|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
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 2200409Ka 4500
controlfield tag 001 002063144
007 cr mnu|||uuuuu
008 100315s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003056
Niemann, Michael Ulrich.
Development and investigation of novel nanostructures and complex hydrides for hydrogen storage
h [electronic resource] /
by Michael Ulrich Niemann.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 178 pages.
Dissertation (Ph.D.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
ABSTRACT: Over the past few years, the need for a clean and renewable fuel has sharply risen. This is due to increasing fossil fuel costs and the desire to limit or eliminate harmful by-products which are created during the burning of these fuels. Hydrogen is the most abundant element in the universe and can be used in either fuel cells or traditional internal combustion engines to produce energy with no harmful emissions. One of the main obstacles facing the implementation of a hydrogen economy is its storage. Classical methods of storage involve either high and unsafe pressures or liquid storage involving a large amount of energy. Two alternative hydrogen storage methods are investigated physisorption, which is the weak chemical bonding to a material, as well as chemisorption, which is a strong chemical bond of hydrogen to a host material.Polyaniline, a conducting polymer, is investigated in both its bulk form as well as in nanostructured forms, more precisely nanofibers and nanospheres, to store hydrogen via physisorption. It is found the bulk form of polyaniline can store only approximately 0.5wt.% hydrogen, which is far short of the 6wt.% required for practical applications. Nanofibers and nanospheres, however, have been developed, which can store between 4wt.% and 10wt.% of hydrogen at room temperature with varying kinetics. A new complex metal hydride comprised of LiBH, LiNH and MgH has been developed to store hydrogen via chemisorption.While the parent compounds require high temperatures and suffer of slow kinetics for hydrogen sorption, the work performed as part of this dissertation shows that optimized processing conditions reduce the hydrogen release temperature from 250C to approximately 150C, while the addition of nano sized materials has been found to increase the kinetics of hydrogen sorption as well as further decrease the hydrogen release temperature, making this one of the first viable hydrogen storage materials available. This is the first time that nanostructured polyaniline has been investigated for its hydrogen performance. Additionally, the thorough investigation of the effects of nano sized additives and processing parameter optimization of the multinary hydride are first reported in this dissertation.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Co-advisor: Ashok Kumar, Ph.D.
Co-advisor: Elias K. Stefanakos, Ph.D.
x Mechanical Engineering
t USF Electronic Theses and Dissertations.
Development and Investigation of Novel Nanostructures and Complex Hydrides for Hydrogen Storage by Michael Ulrich Niemann A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Mechanical Engineering College of Engineering University of South Florida Co-Major Professor: Ashok Kumar, Ph.D. Co-Major Professor: Elias K. Stefanakos, Ph.D. Matthias Batzill, Ph.D. Delcie Durham, Ph.D. D. Yogi Goswami, Ph.D. Autar Kaw, Ph.D. Sesha S. Srinivasan, Ph.D. Date of Approval: May 26, 2009 Keywords: polyaniline, chemisorption, ki netics, complex hydride, physisorption Copyright 2009, Michael Ulrich Niemann
Dedication This dissertation is dedicated to my pare nts, Dieter Horst Kunibert and Ursula Luzia Jurczyk as well as my wife Christine Victoria Niemann, my son Sebastian Eduard Gerhard Niemann and any other ch ildren I will hopefully have. Diese Dissertation ist meinen Eltern, Di eter Horst Kunibert und Ursula Luzia Jurczyk, als auch meiner Frau Christine Victoria Niemann, meinem Sohn Sebastian Eduard Gerhard Niemann, als auch meinen anderen Kindern, die ich hoffentlich eines Tages haben werde.
Acknowledgements I would like to thank my research advi sors, Dr. Ashok Kumar and Dr. Elias K. Stefanakos for providing me with the extrao rdinary opportunity of pursuing this research as well as for all the guidance they lent to me over the years. I would also like to thank Dr. Sesha S. Srinivasan for his invaluable he lp throughout the entire research as well as for helping me with anything I ever needed help with. Additionally, I would like to thank Dr. D. Yogi Goswami, Dr. Delcie Durham, Dr. Autar Kaw, and Dr. Matthias Batzill for being part of my dissertation committee, as well as Dr. John Wolan for being the chair of my Ph.D. defense. Additionally, I am very grateful to th e Department of Energy that provided funding for this research as well as Quan tumSphere, Inc. for additional funding and material support. I would also like to thank th e outstanding people at the department of mechanical engineering, specifically, Sue Britten and Sh irley Trevor for the many, many things they have done for me. Finally, I would like to thank my fam ily and friends for their patience and emotional support throu ghout this endeavor.
i Table of Contents List of Tables ................................................................................................................ ..... ivList of Figures ............................................................................................................... ...... vAbstract ...................................................................................................................... ........ xiChapter 1 Introduction ...................................................................................................... 11.1. Hydrogen A Clean and Renewable Fuel ............................................................. 11.2. Obstacles Facing the St orage of Hydrogen ............................................................ 41.3. Current State of Hydrogen Storage ........................................................................ 71.3.1. Compressed Hydrogen Gaseous Hydrogen Storage ................................... 81.3.2. Liquid Hydrogen Storage .............................................................................. 91.3.3. Physisorption............................................................................................... 101.3.4. Chemisorption ............................................................................................. 111.4. Dissertation Outline ............................................................................................. 121.5. Dissertation Goals and Si gnificance of the Study ................................................ 12Chapter 2 Experimental Equipment ............................................................................... 182.1. Synthesis and Materials ....................................................................................... 182.1.1. Glove Box ................................................................................................... 192.1.2. Polymerization ............................................................................................ 192.1.3. Ball Milling ................................................................................................. 202.2. Thermal Characterization ..................................................................................... 222.2.1. Thermogravimetric Analysis (TGA) ........................................................... 222.2.2. Differential Scanning Calorimetry (DSC) .................................................. 262.2.3. Thermal Programmed Desorption (TPD) ................................................... 272.3. Structural Characterization .................................................................................. 302.3.1. X-Ray Diffraction (XRD) ........................................................................... 312.3.2. Fourier Transform Infrared Spectroscopy (FTIR) ...................................... 332.3.3. Scanning Electron Microscopy (SEM) ....................................................... 342.4. Hydrogen Sorption Measurements (PCT) ........................................................... 35Chapter 3 Physisorption in Polyaniline .......................................................................... 403.1. Introduction .......................................................................................................... 403.2. Bulk Polyaniline................................................................................................... 443.2.1. Synthesis of Bulk Polyaniline ..................................................................... 463.2.2. Bulk Polyaniline Characterization .............................................................. 473.2.3. FTIR Characterization Results .................................................................... 49
ii 3.2.4. Thermogravimetric Results ......................................................................... 503.2.5. Scanning Electron Mi croscopy Results ...................................................... 523.2.6. Hydrogen Sorption Measurements ............................................................. 533.2.7. Bulk PANI Summary .................................................................................. 583.3. PANI Nanospheres............................................................................................... 593.3.1. Synthesis of Polyaniline Nanospheres ........................................................ 603.3.2. Scanning Electron Mi croscopy Results ...................................................... 613.3.3. FTIR Characterization Results .................................................................... 633.3.4. Hydrogen Sorption Results ......................................................................... 643.3.5. Hydrogen Cycling Effect s on PANI Nanospheres ...................................... 673.3.6. PANI Nanospheres Summary ..................................................................... 683.4. Chemically Grown PANI Nanofibers .................................................................. 693.4.1. Synthesis of Polyaniline Nanofibers ........................................................... 703.4.2. FTIR Characterization Results .................................................................... 713.4.3. Scanning Electron Mi croscopy Results ...................................................... 733.4.4. Hydrogen Sorption Results ......................................................................... 753.4.5. Hydrogen Cycling Effect s on PANI Nanofibers ........................................ 803.4.6. PANI Nanofibers (CM) Summary .............................................................. 823.5. Electrospun PANI Nanofibers ............................................................................. 843.5.1. Synthesis of Electrospun Polyaniline Nanofibers ....................................... 843.5.2. Scanning Electron Mi croscopy Results ...................................................... 873.5.3. FTIR Characterization Results .................................................................... 883.5.4. Thermogravimetric Results ......................................................................... 903.5.5. Hydrogen Sorption Results ......................................................................... 913.5.6. Hydrogen Cycling Effects on El ectrospun PANI Nanofibers .................... 973.5.7. Electrospun PANI Nanofibers Summary .................................................... 993.6. Polyaniline Hydrogen Storage Summary ........................................................... 101Chapter 4 Complex Hydrides LiBH4/LiNH2/MgH2 ..................................................... 1064.1. Introduction ........................................................................................................ 1064.2. Quaternary Complex Hydride LiBH4/LiNH2 (LiBNH) ..................................... 1124.2.1. Synthesis of the Quaternary LiBNH ......................................................... 1124.2.2. Hydrogen Characteristics of the Quaternary LiBNH ................................ 1164.2.3. Activation Energy of the Quaternary LiBNH ........................................... 1194.2.4. Destabilization of the Quaternary LiBNH with Nano Sized Additives .... 1224.2.5. Quaternary LiBNH Summary ................................................................... 1294.3. Destabilization of LiBH4/LiNH2 with MgH2 ..................................................... 1304.3.1. Synthesis of the Multinary Hydrides ........................................................ 1324.3.2. X-ray Diffraction Results .......................................................................... 1344.3.3. FTIR Characterization Results .................................................................. 1354.3.4. Thermal Programmed De sorption Results ................................................ 1374.3.5. Activation Energy Results ........................................................................ 1384.3.6. Pressure-Composition-Temperature Isotherms ........................................ 1414.3.7. Crystallite Size Effects on Hydrogen Capacity ........................................ 1444.3.8. Surface Morphology ................................................................................. 146
iii 4.3.9. MgH2 Destabilized LiBH4/LiNH2 Summary ............................................ 1494.4. Destabilization of LiBH4/LiNH2/MgH2 with Nano Sized Additives ................. 1504.4.1. Synthesis of Destabilized Multinary Hydride ........................................... 1514.4.2. Thermal Programmed De sorption Results ................................................ 1514.4.3. Hydrogen Sorption Screening Results ...................................................... 1534.5. Complex Hydrides Storage Summary ............................................................... 157Chapter 5 Conclusions and Future Work Recommendation ........................................ 1615.1. Overview ............................................................................................................ 1615.2. Physisorption in Polyaniline Summary and Conclusions ............................... 1615.3. Complex Hydride Summ ary and Conclusions ................................................ 1665.4. Future Work Recommendation .......................................................................... 168References .................................................................................................................... ... 170 About the Author .................................................................................................. End Page
iv List of Tables Table 1.1: Department of Energy hydroge n storage system requirements .. ...........5Table 1.2: Comparison of some hydrogen storage methods . ................................7Table 2.1: Materials used as part of this dissertation w ith their respective purity. .........................................................................................................18Table 3.1: Hydrogen sorption capacity of electrospun nanofibers at various temperatures and sorption cycles. ............................................................101Table 4.1: Physical charact eristics of some previously investigated complex hydrides. ...................................................................................................106Table 4.2: Hydrogen performance of quaternary LiBNH with and without nano Zn and nano Ni. ...............................................................................129Table 4.3: Comparison of the results for the multinary complex hydrides developed by different processing conditions (the best results are shown in bold)..........................................................................................150
v List of Figures Figure 1.1: The zero emission, clean and renewable hydrogen cycle showing production, storage and use by either fuel cells (FC) or hydrogen internal combustion engine (HICE). ............................................................4Figure 1.2: Schematic representation of physisorption and chemisorption with catalyst materials. .......................................................................................14Figure 1.3: Hydrogen storage mechanism in complex hydrides and the effect of particle size. ...........................................................................................16Figure 1.4: Particle size e ffects on hydrogen storage. ..................................................17Figure 2.1: Innovative Technol ogy System One glove box. ........................................19Figure 2.2: Stainless steel bowl with stainless steel balls (left), Fritsch Pulverisette P5 (right). ...............................................................................21Figure 2.3: Thermal Analysis SD T Q600 simultaneous DSC and TGA ..................23Figure 2.4: Schematic representation of the most important parts of a TGA. ..............24Figure 2.5: Typical TGA profile Calc ium oxalate exhibiting weight loss due to water, carbon monoxide and finally carbon dioxide loss. .....................25Figure 2.6: Thermal Analysis DSC Q10. .....................................................................26Figure 2.7: Typical DSC plot showi ng endothermic glass transition and melting as well as exothermic recr ystallization and decomposition. .........27Figure 2.8: Thermal programmed de sorption (TPD) with optional TCD controller. ...................................................................................................28Figure 2.9: Typical TPD pl ot showing three-step hydrogen release of LiAlH4. ..........30Figure 2.10: Philips X'Pert XRD used for characterizing the chemical structure of processed complex hydrides. .................................................................31Figure 2.11: (a) X-rays not in phase ca using a base line of XRD signal, (b) xrays in phase causing a peak in XRD signal. .............................................32
vi Figure 2.12: Perkin Elmer Spectrum One FTIR. ............................................................33Figure 2.13: Hitachi S800 SEM. ....................................................................................34Figure 2.14: HyEnergy's PCT Pro 2000 used for hydrogen sorption PCT and kinetics measurements. ..............................................................................35Figure 2.15: Desorption PCT plot of LiAlH4 showing plateau pressure around 15bar. .........................................................................................................38Figure 2.16: Kinetics plot of LiAlH4 showing absorption kinetics. ...............................39Figure 3.1: Chemical structure of emeraldine base polyaniline. ..................................40Figure 3.2: Previous work on hydrogen storage in polymers, mainly in polymers of intrinsic microscopy (PIM). ...................................................44Figure 3.3: Hydrogen sorption mechanism of standard polyaniline composite material. .....................................................................................................45Figure 3.4: FTIR comparing standard polyaniline (PANI-S TD), polyaniline with 10wt.% SnO2, polyaniline with 10wt.% MWCNT and polyaniline with 30wt.% aluminum powder. .............................................49Figure 3.5: Thermogravimetric analysis of standard polyaniline, polyaniline with 10wt.% multiwall carbon nanot ubes, polyaniline with 10wt.% tin oxide, and polyaniline after hydr ogen sorption measurements. ...........51Figure 3.6: Scanning electron microscopy images of standard polyaniline ((a) and (b)) before and ((c) and (d)) after hydrogen interaction indicating a ballooning caus ed by hydrogen cycling. ................................53Figure 3.7: Hydrogen sorption measurem ents of standard polyaniline at different temperatures. ...............................................................................55Figure 3.8: Hydrogen sorption measurements of polyaniline with 10wt.% SnO2 at different temperatures and increasing pressure. ...........................56Figure 3.9: Hydrogen sorption measurements of polyaniline with 10wt.% multiwall carbon nanotubes at different temperatures and increasing pressure. ....................................................................................57Figure 3.10: Hydrogen sorption measurements of polyaniline with 30wt.% fine aluminum powder at different temperatures and increasing pressure. .....................................................................................................58Figure 3.11: Synthesis of PANI nanospheres. ................................................................61
vii Figure 3.12: SEM of PANI nanosphe res at a magnification of 25000. ..........................62Figure 3.13: FTIR comparison of bul k PANI with PANI nanospheres. ........................63Figure 3.14: Absorption and desorption kinetics of PANI nanospheres at 30oC. ..........65Figure 3.15: Absorption kinetics of PANI nanospheres at 90oC. ...................................66Figure 3.16: SEM image of PANI na nospheres after hydrogen cycling. .......................68Figure 3.17: Flow chart for the synthesi s of polyaniline nanofibers prepared in an aqueous medium with different surfactants and using ammonium persulfate as oxidizing agent. .................................................71Figure 3.18: FTIR spectra of PANI nanof iber and standard sample indicating that the major bonding environment remains unchanged for both standard and nanofibrous polyaniline structures. .......................................72Figure 3.19: Scanning electron micrographs of polyaniline nanofibers grown at room temperature in aqueous medium with different surfactants (a) dodecyl benezene sulfonic acid, (b) acrymethylpropyl sulfonic acid, and (c) camphorosulfonic acid us ing ammonium persulfate as oxidizing agent. ..........................................................................................74Figure 3.20: Hydrogen absorption and desorp tion kinetics of PANI-NF in the 1st and 13th cycle. ............................................................................................76Figure 3.21: Pressure-Composition Isotherms (PCT) of PANI-NF at room temperature from 2nd to 6th (a) absorption (b) desorption cycles. ..............78Figure 3.22: Hydrogen sorption kinetics at room temperature from 14th cycle to 25th cycle showing little degradation in kinetics and no degradation in the storage capacity with in the measured cycles. ..................................80Figure 3.23: SEM image of PANI na nofibers after hydr ogen cycling. ..........................81Figure 3.24: BET surface area analysis of PANI nanofibers (CM) before and after hydrogen cycling. ..............................................................................82Figure 3.25: Flow chart for the synthesi s of electrospun polyaniline nanofibers. .........85Figure 3.26: Schematic diagram of electrospinning method to produce polyaniline nanofibers. ...............................................................................86Figure 3.27: SEM image of electrospun PANI nanofiber s at a magnification of (a) 1000 and (b) 5000. ................................................................................88
viii Figure 3.28: FTIR spectra of bulk PANI and electrospun PANI nanofibers. ................89Figure 3.29: Thermogravimetric weight lo ss analysis of PANI samples in both bulk and nanofiber form. ............................................................................91Figure 3.30: Hydrogen sorption kinetics cu rves for the PANI-NF-ES at 30 and 100oC..........................................................................................................93Figure 3.31: Hydrogen adsorption PCT curves for the PANI-NF-ES at different temperatures. ..............................................................................................95Figure 3.32: Hydrogen desorption PCT curves for the PANI-NF-ES at different temperatures. ..............................................................................................96Figure 3.33: SEM micrographs of PANI-N F (a) before and (b) after hydrogen sorption cycling. .........................................................................................98Figure 3.34: FTIR spectra of electros pun PANI nanofibers before and after hydrogen cycling. .......................................................................................99Figure 3.35: Comparison of novel PANI results with previous polymer hydrogen storage research. .......................................................................105Figure 4.1: Boron (top) and lithium ( bottom) coordinations in orthorhombic LiBH4 at room temperature . .............................................................108Figure 4.2: Hydrogen sorption capacity a nd temperature of selected complex hydrides and chemical hydrides with DOE target range highlighted. ..............................................................................................109Figure 4.3: Schematic mechano chemi cal synthesis approach for complex hydrides to reduce particle size and achieve a homogenous mixture of parent compounds. ...............................................................................111Figure 4.4: Destabilization of complex hydride through a dditives or catalysts. ........112Figure 4.5: XRD pattern comp aring parent compounds LiBH4 and LiNH2 for different milling durations. ......................................................................114Figure 4.6: FTIR comparison of parent compounds LiBH4 and LiNH2 with selected milling durations. .......................................................................116Figure 4.7: Hydrogen PCT desorp tion of quaternary LiBNH. ...................................117Figure 4.8: Hydrogen PCT absorp tion of quaternary LiBNH. ...................................118Figure 4.9: Absorpti on kinetics at 250oC of LiBNH after dehydrogenation. .............119
ix Figure 4.10: TPD plot of the quaternar y LiBNH for heating rates of 1, 5, and 10K/min. ..................................................................................................120Figure 4.11: Kissinger plot of th e quaternary structure LiBNH. ..................................121Figure 4.12: TGA and DSC data of the quaternary with 1, 3, and 5 mol% of nano Ni. ....................................................................................................123Figure 4.13: Hydrogen PCT desorption of LiBNH with 3mol% nano Zn. ..................124Figure 4.14: Hydrogen PCT absorption of LiBNH with 3mol% nano Zn. ..................125Figure 4.15: Hydrogen PCT performan ce of LiBNH with 5mol% nano Ni. ...............126Figure 4.16: Hydrogen PCT performance of LiBNH with 3mol% nano Ni and 3mol% nano Zn. .......................................................................................127Figure 4.17: Hydrogen PCT performance of LiBNH with 5mol% nano Ni and 3mol% nano Zn. .......................................................................................128Figure 4.18: Hydrogen performance of quaternary LiBNH with and without nano Ni and nano Zn. ...............................................................................130Figure 4.19: Processing condition flow char t of the five samples investigated showing the two main processing schemes employed. ............................134Figure 4.20: XRD profile of the five diffe rently processed materials as well as the parent compounds, LiBH4, LiNH2, commercial MgH2, and nano MgH2. ..............................................................................................135Figure 4.21: FTIR spectra of the variously processed multinary hydrides. .................137Figure 4.22: TPD comparison of investigat ed processing variations showing the two main hydrogen release regions around 160oC and 300oC. ................138Figure 4.23: Activation energy, as calc ulated from the TPD data using Kissingers method, compared with the first peak hydrogen release temperature. .............................................................................................140Figure 4.24: Activation energy, as calc ulated from the TPD data using Kissingers method, compared w ith the main peak hydrogen release temperature. .................................................................................141Figure 4.25: Comparison of the hydrogen sorp tion characteristics of the various processing conditions at the lowest hydrogen release temperature. ........144
x Figure 4.26: Comparison of hydrogen concen tration and crystallite size of the quaternary LiBNH and MgH2 phases after milling. ................................146Figure 4.27: Surface morphology of the fi ve samples at 2200x magnification of (a) LicMgBNH, (b) LinMgBNH, (c) 10hr LicMgBNH, (d) LiBNH + cMgH2, and (e) LiBNH + nMgH2. .......................................................148Figure 4.28: TPD comparison of LiBNH+nMgH2 without additive and with 2mol% Ni, Cu, Mn, Co and Fe at a constant ramping rate of 1oC/min. ...................................................................................................152Figure 4.29: Ramping kinetics measurem ents of LiBNH+nMgH2 without and with 2mol% nano Mn, Fe, Co, Cu, Ni and Fe+Ni. ..................................154Figure 4.30: Comparison of hydrogen rele ase temperature and hydrogen release rate of the standard LiBNH+nMgH2 and LiBNH+MgH2 with 2mol% of various nano additives. ............................................................155Figure 4.31: Ramping desorption kinetic s of LiBNH+nMgH2 without and with 2mol%, 4mol% and 10mol% nano nickel. ...............................................156Figure 4.32: Ramping desorption kinetics comparing LiBNH+nMgH2 without and with 2mol% and 10mol% nano iron..................................................157Figure 4.33: Complex hydride storage materials comparison with other published materials. .................................................................................159
xi Development and Investigati on of Novel Nanostructures and Complex Hydrides for Hydrogen Storage Michael Ulrich Niemann ABSTRACT Over the past few years, the need for a cl ean and renewable fuel has sharply risen. This is due to increasing fossil fuel costs a nd the desire to limit or eliminate harmful byproducts which are created duri ng the burning of these fu els. Hydrogen is the most abundant element in the universe and can be used in either fuel cells or traditional internal combustion engines to produce energy with no ha rmful emissions. One of the main obstacles facing the implementation of a hydrogen economy is its storage. Classical methods of storage involve e ither high and unsafe pressures or liquid storage involving a large amount of energy. Two alternative hydrogen st orage methods are inves tigated physisorption, which is the weak chemical bonding to a mate rial, as well as chemisorption, which is a strong chemical bond of hydrogen to a host material. Polyaniline, a conducting polymer, is investigated in both its bulk form as well as in nanostructured forms, more precisely nanofibers and nanospheres, to store hydroge n via physisorption. It is found the bulk form of polyaniline can store only approxima tely 0.5wt.% hydrogen, which is far short of the 6wt.% required for practical applications Nanofibers and nanospheres, however, have
xii been developed, which can store between 4wt.% and 10wt.% of hydrogen at room temperature with varying kinetics. A new complex metal hydride comprised of LiBH4, LiNH2 and MgH2 has been developed to store hydrogen via chemisorption. While the parent compounds require high temperatures and suffer of slow kinetics fo r hydrogen sorption, the work performed as part of this dissertation s hows that optimized processi ng conditions reduce the hydrogen release temperature from 250oC to approximately 150oC, while the addition of nano sized materials has been found to increase the kine tics of hydrogen sorpti on as well as further decrease the hydrogen release temperature, ma king this one of the first viable hydrogen storage materials available. This is the first time that nanostructured polyaniline has been investigated for its hydrogen performance. Additionally, the thoro ugh investigation of the effects of nano sized additives and processing parameter op timization of the multinary hydride are first reported in this dissertation.
1 Chapter 1 Introduction 1.1. Hydrogen A Clean and Renewable Fuel Whether one believes that earth has r eached peak oil or that humans are responsible for global warming due to the em issions caused by the burning of fossil fuels, or even if one just thinks that it is impor tant to become energy independent to prevent future conflicts, there is consensus that the search for alternative clean and renewable energy should be a prerogative in the near futu re . One of the many options is to use hydrogen as a fuel . Hydrogen, provided it is produced using clean and renewable energy sources, such as solar energy, can eith er be combusted in an internal combustion engine or used in a fuel cell [3, 4] to produce energy that is free of any pollutant byproducts, producing solely energy and wate r. Though there are many barriers towards realizing a hydrogen economy, one of the biggest challenges is to find a safe and efficient means of storing the hydrogen for use in mobile applications . Current options include storing hydrogen in its liquid form or as a compressed gas. Both methods require a large amount of energy and can pose serious safety risks. Therefore, there is a push to find a material to chemically store hydrogen using, for example, metal hydrides  or complex hydrides . There are, however, many challe nges that these materials must overcome. Specifically, these are to have fast kine tics, a high capacity, e.g. more than 6wt.% hydrogen, and to be reusable for at least 1000 cycles .
2 Over the past few years, the scientific evidence for the inevitability of the socalled peak-oil has become almost overwhelm ing. Even the optimists  amongst the scientists studying peak-oil agr ee that alternatives to oil need to be found. And the sooner this happens, the better it will be for hum anity. If one imagines the oil supply as a clock, peak oil would occur at roughly si x oclock . And, according to many scientists, it is approximately 5:30 peak oil is near. It is no longer a question of whether oil will run out, but when it will run out. Th is is only one reason for the need for a substitute to fossil fuels. In addition to the peak-oil problem, the idea of global warming is finding more supporters in the scientific community than it ever has. It is widely agreed that an alternative to fossil fuels must be explored . Hydrogen is one viable alternative to the black gold, and has been shown to successfully power both fuel cells as well as internal combustion engines . All major automobile manufacturers such as BMW, Daimler, Volkswagen, Honda, General Motors, and For d, to name a few, have successfully demonstrated the use of hydrogen as a chemic al or ignition fuel, though the companies vary on the way the hydrogen is used. BMW, for example, is one of the few automobile manufacturers that uses a modi fied internal combustion engine as opposed to a fuel cell. One of the main problems with the utiliz ation of hydrogen as a fuel is its onboard storage. The Department of Energy of the Un ited States of America and the Freedom Car Technology Team have set forth very specific gu idelines as to the requirements that need to be met by a hydrogen storage material to su ccessfully compete, and eventually replace, fossil fuels. The specific targets, as they apply to this dissertation are discussed in Section
3 1.2. A thorough review of current hydrogen storage research is given by Schlapbach and Zttel . Hydrogen, the most abundant element in th e universe, is a clean and renewable energy carrier. It can be used in a fuel cell (FC) to produce el ectricity as well as burned in a hydrogen internal combustion engine (HICE), provided it is adjusted for use of hydrogen as the main fuel. Figure 1.1 is a sc hematic representation of the ideal clean and renewable hydrogen cycle. It also represents the three main challenges facing the use of hydrogen as a fuel source: (1) A clean and effici ent production of hydrogen (2) An efficient and safe storage system for hydrogen (3) An efficient and inexpensive mean s of utilizing hydrogen as a fuel The ideal production of hydrogen is achie ved utilizing clean and renewable energy sources, such as wind energy, solar energy or geo-thermal means to produce electricity, which in turn can split water into oxygen and hydrogen, where the hydrogen is then processed and stored while the oxygen is ei ther released to the atmosphere or bottled and sold for medical purposes, etc. The storage of the hydrogen on board of an automobile is the topic of this dissertation and will not be further elaborated in this section. Finally, the hydrogen needs to be used to produce energy. This can be achieved either by using a hydrogen fuel cell, wher e the hydrogen is used to produce electricity using, for example, a polymer electrolyte fu el cell or an internal combustion engine, which is modified to run on hydrogen instead of gasoline. Both of these technologies have their advantages as well as drawbacks and are still being optim ized. In the end, the
4 water produced during either the burning of hydroge n or in a fuel cell is simply released back to the environment or collected for further use. 1.2. Obstacles Facing the Storage of Hydrogen Under ambient conditions hydrogen is a gas, which is highly reactive with other compounds. Therefore, it is not generally found in its molecular form. In order to utilize hydrogen as a practical means for replacemen t of gasoline in auto mobiles, the U.S. Department of Energy (DOE) has established crite ria that need to be met. These criteria Hydrogen Hydrogen Oxygen Oxygen Water Storage FC/HICE Production Figure 1.1: The zero emission, clean and re newable hydrogen cycle showing production, storage and use by either fuel cells (FC) or hydrogen internal combustion engine (HICE).
