Li linkages that can be formed when amino acid zwitterions form cocrystals with lithium salts is described. The strategy described herein is based upon generating compounds in whic h there is a stoichiometric ratio of one lithium cation and two carboxylate anions. That carboxylate moieties can sustain dia nets is exemplified in a series of divalent metal formates that naturally possess the requi red 2:1 ratio of linker to node185-186. Indeed, such structures can even exhibit the rarely encountered lonsdaleite177-178 ( lon ) topology. However, that lithium is monovalent means that the requisite 2:1 ratio of linker to node will be very difficult to achieve with anionic linkers. Again, Bu and coworkers circumvented th is issue by supplemen ting neutral lithium

PAGE 111

92 imidazolates with neutral bifunctional N-donor ligands.187 The charge of the linker will not be an issue by using the carboxylate moietie s in amino acids, i.e. neutral zwitterions, to bridge two lithium cations. The target is a new class of compound: 2:1 cocrystals of amino acids and inorganic lithium salts. The use of amino acids provides numerous advantages and opportunities: 1) many amino acids are commercially available and they are typically inexpensive; 2) Figure 5 reveals that lithiu mcarboxylate-lithium angles offer the requisite diversity needed to generate a wide range of exte nded structures; 3) the lithium-carboxylate bond is robust even in the presence of water; 4) amino acids possess functionalized side chains that facilitates fine-tuning of the resu lting structures through pre-synthetic methods; 5) the abundance of homochiral amino acids means that homochiral crystals with optical activity an d bulk polarity are guaranteed; 6) most amino acids are soluble in water, therefore facilita ting green synthesis; 7) the charge of the network is inverted compared to zeolites because the framework is cationic and the required counterions are anions, i.e. anion exchange becomes feasible. The remarkable range of Li-carboxylate-Li angles stems pa rtly from the tendency of the carboxylate ligand to exhibit eith er endodentate or exode ntate bridging modes.

PAGE 112

93 Figure 4-5. Distribution of the Li-C-Li angle in crystal structures that contain lithium cations bridged by carboxylate mo ieties (data obtained using CSD version 5.31). 4.2 Results and Discussion Table 4-2. Table of zwitte rionic cocrystal formers. H2 +N OO SAR NH+ OO DMG N+ OO BTN +H3N OO BAL OO +H3N ABA N H2 + O OPRO The amino acids listed in table 4-2 are co mbined with lithium sa lts in two distinct stoichiometries to achieve a wide variety of ionic cocrystals. 6080100120140160180 0 10 20 30 40 FrequencyLi-C-Li bridging angle

PAGE 113

94 4.2.1 1:1 ionic cocrystals of LiX and amino acids: A discrete supermolecule and linear chains A series of 1:1 ionic cocrystals and cocr ystal hydrates of lithium salts (lithium chloride, LIC, lithium bromide, LIB and lithium nitrate, LIN) and amino acids (sarcosine, SAR, N,N-dimethylglycine DMG, betaine, BTN, -alanine, BAL, 4-aminobutyric acid and L-proline, PRO) are synthesized and characterized. A discrete assembly and 1periodic chains were obtained from these 1:1 cocrystals and cocrystal hydrates. Figure 4-6. A chain of fused six-at om rings is observed in LICSAR. LICSAR: The crystal structure reveals that LICSAR is a 1:1 lithium chloride and sarcosine cocrystal monohydrate. Each lithiu m cation is bridged by three carboxylates [Li-O: 1.895(12) , 1.932(13) and 1.959(13) ] with one coordinated water molecule [Li-O: 1.960(12) ] to fulfill the 4-coordinate tetrahedral environment. It forms a chain of fused six-atom rings. Adjacent chains are linked by hydrogen bonding interaction

PAGE 114

95 between the chloride anions and the ammonium groups [NH•••Cl: 3.145(5) and 3.114(5) ] and OH(water) •••O(carboxyl ate) interaction [2.951(7) ]. Figure 4-7. The nitrate anions coordina te to the lithium cations in LINSAR. LINSAR: Analysis of the crystal structure reveals a 1:1 cocrystal of lithium nitrate and sarcosine. Each lithium cati on is bridged by three carboxylates [Li-O: 1.939(4) , 1.952(4) and 1.956(5) ] wi th one coordinated nitrate anion [Li-O: 2.017(5) ] to fulfill the 4-coordinate tetrahe dral environment. It forms a chain of fused six-atom ring. Adjacent chains are linke d by hydrogen bonding interaction between the nitrate anions and the ammonium groups [NH•••O: 2.896(3) and 2.918(3) ].

PAGE 115

96 Figure 4-8. Li2(DMG)2 clusters are hydrogen bonded in to a 1-periodic structure. LICDMG: The X-ray structure reveals a dihydrate of a 1:1 cocrystal of lithium chloride and N,N-dimethylglycine. Each lithium cati on is bridged by the same pair of carboxylates [Li-O: 1.926(3) and 1.932(2) ] with two coordinated water molecules [Li-O: 1.909(3) and 1.943(3) ] to estab lish a discrete supermolecular assembly. OH(water)•••O(carboxylate) [2.7543(14) ] interact ions align these disc rete units into a linear chain. The chloride anions act as hydrogen bond acceptors towards three water molecules [OH•••Cl: 3.1264(12) , 3.1452(12) and 3.1686(11) ] and an ammonium group [NH•••Cl: 3.2204(12) ] to tether adjacent chains together.

PAGE 116

97 Figure 4-9. The nitrate anions are coordi nated to the lithium cations in LINDMG. LINDMG: The crystal structure reveals a 1:1 cocrystal of lithium nitrate and N,Ndimethylglycine.. Each lithium cation is b ridged by three carboxylat es [Li-O: 1.919(3) , 1.934(3) and 1.937(3) ] with one coordi nated nitrate anion [Li-O: 1.977(3) ] to fulfill the 4-coordinate tetrahedral environment. It forms a chain of fused six-atom ring. Adjacent chains are linked by hydrogen bonding interaction between the nitrate anions and the ammonium groups [NH•••O: 2.8349(1) ]. Figure 4-10. An undulating chain is co nstructed by bridging discrete (LiCl)2(H2O)2 clusters with betaine zwitterions.

PAGE 117

98 LICBTN: The X-ray structure reveals a 2: 1 lithium chloride and betaine cocrystal dihydrate. Each lithium cation is coordina ted by a bridging carboxylate [Li-O: 1.849(5) and 1.865(6) ], two water molecules [Li-O: 1. 997(5) , 2.023(5) and 2.011(5) ] and a coordinated chloride anion [Li-Cl: 2.319(4) and 2.329(4) ]. Two lithium cations and two bridging water molecules form a four-atom ring cluster. The four-atom ring cluster alternates with bridging carboxylates to fo rm an undulating chain. Each chloride anion accepts two hydrogen bonds from bridging wa ter molecules [OH•••Cl: 3.065(2) , 3.117(2) , 3.120(2) and 3.178(2) ] to cr osslink adjacent chains. Crystals of LICBTN deliquesce within minutes and do not allow further analysis to be possible. Figure 4-11. A chain of fused alternating eight-atom rings and four-atom rings in LICBAL-anhydrate. LICBAL-anhydrate: X-ray analysis reveal s a 1:1 cocrystal of lithium chloride and -alanine. Each lithium cation is bridged by three carboxylates [Li-O: 1.915(3) , 1.920(3) and 1.942(3) ] with a coordina ted chloride anion [Li-Cl: 2.304(3) ] to attain tetrahedral coordination environment. An eight-atom ring is constructed from two

PAGE 118

99 lithium cations and two bridging carboxylates. Adjacent eight-atom rings are connected by forming a four-atom ring to yield a chai n. Adjacent chains are linked by hydrogen bonding interaction between the ammonium gr oups with chloride anions [NH•••Cl: 3.1983(14) and 3.2433(14) ] and adjacen t carboxylates [NH•••O: 2.8251(17) ]. Figure 4-12. Water molecules coordinate to the lithium cations in LICBAL. LICBAL: The crystal structure reveals a 1:1 lithium chloride and -alanine cocrystal monohydrate. Each lithium cation is bridged by three carboxylates [Li-O: 1.928(3) , 1.933(3) and 1.943(3) ] with one coordinated water molecule [Li-O: 2.002(3) ] to fulfill the 4-coordinate tetrahe dral environment. It forms a chain of fused six-atom rings. Adjacent chains are linke d by hydrogen bonding interaction between the chloride anions and the ammonium groups. Each chloride anion acts as hydrogen bond acceptors towards two water molecules [O H•••Cl: 3.1262(17) and 3.2458(14) ] and two ammonium groups [NH•••Cl: 3.222(2) and 3.2475(15) ].

PAGE 119

100 Figure 4-13. The chloride anions are coordi nated to the lithium cations in LICABA. LICABA: Analysis of the crystal struct ure reveals a 1:1 cocrystal of lithium chloride and 4-aminobuytric acid. Each lith ium cation is bridged by three carboxylates [Li-O: 1.900(3) , 1.965(3) and 1.976(3) ] with one coordinated chloride anion [LiCl: 2.289(2) ] to achieve the tetrahedral coordination geometry. It forms a chain of fused six-atom rings. Adjacent chains ar e crosslinked by hydrogen bonding interaction between the ammonium groups acting as hydrogen bond donors towards neighbouring chloride anions [NH•••Cl: 3.1516(13) a nd 3.1886(13) ] and ad jacent carboxylates [NH•••O: 2.7970(16) ]. LICPRO: The crystal structure reveals a 1:1 lithium chloride and L-proline cocrystal monohydrate. Each lithium cation is bridged by three carboxylates [Li-O: 1.956(5) , 1.962(7) and 1.993(8) ] with one coordinated water molecule [Li-O:

PAGE 120

101 1.996(5) ] to fulfill the 4-coordinate tetrahe dral environment. It forms a chain of fused six-atom rings. Adjacent chains are crossl inked by hydrogen bonding interaction between the chloride anions and the ammonium groups Each chloride anion acts as hydrogen bond acceptors towards two water molecules [O H•••Cl: 3.194(3) and 3.234(3) ] and two ammonium groups [NH•••Cl: 3.187(3) ]. Figure 4-14. A chain of fused six-atom rings in LICPRO (top) and LIBPRO (bottom).

PAGE 121

102 LIBPRO: X-ray analysis reveals a monohydrate of 1:1 cocrystal of lithium bromide and L-proline. Each lithium cation is bridged by three carboxylates [Li-O: 1.937(4) , 1.941(4) , 1.950(5), 1.970(6) an d 1.987(6) ] with one coordinated water molecule [Li-O: 1.939(4) and 1.971(4) ] to fulfill the 4-coordinate tetrahedral environment. It forms a chain of fused six-atom rings. Adjacent chains are linked by hydrogen bonding interaction between the br omide anions and the ammonium groups. Each bromide anion acts as hydrogen bond acceptors towards two water molecules [OH•••Br: 3.312(3) , 3.322(3) , 3.345(3) and 3.353(3) ] and two ammonium groups [NH•••Br: 3.277(3) , 3.289(3) and 3.326(3) ]. Figure 4-15. An adamantane cage (left) is constructed with L-proline zwitterions and nitrate anions as bridging ligands, hexagonal channels in the dia net (right) are occupied by the pyrrolidinium ring of the zwitterions to mitigate interpenetration. LINPRO: The X-ray structure reveals a 1:1 lithium nitrate and L-proline cocrystal. Each lithium cation is bri dged by two carboxylates [Li-O: 1.882(3) and 1.888(3) ]

PAGE 122

103 and two nitrate anions [Li-O: 1.974(3) and 1.998(3) ] to form a neutral mixed ligand dia net with distorted hexagonal channels exhibiting diameters ranging from 5.095 to 9.288 , populated by the pyrrolidinium ring of the zwitterions [figur e 4-15(right)]. The framework is reinforced by hydrogen bonding be tween the ammonium of an amino acid with an adjacent carboxylate [N-HO: 2.898( 2) ] of another amino acid and a neighbouring nitrate anion [N-HO: 2.842(2) ]. The presence of pairs of zwitterions in these hexagonal channels interacting with the framework renders interpenetration impossible. 4.2.2 1:2 ionic cocrystals of LiX and amino acids: Square grids, Diamondoids and a Zeolitic ABW net In principle, 1:2 cocrystals of lithium salts and amino acids fit the criteria for formation of dia and/or zeolitic frameworks. Howeve r, another structural feature of zeolites is the presence of one or more rings, typically 4-, 6or 8-membered MnOn rings. 4-membered Li4(carboxylate)4 rings have previously been ob served in a 1:2 cocrystal of lithium nitrate and glycine demonstrating a square grid network154 (Refcode ROZTUW, Li-C-Li angles 115.88, 117.83). With the struct ural prerequisites for self-assembly of zeolitic and/or dia networks in mind, a series of 1:2 io nic cocrystals of lithium salts (LIC, LIB and LIN) and amino acids (SAR, DM G, BTN and PRO) are synthesized and characterized. Cocrystals of the desired stoi chiometry were prepared by slow evaporation of aqueous solutions of LIC, LIB or LIN and two equivalents of the amino acid at ca 80oC. These cocrystals are stable to at least 200 C and are freely soluble in water.

PAGE 123

104 Crystallographic analysis of the products revealed that th ree distinct networks were observed: square grids based upon only 4-membered Li4(carboxylate)4 rings; diamondoid networks based upon only Li6(carboxylate)6 rings; a zeolitic ABW network based upon 4membered, 6-membered and 8-membered Lin(carboxylate)n (n=4, 6 and 8) rings. Figure 4-16. Structural diversity in the netw orks formed by 1:2 cocrystals of lithium salts with amino acids: (a) square grids; (b) diamondoid LiDMOM nets; (c) zeolitic ABW topology in the first LiZMOM. Hydrogen atoms and counteranions are omitted for clarity. Figure 4-17. Undulating square grid (lef t) and bilayered packing arrangement (right) in LICSAR2. (a) (b) (c)

PAGE 124

105 LICSAR2: Each lithium cation is bridged by four carboxylates [Li-O: 1.905(3) , 1.916(3) , 1.928(3) and 1.966(3) ] to form an undulating square grid [figure 4.17(left)], while the opposite ends of the amino acids point away (above and below) from the square grid to establish a bilayer packing arrangement [figure 4.17(right)]. The chloride anions reside at the interface of the ammonium groups, sustained by N-HCl [3.1011(1) , 3.1549(1) ] interactions. Thes e square cavities are about 5.0 by 6.0 and form undulating sheets that st ack in a roughly eclipsed manner. Figure 4-18. Undulating square grid (lef t) and bilayered packing arrangement (right) in LINBTN2. LINBTN2: Each lithium cation is bridged by four carboxylates [Li-O: 1.931(3) , 1.933(3) , 1.958(3) and 1.973(3) ] to form an undulating square grid [figure 418(left)], while the opposite ends of the amino acids point away (above and below) from the square grid to establish a bilayer packi ng arrangement [figure 4-18 (right)]. The nitrate anions reside at the interface of the amm onium groups, sustained by C-HO [3.154(2)

PAGE 125

106 to 3.606(2) ] interactions. These square cavities are about 5.5 by 5.7 and form undulating sheets that stack in a roughly eclipsed manner. Figure 4-19. Hexagonal channels in dia nets of LICDMG2 (left) and LIBDMG2 (right). LICDMG2: Each lithium cation is bridge d by four carboxyla tes [Li-O distances: 1.898(3) and 1.910(3)] to form a cationic dia net with hexagonal channels exhibiting diameters ranging from 10.6 to 12.0 , popul ated by the chloride anions [figure 4-19 (left)]. The framework is reinforced by hydrogen bonding between the carboxylate of one amino acid and the ammonium of an adja cent amino acid [N-HO, 2.742(2) ]. The presence of pairs of chloride anions [C-HC l, 3.628(3) , 3.633(2) and 3.635(2) ] in these hexagonal channels interacting wi th neighbouring methyl groups renders interpenetration impossible.

