MStS t/m/i lr‘l‘ |||||l|ll|liliilfiiiliiliililiilliiiillllllllll 31293 017721170 This is to certify that the dissertation entitled HYDROGEN INHIBITION IN STEAM GASIFICATION OF ANNEALED SARAN CHAR presented by Michael Gerard Lussier Jr. has been accepted towards fillfillment of the requirements for Ph. D. degree in Chemical Engineering Major p gessor Date iO/Zi/IS’ MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY M‘ch'gan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE ' fizlci‘fl‘ 1M Gama-9.14 HYDROGEN INHIBITION IN STEAM GASIFICATION OF ANNEALED SARAN CHAR by Michael Gerard Lussier Jr. A DISSERTATION Submitted to Michigan State University in partial fulfiliment of the requirements fer the degree of DOCTOR OF PHILOSOPHY Department of Chemical Engineering 1998 ABSTRACT HYDROGEN INHIBITION IN STEAM GASIFICATION OF ANNEALED SARAN CHAR by Michaei Gerard Lussier Jr. Anneaied Saran and coai chars were gasified in mixtures of HXNWh/Ar at 1123 K and varying pressures to varying extents of conversion, foiiowed by transient kinetic desorption and TPD to 1773 K. in order to characterize hydrogen adsorbed onto char surfaces during gasification and to identify the mode(s) of hydrogen inhibition at varying extents of char conversion. Adsorbed hydrogen concentration on anneaied Saran char was found to be independent of reactant gas composition and pressure, to increase from an initiai surface concentration of 3x10‘5 to 1.5x10'3 mmoiH2(STP)/nfi over the first 1% conversion. and to increase very graduaiiy after this. Gasification rate deciines significantiy over the initiai 1% carbon conversion and is inhibited mainiy by dissociative hydrogen adsorption over this range. ., F" " r‘ -a n ,- .-‘ ‘, (I ‘ w. IF a 1‘ ' ' i.o-‘ ' C t ‘_ - '0- "..o-' f A I P , ' --' v a -T .u. .l' _ . a g.- .,.,.- . .. g ..o'- . “~-.~. «.._ "r-o. <‘ l . .. v. Q.- Linearized Langmiur-Hinsheiiwood type rate expressions based on the three primary modes of hydrogen inhibition have been deveioped for aii gasification data above 1% char conversion. The expression which indicates reverse oxygen exchange or “associative" hydrogen adsorption fits the data weii. whiie the expression for dissociative hydrogen adsorption does not. Caicuiation of the equiiibrium constant for oxygen exchange (kl/kd=0.029) indicates a 10w fractionai coverage of adsorbed oxygen compiexes (C(O)). whiie the equiiibrium constant for “associative” hydrogen adsorption (k3/ke=425 MPa4) stipuiates a high fractionai coverage of “associativeiy” adsorbed hydrogen. Because no “associativeiy” bound hydrogen was detected and because 10w concentrations of surface oxides were found during gasification, it is conciuded that reverse oxygen exchange is the primary mode of hydrogen inhibition past 1% char conversion for Saran char. Active site propagation aiong graphitic zig-zag edges is proposed as the main source of surface carbon consumption for steady-state char gasification in steam. ...¢, - I dedicate my PhD. Dissertation to my parents Michaei and Roseann Lussier and my sister Amy Lussier, and to Jennifer Durst. iv ACKNOWLEDGEMENTS I wouid iike to thank my graduate advisor and project principai investigator, Dr. Dennis J. Miiier, for his patience and high degree of invoivement in my Doctorate research and dissertation work. Dr. Lawrence T. Drzai, Dr. Martin C. Hawiey, Dr. Thomas J. Pinnavaia, and Dr. Robert E. Buxbaum must aiso be thanked for accepting the task of being my committee members. I give thanks to Zhigang Zhang for aii of his effort on the char gasification project, inciuding work on TPD, nitrogen adsorption, unanneaied char gasification, and data deconvoiution programs. Thanks must aiso go to Radu Craciun for his heip with using the X-ray diffractometer. Juiie Caywood, Faith Peterson, and Candice McMaster must aii be thanked for their exceptionai secretariai support. as weii as Rob Seiden for fabricating severai intricate components of the experimentai apparatus used in this investigation. Finaiiy, I wouid iike to thank the United States Department of Energy for their financiai support through grant #DE-F622-93PC93213. ‘F i, .