5 address not only the cost of the hydrogen stor age system, but also the gravimetric and volumetric hydrogen storage requirements of a ny material to be used. The fueling time and delivery temperatures of hydrogen are al so of utmost importance. Table 1.1 provides an overview of the requirements of any hydrogen storage system. Table 1.1: Department of Energy hydroge n storage system requirements . Storage Parameter Units 2007 2010 2015 System Gravimetric Capacity kg H2/kg system 0.045 0.06 0.09 System Volumetric Capacity kg H2/L system 0.036 0.045 0.081 Storage system cost $/kg H2 200 133 67 Min/max delivery temperature oC -30/85 -40/85 -40/85 Cycle life Cycles 500 1000 1500 Min/max delivery pressure from tank atm (abs) 8/100 4/100 3/100 System fill time (5kg) min 10 3 2.5 These specific targets were created to en able the use of hydrogen as a fuel for automotive applications by replacing an av erage gasoline tank with a hydrogen storage system. A gravimetric density of 6wt.% of hydr ogen (0.06 kg H2/kg system) is based on a driving range of approximately 300 miles w ith 6kg of hydrogen. Of course, the actual fuel economy of the hydrogen powered vehicle will vary based on whether it utilizes an internal combustion engine or a fuel cell to power the vehicle as well as the weight of the car, just like a gasoline powered vehicle. Si nce the hydrogen storage system includes the container that is used to hold the hydrogen storage materi al, the actual gravimetric density of the material should be at least 2wt.% higher. This means that the 2010 target really is considered to be 8wt.% of hydrogen for th e hydrogen storage material. For
6 comparison, a typical gasoline tank onboard a car currently weighs approximately 74kg, of which 55.4kg is gasoline and th e rest the act ual canister. Another important criterion for a practical hydrogen storage material is that the hydrogen should be absorbed and released between -40oC and 85oC. More specifically, this means that the temperature of the hydroge n should be within this range, as these are the practical temperature ranges for the cu rrent state of the art fuel cells. In all practicality, this means that a higher temp erature can be used to release the hydrogen from its material, as long as the hydrogen is then cooled to within the previously mentioned temperature range. When using a hy drogen fuel cell, howev er, it is important to maintain the temperature of the fuel cell above 0oC since the water that is produced as a by-product would otherwise freeze and destr oy the membrane within the fuel cell. The refueling time of hydrogen into th e system is equally as important. Approximately 5kg of hydrogen should be absorbed by the storage system in less than 5 minutes. It is estimated that approximately 5 to 13kg of hydrogen is required to power light duty vehicles. The specific system fill time listed in Table 1.1 is for 5kg, which means that the time to fill the system w ith more hydrogen would accordingly be scaled up. While there are other targets that need to be met, the combination of the gravimetric density, hydrogen delivery temperature, and fueling time are the most difficult to combine for a hydrogen storage sy stem. There has been success in meeting one or even two of these criteria in a hydrogen storage system, but so far no success in meeting all three of these. The most importa nt developments are listed in Section 1.3.
7 1.3. Current State of Hydrogen Storage Hydrogen storage by itself is nothing new. In fact, hydrogen does not occur naturally as an element or molecule, but in stead is always bound chemically to another material. These chemical compounds can be ga ses such as methane, liquids such as water, or solids such as metal hydrides. Of course, hydrogen can easily be separated from its chemically bonded counterpart s and stored in molecular or atomic form. NASA has been using liquid hydrogen, for example, to propel its shuttles into space for decades now. Hydrogen can be stored by four main met hods as a compressed gas, as a liquid, or via physisorption and chemisor ption, the latter two of whic h are considered chemical storage. These four main methods of storag e are discussed in fu rther detail in the following subsections. Table 1.2 gives an overv iew of some of the most commonly used methods for hydrogen. Table 1.2: Comparison of some hydrogen storage methods  Storage method Gravimetric density (wt.%) Volumetric density (kg H2/m3) Temp. (oC) Press. (bar) Phenomena and remarks Compressed 13 <40 Room Temp. 3001000 Compressed gas in light weight composite cylinders Liquid Size dep. 70.8 -252 1 Liquid H2, boil off Adsorbed hydrogen ~2 20 -80 100 Physisorption large spec. surface area, reversible Absorbed on interstitial ~2 150 Room Temp. 1 Atomic H2 intercalation in host sites in metal, reversible Complex compounds <18 150 >100 1 High Temp., high Press. Complexes with water <40 >150 Room Temp. 1 Chemical oxidation of metals with water
8 1.3.1. Compressed Hydrogen Gaseous Hydrogen Storage One of the easiest ways of storing hydroge n is in the form of compressed gas. Since hydrogen has a density of approximate ly 0.09g/L, high pressures are required to store enough hydrogen in a reasonable volum e for practical use. Thus, compressed hydrogen is generally stored with a pressure of approximately 300 to 1000bar, which can pose not only engineering design issues, but also safety risks. If hydrogen is stored at a pressure of 800bar at room temperature, th e overall density of hydrogen is approximately 13wt.%, which does meet the DOE safety targ ets, but only has a vol umetric density of less than 40kg H2/m3, which ultimately is less than the 81kg H2/m3 required by the DOE. Nevertheless, there is still a lot of research going on in the development of light weight and strong hydrogen cylinders by utilizing, for ex ample, light weight composite materials with a tensile strength of 2000MPa. Ultimately, though, the relatively low volumetric density combined with the unsafe high pres sure simply does not make compressed hydrogen a viable option for use in automotive applications. Compressed hydrogen might, however, still be used for large scale transportation, such as hydrogen tankers, as well as storage for hydrogen gas stations places where there can be enough safety guaranteed by enclosing the tank to provide added safety. For automotive applications, however, compressed hydrogen simply cannot be considered as a viable option, mainly due to the safety aspects as well as the stri ct restrictions on volume and weight of the storage system.
9 1.3.2. Liquid Hydrogen Storage Hydrogen can, of course, also be stored as a liquid. Since the boiling point of hydrogen is -252.87oC, the hydrogen must be cooled significantly. C ooling the hydrogen of course means that a very large amount of energy is requir ed to do this. If, however, one were to ignore the energy re quired to achieve this, there ar e still several issues with using liquid hydrogen. One of th e main issues is the so-c alled boil off . Boil off occurs when the hydrogen is not uniformly maintained below the boiling point, which can easily occur when the hydrogen is not consta ntly stirred to ensure that there are no hot spots within the liquid storage contai ner. Once part of the hydrogen boils, the hydrogen gas can then easily permeate the cont ainer walls and simply dissipate into the environment. The most advanced liquid stor age tanks available to date still have significant boil off problems. BMW, for exam ple, in its hydrogen powered test fleet, claims that a full tank of liquid hydrogen will be empty within approximately 7 days by simply sitting in the parking lot. In fact, boil off is such an issue, that BMW strongly discourages people from parking the cars inside a garage, as the hydrogen could accumulate and ignite once a source of ignition, as can be created from a light switch when it is turned on, is present. Liquid hydrogen does have a few advantages , when compared with the DOE targets set forth. The pressure at which liquid hydrogen is stored is 1atm, which, in terms of pressure requirements, is an ideal pressu re and also safe. The volumetric density of liquid hydrogen is 70.8kg H2/m3, which almost meets the 2015 DOE targets. The gravimetric density, however, wi ll depend on the size and type of container used to store. The cooling unit as well as any valves and tubing that might be used needs to be taken
10 into account when obtaining the gravimetri c density. Since liquid hydrogen requires constant cooling as well as insulative ma terial, the system weight quickly adds up, thereby greatly decreasing the gravimetric density. 1.3.3. Physisorption Another commonly investigated met hod of storing hydrogen is by using physisorption. Through this process, gas molecu les bond to the surface of the material via weak van der Waals bonds. Therefore, the amount of hydrogen that can be stored in a physisorption material is proportional to the surface area of the material. Theoretically, therefore, any material with a large specifi c surface area can be used to store hydrogen through the weak van der Waals bonds. The ma in class of material that has been investigated for physisorption is carbon si nce it possesses unique properties that are interesting for storing hydrogen. Due to th e abundance of carbon and the simplicity of which one can produce various forms of it, such as carbon nanotubes, carbon nanofibers, nanobells, or grapheme [17, 18], there is virt ually an unlimited amount of materials that can be chosen for hydrogen storage. One of the most important advantages of carbon based materials is that they do not need to be kept in an inert atmosphere and also do not require a high pressure or strong cylinders to store the material. Virtually all carbon materials are environmentally friendly and benign to humans, which allows for a high safety aspect. Some work has been performe d on investigating car bon-based materials for hydrogen storage. Graphite nanofibers, in it s herringbone structure, have been shown to store hydrogen , though the exact structur e required to store the hydrogen is still being investigated. Hydrogen storage in carbon nanotubes , carbon nanobells ,
11 carbon nanofibers  and especially doped carbon nanotubes  have shown repeatable hydrogen storage of as much as 14wt.%, at least th eoretically. Carbon materials can be readily decora ted or modified to allow th e modification of kinetics as well as capacity of these materials for use w ith hydrogen storage. Ca rbon-based materials possess another important feature, namely a hydrogen binding energy between 10 and 50kJ/mol H2. This is an ideal value for the bind ing of hydrogen, since a material with hydrogen bonds less than this is not a suitable candidate material, as the hydrogen can be readily and accidentally released. More sp ecific physisorption background information, as related to this work, is discu ssed at the beginn ing of chapter 3. 1.3.4. Chemisorption Chemisorption of hydrogen is achieved when a material, such as a metal, forms a primary atomic bond with hydrogen, as is th e case for metal hydrides. Metal hydrides, also known as interstitial hydrides, generally are metallic or graphite-like in appearance and generally are good conductors . These materials bind the hydrogen in the form of strong primary chemical bonds. By changing the pressure or temperature of the metal hydrides, hydrogen is generally released. Metal hydrides, though investigated for decades, still do not provide an easy, reliab le method for hydrogen st orage. Almost all metal hydrides require an inert atmosphere from production to disposal to ensure that the material does not react with water or oxyge n, which destroys or at the very least deteriorates the material. Often, a temperature of more than 250oC is required to release the hydrogen. Additionally, the hydrogen sorption kinetics are often very slow with low usable capacity (<2wt.%). Metal hydrides, however, do have a much higher volumetric
12 density than even liquid hydrogen. Mg2FeH6, for example, has a volumetric density of 150kg H2/m3. Since the maximum hydrogen to metal ra tio known is 4.5, as is the case for BaReH9 , it is essentia l to investigate only metal hydrides that consists of elements with an average molecular weight of less th an 51.8g/mol to obtain a gravimetric density that is higher than 8wt.%, as required by the DOE. More specific examples of chemisorption, as it pertains to this di ssertation are discussed in chapter 4. 1.4. Dissertation Outline Chapter 2 discusses the detail s of the experimental part of this dissertation, more specifically, the details of the synthesis apparatuses, as well as the characterization techniques employed. The actual work performe d as part of this dissertation can be separated into two parts the investig ation of physisorption phenomena and the investigation of chemisorption. Polyanilin e, a conducting polymer, was chosen as the material to be investigated for physisorption a nd is described in chap ter 3 both in its bulk form as well as in nanostructured form, more specifically nanofibers and nanospheres. Chapter 4 discusses the inve stigation and optimization of hydrogen performance of complex hydride structures composed of LiBH4, LiNH2, and MgH2 with and without nano sized additives. Chapter 5 finally summarizes the work completed as part of this dissertation and offers some suggestions for future work. 1.5. Dissertation Goals and Significance of the Study As traditional fossil fuel supplies are dwindling and carbon emissions derived from burning these fuels are being blamed fo r global weather changes, it is becoming increasingly important to find alternativ e energy sources. While there are clean and
13 renewable energy production methods, such as wind and solar energy, there is yet to be found a clean and safe means of propelling au tomobiles. Hydrogen is an ideal candidate since it can easily be refueled in automobile s in a manner similar to gasoline. Its onboard storage, however, is a signifi cant barrier for utilizing hydrog en as a fuel. The materials developed as part of this dissertation pr ovide a significant improvement in the hydrogen storage properties of solid state storage. Two main approaches are inve stigated as part of this dissertation. The first is to tailor the nanostructure of polyaniline in its emeraldine form, a conductive polymer, so that the surface area of the polymer is incr eased, by creating nanospheres and nanofibers. The bulk form of polyaniline is also investigat ed for its hydrogen sorption properties as a comparison. The reason behind using a polymer for hydrogen storage is that polymers contain many hydrogen atoms which allow for the formation of weak secondary bonds between the hydrogen that is part of the polymer and the hydrogen that is meant to be stored. Since polyaniline is easily synthesized and is rather inexpensive, it is an excellent choice as a hydrogen st orage material. The alteration of the physical structure of polyaniline into nanospheres and nanofibers allows for an increase in surface area, thereby exposing more material as pot ential bonding sites for the hydrogen. Since polyaniline is composed of quinoid and benz enoid rings and the emeraldine form is terminated with Clions, additional hydrogen bonding sites, in the form of stronger chemisorption of hydrogen, is made avai lable. The advantage of having both chemisorption and physisorption sites is s hown schematically in Figure 1.2. Hydrogen, as a molecule, can bond to the material through th ree main mechanisms. The simplest form is to simply bond weakly to the host material via physisorption as a molecule. This can
14 generally be achieved if the temperature is low enough, and is the main mechanism of storage for physisorption materi als, but generally requires te mperatures of approximately 77K. If additive materials, such as catalysts, are present, these materials can break up the molecule into ions which th en allows for the diffusion th rough the relatively porous material to then chemically bond with the host material. Similarly, hydrogen molecules can also bond the host material. By increasing the surface area as well as the porosity, two main events can occur. The first is that the hydrogen has more bonding sites to the host material. The second is that the diffusion of hydrogen into the material is made possible. Since polyaniline cannot Hostmaterial H atom H 2 molecule p h y sisorbe d chemisorbe d Atomic chemisorbed Catal y st atoms Figure 1.2: Schematic representation of physisorption and chemisorption with catalyst materials.
15 be simply modified to have smaller molecula r size, the nanostructure can be altered to increase the surface area and also to increase the hydrogen diffusion pathways. For the first time, polyaniline is shown to possess outstanding hydrogen storage properties by carefully modifying the nanostr ucture of the material. It is found that polyaniline combines both physisorption as we ll as chemisorption for storing hydrogen, which provides a significant improvement of hy drogen storage properties, since a usable capacity is achieved. The second mechanism investigated as part of this dissertation is the route of chemisorption, or strong hydrogen atomic bonding, to complex hydride materials. Complex hydrides generally require a hi gh temperature for hydrogen release as the hydrogen bonds are very strong. By reducing th e crystallite size of the host material, specifically LiBH4 and LiNH4 and by further destabilizi ng the structure with MgH2, the temperature can be reduced to allow for reversible hydrogen storage at a lower temperature. The reduction of crystallite size, as well as the exact processing technique is shown to be very important and have a large effect on the storage capacity. Figure 1.3 shows general hydrogen ab sorption and desorption of complex hydrides. The unhydrided material absorbs hydroge n from the outside in until it is fully hydrided. By reducing the pressure on the samp le or by increasing the temperature, the hydrogen is then released from the outside first until the material is fully unhydrided again. When the particles are too large, though, a hydrogen passivation layer can form during the initial hydrogen uptake, th ereby reducing any further hydrogenation. Additionally, the kinetics, or rate, of hydrogen so rption is increased with particle size, as this means that the hydrogen has a larger distance to diffuse th rough. The effects of
16 particle size on hydrogen storage are s hown in Figure 1.4. By employing mechano chemical milling, the particle s are not only reduced in size, but dislocations and vacancies are created, thereby increasing the kinetics and capacity of the material. The interaction of the various part icles that make up the hydrogen st orage material are also of great importance, as this dete rmines the interaction of the various compounds and either facilitates or hinders successful hydr ogen sorption as will be shown. The complex hydrides that are develope d are carefully inve stigated employing varying processing techniques, a deviation of the traditional means of producing complex hydrides. The nanostructure create d with the varying techniques is carefully analyzed and Figure 1.3: Hydrogen storage mechanism in comp lex hydrides and the effect of particle size. absorption desorption unhydrided hydrided
17 correlated with the materials hydrogen performance. Addi tionally, the hydrogen storage properties are significantly improved by us ing optimized quantities of nano sized additives. Time H y dro g en Conentratio n large particles small particles Figure 1.4: Particle size e ffects on hydrogen storage.
18 Chapter 2 Experimental Equipment This chapter is intended to give a general overview of the experimental equipment, materials, as well as some basic background information on these, employed within the scope of the dissertation work. More detailed information is presented in the beginning of chapter 3 and chapter 4, as any specifics may relate to the polymer or complex hydride. 2.1. Synthesis and Materials With the exception of the starting materi als, which are listed in Table 2.1, all materials that are described in this dissertati on were synthesized in the laboratory in order to achieve the highest amount of flexibility on the tailoring of the ma terials properties to optimize the hydrogen performance of said materials. Table 2.1: Materials used as part of this dissertation with their respective purity. Material Company Purity Aniline Sigma Aldrich 99% Ammonium Persulfate Sigma Aldrich 99% Hydrochloric Acid Sigma Aldrich 37% Aluminum Sigma Aldrich 98% SnO2 Sigma Aldrich 98% LiBH4 Sigma Aldrich 98% LiNH2 Sigma Aldrich 98% MgH2 Sigma Aldrich 98% Cu QSI 99% Co QSI 99% Ni QSI 99% Mn QSI 99%
19 2.1.1. Glove Box The materials used for the complex hydrides were all stored in an Innovative Technology System One glove box. The glove box, shown in Figure 2.1, is filled with a 99.999% pure nitrogen atmosphere to protec t all starting materials from oxidation and contamination. The atmosphere is constantly monitored and controlled so that the oxygen and moisture levels are constantly kept be low 1ppm by providing a continuous flow of nitrogen. All sample preparation, whenever possible, was performed inside the glove box. This ensures that the hydrides do not react with any moisture or oxygen and also that the materials do not pick up any contamin ation, as might otherwise occur. Figure 2.1: Innovative Technology System One glove box. 2.1.2. Polymerization The polyaniline samples that are investigated as part of this research were synthesized in the laboratory ra ther than purchased. This wa s done to ensure the utmost flexibility in tailoring the characteristics of the polyaniline. The exact polymerization
20 procedure is described in detail in chapter 3. The polymerization was carried out inside of a fume hood, rather than the glove box, sinc e the vapors emitted during polymerization could have damaged the glove box as well as brought the moisture and oxygen levels of the glove box to an unsafe amount, jeopardizing the purity of the materials stored within the glove box. The only experimental equipment used was glass ware and a stirring plate, as well as an electrospi nning setup, described more closely in chapter 3. 2.1.3. Ball Milling The complex hydrides were synthesized using ball milling, also known as mechano chemical synthesis. More specifically, a planetary ball mill, the Fritsch Pulverisette P5, was used for synthesis. The Fritsch Pulverisette P5, shown in Figure 2.2, contains a holder for a stainless steel bowl, which in turn contains 20 smaller stainless steel balls. When the ball mill runs, the bowl rotates in a clockwise direction, while the base of the bowl rotates in the oppos ite direction. The sample to be milled is placed in the stainless steel bowl and then sealed. When the ball mill is running, the planetary action causes centrifugal forces that add and subtr act alternatingly. Th rough these alternating forces, the balls roll halfway around the bowl an d then are thrown across the container. This rotation, combined with translation of the balls, provides for a high impact of the grinding balls with the materi al that is being milled.
21 Figure 2.2: Stainless steel bowl w ith stainless steel balls (lef t), Fritsch Pulverisette P5 (right). This means that ball milling allows not only for physical but also chemical changes of the material that is being processed. The physical aspects include: (1) particle size reduction through crushing (2) physical combination of materials (3) strengthening of pa rticles through grai n size reduction (4) specific surface area increase through crushing of the particles The ball mill also allows for chemical interaction of the materials that are in the stainless steel container. Due to localized temperature increases created by the grinding action of the stainless steel balls with the mate rial that is being milled, the materials that are being milled can form new chemical co mpounds. The stainless st eel container used during synthesis also has an inlet and outlet vent, allowing for the addition of various processing gases, which is essential to the synthesis of the materials.
22 2.2. Thermal Characterization Due to the nature of the ma terials investigated for hydroge n storage, the effect of temperature on the materials characteristics is extremely important. The complex hydrides, for example, use temperature changes as a driving force for release. Therefore, it is extremely important to investigate the effects of temperature increases on the material. One of the first steps after synt hesis of any new sample was therefore to perform thermal characterization. The three main techniques employed were thermogravimetric analysis (TGA), differen tial scanning calorimetry (DSC) and thermal programmed desorption (TPD), which are all discussed in this subsection. 2.2.1. Thermogravimetric Analysis (TGA) Since the materials synthesized as part of this dissertation are all in their hydrided phase, a quick analysis of the material em ploying TGA can be very valuable. The data obtained from this quick screening tool can be used to obtain insight into the thermal stability, e.g. the usable temperature range of the material, as well as any phase changes that might be present such as melting. The TGA employed is a Thermal Analysis SDT (Simultaneous DSC and TGA) Q600, shown in Figure 2.3.
23 Figure 2.3: Thermal Analysis SD T Q600 simultaneous DSC and TGA The Q600 possesses two alumina cantilever b eams with built in thermocouples, as shown in Figure 2.4. This enables the simultaneous measurement of heat flow in a sample as well as weight loss, which might occur due to hydrogen release. An empty alumina beam is placed on one of the cantilever beam s, while the sample is placed in another alumina pan on the other cantilever beam. Be fore loading the sample, both sample and reference pan are zeroed out to obtain the proper weight of sample. Approximately 5mg of sample was loaded into the sample pan just enough to allow for an accurate measurement, but not enough for the sample to overflow and cause damage to the TGA. The TGA was placed in another glove box in order to minimize contact of the sample with the environment in order to obtain the most accurate results. While the furnace of the TGA can go up to 1200oC, none of the samples were ever heated above 500oC for two reasons. The first reason is to ensure that the TGA does not get damaged due to harmful vapor emissions. The other reason is that it would simply be impractical to use a hydrogen storage material at temperatures above 500oC.
24 Figure 2.4: Schematic representation of the most important parts of a TGA. As Figure 2.4 shows, the main parts of the TGA are the two cantilever beams, a tube furnace, and a dual-balance mechanism coupled with photodiodes and sensors that actually measure the deflection of the beams due to weight loss or weight gain. In general, the samples were all analyzed in an argon atmosphere with a heating rate of 1oC/min to allow the sample to reach localized equilibrium and to obtain the most accurate material performance. Argon was chos en over nitrogen since it is an inert gas and since some of the samples that were ch aracterized contained nitrogen, which might cause the sample to react w ith the nitrogen atmosphere, th ereby giving inaccurate results. The TGA was calibrated in regular intervals fo r both heat flow as well as weight loss with standardized samples to provide the utmost accuracy.
25 A typical TGA plot, in this case for calcium oxalate, is shown in Figure 2.5 below. Calcium oxalate is often chosen as a reference material fo r calibrating the TGA since it exhibits three distinct weight loss regions. Typically, a weight loss of approximately 12.3% is observed below 200oC. This is then followed by a weight loss of 18.5% due to CO emission below approximately 500oC. Finally, the calcium oxalate loses approximately 30.3% of CO2 below 800oC. The TGA measurements of the hydrogen storage materials are therefore used to determine the amount of gas released as a function of temperature. The slope of the weight loss curve can give an indirect indication of the rate at which the gas is released by the material, while the amount of weight loss can be directly correlat ed to the amount of gas released. Figure 2.5: Typical TGA profile Calcium oxala te exhibiting weight loss due to water, carbon monoxide and finally carbon dioxide loss.
26 2.2.2. Differential Scanning Calorimetry (DSC) Even though the TGA contains a built in DSC, several samples were characterized using Thermal Analysis DSC Q10, shown in Figure 2.6, for more accurate measurements. Similar to the TGA, two pans are used one is an empty aluminum reference pan, the other is an aluminum pa n with approximately 5mg of sample. The DSC then heats the sample and reference with a heating rate of approximately 1oC/min in a helium atmosphere. The DSC instrument ensure s that both the sample pan as well as the reference pan is kept at the same temperature. If there is a temperature difference, due to an endoor exothermic reacti on, the instrument adds more en ergy to the cool er of the two pans and records this difference in heat flow. Figure 2.6: Thermal Analysis DSC Q10. A typical DSC profile is shown in Figure 2.7. It can be seen that there are four distinct thermal events that the material e xperiences while it is being heated. There are two endoand two exothermic events. The glas s transition, a phase change essentially, as well as the melting of the material are both en dothermic events. The recrystallization and
27 decomposition of the material, however, are exothermic events. DSC analysis, combined with TGA analysis of the hydrogen storage mate rials is very important since it provides information as to the amount of energy requi red to release the hydr ogen. Unfortunately, neither the TGA nor the DSC is capable of measuring the thermal performance of the material under high pressures of hydrogen. Theref ore, it is not possible to determine the amount of energy required for re hydrogenation of the material. 2.2.3. Thermal Programmed Desorption (TPD) The third and final thermal analysis of the hydrogen storage materials was performed using thermal programmed desorptio n, TPD. Quantachromes Autosorb 1 was used with the optional TCD (thermal conduc tivity detector) attachment, as shown in Figure 2.8. TPD works by heating the sample to be investigated within a u-shaped quartz exothermic endothermic Heat flow Temperature Glass Transition Recrystallization Melting Decomposition Figure 2.7: Typical DSC plot showing endothermic glass tr ansition and melting as well as exothermic recrystallization and decomposition.
28 tube at a controlled rate generally 1oC/min was used for the work described in this dissertation. The mass of the sample is measured and then the sample is loaded inside the glove box. The quartz tube is then sealed with Parafilm to prevent the sample from picking up moisture or oxygen. Once the quartz tube is installed in the TPD, the measurement is started. This includes the init ial purging of any residua l gas from the tube using helium for 20 minutes. After this, the sample is heated to 50oC and purged with nitrogen for another 30 minutes. After the purge, the sample is h eated at a constant rate of 1oC/min up to a temperature of 350oC. Nitrogen was used as the purge gas since hydrogen and helium have similar thermal conductivities. Figure 2.8: Thermal programmed desorption (TPD) with optional TCD controller.
29 The TPD measurements are used as a quick screening tool to obtain information about the temperature at which hydrogen, or any other gas, is released. The data can be used to verify any TGA or DSC data previ ously obtained. This provides an accurate and quick screening method for the hydrogen de sorption properties of the investigated materials. Additionally, TPD data is collected at varying heating rates in order to obtain information about the activation energy asso ciated with hydrogen release by employing Kissingers method . A TPD plot showing the three st ep release of hydrogen of LiAlH4 is shown in Figure 2.9. It is clear that there are three distinct peak s obtained at very distinct temperatures. This TPD data is easily confir med either via literatu re or through other thermal characterization techniques such as TGA or DSC. Generally, the data is plotted as TCD signal versus temperature since TPD measurements are performed with a constant temperature ramping rate. One can, however, plot signal versus time to obtain some basic information about the kinetics of hydrogen release.