PAGE 126

107 LIBDMG2: Each lithium cation is bridge d by four carboxylates [Li-O: 1.908(3) and 1.942(3) ] to form a cationic dia net with hexagonal channels exhibiting diameters ranging from 10.7 to 12.1 , populated by the bromide ani ons [figure 4-19(right)]. The framework is reinforced by hydrogen bonding be tween the carboxylate of one amino acid and the ammonium of an adjacent amino acid [N-HO, 2.747(2) ]. The presence of pairs of bromide anions [C-HBr, 3.716(3) , 3.731(2) and 3.772(2) ] in these hexagonal channels interacting with neighbour ing methyl groups rende rs interpenetration impossible. LIBDMG2 is isostructural with LICDMG2. LICPRO2: Each lithium cation is bri dged by four carboxylates [Li-O: 1.9341(18) and 1.9536(19) ] to form a cationic dia net with hexagonal channels exhibiting diameters ranging from 10.1 to 12.5 , popu lated by the chloride anions [figure 420(left)]. The framework is reinforced by hydrogen bonding between the carboxylate of one amino acid and the ammonium of an adjacent amino acid [N-HO, 2.7404(15) ]. The presence of pairs of chloride anions [N-HCl, 3.1322(12) ] in these hexagonal channels interacting with neighbouring ammonium groups renders interpenetration impossible.

PAGE 127

108 Figure 4-20. Hexagonal channels in dia nets of LICPRO2 (left) and LIBPRO2 (right). LIBPRO2: Each lithium cation is bridged by four carboxylates [Li-O: 1.939(2) and 1.974(3) ] to form a cationic dia net with hexagonal channels exhibiting diameters ranging from 10.2 to 12.6 , populated by the bromide anions [figure 4-20(right)]. The framework is reinforced by hydrogen bonding be tween the carboxylate of one amino acid and the ammonium of an adjacent amino aci d [N-HO, 2.737(2) ]. The presence of pairs of bromide anions [N-HBr: 3.2769(15) ] in these hexagonal channels interacting with neighbouring ammonium gr oups renders interpenetration impossible. LIBPRO2 is isostructural with LICPRO2.

PAGE 128

109 Figure 4-21. Hexagonal channels in LINPRO2 (dia, top) and octagonal channels in LINPRO2 (ABW, bottom). LINPRO2 (dia): Each lithium cation is bridged by four carboxylates [Li-O: 1.921(12) , 1.937(12) , 1.959(13) a nd 1.965(12) ] to form a cationic dia net with hexagonal channels exhibiting diameters ranging from 10.6 to 12.3 , populated by the

PAGE 129

110 nitrate anions [figure 4-21(top)]. The fram ework is reinforced by hydrogen bonding between the carboxylate of one amino acid and the ammonium of an adjacent amino acid [N-HO, 2.751(5) , 2.762(5) ]. The presence of pairs of nitrate anions [N-HO, 2.741(6) , 2.870(6) , 3.046(6) ] in these hexagonal channels interacting with neighbouring ammonium groups renders interpenetrati on impossible. LINPRO2 (ABW): A combination of Li4(carboxylate)4 and Li8(carboxylate)8 rings generate channels that lie parallel to the crystallographic a axis. When viewed down the crystallographic c axis, the presence of Li6(carboxylate)6 rings is evident. Li-O bond distances in the range of 1.913(3) -1.976(3) occur in the 4and 8-membered rings while Li-O bond distances in th e range of 1.920(2) -1.976(3) are observed in the 6membered rings. Pairs of nitrate anions occupy the 8-membered ring channels and they are crystallographically ordered thr ough N-HO hydrogen bonding interactions [2.735(8), 2.885(8), 3.005(8)]. The dimensi ons of the largest 8-membered ring channel are about twice that of an AB W zeolite as illustrated in figure 4-22. Although the ABW form of LINPRO2 is stable at elevated temperatures, it converts to the diamondoid form, LINPRO2( dia ), upon standing in mother liquor under ambient conditions. Grinding of the plate-like crystals of LINPRO2(ABW) also results in conversion to LINPRO2( dia ) as determined by powder X-ray diffraction (PXRD). Interestingly, ABW zeolite has been observed as an intermediate phase in inorganic zeolite synthesis188.

PAGE 130

111 Figure 4-22. Significant expansion of dimensions occurs because of the ditopic carboxylate linker in the ABW-type LiZMOM LINPRO2 when compared to prototypal ABW zeolite. Whereas the Li-carboxylat e bond distances observed in the structures reported herein exhibit a relatively narrow range, th e Li-carboxylate-Li angles range from 117.78o [Li4(carboxylate)4 ring in LICSAR2] to 180o [Li8(carboxylate)8 ring in LINPRO2(ABW)] and are detailed in Table 4-3. The majority of angles cluster around 150o, intermediate between those for linear and tetrahedral geometry, which is consistent with what would be needed to form a wider range of zeolitic structures.

PAGE 131

112 Table 4-3. Analysis of Li-C-Li angles form ing 4-, 6and 8-membered ring motifs. 4-membered ring motif 6-membered ring motif 8-membered ring motif LICSAR2 117.78, 157.08 LINBTN2 122.77, 144.37 LICDMG2 153.27 LIBDMG2 153.27 LICPRO2 156.00 LIBPRO2 155.50 LINPRO2 (dia) 154.84, 158.77 LINPRO2 (ABW) 122.13, 148.72 148.72, 155.18, 158.97 122.13, 148.72, 155.18, 158.97 ABW 149.4, 156.8 156.8, 156.9, 180149.4, 180 4.3 Conclusion 19 crystal forms (13 cocrystals and 6 cocrys tal hydrates) of lithi um salts (chloride, bromide, nitrate) have been presented here in. Hydration appears to be a phenomenon in these ionic cocrystals, in particular, the 1: 1 cocrystals. That the 1:1 cocrystals with lithium nitrate and the 1:2 cocrystals escape hydration suggests that perhaps hydration is predictable in this situation. Given their ub iquity, relevance and that many hydrates are discovered due to the adventitious water mo lecules, it should be unsurprising that a number of researchers have addressed the frequency189, formation190 and the water environment of organic crystalline hydrates191-194 and crystalline hydrates have been classified according to their structure or energetics.194 Morris and Rodriguez-Hornedo195 proposed a classification system whereby hydrat es are subdivided into three classes ( 1 ) channel hydrates in which water molecules in teract with each other to form tunnels within the crystal lattice, ( 2 ) isolated site hydrates in which water molecules are not

PAGE 132

113 directly hydrogen bonded to each other, ( 3 ) metal ion associated hydrates in which water molecules form strong interac tions with transition metals or alkali metals. The 1:1 cocrystal hydrates fall into the third categor y where water molecules are coordinated to lithium cations, as evidenced from the crystal structures. The isolation and characterization of both the anhydrate and hydr ate will be advantageous to discern the hydration trend. In fact, bot h the anhydrate and monohydrate of LICBAL are isolable. Whereas a coordinated chloride anion [Li-Cl: 2.304 (3) ] is observe d in the anhydrate, a water molecule displaces the chloride and coordinates to the lithium cation [Li-O: 2.002(3) ] in the monohydrate. A comparison of the respective Li-Cl distance and Li-O distance confirms the strength of the Li-O bond over the Li-Cl bond. This is expected from the oxophilic character of the lithium ca tion and also explains why lithium chloride is a hygroscopic solid. The c oordination sphere of lithium cations is surrounded by 4 oxygen atoms in the 1:1 cocrystals with lithium nitrate and the 1:2 cocrystals and therefore hydration cease in the solid state. These crystal forms of lithium salts have generated a rich diversity of structure types from 0-periodic to 3-periodic. All 1: 1 cocrystals are asse mbled into 1-periodic chains, including a discrete supermolecule that is hydrogen bonded into a chain in LICDMG. The only exception is LINPRO which exists as a mixed ligand dia net, where the nitrate anion plays the same role of a bridging ligand as the carboxylate moiety of the zwitterion. The architectures constructed from 1:2 cocrystals feature 2-periodic and 3periodic nets, composed of 4-, 6and 8-membered Lin(carboxylate)n (n=4, 6, 8) rings, reminiscent of square grids, dia nets and a zeolitic ABW net. This approach consists of one-step synthesis from readily available star ting materials and shoul d be general because

PAGE 133

114 of the modular nature of these compounds and the ready availability of amino acids and counteranions. The use of larger anions an d/or templates coupled with the expected diversity of Li-carboxylate-Li angles will en visage a range of LiZMOMs that parallels the structural diversity seen in inorganic zeolites and zeolitic metal-organic materials. 4.4 Materials and Methods All chemicals purchased were used as received without further purification. 4.4.1 Synthesis of Crystal Forms of Lithium Salts LICSAR: Lithium chloride (100.0 m g, 2.36 mmol) and sa rcosine (211.0 mg, 2.36 mmol) were dissolved in 0.5 mL of deioni sed water and left for slow evaporation. Colourless plate-like crysta ls (200 mg, m.p. 245C) were harvested after three days. LINSAR: Lithium nitrate (413.4 m g, 6.0 mmol) and sarcosine (534.5 mg, 6.0 mmol) were dissolved in 1 mL of deionise d water. It was mainta ined on the hot plate until crystals emerged from the hot solution. Colorless crystals (368.4 mg, 208C) were collected from the hot solution. LICDMG: Lithium chloride (50.0 mg, 1.18 mmol) and N,N -dimethylglycine (122.0 mg, 1.18 mmol) were dissolved in 0.75 mL of hot deionised water. It was left for slow evaporation. Colourless block crystals (101 mg, 272C) were harvested after one month.

PAGE 134

115 LINDMG: Lithium nitrate (413.4 mg, 6.0 mmol) and N,N -dimethylglycine (618 mg, 6.0 mmol) were dissolved in 3 mL of hot deionised water. It was maintained on the hot plate until crystals emerged from the hot solution. Colo rless plates (595 mg, 202C) were collected from the hot solution. LICBTN: Lithium chloride (1.00 g, 23.6 mmol) and betaine (2.76 g, 23.6 mmol) were dissolved in 1 mL of hot deionised wate r. It was maintained on the hot plate until crystals emerged from the hot solution. Colorless block crysta ls were collected from the hot solution. The crystals were very hygrosc opic and deliquesced into liquid rapidly once out of the mother liquor. LICBAL-anhydrate: Lithium ch loride (1.00 g, 23.6 mmol) and -alanine (2.11 g, 23.6 mmol) were dissolved in 2 mL of hot dei onised water. It was maintained on the hot plate until crystals emerged from the hot solution. Colourless r od crystals (2143 mg, 252C) were collected from the hot solution. LICBAL: Lithium chloride (1.00 g, 23.6 mmol) and -alanine (2.11 g, 23.6 mmol) were dissolved in 2 mL of hot deio nised water. It was maintained on the hot plate until crystals emerged from the hot solu tion. The contents were allowed to cool. The mother liquor was decanted and 0.5 mL of deionised water was added. The colourless plate crystals (1910 mg, 252C) were harvested the next day. LICABA: Lithium chloride (1.00 g, 23.6 mmol) and 4-aminobut yric acid (2.44 g, 23.6 mmol) were dissolved in 2 mL of hot dei onised water. It was maintained on the hot plate until crystals emerged from the hot solution. Colourless r od crystals (2244 mg, 241C) were collected from the hot solution.

PAGE 135

116 LICPRO: Lithium chloride (200.0 mg, 4.72 mmol) and L-proline (544.0 mg, 4.72 mmol) were dissolved in 0.75 mL of dei onised water and left for slow evaporation. Colourless plate crystals (609 mg, 281C) were harvested after three days. LIBPRO: Lithium bromide (1.00 g, 11.5 mmol) and L-proline (1.33 g, 11.5 mmol) were dissolved in 2 mL of hot deionised wate r. It was maintained on the hot plate until crystals emerged from the hot solution. Col ourless block crystals (1427 mg, 280C) were harvested from the hot solution. LINPRO: Lithium nitrate (413.4 mg, 6.0 mmol) and L-proline (690.0 mg, 0.599 mmol) were dissolved in 1.5 mL of dei onised water and was maintained on the hot plate until crystals emerged from the hot solution. Colorless plates (473.8 mg, 232C) were collected from the hot solution. LICSAR2: Lithium chloride (0.5 g, 11.8 mmol) and sarcosine (3.15 g, 35.4 mmol) were dissolved in 2 mL of hot deionised wate r. It was maintained on the hot plate until crystals emerged from the solution. Colourless block crystals (1213 mg, 247oC) were harvested from the hot solution. LINBTN2: Lithium nitrate (413.4 mg, 6.0 mmol) and betaine (1405.6 mg, 12.0 mmol) were dissolved in 2 mL of hot deio nised water. It was maintained on the hot plate until crystals emerged from the hot solution. Colorless plates (357 mg, 286oC) were collected from the hot solution. LICDMG2: Lithium chloride (0.5 g, 11.8 mmol) and N,N -dimethylglycine (2.44 g, 23.6 mmol) were dissolved in 3 mL of hot deionised water. It was maintained on

PAGE 136

117 the hot plate until crystals emerged from the hot solution. Colourless block crystals (1488 mg) were collected from the hot solution. LIBDMG2: Lithium bromide (0.5 g, 5.76 mmol) and N,N -dimethylglycine (1.19 g, 11.5 mmol) were dissolved in 2 mL of hot deionised water. It was maintained on the hot plate until crystals emerged from the hot solution. Colourless block crystals (696 mg, 289C) were collected from the hot solution. LICPRO2: Lithium chloride (0.50 g, 11.8 mmol) and L-proline (2.72 g, 23.6 mmol) were dissolved in 2 mL of hot deio nised water. It was maintained on the hot plate until crystals emerged from the hot solution. Colourless block crystals (1266 mg, 250oC) were harvested from the hot solution. LIBPRO2: Lithium bromide (0.50 g, 5.76 mmol) and L-proline (1.33 g, 11.5 mmol) were dissolved in 2 mL of hot deio nised water. It was maintained on the hot plate until crystals emerged from the hot so lution. Colourless block crystals (861 mg, 257oC) were collected from the hot solution. LINPRO2: Lithium nitr ate (413.4 mg, 6.0 mmol) and L-proline (1381.2 mg, 12.0 mmol) were dissolved in 2 mL of hot deio nised water. It was maintained on the hot plate until crystals emerged from the hot solution. Colorless plat es (LINPRO2, ABW) were collected from the hot solution wh ich converted to rhombohedral crystals (LINPRO2, dia 603 mg) once they are removed from the hot plate.

