GAS SORPTION INENGINEERED CARBONNANOSPACESA Dissertationpresented tothe Faculty of the Graduate Schoolat the University of MissouriIn Partial Fulfillmentof the Requirements for the DegreeDoctor of PhilosophybyJACOB W. BURRESSDr. Peter Pfeifer, Dissertation SupervisorAUGUST 2009

Copyright by Jacob W. Burress 2009All Rights Reserved

The undersigned, appointed by the Dean of the Graduate School, have examined theDissertation entitledGAS SORPTION IN ENGINEEREDCARBON NANOSPACESpresented by Jacob Burress, a candidate for the Degree of Doctor of Philosophy, andhereby certify that, in their opinion, it is worthy of acceptance.Professor Peter PfeiferAssociate Professor Carlos WexlerProfessor Galen SuppesProfessor Paul MiceliProfessor Haskell TaubProfessor Frederick Hawthorne

I dedicate this thesis to my loving wife, Molly.Trust in the Lord with all your heartAnd lean not on your own understanding;In all your ways acknowledge him,And he will make your paths straight.Proverbs 3:5, 6

ACKNOWLEDGEMENTSAcknowledgement goes first and foremost to the members of my committee: Drs.Peter Pfeifer, Carlos Wexler, Galen Suppes, Paul Miceli, Haskell Taub, and FrederickHawthorne.The research presented in this dissertation was funded in part by the followinggrants: U.S. Department of Defense, Navy, NAVSEA Warfare Center, N00164-07-P1306, U.S. Department of Energy, Basic Energy Sciences, DE-FG02-07ER46411, U.S. Department of Energy, Argonne National Laboratory, Advanced PhotonSource, DE-AC02-06CH11357, U.S. Department of Energy, Energy Efficiency & Renewable Energy, DE-PS3608GO98004P, National Science Foundation, Partnership for Innovation, PFI-0438469, University of Missouri, MU Research Board, RB-06-040, and California Energy Commission, Contract 500-08-022.I would also like to acknowledge the following individuals who have worked onvarious aspects of this research: Lauren Aston, Lin Bai, Sarah Barker, Matthew Beckner,Michael Benham, Sam Bowman, Phil Buckley, Sara Carter, Raina Cepel, Joe Clement,Anne Dillon, Elmar Dohnke, Carol Faulhaber, Lucy Firlej, John Flavin, MonikaGolebiowska, Michael Gordon, Tushar Gosh, Lacy Hardcastle, Frederick Hawthorne,Steve Hubbard, Jan Ilavsky, Satish Jalisatgi, Michael Kraus, Bogdan Kuchta, Nickii

Kullman, Cintia Lapilli, Han Baek Lee, Mark Lee, Erik Leimkuehler, Erik Nordwald,Patrick O’Keefe, Philip Parilla, Peter Pfeifer, Jeffrey Pobst, Sam Potts, Jimmy Romanos,Lou Ross, Bryan Sawyer, Parag Shah, Yuchoong Soo, Galen Suppes, Demetrius Taylor,Matthew Taylor, Ali Tekeei, Carlos Wexler, Henry White, and Michael Wood. Theseindividuals represent the Departments of Physics and Astronomy, Chemical Engineering,and Chemistry at the University of Missouri; the Midwest Research Institute; theNational Renewable Energy Laboratory; Quantachrome Instruments Inc. and HidenIsochema Ltd.I would like to give special thanks to Dr. Peter Pfeifer, Dr. Galen Suppes, PhysicsMachine Shop Supervisor Sam Potts, and Dr. Carlos Wexler. Dr. Wexler and Dr. Suppeshave been a tremendous help in advancing the research presented through their leadershipof the theoretical and chemical engineering aspects of the project, respectively. SamPotts has been a tremendous help from day one of my joining the project. He entertainedevery experimentalist idea I had and was always able to take the ideas from paper toapparatus. He is and will continue to be the master of metal. Dr. Pfeifer has given hisinvaluable skills and time to this project as the leader of ALL-CRAFT and primaryadvisor to this research. This research would not have been successful if not for histremendous foresight and leadership.iii