30 2.3. Structural Characterization In addition to detailed thermal studies of the materials, the structural characteristics are of utmost importance. It is essential, for example, to confirm that the polymerization of polyaniline is indeed successful. The main tool employed for this is Fourier transform infrared spectroscopy (FTIR) Additionally, the micr ostructure of the materials for hydrogen storage is extremely important. As will be demonstrated, the surface area as well as the crystallite size of the material has a large effect on the hydrogen storage properties. In order to char acterize this, scanning electron microscopy (SEM) is employed. Additionally, it is essen tial to know the nature of the chemical structure of the material after ball m illing, which is determined employing x-ray Figure 2.9: Typical TPD pl ot showing three-step hydrogen release of LiAlH4.
31 diffraction (XRD). XRD allows for the confir mation of any chemical reaction that might occur during ball milling. 2.3.1. X-Ray Diffraction (XRD) The chemical structure of the processe d complex hydrides was analyzed using a Philips XPert XRD system, as shown in Fi gure 2.10. The polymer samples synthesized were not analyzed using the XRD since XRD can only be used to characterize and identify crystalline materials. Figure 2.10: Philips X'Pert XRD used for characterizing the chemical structure of processed complex hydrides. William Henry Bragg and William Lawrence Bragg won a Nobel Prize in 1915 after deriving a relationship between x-rays and crystal structure. This relationship is still used today and is known as Braggs law. Br aggs law gives a relationship between the angle of the incident x-rays, their wavelength, and the spacing between planes in an
32 atomic crystal lattice, d, as well as an intege r determined by the order of the x-rays, n, as described in Equation 2.1: n = 2d sin (2.1) The principle of XRD is rather simple and ingenious. One takes a crystalline material and uses an x-ray source to provide a concentrated beam of x-rays, which then hit the surface at a cert ain incident angle, This angle is then increased at a fixed rate while a receptor plate with an x-ray sensor is moved around to record the intensity of the emitted x-rays. When the x-rays do not hit atom s of several planes at the same time, the x-rays are out phase, causing a base line of th e spectrum, as shown in Figure 2.11 (a). If, however, the x-rays hit atoms of several plan es at the same time, the resulting x-rays are in phase and cause a spike in the x-ra y signal, as shown in Figure 2.11 (b). Figure 2.11: (a) X-rays not in phase causing a base line of XRD signal, (b) x-rays in phase causing a peak in XRD signal. Therefore, it is easy to see why only crystalline materials, with a periodic arrangement of atoms, can be successfu lly investigated using XRD. Amorphous materials, with randomly arranged atoms, will always cause an annihilation of the x-ray signal, causing this technique to be useless.
33 2.3.2. Fourier Transform Infrared Spectroscopy (FTIR) Since the polymer samples that were synt hesized were of amorphous nature, FTIR was employed to confirm that polyaniline actually was synthesized and to obtain information about its chemical nature, such as the presence of benzenoid and quinoid rings. A Perkin Elmer Spectrum One, as show n in Figure 2.12, was used to measure the FTIR spectra of the various samples. Si nce the polymer sample s do not require any special consideration when performing measur ements, as the complex hydrides require, all polyaniline samples were simply measur ed, after performing background scans, using the ATR powder accessory, which is also s hown installed in Figure 2.12. Some of the complex hydrides were also characterized using the FTIR in order to determine the presence of BHions, for example. Since the comple x hydrides react with moisture and oxygen, the samples were pelletized after mixing with KBr powder and wedged between KBr windows to prevent contam ination of the samples. Figure 2.12: Perkin Elmer Spectrum One FTIR.
34 2.3.3. Scanning Electron Microscopy (SEM) The surface morphology of both the polyani line and complex hydride samples are of utmost importance to this research. Theref ore, the materials were characterized using the Hitachi S800, shown in Figure 2.13. The polyaniline samples were simply placed on the sample holder using carbon tape and it wa s ensured that no excess material was loose, which might contaminate the SEM. The complex hydrides require more attention since these materials can react with moisture and oxygen. Therefore, all complex hydride samples were prepared inside the glove box and secured to the sa mple holder using carbon tape. The sample holder was then placed in a desiccator and transferred to the SEM. The only time the sample was exposed to the atmosphere was when loading the sample inside the SEM. The exposure was limited to a few seconds, which caused only minimal contamination. The main feature th at was explored was the surface morphology. Figure 2.13: Hitachi S800 SEM.
35 2.4. Hydrogen Sorption Measurements (PCT) The most important charac terization was the measurem ent of hydrogen sorption using a HyEnergy PCT Pro 2000, shown in Fi gure 2.14. The PCT Pro 2000 consists of a sample holder and a heater jacket for the sample holder along with carefully calibrated reservoirs for hydrogen. The equipment is capable of measuring hydrogen release and uptake from room temperature to 500oC with hydrogen pressure varying from vacuum to 200bar, though, for safety reasons, the highes t pressure used was approximately 80bar. The PCT Pro 2000 is capable of performing tw o different types of hydrogen sorption: (1) So-called PCT measurements, where the hydr ogen is incrementally increased for absorption runs or decreased for desorption runs. (2) Kinetics measurements where hydrogen absorption or desorption is measured with a constant pressure. Figure 2.14: HyEnergy's PCT Pro 2000 used for hydrogen sorption PCT and kinetics measurements.
36 Most measurements are performed unde r isothermal conditions, though some of the measurements were run while increasing the temperature at a constant rate. The samples that were measured for their hydr ogen performance gene rally had a mass of approximately 1 to 2g. The complex hydrides we re loaded inside th e glove box, while the polyaniline samples were loaded into the samp le holder outside of the glove box. Before every measurement, the sample holder was purged with helium several times to remove any air or nitrogen that might be present. After the purging process, the volume of the sample was calibrated using a built in procedur e. The accuracy of this calibration is of extreme importance since the equipment directly measures pressure and pressure drop. The change in pressure and the volume of the hydrogen was then correlated to the amount of hydrogen the material either abso rbed or desorbed using the van der Waals equation, a modification of the ideal gas law. The van der Waals equation can be applied to fluids that is composed of particles wi th a non-zero size and th at have a pairwise attractive force between part icles. The equation is described by Equation 2.2: p+n2a V2 V-nb =nRT (2.2) where p is the hydrogen pressure (measured directly), V is the volume that the hydrogen has available to expand into (obtained from the volume calibra tion), R is the gas constant (8.314J/(mol K)), a is the dipole interaction or repulsion constant (2.476x10-2m6Pa/mol2), b is the volume occupied by the hydrogen molecules (2.661x105 m3/mol), n is the number of moles of hydrogen, and T is the temperat ure of the sample during measurement in Kelvin. The temperature and the pressure are measured directly by the PCT Pro 2000. Since a, b, and R are constants, the measured data can be used to obtain the number of
37 moles of hydrogen that was absorbed or desorb ed by the material. Multiplying this with the molecular weight of hydrogen (1.0079g/mol) yields the equivalent mass of hydrogen that was sorbed. This can then be used al ong with the mass of the sample to obtain the amount of hydrogen in terms of weight pe rcent (wt.%) as indicated by Equation 2.3: H2 concentration wt.% = mhydrogenmhydrogen+msample x100 (2.3) where mhydrogen is the mass of the hydroge n that is sorbed and msample is the mass of the unhydrided material that sorbs the hydrogen. As previously mentioned, the two main hydrogen sorption measurements performed on the samples are PCT and kineti cs measurements. The PCT measurement is performed at a constant temperature with varying pressure. A basic absorption PCT is conducted by starting out with a low hydrogen pressure (0 to 1bar) on the sample. A precalibrated reservoir is then filled w ith the same amount of hydrogen and the measurement is started. When the valve that separates the sample from the reservoir is opened, the change in pressure is recorded. From the gas law it is known what the equilibrium pressure should be. Any variation of the pressure is therefore due to the interaction of hydrogen with the material. If hydrogen is absorbed by the material, the pressure drops; if it is desorbed, or release d, by the material, the pressure increases. After pressure equilibrium is reached, the valve betw een the reservoir and the sample is closed and the reservoir is filled with a certain amount of hydrogen either more for absorption or less for desorption. The valve is then opened again and the previous procedure repeated. This is done with a variation of hydrogen pressure from 0 to 80bar for absorption or from 80 to 0bar for desorption in either 5 or 3bar increments. The data that
38 is collected is then plotted as hydrogen pr essure versus hydrogen concentration. Some materials exhibit a so-called plateau pressure a pressure at which an optimal amount of hydrogen is released or absorbed. An ex ample of a desorption PCT measurement exhibiting a plateau pressure around 15bar is shown in Figure 2.15. Figure 2.15: Desorption PCT plot of LiAlH4 showing plateau pressure around 15bar. The other main hydrogen sorption measurement performed is a kinetics measurement. For an absorption measurement, an example of which is shown in Figure 2.16, the sample is first either evacuat ed or charged with 1bar of hydrogen. A precalibrated reservoir is then filled with 80bar of hydrogen. When the valve between the reservoir and the sample is opened, the hydr ogen immediately reac hes an equilibrium pressure, which is determined automatically using the gas law. If hydrogen is absorbed,
39 the pressure within the sample holder drops until it finally reaches equilibrium. This pressure data is then used to obtain the amount of hydrogen that the material absorbed. A desorption kinetics measurement is performed the same way, except that the sample is first placed under 80bar of hydrogen pressure and the reservoir evacuated. If hydrogen is released, the pressure increases above the expected equilibriu m and can then be related to the amount of hydrogen desorbed. Kinetics m easurements are generally performed at constant temperatures to obtain information about the rate of hydrogen sorption. However, kinetics measurements can also be performed with increasing or decreasing temperature to obtain a relationship between the temperature and the amount of hydrogen sorbed. Figure 2.16: Kinetics plot of LiAlH4 showing absorption kinetics.
40 Chapter 3 Physisorption in Polyaniline 3.1. Introduction As mentioned in chapter 1, two main hydrogen storage approaches are investigated. This chapter i nvestigates hydrogen storage vi a physisorption in polyaniline (PANI). PANI is a conducting polymer that has been well characterized. Its chemical structure is shown in Figure 3.1. The emeraldine base form of PANI consists of quinoid and benzenoid rings. This means that the theo retical hydrogen capacity of polyaniline is approximately 6wt.%. While this is too lo w to exceed the DOE targets mentioned in chapter 1, it is not the chemisorption aspect of PANI that is of interest, but instead the physisorption aspect. Figure 3.1: Chemical structure of emeraldine base polyaniline. To date, there has been very little res earch into hydrogen stor age using polymers. Since polyaniline is rather simple to synthesize in a laboratory setting, while being relatively inexpensive, it was deci ded to investigate polyaniline in its bulk form as well as in nanostructured form with various additi ves for hydrogen storage. This nanocomposite material consists of a polyaniline matrix, wh ich can be functionalized by either catalytic doping or incorporation of nano variants. It was reported that polya niline could store as
41 much as 6 to 8wt.% of hydrogen , which, ho wever, another team of scientists could not reproduce . Yet another recent study revealed that a hydrogen uptake of 1.4 to 1.7wt.% hydrogen has been reported for pol ymers of intrinsic microscopy . Most hydrogen storage invest igations of polymers that have been conducted were performed at 77K and only involved pure physisorption, mainly by increasing the surface area of the polymer. One form of hydrogen st orage in polymers is to use a polymeric foam for hydrostatic pressure retainment of hydrogen . In this method of storing hydrogen, the ideal tank contains spherical cells that act as microscopic pressure vessels, ideally in a homogeneous manner. However, no significant amount of hydrogen has been experimentally proved to be stored. As of the time of writing this dissertation, only theoretical work has been performed on hydrogen storage in polymer foams. Hypercrosslinked polymers, such as styrenic polymer produced from polyvinylbenzylchloride, have also been investigated for th eir hydrogen sorption properties . It was found that approxima tely 3wt.% hydrogen c ould be stored at 15bar, although a temperature of 77K was re quired to ensure that the weak hydrogen bonds created through physisorption did not break, thereby rele asing the hydrogen. Polychloromethylstyrene-co-divinylbenzene , another hypercrosslinked polymer, was shown to reversibly store 1.6wt.% hydrogen at a temperature of 77K and a pressure of 12bar. Again, the surface area of the material played an important part in storing the hydrogen, as is expected for physisorption materials. The hydrogen storage mechanism was purely physisorption and lacked any chemis orption of hydrogen, an attribute that is needed to get away from the low temperatures required for pure physisorption materials.
42 Another polymer class that has been investigated for its hydrogen storage properties are polymers of intrinsic micr oporosity, known as PIMs. PIMs are polymers that have very large surface areas, generally in the range of 500 to 900m2/g and are generally prepared by using a benzodioxa ne formation reaction between suitable monomers. It has been shown, for example, that HATN-network-PIM can store about 1.6wt.% hydrogen at a pressure of 10bar and a temperature of 77K [28, 32]. While this value is higher than that of metal organic fr ameworks, it is still to o low for practical use and of course requires a temperature which is not only impractical but also requires a large amount of energy. Polyaniline is a conductive polymer, with conductivity on the order of 100S/cm. This is higher than that of typical non-conducting polymers, but much lower than that of metals . In addition to it s conductivity, po lyaniline in its emeral dine base (EB) form is very simple and inexpensive to polymerize. It is because of this simplicity that it was chosen as a matrix material for the nanocom posite structure discussed in this paper. Conducting polymer nanostructures combine the advantages of organic conductors and low dimensional systems havi ng interesting physicoc hemical properties [33-36] and useful applications [37-39]. Am ong the conducting polymers, polyaniline was considered important because of its extrao rdinary properties of electrical and optical behavior. It was recently reported that polyanili ne could store as much as 6 to 8wt.% of hydrogen , which was later refuted by Pa nella et al. . Polypyrrole, another polymer investigated for hydrogen storage had a similar fate, in that it was cl aimed to be able to store as much as 10wt.% at room temperature , though this was also refuted . Though many controversial results were reported in terms of hydrogen uptake [40-
43 43] in polymer nanocomposites, there are st ill a number of parameters, tailor-made properties, surface morphologies and their corr elation with hydrogen sorption behavior to be investigated before these materials can be commercially deployed for onboard hydrogen storage. Similarly, nanotubes [20, 44], nanofibers  and nanospheres  have attracted more interest because of their novel properties and wide potential application for nanometer-scale engineerin g applications. It is known that the nanofibrallar morphology signi ficantly improves the performance of polyaniline in many conventional applications involving polymer inte ractions with its e nvironment . This leads to faster and more responsive chemi cal sensors [39, 48], new organic/polyaniline nanocomposites  and ultra-fast non-volatile memory devices . A graphical comparison of polymer structur es investigated to date is shown in Figure 3.2. It can be seen that there has not been a lot of work performed on polymeric hydrogen storage, and that the wo rk that has been reported to date mostly is performed at low temperatures of 77K. The polypyrrole an d polyaniline with high capacity that are shown in that figure were later refuted by other teams of scien tists, as previously mentioned. The polyaniline structures invest igated in the following sections are all evaluated for their hydr ogen performance between room temperature and 125oC, which correlates approximately with the temperatur e range required by the DOE for practical hydrogen storage.
44 Figure 3.2: Previous work on hydrogen storag e in polymers, mainly in polymers of intrinsic microscopy (PIM). 3.2. Bulk Polyaniline The first form of the emeraldine base polyaniline (PANI) to be investigated was bulk polyaniline. Bulk simply refers to the nanostructure of PANI, indicating that the PANI is used as-synthesized in its non nanostructured form. This means that the surface morphology was not adjusted into nanofibrous form. After some baseline characterization and hydrogen sorption experiments, filler mate rials could be added to the PANI. The addition of filler materials to the matrix ma terial could be used to tailor an ideal nanocomposite material for hydrogen storage. Se veral different filler materials in various concentrations were added to the nanocomposite and its effect on the characteristics of
45 the material determined. A schematic represen tation of the use of filler materials and the mechanism of hydrogen storage for bulk polyaniline can be seen in Figure 3.3. Figure 3.3: Hydrogen sorption mechanism of st andard polyaniline composite material. Since polyaniline is composed primar ily of hydrogen and carbon atoms, it possesses a large number of both primary a nd secondary hydrogen bonding sites. Figure 3.3 shows a combination of both primary and secondary bonding sites as follows: (1) Hydrogen gas enters the hydrogen storage tank containing the polyaniline matrix material as hydrogen molecules in gaseous form. (2) Any added materials, such as SnO, then react with hydrogen molecule to break it up into hydrogen ions to facilita te the bonding to the host matrix. Conducting Polymer Composite Catalyst Hydrogen ion Hydrogen molecule Carbon nanotube Fullerene Bond H2input H2output 1 3a 2 3b 3c 4 5
46 (3) The hydrogen ions then bond to either f iller materials such as 3a) carbon nanotubes, 3b) fullerenes, or simply to 3c) the host material through either primary atomic bonds, where and impurities or vacancies are pres ent or to other primarily bonded hydrogen through weak secondary hydrogen bonding. (4) By reducing the hydrogen pressure on the material or heating up the host matrix, the hydrogen is then released from storage. (5) Finally, after recombining to hydrogen molecules, the hydrogen is used in either an internal combustion engi ne or in a fuel cell. 3.2.1. Synthesis of Bulk Polyaniline The polyaniline used for this experime nt was synthesized by following a well established method for the synthesis of the emeraldine base (EB) form of polyaniline . 5.71g of ammonium persul fate were dissolved in a 100m L beaker with 50mL of deionized water. 5mL of aniline were mixed with approximately 35mL of de-ionized water. These two solutions were then slowly mixe d with 10mL of 37% HCl acid in a large beaker. After the temperature rises to approximately 45oC, the polymerization is considered finished. The sample, however, is magnetically stirred for approximately 24 hours to ensure complete polymerization. After magnetic stirring, the sample was vacuum filtered through an 11 m filter, after which it was washed twice with HCl to ensure termination of any bonds and methanol. The sample was then vacuum dried at 100oC for approximately 24 hours. While most samples were polymerized at room temperature, which was approximately 23oC, a few samples, which will be discussed separately, were polymeriz ed at approximately 0oC by placing the chemicals in a freezer
47 for several hours before polymerization and then mixing the chemicals inside an ice bath cooled beaker. The so-called filler materials were added during the polymerization process, so as to ensure uniform distribution within the samp le. Some of the filler materials added to the nanocomposite material included SnO2 (obtained from Sigma Aldrich in nanopowder form), multi wall carbon nanotubes (with a purity of at least 60% before purification), as well as aluminum powder (obtained from Ried el-de Han as fine powder with a purity greater than 93%). These materials were adde d in various concentrations, which will be discussed in the individual results section. All materials were adde d in terms of weight percent based on the expected yield. 3.2.2. Bulk Polyaniline Characterization The as-synthesized nanocomposite samples were characterized before and after hydrogen sorption employing Fourier tran sform infrared spectroscopy (FTIR), differential scanning calorimetry (DSC), ther mogravimetric analysis (TGA), as well as scanning electron microscopy (SEM). FTIR, which was performed on a Perkin Elmer Spectrum One, was utilized to ensure the proper polymerization of the sample as well as determine any chemical effect a filler materi al might have had. TGA was carried out on a TA SDT Q600, while DSC measurements were made with a TA DSC Q10P. TGA was employed mainly to determine the useful temperature range of the sample material, as well as to determine any hydrogen that might have bonded to the material chemically. The DSC measurement was used to determine any phase transitions that might occur as a result of the heat ramping process. Fina lly, the SEM was utilized to determine any
48 surface morphological changes of the materials, such as the formation of clusters as a result of hydrogen sorption. As described, the polyaniline based nanocomposite materials were synthesized according to a standardized method. Initially, so lely polyaniline in its emeraldine base form was synthesized. This sample is referred to as PANI-STD. It wa s decided to include multiwall carbon nanotubes into the polyanilin e matrix material, as these have been shown to increase the conductivity of the conducting polymer  This sample is referred to as PANI-MWCNT. Due to the availa bility and relatively low cost of multiwall carbon nanotubes, as compared to single wa ll carbon nanotubes, it was decided to only use multiwall carbon nanotubes, though there has been significant research on hydrogen storage in single wall carbon nanotubes [ 20]. The hydrogen sorption measurements for carbon nanotubes, however, are usually performed at a temp erature of 77K . 10wt.% tin oxide was also added to the matr ix material, since it was expected to act as a catalyst material in the capture of hydrogen, breaking the hydrogen molecule into ions. This sample is referred to as PANI-SnO2 throughout this paper. Finally, fine aluminum powder was added to the matrix material during polymerization to determine any effect of the aluminum on hydrogen sorption. This sample is referred to as PANI-Al. It was thought that the aluminum would fo rm alane or alanate, thus increasing the hydrogen sorption of the nanocomposite. All samples were prepared with varying concentrations of the filler materials. Therefore, the samples will be denoted not only by their abbreviation, but also la beled with the appr opriate concentration of the material added to the nanocomposite in weight percent.
49 3.2.3. FTIR Characterization Results FTIR spectrum analysis was performed to identify the bondi ng environment of the as-synthesized materials, and to determine whether the filler materials chemically reacted with the polyaniline matrix material, or simply physically mixed with the polyaniline. Figure 3.4 shows a comparison of standard pol yaniline, polya niline with multiwall carbon nanotubes, polyaniline with SnO2, and finally polyaniline with aluminum powder. Figure 3.4: FTIR comparing st andard polyaniline (PANI-STD), polyaniline with 10wt.% SnO2, polyaniline with 10wt.% MWCNT and polyaniline with 30wt.% aluminum powder. All four samples exhibit the same be nzenoid and quinoid ri ng vibrations around 1500cm-1 and 1600cm-1, respectively. Figure 3.4 shows that the polyaniline that was synthesized is in fact polyan iline in its emeraldine base . Additionally, all samples
50 exhibit peaks around 1306cm-1, which is attributed to the -electron delocalization induced in the polymer by protonation. The peak around 1375cm-1 is attributed to the CN stretch in the base. The peak around 822cm-1 is due to C-H bond out of plane . It can also be seen that the addition of tin oxide mainly caused a phys ical rather than a chemical reaction between the polyaniline an d the tin oxide, as the FTIR spectrum is essentially unchanged. The addition of mu ltiwall carbon nanotubes to the polyaniline matrix material also shows more of a physical mixing rather than a chemical mixing. This explains the results that are obtained in the hydrogen sorption measurements, to be discussed in the next section. The addition of aluminum to the sample, however, clearly shows a chemical reaction between the polya niline and the aluminum powder. The large peak around 1150cm-1 is a sign of increased conductiv ity and was described by Quillard et al.  as the electron-like band. 3.2.4. Thermogravimetric Results Once the successful polymerization of the sample was confirmed, a TGA measurement was performed on the samples. The sample was heated from room temperature up to 400oC. The main purpose for perfor ming the TG analysis was to determine the maximum temperature possibl e for hydrogen sorption measurements, as well as to confirm the useful thermal range of the samples. It can be seen in Figure 3.5 that all samples are stable up to at least 200oC. The initial weight loss in the samples at around 100oC is simply due to moisture contained within the nanocomposites. It is interesting to note that the sample with the multiwall carbon nanotubes exhibits a highe r weight loss than the other samples. This could be
51 explained by the fact that the multiwall ca rbon nanotubes were not of 100% purity, but instead contained soot or other forms of carbon, which were burned off during the temperature increase. Figure 3.5: Thermogravimetric analysis of st andard polyaniline, polyaniline with 10wt.% multiwall carbon nanotubes, polyaniline with 10wt.% tin oxide, and polyaniline after hydrogen sorption measurements. While care was taken to purify the carbon na notubes as much as possible before synthesis of the nanocomposite material by using a combination of ultrasonication, washing with de-ionized water, drying in ai r and washing with HCl , some impurities must have remained in the sample, causi ng the observed drop. Additionally, the PANIMWCNT sample contained a higher amount of mo isture, as it was ob served that this
52 sample readily absorbed moisture, as the sample were handled and stored in atmospheric conditions. 3.2.5. Scanning Electron Microscopy Results Figure 3.6 shows some SEM images of st andard polyaniline before and after hydrogen sorption measurements. Figure 3.6 (a) a nd (b) clearly show the rough surface of the standard polyaniline. The images show a ro ck like surface structure, with sharp edges. Figure 3.6 (c) and (d), on the other hand, reveal that hydrogen interacted with the polyaniline during the PCT hydrogen sorpti on measurements. The interaction of hydrogen with the polyaniline caused a type of ballooning effect, in which the hydrogen seeped into the material and was stored in small pockets. This indi cates that the hydrogen is stored physically, rather than chemically.
53 Figure 3.6: Scanning electron microscopy imag es of standard polyaniline ((a) and (b)) before and ((c) and (d)) after hydrogen in teraction indi cating a ballooning caused by hydrogen cycling. 3.2.6. Hydrogen Sorption Measurements The isothermal volumetric measurements were carried out by HyEnergys PCT Pro 2000 sorption equipment. This fully automated Sieverts type instrument uses an internal PID controlled pressu re regulator with maximum pr essure of 170bar. It also includes five built-in and calibrate d reservoir volumes of 4.66, 11.61, 160.11, 1021.30 and 1169.80mL. The volume calibrati on without and with the sa mple was performed at a constant temperature with an accuracy of 1C using helium. The software subroutines for hydrogen purging cycles, leak test, kinetics, PCT and cycling etc. were performed by
54 the HyDataV2.1 program. The data collected for each run was analyzed using the Igor Pro 5.03 program with a built in HyAnalysis Macro. The amount of material loaded for hydrogen sorption measurements corresponds to approximately 2g. Before the experiment was run, the volume of the sample was determined at the appropriate temperature. The absorption measurement was generally performed at 60bar of hydr ogen pressure at temper atures varying from 25oC up to 125oC. Desorption measurements have been carried out under vacuum. Cho et al.  had reported hydrogen sorption of polyaniline at approximately 77K, which was shown to not be reproducible [2 7]. This low temperature, however, is not feasible for use in mobile applications. It would simply require too much energy to cool and maintain the hydrogen storage material at such a low temperatur es. It was therefore decided to conduct the hydroge n sorption measurements between ambient temperature and 125oC. The absorption of hydrogen was carri ed out at 60bar hydrogen pressure, while the sample was initially just below am bient pressure. For desorption measurements, the sample was held at the pressure resultin g from the absorption measurement, while the reservoir, into which the hydroge n was to flow, was evacuated. Figure 3.7 shows the hydroge n sorption measurements of standard polyaniline. The pressure was raised in aliquots fr om 0bar hydrogen pressu re to 60bar hydrogen pressure at different temperat ures. It can be seen that th e standard polyaniline does not absorb any hydrogen at r oom temperature. At 50oC, however, the sample can store approximately 0.075wt.% hydrogen at 60bar. As th e temperature is increased, the sample can store more hydrogen. At 125oC, the standard polyaniline, as synthesized, is capable of storing 0.35wt.% hydrogen.