PAGE 137

118 4.4.2 Crystal Form Characterization Single crystals suitable for X-ray diffr action were selected using an optical microscope. The X-ray diffraction data were collected using Bruker-AXS SMART-APEX/APEXII CCD diffractomet er with monochromatized Cu K radiation ( = 1.54178 ) for all crystal samples. Inde xing was performed using SMART v5.62537a or using APEX 2008v1-0.37b Frames were inte grated with SaintPlus 7.5138 software package. Absorption correction was performed by multiscan method implemented in SADABS.39 The structures were solved us ing SHELXS-97 (Direct methods) and refined using SHELXL-97 (full matrix nonlinear le ast-squares) contai ned in SHELXTL v6.1040 and WinGX v1.70.0141-43 program packages. All non-hydrogen atoms were refined anisotropically. A Bruker AXS D8 X-ray powder diffract ometer was used for all PXRD measurements with experimental parameters as follows: Cu K radiation ( = 1.54056 ); 40 kV and 30 mA; detector type: scintillat ion type; scanning interval: 3 40 2 ; time per step: 0.5 s. The experimental PXRD pa tterns and calculated PXRD patterns from single crystal structures were compared to confirm the composition of bulk materials. Thermal analysis was performed on a TA Instruments DSC 2920 differential scanning calorimeter. Aluminum pans were us ed for all samples and the instrument was calibrated using an indium standard. For refere nce, an empty pan sealed in the same way as the sample was used. Using inert nitrogen conditions, the samples were heated in the DSC cell from 30C to 300C at a rate of 10C/min.

PAGE 138

119 A Perkin-Elmer STA 6000 Simultaneous Th ermal Analyzer was used to conduct thermogravimetric analysis. Open alumina crucibles were used to heat the samples from 30C to 300C at a rate of 10 C/min under a nitrogen stream. Characterization of the cocrystals by FTIR was accomplished with a Nicolet Avatar 320 FT-IR instrument. Sample amounts of 1-2 mg were used, and spectra were measured over the range of 4000-400 cm-1 and analyzed using EZ Omnic software.

PAGE 139

120 Table 4-4. Selected hydrogen bond parameters for 1:1 ionic cocrystals. Compound Hydrogen bond d (HA) / D (DA) / / LICSAR O(3)-H(1O)•••O(2)#1 2.305 (5) 2.951 (7) 170.6 (5) N(6)-H(1N)•••Cl(1)#7 2.249 (2) 3.145 (5) 168.9 (3) N(6)-H(2N)•••Cl(1) 2.224 (2) 3.114 (5) 167.1 (3) #1 x+1,y,z #2 x+1/2,-y+ 1/2,-z+2 #3 x+1,y-1,z #4 x-1,y,z #5 x-1/2,-y+ 1/2,-z+2 #6 x-1,y+1,z #7 x,y+1,z LINSAR N(1)H(1B)•••O(3) 2.012 (2) 2.896 (3) 167.2 (1) N(1)H(1A)•••O(3) 2.197 (2) 2.918 (3) 136.8 (1) LICDMG N(1)-H(1)•••Cl(1)#2 2.430 (19) 3.2204 (12) 148.8 (15) O(1)-H(2)•••Cl(1)#3 2.41 (3) 3.1452 (12) 175 (2) O(2)-H(5)•••O(3)#2 1.96 (3) 2.7543 (14) 174 (2) O(2)-H(4)•••Cl(1)#2 2.38 (2) 3.1686 (11) 169 (2) O(1)-H(3)•••Cl(1)#4 2.29 (3) 3.1264 (12) 177 (2) #1 -x,-y,-z+1 #2 x-1,y,z #3 -x+1,-y+1,-z+2 #4 x-1,y-1,z LINDMG N(1)H(7)•••O(5) 1.9912 2.8349 (1) 158.49 LICBTN O(4)-H(3)•••Cl(1)#3 2.47 (5) 3.178 (2) 165 (5) O(3)-H(2)•••Cl(2)#4 2.30 (3) 3.120 (2) 168 (3) O(4)-H(4)•••Cl(2)#5 2.37(5) 3.117 (2) 163 (4) O(3)-H(1)•••Cl(1)#6 2.33 (5) 3.065 (2) 168 (4) #1 -x+1,-y,z+1/2 #2 -x+1,y,z-1/2 #3 x-1/2,-y-1/2,z #4 x+1/2,-y+1/2,z #5 x,y-1,z #6 x,y+1,z LICBALanhydrate N(1)-H(1A)•••Cl(1)#4 2.62 3.2433 (14) 128.3 N(1)-H(1B)•••Cl(1)#5 2.32 3.1983 (14) 168.6 N(1)-H(1C)•••O(2)#4 1.95 2.8251 (17) 167.2 #1 -x,-y+2,-z+1 #2 -x+1,-y+2,z+1 #3 -x+1/2,y-1/2,-z+1/2 #4 x+1/2,-y+3/2,z+1/2 #5 x-1/2,-y+3/2,z+1/2 LICBAL O(1)-H(1)•••Cl(1)#5 2.41 (3) 3.2458 (14) 179 (2) O(1)-H(2)•••Cl(1)#6 2.36 (3) 3.1262 (17) 170 (3) N(1)-H(4)•••Cl(1)#7 2.53 (2) 3.2475 (15) 135.9 (18) N(1)-H(3)•••O(3) 2.44 (4) 2.980 (2) 120 (3) N(1)-H(5)•••Cl(1)#9 2.38 (3) 3.222 (2) 167 (3) #1 -x+5/2,y+1/2,-z+1/2 #2 x,y+1,z #3 x,y-1,z #4 -x+5/2,y-1/2,-z+1/2 #5 x+1,y,z #6 x+1,y-1,z #7 -x+1,-y+1,-z+1 #8 x1,y+1,z #9 -x+1,-y,-z+1 LICABA N(1)-H(1A)•••Cl(1)#5 2.32 3.1516 (13) 155.9 N(1)-H(1B)•••Cl(1)#6 2.33 3.1886 (13) 163.0 N(1)-H(1C)•••O(2)#7 1.92 2.7970 (16) 166.6 #1 x,y+1,z #2 -x+5/2,y+1/2,-z+1/2 #3 x,y-1,z #4 -x+5/2,y-1/2,-z+1/2 #5 -x+2,y+ 1,-z+1/2 #6 x-1/2,-y+1/2,z-1/2 #7 -x+2,-y+2,-z

PAGE 140

121 LICPRO O(3)-H(11)•••Cl(1)#5 2.32 3.234 (3) 161.1 O(3)-H(12)•••Cl(1)#6 2.23 3.194 (3) 171.6 N(1)-H(1)•••Cl(1) 2.42 3.187 (3) 144.6 N(1)-H(2)•••Cl(1)#4 2.69 3.336 (3) 128.5 #1 -x+1,y-1/2,-z+2 #2 x,y-1,z #3 -x+1,y+1/2,-z+2 #4 x,y+1,z #5 -x+1,y+1/2,-z+1 #6 -x+1,y-1/2,-z+1 LIBPRO O(17)-H(17A)•••Br(2) 2.63 (6) 3.345 (3) 167 (6) O(17)H(17B)•••Br(2) 2.40 (5) 3.312 (3) 168 (4) O(18)-H(18B)•••Br(1) 2.64 (6) 3.353 (3) 163 (6) O(18)-H(18A)•••Br(1) 2.42 (4) 3.322 (3) 177 (3) N(16)-H(16B)•••Br(1) 2.76 (3) 3.326 (3) 123 (2) N(16)H(16A)•••Br(1) 2.47 (4) 3.277 (3) 154 (3) N(8)-H(8A)•••Br(2) 2.53 (4) 3.289 (3) 144 (3) LINPRO N(1)-H(1A)•••O(12)#5 1.95 (2) 2.842 (2) 165 (2) N(1)-H(1B)•••O(1)#2 2.02 (2) 2.898 (2) 161 (2) #1 x-1/2,-y+3/2,-z #2 -x+2,y-1/ 2,-z+1/2 #3 -x+2,y+1/2,-z+1/2 #4 x+1/2,-y+3/2,-z #5 -x+5/2,-y+1,z+1/2

PAGE 141

122 Table 4-5. Selected hydrogen bond parameters for 1:2 ionic cocrystals. Compound Hydrogen bond d (HA) / D (DA) / / LICSAR2 N(1)-H(1A)•••Cl(1)#5 2.195 (2) 3.1011 (1) 159.9 (1) N(1)-H(1B)•••Cl(1)#6 2.281 (2) 3.1549 (1) 166.4 (1) N(2)-H(4A)•••O(3)#4 1.85 (2) 2.7244 (2) 164.4 (2) N(2)-H(4B)•••O(2)#4 2.01 (2) 2.8878 (2) 161.6 (2) #1 -x+5/2,y+1/2,z #2 -x+5/ 2,y-1/2,z #3 x+1/2,-y+1/2,-z+1 #4 x-1/2,-y+1/2,-z+1 #5 x+1,y,z #6 x+1/2,y,-z+1/2 LICDMG2 N(1)-H(1N)•••O(2)#1 1.79 2.742 (2) 168.5 #1 x-1/4,-y+1/4,z-1/4 LIBDMG2 N(1)-H(1)•••O(2)#5 1.85 2.747 (2) 161.3 #1 -x+3/4,y+1/4,z-1/4 #2 -x+1/2,y+1/2,z #3 x-1/4,-y+3/4,z+1/4 #4 -x+3/4,y-1/4,z+1/4 #5 x+1/4,-y+3/4,z-1/4 LICPRO2 N(1)-H(1B)•••Cl(1)#1 2.22 3.1322 (12) 172.9 N(1)-H(1A)•••O(2)#2 1.82 2.7404 (15) 175.4 #1 x,y+1,z #2 x-1/2,-y+3/2,-z+3/4 LIBPRO2 N(1)-H(1A)•••Br(1)#1 2.36 3.2769 (15) 174.6 N(1)-H(1B)•••O(1)#2 1.82 2.737 (2) 175.1 #1 x,y-1,z #2 x+ 1/2,-y+1/2,-z-1/4 LINPRO2 ( dia ) N(3)-H(3A)•••O(2)#5 1.93 2.741 (6) 145.8 N(3)-H(3B)•••O(5)#1 1.85 2.762 (5) 170.3 N(4)-H(4A)•••O(4)#3 1.85 2.751 (5) 167.0 N(4)-H(4B)•••O(2)#6 2.26 2.870 (6) 123.2 N(4)-H(4B)•••O(1)#6 2.13 3.046 (6) 171.2 #1 x-1/2,-y+3/2,-z+1 #2 -x+2,y-1/2, -z+1/2 #3 -x+2,y+1/2,-z+1/2 #4 x+1/2,-y+3/2,-z+1 #5 -x+3/2,-y+1,z-1/2 #6 -x+5/2,-y+2,z-1/2 LINPRO2 (ABW) N(3)-H(3C)•••O(4)#1 1.88 2.791 (2) 170.4 N(3)-H(3D)•••O(13)#21.92 2.830 (2) 168.9 N(3)-H(3D)•••O(14)#22.37 3.054 (2) 131.0 N(1)-H(1A)•••O(11)#32.10 2.998 (2) 166.4 N(1)-H(1B)•••O(12)#22.36 3.010 (2) 127.4 N(1)-H(1B)•••O(10)#22.36 3.060 (2) 132.5 N(2)-H(2C)•••O(1) 1.91 2.783 (2) 156.9 N(2)-H(2C)•••O(3) 2.19 2.688 (2) 113.1 N(2)-H(2D)•••O(8)#4 1.81 2.722 (2) 169.2 N(4)-H(21A)•••O(10) 2.03 2.849 (2) 147.0 N(4)-H(21A)•••O(9) 2.31 2.971 (2) 128.8 N(4)-H(21B)•••O(5) 1.87 2.775 (2) 166.3 #1 -x,y+1/2,-z+1/2 #2 -x+1,y+1/2, -z+1/2 #3 -x+1/2,-y+2,z+1/2 #4 x-1/2,-y+3/2,-z+1