‘- 'A ‘is ‘9 r .5 I ’A 1‘ . ‘v ‘9 1' .~ .I f Q ~ .A / e o 'i o“ a g h. '0- .-, I . A‘ ‘ '5 x TABLE OF CONTENTS LIST OF TABLES .......................... xiii LIST OF FIGURES .......................... xiv LIST OF NOMENCLATURE ........................ xxi Chapter 1. Introduction ....................... 1 1.1. Background .......................... 1 1.2. Literature Review ....................... 3 1.2.1. Hydrogen Chemisorption on Carbon ............. 3 1.2.2. Hydrogasification of Chars ................ 3 1.2.2.1. Genera] Reaction Phenomena .............. 3 1.2.2.2. RoIe of Oxygen .................... 5 1.2.2.3. StructuraT Effects .................. 5 1.2.3. Steam Gasification of Chars ................ 7 1.2.3.1. Generai Reaction Phenomena .............. 7 1.2.3.2. Roie of Oxygen .................... 9 1.2.3.3. Hydrogen Inhibition ................. 12 1.2.4. Isotopic Studies .................... 13 1.3. Previous Research in Our Laboratory ............. 14 1.4. Research Objectives ..................... 19 vi 1.4.1. Mechanism Identification ................ 20 1.4.2. Isotope Effects ..................... 23 Chapter 2. Experimental ...................... 26 2.1. Starting Materials ..................... 27 2.1.1. Chars .......................... 27 2.1.2. Reactant Gases ..................... 28 2.2. Experimental Apparatus ................... 29 2.2.1. High Pressure Reactor .................. 29 2.2.1.1. Internal Microreactor ................ 29 2.2.1.2. Flange Modifications . . . . . . . . . . . . . . . . 31 2.2.2. Flow Control and Mixing System ............. 34 2.2.2.1. Gas Blending/Rapid Switching System ......... 35 2.2.2.2. Low Dead Volume Steam Trap ............. 36 2.2.3. High Temperature Reactor ................ 38 2.2.4 Mass Spectrometer .................... 38 2.3. Experimental Techniques ................... 40 2.3.1. Gasification ...................... 44 2.3.1.1. Gasification Conditions ............... 44 2.3.1.2. Gasification Procedure ............... 45 2.3.2. Temperature Programmed Desorption ............ 47 2.3.2.1. TPD Conditions ................... 47 vii no- P ./.u ' A v 1 A4. I.- o/ih fi/k a /.\ A ./\ 2.3.2.2. TPD Procedure .................... 48 2.3.3. Gas Detection and Calculation of Effluent Rates ..... 49 2.3.3.1. Mass Spectrometer Calibration ............ 50 2.3.3.1.1. Variation of Pressure at Capillary Inlet . . . . 51 2.3.3 1.2. Linearity of Response .............. 56 2.3 3.1.3. Response as a Function of Time ......... 56 2 3.3.1.4. Response as a Function of Carrier Gas Composition 59 2 3.3.1.5. Fragmentation Investigation ........... 62 2.3.3.1 6. COMMHM Response Ratios ............. 62 2.3.3.2. Mass Spectrometer Data Deconvolution ........ 64 2.3.4. Char Characterization .................. 66 2.3.4.1. Mercury Intrusion .................. 67 2.3.4.2. Nitrogen Adsorption ................. 67 2.3.4.3. X-ray Diffraction .................. 68 Chapter 3. Gasification Experiments ................ 69 3.1. Determination of Appropriate Gasification Temperature . . . . 70 3.2. Char Gasification in Steam ................. 