55 Figure 3.7: Hydrogen sorption measurements of standard polyaniline at different temperatures. When tin oxide was added to the polyani line, no change in the hydrogen sorption capabilities was observed, regardle ss of the amount of tin oxide as seen in Figure 3.8. It is thought that the lack of hydrogen sorpti on increase is due to the measurement temperatures chosen. From hydrogen gas sensin g research, it is know n that a temperature of around 300oC is optimum for hydrogen to bind with SnO2 . The measurements followed the same hydrogen sorption behavior as observed for the standard PANI sample seen in Figure 3.7. An increas e in pressure and temperat ure meant an increase in hydrogen sorption, provided the temper ature of sorption was above 30oC. Regardless of
56 the amount of SnO2 added to the polyaniline matrix, there was no cha nge in hydrogen sorption, hence only the 10wt.% samp le has been presented here. Figure 3.8: Hydrogen sorption measurements of polyaniline with 10wt.% SnO2 at different temperatures and increasing pressure. Similarly, the amount of multiwall carbon nanotubes showed no effect on the hydrogen sorption capabili ties of the material. Figure 3.9 shows the hydrogen sorption results of the polyaniline sample with 10wt.% multiwall carbon nanotubes. It can be seen that the same trend is evident a higher temperature and a higher pressure means more hydrogen can be stored. The mu ltiwall carbon nanotubes cause onl y a very slight increase in the hydrogen sorption capabilities, which can be accredited to an increase in the
57 porosity caused by the inclusio n of nanotubes. Again, a mini mum temperature of just above 30oC is needed for the materi al to absorb hydrogen. Figure 3.9: Hydrogen sorption measurements of polyaniline with 10wt.% multiwall carbon nanotubes at different temper atures and increasing pressure. Figure 3.10 shows the hydrogen sorptio n measurements performed on the polyaniline sample with 30wt.% aluminum powder. It can be seen that this nanocomposite material exhibits a much highe r hydrogen sorption than the other samples. It is interesting to note that hydrogen sorption begins at 75oC, above approximately 50bar hydrogen pressure, as compared to 30oC for most other samples. The exact effect of the aluminum on the hydrogen sorption performanc e of the nanocomposite material is still
58 unclear, as well as the effect of different conc entrations of aluminum. It is however clear that no hydrogen alanate is form ed from this interaction. Figure 3.10: Hydrogen sorption measurements of polyaniline with 30wt.% fine aluminum powder at different temperatur es and increasing pressure. 3.2.7. Bulk PANI Summary All experiments were conducted to mimic th e U.S. Department of Energys goals as closely as possible. Hence, it was decide d to keep the hydrogen sorption temperature range below 130oC, and above 0oC, which is reasonably close to DOE targets. Hydrogen sorption measurements were not performed below 0oC, as it is the au thors opinion that it is simply not feasible for use in automotive applications to store hydrogen at such low
59 temperatures, especially considering the freez ing effects of fuel cells that would result from such a low temperature. It has been shown that a polyaniline nanoc omposite material can store just under 0.5wt.% hydrogen at 175oC. The inclusion of multiwall carbon nanotubes ha s little effect on the hydrogen sorption capabilit ies of the material, though th e increase in porosity of the material causes a slight increase in the hydr ogen sorption capabilities of the material. The carbon nanotubes cannot, however, be s een as the primary hydrogen storage material. SEM analysis shows that the hydroge n is mainly physically absorbed, in forms of small pockets of hydrogen. The inclusion of tin oxide has been shown to have no effect on the hydrogen sorption capabilities of the material either. Finally, the addition of aluminum to the nanocomposite matrix has sh own the greatest effect on the materials hydrogen capacity. The increase in hydrogen so rption capabilities caused by the inclusion of fine aluminum powder, as well as th e exact chemical inte raction between the aluminum and the polyaniline, is still unclear. The use of bulk polyaniline, however, is not a feasible or practica l option for hydrogen storage. 3.3. PANI Nanospheres Since the bulk polyaniline showed promise in that it was able to reversibly sorb hydrogen at slightly elevated temperatures, though with a low capaci ty, it was decided to investigate various nanostructures of polyan iline. As was previously mentioned, a higher surface area allows for more hydrogen bonding sites. Additionally, sm aller particle sizes also mean faster kine tics, at least theoreti cally, since the hydrogen can diffuse through the
60 material with more ease. Hence, the first nano structured polyaniline that was investigated was PANI nanospheres. 3.3.1. Synthesis of Polyaniline Nanospheres The polyaniline nanospheres (PANI-NS) were synthesized by oxidative polymerization of aniline monomer at 0oC in an ice bath using ammonium persulfate as the oxidant in the presence of surfactant. Aniline, ammonium persulfate, polyvinyl pyrroledene, cetyl ammonium bromide, and camphorosulfonic acid are used as received from Sigma-Aldrich. Camphorosulfonic acid surfactant as the dopant and ammonium persulfate as the oxidant were used in the present synthesis of polyaniline nanospheres (see flow chart in Figure 3.11). Calcul ated quantities of aniline monomer (0.005mol) were mixed with 50mL of distilled water and stirred using magnetic stirrer for 10 minutes. Meanwhile, calculated quantities of surfactant (0.75 mol) and oxidant (0.005mol) were dissolved separately in distil led water and stirred for 10 minutes in an ice bath. The surfactant solution was first a dded to the aniline monomer aqueous solution and then the previously cooled oxidant so lution was added drop wise after which the mixture was allowed to react for 10hr in an ice bath. The precipitate was filtered and washed several times with distilled water a nd methanol to terminate the polymerization reaction and then dried in vacuum at room te mperature for 24hr. Late r, the vacuum-dried precipitate was annealed at 100oC for 1hr. The dried polyaniline was characterized and tested for the hydrogen uptake a nd release measurements.
61 3.3.2. Scanning Electron Microscopy Results The microstructure of the as-synthesi zed PANI nanospheres was studied by Hitachi S800 scanning electron microscope (SEM). A fixed working distance of 5mm and a voltage of 5-25kV were used. Sample preparation for the SEM measurement was carried out inside the glove box by cove ring the sample holder with Parafilm for minimal exposure to oxygen while transferring it to the secondary emission chamber. EDAX Genesis software was used to analyze the SEM images. Aniline (0.005mol) + 1M HCl (Stirred and cooled to 0C) Ammonium persulfate (0.005mol) + 1M HCl Stirred and cooled to 0C for 5hr 2wt.% Pol y vin y l py rroledene Cam p horosulfonic aci d Cet y l ammonium b romide Stirred at 20C for 15hr Filtered and washed with DI H2O and CH3OH Annealed at 100C for 1hr Figure 3.11: Synthesis of PANI nanospheres.
62 It was found that the synthesis, as prev iously outlined, produced a virtually 100% yield of polyaniline nanospheres with an aver age diameter of 100nm, as can be seen in Figure 3.12. The nanospheres are not indepe ndent of each other, but instead are agglomerated. This is to be expected as the nanospheres are produced without templating and are simply formed through a chemical method. However, the nanospheres are very uniform and possess a better morphology than the bulk PANI previously discussed. Figure 3.12: SEM of PANI nanosphe res at a magnification of 25000.
63 3.3.3. FTIR Characterization Results The quinoid and benzenoid bond stretches of the PANI-NS were compared to the bulk PANI with a Perkin Elmer Spectrum One FTIR spectrometer. The PANI-NS samples were pelletized and sealed in a specially designed KBr cell for infrared measurements to prevent any further moisture uptake. As Figure 3.13 indicates, the FTIR spectrum of the PANI nanospheres is virtually identical to the spectrum of the bulk counterpart. The relative intensities of th e peaks to each other remain unchanged, with the exception of the out of plane C-H deformation peak observed around 800cm-1. All other peaks, most importantly the characteris tic quinoid and benzenoid peaks are virtually identical, therefore confirming the formati on of the emeraldine base polyaniline. Figure 3.13: FTIR comparison of bul k PANI with PANI nanospheres.
64 3.3.4. Hydrogen Sorption Results The volumetric hydrogen sorption measuremen ts are of paramount importance in understanding the hydrogen storage behavior of PANI-NS. Room temperature hydrogen absorption was executed at a high pressure (H2 ~80 bar) with a pre-calibrated reservoir. These isothermal volumetric measurements were carried out by HyEnergys PCT Pro 2000 sorption equipment. This fully automated Si everts type instrument uses an internal PID controlled pressure regulator with maximu m pressure of 170bar. It also includes five built-in and calibrated reservoir volumes of 4.66, 11.61, 160.11, 1021.30 and 1169.80mL, of which the 160.11mL reservoir was used, prov iding for a roughly 15:1 volumetric ratio of the hydrogen volume to available sample volume. The volume calibration with and without the sample was performe d at a constant temperature with an accuracy of 1C using helium. The software subroutines for hydrogen purging cycles, leak test, kinetics, PCT and cycling were performed by the HyDataV2.1 Lab-View program. The data collected for each run were analyzed usi ng the Igor Pro 5.03 program with a built-in HyAnalysis macro. As with the bulk polyaniline sample, it was decided to perform hydrogen sorption measurements close to room temperature. Since the nanospheres do not contain any hydrogen after synthesis, with the exception of the chemically bonded hydrogen that is part of the polyaniline chemical structure, the first hydrogen measurement performed was absorption. The sample was kept under vac uum and then exposed to 80bar of hydrogen pressure. Each absorption measurement was then followed by a desorption measurement at 30oC and given enough time to reach an equilibrium. The results of the kinetic absorption and desorption measurements are shown in Figure 3.14.
65 Figure 3.14: Absorption and desorption ki netics of PANI nanospheres at 30oC. The first hydrogen absorption measurement at 30oC revealed an uptake of more than 5.5wt.%, a significant improvement over the bulk polyaniline. However, the kinetics of the hydrogen uptake were extremely slow as it took almost 5hr to absorb the full 5.5wt.%. The speed of hydroge n release during the second kinetics measurement was even slower, though, as it took nearly 20hr to release just over 4.5w t.%, meaning that approximately 1wt.% was irreversibly stored in the polymer nanos tructure. Consecutive absorption and desorption cycles showed that the amount of hydrogen that was absorbed was always equal to the amount that was rel eased during the previous desorption cycle. However, the amount desorbed continuously de creased as did the kinetics. After only 3
66 absorption and desorption cycl es, the amount of hydrogen th at was absorbed was well under 1wt.%. Since the previous work w ith the bulk PANI showed that a higher temperature also meant a higher capacity, an absorption kinetic measurement at 90oC was performed. The results of this kinetics meas urement, shown in Figure 3.15, indicate that only approximately 0.6wt.% of hydrogen could be reversible sorbed. This data is in agreement with the best case scenario of the bulk polyani line previously discussed. Figure 3.15: Absorption kinetics of PANI nanospheres at 90oC. The slow kinetics of the hydrogen absorption indicate that chemis orption is taking place, rather than physisorption, as physisor ption is generally characterized by fast, almost instantaneous, hydrogen release or uptak e. This is further confirmed by the fact
67 that the desorption of hydrogen is extremely slow, which is to be expected since the temperature of measurement is relatively low, which means that there is not enough energy to break the hydrogen bonds quickly en ough. Furthermore, since the nanospheres are not separated, but instead are agglomerated, there is a larger hydrogen diffusion path which in turn also means slower kinetics. 3.3.5. Hydrogen Cycling Effects on PANI Nanospheres Since the capacity of the nanospheres d ecreased from 5.5wt.% to 0.6wt.% with consecutive hydrogen cycling, the nanosphere s were investigated using SEM. A representative image can be seen in Figure 3.16. Some remnants of the nanospheres can be seen, though it is evident that the nanostructure was signi ficantly deteriorated. The few remnants of the nanospherical structure that are visible further show that the average diameter has increased to approximately 200nm This swelling confirms that the PANINS chemically interacted with the hydrogen, thereby causing the increase in particle size. Additionally, the nanostructure was broken dow n by the repeated cycling, which is not due to any temperature effects that might cause the breakdown of the structure. Some microcracks can also be seen in the SEM image, which is a typical sign of hydrogen cycling in materials as the hydrogen diffuses through the material, thereby creating these cracks. The PANI-NS morphology observed afte r hydrogen cycling is reminiscent of the bulk PANI morphology. Since the nanospheres ar e virtually identical to the bulk PANI, as evidenced by the FTIR analysis, the deteri oration of the nanostructure explains why the final hydrogen capacity is close to that of the bulk PANI samples.
68 Figure 3.16: SEM image of PANI na nospheres after hydrogen cycling. 3.3.6. PANI Nanospheres Summary Emeraldine base polyaniline nanospheres were successfully synthesized with almost 100% yield and an average diamet er of 100nm. The FTIR characterization confirms that the chemical structure of the PANI is virtually iden tical to the bulk PANI previously investigated. It was found that the nanospheres have an initial uptake of approximately 5.5wt.% hydrogen with slow kine tics requiring approximately 5hr for full absorption. The desorption ki netic measurement at 30oC, however, shows that only 4.5wt.% of hydrogen is released again, leaving approximately 1wt.% contained in the PANI nanospheres. With each further absorption and desorption the capacity diminished
69 further, leaving more hydrogen in the stru cture. The rate of hydrogen absorption and desorption also decreased with each cycle, indicating that that the material essentially became saturated with hydrogen through chemisorption, as evidenced by the slow kinetics while making hydrogen diffusion mo re difficult. SEM analysis of the nanospheres after hydrogen cycling reveal th at the nanostructure of the material deteriorated as a result of chemical bonding of hydrogen to the P ANI material. Swelling of the few remaining nanospheres was also ob served, as were microcracks within the structure, evidence of hydrogen diffusion in the material. Hydrogen cycling at higher temperatures, as was done for the bulk PANI revealed no further significant hydrogen uptake or release, but instead essentially confirmed the hydrogen capacity of bulk PANI. 3.4. Chemically Grown PANI Nanofibers Nanofibers with diameters of tens of nanometers appear to be an intrinsic morphological unit that was found to naturally form in the early stage of chemical oxidative polymerization of aniline. In c onventional polymerization, nanofibers are subject to secondary growth of irregularly shaped particles that form the final granular agglomerates. The key to producing pure nano fibers is to suppress secondary growth. Based on this, two methods (int erfacial polymerization and ra pidly mixed reactions) have been developed that can readily produce pure nanofibers by slightly modifying the conventional chemical synthesis of polyanili ne without the need for any template or structural directing material. With th is nanofiber morphology, dispensability and processibility of polyaniline are now greatly improved. On the other hand, the template synthesis method is an effective way to grow the nanofibers of various conducting
70 polymers [57, 58]. The prepara tion conditions and their e ffect on morphology, size, and electrical properties of nanofib ers have been reported previo usly in literature . 3.4.1. Synthesis of Polyaniline Nanofibers The polyaniline nanofibers (PANI-NF) were synthesized by oxidative polymerization of aniline monomer at 0oC in an ice bath using ammonium persulfate as the oxidant in the presence of surfactant. Aniline, ammonium persulfate, dodecyl benzene sulfonic acid, acrylmethylpropyl sulfonic acid, and camphorosulfonic acid are used as received from Sigma-Aldrich. Sulfonic acid based surfactants as the dopant and ammonium persulfate as the oxidant were used in the present synthesis of polyaniline nanofibers (see flow chart in Figure 3.17). Calculated quantities of aniline monomer (0.05mol) were mixed with 50mL of distilled water and stirred using magnetic stirrer for 10 minutes. Meanwhile, calculated quan tities of surfactant (0.75mol) and oxidant (0.05mol) were dissolved separately in distille d water and stirred for 10 minutes in an ice bath. The Surfactant solution wa s first added into the anilin e monomer aqueous solution and then previously cooled oxidant solution drop wise and the mixture was allowed to react for 15 hours in an ice bath. The precipitate was fi ltered and washed several times with distilled water and methanol to termin ate the polymerization reaction and then dried in vacuum at room temperature for 24 hours. Later, the vacuum-d ried precipitate was annealed at 125oC for 3 hours. The dried polyaniline wa s characterized and tested for the hydrogen uptake and release measurements.
71 Figure 3.17: Flow chart for the synthesis of polyaniline nanofibers prepared in an aqueous medium with different surfactants a nd using ammonium persulfate as oxidizing agent. 3.4.2. FTIR Characterization Results The quinoid and benzenoid bond stretches of the PANI-NF were compared via Perkin Elmer Spectrum One FTIR spectromete r. The PANI-NF samples were pelletized and sealed in a specially designed KBr cell fo r infrared measurements. Fourier transform infrared (FTIR) spectra of standard P ANI and PANI-NF prepared from chemical templating method are shown in Figure 3.18. Aniline (0.01mol) + 1M HCl (Stirred and cooled to 0C) Ammonium persulfate (0.01mol) + 1M HCl (Stirred and cooled to 0oC) Stirred and cooled to 0C for 5hr Dodec y l benzene sulfonic aci d Cam p horosulfonic aci d Acr y l meth y l p ro py l sulfonic aci d Stirred at 0C for 15hr Filtered and washed with DI H2O and CH3OH Annealed at 125C for 3hr
72 Figure 3.18: FTIR spectra of PANI nanofiber and standard sample indicating that the major bonding environment remains unchange d for both standard and nanofibrous polyaniline structures. The major bonding environment remains unchanged for both standard and nanofibrous polyaniline structures. The presen ce of two bands in the vicinity of 1500cm-1 and 1600cm-1 are assigned to the non-symmetric C6 ring stretching modes. The higher frequency vibration at 1600cm-1 is for the quinoid rings, while the lower frequency mode at 1500cm-1 depicts the presence of benzenoid ri ng units. Furthermore, the peaks at 1250cm-1 and at 800cm-1 are assigned to vibrations associated with the C-N stretching vibration of aromatic amine out of plane deformation of C-H of 1,4 disubstituted rings. The aromatic C-H bending in the plane (1167cm-1) and out of plane (831cm-1) for a 1,4
73 disubstituted aromatic ring indicates a linea r structure. Independe nt of surfactant, complete formation of polyaniline has been observed always. 3.4.3. Scanning Electron Microscopy Results The microstructure of the as-synthesized PANI nanofibers was studied by Hitachi S800 scanning electron microscope (SEM). A fixed working distance of 5mm and a voltage of 5-25kV were used. Sample prepar ation for the SEM measurement was carried out inside the glove box by covering the sample holder with Parafilm for minimal exposure to oxygen while transferring it to the sec ondary emission chamber. EDAX Genesis software was used to analyze the SEM images. It is clearly discernible from the s canning electron micrograph, shown in Figure 3.19, that the density of nanofiber formation re mains rather constant, irrespective of the surfactant used during the s ynthesis of PANI-NF. Furthermore, it can be seen from Figure 3.19 that the average nanof iber diameter is approxim ately 250nm. It is important to note that the surface of the nanofibers is rather r ough, which is typical of chemical nanofiber growth. Additionally, the nanofibers are rather inconsistent in length and overall appear to be fragmented. However, there is no evidence of bulk PANI observed, meaning that there is essentially a 100% yi eld of nanofibers. The nanofibers that were synthesized using camphorosulf onic acid, as seen in Figure 3.19 (c), were analyzed for their hydrogen sorption behavior because of their slightly rougher surface, as it was thought that they possess more pot ential hydrogen binding sites.
74 Figure 3.19: Scanning electron micrographs of pol yaniline nanofibers grown at room temperature in aqueous medium with differe nt surfactants (a) dodecyl benzene sulfonic acid, (b) acrymethylpropyl sulfonic acid, a nd (c) camphorosulfonic acid using ammonium persulfate as oxidizing agent.
75 3.4.4. Hydrogen Sorption Results The volumetric hydrogen sorption measuremen ts are of paramount importance in understanding the hydrogen storage behavior of PANI-NF. Room temperature hydrogen absorption was executed at a high hydrogen pre ssure of approximate ly 80bar with a precalibrated reservoir of 160.11mL, providing for roughly a 15:1 volumetric ratio of hydrogen to sample volume. Figure 3.20 demonstrates the initial hydroge n uptake of PANI-NF with respect to time. From this figure it is seen that ra pid absorption was achieved (i.e. 95% of total capacity 3-4wt.% in less than 10min) by th e nanofibrous matrix. The absorbed hydrogen was then desorbed against 1bar and approximately 1-2wt. % hydrogen was released at room temperature (see Figure 3.20).
76 Figure 3.20: Hydrogen absorption and desorp tion kinetics of PANI-NF in the 1st and 13th cycle. In order to confirm and ensure the e ffective hydrogen sorption, the PressureComposition-Temperature (PCT) measurements we re carried out at room temperature for the PANI-NF from the 2nd cycle to the 6th cycle as shown in Fi gure 3.21 (a) and (b). Interestingly, a distinct plateau pressure region was observed, which is usually an identifier for effective hydride formation of chemisorption, around 30bar (up to 1.5wt.% hydrogen uptake) and a linear region (up to 2wt.%) with total abso rption capacity around 3.5wt.%. In the consecutive hydrogenation cycles (3rd-6th), this plateau pressure region diminishes; nevertheless, the hydrogen absorpti on capacity remains the same as shown in Figure 3.21 (a). The initial plateau re gion observed most likely corresponds to
77 chemisorption of hydrogen into the ma terial. The hydrogen initially bonds via physisorption to the nanofibers, but hydrogen also fills any vacancies and unterminated bonds that arose during the synthesis of th e nanofibers, which is very common for chemically synthesized polymers. The fact th at no desorption plateau region is observed further confirms hydrogen sorption via chem isorption, as room temperature does not provide the hydrogen with enough energy to break the hydrogen bonds produced through chemisorption. The reduction in the plateau pressure, as well as the reduction in the amount of hydrogen that is ac tually chemically sorbed during consecutive cycles, indicates that the na nofibers become saturated and terminated with hydrogen. Since hydrogen is still reversibly sorbed in the na nofibers without any plateau region, it is clear that the remaining 3.5wt.% of hydroge n is stored via physisorption. The reversibility of hydrogen absorption and desorption of 3 to 4wt.% at room temperature was unambiguously demonstrated by Figure 3.21(a) and (b). It is also noticeable from this figure that by increasing the sorption temperature from room temperature to 50oC in the 5th cycle, the plateau pressure region disappears, which indicates that these nanofibers react with hydrogen more eff ectively at room temperature than at a moderate temperature of 50oC.
78 Figure 3.21: Pressure-Composition Isotherms (PCT) of PANI-NF at room temperature from 2nd to 6th (a) absorption (b) desorption cycles.
79 Furthermore, it is clear that the main driving force behind hydrogen sorption in these nanofibers is not temperature, but instead pressure. This means that a pressure of approximately 80bar is required to keep the hydrogen stored in the material. If the pressure is reduced from 80bar to near vac uum, the hydrogen is very quickly released, as is evident in Figure 3.22 th e kinetic behavior of the na nofibers. Conversely, if the pressure is increased instantaneously to 80bar, the hydrogen is quickly reabsorbed. After several hydrogen sorption cycles the absorption kinetics at the 13th cycle was recorded and compared with the 1st cycle. It was observed that the reaction kinetics in the 13th cycle remain the same as for the 1st cycle (Figure 3.20). Hydrogen absorption and desorption from the 14th to 25th cycle are shown in Figure 3.22. It is clearly seen from this figure that there was no deterioration in the hydrogen storage capacity of 3-4wt.% during these reversible cycles. Furtherm ore, it can be seen that it only takes approximately 6min to absorb the hydrogen. With increasing cycles, however, there is a slight reduction in the kinetics of the mate rial, most notably in the desorption of hydrogen. While the absorption kinetics stay virtually unchanged, the desorption kinetics slow from 6min to about 30min for fu ll hydrogen release with consecutive hydrogen absorption / desorption cycles.
80 Figure 3.22: Hydrogen sorption kinetics at room temperature from 14th cycle to 25th cycle showing little degradation in kinetics and no degradation in the storage capacity within the measured cycles. 3.4.5. Hydrogen Cycling Effects on PANI Nanofibers This decrease in hydrogen kinetics can be explained when looking at the microstructure of the PANI nanofibers after hydrogen cycling, as seen in Figure 3.23. It is evident that the nanofiber structure that was initially present is no longer intact. The nanofibers are in fact who lly missing. The interaction of the hydrogen with the nanofibers, through repeated pressure changes at close to room temperature first caused chemical bonding to the PANI which resulted in termination of bonds. Additionally, the force caused by the pressure variations of the hydrogen essentia lly resulted in the nanofibers being repeatedly compressed and decompressed, thereby forcing the
81 microstructural changes. As can be seen in the SEM image, however, the PANI surface morphology is still characterized by a porous nature, created by the diffusion of hydrogen in and out of the structure. Since the micros tructure changed so significantly, a surface area analysis was performed on the PANI nanof ibers before and afte r hydrogen cycling to confirm the initial findings of unchanged su rface area. The results of the BET surface area analysis are pres ented in Figure 3.24. Figure 3.23: SEM image of PANI na nofibers after hydrogen cycling.
82 Figure 3.24: BET surface area analysis of P ANI nanofibers (CM) before and after hydrogen cycling. It can be seen that the surface area of the PANI nanofibers before hydrogen cycling was approximately 3.26m2/g and after hydrogen cycling actually increased slightly to 3.56m2/g. This surface area is significantl y lower than most other porous polymers reported in literature, whic h often have areas larger than 100m2/g. The intrinsic nature of polyaniline, however, allows th e hydrogen to bond with many different sites both through chemisorption and physi sorption, as previously shown. 3.4.6. PANI Nanofibers (CM) Summary In summary, polyaniline nanofibers (PANI-NF), synthesized by chemical templating technique using vari ous surfactants as dopants a nd ammonium persulfate as the oxidant, were successfully shown to reve rsibly store hydrogen. Th e characteristics of
83 PANI-NF are understood regard ing their structure, micros tructure, bonding, and thermal stability behaviors. The signature bands of benzenoid and quinoid transitions in the frequency of 1500cm-1 and 1600cm-1, respectively, as evidenced from FTIR, confirms the formation of the polyaniline nano structure irrespective of the surfactants used in the precursor preparation. The rate of hydrogen sorp tion during the initial cycle is very rapid (95% hydrogen storage capacity (3-4wt.%) absorb ed in less than 6min). Moreover, these PANI nanofibers demonstrate excellent reversib ility (up to 25 cycles measured) at room temperature. This behavior has previously not been reported in literature. Another important feature discernible from the PC T isotherms was that during the second hydrogen absorption run, the plateau pressure occurred around 30bar, and it diminished in subsequent cycles. This means that hydrogen first bonded to vacancy sites and unterminated bonds created during the chemical synthesis. The reduction in the plateau pressure as well as the reduction in plateau cap acity with subsequent cycles indicates that the chemisorption sites disappear as hydroge n is chemically bonded. This chemisorption is further proven by the lack of plateau regions in desorption cycles, as the low temperature (around room temperature) does not provide enough en ergy for the hydrogen to break the stronger chemical bonds. Neverthe less, the reversible capacity of 3-4wt.% was maintained throughout 25 cycles through ph ysisorption with pre ssure changes as the main driving force in hydrogen absorption and desorption. Furthermore, it was shown that the nanofibrallar na ture of the PANI nanofib ers disappears through hydrogen cycling, most likely due to repeated compre ssion of the structure from 0 to 80bar of hydrogen pressure. The surface area, however remains virtually unchanged and even increases slightly. This means that the bindi ng sites for the hydrogen remain intact, even
84 though the morphology is changed significantl y. While a capacity of 3-4wt.% is not enough to meet the DOE targets for a pract ical hydrogen storage medium, the rapid kinetics and low temperature required for hydrogen cycling make these chemically synthesized nanofibers an attractive hydrogen storage medium for smaller applications. 3.5. Electrospun PANI Nanofibers In this section, the synthesis and char acterization of polyaniline nanofibers by electrospinning process of polymer solution on a collection substrate is investigated. These as-deposited nanostructures were char acterized using Fourie r transform infrared (FTIR) spectroscopy for dete rmination of stretching modes and vibrations. An important aspect of hydrogen adsorption and desorption be havior in these nanof ibers was estimated by high pressure hydrogen sorption measurem ents. Microstructural changes due to hydrogen sorption were observed by scanning electron microscopy (SEM) in the imaging mode. 3.5.1. Synthesis of Electrospun Polyaniline Nanofibers As with the chemically grown nanofiber s, the polyaniline nanofibers (PANI-NFES) were grown by oxidative polymerization of aniline monomer at 0oC in an ice bath using ammonium persulfate as the oxidant in the presence of surfactant, though with slightly different conditions. Aniline, amm onium persulfate, dodecyl benzene sulfonic acid, acrylmethylpropyl sulfoni c acid, and camphorosulfonic acid are used as received from Sigma-Aldrich. Sulfonic acid based surfactants as the dopant and ammonium persulfate as the oxidant were used in the present synthesis of pol yaniline nanofibers (see flow chart in Figure 3.25).
85 Calculated quantities of aniline monomer (0.01mol) were mixed with 50mL of distilled water and stirred using magnetic stirrer for 10min. Meanwhile, calculated quantities of surfactant (0.75mol) and oxidant (0.01mol) were dissolved separately in distilled water and stirred for 10min in an ice bath. The surfactant solution was first added into the aniline monomer aqueous solution and then the previously cooled oxidant solution was added drop wise after which the mi xture was allowed to react for 15hr in an ice bath. Aniline (0.01mol) + 1M HCl (Stirred and cooled to 0C) Ammonium persulfate (0.01mol) + 1M HCl Stirred and cooled to 0C for 5hr Dodec y l b enzene sulfonic aci d Cam p horosulfonic aci d Acr y l meth y l p ro py l sulfonic aci d Stirred at 0C for 15hr Electrospinning of the polyaniline solution Extracted the PANI-NF and annealed at 125C Characterization and hydrogen storage property Figure 3.25: Flow chart for the synthesi s of electrospun polyaniline nanofibers.