PAGE 142

123 Table 4-6. Crystallographic data for 1:1 ionic cocrystals. LICSAR LINSAR LICDMG LINDMG Formula C3H9ClLiNO3 C3H7LiN2O5 C4H13ClLiNO4 C4H9LiN2O5 MW 149.50 158.05 181.54 172.07 Crystal system Orthorhombic Orthorhombic Triclinic Orthorhombic Space group P212121 P212121 P Fdd2 a () 4.9169 (2) 5.0816 (5) 7.2090 (1) 21.7388 (6) b () 5.2917 (2) 10.7057 (9) 7.5745 (2) 29.0098 (8) c () 27.938 (1) 12.5344 (10) 9.0100 (2) 4.9930 (2) (deg) 90 90 110.746 (1) 90 (deg) 90 90 93.170 (1) 90 (deg) 90 90 103.709 (1) 90 V /3 726.91 (5) 681.9 (1) 441.698 (16) 3148.78 (18) Dc/mg m-3 1.366 1.539 1.365 1.452 Z 4 4 2 16 2 range 3.16 to 67.46 5.43 to 66.13 5.31 to 65.96 5.08 to 67.58 Nref./Npara. 1288 / 85 1153 / 101 1466 / 123 1190 / 145 T /K 100 (2) 100 (2) 100 (2) 100 (2) R1 [I>2sigma(I)] 0.0672 0.0384 0.0293 0.0269 w R2 0.1717 0.1079 0.0828 0.0812 GOF 1.188 0.869 1.393 0.744 Abs coef. 4.186 1.259 3.624 1.137 LICBTN LICBALanhydrate LICBAL LICABA Formula C5H15Cl2Li2NO2 C3H7ClLiNO2 C3H9ClLiNO3 C4H9ClLiNO2 MW 237.91 131.47 149.50 145.51 Crystal system Orthorhomic Monoclinic Monoclinic Monoclinic Space group Pna21 P21/n P21/n C2/c a () 11.8788 (4) 5.0510 (2) 6.6716 (9) 20.0021 (4) b () 6.1712 (2) 13.2649 (5) 5.0400 (8) 4.8815 (1) c () 15.3296 (5) 8.9413 (3) 19.816 (3) 17.0417 (6) (deg) 90 90 90 90 (deg) 90 96.721 (2) 95.888 (7) 124.929 (1) (deg) 90 90 90 90 V /3 1123.76 (6) 594.96 (4) 662.78 (16) 1364.21 (6) Dc/mg m-3 1.407 1.468 1.498 1.417 Z 2 4 4 8 2 range 5.77 to 67.63 6.00 to 67.81 4.49 to 67.91 5.39 to 68.66 Nref./Npara. 1790 / 146 1045 / 74 1156 / 102 1214 / 83 T /K 100 (2) 100 (2) 100 (2) 100 (2) R1 [I>2sigma(I)] 0.0333 0.0284 0.0292 0.0299 w R2 0.0757 0.0776 0.0841 0.0780 GOF 1.080 1.361 1.516 1.104 Abs coef. 5.104 4.909 4.591 4.335

PAGE 143

124 Table 4-7. Crystallographic data fo r 1:1 ionic cocrystals (continued). LICPRO LIBPRO LINPRO Formula C5H11ClLiNO3 C5H11BrLiNO3 C5H9LiN2O5 MW 175.54 220.00 184.08 Crystal system Monoclinic Monoclinic Orthorhombic Space group P21 P21 P212121 a () 7.7692 (12) 11.2473 (3) 9.0947 (4) b () 5.1020 (9) 5.1316 (1) 9.2876 (5) c () 10.3795 (16) 14.9547 (4) 9.5743 (5) (deg) 90 90 90 (deg) 105.458 (9) 104.395 (1) 90 (deg) 90 90 90 V /3 396.54 (11) 836.04 (4) 808.72 (7) Dc/mg m-3 1.470 1.748 1.512 Z 2 4 4 2 range 4.42 to 66.91 3.05 to 67.98 6.64 to 67.56 Nref./Npara. 1097 / 100 2675 / 231 1406 / 124 T /K 100 (2) 100 (2) 100 (2) R1 [I>2sigma(I)] 0.0312 0.0200 0.0290 w R2 0.0862 0.0502 0.0739 GOF 1.147 1.102 1.052 Abs coef. 3.928 6.386 1.151

PAGE 144

125 Table 4-8. Crystallographic data of 1:2 ionic cocrystals. LICSAR2 LINBTN2 LICDMG2 LIBDMG2 Formula C6H14ClLiN2O4 C10H22LiN3O7 C8H18ClLiN2O4 C8H18BrLiN2O4 MW 220.58 303.25 248.63 293.09 Crystal system Orthorhombic Monoclinic Orthorhombic Orthorhombic Space group P bca P 21/c F dd2 F dd2 a () 9.5197 (1) 16.0472 (16) 14.0427 (5) 14.0912 (2) b () 9.9275 (1) 8.4767 (10) 14.6533 (5) 14.9035 (2) c () 21.7783 (2) 10.8836 (11) 12.4822 (4) 12.5426 (2) (deg) 90 90 90 90 (deg) 90 103.731 (6) 90 90 (deg) 90 90 90 90 V /3 2058.20 (4) 1438.2 (3) 2568.49 (15) 2634.05 (7) Dc/mg m-3 1.424 1.401 1.286 1.478 Z 8 4 8 8 2 range 4.06 to 67.95 2.83 to 66.56 5.62 to 68.02 5.58 to 66.35 Nref./Npara. 1838 / 145 2484 / 196 1080 / 76 1114 / 77 T /K 100 (2) 100 (2) 100 (2) 100 (2) R1 [I>2sigma(I)] 0.0274 0.0408 0.0312 0.0178 w R2 0.0786 0.1134 0.0726 0.0464 GOF 1.031 1.023 0.994 1.089 Abs coef. 3.248 0.992 2.660 4.282 LICPRO2 LIBPRO2 LINPRO2 ( dia ) LINPRO2 (ABW) Formula C10H18ClLiN2O4 C10H18BrLiN2O4 C10H18LiN3O7 C10H18LiN3O7 MW 272.65 317.11 299.21 299.21 Crystal system Tetragonal Tetragonal Orthorhombic Orthorhombic Space group P 41212 P 41212 P 212121 P 212121 a () 9.0791 (1) 9.1703 (3) 9.5219 (9) 11.0448 (3) b () 9.0791 (1) 9.1703 (3) 9.5664 (9) 12.0393 (3) c () 15.4104 (2) 15.5694 (14) 15.0812 (12) 20.2019 (5) (deg) 90 90 90 90 (deg) 90 90 90 90 (deg) 90 90 90 90 V /3 1270.28 (3) 1309.30 (14) 1373.8 (2) 2686.28 (12) Dc/mg m-3 1.426 1.609 1.447 1.480 Z 4 4 4 8 2 range 5.66 to 67.91 5.60 to 68.18 2.93 to 65.90 4.27 to 65.93 Nref./Npara. 1150 / 84 1182 / 83 2347 / 191 4505 / 379 T /K 100 (2) 100 (2) 100 (2) 100 (2) R1 [I>2sigma(I)] 0.0229 0.0181 0.0576 0.0314 w R2 0.0662 0.0447 0.1454 0.0799 GOF 1.022 1.078 1.046 1.045 Abs coef. 2.745 4.362 1.038 1.061

PAGE 145

126 Chapter 5. Conclusions and Future Directions 5.1 Conclusions The power of the supramolecular heterosynt hon approach to generate cocrystals is once again demonstrated using ellagic acid. 17 crystal forms (5 cocrystals, 7 cocrystal solvates/hydrates and 5 solvates) of ellagic ac id have been isolated and characterized by solid state analytical technique s. The crystal forms of ellagic acid reveal its compatibility with a diverse variety of functional groups in pyridine, imidazole, pyrazole, carboxylate and carbonyl moieties. The discovery of ellagi c acid solvates while serendipitous is not totally unexpected since el lagic acid has an imbala nce of hydrogen bond donors over acceptors and thus demonstrates its affinity for solvent molecules with superior hydrogen bond acceptor capabilities. Solvation/hydration pe rsists after cocrysta l formation if the imbalance remains. This obeys E tter’s first rule of hydrogen bonding196 that “all good proton donors and acceptors are used in hydro gen bonding”, as originally put forth by Donohue197, is satisfied. Hence the synergistic incorporation of the phenol-carbonyl supramolecular heterosynthon resu lted in the cocrystal solvat es of ellagic acid with isonicotinamide and theophylline. This synerg ism put polyphenols such as ellagic acid in a favourable position to be included in a cocr ystal former library for cocrystal screening experiments since APIs routinely contain multiple functional groups. The motivation to diversify the crystal forms of lithium salts stems from the pharmacological properties of li thium that consistently reduce suicidality in recurrent unipolar major depressive disorder and in bipol ar disorder. To engineer ionic cocrystals

PAGE 146

127 of lithium salts, the coordination chemistry of the lithium cation has been immensely helpful. The oxophilic character of lithium poi nts to oxygen based lig ands such as the carboxylate as a good cocrystal former. That th e carboxylate is negati vely-charged means the carboxylate moiety has to be derived from an amino acid so that it is an overall neutral molecule. In addition, that there are ac tive transporters to fa cilitate the movement of amino acids across the blood brain barrier implies that intuitively, amino acids may be the appropriate cocrystal formers to target i onic cocrystals of lithium salts and potentially facilitate the passage of the lithium cations across the blood brain barrier. The choice of the lithium-to-amino acid st oichiometry dictates the periodicity of the crystal structure. Equimolar quantities of lithium salts and amino acids yield 1periodic structures with the exception of LI CDMG (a discrete supermolecule that is hydrogen bonded into a 1-periodic chain) a nd LINPRO (a mixed ligand 3-periodic dia net). The construction of 2-periodic and 3-pe riodic nets is instead observed in 1:2 cocrystals of lithium salts and amino acids to generate cationic square grids, dia nets and a zeolitic ABW net. These architectures are ch aracterized by the formation of 4-, 6and 8-membered Lin(carboxylate)n (n=4, 6, 8) rings with a wi de range of Li-carboxylate-Li angles. Dia nets are prone to interpenetration but th e presence of pairs of counterions and the bulkiness of the amino acid substituents precludes interpenetration in the LiDMOMs. The zeolitic ABW net is the fi rst example of a LiZMOM and features the combination of 4-, 6and 8-membered Lin(carboxylate)n (n=4, 6, 8) rings. The modular nature of these compounds coupled with larger anions will en visage a range of LiZMOMs that parallels the structural diversity seen in inorganic zeolites and zeolitic meta l-organic materials.

PAGE 147

128 5.2 Future Directons While the phenol-pyridine, phenol-imidazole, phenol-pyrazole and phenolcarboxylate supramolecular heterosynthons have been the target of crystal engineering studies, the phenol-carbonyl supramolecular he terosynthon is still relatively unexplored. While -caprolactam (supramolecularly similar to a carboxylic acid) and urea are not the best models for a carbonyl group such as a ketone, the phenol-carbonyl supramolecular heterosynthon is observed nonetheless. Figure 5-1. Triclinic modification of quinhydrone. That the phenol-carbonyl supramolecu lar heterosynthon is not exploited in cocrystal design is perhaps surprising since it is featured in the prototypal cocrystal in quinhydrone198 (1:1 cocrystal of quinone an d hydroquninone) by Whler in 1844. Further, benzophenone was used to prep are a cocrystal with diphenylamine.199 These cocrystals were discovered without the benefit of single cr ystal X-ray crystallography.

PAGE 148

129 The power of X-ray crystallography200-201 undoubtedly elucidates the carbonyl group to be the hydrogen bond acceptor for these cocrystals. Perhaps ellagic acid can form a cocrystal with benzophenone. In line with using pharmaceutically acceptable cocrystal formers, flavone202 may be an excellent choice. Figure 5-2. Carbonyl as the hydrogen bond acceptor in the cocrystal of diphenylamine and benzophenone. All 1:1 cocrystals of lithium nitrate and amino acids are capable of generating mixed ligand dia nets or square grids but only LI NPRO is observed as a mixed ligand dia net. Nature tunes the polymorphic selectivity of silica by varying temperature. Perhaps a similar approach can be used to hunt for other mixed ligand dia nets or square grids. While these may be of pure academic interest, th e ability to control th e crystal structure is an example of crystal engineering at this finest and is potentially useful to the control of

PAGE 149

130 polymorphic selectivity. However, that the mapping of the phase diagram does not guarantee the preparation and isolat ion of a particular crystal form203 must not discourage us in our endeavours but serve as the motivational tool to refi ne our skills and better our knowledge at crystallization. 5.3 Crystal Engineering and Crystallization “… crystal engineering, which is define d as the understandin g of intermolecular interactions in the context of crystal packing and in the u tilization of su ch understanding in the design of new solids with desi red physical and chem ical properties…”204 Gautam R. Desiraju “…, the last step (precipitating good si ngle crystals) remains a black art.”205 Philip Ball “… there is still a great deal of black art (art, not magic!) and I might add skill, experience and chemical in tuition in making them”206 Joel Bernstein I close with some noteworthy quotes I have not deliberated on crystal engineering207 thus far, even though crystal engineer ing is reflected in the title of this dissertation. Using Desiraju’s definition of crystal engineering, the understanding of intermolecular (non-covalent and ionic) interactions ha ve been documented and are exploited to design the molecular and ionic cocr ystals presented in this dissertation. I can identify with the comments by Bernstein (in response to Ba ll) that undoubt edly mirrors one of the toughest challenges I had to overcom e. The variety of organic, inorganic and hybrid solids is ever increasing. Crystal engineering is a science but crystallization

PAGE 150

131 remains an art, but no longer a dark ar t. The concerted opera tion of the art of crystallization and the craft of crystal engineering is imminent to achieve the final prize, the desired crystalline material.

PAGE 151

132 References 1. Moulton, B.; Zaworotko, M. J., From molecules to crystal engineering: Supramolecular isomerism and polymorphism in network solids. Chemical Reviews 2001, 101 (6), 1629-1658. 2. Cooper, S. K., The Periodic Table: Mapping the Elements Compass Point Books: 2007. 3. Simpson, J., Oxford English dictionary Oxford University Press: 1999. 4. Walker, G. W., "The application of the kinetic theory of gases to the electric, magnetic, and optical properties of diatomic gases.". Proceedings of the Royal Society of London Series B-Containing Papers of a Biological Character 1902, 69 (456), 394-398. 5. Wibaut, J. P.; Hoog, H.; Langedijk, S. L.; Overhoff, J.; Smittenberg, J., A study on the preparation and the physical constants of a number of alkanes and cycloalkanes. Recueil Des Travaux Chimiques Des Pays-Bas 1939, 58 329-377. 6. Doolittle, A. K.; Peterson, R. H., Preparat ion And Physical Properties Of A Series Of Normal-Alkanes. Journal of the American Chemical Society 1951, 73 (5), 2145-2151. 7. Dunitz, J. D., Phase-Tr ansitions In Molecular Crystals From A Chemical Viewpoint. Pure and Applied Chemistry 1991, 63 (2), 177-185. 8. Desiraju, G. R., Crystal engineer ing. From molecules to materials. Journal of Molecular Structure 2003, 656 (1-3), 5-15. 9. Haleblian, J.; McCrone, W., Pharmaceuti cal Applications Of Polymorphism. Journal of Pharmaceutical Sciences 1969, 58 (8), 911-&. 10. Donnet, J. B., Carbons allotrops: a large family. Actual Chim. 2006 115-118.