72 3.2.1. Evolution of Char Surface Area ............. 72 3.2.2. Rate Dependence on Pressure, Composition, and Conversion 74 3.2.2.1. Annealed Char Steam Gasification .......... 76 3.2.2 1.1. Annealed Saran Char Steam Gasification ..... 76 viii 3.2.2.1.2. Annealed Coal Char Gasification ......... 101 3.2.2.2. Unannealed Char Steam Gasification ......... 105 3.2.3. Isotopic Studies .................... 108 3.3. Adsorbed Hydrogen Concentration ............... 108 3.3.1. Transient Hydrogen Desorption .............. 108 3.3.2. Adsorbed Hydrogen: Unit Weight Basis .......... 110 3.3.3. Adsorbed Hydrogen: Unit Area Basis ........... 118 Chapter 4. Model Fitting — Linear Regression Analysis ....... 120 4.1. Char Gasification in Steam ................. 122 4.1.1. Reverse Oxygen Exchange ................. 123 4.1.1.1. Reverse Oxygen Exchange ............... 123 4.1.1.2. Rapid Equilibrium Reverse Oxygen Exchange ...... 127 4.1.2. “Associative” Hydrogen Adsorption ............ 128 4.1.2.1. “Associative" Hydrogen Adsorption Only ....... 131 4.1.2.2. “Associative” Hydrogen Adsorption and Reverse Oxygen Exchange ..................... 132 4.1.3. Dissociative Hydrogen Adsorption ............ 138 4.1.3.1. Dissociative Hydrogen Adsorption Only ....... 138 4.1.3.2. Dissociative Hydrogen Adsorption and Reverse Oxygen Exchange .................... 142 4.2. Comparison of Hydrogen Inhibition Models ......... 146 4.2.1. Linear Regression Parameters ............. 146 ix - I - o i c o pro-I o ‘ 0-0.- . . u . . - . I v - - n I ’ . .. . ‘ U . Q . - s ‘ Q . ' ‘ ’ v- u - - n n ' ‘ . I . U.~ _ n c , . . . u I v ‘- . Q n H O A o l H b - . s l o C P. u l‘. U. -' 4.2.2. Calculated Rate Constants ............... 149 4.2.3. Theoretical Rate Curves ................ 153 4.3. Methane Formation Only .................. 160 (impter 5. Char Properties During Gasification ........... 162 5.1. Hydrogen Adsorption ..................... 162 5.1.1. Initial Rapid Adsorption ................ 163 5.1.2. Gradual Adsorption over Time .............. 165 5.2. Char Structure ....................... 165 5.2.1. Total Surface Area and Pore Structure .......... 166 5.2.2. Domain Sizes via H/C Atom Ratio ............. 167 5.2.3. X-ray Analysis ..................... 168 5.2.4. Char Morphology ..................... 175 5.2.5. Char Active Sites .................... 177 5.2.5.1. Etch Pit Analysis .................. 178 5.2.5.2. Active Site Propagation ............... 181 5.2.5.3. Active Site Behavior with Conversion ........ 184 5.2.5.4. Comparison to the “Universal" Mechanism ....... 188 5.3. Rate Enhancement ...................... 189 5.3.1. Partial Combustion ................... 190 5.3.2. Catalysis ........................ 191 (impter 6. Conclusions/Recommendations ............... 193 6.1. Char Gasification Mechanism Identification ......... 193 6.2. Char Structure During Gasification ............. 195 6.3. Recommendations ....................... 198 Appendices ............................. 200 A-I. Linearized Rate Expression for Dissociative Hydrogen Adsorption Including Explicit Adsorbed Hydrogen Term . . . 202 A-2. Linearized Rate Expression for Reverse Oxygen Exchange and Dissociative Hydrogen Adsorption Including Explicit Adsorbed Hydrogen Term .................. 203 8. Mass Spectrometer Controller Settings ............. 204 C. Mass Spectrometer Data Deconvolution Programs ......... 205 D. Annealed Saran Char Generalized Modulus Calculation ...... 213 E-l. Linearized Rate Expression Derivation for Reverse Oxygen Exchange ......................... 