86 The obtained polymer solution was then sprayed by the electrospun method schematically represented in Figure 3.26. In th e electrospinning process, a high electric field is applied to the viscous polymer so lution, held in a capillary tube, inducing a charge density on the liquid surface. The distance between the tip of the needle and the substrate was optimized to 20cm with a voltage to 15kV. Mutual charge repulsion causes a force directly opposite to the surface tension. When the electric field is sufficiently high, th e surface of the solution in proximity of the tip of the capillary tube elongate s and forms a cone, named Taylor cone, which in turn Figure 3.26: Schematic diagram of electro spinning method to produce polyaniline nanofibers. (+) Positive Ground Positive high voltage is charged to a polymer solution As voltage increases, a polymer cone is formed on the capillary tip of the syringe Nanofiber is accumulated on the surface of the grounded copper collector
87 allows for the formation of randomly arranged nanofibers These polyaniline nanofibers are collected and then dried in vacuum at room temperatur e for 24hr. Later, the vacuumdried sample is annealed at 125oC for 3hr. Finally, the dried electrospun nanofibers are characterized and tested for the hydr ogen uptake and release measurements. 3.5.2. Scanning Electron Microscopy Results The microstructure of the as-synthesized PANI nanofibers was studied by Hitachi S800 scanning electron microscope (SEM). A fixed working distance of 5mm and a voltage of 5-25kV were used. Sample prepar ation for the SEM measurement was carried out inside the glove box by covering the sample holder with Parafilm for minimal exposure to oxygen while transferring it to the sec ondary emission chamber. EDAX Genesis software was used to analyze the SEM images. The electrospun nanofibers were successfu lly grown and can be seen in Figure 3.27. It is important to note that the nanofibers are more or less of the same dimensions. The average diameter is approximately 1.5 m, thereby not technically nanofibers. The fibers have very large aspect ratios, with lengths of several m. Additionally, the fibers are very smooth in appearance, which is ty pical of electrospun grown fibers. No bulk PANI is visible from the SEM images, meani ng that there is essent ially a 100% yield of nanofibers.
88 Figure 3.27: SEM image of electrospun PANI nanofibers at a magnification of (a) 1000 and (b) 5000. 3.5.3. FTIR Characterization Results Fourier transform infrared (FTIR) spect ra of electrospun polyaniline nanofibers (PANI-NF-ES) before hydrogen sorption are shown in Figure 3.28. The major bonding environment remains unchanged for both the pristine and hydrogenated PANInanofibrous structures. The presence of two bands in the vicinity of 1500cm-1 and (b) (a)
89 1600cm-1 are assigned to the non-symmetric C6 ring stretching modes. The higher frequency vibration at 1600cm-1 is for the quinoid rings, while the lower frequency mode at 1500cm-1 depicts the presence of benzenoid ri ng units. Furthermore, the peaks at 1250cm-1 and at 800cm-1 are assigned to vibrations associated with the C-N stretching vibration of aromatic amine out of plane de formation of C-H of 1,4 disubstituted rings. The aromatic C-H bending in the plane (1167cm-1) and out of plane (831cm-1) for a 1,4 disubstituted aromatic ring i ndicates a linear structure. Figure 3.28: FTIR spectra of bulk PANI and electrospun PANI nanofibers.
90 It is interesting to note that the quinoid ring stretch around 1650cm-1 is significantly more pronounced re lative to the other peaks and also shifted by 50cm-1 as compared to the bulk polyaniline. Additi onally, the benzenoid ring stretch around 1450cm-1 is not as broad, but instead composed of several individual smaller peaks. Overall, the FTIR spectrum confirms that polyaniline in its emeraldine form is present, as all the major characteristic peaks are observed. However, there is a definite difference in the FTIR spectra of the electr ospun nanofibers as compared to even the chemically grown nanofibers, which essentially had an identic al spectrum to the bulk chemically grown polyaniline. 3.5.4. Thermogravimetric Results The thermal stability of PANI nanofibe rs (both CM and ES methods) and the standard samples were charac terized using TA Instruments simultaneous DSC and TGA (SDT-Q600) tool. A pre-weighed sample was loaded into the ceramic pan and covered with the ceramic lid inside th e glove box to prevent moisture from getting into the sample during transfer. The ramp rate of 5C/min was used for all the measurements. TAs Universal Analysis 2000 software program was used to analyze the TGA and DSC profiles. This TGA comparison is shown in Figure 3.29. The initial wei ght loss up to 15% is due to moisture that was absorbed during regular storage, since the polyaniline samples were all kept in sealed, but non-inert containe rs. After the release of the stored moisture, it can be seen that the PANI nanofibers (ES) are stable up to 150oC. This is to be expected, as the sample preparation includes annealing of the PANI at 125oC. A
91 temperature of higher than 150oC is not required any way as this would be too high a temperature for practical hydrogen storage. It should be noted, however, that the electrospun nanofibers are less thermally stable than the bulk or chemically grown counterparts. This is most likely due to the sl ight chemical differences that were observed during FTIR analysis, as previously mentioned. Figure 3.29: Thermogravimetric weight loss an alysis of PANI samples in both bulk and nanofiber form. 3.5.5. Hydrogen Sorption Results The electrospun PANI nanofibers were subjected to both kinetic and PCT hydrogen measurements. Figure 3.30 demonstr ates the hydrogen sorption life cycle kinetic curves of PANI-NF at tw o different temperatures, 30 and 100oC. As with previous 70 75 80 85 90 95 100 2575125175225275Weight (% ) Temperature (oC) PANI-Nanofiber C M PANI-Standar d PANI Nanofiber ES
92 samples, the first hydrogen sorption measurement was performed at 30oC. However, when the nanofibers were subjected to a hi gh hydrogen pressure of 80bar th e hydrogen uptake was found to be less than 0.75wt.%, which is reminiscent of the bulk PANI samples. Additionally, there was no real desorp tion observed at this low temperature. It was therefore decided to perform the hydroge n measurements at va rious temperatures, and finally at 100oC (see Figure 3.30) hydrogen was absorbed. Surprisingly, a hydrogen uptake (A1 in Figure 3.30) of close to 11wt.% with rapid adsorption kinetics (approximately 95% hydrogen adsorption within 5-10min) was observed at 100oC. The desorption cycle of the same sample at 100oC then exhibited a hydr ogen release of close to 8wt.%. This means that the other 3wt. % of hydrogen that was previously absorbed remained in the structure. Interestingly, the desorption (labeled as D1 in the figure) was composed of both a rapid hydrogen release step in which approximately 4wt.% was released within a few minutes and a slower hydrogen release step, accounting for another 4wt.% in approximately 45min. When the desorption equilibrium pressure was reached, hydrogen was reabsorbed into the sample. Th is time (A2) the fu ll 8wt.% that was previously released was reab sorbed as rapidly as had oc curred in the first hydrogen absorption. After this, only about 5wt.% hydr ogen was released, again in a two-step process. Only this time the slow er release step released only about 1wt.%, while the fast release regime remained at a capacity of 4w t.%. The full 5wt.% was then reabsorbed, though this time the kinetics of the hydroge n uptake decreased from about 5min to 25min.
93 Figure 3.30: Hydrogen sorption kinetics curv es for the PANI-NF-ES at 30 and 100oC. After this cycle, the hydrogen capacity and kinetics dropped very quickly until finally virtually no hydrogen co uld be sorbed. The two-step release observed for the electrospun nanofibers is clearly evidence that both physisorption as well as chemisorption are taking place in the structure, exactly as was hoped at the outset of this work. The physisorption, again driven mainly by pressure changes, allowed for very rapid hydrogen release and uptak e, while the slower second step during release was due to chemisorption, as more energy or more time in lack of a higher temperature, is required to release the hydrogen. The decrease in the hydrogen absorption capacity from cycle to cycle is clearly due to chemisorption of the hydrogen to the nanofibers. Once the -8 -6 -4 -2 0 2 4 6 8 10 12 051015202530 Time (hr)Hydrogen Concentration (wt% ) Life Cycle Kinetics @ 30C Life Cycle Kinetics @ 100C Ph y sisor p tion Chemisorption A1 A2 A3 A4 D3 D2 D1
94 strong primary bond between the structure a nd the hydrogen is formed, it is virtually impossible to get the hydrogen back out without destroying the polymeric structure, as a temperature of more than 150oC has been shown to start the decomposition process. Since only kinetic measurements were pe rformed on the nanofibers and also in order to confirm the initial hydrogen uptake of 10wt.% that resulted from these measurements, a new batch of electrospun nanof ibers was loaded into the PCT reactor. The hydrogen PCT absorption and desorption beha vior of this batch is shown in Figure 3.31 and Figure 3.32, respectively. It is importa nt to note that only PCT measurements were made on this batch and no kinetics measurements. It was observed that virtua lly no hydrogen was absorb ed or desorbed until a measurement temperature of approximately 50oC. After the initial hydrogen uptake of about 2.5wt.% at 50oC, the hydrogen capacity was found to increase w ith increasing temperature up to 125oC in various cycles as seen in Figure 3.31. At approximately 100 to 125oC a two fold increase of capacity (6-8wt .%) was invariably obtained at various hydrogenation cycles. At the end of each adso rption PCT, desorption experiments were performed and the results are depicted in Figure 3.32. A hydrogen st orage capacity of 210wt.% was obtained at te mperature range of 50-125oC.
95 Figure 3.31: Hydrogen adsorption PCT curves for the PANI-NF-ES at different temperatures. While it might appear that the amount of hydrogen that is released from the nanofibers is larger than the amount that is absorbed, this is in fact not the case. The data presented in Figure 3.32 is actually the amount of hydrogen released without taking the hydrogen itself into account which means that the numbers are slightly higher than they should be. It should be regarded as raw data When one actually takes into account the weight of the material plus the weight of hydrogen stored in it, the maximum amount of hydrogen released changes from 8.5wt.% to approximately 7.5wt.%, thereby corresponding to the amount of hydrogen previously absorbed.
96 Figure 3.32: Hydrogen desorption PCT curves for the PANI-NF-ES at different temperatures. The nanofibers that were tested for th eir capacity using PCT only are from the same synthesis batch and have absolutely no difference in terms of their chemical or physical composition to the batch of sample that was subjected only to kinetics measurements. However, the type of meas urement that was employed made a large difference in capacity. As previously me ntioned, hydrogen through physisorption or chemisorption is often based on diffusion. Wh en one performs a kinetics measurement, the sample is subjected to either a high ( 80bar) or a low (0bar) hydrogen pressure. A PCT measurement, however, allows the sample to experience gradual pressure increase as opposed to this sudden pressure change. The hydrogen, therefore, has enough time to
97 diffuse through the material as the pressure is incrementally increased or decreased for absorption or desorption, respectively. While th is has not affected any previous samples, the electrospun nanofibers were much more se nsitive to the type of measurement. Most likely, this is due to the difference in chem ical structure, as was shown by the FTIR spectrum in Figure 3.28. 3.5.6. Hydrogen Cycling Effects on Electrospun PANI Nanofibers In order to confirm the hypothesis of chemisorbed hydrogen from the kinetic measurements, scanning electron microscopy was performed on the cycled nanofibers. The SEM image of the electrospun PANI na nofibers before hydrogen sorption, shown in Figure 3.33 (a) is compared to the SEM imag e of the nanofibers after hydrogen cycling, as shown in Figure 3.33 (b). It becomes eviden t that the density of the nanofibers was not altered significantly. The SEM micrographs of the hydrogenated and dehydrogenation samples show nanofibrallar sw elling, breakage of fiber lengt h and precipitations, which is most likely the cause for hydrogen loading sa turation and poor cyclic reversibility. In addition to the swelling of the nanofibers du e to the chemical bonding of hydrogen to the PANI, it is clear that the nanofibers also break into smaller pieces, thereby losing the high aspect ratio and returning more to the bulk form of polyan iline previously described.
98 Figure 3.33: SEM micrographs of PANI-NF (a ) before and (b) after hydrogen sorption cycling. After hydrogen cycling, FTIR was performe d to observe any possible changes in the structure of the nanofibers. As can be s een from Figure 3.34, there is absolutely no change in the nature of th e chemical bonds. The quinoid and benzenoid ring vibrations and stretches are all intact. Due to the nature of sample preparation, though, the intensity of the transmittance does vary between the measurements. The relative intensities of the individual peaks remains constant, though, i ndicating no chemical change after hydrogen interaction. a) b)
99 Figure 3.34: FTIR spectra of electrospun P ANI nanofibers before and after hydrogen cycling. 3.5.7. Electrospun PANI Nanofibers Summary Conducting polyaniline nanofibers were s ynthesized using chemical templating method followed by an electrospinning proce ss. The FTIR spectra of these PANI nanofibers reveal the presence of tw o bands in the vicinity of 1500 cm-1 and 1600 cm-1, which are assigned to the non-symmetric C6 ri ng stretching modes. The lower frequency vibration at 1500cm-1 is for the benzenoid rings, while the higher frequency mode at 1600cm-1 corresponds to benzenoid to quinoid transition indicating the synthesized polymer was polyaniline in oxidized form. Th ese nanofibers have been compared with their standard bulk counterpart and found to be stable up to 150oC, slightly lower than
100 their bulk chemical counterparts. It was found that electrospun polyaniline nanofibers have a high hydrogen uptake of 10wt.% at around 100oC in the first absorption run. However, in the consecutive hydrogenation and dehydrogenation cy cles, the capacity diminishes. This is due to a combinati on of chemisorption and physisorption during kinetic measurements where the sample is exposed to varying extreme pressures. Eventually, the electrospun PANI nanofiber s become saturated with hydrogen during kinetic loading. When the nanofibers are exposed to PCT measurements, though, where the pressure is either gradually increased or decreased, the nanofibers exhibit a reversible hydrogen storage capacity of 3-8wt.% at differe nt temperatures. This is due to the fact that the hydrogen is given enough time to diffuse and reach local equilibrium at the varying pressures. Unfortunately, this also mean s that a larger time is required to either absorb or desorb the hydrogen. The surface morphologies before and after hydrogen sorption on these PANI nanofibers encompass significant changes in the microstructure such as nanofibrallar swelling and slight de terioration in the le ngth of the fibers. It was confirmed that the hydrogen data is reproducible with high hydrogen storage capacity of 3-8wt.% at different temperatures, as summarized in Table 3.1. From this table, it is discernibl e that the hydrogen capacity incr eases with an increase of temperature. However, in order to attain th is higher capacity, the nanofibers need to be exposed to gradual pressure changes rather than be exposed to large pressure changes, as would be typical with kinetics measurements.
101 Table 3.1: Hydrogen sorption capacity of electrospun nanofibers at various temperatures and sorption cycles. Cycle Number Adsorption/ Desorption Temperature (oC) Hydrogen Concentration (wt.%) 7 Adsorption 100 6.5 8 Desorption 100 6.5 13 Adsorption 50 3 14 Desorption 50 3 24 Adsorption 125 8.5 25 Desorption 125 8.5 3.6. Polyaniline Hydrogen Storage Summary Polyaniline, in its emeraldine form was successfully synthesized through chemical means. The bulk, or as synthe sized form, as well as three different nanostructures of PANI were characterized for physical, ch emical and hydrogen sorption characteristics. It was found that the inclus ion of additives to the bulk form PANI had virtually no effect on its hydrogen sorption char acteristics. Bulk PANI was found to have a hydrogen capacity of less than 0.5 wt.% at a temperature of 100oC. It was observed that the capacity of hydrogen in th is standard sample increase d slightly with increasing temperature. This increase with temperature is due to the expansion of PANI, which in turn allows for the hydrogen to diffuse r eadily into the polymer While others have claimed that standard PANI can absorb as mu ch as 8wt.%, this was refuted by Panella et al., as well as this research. The nanostructured PANI nanospheres, with an average diameter of 100nm, were synthesized by slightly modify ing the chemical synthesis technique used to produce the standard PANI. FTIR analysis, however, show s that the nanospheres do possess the same chemical composition as the standard PANI. While an initial hydroge n uptake of 5.5wt.%
102 of hydrogen was observed at 30oC, the capacity of each cons ecutive desorption cycle was reduced by roughly 1wt.%, t hough the desorbed hydrogen wa s reabsorbed. Since the kinetics of hydrogen uptake were on the order of several hours, with slower kinetics for each consecutive cycle, it was found that th e main driving force behind the hydrogen sorption of the nanospheres was chemisorp tion, most likely in the form of hydrogen reacting with unterminated bonds that were a resu lt of the synthesis. Fi nally, a capacity of only 0.6wt.%, virtually identical to that of bulk PANI, was observed at a temperature of 90oC. SEM analysis showed that the repeat ed hydrogen cycling caused the nanospheres to swell after hydrogen r eaction and also to agglomerate unti l it finally had an appearance reminiscent of bulk polyaniline. The interacti on of hydrogen with the sample also led to the formation of microcracks, which are cau sed by the diffusion of hydrogen through the material. By again slightly modifying the chemical synthesis technique, chemically grown PANI nanofibers were created. FTIR investig ation again proved that the nanofibers were of the emeraldine form. The fibers had an average diameter of 250nm with varying lengths. The nanofibers were also characte rized by a rough surface. Hydrogen cycling of the nanofibers showed that an initial capacity of 3-4wt.% was absorbed in less than 10min, while approximately 1-2wt.% hydrogen wa s released at room temperature. PCT measurements performed on the nanofiber s showed that roughly 3.5wt.% hydrogen was reversibly stored in the nanostructures. The initial hydrogen absorption showed a hydrogen plateau pressure around 30bar, indica tive of chemisorption. Through repeated cycling, however, the plateau pressure re gion disappeared, though the capacity remained unchanged at 3.5wt.% for many cycles. SEM analysis of the chemically grown
103 nanofibers after hydrogen cycling showed no evidence of the initial nanofibrous morphology. Instead, the fibers had virtually disappeared a nd a rather porous material remained. Surface area analysis revealed th at the surface area re mained unchanged after hydrogen cycling. Unlike the nanospheres, whose structure also disappeared, the nanofibrous nature of the PANI allowed fo r it to agglomerate into a porous material, thereby losing the initia l physical characteristics, but st ill maintaining the surface area. This is an important factor for hydrogen stor age materials, as was previously discussed. Finally, electrospun nanofibers were investigated for th eir hydrogen performance. By using a well known electrospinning techniqu e, the nanofibers were synthesized with average diameters of 1.5 m and lengths of several microns. The nanofibers were, unlike the chemically grown nanofibers, rather smooth. While FTIR again confirmed the presence of the typical benzenoid and quinoi d rings of emeraldine base PANI, it was observed that the quinoid peak was much larger in relation to the other FTIR peaks. The quinoid peak was also shif ted by approximately 50cm-1, making it unique among all the polyaniline structures inves tigated. Hydrogen kinetic meas urements on the electrospun nanofibers at 30oC showed no hydrogen sorption, but absorption at 100oC revealed a very promising initial upta ke of over 10wt.% in less than 10m in. The desorption that followed showed an almost instantaneous release of 4wt.% hydrogen and another 4wt.% with much slower kinetics. Each consecutive absorption cycle reabsorbed the previously released hydrogen with slightly lower kinetics, but still in a matter of minutes rather than hours. The desorption capacity decreased by roughly 2wt.% for each cycle, but the fast hydrogen desorption remained until finally, afte r approximately 3 absorption / desorption
104 cycles, virtually no hydrogen was sorbed. Wh ile the fast kinetics of the absorption measurements clearly indicate that physisorption is taking place, the desorption cycles are clearly a combination of both chemisorption (slower kinetics) and physisorption (faster kinetics). The combination of physisorption and chemisorption is also evidenced by the breakage and swelling of the nanofibers, though overall they do remain intact. Hydrogen PCT measurements on a new batch of this sa mple, however, revealed that approximately 8wt.% hydrogen could be reversib ly stored in the electrosp un nanofibers as long as the pressure was gradually increased or decr eased. This slow exposure to hydrogen is essential in that the nanofibers are not exposed to pressure extremes, which can result in the destruction of the nanofiber mor phology. Additionally, the hydrogen is allowed enough time to diffuse into the material and bond. While this method of storage requires a higher amount of time, it is still the most preferential method of storing hydrogen in polyaniline. The capacity of 8wt.% along w ith the relatively low temperature of 100oC makes the electrospun polyaniline an id eal candidate for hydrogen storage. It appears that the main factor in storing hydrogen in polyaniline is less the surface area, as was previously thought for po lymers, but in fact lie s more in the exact chemical composition. Unterminated bonds, which provide chemisorption sites for hydrogen are an excellent initial hydrogen stor age site, but require an amount of energy that is simply too high for practical purposes therefore requiring high temperatures that would in fact destroy the chemical structure of the polyaniline. The relative intensity of the quinoid ring vibration with respect to the ot her FTIR peaks appears to play a vital role in the hydrogen storage behavi or of polyaniline. Also, th e form of hydrogen storage, namely exposure of the material to hydrogen, is extremely important when dealing with
105 relatively weak materials such as PANI. T hough it might require more time to store and release the hydrogen, a gradual pressure change is essential to maintaining the hydrogen characteristics of the material. The work on these polymers is the first to investigate various nanostructures of polymers at temperatures above 77K. A comp arison of the re sults obtained from this research with other work that was previ ously conducted on hydrogen storage in polymers is shown in Figure 3.35. The significant resu lts described within this chapter are highlighted in red boxes, showing a clos ing in on the DOE goals with the PANI nanofibers produced via electrospinning me thod almost meeting the DOE targets. Figure 3.35: Comparison of novel PANI results with previous polymer hydrogen storage research.
106 Chapter 4 Complex Hydrides LiBH4/LiNH2/MgH2 4.1. Introduction As briefly mentioned in chapter 1, only complex hydrides composed of elements with less than an average molecular weight of 51.8g/mol (Cr) provide a sufficient enough gravimetric hydrogen density to meet or exceed the DOE targets for a practical hydrogen storage system. The group I, II and III light meta ls, such as Li, K, Be, Na, Mg, B, Ca and Al are excellent candidates as they form a large variety of me tal hydrogen complexes. The most common of these metal hydrogen complexes consists of borohydride (BH4 -), amide (NH2 -) and alanate (AlH4 -) ions, which are accompanied by cations such as Li+ or Na+. The hydrogen is generally stored in th e corners of a tetrahedron within these systems. Some previously investigated materi als in this class are listed in Table 4.1. Table 4.1: Physical characteristics of some previously investigated complex hydrides. Material Molecular Weight (g/mol) Density (g/cm3) Melting Temperature (oC) H2 Release Temperature (oC) H2 Capacity (wt.%) LiBH4 [ 60 ] 21.784 0.66 268 380 18.4 NaBH4 [ 61 ] 37.83 1.074 505 400 10.6 LiAlH4 [ 61 ] 37.95 0.917 125 125 9.5 KBH4 [ 61 ] 53.94 1.178 585 500 7.4 NaAlH4 [ 62 ] 54.0 1.27 178 210 7.4 Mg2NiH4 [ 63 ] 111.3 2.72 280 3.6 Mg2FeH6 [ 24 ] 110.5 2.72 320 5.4 Mg3MnH7 [ 24 ] 134.9 2.30 280 5.2 BaReH9 [ 24 ] 332.5 4.86 100 2.7
107 Most alanates are extremely stable and often decompose in twoor three-step reactions, limiting the amount of hydrogen that can be released, since often a higher temperature is required for th e release of hydrogen from the second step. Borates are also extremely stable and therefore often require high temperatures for hydrogen release, but do so in a one-step reaction, thereby eliminati ng the need for extremely high temperatures and also increasing the kine tics of hydrogen release. Lithium borohydride (LiBH4) and lithium amide (LiNH2) were chosen as the two primary hydride materials to be investigated for hydrogen storage in the form of complex hydrides. These two materials were chosen due to their high gravimetric hydrogen densities and their initial invest igation for hydrogen storage. LiBH4 has been studied extensively for its chemical properties as well as its hydrogen storage characteristics. It has been determined that LiBH4 possesses an orthorhombic crys tal structure , as is shown in Figure 4.1. Each Li+ ion is surrounded by four BH4 ions in a tetrahedral configuration. The gravimetric hydrogen density of LiBH4 is 18.5wt.%, significantly more than the material requirements set forth by the DOE. While hydrogen is generally not released until above 470oC , it was found that additives can reduce the hydrogen release temperature to as low as 200oC as is the case for SiO2 , but the kinetics of the reaction are very slow, making the material impractical for hydrogen storage use in automobiles.
108 Figure 4.1: Boron (top) and lithium (bottom) coordina tions in orthorhombic LiBH4 at room temperature . As early as 1910, it was found that Li3N reacts with hydrogen to form LiNH2 , though LiH is also formed as a by-product. LiNH2 has a theoretical hydrogen capacity of 8.1wt.% and releases hydroge n after melting at a temperature of 380oC. While the temperature is too high for hydrogen releas e, there has been significant improvement in the hydrogen behavior of LiNH2 by using either catalysts to reduce the hydrogen release temperature or by destabilizing LiNH2 with other compounds. Additionally, the release of ammonia (NH3) poses a large problem since NH3 can poison fuel cells. However, the formation of ammonia can be suppressed through the addition of either catalysts or by de stabilizing LiNH2 with other compounds. It was shown, for example, that the addition of LiH to LiNH2 can suppress any ammonia formation. This is accomplished due to the LiH reacting with NH3, which is a very fast reaction . Furthermore, the temperature of hydrogen release of LiNH2 has been reduced to approximately 150-250oC through the addition of TiCl3 .
109 There has been a lot of work in findi ng complex hydrides for hydrogen storage, the most important and recent of which ar e summarized in Figure 4.2 [66, 70-95]. While this figure certainly represents only a sm all portion of research performed on these systems, it nevertheless clearly illustrates that most materials either require temperatures that are too high for practical use of simply have a capacity that is too low. This figure summarizes mainly borohydrideand amide-base d materials for hydrogen storage, as this is the main focus of this dissertation. Figure 4.2: Hydrogen sorption capacity and te mperature of selected complex hydrides and chemical hydrides with DOE target range highlighted. While it appears that there are severa l materials that would meet the DOE guidelines, the amidoborane samples , such as LiNH2BH3 or NaNH2BH3 are non-
110 reversible, thereby making the systems impracti cal for mobile use, as required by the DOE. Mg(NH2)2 + 2MgH2 was found to release 7.6wt. % around room temperature during ball milling , but exhibited such a low enthalpy that a high pressure (much higher than that required by the DOE) would be required to rehydrog enate the material, thereby also making the material impractical for use as a reversible hydrogen storage system. A promising system for hydroge n storage has been magnesium amide (Mg(NH2)2) with a capacity of between 5.6wt.% and 9.2wt.% [79, 82, 94], though all of these systems require temperatures of close to 200oC with a reduction in capacity directly proportional to the reduc tion in temperature. By combining the advantages of some of these systems, namely the borohydride family of materials with the magnesium amid e systems, it is thought that a combinatorial effect can be achieved with a reduction in hydrogen sorption temperature, reversibility, as well as a high hydrogen capacity. The overall goal of the investigation of complex hydrides for hydrogen storage is to reduce the hydrogen release temperature, which can be accomplished by either reducing the pa rticle size, as is the case for MgH2 or by destabilizing the material through the addition of catalysts or other additives. Ball milling is the chosen processing tech nique, as this combines both chemical and mechanical synthesis of the material. By ball milling, a homogenous mixture with reduced particle size can be achieved, as schematically indi cated by Figure 4.3. The parent compounds are combined in the ball mill container and th rough milling at high speeds the materials grind each other down to smaller pa rticle size and produce a homogenous mixture, possibly with a new chemical composition.