TABLE OF CONTENTSAcknowledgements . iiTable of Contents . ivList of Figures . ixList of Tables . xiiList of Symbols and Abbreviations . xiiiAbstract . xviChapter 1: Motivation and Research Problem . 11.1 Motivation . 11.2 Statement of Research Problem . 5Chapter 2: Background . 72.1 Physisorption for Methane and Hydrogen Storage . 72.2 Activated Carbons for Gas Storage . 112.3 ALL-CRAFT . 12Chapter 3: Conceptual Framework . 153.1 Interaction of Gases with Solid Surfaces . 153.1.1 Intermolecular forces . 153.1.2 Physisorption Modeling. 173.1.3 Sorption inside Pores . 193.2 Langmuir Theory of Gas Adsorption . 243.3 Brunauer, Emmett, Teller Theory . 283.4 Density Functional Theory . 30iv

3.4.1 Non-local Density Functional Theory . 313.4.2 Quenched Solid Density Functional Theory . 323.5 Hypothesis I: Carbon Precursors . 333.6 Hypothesis II: KOH Activation Effect on Pores . 333.7 Hypothesis III: Pore Optimization for Gas Storage . 333.8 Hypothesis IV: Pore Optimization for Surface Area . 343.9 Hypothesis V: Adsorption Physics . 34Chapter 4: Methods and Procedures . 364.1 Electron Microscopy Methods . 364.2 Chemical Composition Analysis Methods . 384.3 General Features of Gas Sorption Measurements . 394.3.1 Manometry and Gravimetry . 394.3.2 Sorption Measurement Quantities . 404.3.3 Volumes and Densities . 424.4 Manometric Subcritical Nitrogen Sorption Methods . 454.4.1 Autosorb: Sample Preparation . 454.4.2 Autosorb: Dosing Procedure . 474.4.3 Autosorb: Critical Features. 494.4.4. Autosorb: Sample Characteristics Determined . 514.5 Manometric Supercritical Hydrogen Sorption Methods . 524.5.1 Rapid-screening Hydrogen Sorption Instrument . 534.5.2 Hiden Isochema HTP1-V Mark II . 554.5.3 HTP1: Sample Preparation . 564.5.4 HTP1: Dosing Procedure . 57v

4.5.5 HTP1: Critical Features . 584.5.6 HTP1: Sample Characteristics Determined . 594.6 Manometric Supercritical Methane Sorption Methods . 594.6.1 Manometric Methane: Sample Preparation . 604.6.2 Manometric Methane: Reservoir Procedure . 604.6.3 Manometric Methane: Critical Features . 614.6.4 Manometric Methane: Sample Characteristics Determined . 624.7 Gravimetric Supercritical Methane Sorption Methods . 634.7.1 Gravimetric Methane: Sample Preparation . 644.7.2 Gravimetric Methane: Gravimetric Method . 644.7.3 Gravimetric Methane: Critical Features . 654.7.4 Gravimetric Methane: Sample Characteristics Determined . 66Chapter 5: Experimental Results . 685.1 Electron Microscopy Results . 685.2 Chemical Composition Analysis Results . 695.3 Manometric Subcritical Nitrogen Sorption Results . 725.4 Supercritical Hydrogen Sorption Results . 765.5 Manometric Supercritical Methane Sorption Results . 795.6 Gravimetric Supercritical Methane Sorption . 80Chapter 6: Conclusions . 846.1 Conclusion I: Carbon Precursors . 846.2 Conclusion II: KOH Activation Effect on Pore Structure . 856.3 Conclusion III: Pore Optimization for Gas Storage . 896.4 Conclusion IV: Pore Optimization for Surface Area . 93vi

6.5 Conclusion V: Adsorption Physics . 946.5.1 Hydrogen: Adsorption Equilibria . 956.5.2 Hydrogen: Binding Energies . 966.5.3 Hydrogen: Experimental Verification of Cross-sectionalAdsorption Areas . 996.5.4 Hydrogen: Sorption Kinetics . 1006.6 General Conclusions . 102Chapter 7: Future Research and Recommendations . 1037.1 New Projects . 1037.1.1 Boron-doping . 1037.1.2 Fission Tracks . 1047.2 New Instrumentation . 1047.2.1. Manometric Supercritical Hydrogen Sorption Apparatus . 1047.2.2. Manometric Supercritical Methane Sorption Apparatus . 1057.2.3 Advanced Outgassing Apparatus . 1057.3 Alternative Experimental Methods . 1067.3.1 Manometric Subcritical Nitrogen Sorption . 1067.3.2 Manometric Supercritical Hydrogen Sorption: Rapidscreening Hydrogen Sorption Instrument . 1067.3.3 Manometric Supercritical Hydrogen Sorption: HidenIsochema HTP1-V Mark II . 1067.3.4 Manometric Supercritical Methane Sorption . 1077.3.5 Gravimetric Supercritical Methane Sorption. 107Appendix A: Determination of Porosity . 109Appendix B: Gravimetric Methane Procedure . 111vii