217 E-2. Linearized Rate Expression Derivation for Rapid Equilibrium Reverse Oxygen Exchange .................. 218 E-3. Linearized Rate Expression Derivation for “Associative” Hydrogen Adsorption .................... 219 E~4. Linearized Rate Expression Derivation for “Associative" Hydrogen Adsorption and Reverse Oxygen Exchange ...... 221 E-5.Linearized Rate Expression Derivation for Dissociative Hydrogen Adsorption .................... 223 E-6.Linearized Rate Expression Derivation for Dissociative Hydrogen Adsorption and Reverse Oxygen Exchange ...... 225 E-7.Linearized Rate Expression Derivation for Methane Formation Only - “Associative" Hydrogen Adsorption ......... 227 xi E—8. Linearized Rate Expression Derivation for Methane Formation Only - Dissociative Hydrogen Adsorption .......... 228 References ............................. 229 xii LIST OF TABLES Table 1 - Char Ultimate Analysis .................. 28 Table 2 - Number of Parameters in Various Kinetic Models of FbO/Hz Gasification of Annealed Saran Char .......... 146 Table 3 - Degrees of Freedom for Linear Regression of Various Kinetic Models of Hxhih Gasification of Annealed Saran Char as a Function of Conversion ............ 147 Table 4 - Regression Results for Linearized Rate Expressions . . . . 150 Table 5 - Comparison of Rate Constant Groups ............ 152 xiii n p I u a . P119 a Q F./s PilV li.‘ I n u. 4 ff.. 5 . -.J .41 o c . a c a . o a . a . ... .—. ;—» an. .—. .~. .~- .- ._. .n. .5. .P- a a r r r r r r .v r u. r 4 .- 4 o a A a J J a a t-. .c. ~ .- . .q .Iu «u. .0. RI. ~ .- flu. .. ... o.» c.- .6 .o v.. n. e.» Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 LIST OF FIGURES Various Edge Configurations on the Graphite Basal Plane . . 6 Edge Carbon Atoms Configurations of the Graphite Basal Plane ......................... 10 Universal Gasification Mechanism Proposed by Chen, Yang. Kapteijn, and Moulijn [56] .............. 11 Hydrogasification Rate Curves for Saran Char ...... 15 Hydrogasification Rate Curves for Coal Char ....... 16 Hydrogasification Rate Curves for Chars Based on Total Surface Area ..................... 17 Overall Experimental Apparatus ............. 30 High Pressure Vessel Internal Microreactor ....... 32 High Pressure Vessel Flange Modifications . . . j . . . . 33 Low Dead Volume Steam Trap ............... 37 High Temperature Reactor ................ 39 Experimental Technique for Gasification and Subsequent Temperature Programmed Desorption of Chars ...... 41 High Pressure System Transient Response ........ 43 Vacuum Chamber Pressure vs. Capillary Inlet Pressure with Argon Purge Gas ................. 52 Background Partial Pressure of Various Masses vs. Vacuum Chamber Pressure (Argon) ............... 54 xiv Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 Figure 22 Figure 23 Figure 24 Figure 25 Figure 26 Figure 27 Figure 28 Background Partial Pressure of Various Masses vs. Vacuum Chamber Pressure (Ar/Kr) ............... 55 Linearity of Mass Spectrometer Response with Species Concentration .................... 57 Change in Mass Spectrometer Peak Heights over Five Hour Time Interval .................... 58 Partial Pressures of Carrier Gases in Vacuum Chamber vs. Carrier Gas Composition ............... 60 Partial Pressures of Key Components in Vacuum Chamber vs. Carrier Gas Composition ............... 61 Partial Pressure of COziand CO Fragments vs. Carrier Gas Composition and Inlet Percent CO2 .......... 63 Arrhenius Plot of Annealed Saran Char Steam Gasification at 1.0 MPa ...................... 71 Total Surface Area of Annealed Chars by Nitrogen Adsorption after Gasification in Steam at 1.0 MPa . . 