111 Figure 4.3: Schematic mechano chemical s ynthesis approach for complex hydrides to reduce particle size and achieve a hom ogenous mixture of parent compounds. Furthermore, by combining several materials, whether they are in small quantities, so as to count as a catalyst, or in larger quantities, to be considered destabilizers, the activation energy for hydrogen release or absorption can be altered and ideally brought to a point where a low temp erature is enough to release the hydrogen. When a material is destabilized, it can react with the additive during dehydrogenation to form a new compound, one that requires a lo wer energy, as schematically shown in Figure 4.4. An additive, B, can allow th e hydrogenated material, A, to form an intermediate compound, AB, wh ile releasing hydrogen, ther eby lowering the activation energy required for desorption. Various nano additives are investigated for the complex hydride LiBNH as well as MgH2, which is added in larger quantities. The details are described in the specific upcom ing sections, as appropriate.
112 Figure 4.4: Destabilization of complex hydride through a dditives or catalysts. 4.2. Quaternary Complex Hydride LiBH4/LiNH2 (LiBNH) It has previously b een shown that LiBH4 combines with LiNH2 to form a new quaternary structure when either heated for several hours or milled for several hours. This new quaternary structure, identified as Li4BN3H10 is formed when LiNH2 and LiBH4 are combined in a 3:1 molar ratio . Wh ile the material was found to release approximately 6wt.% hydrogen at a temperature of well over 250oC, it was also observed that the structure is non-reversible, theref ore disqualifying it as a practical hydrogen storage system. Since this new structure exhi bits hydrogen release at a temperature lower than its parent compounds, it was chosen as the ideal material for investigation, as the addition of various materials as well as various processing conditions had not been previously reported. 4.2.1. Synthesis of the Quaternary LiBNH As mentioned in chapter 2, mechano chemical synthesis, or ball milling, is used to synthesize all complex hydrides investigated within this chapter. Since ball milling has A + H2 AH2 + xB ABx + H2 Dehydrogenated state activation energy to release hydrogen high Hydrogenated state Stabilized intermediate state activation energy to release hydrogen lower Activation Energy
113 many variables, such as rotati onal speed, sample mass to ball ratio, the type of purge gas (or lack thereof), as well as milling duration, these parameters had to be optimized before commencing milling. The milling speed was kept at a constant sp eed of 300 rpm, since this is the most commonly used milling speed employed and therefore the sample properties can be easily compared with other complex hydrides synthesized. The milling container used is made of stainless steel and has a volume of 80mL and contains 30 10mm diameter stainless steel balls. It was found that a total samp le mass of 2g was the optimum amount for milling. This amount was chosen as it was found that more than 2g of sample leads to agglomeration of sample in the mill, producing a smaller yield and incomplete mixing. Additionally, it was found that using an Ar/H2 (95%/5%) gas mixture with which the container was purged before milling commen ced and then every two hours during milling further led to a reduction in agglomeration of sample, thereby increasing the yield to virtually 100%. The use of this Ar/H2 gas mixture also enables the synthesis of larger sample masses, up to 4g, though the 2g limit was kept to keep the sy nthesis parameters constant. Finally, the most important milling variable was the milling time. If a sample is milled for a short time, the parent compounds only physically mix, which is used for catalyst addition, but there is no chemical r eaction. Since it was desired to produce a new chemical structure to eliminate the poor hydrogen sorption properties of LiBH4 and LiNH2, optimal milling duration was needed. Fi gure 4.5 shows the effect of milling duration on the quaternary material investig ated. The bottom two curves show the XRD spectrum of the parent compounds LiBH4 and LiNH2 in their as-receiv ed (non-purified)
114 form. After milling 2LiNH2 with LiBH4 for 30min, a new peak at approximately 29o and 48o emerge, indicating the formation of a new chemical structure. However, LiNH2 is still clearly visible. Therefore, the milling duration was increased to 1hr and finally to 5hr, as the 1hr milling still showed the presence of LiBH4. After 5hr of milling, there was no trace of the parent compounds, but new peaks we re observed, verifying the formation of a new chemical structure. A milling duration of more than 5hr did not yield any changes in the chemical structure of the compound, and therefore would only le ad to more costly and time-intensive synthesis. Figure 4.5: XRD pattern comp aring parent compounds LiBH4 and LiNH2 for different milling durations.
115 Fourier transform infrared spectroscopy was used to further characterize the effect of milling duration on the sample s. The parent compounds LiBH4 and LiNH2 are shown as the top two data sets in Fi gure 4.6. The FTIR spectrum of LiBH4 shows BH2 deformation bands at around 1100 and 1200cm-1 and B-H bonding stretches around 2300cm-1. LiNH2 has both symmetric and asymmetric amide stretches at around 3250 and 3300cm-1. After milling for 5hr, the new quaterna ry structure still exhibits the BH2 deformation bands, the B-H bond stretch an d the symmetric and asymmetric amide stretch. This indicates that th e new structure that is formed, a new chemical compound as was shown in Figure 4.5, is co mposed of borohydride and amides, but is a new chemical structure, meaning neither of the parent compounds are pres ent in their initial form. Additionally, it can be seen that the B-H stre tch peak widens significantly with increased milling time. This is advantageous to hydrogen sorption as there is essentially more space for the hydrogen to bond to the material.
116 Figure 4.6: FTIR comparison of parent compounds LiBH4 and LiNH2 with selected milling durations. 4.2.2. Hydrogen Characteristics of the Quaternary LiBNH The optimum quaternary structure, referred to from here on out as LiBNH, which consists of a 2:1 molar ratio of LiNH2:LiBH4, ball milled for 5hr, was investigated for its hydrogen sorption characteristics. As Figure 4.5 indicates, there is no trace of either of the parent compounds, LiBH4 and LiNH2, present and instead a new quaternary phase consisting of Li, B, N, and H was formed. BH2 deformation band B-H stretch Symmetric / asymmetric amide stretch
117 Figure 4.7: Hydrogen PCT desorp tion of quaternary LiBNH. After following the typical hydrogen sorp tion measurement procedures outlined in chapter 2, the hydrogen desorption and abso rption of the structure was investigated starting at a temperature of 175oC. As can be seen from Figure 4.7, there is no significant hydrogen release until approximately 250oC. Only approximately 0.5wt.% of hydrogen is released at 175oC and 200oC, while 0.7wt.% of hydrogen is released at approximately 225oC. Finally, at 250oC, about 3.5wt.% of hydrogen is rele ased from the material. It is interesting to note the presence of two hydrogen plateau regions, which most likely results from the release of hydrogen from borohydride and amide part of the new quaternary structure.
118 Figure 4.8: Hydrogen PCT absorp tion of quaternary LiBNH. Each desorption measurement was followed by an absorption PCT measurement at the same temperature. However, as can be seen from Figure 4.8, only less than 0.3wt.% of hydrogen could be reabsorbed, w ith more absorption occurring at lower temperatures, as should be expe cted. This is because a highe r temperature means that the material contains more energy, and therefor e the hydrogen bonds are broken. Because of the requirements of the DOE, it was decided not to investigate the hydrogen sorption of any material discussed as pa rt of this work above 250oC. Repetition of these measurements with other batches yielded the same results as those presented in Figure 4.7 and Figure 4.8. This means that the quatern ary structure that was formed is non reversible, thereby making it unusable for automotive applications. Additionally, the
119 kinetics of absorption, as can be seen in Fi gure 4.9, are too slow to be useful. It takes almost 2 hours to reabsorb only 0.7wt.% of hydrogen. Figure 4.9: Absorption kinetics at 250oC of LiBNH after dehydrogenation. 4.2.3. Activation Energy of the Quaternary LiBNH To understand why this new quaternary structure does not reversibly store hydrogen, it was decided to investigate the ac tivation energy of the structure. The TPD was used along with various heating rates to obt ain peak data of the material which could then be correlated using Kissinger's method, as described in chapter 2, to the activation energy of the material. Figure 4.10 shows the data obtained for heating rates of 1, 5, and 10K/min. It can be seen that there is no si gnificant hydrogen releas e until approximately
120 250oC, which verifies the PCT data previously discussed. As is typical of increased heating rates, the peak of the hydrogen evol ution shifts upward as the material does not have enough time to reach localized equilibrium. Figure 4.10: TPD plot of the quaternary LiB NH for heating rates of 1, 5, and 10K/min. The temperature of the peak of each h eating rate was then correlated to the activation energy by plotting the ln(heating rate/peak temperature) versus 1/temperature, as can be seen in Figure 4.11. This is ba sed on Kissingers equation, which relates the heating rate to the activa tion energy of sample as: d ln T m 2 d 1 T m = -E R (4.1)
121 where is the heating rate of the sample in K/min, Tm is the peak temperature in K, E is the activation energy of the hydrogen release, and R is the gas consta nt. By obtaining the slope of the best fit linear line, it was f ound that the activation energy for the hydrogen release of the quaternary structure is ap proximately 145kJ/mol. As was previously described, this activation energy is too high to allow for re versible hydrogen storage at usable temperatures. Therefore, the new qua ternary structure had to be destabilized further. As the ball milling was already optim ized, it was decided to investigate various nano sized additives to the material, as is described in the next section. Figure 4.11: Kissinger plot of th e quaternary structure LiBNH.
122 4.2.4. Destabilization of the Quaternary LiBNH with Nano Sized Additives Since the activation energy for the release of hydrogen of the quaternary structure was found to be 145kJ/mol, a value which is too high for practical reversible hydrogen sorption at usable temperatures, it was decided to investigate the eff ect of various nano sized additives to the quaternary structure. Si nce particle size plays an important role in hydrogen sorption, as previously mentioned, it was decided to use pre-prepared nano sized additives such as nickel (Ni) and zinc (Zn) in various concentrations as well as the co-addition of these additives. Initially, TGA measurements were performed to get an insight into the role that these nano additive s have on the quaternary structure and their hydrogen release. Figure 4.12 shows the TGA and DSC data of 1, 3, and 5 mol% of Ni. It can be seen that the amount of catalys t is important on the hydroge n release temperature of the material. While the DSC data does not show si gnificant difference, it can be seen that the addition of 5mol% of Ni provides for a slight decrease in hydrogen release temperature, as compared to 1 and 3mol%. All samples we re prepared in the sa me manner the nano additive was added to the bul k quaternary structure in the previously described manner for synthesis and, after purging with an Ar/H2 mixture, milled for 15 minutes. Since obtaining reliable TGA data prove d difficult due to contamination of the material with either oxygen or moisture, it was decided to use TPD to obtain information about the effect of additives to the hydrogen storag e material for any further materials.
123 Figure 4.12: TGA and DSC data of the quate rnary with 1, 3, and 5 mol% of nano Ni. One of the first samples that was inves tigated for its hydrogen storage behavior was the quaternary structure (LiBNH) with 3mol% of nano sized Zn. Figure 4.13 shows the desorption of hydrogen from the materi al with increasing temperature. Each desorption measurement was followed by a PC T absorption measurement at the same temperature. It can be seen that virtually no hydrogen is released from the material until approximately 250oC. At this temperature, however, 6wt.% of hydrogen is released in a three step reaction, compared to 3.5wt.% in a two-step reaction as was found for the quaternary structure, shown in Figure 4.7. Af ter this desorption, however, no hydrogen is reabsorbed, as can be seen in Figure 4.14. Conversely, no significant amount of hydrogen
124 is released in subsequent cycles, as show n by the second desorption PCT measurement in Figure 4.13. Figure 4.13: Hydrogen PCT desorption of LiBNH with 3mol% nano Zn.
125 Figure 4.14: Hydrogen PCT absorption of LiBNH with 3mol% nano Zn. According to the TGA optimization of nano sized Ni in Figure 4.12, the effect of the addition of 5mol% of nano Ni on the qua ternary LiBNH was investigated. Again, the hydrogen sorption measurements were investigated starting at 100oC, with each desorption followed by an absorption measurem ent at the same temperature. Figure 4.15 shows that the addition of 5mol% of Ni greatly reduces the temperature required for hydrogen release. At 175oC approximately 5.6wt.% of hydrogen is released. As can be seen, though, two separate desorption measuremen ts had to be performed. This is because the maximum time allowed for a desorpti on measurement is 24hr. Hence, while the temperature of hydrogen releas e was reduced from 250 to 175oC, the kinetics were also significantly reduced to well over 24hr for th e desorption of hydrogen. However, it can
126 be seen that, compared to the addition of Zn, the hydrogen release is performed in a onestep process with a plateau pressure less than 5bar of hydrogen pressure. As with the previous quaternary samples, however, th ere is virtually no r eabsorption of hydrogen within the material. Figure 4.15: Hydrogen PCT performan ce of LiBNH with 5mol% nano Ni. Since Ni not only reduced the temperatur e for hydrogen release by approximately 75oC, but also reduced the kine tics of hydrogen releas e, it was decided to investigate the additive effect of using both nano Ni a nd nano Zn. Initially, 3mol% of nano Zn and 3mol% of nano Ni were added to the quatern ary structure. Figure 4.16 shows that the addition of 3mol% of nano Ni and 3mol% of nano Zn does not improve the temperature of hydrogen release, as is seen for nano Ni. However, the three-step hydrogen release that
127 was observed with just nano Zn is preser ved, though the plateau pressures are reduced from 60, 35, and 2bar to 25, 18, and 5ba r of hydrogen pressure, an important improvement. However, the addition of Ni al ong with Zn did allow for almost 8.5wt.% of hydrogen to be released a significant impr ovement compared to any of the other materials. Despite this improvement, though, the ma terial is still not able to reabsorb the hydrogen that is released initially. The reab sorption of hydrogen essentially followed that of the quaternary structure, as depicted prev iously in Figure 4.8 or that of the Zn-doped quaternary that was shown in Figure 4.14. Figure 4.16: Hydrogen PCT performance of LiBNH with 3mol% nano Ni and 3mol% nano Zn.
128 Since a concentration of 5mol% Ni was found to greatly reduce the temperature for hydrogen release, while allowing for a on e-step hydrogen rele ase at about 5bar, another sample, this time with 5mol% Ni and 3mol% Zn was investigated for its hydrogen characteristics. Figure 4.17 shows that hydrogen is released in a one-step reaction at approximately 3bar of hydrogen pres sure with a total cap acity of 6.5wt.% at 200oC. Hence, the larger amount of Ni took over the hydrogen performance of the material and allowed for the reduction of temperature from 250 to 200oC while the Zn allowed for the increased kinetics, as compar ed to the quaternary structure containing only nano Ni. Figure 4.17: Hydrogen PCT performance of LiBNH with 5mol% nano Ni and 3mol% nano Zn.
129 4.2.5. Quaternary LiBNH Summary Through XRD measurements, it was found that a 2:1 molar ratio of LiNH2:LiBH4 formed a new quaternary compound after approx imately 5hr of ball milling. Furthermore, a purge with Ar/H2 (95%/5%) before milling, as well as every 2hr of milling ensures that the resultant material does not agglomerat e, thereby producing a homogenous yield of sample, provided that only 2g of sample are milled at a time. PCT hydrogen sorption measurements of the quaternary structure, LiBNH, indicate that a temperature of 250oC is required to release approximately 4wt.% of hydrogen in a two-step release at pressure of 25 and 15bar. The addition of 5mol% nano sized Ni was shown to lower the hydroge n release temperature from 250 to 175oC in a one-step release at a pressure of 5bar, but with kinetics that require more than 24hr to release the hydrogen. The addition of 3mol % nano sized Zn had no effect on the temperature, but allowed for a re lease of 6wt.% of hydrogen at 250oC at the cost of hydrogen desorption kinetics. Finally, the ad dition of 3mol% nano sized Zn and 3mol% nano sized Ni enabled the quaternary stru cture to release 9wt.% of hydrogen at 250oC in a three-step process. The main hydrogen results of the quaternary structure are summarized in Table 4.2 and in Figure 4.18. Table 4.2: Hydrogen performance of quaternar y LiBNH with and without nano Zn and nano Ni. Catalyst Hydrogen release temperature (oC) Hydrogen release pressure (bar) Hydrogen capacity (wt.%) None 250 25, 15 4 5mol% Ni 175 5 5.8 3mol% Zn 250 60, 35, 1 6 3mol% Zn + 3mol% Ni 250 28, 18, 5, 1 9 3mol% Zn + 5mol% Ni 200 2 5.2
130 Figure 4.18: Hydrogen performance of quatern ary LiBNH with and without nano Ni and nano Zn. 4.3. Destabilization of LiBH4/LiNH2 with MgH2 Magnesium hydride, MgH2, has a theoretical hydrogen capacity of 7.6wt.%, an ideal value for practical hydroge n storage applications. Howeve r, a temperature of 350 to 400oC is required to release hydrog en from this material. Additionally, the kinetics of release and uptake of hydrogen are too slow for practical use . It was found that hydrogen pressure change on MgH2 is the driving force for hydrogen absorption . This means that a higher pressure leads to a higher rate of absorp tion of hydrogen by pure magnesium. However, a limiting factor in th e rate of hydrogen, as well as the final capacity of hydrogen absorbed, is the formati on of a surface shell of magnesium hydride,
131 essentially a diffusion barrier layer, which pr events any further hydrogen uptake. This is found to especially be the case for pressu res above 30bar, where the rate of hydrogen uptake is found to be a maximum . The hydrogen absorption kinetics were found to be controlled by diffusion of hydrogen atoms [ 98], especially the diffusion of hydrogen in the hydride-metal interface. If th e hydride layer exceeds 30 to 50 m, hydrogen diffusion, and therefore uptake, is found to decrease du e to the coalescence of the hydride nuclei on the magnesium surface which form a compact hydride layer . In order to prevent this passivation laye r, which not only slows the uptake of hydrogen, but also prevents full hydrogenation of the magnesium, the particle size of MgH2 can be reduced so as to prevent the forma tion of this hydride laye r. If the particles are smaller than 30 m, the hydrogen diffusion should th erefore be able to continue allowing for more rapid and full hydrogen uptak e of magnesium. As already mentioned, the use of ball milling allows for the reduction of particle sizes. Henc e, it was decided to investigate the particle size reduction a nd some basic thermal studies of MgH2 that result from this purely physical aspect of mechano chemical synthesis. In order to keep the MgH2 hydrided during the milling process, the sample was kept in a hydrogen atmosphere by purging the sample holder with hydrogen before milling as well as every 2hr of milling for 15min. Solid state synthesis pertaining to the destabilization of LiBH4 and LiBH4/LiNH2  with MgH2 has been found to enhance th e reversible hydrogen storage characteristics. The multinary complex hydr ide Li-Mg-B-N-H possesses a theoretical hydrogen capacity of approximately 8 to 10wt.%. However, it has been reported that only about 3wt.% of hydrogen was reve rsibly released between 160-200oC [101, 102]. It was
132 reported that the MgH2 acts as a catalyst and assists in self-catalyzing the material to release hydrogen with three main reactions: ~ 175oC: 2Li4BN3H10 + 3MgH2 3Li2Mg(NH)2 + 2LiBH4 + 6H2 (4.2) ~ 200oC: Mg(NH2)2 + 2LiH Li2Mg(NH)2 + 2H2 (4.3) ~ 300oC: 3Li2Mg(NH)2 + 2LiBH4 2Li3BN2 + Mg3N2 + 2LiH + 6H2 (4.4) Keeping these aspects in view, the current study aims to develop high capacity nanocomposite multinary hydrides (e.g. Li-M g-B-N-H) by an inexpensive mechanochemical process. The effects of the proce ssing technique used to create the multinary hydrogen storage system have so far not been investigated. Therefore, the effects of commercial and nanocrystalline forms of MgH2 on the multinary hydride structure formation and overall hydrogen decomposition ch aracteristics are investigated to obtain a better understanding of the effects that pr ocessing conditions have on the hydrogen performance of the material. Additionally, the synergistic effects of nanocrystallinity on these hydrides have been investigated with a view to establis h structure-property relations. Extensive analytical tools, as we ll as activation energy evaluations, have been employed to obtain insights in to the nanocrystalline enha ncement of the hydrogenation properties in these solid state multinary hydrides. 4.3.1. Synthesis of the Multinary Hydrides The parent compounds, LiBH4 and LiNH2, were purchased from Sigma Aldrich with a purity of at least 95%, while MgH2 was obtained from Alfa Aesar with a purity of at least 98%. All materials were kept in the inert atmosphe re of the glove box and used without further purification. The investigated sample s were created in 4g batches with a
133 constant molar ratio of 2LiNH2:LiBH4:MgH2, while taking into account the purity of the parent compounds, by employing high energy ball milling (Fritsch Pulverisette 6) for 5hr at 300rpm with intermittent hydrogen/argon (5%/95%) purges for 20min before milling and after 2 and 4hr. This was done to ensure that as little hydr ogen as possible was released during the milling process and to re duce the agglomeration of the hydride that occurs when pure hydrogen is used as compared to the hydrogen/argon mixture, as was previously mentioned. The MgH2 was either added as received or was added as a socalled nano MgH2. The nano MgH2 (nMgH2) was created by ball milling the commercial MgH2 (cMgH2) for 12hr with intermittent hydrogen/argon purges every 2hr. This ensured the reduction of particle size as well as th e decrease in hydrogen release temperature, as previously reported . The two main proces sing schemes that were used are shown in Figure 4.19. The first processing scheme was to add all parent compounds and mill for 5 or 10hr using either commercial or nano MgH2. This is the scheme th at is generally used in reported literature and the materials serve as a sort of referenc e material. The second processing scheme was to first create the quaternary structure Li4BN3H10 (referred to as LiBNH in the rest of the text) by milling LiBH4 with 2LiNH2 for 5hr and then adding either commercial or nano sized MgH2, after which the quaternary and the MgH2 were milled for an additional 5hr. All milling was ca rried out in an inert atmosphere and the samples were purged with the hydrogen/argon mi xture every 2hr. In total, five different samples were created. The samples are referred to in this chapter according to the naming convention shown in the bold boxes of Figure 4.19.
134 Figure 4.19: Processing condition fl ow chart of the five samples investigated showing the two main processing schemes employed. 4.3.2. X-ray Diffraction Results Figure 4.20 shows the XRD pattern compar ing the five differently processed complex hydrides. The parent compounds, LiBH4, LiNH2, as well as both commercial and nano MgH2 are in the lower half of the figur e as a reference. The peak around 21o is from the Parafilm used to protect the samples during measurement. Neither LiBH4 nor LiNH2 peaks are observed in any of the five samples. This confirms that these two materials are fully consumed during the milling process and actually form a new quaternary structure, referred to as LiBNH. Th e quaternary structure has been reported to be Li4BN3H10 . The addition of commercial MgH2 does not cause the formation of a new complex structure, but instead indicates th at the quaternary structure is preserved, while the MgH2 simply intermixes with the ma terial. When the nano sized MgH2 is added
135 to LiBH4 and LiNH2 or to the quaternary LiBNH, the MgH2 peaks are barely picked up by the XRD. This indicates th at the small size of the MgH2 causes the material to intermix and fill voids of the quaternary st ructure, which results in a nanocrystalline particle distribution, while still preserving the quaternary structure formed by the LiNH2 and LiBH4. All samples are a physical, rather than a chemical, mixture of the quaternary structure LiBNH with MgH2. Figure 4.20: XRD profile of the five differently processed materials as well as the parent compounds, LiBH4, LiNH2, commercial MgH2, and nano MgH2. 4.3.3. FTIR Characterization Results The five differently processed samples we re characterized using FTIR to obtain information about the B-H and NH2 stretches and BH2 deformation, as seen in Figure
136 4.21. The data clearly shows that the amide, NH2 -, and borohydride, BH4 -, anions remain intact, as observed previous ly for the quaternary LiBNH sa mples . The peaks of the symmetric and asymmetric amide anions are shifted from the expected 3312 and 3259cm-1 to 3302 and 3244cm-1, respectively. Furthermore, the peak around 1560cm-1 is characteristic of th e amide ion. The B-H stretc hes, usually found at 2225, 2237, 2293, and 2387cm-1, overlap in the samples to form one large B-H stretch with a peak around 2320cm-1. Finally, the BH2 deformation peaks found at 1120 and 1092cm-1 in LiBH4 are observed at 1120 and 1082cm-1, respectively, though the peak around 1120cm-1 is extremely weak. However, there is no observable shift in any of the main stretches between the different samples, indicating that the chemical composition of the quaternary hydride is kept in tact, and there is in fact no formation of a new compound other than the previously reported structure . There is no evidence in the FTIR data indicating the r eaction of MgH2 with either the amide or borohydride, further confirming the XRD data shown in Figure 4.20.
137 Figure 4.21: FTIR spectra of the variously processed multinary hydrides. 4.3.4. Thermal Programmed Desorption Results Upon producing the complex hydrides, each sample was characterized for its thermal characteristics using TPD w ith a heating rate of 1, 5, 10, and 15oC/min. As compared to the quaternary structure, the multinary structure containing MgH2 showed a three-step hydrogen release mechanism, as is shown in Figure 4.22. This three-step release has previously been reported for the sample that is referred to as LicMgBNH in this paper. Our TPD analysis confirms the pr eviously reported data but also shows that the processing condition of the material does ha ve an effect on the thermal decomposition characteristics. The first hydr ogen release peaks between 157.7oC for the 10hr LicMgBNH and 165.2oC for LinMgBNH, which is a relatively small difference in
138 temperature. When investigating the second or main peak, of the various samples, it is interesting to note that the temperature range for main hydrogen release varies from 287oC for the 10hr LicMgBNH and 306.6oC for LinMgBNH. Figure 4.22: TPD comparison of investigated pr ocessing variations s howing the two main hydrogen release regions around 160oC and 300oC. 4.3.5. Activation Energy Results The activation energy of each sample was experimentally determined using Kissinger's method, based on TPD data taken at 1, 5, 10, 15oC/min for the two peaks and correlated to the hydrogen rele ase temperature. When invest igating the first peak, around 160oC, it is interesting to note that the 10hr LicMgBNH sample has the lowest activation
139 energy (109.8 kJ/mole) at 157.7oC, but LiBNH+nMgH2 has the lowest peak hydrogen release temperature (153.3oC), as seen in Figure 4.23. Although these samples exhibit comparable decomposition temperatures, their activation energies vary by ~20 kJ/mol. A plausible reason is that the reaction pathways of ad-mixing MgH2 either in the first place (10hr LicMgBNH) or after the quaternary formation (LiBNH+nMgH2) proceeds with fine distribution of nanocrystalline MgH2 in the host matrix of multinary hydrides. It has been recently claimed that the nanocrystallization of MgH2 has significant impact on lowering the enthalpy of formation and enhancement of the reaction kineti cs [46, 104]. The high temperature main hydrogen release peak (300oC) for all the processed materials and the reference LiBNH quaternary hydride are s hown in Figure 4.22. While comparing the activation energies and decomposition temperatur es of all the samples, it can be clearly inferred that the quaternary hydrides LiB NH combined with either commercial (LiBNH+cMgH2) or nanocrystalline MgH2 (LiBNH+nMgH2) milled for 5 hours show lower values, e.g. 145-148 kJ/mol at ~300oC. At this juncture, it is slightly difficult to justify from both Figure 4.23 and Figure 4.24, wh ich sample or the processed material possesses an optimum hydrogen release characte ristic at these two main decompositions. Again, it is undoubtedly clear that both the st eps occur at two different temperature regimes such as 160 and 300oC for drive-off the hydrogen eith er surface adsorbed or bulk absorbed species. At the low temperatur e first step hydrogen release, the nanoMgH2 acts as a catalyst to speed up the reaction, hence 10hr LicMgBNH and LiBNH+nMgH2 materials demonstrated lower activation energies (Figure 4.2 3). On the other hand, in the high temperature main hydrogen releas e (Figure 4.24), temperature of 300oC act as a driving force to release hydrogen from th e bulk structures of both LicMgBNH and
140 LinMgBNH milled only for 5hr. Hence, these materials exhibit lower activation energies which are comparable to pristine LiBNH. Based on the detailed analysis, we draw the conclusion that an additional 5hr of ball milling, either of the all-in-one hydride (10hr LicMgBNH) or the quaternary/nanocrys talline hydride mixture (LiBNH+nMgH2), will alter the decomposition characteristics, especi ally the activation energy which is very vital for hydrogen storage. Figure 4.23: Activation energy, as calculated from the TPD data using Kissingers method, compared with the first pe ak hydrogen release temperature.