Appendix C Processing of Gravimetric Data . 122Appendix D: Processing of Manometric Data . 134Appendix E: Hydrogen Test Fixture Procedure . 138Appendix F: Buoyancy Corrections . 143References . 146Vita . 150viii

LIST OF FIGURESFigurePage1.1.1. Breakdown of 2007 U.S. Energy consumption .21.1.2. Sources and uses for natural gas .42.1.1. Summary of hydrogen sorption materials as of 2009 DOE AMR meeting. .112.3.1. Overview of methane project. Image Courtesy of the National ScienceFoundation, by Nicole Rager Fuller.132.3.2. Kansas City fleet vehicle and carbon monolith .143.1.1. Plot of a model Lennard-Jones potential .163.1.2. Three phase approximation for adsorption. .183.1.3. Model slit-shaped pores .193.1.4. Cartoon depicting methane molecules adsorbed in slit-shaped carbon .213.1.5. Cartoons depicting 77 K hydrogen molecules adsorbed in slit-shaped carbon .223.1.6. IUPAC isotherm and hysteresis classifications .233.2.1. Chi dependence of Langmuir adsorption isotherms .253.2.2. Corrugated surface potential with adsorbate energies for localizedand mobile sorption.263.4.1. Pore size distributions from NLDFT and QSDFT for samples S-33/kand Batch 5.1 .324.1.1. Comparison of secondary and backscattered electron detectionin scanning electron microscopy.374.3.1. Comparison of adsorption definitions; absolute, excess and stored .414.3.2. Cartoon of carbon grains showing volume definitions .43ix

4.4.1. Cartoon depicting the functionality of the Autosorb 1-Cand photo of the Autosorb 1-C.484.5.1. Cartoon depicting the functionality of the rapid-screening instrumentand photo of the instrument .544.5.2. Cartoon depicting the functionality of the HTP1 and photo of the HTP1 .574.6.1. Picture of carbon monolith and corncob precursor. .604.6.2. Cartoon depicting the functionality of the methane test fixtureand photo of the methane test fixture.614.7.1. 35 bar gravimetric methane apparatus .634.7.2. 250 bar gravimetric methane apparatus .645.1.1. SEM images of sample S-33/k .685.1.2. SEM and TEM images of microtomed S-56.695.2.1. Backscatter electron image of briquette 46 .705.2.2. EDS analysis of briquette 46.715.3.1. Subcritical nitrogen sorption isotherms and QSDFT analysis ofsamples S7 and S-33/k .755.3.2. QSDFT PSD’s for three carbons produced using the same activationprocess using corncob precursor. .765.4.1. Results from the rapid-screening hydrogen sorption apparatuson S-33/k .775.4.2. Excess and stored hydrogen uptake isotherms on S-33/k and Batch 5.1 .785.5.1. Manometric, room temperature stored methane isotherm on briquette 46 .795.6.1. Gravimetric, room temperature excess methane adsorbed isothermson samples B-21/k and S-33/k and summary of ANG versus CNG .825.6.2. Comparison of methane and nitrogen NLDFT pore size distributionsof S-33/k.836.1.1. Comparison of nitrogen sorption analysis of 3K and S4 .856.2.1. KOH activation impact on nanoporosity and porosity .86x

6.2.2. Subcritical nitrogen isotherms of samples with differingKOH concentrations.876.2.3. QSDFT PSD’s for carbons of increasing KOH ratio .886.2.4. Nominal BET surface area dependence on KOH concentration .896.3.1. Comparison of pore characteristics and excess methane uptake .906.3.2. Excess mass adsorbed methane for mass sample compared with surface area .916.3.3. Comparison of KOH:carbon ratio and methane uptake .926.4.1. Nominal surface area compared to other pore characteristics .936.5.1. Theoretical excess isotherms for localized and mobile hydrogensorption compared to experiment.956.5.2. Subcritical nitrogen sorption analysis on S-33/k and Batch 5.1 .966.5.3. Hydrogen single and two binding energy fits for S-33/k and Batch 5.1.987.2.1. New room temperature manometric supercritical hydrogensorption apparatus .105xi