PAGE 152

133 11. Kroto, H. W.; Heath, J. R.; O'Brien, S. C.; Curl, R. F.; Smalley, R. E., C60: Buckminsterfullerene. Nature 1985, 318 (6042), 162-163. 12. Lindsay, R. B., The carbon atom mo del and the structure of diamond. Phys. Rev. 1927, 29 (4), 0497-0509. 13. Dienwiebel, M.; Verhoeven, G. S.; Prad eep, N.; Frenken, J. W. M.; Heimberg, J. A.; Zandbergen, H. W., Superlubricity of Graphite. Physical Review Letters 2004, 92 (12), 126101. 14. Whler; Liebig, Untersuchungen ber das Radikal der Benzoesure. Annalen der Pharmacie 1832, 3 (3), 249-282. 15. Penfold, B. R.; White, J. C. B., The crys tal and molecular structure of benzamide. Acta Crystallographica 1959, 12 (2), 130-135. 16. David, W. I. F.; Shankland, K.; Pulham, C. R.; Blagden, N.; Davey, R. J.; Song, M., Polymorphism in Benzamide. Angewandte Chemie International Edition 2005, 44 (43), 7032-7035. 17. Blagden, N.; Davey, R.; Dent, G.; Song, M.; David, W. I. F.; Pulham, C. R.; Shankland, K., Woehler and Liebig Revisited: A Small Molecule Reveals Its Secrets The Crystal Structure of the Unstable Poly morph of Benzamide Solved after 173 Years. Crystal Growth & Design 2005, 5 (6), 2218-2224. 18. Thun, J.; Seyfarth, L.; Butterhof, C.; Senker, J.; Dinnebier, R. E.; Breu, J., Whler and Liebig Revisited: 176 Years of Polymo rphism in Benzamide and the Story Still Continues!†. Crystal Growth & Design 2009, 9 (5), 2435-2441.

PAGE 153

134 19. Amidon, G. L.; Lennerns, H.; Shah, V. P. ; Crison, J. R., A Theoretical Basis for a Biopharmaceutic Drug Classification: Th e Correlation of in Vitro Drug Product Dissolution and in Vivo Bioavailability. Pharmaceutical Research 1995, 12 (3), 413-420. 20. Chemburkar, S. R.; Bauer, J.; Deming, K.; Spiwek, H.; Patel, K.; Morris, J.; Henry, R.; Spanton, S.; Dziki, W.; Porter, W.; Quick, J.; Bauer, P.; Donaubauer, J.; Narayanan, B. A.; Soldani, M.; Riley, D.; McFarland, K., Dealing with the Impact of Ritonavir Polymorphs on the Late Stages of Bulk Drug Process Development. Organic Process Research & Development 2000, 4 (5), 413-417. 21. Bauer, J.; Spanton, S.; Henry, R.; Quick, J.; Dziki, W.; Porter, W.; Morris, J., Ritonavir: An Extraordinary Exampl e of Conformational Polymorphism. Pharmaceutical Research 2001, 18 (6), 859-866. 22. Morissette, S. L.; Souk asene, S.; Levinson, D.; Ci ma, M. J.; Almarsson, ., Elucidation of crystal form diversity of the HIV protease inhi bitor ritonavir by highthroughput crystallization. Proceedings of the National Academy of Sciences of the United States of America 2003, 100 (5), 2180-2184. 23. Yu, L., Polymorphism in Molecular Solid s: An Extraordinary System of Red, Orange, and Yellow Crystals. Accounts of Chemical Research 2010, 43 (9), 1257-1266. 24. Bernstein, J.; Davey, R. J.; Henck, J. O., Concomitant polymorphs. Angewandte Chemie-International Edition 1999, 38 (23), 3440-3461. 25. Yu, L., Inferring thermodynamic stabil ity relationship of polymorphs from melting data. Journal of Pharmaceutical Sciences 1995, 84 (8), 966-974.

PAGE 154

135 26. Yu, L.; Huang, J.; Jones, K. J., M easuring Free-Energy Difference between Crystal Polymorphs through Eutectic Melting. The Journal of Physical Chemistry B 2005, 109 (42), 19915-19922. 27. Burger, A.; Ramberger, R., Polymo rphism Of Pharmaceuticals And Other Molecular-Crystals .1. Theory Of Thermodynamic Rules. Mikrochimica Acta 1979, 2 (34), 259-271. 28. Burger, A.; Ramberger, R., Polymo rphism Of Pharmaceuticals And Other Molecular-Crystals .2. Applicabil ity Of Thermodynamic Rules. Mikrochimica Acta 1979, 2 (3-4), 273-316. 29. Sun, E. In Excerpts from news conference, 15 October 1998 International Association of Physicians in AIDS Ca re, 225 West Washington Street, Suite 2200, Chicago, IL 60606, 225 West Washington Street, Suite 2200, Chicago, IL 60606, 1998. 30. Aaltonen, J.; Alleso, M.; Mirza, S.; Koradia, V.; Gordon, K. C.; Rantanen, J., Solid form screening A review. European Journal of Pharmaceutics and Biopharmaceutics 2009, 71 (1), 23-37. 31. Garad, S. D., How to Improve the Bioa vailability of Poorly Soluble Drugs. American Pharmaceutical Review 2004 80-85. 32. Cooke, M. W.; Stanton, M.; Shimanovich, R.; Bak, A., Co-crystals to Enhance the Exposure of Insoluble Molecules. American Pharmaceutical Review 2007 54-59. 33. Lipinski, C. A., Poor Aqueous Solubili ty an Industry Wide Problem in Drug Discovery. American Pharmaceutical Review 2002 82-85.

PAGE 155

136 34. Lipinski, C. A., Drug-like properties and the causes of poor solubility and poor permeability. Journal of Pharmacological and Toxicological Methods 2000, 44 (1), 235249. 35. Meanwell, N. A., Chapter 22 The Emer ging Utility of Co-Crystals in Drug Discovery and Development. In Annual Reports in Medicinal Chemistry John, E. M., Ed. Academic Press: 2008; Vol. Volume 43, pp 373-404. 36. Childs, S. L.; Chyall, L. J.; Dunlap, J. T.; Smolenskaya, V. N.; Stahly, B. C.; Stahly, G. P., Crystal Engineering Appr oach To Forming Cocrystals of Amine Hydrochlorides with Organic Acids. Mol ecular Complexes of Fl uoxetine Hydrochloride with Benzoic, Succinic, and Fumaric Acids. Journal of the American Chemical Society 2004, 126 (41), 13335-13342. 37. Takagi, T.; Ramachandran, C.; Bermejo, M.; Yamashita, S.; Yu, L. X.; Amidon, G. L., A Provisional Biopharmaceutical Classi fication of the Top 200 Oral Drug Products in the United States, Great Britain, Spain, and Japan. Molecular Pharmaceutics 2006, 3 (6), 631-643. 38. Shan, N.; Zaworotko, M. J., Polymorphic crystal forms and cocrystals in drug delivery (crystal engineering). In Burger's Medicinal Chemistry, Drug Discovery, and Development, Seventh Edition Abraham, D. J.; Rotella, D. P., Eds. John Wiley & Sons, Inc.: 2010; pp 187-218. 39. Childs, S. L.; Hardcastle, K. I., Cocr ystals of chlorzox azone with carboxylic acids. CrystEngComm 2007, 9 (5), 364-367. 40. Sun, C. C.; Hou, H., Improving Mechanical Properties of Caffeine and Methyl Gallate Crystals by Cocrystallization. Crystal Growth & Design 2008, 8 (5), 1575-1579.

PAGE 156

137 41. Trask, A. V.; Motherwell, W. D. S.; Jone s, W., Pharmaceutical Cocrystallization: Engineering a Remedy fo r Caffeine Hydration. Crystal Growth & Design 2005, 5 (3), 1013-1021. 42. Noyes, A. A.; Whitney, W. R., The Rate Of Solution Of Solid Substances In Their Own Solutions. Journal of the Americ an Chemical Society 1897, 19 (12), 930-934. 43. Dokoumetzidis, A.; Macheras, P., A centu ry of dissolution re search: From Noyes and Whitney to the Biopharmaceutics Classification System. International Journal of Pharmaceutics 2006, 321 (1-2), 1-11. 44. Remenar, J. F.; Peterson, M. L.; Stephens, P. W.; Zhang, Z.; Zimenkov, Y.; Hickey, M. B., Celecoxib: Nicotinamide Dissociation: Using Excipients To Capture the Cocrystal's Potential. Molecular Pharmaceutics 2007, 4 (3), 386-400. 45. Schultheiss, N.; Newman, A., Pharmaceutical Cocrystals and Their Physicochemical Properties. Crystal Growth & Design 2009, 9 (6), 2950-2967. 46. Good, D. J.; Rodrguez-Hornedo, N. r., So lubility Advantage of Pharmaceutical Cocrystals. Crystal Growth & Design 2009, 9 (5), 2252-2264. 47. Walsh, R. D. B.; Bradner, M. W.; Fleischman, S.; Morales, L. A.; Moulton, B.; Rodriguez-Hornedo, N.; Zaworotko, M. J., Cr ystal engineering of the composition of pharmaceutical phases. Chemical Communications 2003, (2), 186-187. 48. Desiraju, G. R., Supramolecular S ynthons in Crystal Engineering—A New Organic Synthesis. Angewandte Chemie Interna tional Edition in English 1995, 34 (21), 2311-2327. 49. Remenar, J. F.; Morissette, S. L.; Peterson, M. L.; Moulton, B.; MacPhee, J. M.; Guzman, H. R.; Almarsson, ., Crystal Engine ering of Novel Cocrystals of a Triazole

PAGE 157

138 Drug with 1,4-Dicarboxylic Acids. Journal of the American Chemical Society 2003, 125 (28), 8456-8457. 50. Li, Z. J.; Abramov, Y.; Bordner, J.; Leonard, J.; Medek, A.; Trask, A. V., SolidState Acid Base Interactions in Complexes of Heterocyclic Bases with Dicarboxylic Acids: Crystallography, Hydrogen Bond Analysis, and 15N NMR Spectroscopy. Journal of the American Chemical Society 2006, 128 (25), 8199-8210. 51. Basavoju, S.; Bostrm, D.; Velaga, S. P ., Pharmaceutical Cocr ystal and Salts of Norfloxacin. Crystal Growth & Design 2006, 6 (12), 2699-2708. 52. Bak, A.; Gore, A.; Yanez, E.; Stanton, M.; Tufekcic, S.; Syed, R.; Akrami, A.; Rose, M.; Surapaneni, S.; Bostick, T.; King, A. ; Neervannan, S.; Osto vic, D.; Koparkar, A., The co-crystal approach to improve th e exposure of a water-insoluble compound: AMG 517 sorbic acid co-crystal ch aracterization and pharmacokinetics. Journal of Pharmaceutical Sciences 2008, 97 (9), 3942-3956. 53. Stanton, M. K.; Bak, A., Physicochemi cal Properties of Pharmaceutical CoCrystals: A Case Study of Ten AMG 517 Co-Crystals. Crystal Growth & Design 2008, 8 (10), 3856-3862. 54. Stanton, M. K.; Tufekcic, S.; Morgan, C.; Bak, A., Drug Substance and Former Structure Property Relationships in 15 Diverse Pharmaceutical Co-Crystals. Crystal Growth & Design 2009, 9 (3), 1344-1352. 55. Aitipamula, S.; Chow, P. S.; Tan, R. B. H., Trimorphs of a pharmaceutical cocrystal involving two active pharmaceutical ingredients: potential relevance to combination drugs. CrystEngComm 2009, 11 (9), 1823-1827.

PAGE 158

139 56. Aakery, C. B.; Forbes, S.; Desper, J., Using Cocrystals To Systematically Modulate Aqueous Solubility and Melt ing Behavior of an Anticancer Drug. Journal of the American Chemical Society 2009, 131 (47), 17048-17049. 57. Cheney, M. L.; Shan, N.; Healey, E. R.; Hanna, M.; Wojtas, L.; Zaworotko, M. J.; Sava, V.; Song, S.; Sanchez-Ramos, J. R., Effects of Crystal Form on Solubility and Pharmacokinetics: A Crystal Engine ering Case Study of Lamotrigine. Crystal Growth & Design 2009, 10 (1), 394-405. 58. Lee, T.; Wang, P. Y., Screening, Manufacturing, Photoluminescence, and Molecular Recognition of Co-Crystals: Cytosine with Dicarboxylic Acids. Crystal Growth & Design 2010, 10 (3), 1419-1434. 59. Fremont, L., Minireview Biological effects of resveratrol. Life Sciences 2000, 66 (8), 663-673. 60. Schultheiss, N.; Bethune, S.; Henck, J. -O., Nutraceutical cocrystals: utilizing pterostilbene as a cocrystal former. CrystEngComm 2010, 12 (8), 2436-2442. 61. Trask, A. V., An Overview of Pharmaceutical Cocrystals as Intellectual Property†. Molecular Pharmaceutics 2007, 4 (3), 301-309. 62. Shattock, T. R.; Arora, K. K.; Vishwesh war, P.; Zaworotko, M. J., Hierarchy of Supramolecular Synthons: Persistent Car boxylic AcidPyridin e Hydrogen Bonds in Cocrystals That also Contain a Hydroxyl Moiety. Crystal Growth & Design 2008, 8 (12), 4533-4545. 63. Bis, J. A.; Vishweshwar, P.; Weyna D.; Zaworotko, M. J., Hierarchy of Supramolecular Synthons: Persistent HydroxylPyridine Hydrogen Bonds in Cocrystals That Contain a Cyano Acceptor. Molecular Pharmaceutics 2007, 4 (3), 401-416.

PAGE 159

140 64. Kavuru, P.; Aboarayes, D.; Arora, K. K.; Clarke, H. D.; Kennedy, A.; Marshall, L.; Ong, T. T.; Perman, J.; Pujari, T.; Wojtas, .; Zaworotko, M. J., Hierarchy of Supramolecular Synthons: Persistent H ydrogen Bonds Between Carboxylates and Weakly Acidic Hydroxyl Moieties in Cocrystals of Zwitterions. Crystal Growth & Design 2010, 10 (8), 3568-3584. 65. Datta, S.; Grant, D. J. W., Crystal struct ures of drugs: advances in determination, prediction and engineering. Nat Rev Drug Discov 2004, 3 (1), 42-57. 66. Gardner, C. R.; Walsh, C. T.; Almarsson, ., Drugs as materials: valuing physical form in drug discovery. Nat Rev Drug Discov 2004, 3 (11), 926-934. 67. Maddox, J., Crystals from first principles. Nature 1988, 335 (6187), 201-201. 68. Newman, A. W.; Childs, S. L.; Cowans, B. A., Pharmaceutical Sciences Encyclopedia: Drug Discovery, Development, and Manufacturing John Wiley & Sons, Inc.: 2007; p 455-481. 69. Childs, S. L.; Zaworotko, M. J., The Reemergence of Cocrystals: The Crystal Clear Writing Is on the Wall Introduction to Virtual Special Issue on Pharmaceutical Cocrystals. Crystal Growth & Design 2009, 9 (10), 4208-4211. 70. Childs, S. L.; Rodriguez-Hornedo, N.; Reddy, L. S.; Jayasankar, A.; Maheshwari, C.; McCausland, L.; Shipplett, R.; Stahly, B. C., Screening strategies based on solubility and solution composition generate pharmaceutically acceptable cocrystals of carbamazepine. CrystEngComm 2008, 10 (7), 856-864. 71. Childs, S. L.; Hardcastle, K. I., Cocrysta ls of Piroxicam with Carboxylic Acids. Crystal Growth & Design 2007, 7 (7), 1291-1304.