141 Figure 4.24: Activation energy, as calculated from the TPD data using Kissingers method, compared with the main peak hydrogen release temperature. 4.3.6. Pressure-Composition-Temperature Isotherms Figure 4.25 represents the PCT isotherms of multinary complex hydrides created with different processing conditions. Th e dehydrogenation PCT of LiBNH quaternary hydride is plotted for reference. The PCT studi es of the multinary sa mples are carried out under the following conditions: temperature, T=150-175oC; pressure difference between aliquots, P=3bar; absorption pressure limit, Pa=80bar; desorption pressure limit, Pd=0bar; and reservoir volume, Vr=160cm3. Since all these samples are in hydride phases, the dehydrogenation experiment was followed by the rehydrogenation for at least
142 10hr. The PCT characteristics and their observati ons are given with respect to the sample processing conditions as follows. (1) Quaternary hydride, LiBNH: A dehydrogena tion capacity of ~4wt.% was found at 250oC. A two step plateau pressure region, Pp (low plateau at P<20bar and high plateau at P<30bar) was observed and might pertain to the tw o phase components of LiBH4 and LiNH2. Although the hydrogen releas e capacity of 4wt.% at 250oC seems promising, these quaternary hydrid es are not reversible at these temperatures. (2) LicMgBNH and LinMgBNH: The multinary complex hydrides processed with either commercial or nanocrystalline MgH2 and milled all-in-one for 5hr reveal reproducible hydrogen capacity of 3-4wt.%. It is noteworthy to mention that LinMgBNH possesses at least 1wt.% higher capacity and 25oC reduction in temperature as compared to the LicMgB NH counterpart. This could be achieved because of the uniform distribution of fine MgH2 nanoparticles which might act as catalytic centers for loweri ng the hydrogen dissociation temperatures. Yet another difference between these two processed mate rials is the tailoring of the plateau pressure (hydrogen/hydride equilibrium region), which is crucial for a hydrogen storage system to be viable for mob ile applications. The LinMgBNH material exhibits reduction in the absorption plat eau pressure by 20bar in contrast to the LicMgBNH due to nanoparticulate formation. (3) LiBNH+cMgH2 and LiBNH+nMgH2: Surprisingly, a greater reversible hydrogen storage capacity of 5.3-5.8wt.% was found at temperatures of 150-175oC for the quaternary hydrides LiBNH either milled with commercial or nano MgH2 for 5hr.
143 The nano MgH2 loaded LiBNH outperformed its commercial counterpart with a higher hydrogen capacity of 5.8wt.% at 150oC as compared to 175oC. There exist inflections of plateau pressure regions; how ever, they are not as clearly defined as the LicMgBNH and LinMgBNH samples. Moreover, the sorption plateau of these samples resembles greatly the pristine LiB NH which is the precursor material for the multinary hydride formation confir ming XRD data. Overall, it is unambiguously claimed that LiBNH admixe d either with commercial or nano MgH2 and milled for 5hr, exhibits a high reversible hydrogen storage capacity of ~6wt.% at temperatures less than 175oC. (4) 10hr LicMgBNH: The extended milling dura tion of 10hr for the three component systems, 2LiNH2+cMgH2+LiBH4 show poor hydrogen performance as depicted in Figure 4.25. A low hydrogen desorption capacity of 2wt.%, a low plateau pressure region of less than 5bar with less or no re versibility was obtained in this material. The crystallite agglomeration or th e amorphous phase during the prolonged milling is expected to be the limiting factor for the absence of plateau and overall storage capacity. Though these material s exhibit lower activation energy (~109kJ/mol), in the first hydrogen re lease (see Figure 4.23), the effective hydrogenation needs systematic optimization strategies which are currently under investigation.
144 Figure 4.25: Comparison of the hydrogen sorp tion characteristics of the various processing conditions at the lowest hydrogen release temperature. 4.3.7. Crystallite Size Effects on Hydrogen Capacity In order to better understand the hydr ogen performance of the differently processed materials, the hydrogen capacity was i nvestigated with respec t to the crystallite sizes of the quaternary pha se, LiBNH, and the MgH2. The crystallite sizes were calculated from the XRD data (Figure 4.20) of each material using Scherrers method . The initial crys tallite sizes of LiNH2, LiBH4, MgH2, nano MgH2 and LiBNH were determined to be 138, 152, 212, 27, and 60nm, respectively. As seen from Figure 4.25, the nano size MgH2 has an effect on the initial hydr ogen release temperature. Both samples synthesized with nano MgH2 release hydrogen at 150oC as compared to 175oC
145 for all the other samples, which were synt hesized with its comme rcial counterpart. The MgH2 crystallite size for the nano MgH2 samples are both approximately 10nm, whereas the crystallite size of the commercial MgH2 samples vary from 35-75nm, as seen in Figure 4.26. It is important to note th at the crystallite size of both MgH2 and LiBNH are largest for the 10hr LicMgBNH sample, whic h explains the poor hydrogen performance of the sample, since it is well known that larger particles, an d therefore a smaller surface area, correspond to poorer hydrogen performance (less than 2wt.% capacity). A milling duration of more than 5h r is in fact counterpr oductive and allows for the crystall ite size to increase, as both the LiBNH and MgH2 particles agglomerate. When looking at the correlation between crystallite size and hydroge n concentration, as s hown in Figure 4.26, it becomes evident that the size of the LiBNH crystallites plays an important role on the hydrogen concentration. If the LiBNH crystall ites have a size of approximately 28nm and the MgH2 crystallites are approximately within 15 nm (13-43nm) of this size, the highest possible hydrogen concentration is achieved (5.5wt.%). When the MgH2 and LiBNH crystallites are either too similar in size, as in the case for the LicMgBNH sample (3.3nm difference), or if they are t oo different in size, as in th e case for the 10hr LicMgBNH sample (32.9nm difference), the hydrogen conc entration of the sample is reduced.
146 Figure 4.26: Comparison of hydrogen concentrati on and crystallite size of the quaternary LiBNH and MgH2 phases after milling. 4.3.8. Surface Morphology In order to perform SEM imaging, the sample s were pressed into pellet form with a uniform thickness and pressure. The SEM images at 2,200 magnification are shown in Figure 4.27 for the five main samples. Figur e 4.27(a) and Figure 4.27(b) show the two samples that were milled for a total duration of 5hr. The sample prepared using the commercial MgH2 (Figure 4.27(a)) contains visibly la rger particles than the sample prepared using nano MgH2 (Figure 4.27(b)). This is due mainly to the fact that the commercial MgH2, which does not form a chemical b ond with the quaternary structure, but instead is intermixed with the material, is made up of larger particles than the nano
147 MgH2. When the sample is milled for 10hr, however, as seen in Figure 4.27(c), the material is composed of uniformly agglomerat ed nanoparticles. This is surely due to the fact that the commercial MgH2 has enough time, upon forma tion of the quaternary structure, to be ground into smaller particles and then given enough time to coalesce. When MgH2 was added to the quaternary structure LiBNH, the commercial MgH2 produced comparable particle size distributi on (Figure 4.27(d)) to th e sample containing nano MgH2 (Figure 4.27(e)).
148 Figure 4.27: Surface morphology of the five samples at 2200x magnification of (a) LicMgBNH, (b) LinMgBNH, (c) 10hr LicMgBNH, (d) LiBNH + cMgH2, and (e) LiBNH + nMgH2.
149 4.3.9. MgH2 Destabilized LiBH4/LiNH2 Summary A reversible hydrogen storage capacity of ~5-6wt.% was achieved at 150oC in the multinary complex hydrides comprising LiBH4, LiNH2 and MgH2. Various processing conditions and optimi zation strategies were adapted to prepare these complex hydrides in a solid state. Among the vari ous reaction pathways, the pre-processed quaternary hydride LiBNH ad -mixed with either bulk or nanocrystalline MgH2 (LiBNH+cMgH2 and LiBNH+nMgH2) milled for 5hr shows pronounced hydrogen storage characteristics in reversible sorpti on cycles and lower act ivation energy of 145148kJ/mol at ~300oC. Irrespective of the processing scheme employed, it was found that all samples were intimate mixtures of the quaternary st ructure, referred to as LiBNH, with MgH2 and that no new chemical compound was formed. It was found that the samples prepared with nano MgH2 exhibited MgH2 crystallites sizes of approximately 10nm, as compared to sizes of 35-75nm for those samples prepared with commercial MgH2. The small MgH2 crystallites enable the sample to releas e hydrogen at temperatures as low as 150oC as compared to 175oC for the larger MgH2 crystallites. Furthermor e, it was found that the size of the LiBNH crystallites plays an im portant role on the hydr ogen concentration. If the LiBNH crystallites ha ve a size of approxima tely 28nm and the MgH2 crystallites are approximately within 15nm (13-43nm) of this size, the highest possible hydrogen concentration is achieved. When the MgH2 and LiBNH crystallites are either too similar in size, as in the case of the LicMgBNH sa mple (3.3nm difference), or if they are too different in size, as in the case of the 10hr LicMgBNH sample (32.9nm difference), the hydrogen concentration of th e sample is reduced.
150 The structural, microstructural and th ermal desorption and activation energy calculations of all processed materials, as tabulated in Table 4.3, cumulatively suggest that synergistic effects of destabilization, nanocrystalli zation and process optimization lead to high hydrogen capacity materi als for tomorrows fuel cell cars. Table 4.3: Comparison of the results for the multinary complex hydrides developed by different processing cond itions (the best result s are shown in bold). LiBNH LicMgBNH LinMgBNH 10hr LicMgBNH LiBNH + cMgH2 LiBNH + nMgH2 TPD Temp. (oC) 1st Peak N/A 159.2 165.2 157.7 162.1 153.3 Main Peak 308.5 303.6 306.6 287.0 298.8 300.3 Activation Energy (kJ/mol) 1st Peak N/A 140.3 162.2 109.8 162.5 123.6 Main Peak 144.6 197.3 245.9 237.7 145.0 148.6 H2 Capacity (wt.%) 4.0 3.0 4.0 2.0 5.2 5.6 H2 Release Temp. (oC) 250.0 175.0 150.0 175.0 175.0 150.0 Plateau Pressure (bar) Abs. N/A <40 20 N/A <10 21 Des. 20-30 <10 10 <5 <5 <5 Reversibility No Yes Yes Less/No Yes Yes Crystallite Size (nm) LiBNH 60.0 38.7 21.0 42.1 29.0 28.1 MgH2 N/A 35.4 10.6 75.0 41.2 9.3 4.4. Destabilization of LiBH4/LiNH2/MgH2 with Nano Sized Additives Based on the previous experimental work on the optimization of the processing conditions as well as some previous insight in to the role of nano si zed additives on the quaternary LiBNH structure, it was decided to systematically investigate the effect of various nano sized additives on the multinary structure LiBNH + nMgH2, as described in the previous section.
151 4.4.1. Synthesis of Destabilized Multinary Hydride After the multinary hydride was synthesized, as described in chapter 4.3.1, various concentrations of nano additives were added to 0.4g of sample. The samples were loaded in the glove box and the same stai nless steel milling container used for other synthesis was employed. The samples were then purged with the same hydrogen/argon mixture, as previously described, for 15min before being milled. Initially, the samples were milled at 300rpm for 15min to obtain a comparison of the effect of the nano additives on the multinary hydride. This m illing duration allows for a thorough mixing of the parent compound with the na no additive without allowing the two to react and form a novel chemical structure. After a quick sc reening comparison of the various nano additives available, the samples that showed an increase in kinetics and the samples that showed a decrease in hydrogen release temp erature were chosen to be optimized. Initially, 2mol% of various nano sized additi ves, obtained from Q unatumSphere Inc., was investigated. The materials available were nickel, copper, manganese, cobalt and iron. 4.4.2. Thermal Programmed Desorption Results TPD was used to obtain information about changes in the parent materials hydrogen characteristics. Specifically, the te mperature of hydrogen release as well as some general information about the kinetics of hydrogen release ca n be ascertained from this type of measurement. The peak temperature indicates the optimal hydrogen release temperature, whereas the width of the peak can be used to get insight into the rate at which hydrogen is released, at least qualitatively. A wide peak indicates a low rate of hydrogen release, whereas a narrow and sharp peak indicates rapid hydrogen release. As
152 can be seen from Figure 4.28, it is clear that all of the additives allow for a lower hydrogen release temperature. As previous ly described, the parent compound, LiBNH + nMgH2, exhibits a three-step hydrogen release. While the TPD measurements are used for quick-screening the effect of the addi tives on the hydrogen performance of the material, it can be seen that each additive material either affects the rate of hydrogen release, as depicted by a sharp and narrow peak (especially iron) or significantly lowers the temperature required for hydrogen release. Figure 4.28: TPD comparison of LiBNH+nMgH2 without additive and with 2mol% Ni, Cu, Mn, Co and Fe at a constant ramping rate of 1oC/min.
153 4.4.3. Hydrogen Sorption Screening Results Since the TPD measurements only give an indication of the hydrogen sorption results, ramping kinetic measurements, wher e approximately 0.1g of sample are loaded into the PCT and then ramped at a rate of 1oC/min, were performed on all samples. Figure 4.29 shows the more detailed hydroge n performance of the standard sample, LiBNH+nMgH2 without any additives, as well as with 2mol% of the aforementioned additives. It becomes clear that while cobalt seemed promising from the TPD data, it in fact has such slow kinetics, that cobalt is no longer of interest as an additive. The kinetics measurement do confirm the TPD data in th at manganese and ir on have the fastest kinetics, as indicated by the slope of the de sorption curves. Furthermore, the significant reduction in hydrogen release temper ature of nickel is confirmed.
154 Figure 4.29: Ramping kinetics measurements of LiBNH+nMgH2 without and with 2mol% nano Mn, Fe, Co, Cu, Ni and Fe+Ni. Figure 4.30 shows a comparison of th e hydrogen release rate and hydrogen release temperature of the standard samp le without and with 2mol% of nano sized additives. Since the nano sized nickel show ed the lowest hydrogen release temperature of just under 200oC and nano sized iron showed the highest release rate (0.2wt.%/min) at a comparatively low temperature of 245oC, these two additives were chosen to be optimized in terms of their concentration.
155 Figure 4.30: Comparison of hydrogen release te mperature and hydrogen release rate of the standard LiBNH+nMgH2 and LiBNH+MgH2 with 2mol% of various nano additives. TPD measurements were used to obtain information about the effect of various concentrations of nano nickel and nano ir on, but are not shown here, as the optimum concentration of these additives were investigated for their hydrogen sorption properties. The effect of 2, 4, and 10mol% of nano nickel on the material can be seen in Figure 4.31. While 2mol% reduce the hydrogen release temp erature as previously mentioned, any further amount of nickel does not further lo wer this temperature, but instead leads to much slower kinetics of hydrogen release, as is evident by the hydrogen desorption of the standard sample with 10mol%.
156 Figure 4.31: Ramping desorption kinetics of LiBNH+nMgH2 without and with 2mol%, 4mol% and 10mol% nano nickel. Nano sized iron, on the othe r hand, maintains its kinetic advantage, but the hydrogen release temperature is further reduced with increasi ng iron amounts, as seen in Figure 4.32. A capacity of 6wt. % of hydrogen can be achieved with 0.2wt.%/min at a temperature of around 200oC for 10mol% nano iron, as compared to 6wt.% at 300oC at a rate of 0.08wt.%/min without any additive. This is a significant improvement and is most likely due the interaction of iron with ma gnesium which allows for a rapid hydrogen release at lower temperatures as the iron bonds with the magnesium, thereby forming a magnesium-iron alloy.
157 Figure 4.32: Ramping desorption kinetics comparing LiBNH+nMgH2 without and with 2mol% and 10mol% nano iron. Further detailed hydrogen cycling measur ements were being carried out on the 2mol% nano nickel and 10mol% nano iron samples, but the data is not presented in this dissertation due to time constraints. The data will be submitted to the International Journal of Hydrogen Energy and can be f ound when referencing the authors name. 4.5. Complex Hydrides Storage Summary A novel complex hydride system consisting of LiBH4, LiNH2, and MgH2 was developed and presented in this chapter. All starting compounds had a capacity higher than the DOE targets so that the combinati on of the three compounds could only lead to a material with a higher theore tical value than that require d for practical applications.
158 Initially, a new quaternary compound, referred to as LiBNH, was developed by optimizing the synthesis conditions of the so lid state mechano chemical milling. It was found that a 2:1 molar ratio of LiNH2:LiBH4 produced a new chemical compound when milled for 5hr at 300rpm in an 80mL stainles s steel container. The samples were purged before milling and every 2hr of milling with a 95%/5% Ar/H2 gas mixture which was found to reduce agglomeration of sample dur ing milling and to ensure the hydrogenation of the material. FTIR and XRD spectra of the samples indicate that the borohydride and amide stretches and bands are still intact, but that neither of the parent compounds are present. The quaternary struct ure exhibited hydrogen release at temperatures as low as 175oC, when destabilized with nano sized nickel, but irrespective of any additive, the quaternary structure was only able to reabsorb approximately 0.75wt.%. The quaternary structure was destabilized further with the addition of MgH2. Two main processing schemes were investigated: (1) The addition of nano sized MgH2 or commercial MgH2 to the previously formed quaternary structure. (2) The addition of nano sized MgH2 or commercial MgH2 to LiBH4 and LiNH2 in the stoichiometric ratio necessary to form the quaternary structure. It was found that the addition of MgH2 in a molar ratio of 1:1 LiBNH:MgH2 did not form any new chemical compound with the qua ternary structure, but that even in the case of the second processing scheme, the qua ternary structure was formed alongside the MgH2. All samples created with MgH2 were found to reversibly store hydrogen at temperatures as low as 150oC with capacities ranging from 2-5.5wt.%. The main driving force behind the hydrogen sorption properties of these materials was determined to be the
159 interaction of crystallite sizes of both the quaternary phase and the MgH2. In fact, a size difference of approximately 15nm was found to be optimum and was observed in the LiBNH+nMgH2 sample milled for 5hr. The second best performing material was found to be LinMgBNH milled for 5hr. Both samples with the nano sized MgH2 were able to release hydrogen at temp eratures as low as 150oC as compared to 175oC for the commercial MgH2 samples. A comparison of the materials created as part of this dissertation with previously publishe d results is shown in Figure 4.33. Figure 4.33: Complex hydride storage materi als comparison with other published materials. The addition of various nano-sized additive s, especially nickel and iron, further improve the hydrogen sorption behavior of the multinary hydride. While iron mainly
160 enhances the kinetics of hydroge n release in small (2mol%) co ncentrations, it can also significantly reduce the temper ature required for hydrogen release to close to 200oC. Nickel, on the other hand, has a large effect on reducing the temperature of hydrogen release in small concentrations (2mol%) but ha s no effect on the rate of hydrogen release. A larger concentration of nick el was found to have no further effect on the materials properties, but did lead to a significant re duction in hydrogen release kinetics due to the formation of a magnesium-nickel alloy.
161 Chapter 5 Conclusions and Future Work Recommendation 5.1. Overview A systematic study of two potential hydroge n storage systems has been carried out. The first system, polyaniline, was synt hesized in its bulk form as well as in nanostructured forms of nanofibers and nanosphere s. This system was investigated for its physisorption of hydrogen th rough weak secondary atom ic bonds. The second main system was investigated for its strong pr imary atomic bonding of hydrogen and consisted of LiBH4, LiNH2, and MgH2. Various nano sized additives were added to both the polyaniline and the complex hydride systems, though mainly the complex hydride system was found to destabilize to al low for hydrogen storage at lo wer temperatures and with faster kinetics. 5.2. Physisorption in Polyaniline Summary and Conclusions Polyaniline, in its emeraldine form was successfully synthesized through chemical means. The bulk, or as-synthesized form, as well as three different nanostructures of PANI were characterized for physical, ch emical and hydrogen sorption characteristics. It was found th at the inclusion of additives to the bulk form of bulk PANI had virtually no effect on its hydrogen sorptio n characteristics. Bu lk PANI was found to have a hydrogen capacity of less th an 0.5wt.% at a temperature of 100oC. It was observed that the capacity of hydrogen in this standard sample increased slightly with increasing
162 temperature. This increase with temperature is due to the expansion of PANI, which in turn allows for the hydrogen to diffuse r eadily into the polymer While others have claimed that standard PANI can absorb as much as 8wt.%, this was refuted by another group , as well as this research. The nanostructured PANI, in the form of nanospheres with an average diameter of 100nm, was synthesized by sligh tly modifying the chemical s ynthesis technique used to produce the standard PANI. FTIR analysis however, shows that the nanospheres do possess the same chemical composition as the standard PANI. While an initial hydrogen uptake of 5.5wt.% hydroge n was observed at 30oC, the capacity of each consecutive desorption cycle was reduced by roughly 1wt.%, though the desorbed hydrogen was reabsorbed. Since the kinetic s of hydrogen uptake were on th e order of several hours, with slower kinetics for each consecutive cycl e, it was found that the main driving force behind the hydrogen sorption of the nanosphere s was chemisorption, most likely in the form of hydrogen reacting with unterminated bonds that were a resu lt of the synthesis. Finally, a capacity of only 0.6wt.%, virtua lly identical to that of bulk PANI, was observed at a temperature of 90oC. SEM analysis showed th at the repeated hydrogen cycling caused the nanospheres to swell after hydrogen reaction and also to agglomerate until the sample finally had an appearance simi lar to bulk polyaniline. The interaction of hydrogen with the sample also led to the form ation of microcracks, as shown in Figure 3.16, which are caused by the diffusion of hydrogen through the material. By again slightly modifying the chemical synthesis technique, chemically grown PANI nanofibers were created. FTIR investig ation proved that again the nanofibers were of the emeraldine form. The fibers had an average diameter of 250nm with varying
163 lengths. While this does not technically qualify them as nanofibers, the term is still employed to indicate the sub micron size of the fibers. The nanofibers also exhibited a rough surface. Hydrogen cycling of the nanofibers showed that an initial capacity of 34wt.% was absorbed in less than 10min, while approximately 1-2wt.% hydrogen was released at room temperature. PCT measur ements performed on the nanofibers showed that roughly 3.5wt.% hydrogen was reversibly st ored in the nanostructures. The initial hydrogen absorption showed a hydrogen plat eau pressure around 30b ar, indicative of chemisorption. Through repeated cycling, though, the plateau pressure disappeared, though the capacity remain unchanged at 3.5w t.% for many cycles. SEM analysis of the chemically grown nanofibers after hydrogen cy cling showed no evidence of nanofibers. Instead, the fibers had virtua lly disappeared and a rather porous material remained. Surface area analysis revealed that the su rface area in fact remained unchanged after cycling. Unlike the nanospheres whose struct ure also disappeared, th e nanofibrous nature of the PANI allowed for it to agglomerate into a porous material, thereby losing the physical characteristics, but still keeping the su rface area. This is an important factor for hydrogen storage materials as was previously discussed. Finally, electrospun nanofibers were investigated for th eir hydrogen performance. By using a well know electrospi nning technique, the nanofib ers were synthesized with average diameters of 1.5 m and lengths of several microns. The nanofibers were, unlike the chemically grown nanofibers, rather smooth. While FTIR again confirmed the presence of the typical benzenoid and quinoi d rings of emeraldine base PANI, it was observed that the quinoid peak was much larger in relation to the other FTIR peaks. The
164 quinoid peak was also shif ted by approximately 50cm-1, making it unique among all the polyaniline structures investig ated. Hydrogen kinetics meas urements on the electrospun nanofibers at 30oC showed no hydrogen sorption, but absorption at 100oC revealed a very promising initial upta ke of over 10wt.% in less than 10m in. The desorption that followed showed an almost instantaneous release of 4wt.% hydrogen and another 4wt.% with much slower kinetics. Each consecutive absorption cycle reabsorbed the previously released hydrogen with slightly lower kinetics, but still in a matter of minutes rather than hours. The desorption capacity decreased by roughly 2wt.% for each cycle, but the fast rate of hydrogen release remained until the capacity of hydrogen sorption decreased to virtually zero after approximately 3 absorpti on / desorption cycles. While the fast kinetics of the absorption measurements clearly indicate that physisorption is taking place, the desorption cycles are clearly a combination of both chemisorption (slower kinetics) and physisorption (faster ki netics). Furthermore, breakage a nd swelling of the nanofibers was observed, though overall they do remain int act. Hydrogen PCT measurements on a new batch of this sample, however, revealed that approximately 8wt.% hydrogen could be reversibly stored in the el ectrospun nanofibers as long as the pressure was gradually increased or decreased. This slow exposure to hydrogen is essential in that the nanofibers are not exposed to pressure extremes, which can result in the destruction of the nanofiber morphology. Additionally, the hydrogen is allowed enough time to diffuse into the material and bond. While this method of storage requires a higher amount of time, it is still the most preferential method of stori ng hydrogen in polyani line. The capacity of 8wt.% along with the relatively low temperature of 100oC makes the electrospun polyaniline an ideal candidate for hydrogen storage.
165 It appears that the main factor in storing hydrogen in polyaniline is less the surface area, as was previously thought for po lymers, but in fact lie s more in the exact chemical composition. Unterminated bonds, which provide chemisorption sites for hydrogen are an excellent initial hydrogen stor age site, but require an amount of energy that is simply too high for practical purpo ses, therefore requiring high temperatures which would in fact destroy the chemical structure of the polyaniline. The relative intensity of the quinoi d ring vibration with respect to the other FTIR peaks appears to play a vital role in the hydrogen storage be havior of polyaniline as quinoid rings allow for a more readily delocalization of elect rons, thereby enabling the attachment of hydrogen atoms to the material. Also, the form of hydrogen storage, namely exposure of the material to hydrogen, is extremely impor tant when dealing with relatively weak materials such as PANI. Though it might requi re more time to store and release the hydrogen, a gradual pressure change is essential to maintaining the hydrogen characteristics of the material. The work on these polymers is the first to investigate various nanostructures of polymers at temperatures above 77K. A comp arison of the re sults obtained from this research with other work that was previ ously conducted on hydrogen storage in polymers is shown in Figure 3.35. While all the previously published polymers were investigated for pure physisorption at room temperature, the polymers investigated as part of this dissertation were also inves tigated for combinatorial physis orption / chemisorption of hydrogen. It was found that the many untermi nated bonds of the as-synthesized polymers allow for a rather large amount of hydrogen to be stored as part of the polymer. When adjusting the hydrogen cycling conditions, the hydrogen sorption behavior of the
166 materials was affected significantly. Therefore, it is important to not only consider the chemical structure of the material but also the testing methodology. 5.3. Complex Hydride Su mmary and Conclusions A reversible hydrogen storage capacity of 5-6wt.% was achieved at 150oC in the multinary complex hydrides comprising LiBH4, LiNH2 and MgH2. Through XRD measurements, it was found that a 2:1 molar ratio of LiNH2:LiBH4 formed a new quaternary compound after approximately 5 hour s of ball milling. Furthermore, a purge with Ar/H2 (95%/5%) before milling, as well as ev ery 2 hours of milling ensures that the resultant material does not agglomerate, thereby producing a homogenous yield of sample, provided that only 2g of sample are milled at a time. PCT hydrogen sorption measurements of the quaternary structure, LiBNH, indicates that a temperature of 250oC is required to releas e approximately 4wt.% of hydrogen in a two-step release at pressure of 25 and 15bar. The addition of 5mol% nano Ni was shown to lower the hydrogen release temperature from 250oC to 175oC in a onestep release at a pressure of 5bar, but with kinetics that require more than 24 hours to release the hydrogen. Th e addition 3mol% nano Zn had no effect on the temperature, but allowed for a release of 6wt.% of hydrogen at 250oC. Finally, the addition of 3mol% nano Zn and 3mol% nano Ni enabled the qua ternary structure to release 9wt.% of hydrogen at 250oC in a three-step process, requiring approximately 24hr. The main hydrogen results of the quaternary structure are summarized in Table 4.2 and in Figure 4.18.