LIST OF TABLESTablePage2.1.1. Summary of DOE targets for hydrogen as of March 24, 2009 .82.1.2. Summary of DOE targets for methane and hydrogen .92.1.3. Comparative characteristics of large ANG projects up to 2006 .104.3.1. Names of densities used, and the volumes included in that density .445.2.1. ICP-AES and XRF results on sample S-33/k .725.3.1. List of most carbons analyzed using subcritical nitrogen adsorption .735.3.2Comparison of the three analysis techniques: BET, NLDFT and QSDFTon S-33/k and Batch 5.1 .765.4.1. Validation of H2 storage in new corncob-based nanoporous carbonfrom three independent laboratories .785.6.1. Methane uptake results for 70 samples at 35 bar and room temperature.806.5.1. Low and high binding energies and respective fractions of sitesfor S-33/k and Batch 5.1 .996.5.2. Values of (T ) from computer simulations, and experimentalupper bounds .996.5.3. Characteristic times of adsorption and desorption at 77 K androom temperature for different binding energies. .101xii

LIST OF SYMBOLS AND ABBREVIATIONSSymbol orAbbreviationUnitsDescriptionAGLARGAtlanta Gas Light Adsorbent ResearchGroupANGAdsorbed natural gas, typically 500 psigmaximum tank pressure (T )Å2 or nm2BETBET constantCompressed natural gas, typically 3,600psig maximum tank pressureCCNGDDOEEBSurface area per site, Area per adsorbedmoleculeBrunauer, Emmett, TellerÅ or nmkJPore dimension (width of slit)Department of Energy, United States ofAmericaBinding energymolEDSEnergy dispersive spectroscopyf(D)FESEMRelative distribution of pore widthsField emission scanning electronmicroscopyPlanck’s constanth6.626 10 34 m2 kg s 1HTP-1VIGAk1.381 10 23 m2 kg s 2 K 1Hydrogen manometric instrument, HTP1V Mark II, produced by Hiden IsochemaHydrogen gravimetric instrument,Intelligent Gravimetric Instrument,produced by Hiden IsochemaBoltzmann’s constantLAILangmuir adsorption isothermLN2Liquid nitrogenMass of absolute adsorbed gasM Adsgxiii

MCM C,GggM C,SgM C,S,GggMass of evacuated chamberMass of sample cell with gas to beanalyzed (G H2, CH4) at desiredtemperature and pressureMass of evacuated sample cell withoutgassed sampleMass of chamber, sample and gasMass of excess adsorbed gasM H2gMass of the H2 moleculeM MonogMass of one monolayer of adsorbed gasExM AdsMOFMSM StoredMetal organic frameworksggSample massMass of gas stored in the pores, includesboth adsorbed and non-adsorbed gasNGNumber of gas admolecules on theadsorbentAvogadro’s constantNumber of admolecules constituting amonolayerNatural gasNGVNatural gas vehicleNLDFTNon-local density functional theoryPressure of gasNNANMonopbarp0barSaturation pressure of the gas species insubcritical adsorption (coexistencepressure of bulk liquid/gas at a giventemperature)PdPalladiumpsiPounds per square inchPSD NanoPore size distribution(s)PorosityNanoporosityDensity of adsorbed filmApparent density, all pores included (alsoreferred to as monolith density)Density of the bulk gas (adsorptive)Skeletal densityDensities of the gas in the supply tankbefore and after adsorption, respectively Ads Apparent Gas SkeletalSTST Initial, Finalg/cm3g/cm3g/cm3g/cm3g/cm3xiv

TT Finalg/cm3Density of the final gas in the test tankm2/gKQuenched solid density functional theoryScanning electron microscopySpecific surface area for mass sampleTemperatureTransmission electron microscopyLAI CoverageQSDFTSEM TTEM ( p, T )VApparentcm3VAdscm3Apparent volume, taken as convex hull ofmonolithVolume of the adsorbed filmVNanoporecm3Volume of nanoporesVSkeletal x , y , zV ( x, y, z )cm3cm3cm3cm3HzkJ (T )bar-1Skeletal volumeVolume of the supply tankTotal open pore volumeVolume of the test tankVibrational frequencies of admoleculeAdsorption potential at position (x,y,z),where z is the coordinate perpendicular tothe surface, and x, y are the coordinatesparallel to the surfaceLangmuir constantVSTVTotalPoreV TTxv