PAGE 160

141 72. Cheney, M. L.; Weyna, D. R.; Shan, N.; Hanna, M.; Wojtas, L.; Zaworotko, M. J., Supramolecular Architectures of Meloxicam Carboxylic Ac id Cocrystals, a Crystal Engineering Case Study. Crystal Growth & Design 2010, 10 (10), 4401-4413. 73. McNamara, D.; Childs, S.; Giordano, J.; Iarriccio, A.; Cassidy, J.; Shet, M.; Mannion, R.; O'Donnell, E.; Park, A., Use of a Glutaric Acid Cocrystal to Improve Oral Bioavailability of a Low Solubility API. Pharmaceutical Research 2006, 23 (8), 18881897. 74. Kastelic, J. e.; Hodnik, Z. I.; Šket, P. ; Plavec, J.; Lah, N.; Leban, I.; Pajk, M.; Planinšek, O.; Kikelj, D., Fluconazole Cocrystals with Dicarboxylic Acids. Crystal Growth & Design 2010, 10 (11), 4943-4953. 75. Schultheiss, N.; Roe, M.; Boerrigter, S. X. M., Cocrystals of nutraceutical pcoumaric acid with caffeine and theophylline : polymorphism and so lid-state stability explored in detail usi ng their crystal graphs. CrystEngComm 2011, 13 (2), 611-619. 76. Reddy, L. S.; Bethune, S. J.; Kampf, J. W.; Rodrguez-Hornedo, N., Cocrystals and Salts of Gabapentin: pH Dependent Cocrystal Stability and Solubility. Crystal Growth & Design 2008, 9 (1), 378-385. 77. Schultheiss, N.; Lorimer, K.; Wolfe, S. ; Desper, J., Attempted construction of minoxidil: carboxylic acid cocrystals; 7 salts and 1 cocrystal resulted. CrystEngComm 2010, 12 (3), 742-749. 78. Aitipamula, S.; Chow, P. S.; Tan, R. B. H., Polymorphs and Solvates of a Cocrystal Involving an Analgesic Drug, Et henzamide, and 3,5-Dinitrobenzoic Acid. Crystal Growth & Design 2010, 10 (5), 2229-2238.

PAGE 161

142 79. Aitipamula, S.; Chow, P. S.; Tan, R. B. H., Conformational and enantiotropic polymorphism of a 1: 1 cocrystal invo lving ethenzamide and ethylmalonic acid. CrystEngComm 2010, 12 (11), 3691-3697. 80. Database of Select Committee on GRAS Substances (SCOGS) Reviews. http://www.accessdata.fda.gov/scripts/fcn/ fcnNavigation.cfm?rpt=scogsListing (accessed 15 March 2011). 81. Manach, C.; Scalbert, A.; Morand, C.; Remesy, C.; Jimenez, L., Polyphenols: food sources and bioavailability. American Journal of Clinical Nutrition 2004, 79 (5), 727-747. 82. Yang, C. S.; Wang, Z. Y., Tea And Cancer. Journal of the National Cancer Institute 1993, 85 (13), 1038-1049. 83. Soleas, G. J.; Diamandis, E. P.; Goldbe rg, D. M., Wine as a biological fluid: History, production, and role in disease prevention. Journal of Clinical Laboratory Analysis 1997, 11 (5), 287-313. 84. Reed, J. D., Nutritional Toxicology Of Tannins And Related Polyphenols In Forage Legumes. Journal of Animal Science 1995, 73 (5), 1516-1528. 85. Aggarwal, B. B.; Shishodia, S., Molecula r targets of dietary agents for prevention and therapy of cancer. Biochemical Pharmacology 2006, 71 (10), 1397-1421. 86. Lockwood, B., Nutraceuticals Pharmaceutical Press: London, UK, 2007. 87. Espin, J. C.; Garcia-Conesa, M. T.; Toma s-Barberan, F. A., Nutraceuticals: Facts and fiction. Phytochemistry 2007, 68 (22-24), 2986-3008. 88. Clifford, M. N.; Scalbert, A., Ellag itannins nature, oc currence and dietary burden. Journal of the Science of Food and Agriculture 2000, 80 (7), 1118-1125.

PAGE 162

143 89. Gil, M. I.; Tomas-Barberan, F. A.; Hess-Pi erce, B.; Holcroft, D. M.; Kader, A. A., Antioxidant activity of pomegranate juice a nd its relationship with phenolic composition and processing. Journal of Agricultural and Food Chemistry 2000, 48 (10), 4581-4589. 90. Wood, A. W.; Huang, M. T.; Chang, R. L. ; Newmark, H. L.; Lehr, R. E.; Yagi, H.; Sayer, J. M.; Jerina, D. M.; Conney, A. H., Inhibition Of The Mutagenicity Of BayRegion Diol Epoxides Of Polycyclic Arom atic-Hydrocarbons By Naturally-Occurring Plant Phenols Exceptional Activity Of Ellagic Acid. Proceedings of the National Academy of Sciences of the United St ates of America-Biological Sciences 1982, 79 (18), 5513-5517. 91. Queimada, A. n. J.; Mota, F. t. L.; Pinho, S. o. P.; Macedo, E. n. A., Solubilities of Biologically Active Phenolic Comp ounds: Measurements and Modeling. The Journal of Physical Chemistry B 2009, 113 (11), 3469-3476. 92. Stoner, G. D.; Sardo, C.; Apseloff, G.; Mullet, D.; Wargo, W.; Pound, V.; Singh, A.; Sanders, J.; Aziz, R.; Casto, B.; Sun, X. L., Pharmacokinetics of anthocyanins and ellagic acid in healthy volunteers fed freeze-d ried black raspberries daily for 7 days. Journal of Clinical Pharmacology 2005, 45 (10), 1153-1164. 93. Cerda, B.; Tomas-Barberan, F. A.; Espi n, J. C., Metabolism of antioxidant and chemopreventive ellagitannins from strawber ries, raspberries, walnuts, and oak-aged wine in humans: Identification of bi omarkers and individual variability. Journal of Agricultural and Food Chemistry 2005, 53 (2), 227-235. 94. Seeram, N. P.; Henning, S. M.; Zhang, Y. J.; Suchard, M.; Li, Z. P.; Heber, D., Pomegranate juice ellagitannin metabolites are present in human plasma and some persist in urine for up to 48 hours. Journal of Nutrition 2006, 136 (10), 2481-2485.

PAGE 163

144 95. Seeram, N. P.; Zhang, Y.; McKeever, R.; Henning, S. M.; Lee, R. P.; Suchard, M. A.; Li, Z. P.; Chen, S.; Thames, G.; Zerlin, A.; Nguyen, M.; Wang, D.; Dreher, M.; Heber, D., Pomegranate juice and extracts pr ovide similar levels of plasma and urinary ellagitannin metabolites in human subjects. Journal of Medicinal Food 2008, 11 (2), 390394. 96. Gonzalez-Barrio, R.; Borges, G.; Mullen, W.; Crozier, A., Bioavailability of Anthocyanins and Ellagitannins Following Consumption of Raspberries by Healthy Humans and Subjects with an Ileostomy. Journal of Agricultural and Food Chemistry 2010, 58 (7), 3933-3939. 97. Seeram, N. P.; Lee, R.; Heber, D., Bioava ilability of ellagic acid in human plasma after consumption of ellagitannins from pomegranate (Punica granatum L.) juice. Clinica Chimica Acta 2004, 348 (1-2), 63-68. 98. Mertens-Talcott, S. U.; Jilma-Stohlawet z, P.; Rios, J.; Hingorani, L.; Derendorf, H., Absorption, metabolism, and antioxidant effects of pomegrana te (Punica granatum L.) polyphenols after ingestion of a standardized extract in healthy human volunteers. Journal of Agricultural and Food Chemistry 2006, 54 (23), 8956-8961. 99. Allen, F., The Cambridge Structural Da tabase: a quarter of a million crystal structures and rising. Acta Crystallographica Section B 2002, 58 (3 Part 1), 380-388. 100. Clarke, H. D.; Arora, K. K.; Bass, H.; Kavuru, P.; Ong, T. T.; Pujari, T.; Wojtas, L.; Zaworotko, M. J., Structure Stability Relationships in Cocrystal Hydrates: Does the Promiscuity of Water Make Crystalline Hydrates the Nemesis of Crystal Engineering? Crystal Growth & Design 2010, 10 (5), 2152-2167.

PAGE 164

145 101. McL Mathieson, A.; Poppleton, B. J., Th e crystal structure of ellagic acid. Acta Crystallographica Section B 1968, 24 (11), 1456-1461. 102. Rossi, M.; Erlebacher, J.; Zacharias, D. E.; Carrell, H. L.; Iannucci, B., The Crystal And Molecular Structure Of Ella gic Acid Dihydrate A Dietary Anticancer Agent. Carcinogenesis 1991, 12 (12), 2227-2232. 103. Hasegawa, M.; Terauchil, M.; Kikuchi, Y.; Nakao, A.; Okubo, J.; Yoshinaga, T.; Hiratsuka, H.; Kobayashi, M.; Hoshi, T., De protonation processes of ellagic acid in solution and solid states. Monatshefte Fur Chemie 2003, 134 (6), 811-821. 104. Hunter, C. A.; Sanders, J. K. M., The Nature Of pi-pi Interactions. Journal of the American Chemical Society 1990, 112 (14), 5525-5534. 105. Wolf, K. L.; Wolff, R., bermolekeln. Angewandte Chemie 1949, 61 (5), 191201. 106. MacGillivray, L. R.; Reid, J. L.; Ripmees ter, J. A., Supram olecular Control of Reactivity in the Solid State Using Linear Molecular Templates. Journal of the American Chemical Society 2000, 122 (32), 7817-7818. 107. Vishweshwar, P.; Nangia, A.; Lynch, V. M., Supramolecular synthons in phenolisonicotinamide adducts. CrystEngComm 2003, 5 (31), 164-168. 108. Trask, A. V.; Motherwell, W. D. S.; J ones, W., Pharmaceutical Cocrystallization: Engineering a Remedy fo r Caffeine Hydration. Crystal Growth & Design 2005, 5 (3), 1013-1021. 109. Trask, A. V.; Motherwell, W. D. S.; Jones, W., Physical stability enhancement of theophylline via cocrystallization. International Journal of Pharmaceutics 2006, 320 (12), 114-123.

PAGE 165

146 110. Friš i T.; Trask, A. V.; Motherwell, W. D. S.; Jones, W., Guest-Directed Assembly of Caffeine and Succinic Acid in to Topologically Different Heteromolecular Host Networks upon Grinding. Crystal Growth & Design 2008, 8 (5), 1605-1609. 111. Baxter, N. J.; Williamson, M. P.; Lilley, T. H.; Haslam, E., Stacking interactions between caffeine and methyl gallate. Journal of the Chemical Society, Faraday Transactions 1996, 92 (2), 231-234. 112. Chattoraj, S.; Shi, L.; Sun, C. C., Unde rstanding the relations hip between crystal structure, plasticity and compaction behaviour of theophylline, methyl gallate, and their 1:1 co-crystal. CrystEngComm 2010, 12 (8), 2466-2472. 113. Boldog, I.; Rusanov, E. B.; Sieler, J. ; Domasevitch, K. V., Cooperative association of pyrazoles and pheno ls: A versatile binary system. New Journal of Chemistry 2004, 28 (6), 756-759. 114. Creighton, T. H.; Freeman, W. H., Proteins: Structure and Molecular Properties Second Edition Freeman, W. H. & Company: New York, 1993. 115. Fennema, O. R., Food Chemistry Third Edition CRC Press: Boca Raton, FL, 1996; p 327-328. 116. Hu, Z.-Q.; Xu, D.-J.; Xu, Y.-Z., (S)-Alanine-(S)-mandelic acid (1/1). Acta Crystallographica Section E 2004, 60 (2), o269-o271. 117. Fujii, I.; Baba, H.; Takahashi, Y., Crys tal Structures of L -(R)-Cysteine-Mandelic Acid Diastereomers. Analytical Sciences X-ray Structure Analysis Online 2005, 21 175176.