73 Equilibrium Composition of the Shift Reaction and Reactor Effluent Composition vs. Temperature ..... 75 Equilibrium Composition of the Methanation Reaction and Reactor Effluent Composition vs. Temperature ..... 77 C0 + C02 Evolution Rate from Several Runs of Annealed Saran Char Steam Gasification at 1.0 MPa in 40/0/60 H20/H2/Ar ...................... 78 C0 + C02 Evolution Rate from Annealed Saran Char Steam Gasification at 1.0 MPa in 40/0/60 HAMNb/Ar: Several Runs at 2% Carbon Conversion ............. 80 CO +-CIk Evolution Rate from Annealed Saran Char Steam Gasification at 3.1 MPa ............... 81 XV a,- . - a 04- ' ,. .r' 4 .- ., - ~- .. . .p- . -p . ,. . . .— ' Fa . . «0 . g. ,. I O U. ‘ .~ .- ‘- -» ,-- r- -. - or 'J I ‘ U. «Qr‘ A- .- I‘ ‘ i ‘ -: . " r. *- h '4 ‘. . '- . .‘fi r. t'.. n‘ — '4 ' Q 'U . W I. I' '1 ' . ~ Figure 29 Figure 30 Figure 31 Figure 32 Figure 33 Figure 34 Figure 35 Figure 36 Figure 37 Figure 38 Figure 39 Figure 40 CH4 Evolution Rate from Annealed Saran Char Steam Gasification at 3.1 MPa ............... 82 CO +-CXk Evolution Rate from Annealed Saran Char Steam Gasification at 1.0 MPa ............... 83 (it Evolution Rate from Annealed Saran Char Steam Gasification at 1.0 MPa ............... 84 CO +-CI» Evolution Rate from Annealed Saran Char Steam Gasification at 0.3 MPa ............... 85 ()h Evolution Rate from Annealed Saran Char Steam Gasification at 0.3 MPa ............... 86 CO +-(Ih Evolution Rate from Annealed Saran Char Steam Gasification ..................... 88 CH4 Evolution Rate from Annealed Saran Char Steam Gasification ..................... 89 CO +-(Ih Evolution Rate from Annealed Saran Char Steam Gasification at 1.0 MPa with Cycled Reactant Gas Composition ..................... 91 Quadratic Polynomial Least Squares Fit to Annealed Saran Char Total Surface Area Following Gasification in Steam ........................ 93 Quadratic Polynomial Least Squares Fit to Annealed Coal Char Total Surface Area Following Gasification in Steam ........................ 94 CO +-(Xk Evolution Rate from Annealed Saran Char Steam Gasification at 1.0 MPa; Rate Per Unit Weight and Unit Surface Area Basis .................. 95 (1h Evolution Rate from Annealed Saran Char Steam Gasification at 1.0 MPa: Rate Per Unit Weight and Unit Surface Area Basis .................. 96 xvi ‘ Figure 41 Figure 42 Figure 43 Figure 44 Figure 45 Figure 46 Figure 47 Figure 48 Figure 49 Figure 50 Figure 51 CO +-(Ib Evolution Rate from Annealed Saran Char Steam Gasification to Low Conversion: Rate Per Unit Weight and Unit Surface Area Basis ............. 97 CH4 Evolution Rate from Annealed Saran Char Steam Gasification to Low Conversion: Rate Per Unit Weight and Unit Surface Area Basis ............. 98 Evolution Rate from Annealed Saran Char Steam Gasification at 1.0 MPa; Rate Per Unit Weight and Unit Surface Area Basis .................. 100 Product Gas Evolution Rate from Annealed Coal Char Steam Gasification at 1.0 MPa ............... 102 CO +-CIb Evolution Rate from Annealed Coal Char Steam Gasification at 1.0 MPa; Unit Weight and Unit Surface Area Basis ...................... 103 CH4 Evolution Rate from Annealed Coal Char Steam Gasification at 1.0 MPa: Unit Weight and Unit Surface Area Basis ...................... 104 Product Gas Evolution Rate from Unannealed Saran Char Steam Gasification at 1000 K and 3.1 MPa ....... 106 Product Gas Evolution Rate from Unannealed Coal Char Steam Gasification at 1000 K and 3.1 MPa ....... 107 C0 + C02 Evolution Rate from Annealed Saran Char H20 or [k0 Gasification at 3.1 MPa ............. 