AbstractVehicular storage of gaseous fuels is a key enabling technology for the two pillarsof a non-petroleum based transportation economy as envisaged under various federal andstate alternative fuel plans: 1) natural-gas vehicles (internal combustion engines) and 2)hydrogen fuel cell vehicles—. My research focuses on the development of nanoporouscarbons as high-capacity storage materials for natural gas (methane) and molecularhydrogen in on-board fuel tanks for next-generation clean vehicles. The carbons areproduced in a multi-step process from corncob, have surface areas of up to 3500 m2/g,porosities of up to 0.8, and store, by physisorption, exceptional amounts of methane andhydrogen. Adsorbent-based storage materials are attractive due to their low operatingpressure (relative to compressed gas), reversibility, ease of fueling, and absence ofthermal management issues. Nanopores generate high storage capacities by high surfacearea to volume ratios, and by hosting deep potential wells through overlapping substratepotentials from opposite pore walls, with binding energies approximately twice thebinding energy in wide pores.The best gravimetric and volumetric storage capacities, among the carbonsinvestigated in this thesis, are 1) 250 g CH4/kg carbon and 130 g CH4/liter carbon (199V/V) at 35 bar and 293 K and 2) 80 g H2/kg carbon and 47 g H2/liter carbon at 50 bar and77 K. This is the first time the DOE methane storage target of 180 V/V at 35 bar andambient temperature has been reached and exceeded. The hydrogen values comparefavorably with the 2010 DOE targets for hydrogen, excluding cryogenic components.xvi

Methane and hydrogen storage capacities were measured on custom-built gravimetric andmanometric instruments, and validated on instruments in external laboratories.To shed light on the mechanisms leading to the exceptional storage capacitiesattained, extensive characterization of the surface and pore structure of samples wasperformed.Characterizations include surface areas and pore-size distributions fromnitrogen adsorption at 77 K (subcritical adsorption, 1 bar); pore-size distribution frommethane adsorption at 293 K (supercritical adsorption, 35 bar); spatial organization ofpores from small-angle x-ray scattering, scanning electron microscopy, and transmissionelectron microscopy; and chemical composition from energy-dispersive x-rayspectroscopy. Surface areas and pore volumes in selected intervals of pore size werecorrelated (a) with process parameters for the carbon production, and (b) with methaneand hydrogen storage capacities, to identify carbon production procedures that lead tosuperior storage capacities.For hydrogen, surface areas and pore volumes were used to convert experimentalstorage isotherms into surface coverage.Results were compared with statisticalmechanical models for supercritical adsorption of hydrogen on carbon surfaces and withmolecular dynamics simulations.From fits of the experimental coverage to thetheoretical coverage, key parameters controlling adsorption at the atomic level—such asbinding energies and the nature of adsorption sites—can be deduced. In two case studies,it was found that 40% of surface sites reside in pores of width 0.7 nm and bindingenergy 9 kJ/mol, and 60% of surface sites reside in pores of width 1.0 nm and bindingenergy 5 kJ/mol. The findings, including the prevalence of only two distinct bindingenergies, are in excellent agreement with results from molecular dynamics simulations. Itxvii

was furthermore found that we can experimentally distinguish between the situation inwhich molecules do (mobile adsorption) or do not (localized adsorption) move parallel tothe surface, how such lateral dynamics affect the hydrogen storage capacity, and how thetwo situations are controlled by the vibrational frequencies of adsorbed hydrogenmolecules parallel and perpendicular to the surface: in the two case studies, adsorption ismobile at 293 K, and localized at 77 K. These findings support the hypothesis thathydrogen storage capacities in nanoporous carbons can be optimized by suitableengineering of the nanopore space and demonstrate potential engineering methods.xviii

Chapter 1: Motivation and Research Problem1.1 MotivationIn recent years, the search for inexpensive, abundant energy sources has been aprominent public policy concern. The challenges of energy shortages, greenhouse gasesand air pollution have compelled the public and private sectors to invest in thedevelopment of alternative energy sources. The U.S. Department of Energy (DOE) hasmandated a transition from conventional sources of energy (fossil fuels, uranium) to nonconventional sources (biofuels, non-fossil natural gas, hydrogen) [1]. There is broadconsensus that the use of a combination of these non-conventional fuel sources will berequired to meet energy demand[2].The transportation sector is a major consumer of energy in the U.S., constituting29.1% of energy consumption in 2007, see figure 1.1.1. Therefore, alternative fuelresearch for vehicular use has become a keystone in the future U.S. energy economy.According to the State Alternative Fuels Plan of the California Air Resources Board andCalifornia Energy Commission, adopted Octob

I would like to give special thanks to Dr. Peter Pfeifer, Dr. Galen Suppes, Physics Machine Shop Supervisor Sam Potts, and Dr. Carlos Wexler. Dr. Wexler and Dr. Suppes have been a tremendous help in advancing the research presented through their leadership of the theoretical and chemical engineering aspects of the project, respectively. Sam