167 Various processing conditions and optimizatio n strategies were adapted to prepare these complex hydrides in a solid state. Am ong the various reaction pathways, the preprocessed quaternary hydride LiBNH ad-mixed with either bulk or nanocrystalline MgH2 (LiBNH+cMgH2 and LiBNH+nMgH2) milled for 5 hours shows pronounced hydrogen storage characteristics in reversible sorpti on cycles and lower act ivation energy of 145148 kJ/mole at ~300oC. Irrespective of the processing scheme employed, it was found that all samples were intimate mixtures of Li4BN3H10 with MgH2 and that no new chemical compound was formed. It was found that the samples prepared with nano MgH2 exhibited MgH2 crystallites sizes of approximately 10nm, as compared to sizes of 35nm to 75nm for those samples prepared with commercial MgH2. The small MgH2 crystallites enable the sample to release hydrogen at temperatures as low as 150oC as compared to 175oC for the larger MgH2 crystallites. It was found that the size of the LiBNH cr ystallites is an essential factor in determining the amount of hydrogen that can be released at any given temperature. If the LiBNH crystallites have a size of approximately 28nm and the MgH2 crystallites are approximately within 15nm (13nm to 43nm) of this size, the hi ghest possible hydrogen concentration is achieved. When the MgH2 and LiBNH crystallites are either too similar in size, as in the case of the LicMgBNH sa mple (3.3nm difference), or if they are too different in size, as in the case of the 10hr LicMgBNH sample (32.9nm difference), the hydrogen concentration of the sample is re duced. Unlike most other research, which focuses on quick screening materials for th eir hydrogen capacity, the work described within this dissertation shows that even theoretical investigation of hydrogen storage
168 materials is insufficient in finding the optimum material. Factors such as the crystallite size are routinely ignored, ev en though they can double the amount of hydrogen stored by the material. The structural, microstructural and th ermal desorption and activation energy calculations of all processed materials, as tabulated in Table 4.3, cumulatively suggest that synergistic effects of destabilization, nanocrystalli zation and process optimization lead to high hydrogen capacity materi als for tomorrows fuel cell cars. 5.4. Future Work Recommendation While significant improvements have been made on the hydrogen sorption characteristics of the polyaniline system as well as the complex hydride system, more detailed chemical analysis into the exact mechanism of its hydrogen sorption characteristics are needed. Additionally, there has been some indication that NB2O5 can be used to further destabilize MgH2, which would be an interesti ng topic of research as an additive to the complex hydride that has been investigated and devel oped as part of this dissertation. More detailed investigation of polyanilin e and its various forms should be carried out to obtain proper insight in to the role of processing co nditions on the structure, both chemically as well as physically, as well as the hydrogen storage char acteristics of PANI. The complex hydrides investigated as part of this dissertation should definitely be investigated further and destab ilized, as all the storage characteristics appear to be extremely promising. Furthermore, a scaledup process should be designed and built to test the materials in real-life conditions.
169 Finally, theoretical modeling of both systems using state of the art molecular dynamics simulation would further be of great value to further understand the elaborate processes involved in hydrogen storage and thereby allowing for a quick screening and development of practical hydrogen storage systems.
170 References  V. S. Arunachalam and E. L. Fleischer, The global energy landscape and materials innovation, MRS Bulletin 33 (2008), pp. 264-275.  G. W. Crabtree and M. S. Dresselh aus, The hydrogen fuel alternative, MRS Bulletin 33 (2008), pp. 421-428.  E. Fontes and E. Nilsson, Modeling th e Fuel Cell, Industrial Physicist 7 (2001), pp. 14-16.  E. K. Stefanakos, D. Y. Goswami, S. S. Srinivasan, and J. T. Wolan, Hydrogen Energy, in Environmentally Conscious A lternative Energy Production, Kutz and Myer, Eds.: John Wiley & Sons, Inc., 2007, pp. 165-206.  S. Satyapal, J. Petrovic, and G. T homas, Gassing up with Hydrogen, Scientific American 296 (2007), pp. 80-87.  L. Schlapbach and A. Zuttel, Hydrogen-s torage materials for mobile applications, Nature 414 (2001), p. 353.  W. Grochala and P. P. Edwards, Ther mal Decomposition of the Non-Interstitial Hydrides for the Storage and Produc tion of Hydrogen, Chemical Reviews 104 (2004), pp. 1283-1316.  S. Satyapal, J. Petrovic, C. Read, G. Thomas, and G. Ordaz, The U.S. Department of Energy's National Hydrogen Storag e Project: Progress towards meeting hydrogen-powered vehicle requiremen ts, Catalysis Today 120 (2007), pp. 246256.  D. L. Greene, J. L. Hopson, and J. Li, Have we run out of oil yet? Oil peaking analysis from an optimist's perspective, Energy Policy 34 (2006), pp. 515-531.  R. W. Bentley, Global oil & gas deplet ion: an overview, Energy Policy 30 (2002), pp. 189-205.  R. A. Kerr, CLIMATE CHANGE: A Wo rrying Trend of Less Ice, Higher Seas, Science 311 (2006), pp. 1698-1701.
171  C. M. White, R. R. Steeper, and A. E. Lutz, The hydrogen-fueled internal combustion engine: a technical review, In ternational Journal of Hydrogen Energy 31 (2006), pp. 1292-1305.  L. Schlapbach and A. Zuttel, Hydrogen -storage materials for mobile applications, Nature 414 (2001), pp. 353-358.  A. Zttel, Materials for hydrogen st orage, Materials Today 6 (2003), pp. 24-33.  W. Hettinger, F. Michel, P. Ott, and F. Theissen, Refueling equipment for liquid hydrogen vehicles, International Journal of Hydrogen Energy 23 (1998), pp. 943947.  J. Wolf, Liquid-hydrogen technology fo r vehicles, MRS Bulle tin 27 (2002), pp. 684-687.  A. D. Lueking and R. T. Yang, Hydr ogen spillover to enhance hydrogen storage-study of the effect of carbon physicochemi cal properties, Applied Catalysis A: General 265 (2004), pp. 259-268.  R. Strbel, J. Garche, P. T. Mosele y, L. Jrissen, and G. Wolf, Hydrogen storage by carbon materials, Journal of Po wer Sources 159 (2006), pp. 781-801.  A. Chambers, C. Park, R. T. K. Baker, and N. M. Rodriguez, Hydrogen Storage in Graphite Nanofibers, The Journal of Physical Chemistry B 102 (1998), pp. 42534256.  A. C. Dillon, K. M. Jones, T. A. Bekk edahl, C. H. Kiang, D. S. Bethune, and M. J. Heben, Storage of hydrogen in singl e-walled carbon nanotubes, Nature 386 (1997), pp. 377-379.  X. Ma, E. Wang, W. Zhou, D. A. Jeffers on, J. Chen, S. Deng, N. Xu, and J. Yuan, Polymerized carbon nanobells and their field-emission properties, Applied Physics Letters 75 (1999), pp. 3105-3107.  Y.-Y. Fan, B. Liao, M. Liu, Y.-L. Wei, M.-Q. Lu, and H.-M. Cheng, Hydrogen uptake in vapor-grown carbon nanof ibers, Carbon 37 (1999), pp. 1649-1652.  P. Chen, X. Wu, J. Lin, and K. L. Tan, High H2 Uptake by Alkali-Doped Carbon Nanotubes Under Ambient Pressure and Moderate Temperatures, Science 285 (1999), pp. 91-93.  K. Yvon, Complex transition-meta l hydrides, Chimia 52 (1998), pp. 613-619.
172  H. E. Kissinger, Variatio n of Peak Temperature with Heating Rate in Differential Thermal Analysis, Journal of Research of the National Bureau of Standards 57 (1956), pp. 217-221.  S. J. Cho, K. S. Song, J. W. Kim, T. H. Kim, and K. Choo, Hydrogen sorption in HCl-treated polyaniline and polypyrrole: New potential hydrogen storage media, ACS Division of Fuel Chemistry, Preprints 47 (2002), pp. 790-791.  B. Panella, L. Kossykh, U. Dettlaff-Weg likowska, M. Hirscher, G. Zerbi, and S. Roth, Volumetric measurement of hydroge n storage in HCl-treated polyaniline and polypyrrole, Synthetic Metals 151 (2005), pp. 208-210.  N. B. McKeown, B. Gahnem, K. J. Msayib, P. M. Budd, C. E. Tattershall, K. Mahmood, S. Tan, D. Book, H. W. Langmi, and A. Walton, Towards PolymerBased Hydrogen Storage Materials: E ngineering Ultramicroporous Cavities within Polymers of Intr insic Microporosity13, Angewandte Chemie International Edition 45 (2006), pp. 1804-1807.  G. A. Banyay, M. M. Shaltout, H. Tiwari, and B. V. Mehta, Polymer and composite foam for hydrogen storage appli cation, Journal of Materials Processing Technology 191 (2007), pp. 102-105.  J.-Y. Lee, C. D. Wo od, D. Bradshaw, M. J. Ro sseinsky, and A. I. Cooper, Hydrogen adsorption in microporous hype rcrosslinked polymers, Chemical Communications (2006), pp. 2670-2672.  J. Germain, J. Hradil, J. M. J. Frechet, and F. Svec, High Surface Area Nanoporous Polymers for Reversible Hydr ogen Storage, Chemistry of Materials 18 (2006), pp. 4430-4435.  N. B. McKeown and P. M. Budd, Poly mers of intrinsic microporosity (PIMs): organic materials for membrane separa tions, heterogeneous catalysis and hydrogen storage, Chemical So ciety Reviews 35 (2006), pp. 675-683.  J. Stejskal and R. G. Gilbert, Po lyaniline. Preparation of a conducting polymer (IUPAC Technical Report), Pure and Applied Chemistry 74 (2002), pp. 857-867.  A. G. MacDiarmid and A. Epstein, J., Polyanilines: a novel class of conducting polymers, Faraday Discuss. Chem. Soc. 88 (1989), pp. 317-332.  J. Stejskal, P. Kratochv l, and A. D. Jenkins, The form ation of polyaniline and the nature of its structures, Polymer 37 (1996), pp. 367-369.  D. C. Trivedi and H. S. Nalwa, Handbook of Organic Conductive Molecules and Polymers. vol. 2 Chichester, England: Wiley, 1997.
173  F. Fusalba, P. Gouerec, D. Vill ers, and D. Belanger, Electrochemical Characterization of Polyaniline in Nona queous Electrolyte and Its Evaluation as Electrode Material for Electrochemi cal Supercapacitors, Journal of The Electrochemical Society 148 (2001), pp. A1-A6.  K. Roberg, G. Paasch, L. Dunsch, a nd S. Ludwig, The influence of porosity and the nature of the charge storage cap acitance on the impedance behaviour of electropolymerized polyaniline films, J ournal of Electroanal ytical Chemistry 443 (1998), pp. 49-62.  S. Virji, R. B. Kaner, and B. H. Weiller, Hydrogen Sensors Based on Conductivity Changes in Polyaniline Nanofibers, The Journal of Physical Chemistry B 110 (2006), pp. 22266-22270.  S. J. Cho, K. Choo, D. P. Kim, and J. W. Kim, H2 sorp tion in HCl-treated polyaniline and polypyrrole, Cataly sis Today 120 (2007), pp. 336-340.  J. Germain, J. M. J. Frechet, and F. Svec, Hypercrosslinked polyanilines with nanoporous structure and high surface area: potential adsorbents for hydrogen storage, Journal of Materi als Chemistry 17 (2007), pp. 4989-4997.  M. U. Jurczyk, A. Kumar, S. Sriniv asan, and E. Stefanakos, Polyaniline-based nanocomposite materials for hydrogen storag e, International J ournal of Hydrogen Energy 32 (2007), pp. 1010-1015.  N. B. McKeown, P. M. Budd, and D. Book, Microporous Polymers as Potential Hydrogen Storage Materials, Macromol ecular Rapid Communications 28 (2007), pp. 995-1002.  A. Nikitin, X. Li, Z. Zhang, H. Ogas awara, H. Dai, and A. Nilsson, Hydrogen Storage in Carbon Nanotubes through the Formation of Stable C-H Bonds, Nano Letters 8 (2008), pp. 162-167.  J. Y. Hwang, S. H. Lee, K. S. S im, and J. W. Kim, Synthesis and hydrogen storage of carbon nanofibers, Synt hetic Metals 126 (2002), pp. 81-85.  M. U. Niemann, S. S. Srinivasan, A. R. Phani, A. Kumar, D. Y. Goswami, and E. K. Stefanakos, Nanomaterials for Hydr ogen Storage Applications: A Review, Journal of Nanomaterials 2008 (2008).  Y. Wang and X. Jing, Formation of Polyaniline Nanofibers: A Morphological Study, The Journal of P hysical Chemistry B 112 (2008), pp. 11571162.
174  A. Z. Sadek, A. Trinchi, W. Wlodarski, K. Kalantar-Zadeh, K. Galatsis, C. Baker, and R. B. Kaner, A room temperature polyaniline nanofiber hydrogen gas sensor, IEEE Sensors 1 (2005), pp. 207-210.  A. A. Athawale and S. V. Bhagwa t, Synthesis and char acterization of novel copper/polyaniline nanocomposite and appl ication as a catalyst in the Wacker oxidation reaction, Journal of Applie d Polymer Science 89 (2003), pp. 2412-2417.  Y. Yang, J. Ouyang, L. Ma, R. J.-H Tseng, and C.-W. Chu, Electrical Switching and Bistability in Organic/ Polymeric Thin Films and Memory Devices, Advanced Functional Material s 16 (2006), pp. 1001-1014.  J. Deng, X. Ding, W. Zh ang, Y. Peng, J. Wang, X. Lon g, P. Li, and A. S. C. Chan, Carbon nanotube-polyaniline hybrid materials, European Polymer Journal 38 (2002), pp. 2497-2501.  R. Deshpande, A. C. Dillon, A. H. Mahan, J. Alleman, and S. Mitra, Hydrogen adsorption in single-walled and multi-wa lled carbon nanotubes grown in a hotwire CVD (Cat-CVD) reactor, Th in Solid Films 501 (2006), pp. 224-226.  G. J. Li, X. H. Zhang, and S. Kawi, Relationships between se nsitivity, catalytic activity, and surface areas of SnO2 gas sensors, Sensors and Actuators B: Chemical 60 (1999), pp. 64-70.  M. Trchov, I. Sedenkov, E. Tobolkov and J. Stejskal, FTIR spectroscopic and conductivity study of the thermal degrad ation of polyaniline films, Polymer Degradation and Stability 86 (2004), pp. 179-185.  S. Quillard, G. Louarn, S. Lefrant, and A. G. Macdiarmid, Vibrational analysis of polyaniline: A comparative study of leucoemeraldine, emeraldine, and pernigraniline bases, Physical Review B 50 (1994), p. 12496.  Y.-Y. Fan, A. Kaufmann, A. Mukasya n, and A. Varma, Singleand multi-wall carbon nanotubes produced using the fl oating catalyst method: Synthesis, purification and hydrogen up-ta ke, Carbon 44 (2006), pp. 2160-2170.  A. Huczko, Template-based synthesi s of nanomaterials, Applied Physics A: Materials Science & Pr ocessing 70 (2000), p. 365.  A. Rahy and D. J. Ya ng, Synthesis of highly conductive polyaniline nanofibers, Materials Letters 62 (2008), pp. 4311-4314.  D. Zhang and Y. Wang, Synthesis a nd applications of one-dimensional nanostructured polyaniline: An overview, Ma terials Science and Engineering: B 134 (2006), pp. 9-19.
175  H. I. Schlesinger and H. C. Brown, Metallo Borohydrides. III. Lithium Borohydride, Journal of the American Chemical Society 62 (1940), pp. 34293435.  O. Knacke, O. Kubaschewski, and K. Hesselmann, Thermochemical properties of inorganic substances, 2nd ed Berlin: Springer-Verlag, 1991.  B. Bogdanovic, R. A. Brand, A. Marja novic, M. Schwickardi, and J. Tlle, Metaldoped sodium aluminium hydrides as poten tial new hydrogen storage materials, Journal of Alloys and Compounds 302 (2000), pp. 36-58.  G. Sandrock and G. Thomas, The IEA/DOE/SNL on-line hydride databases, Applied Physics A: Materials Scie nce & Processing 72 (2001), pp. 153-155.  J. P. Souli, G. Renaudin, R. Cern, and K. Yvon, Lithiu m boro-hydride LiBH4: I. Crystal structure, Journal of Alloys and Compounds 346 (2002), pp. 200-205.  D. S. Stasinevitch and G. A. Egore nko, Russian Journal of Inorganic Chemistry 13 (1968), pp. 341-343.  A. Zttel, S. Rentsch, P. Fischer, P. Wenger, P. Sudan, P. Mauron, and C. Emmenegger, Hydrogen storage propertie s of LiBH4, Journal of Alloys and Compounds 356-357 (2003), pp. 515-520.  F. W. Dafert and R. Miklauz, Uebe r einige neue Verbi ndungen von Stickstoff und Wasserstoff mit Lithium., Diese Sitzungsberichte CXVIII (1910), pp. 981-996.  Y. H. Hu and E. Ru ckenstein, Ultrafast Reaction between LiH and NH3 during H2 Storage in Li3N, The Journal of Physical Chemistry A 107 (2003), pp. 97379739.  T. Ichikawa, S. Isobe, N. Hanada, and H. Fujii, Lithium nitride for reversible hydrogen storage, Journal of All oys and Compounds 365 (2004), pp. 271-276.  M. Aoki, K. Miwa, T. Noritake, G. Kitahara, Y. Nakamori, S. Orimo, and S. Towata, Destabilization of LiBH4 by mixing with LiNH2, Applied Physics A: Materials Science & Pr ocessing 80 (2005), pp. 1409-1412.  P. Chen, Z. Xiong, J. Luo, J. Lin, a nd K. L. Tan, Interaction of hydrogen with metal nitrides and imides, Nature 420 (2002), pp. 302-304.  Roshan P. S. Himashinie V. K. Di yabalanage, Troy A. Semelsberger, Brian L. Scott, Mark E. Bowden, Benjamin L. Davis, Anthony K. Burrell,, Calcium Amidotrihydroborate: A Hydrogen Stor age Material, Angewandte Chemie International Edition 46 (2007), pp. 8995-8997.
176  J. Hu, G. Wu, Y. Liu, Z. Xiong, P. Chen, K. Murata K. Sakata, and G. Wolf, Hydrogen Release from Mg(NH2)2-MgH 2 through Mechanochemical Reaction, The Journal of Physical Chemistry B 110 (2006), pp. 14688-14692.  E. Jeon and Y. Cho, Mechanochemical synthesis and thermal decomposition of zinc borohydride, Journal of All oys and Compounds 422 (2006), pp. 273-275.  A. S. Jun Yang, Donald J. Siegel, De vin Halliday, Andrew Drews, Roscoe O. Carter III, Christopher Wolverton, Gregor y J. Lewis, J. W. Adriaan Sachtler, John J. Low, Syed A. Faheem, David A. Lesch, Vidvuds Ozolinscaron,, A SelfCatalyzing Hydrogen-Storage Material 13, Angewandte Chemie International Edition 47 (2008), pp. 882-887.  J.-H. Kim, S.-A. Jin, J.-H. Shim and Y. W. Cho, Thermal decomposition behavior of calcium borohydride Ca(BH4 )2, Journal of Alloys and Compounds 461 (2008), pp. L20-L22.  J.-H. Kim, S.-A. Jin, J.-H. Shim, and Y. W. Cho, Reversible hydrogen storage in calcium borohydride Ca(BH4)2, Script a Materialia 58 (2008), pp. 481-483.  Y. Kojima, M. Matsumoto, Y. Kawai, T. Haga, N. Ohba, K. Miwa, S.-i. Towata, Y. Nakamori, and S.-i. Orimo, Hydrogen Absorption and Desorption by the LiAl-N-H System, The Journal of Physical Chemistry B 110 (2006), pp. 9632-9636.  H. Y. Leng, T. Ichikawa S. Hino, N. Hanada, S. Is obe, and H. Fujii, New MetalN-H System Composed of Mg(NH2)2 and LiH for Hydrogen Storage, The Journal of Physical Chemistry B 108 (2004), pp. 8763-8765.  Y. Chen, P. Wang, C. Liu, and H.-M. Cheng, Improved hydrogen storage performance of Li-Mg-N-H material s by optimizing composition and adding single-walled carbon nanotubes, Interna tional Journal of Hydrogen Energy 32 (2007), pp. 1262-1268.  Y. Nakamori, H. Li, K. Miwa, S.-i Towata, and S.-i. Orimo, Syntheses and Hydrogen Desorption Properties of Me tal-Borohydrides M(BH4)n (M=Mg, Sc, Zr, Ti, and Zn; n=2-4) as Advanced Hydrogen Storage Mate rials, Materials Transactions 47 (2006), pp. 1898-1901.  Y. Nakamori and S.-i. Orimo, Destab ilization of Li-based complex hydrides, Journal of Alloys and Compounds 370 (2004), pp. 271-275.  S.-i. Orimo, Y. Nakamori, J. R. Elis eo, A. Zuttel, and C. M. Jensen, Complex Hydrides for Hydrogen Storage, Chemical Reviews 107 (2007), pp. 4111-4132.
177  F. E. Pinkerton, G. P. Meisner, M. S. Meyer, M. P. Balog h, and M. D. Kundrat, Hydrogen Desorption Exceeding Ten Weight Percent from the New Quaternary Hydride Li3BN2H8, The Journal of P hysical Chemistry B 109 (2005), pp. 6-8.  G. Soloveichik, J.-H. Her, P. W. St ephens, Y. Gao, J. Rijssenbeek, M. Andrus, and J.-C. Zhao, Ammine Magnesium Borohydride Complex as a New Material for Hydrogen Storage: Structure and Pr operties of Mg(BH4)2-NH3, Inorganic Chemistry 47 (2008), pp. 4290-4298.  D. S. Stasinevich and G. A. Egorenk o, Russian Journal of Inorganic Chemistry 13 (1968), p. 341.  F. H. Stephens, V. Pons, and R. T. Baker, Ammonia-borane: the hydrogen source par excellence?, Dalton Tr ansactions (2007), pp. 2613-2626.  Todd B. Marder, Will We Soon Be Fueling our Automobiles with AmmoniaBorane?, Angewandte Chemie Inte rnational Edition 46 (2007), pp. 8116-8118.  J. J. Vajo, Journal of Physical Chemistry B 109 (2005), p. 3219.  G. Wu, Z. Xiong, T. Liu, Y. Liu, J. H u, P. Chen, Y. Feng, and A. T. S. Wee, Synthesis and Characterization of a Ne w Ternary Imide Li2Ca(NH)2, Inorganic Chemistry 46 (2007), pp. 517-521.  Z. Xiong, C. K. Yong, G. Wu, P. Chen, W. Shaw, A. Karkamkar, T. Autrey, M. O. Jones, S. R. Johnson, P. P. Edwards, and W. I. F. David, High-capacity hydrogen storage in lithium and sodium amidoboranes, Nat Mater 7 (2008), pp. 138-141.  Z. T. Xiong, J. J. Hu, G. T. Wu, Y. F. Liu, and P. Chen, Large amount of hydrogen desorption and stepwise phase tr ansition in the chemical reaction of NaNH2 and LiAlH4, Cataly sis Today 120 (2007), pp. 287-291.  T. L. Yongfeng Liu, Zhitao Xiong, Jian jiang Hu, Guotao Wu, Ping Chen, Andrew T. S. Wee, Ping Yang, Kenji Murata, Ko Sakata,, Synthesis and Structural Characterization of a New Alkaline Eart h Imide: MgCa(NH)2, European Journal of Inorganic Chemistry 2006 (2006), pp. 4368-4373.  G. W. Z. Xiong, J. Hu, P. Chen,, Ternar y Imides for Hydrogen Storage, Advanced Materials 16 (2004), pp. 1522-1525.  G. W. Z. Xiong, J. Hu, Y. Liu, P. Ch en, W. Luo, J. Wang,, Reversible Hydrogen Storage by a Li-Al-N-H Co mplex, Advanced Functiona l Materials 17 (2007), pp. 1137-1142.
178  A. Zaluska, L. Zalusk i, and J. O. Strm-Olsen, Na nocrystalline magnesium for hydrogen storage, Journal of All oys and Compounds 288 (1999), pp. 217-225.  B. Vigeholm, K. Jensen, B. Larsen, and A. S. Pedersen, Elements of hydride formation mechanisms in nearly spheri cal magnesium powder particles, Journal of the less-common metals 131 (1987), pp. 133-141.  B. Vigeholm, J. Kjoller, B. Lars en, and A. S. Pedersen, Formation and decomposition of magnesium hydride, J ournal of the less-common metals 89 (1983), pp. 135-144.  J. Ryden, T. Hjorvarsson, T. Ericss on, E. Karlsson, A. Krozer, and B. Kasemo, Unusual kinetics of hydride formation in magnesium-palladium sandwiches, studied by hydrogen profiling and quartz crystal microbalance measurements, Zeitschrift fuer Physikalische Chemie 164 (1989), pp. 1259-1260.  P. A. Chater, P. A. Anderson, J. W. Prendergast, A. Walton, V. S. J. Mann, D. Book, W. I. F. David, S. R. Johnson, and P. P. Edwards, Synthesis and characterization of amide-borohydrides: Ne w complex light hydrides for potential hydrogen storage, Journal of Alloys and Compounds 446-447 (2007), pp. 350354.  G. J. Lewis, J. W. A. Sachtler, J. J. Low, D. A. Lesch, S. A. Faheem, P. M. Dosek, L. M. Knight, L. Hallo ran, C. M. Jensen, J. Yang, A. Sudik, D. J. Siegel, C. Wolverton, V. Ozolins, and S. Zha ng, High throughput screen ing of the ternary LiNH2-MgH2-LiBH4 phase diagram, Jour nal of Alloys and Compounds 446-447 (2007), pp. 355-359.  J. Yang, A. Sudik, D. J. Siegel, D. Halliday, A. Drew s, R. O. Carter Iii, C. Wolverton, G. J. Lewis, J. W. A. Sachtler, J. J. Low, S. A. Faheem, D. A. Lesch, and V. Ozolins, Hydrogen storage pr operties of 2LiNH2 + LiBH4 + MgH2, Journal of Alloys and Co mpounds 446-447 (2007), pp. 345-349.  P. A. Chater, W. I. F. David, S. R. Johnson, P. P. Edwards, and P. A. Anderson, Synthesis and crystal structure of Li4BH4(NH2)3, Chemical Communications 23 (2006), pp. 2439-2441.  V. Berube, G. Chen, and M. S. Dre sselhaus, Impact of nanostructuring on the enthalpy of formation of metal hydrid es, International Journal of Hydrogen Energy 33 (2008), pp. 4122-4131.  P. Scherrer, Bestimmung der Groesse und der inneren Struktur von Kolloidteilchen mittels Roentgenstrahlen, Goettinger Nachrichten 2 (1918), pp. 98-100.
About the Author Michael Ulrich Niemann was born February 24th, 1980 in Bblingen (BadenWrttemberg), Germany as the fifth son of Dieter Horst Kunibert Jurczyk and Ursula Luzia Jurczyk (ne Niemann). He atte nded Kindergarten in Schopfloch (Kr. Freudenstadt) and then Higgins Elementa ry School in Poughkeepsie, NY until 3rd grade after which he finished the Grundschule in Schopfloch and the Progymnasium in Dornstetten, followed by 11th grade at the Wirtschaftsgymnasium in Freudenstadt, Germany. He obtained his Abitur equivalence at Venice High School in Venice, FL and then went on to the University of South Flor ida, Tampa, FL, USA. There, he obtained his Bachelor of Science degree in Mechanical Engineering and completed his Masters of Science in Mechanical Engineering with a th esis entitled Shape Based Stereovision Assistance in Rehabilitation Robotics under guidance of Dr. Rajiv Dubey. He joined Dr. Kumar and Dr. Stefanakos to pursue his Ph.D. in mechanical engineering, of which this dissertation is the culmination. Michael married Christine Victoria Woodard during his Ph.D. studies on January 5th, 2005 in Bogota, Colombia. Both Christin e and Michael are now proud parents of Sebastian Eduard Gerhard Niemann, who was born on February 27th, 2009 in Tampa, Florida. Upon graduation, Michael plans to return to his native Germany to pursue a career in industry, though a position in academia is not out of the question.