109 Step Change in Feed Gas Composition During Annealed Saran Char DzO Gasification ............. 111 Temperature Programmed Desorption Profile of Hydrogen from Annealed Saran Char Following Gasification in Steam at 1123 K ................... 112 xvii OF U Ihw ( Hydrogen TPD Profile of Annealed Saran Char Following Hydrogasification at 3.1 MPa and Cooling in Argon or Hydrogen ....................... 114 Figure 52 Adsorbed Hydrogen Concentration from Annealed Chars Following Gasification in Steam at 1123 K ...... 115 Figure 53 Product Gas Evolution Rate and Adsorbed Hydrogen Concentration from Annealed Saran Char Steam Gasification at Low Conversion ............ 117 Figure 54 Adsorbed Hydrogen Concentration on Surface Area Basis from Annealed Chars Following Steam Gasification at 1123 K ........................ 119 Figure 55 Correlation Coefficient for Linear Regression of Annealed Saran Char Steam Gasification - Reverse Oxygen Exchange (ROE) ................ 125 Figure 56 F Statistic for Linear Regression of Annealed Saran Char Steam Gasification - Reverse Oxygen Exchange (ROE) ........................ 126 Figure 57 Correlation Coefficient for Linear Regression of Annealed Saran Char Steam Gasification - ROE. and Rapid Reverse Oxygen Exch. (R-ROE) .......... 129 Figure 58 F Statistic for Linear Regression of Annealed Saran Char Steam Gasification - ROE. and Rapid Reverse Oxygen Exch. (R-ROE) ................. 130 Figure 59 Figure 60 Correlation Coefficient for Linear Regression of Annealed Saran Char Steam Gasification - “Associative” Hydrogen Adsorption (AHA) .............. 133 Figure 61 F Statistic for Linear Regression of Annealed Saran Char Steam Gasification - "Associative” Hydrogen Adsorption (AHA) ................... 134 xviii .r‘ r . .. up ' :0 .o- P‘ ‘1 ~ g. 'l i,” ' -. .4 - v ,- ~,. f“ ‘t - I- ., I‘ve - Pr " v I. .‘ D§ ,. . r C) v 1 ‘A II ' n '4 9‘- . ' v.1 w.” f- .‘— Va . s v. d . .‘s '. ~- - . '4 fi ‘- . .‘s r, .. - ‘a . \ § . (Ir. ‘ ' fi " / . s ‘L . 's f:- ‘A 1".- ‘ . ‘ K. Figtire 62 Figure 63 Figure 64 Figure 65 Figure 66 Figure 67 Figure 68 Figure 69 Figure 70 Figure 71 Figure 72 Figure 73 - Correlation Coefficient for Linear Regression of Annealed Saran Char Steam Gasification - ROE. AHA, and ROE + AHA ...................... 136 — F Statistic for Linear Regression of Annealed Saran Char Steam Gasification - ROE. AHA, and ROE + AHA . . 137 - Correlation Coefficient for Linear Regression of Annealed Saran Char Steam Gasification - Dissociative Hydrogen Ads. (DHA). and ROE ............. 140 - F Statistic for Linear Regression of Annealed Saran Char Steam Gasification - Dissociative Hydrogen Ads. (DHA). and ROE .................... 141 - Correlation Coefficient for Linear Regression of Annealed Saran Char Steam Gasification - ROE. AHA, and ROE + DHA ...................... 144 - F Statistic for Linear Regression of Annealed Saran Char Steam Gasification - ROE, AHA. and ROE + DHA . . 145 - CO +-(Ik Evolution Rate from Annealed Saran Char Steam Gasification at 3.1 MPa and “n=1" Model ....... 154 - CO +-(Ik Evolution Rate from Annealed Saran Char Steam Gasification at 1.0 MPa and “n=1” Model ....... 155 — C0 +-CIk Evolution Rate from Annealed Saran Char Steam Gasification at 0.3 MPa and “n=1” Model ....... 156 - CO +-CIk Evolution Rate from Annealed Saran Char Steam Gasification at 3.1 MPa and “n=0.5" Model ...... 157 - C0 +-CIk Evolution Rate from Annealed Saran Char Steam Gasification at 1.0 MPa and “n=0.5" Model ...... 158 - CO +-CIh Evolution Rate from Annealed Saran Char Steam Gasification at 0.3 MPa and "n=0.5" Model ...... 