PAGE 166

147 118. Samas, B.; Wang, W.; Godrej, D. B., 1:1 Cocrystal of (S)-3-(ammoniomethyl)-5methylhexanoate and (S)-mandelic acid. Acta Crystallographica Section E 2007, 63 (10), o3938. 119. Wenger, M.; Bernstein, J., Designing a Cocrystal of -Amino Butyric Acid. Angewandte Chemie International Edition 2006, 45 (47), 7966-7969. 120. Krishnakumar, R. V.; Natarajan, S., Crysta l structure of an adduct of sarcosine with sucrose. Carbohydrate Research 1996, 287 (1), 117-122. 121. Alagar, M.; Krishnakumar, R. V.; Raja gopal, K.; Subha Nandhini, M.; Natarajan, S., L-Phenylalanine fumaric acid. Acta Crystallographica Section E 2003, 59 (7), o952o954. 122. Herbstein, F. H.; Kapon, M.; Maor, I.; Reis ner, G. M., The structure of trimesic acid, its hydrates and complexes. VI. Glycine-trimesic acid monohydrate, H3N+CH2COO-.C9H6O6.H2O. Acta Crystallographica Section B 1981, 37 (1), 136-140. 123. Klussmann, M.; Izumi, T.; White, A. J. P.; Armstrong, A.; Blackmond, D. G., Emergence of Solution-Phase Homochirality via Crystal Engineeri ng of Amino Acids. Journal of the American Chemical Society 2007, 129 (24), 7657-7660. 124. Dega-Szafran, Z.; Dutkiewicz, G.; Ko sturkiewicz, Z.; Petryna, M., Two nonequivalent complexes of 1-piperidineacetic acid with 2,4-dinitrophenol studied by X-ray diffraction, PM3 and SAM1 methods and FTIR. Journal of Molecular Structure 2004, 697 (1-3), 151-159. 125. Timofeeva, T. V.; Kuhn, G. H.; Nester ov, V. V.; Nesterov, V. N.; Frazier, D. O.; Penn, B. G.; Antipin, M. Y., Cocrystal of 1,1-Dicyano-2-(4-hydr oxyphenyl)-ethene with

PAGE 167

148 L-Proline and Induced Conformational Polymo rphism of 1,1-Dicyano-2-(4-hydroxy-3methoxyphenyl)-ethene. Crystal Growth & Design 2003, 3 (3), 383-391. 126. Sander, J. R. G.; Bu ar, D.-K.; Henry, R. F.; Baltrusaitis, J.; Zhang, G. G. Z.; MacGillivray, L. R., A re d zwitterionic co-crystal of acetaminophen and 2,4pyridinedicarboxylic acid. Journal of Pharmaceutical Sciences 2010, 99 (9), 3676-3683. 127. Aakery, C. B.; Bahra, G. S.; Brown, C. R.; Hitchcock, P. B.; Patel, Y.; Seddon, K. R., L-Proline 2,5-Dihydroxybenzoic Acid (1/1): A Zwitterion Co-crystal. Acta Chemica Scandinavica 1995, 49 762-767. 128. Harris, K. D. M.; Thomas, J. M., Stru ctural aspects of ur ea inclusion compounds and their investigation by X-ray di ffraction: a general discussion. Journal of the Chemical Society, Faraday Transactions 1990, 86 (17), 2985-2996. 129. Lee, S.-O.; Kariuki, B. M.; Harris, K. D. M., Hydrogen-bonded chains of diaminoalkane and -dihydroxyalkane gu est molecules lead to disrupted tunnel structures in urea inclusion compounds. New Journal of Chemistry 2005, 29 (10), 12661271. 130. Bis, J. A.; McLaughlin, O. L.; Vi shweshwar, P.; Zaworotko, M. J., Supramolecular Heterocatemers and Their Role in Cocrystal Design. Crystal Growth & Design 2006, 6 (12), 2648-2650. 131. Papaefstathiou, G. S.; MacGillivray, L. R. Discrete versus Infinite Molecular Self-Assembly: Control in Crystalline Hydrogen-Bonded Assemblies Based on Resorcinol. Organic Letters 2001, 3 (24), 3835-3838. 132. McMahon, J. A.; Bis, J. A.; Vishweshwar, P.; Shattock, T. R.; McLaughlin, O. L.; Zaworotko, M. J., Crystal engineering of th e composition of pharmaceutical phases. 3.

PAGE 168

149 Primary amide supramolecular heterosynth ons and their role in the design of pharmaceutical co-crystals. Zeitschrift fr Kristallographie 2005, 220 (4-2005), 340-350. 133. Babu, N. J.; Reddy, L. S.; Nangia, A., Amide N-Oxide Heterosynthon and Amide Dimer Homosynthon in Cocrystals of Carboxamide Drugs and Pyridine NOxides. Molecular Pharmaceutics 2007, 4 (3), 417-434. 134. Leiserowitz, L.; Hagler, A. T., The Gene ration of Possible Crys tal Structures of Primary Amides. Proceedings of the Royal Societ y of London. A. Mathematical and Physical Sciences 1983, 388 (1794), 133-175. 135. Viswamitra, M. A.; Radhakrishnan, R.; Bandekar, J.; Desiraju, G. R., Evidence for O-H•••C and N-H•••C hydrogen bonding in crystalline alkynes, alkenes, and aromatics. Journal of the Americ an Chemical Society 1993, 115 (11), 4868-4869. 136. Bohmer, V., Calixarenes, Macrocycles With (Almost) Unlimited Possibilities. Angewandte Chemie-International Edition in English 1995, 34 (7), 713-745. 137. Saenger, W., Cyclodextrin InclusionCompounds In Research And Industry. Angewandte Chemie-International Edition in English 1980, 19 (5), 344-362. 138. Kessler, R. C.; Berglund, P.; Demler, O.; Jin, R.; Koretz, D.; Merikangas, K. R.; Rush, A. J.; Walters, E. E.; Wang, P. S., The epidemiology of major de pressive disorder Results from the National Comorbidity Survey Replication (NCS-R). Jama-Journal of the American Medical Association 2003, 289 (23), 3095-3105. 139. Beers, M. H.; Porter, R. S., The Merck Manual of Diagnosis and Therapy 18th Edition: 12-Copy Display Merck & Co: 2006.

PAGE 169

150 140. Klein, P. S.; Melton, D. A., A molecula r mechanism for the effect of lithium on development. Proceedings of the National Academy of Sciences of the United States of America 1996, 93 (16), 8455-8459. 141. Tondo, L.; Isacsson, G.; Baldessarini, R. J., Suicidal behaviour in bipolar disorder Risk and prevention. CNS Drugs 2003, 17 (7), 491-511. 142. Ebadi, M. S.; Simmons, V. J.; Hendricks on, M. J.; Lacy, P. S., Pharmacokinetics of lithium and its regional distribution in rat brain. European Journal of Pharmacology 1974, 27 (3), 324-329. 143. Shan, N.; Zaworotko, M. J., The role of cocrystals in pharm aceutical science. Drug Discovery Today 2008, 13 (9-10), 440-446. 144. Smith, Q. R.; Momma, S.; Aoyagi, M. ; Rapoport, S. I., Kinetics Of Neutral Amino-Acid-Transport Across The Blood-Brain-Barrier. Journal of Neurochemistry 1987, 49 (5), 1651-1658. 145. Braga, D.; Grepioni, F.; Maini, L.; Prosperi, S.; Gobetto, R.; Chierotti, M. R., From unexpected reactions to a new family of ionic co-crystals: the case of barbituric acid with alkali bromides and caesium iodide. Chemical Communications 2010, 46 (41), 7715-7717. 146. Olsher, U.; Izatt, R. M.; Bradshaw, J. S.; Dalley, N. K., Coordination chemistry of lithium ion: a crystal and molecular structure review. Chemical Reviews 1991, 91 (2), 137-164. 147. Pearson, R. G., Hard and Soft Acids and Bases. Journal of the American Chemical Society 1963, 85 (22), 3533-3539.

PAGE 170

151 148. Pearson, R. G., Hard and soft acids and bases, HSAB, part 1: Fundamental principles. Journal of Chemical Education 1968, 45 (9), 581-587. 149. Pearson, R. G., Hard and soft acids and bases, HSAB, part II: Underlying theories. Journal of Chemical Education 1968, 45 (10), 643-648. 150. Mller, G.; Maier, G.-M.; Lutz, M., Lithium coordination to amino acids and peptides. Synthesis, spectroscopic characterization and structure determination of lithium complexes of neutral and ani onic glycine and diglycine. Inorganica Chimica Acta 1994, 218 (1-2), 121-131. 151. Devi, T. U.; Lawrence, N.; Ramesh Ba bu, R.; Selvanayagam, S.; Stoeckli-Evans, H.; Ramamurthi, K., Synthesis, Crys tal Growth and Characterization of L-Proline Lithium Chloride Monohydrate: A New Se miorganic Nonlinear Optical Material. Crystal Growth & Design 2009, 9 (3), 1370-1374. 152. Declercq, J. P.; Meulemans, R.; Piret, P.; Van Meerssche, M., Structure cristalline de derives d'acides amines. II. Le compose d'addition LiBr.L-alanylglycine.2H2O. Acta Crystallographica Section B 1971, 27 (3), 539-544. 153. Gonzlez-Valenzuela, R.; Hernndez-Pa redes, J.; Medrano-Pesqueira, T.; Esparza-Ponce, H. E.; Jess-Castillo, S. ; Rodriguez-Mijangos, R.; Terpugov, V. S.; Alvarez-Ramos, M. E.; Duarte-Mller, A., Gl ycine lithium nitrate crystals: growth and optical properties. Radiation Effects and Defects in Solid s: Incorporating Plasma Science and Plasma Technology 2009, 164 (9), 523 532. 154. Baran, J.; Drozd, M.; Ratajczak, H.; Pietra szko, A., Bis(glycine) lithium nitrate A new non-centrosymmetric crystal: X-ray st ructure, vibrationa l spectra and DSC investigations. Journal of Molecular Structure 2009, 927 (1-3), 43-53.

PAGE 171

152 155. Baerlocher, C.; McCusker, L. B.; Olson, D. H., Atlas of Zeolite Framework Types Sixth Revised ed.; Elsevier Science: Amsterdam, 2007. 156. Cundy, C. S.; Cox, P. A., The hydrothermal synthesis of zeolites: History and development from the earliest days to the present time. Chemical Reviews 2003, 103 (3), 663-701. 157. Zaworotko, M. J., Crystal engi neering of diamondoid networks. Chemical Society Reviews 1994, 23 (4), 283-288. 158. Ockwig, N. W.; Delgado-Fr iedrichs, O.; O'Keeffe, M.; Yaghi, O. M., Reticular Chemistry: Occurrence and Taxonomy of Nets and Grammar for the Design of Frameworks. Accounts of Chemical Research 2005, 38 (3), 176-182. 159. Davis, M. E., Ordered porous materials for emerging applications. Nature 2002, 417 (6891), 813-821. 160. Lopez, S.; Kahraman, M.; Harmata, M.; Keller, S. W., Novel 2-fold Interpenetrating Diamondoid Coordina tion Polymers: [Cu(3,3‘-bipyridine)2]X (X= BF4 -, PF6 -). Inorganic Chemistry 1997, 36 (26), 6138-6140. 161. Evans, O. R.; Lin, W. B., Crystal engi neering of NLO materials based on metalorganic coordination networks. Accounts of Chemical Research 2002, 35 (7), 511-522. 162. Keller, S. W., An Acentric, Three-Dime nsional Coordination Polymer: Synthesis and Structure of [Cu(pyrimidine)2]BF4. Angewandte Chemie Inte rnational Edition in English 1997, 36 (3), 247-248. 163. Liang, K.; Zheng, H. G.; Song, Y. L.; La ppert, M. E.; Li, Y. Z.; Xin, X. Q.; Huang, Z. X.; Chen, J. T.; Lu, S. F., Self-asse mbly of interpenetrating coordination nets

PAGE 172

153 formed from interpenetrating cationic and an ionic three-dimentsional diamondoid cluster coordination polymers. Angewandte Chemie-International Edition 2004, 43 5776-5779. 164. Fang, Q.; Zhu, G.; Xue, M.; Sun, J.; We i, Y.; Qiu, S.; Xu, R., A Metal–Organic Framework with the Zeolit e MTN Topology Containing La rge Cages of Volume 2.5 nm3 Angewandte Chemie International Edition 2005, 44 (25), 3845-3848. 165. Copp, S. B.; Holman, K. T.; Sangster, J. O. S.; Subramanian, S.; Zaworotko, M. J., Supramolecular chemistry of [{M(CO)3(3-OH)}4](M = Mn or Re): a modular approach to crystal engineerin g of superdiamondoid networks. Journal of the Chemical Society, Dalton Transactions 1995, (13), 2233-2243. 166. Copp, S. B.; Subramanian, S.; Zaworotko, M. J., Supramolecular chemistry of manganese complex [Mn(CO)3(3-OH)]4: assembly of a cubic hydrogen-bonded diamondoid network with 1,2-diaminoethane. Journal of the Americ an Chemical Society 1992, 114 (22), 8719-8720. 167. Copp, S. B.; Subramanian, S.; Zaworotko, M. J., Supramolecular chemistry of [M(CO)3(3-OH)]4(M = Mn, Re): spontaneous strict self-assembly of distorted superdiamondoid networks that are capabl e of enclathrating acetonitrile. Journal of the Chemical Society, Chemical Communications 1993, (13), 1078-1079. 168. Hayashi, H.; Cote, A. P.; Furukawa, H.; O/'Keeffe, M.; Yaghi, O. M., Zeolite A imidazolate frameworks. Nat Mater 2007, 6 (7), 501-506. 169. Park, K. S.; Ni, Z.; Ct, A. P.; Choi, J. Y.; Huang, R.; Uribe-Romo, F. J.; Chae, H. K.; O’Keeffe, M.; Yaghi, O. M., Exceptional chemical and thermal stability of zeolitic imidazolate frameworks. Proceedings of the National Academy of Sciences 2006, 103 (27), 10186-10191.

PAGE 173

154 170. Huang, X. C.; Lin, Y. Y.; Zhang, J. P.; Chen, X. M., Ligand-Directed Strategy for Zeolite-Type Metal–Organic Frameworks: Zi nc(II) Imidazolates with Unusual Zeolitic Topologies. Angewandte Chemie International Edition 2006, 45 (10), 1557-1559. 171. Zhang, J.; Wu, T.; Zhou, C.; Chen, S.; Feng, P.; Bu, X., Zeolitic Boron Imidazolate Frameworks. Angewandte Chemie 2009, 121 (14), 2580-2583. 172. Liu, Y.; Kravtsov, V. C.; Larsen, R.; Eddaoudi, M., Molecula r building blocks approach to the assembly of zeolite-like me tal-organic frameworks (ZMOFs) with extralarge cavities. Chemical Communications 2006, (14), 1488-1490. 173. Zheng, S.-T.; Zuo, F.; Wu, T.; Irfanoglu, B.; Chou, C.; Nieto, R. A.; Feng, P.; Bu, X., Cooperative Assembly of Three-Ri ng-Based Zeolite-Type Metal–Organic Frameworks and Johnson-Type Dodecahedra. Angewandte Chemie International Edition 2011, 50 (8), 1849-1852. 174. Alkordi, M. H.; Brant, J. A.; Wojtas, L. ; Kravtsov, V. C.; Cairns, A. J.; Eddaoudi, M., Zeolite-like Metal Organic Frameworks (ZMOFs) Based on the Directed Assembly of Finite Metal Organic Cubes (MOCs). Journal of the American Chemical Society 2009, 131 (49), 17753-17755. 175. Navarro, J. A. R.; Barea, E.; Salas, J. M.; Masciocchi, N.; Galli, S.; Sironi, A.; Ania, C. O.; Parra, J. B., H2, N2, CO, and CO2 Sorption Properties of a Series of Robust Sodalite-Type Microporous Coordination Polymers. Inorganic Chemistry 2006, 45 (6), 2397-2399. 176. Sava, D. F.; Kravtsov, V. C.; Nouar, F.; Wojtas, L.; Eubank, J. F.; Eddaoudi, M., Quest for Zeolite-like Metal Organic Frameworks: On Pyrimidinecarboxylate Bis-

PAGE 174

155 Chelating Bridging Ligands. Journal of the American Chemical Society 2008, 130 (12), 3768-3770. 177. Sreenivasulu, B.; Vittal, J. J., A Metal Coordination Polymer with Hexagonal Diamondoid (or Lonsdaleite) Network Topology. Crystal Growth & Design 2003, 3 (5), 635-637. 178. Chen, Z. L.; Jiang, C. F.; Yan, W. H.; Liang, F. P.; Batten, S. R., ThreeDimensional Metal Azide Coordination Poly mers with Amino Carboxylate Coligands: Synthesis, Structure, and Magnetic Properties. Inorganic Chemistry 2009, 48 (11), 46744684. 179. Heaney, P. J.; Prewitt, C. T.; Gibbs, G. V.; America, M. S. O., Silica: physical behavior, geochemistry and materials applications Mineralogical Soci ety of America: 1994. 180. Hasche, S.; Mock, C.; Otto, J.; Schweppe, F.; Kirschbaum, K.; Krebs, B.; Pinkerton, A. A., Polymeric structures cont aining self-assembled Li---dioxane networks; syntheses and crystal stru ctures of [{Li(dioxane)2.5TaCl4S}n]n/2 dioxane and [{Li2(dioxane)3Cl}n][TaCl6]n. Inorganica Chimica Acta 2000, 298 (1), 9-15. 181. Wu, T.; Zhang, J.; Zhou, C.; Wang, L.; Bu, X.; Feng, P., Zeolite RHO-Type Net with the Lightest Elements. Journal of the Americ an Chemical Society 2009, 131 (17), 6111-6113. 182. Abrahams, B.; Grannas, M.; Hudson, T.; Robson, R., A Simple Lithium(I) Salt with a Microporous Structure and Its Gas Sorption Properties. Angewandte Chemie International Edition 2010, 49 (6), 1087-1089.

PAGE 175

156 183. Morris, J. J.; Noll, B. C.; Henderson, K. W., Assembly of 63,66-Pillared Metal Organic Bilayers and Diamondoi d Lattices Using Molecular Li2O2 Ring Dimers as Secondary Building Units. Crystal Growth & Design 2006, 6 (5), 1071-1073. 184. Banerjee, D.; Kim, S. J.; Li, W.; Wu, H.; Li, J.; Borkow ski, L. A.; Philips, B. L.; Parise, J. B., Synthesis and Structural Characterization of a 3-D Lithium Based Metal Organic Framework Showing D ynamic Structural Behavior. Crystal Growth & Design 2010, 10 (6), 2801-2805. 185. Saravanan, K.; Nagarathinam, M.; Balaya, P.; Vittal, J. J., Lithium storage in a metal organic framework with diamondoid topology a case study on metal formates. Journal of Materials Chemistry 2010, 20 (38), 8329-8335. 186. Wang, Z. M.; Zhang, B.; Fujiwara, H.; Kobayashi, H.; Kurmoo, M., Mn3(HCOO)6: a 3D porous magnet of diamond fram ework with nodes of Mn-centered MnMn4 tetrahedron and guest-modulated ordering temperature. Chemical Communications 2004, (4), 416-417. 187. Zheng, S.-T.; Li, Y.; Wu, T.; Nieto, R. A.; Feng, P.; Bu, X., Porous Lithium Imidazolate Frameworks Constructed with Charge-Complementary Ligands. Chemistry – A European Journal 2010, 16 (44), 13035-13040. 188. Aiello, R.; Barrer R, M.; Kerr I, S., St ages of Zeolite Growth from Alkaline Media. In Molecular Sieve Zeolites-I American Chemical Society: 1974; Vol. 101, pp 44-50. 189. Infantes, L.; Fabian, L.; Motherwell, W. D. S., Organic crysta l hydrates: what are the important factors for formation. CrystEngComm 2007, 9 (1), 65-71.

PAGE 176

157 190. Khankari, R. K.; Grant, D. J. W., Pharmaceutical hydrates. Thermochimica Acta 1995, 248 61-79. 191. Infantes, L.; Chisholm, J.; Motherwell, S., Extended motifs from water and chemical functional groups in organic molecular crystals. CrystEngComm 2003, 5 (85), 480-486. 192. Hickey, M. B.; Peterson, M. L.; Manas, E. S.; Alvarez, J.; Haeffner, F.; Almarsson, ., Hydrates and solid -state reactivity: A survey of -lactam antibiotics. Journal of Pharmaceutical Sciences 2007, 96 (5), 1090-1099. 193. Kiang, Y. H.; Xu, W.; Stephens, P. W.; Ba ll, R. G.; Yasuda, N., Layered Structure and Swelling Behavior of a Multiple Hydrate-Forming Pharmaceutical Compound. Crystal Growth & Design 2009, 9 (4), 1833-1843. 194. Bak, A.; Cooke, M.; Kleppe, M., Wate r Environment Affects Hydration and Dehydration Behaviour of Pharmaceutical Compounds. American Pharmaceutical Review 2008, 11 (6), 116-120. 195. Morris, K. R.; Rodriguez-Hornedo, N., Encyclopedia of Pharmaceutical Technology Marcel Dekker, Inc.: New York, 1993; Vol. 7. 196. Etter, M. C., Encoding and dec oding hydrogen-bond patterns of organic compounds. Accounts of Chemical Research 1990, 23 (4), 120-126. 197. Donohue, J., The Hydrogen B ond in Organic Crystals. The Journal of Physical Chemistry 1952, 56 (4), 502-510. 198. Whler, F., Untersuchungen ber das Chinon. Justus Liebigs Annalen der Chemie 1844, 51 (2), 145-163.

PAGE 177

158 199. Lee, H. H.; Warner, J. C., The Sy stems (I) Diphenyl—Diphenylamine, (II) Diphenyl—Benzophenone and (III) Benzophenone—Diphenylamine. Journal of the American Chemical Society 1933, 55 (1), 209-214. 200. Sakurai, T., The crystal structure of the triclinic modification of quinhydrone. Acta Crystallographica 1965, 19 (3), 320-330. 201. Chadwick, K.; Davey, R. J.; Dent, G.; Pr itchard, R. G.; Hunter, C. A.; Musumeci, D., Cocrystallization: A Solu tion Chemistry Perspective a nd the Case of Benzophenone and Diphenylamine. Crystal Growth & Design 2009, 9 (4), 1990-1999. 202. Cao, G. H.; Sofic, E.; Prior, R. L., Antioxidant and prooxidant behavior of flavonoids: Structure-ac tivity relationships. Free Radical Biology and Medicine 1997, 22 (5), 749-760. 203. Guo, K.; Sadiq, G.; Seaton, C.; Davey, R. ; Yin, Q., Co-Crystallization in the Caffeine/Maleic Acid System: Lessons from Phase Equilibria. Crystal Growth & Design 2009, 10 (1), 268-273. 204. Desiraju, G. R., Crystal Engineering: The Design of Organic Solids Elsevier: Amsterdam, 1989. 205. Ball, P., Scandal of crystal design. Nature 1996, 381 (6584), 648-650. 206. Bernstein, J., Polymorphism A Perspective. Crystal Growth & Design 2011, 11 (3), 632-650. 207. Pepinsky, R., Crystal Engineering New Concept In Crystallography. Phys. Rev. 1955, 100 (3), 971-971.

PAGE 178

159 Appendices

PAGE 179

160 Appendix 1. Solid state characterization data for ELANAM 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 638.57 692.40 753.47 776.65 845.52 923.16 972.56 1039.45 1098.37 1188.81 1327.28 1438.25 1601.32 1673.94 1701.59 3350.41 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98%T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 180

161 Appendix 2. Solid state characterization data for ELAINM•1.7H2O 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 640.74 651.64 678.77 752.36 774.90 852.72 923.36 1010.67 1047.78 1106.39 1189.71 1211.61 1225.52 1268.46 1339.91 1421.13 1555.94 1584.78 1603.30 1682.51 1722.53 3175.91 3350.95 3591.77328k EAINTA dihydrate 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 181

162 Appendix 3. Solid state characterization data for ELAINM•4NMP 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 638.18 657.49 761.10 812.85 885.21 915.75 1008.69 1046.03 1092.50 1180.67 1304.11 1343.43 1413.38 1443.17 1505.48 1556.54 1605.96 1666.07 1691.03 1725.44 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 182

163 Appendix 4. Solid state characterization data for ELAINM•4DMA 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 183

164 Appendix 5. Solid state characterization data for ELACAF•H2O 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 741.12 759.47 921.40 1029.80 1042.55 1103.16 1175.67 1195.73 1322.28 1443.62 1583.46 1619.23 1635.51 1697.59 3262.45 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 184

165 Appendix 6. Solid state characterization data for ELATPH•0.13H2O 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative IntensityCalculated Experimental 608.48 619.40 744.30 756.59 831.97 916.72 1042.82 1092.59 1161.00 1179.63 1319.62 1334.37 1440.64 1559.43 1617.93 1636.74 1692.92 1736.39 2861.43 2981.07 3062.10 3356.38 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 185

166 Appendix 7. Solid state characterization data for ELATPH•2DMA 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 186

167 Appendix 8. Solid state characterization data for ELADMP 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 601.71 626.48 663.66 728.76 753.01 769.69 782.63 869.72 919.52 964.81 1019.73 1041.92 1061.40 1104.21 1150.52 1201.88 1268.79 1298.78 1333.42 1376.15 1387.81 1417.82 1450.38 1478.31 1573.82 1604.07 1710.62 3259.64 3455.21319e EADPZ 40 45 50 55 60 65 70 75 80 85 90 95 100 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 187

168 Appendix 9. Solid state characterization data for ELASAR 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 601.71 641.21 653.70 674.26 739.34 753.78 812.95 861.67 879.96 911.48 961.22 1037.21 1057.85 1084.52 1175.92 1221.88 1271.76 1339.64 1363.50 1404.95 1449.32 1504.48 1584.44 1723.90 1738.45 3108.23 72 74 76 78 80 82 84 86 88 90 92 94 96 98 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 188

169 Appendix 10. Solid state characterization data for ELADMG 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 636.49 690.76 740.88 753.04 809.73 863.38 914.11 957.17 995.70 1042.67 1090.40 1172.06 1210.05 1256.23 1305.83 1340.26 1403.18 1433.74 1456.25 1505.73 1592.94 1733.67 3125.22 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 189

170 Appendix 11. Solid state characterization data for ELACAP 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 60522 661.62 719.31 759.17 868.02 920.59 1043.08 1107.36 1185.83 1328.24 1437.98 1583.86 1609.43 1714.93 2932.36 3333.50 55 60 65 70 75 80 85 90 95 %T 1000 2000 3000 Wavenumbers (cm-1)

PAGE 190

171 Appendix 12. Solid state characte rization data for ELAURE•2NMP 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 191

172 Appendix 13. Solid state characterization data for ELADMA4 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 192

173 Appendix 14. Solid state characterization data for ELADMA2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 193

174 Appendix 15. Solid state characterization data for ELADMSO2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 194

175 Appendix 16. Solid state characterization data for ELANMP2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 195

176 Appendix 17. Solid state characterization data for ELAPG2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 196

177 Appendix 18. Solid state characterization data for LICSAR 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 197

178 Appendix 19. Solid state characterization data for LINSAR 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 198

179 Appendix 20. Solid state characterization data for LICDMG 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 618.54 667.71 723.50 856.89 928.14 969.77 990.98 1013.39 1167.06 1227.75 1291.80 1318.61 1368.12 1409.10 1423.84 1448.52 1473.32 1636.11 3036.82 3372.33 3412.09LiDMGH2O50122 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100 102 104 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 199

180 Appendix 21. Solid state characterization data for LINDMG 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 200

181 Appendix 22. Solid state characterization data for LICBAL-anhydrate 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 700.84 798.16 838.56 970.47 1043.49 1294.37 1322.28 1382.19 1426.20 1457.18 1492.11 1578.93 3000.91 3132.11 3395.37T418g LITbetaALA 150degC 55 60 65 70 75 80 85 90 95 100 105 110 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 201

182 Appendix 23. Solid state characterization data for LICBAL 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 659.84 837.86 874.32 949.17 1042.76 1084.66 1124.22 1250.40 1294.87 1321.02 1383.16 1400.26 1422.28 1456.18 1590.10 1649.82 3039.76 3392.66125f LITBAH 50 55 60 65 70 75 80 85 90 95 100 105 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 202

183 Appendix 24. Solid state characterization data for LICABA 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 668.65 699.65 756.65 961.46 984.06 1149.38 1242.40 1261.86 1341.98 1403.97 1438.08 1477.37 1546.32 1611.06 2927.02 2959.58 2985.04 3064.27T418h LITGABA 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 203

184 Appendix 25. Solid state characterization data for LICPRO 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 204

185 Appendix 26. Solid state characterization data for LIBPRO 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 653.15 775.74 844.70 908.63 922.43 945.24 970.25 1030.44 1167.34 1229.56 1276.95 1327.14 1368.13 1421.74 1447.90 1473.06 1522.35 1613.53 3085.99 3186.44 3392.70 3416.85145t LIBPRO 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103%T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 205

186 Appendix 27. Solid state characterization data for LINPRO 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 206

187 Appendix 28. Solid state characterization data for LICSAR2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 611.77 626.53 710.52 866.41 894.31 935.43 964.43 1052.32 1137.33 1265.40 1306.61 1406.97 1447.49 1471.36 1496.22 1606.10 1670.22 2444.12 2705.65 2787.79 2930.08145a LICSAR2 84 86 88 90 92 94 96 98 100 102 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 207

188 Appendix 29. Solid state characterization data for LINBTN2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 208

189 Appendix 30. Solid state characterization data for LICDMG2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 209

190 Appendix 31. Solid state characterization data for LIBDMG2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 706.03 867.69 931.42 967.12 1003.39 1143.82 1170.95 1282.30 1332.63 1383.30 1420.98 1473.73 1605.11146n LIBDMG2 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100 102 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 210

191 Appendix 32. Solid state characterization data for LICPRO2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental

PAGE 211

192 Appendix 33. Solid state characterization data for LIBPRO2 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental 603.70 710.59 762.99 802.07 854.27 908.28 958.43 989.50 1049.02 1078.86 1191.88 1291.35 1335.51 1381.07 1401.66 1449.80 1593.25 1622.27 1644.49 2204.97 2498.98 2572.37 2657.35 2754.82 2926.57 3401.27142e LIBPRO2 50 55 60 65 70 75 80 85 90 95 100 105 %T 1000 2000 3000 4000 Wavenumbers (cm-1)

PAGE 212

193 Appendix 34. Solid state characterization data for LINPRO2 (dia) 051015202530354045 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Relative Intensity2Calculated Experimental