159 xix Figure 74 Figure 75 Figure 76 Figure 77 Figure 78 Figure 79 Figure 80 Figure 81 Figure 82 Figure 83 Figure 84 Dependence of Methane Formation Rate on Hydrogen Partial Pressure in Annealed Saran Char Steam Gasification ..................... 161 X-ray Spectra of 360 mesh Ultra “F” Graphite ...... 170 X-ray Spectra of 325 mesh Alpha Graphite ........ 171 X—ray Spectra of —10+60 mesh Alpha Graphite ...... 172 X-ray Spectra of -60+100 mesh Annealed Saran Char . 173 Magnified View of X-ray Spectra of —60+100 mesh Annealed Saran Char ...................... 174 ()h Evolution Rate from Saran Char Hydrogasification. Temperature at or Normalized to 1123 K via E1=75 Kcal/mol Round and Hexagonal Etch Pits of the Graphite Basal Plane ...... Active Site Propagation along Graphite Zig-Zag Edge Active Site Propagation along Graphite Armchair Edge . . Round Etch Pit and Crystallite Converting to Hexagonal Features with Zig-Zag Edges XX Q. h. ~\ . .IJ. . . o w . . . . . l . . 2 ~ . . \ t i . s . . ~ I a . . _ v n p n w I ii Ii. . I It '1- r ‘ i Ink [“4 J D I n. 6 .nl . 1 . u a . "I a .1- - ~ . vii » CF CT C(H) C(H) C(H)2 C(H)i C(O) d E? Ew FSTAT h k k4 k1 k2 K3 k3 k3 K4 k4 k4 h Km RH n PM H mm mm Qdel ec thrans QHrot 0%6 Qhel ec LIST OF NOMENCLATURE “Free" Active Site Surface Concentration Total Active Site Surface Concentration Hydrogen Surface Concentration Dissociatively Bound Hydrogen Concentration “Associatively” Bound Hydrogen Concentration Initial Hydrogen Surface Concentration Oxygen Surface Concentration Interlayer Spacing Deuterium Zero-Point Energy Hydrogen Zero-Point Energy F-Statistic Planck's Constant Shape Factor Reverse Oxygen Exchange Forward Rate Constant Oxygen Exchange Forward Rate Constant Gasification Rate Constant "Associative” Hydrogen Adsorption Equilibrium Constant “Associative" Hydrogen Adsorption Forward Rate Constant “Associative” Hydrogen Adsorption Reverse Rate Constant Dissociative Hydrogen Adsorption Equilibrium Constant Dissociative Hydrogen Adsorption Forward Rate Constant Dissociative Hydrogen Adsorption Reverse Rate Constant Boltzman's Constant Overall Steam Gasification Rate Constant Shift Reaction Equilibrium Constant Positive Integer Hydrogen Partial Pressure Water Partial Pressure Deuterium Rotational Partition Function Deuterium Vibrational Partition Function Deuterium Electric Partition Function Deuterium Translational Partition Function Hydrogen Rotational Partition Function Hydrogen Vibrational Partition Function Hydrogen Electric Partition Function xxi (Ykrans = Hydrogen Translational Partition Function R = Gas Constant r2 = Linear Regression Coefficient of Correlation rcn = CH4 Evolution Rate rco = C0 Evolution Rate rco.1 = Initial CO Evolution Rate rco“ = C0 Evolution Rate in Hydrogen Containing Species rcoD = C0 Evolution Rate in Deuterium Containing Species 'T = Temperature t = Ordered Domain Size TSA = Total Surface Area x = Percent Carbon Conversion B = Peak Full Width at Half Max A. = X-ray Wavelength L1 = Reduced Mass v = Frequency 9 = X-ray Scatter Angle 93 = X-ray Peak Width at Half Max xxii ‘1! Chapter 1 INTRODUCTION 1.1. Background Increasing demand for pipeline and transportation fuels in the United States, coupled with a dwindling world supply of these fuels. indicates that gasification of coal may be the most viable solution. Future demands on current natural gas supplies will deplete them far more rapidly than current coal supplies. Natural gas is superior to coal as an energy source because it is a pipeline and tranSportation fuel, and burns much cleaner than coal. Benefits of coal gasification .0‘ ‘- "J _-— .'- . . ‘O. ‘ .r" T'7 I. . a. ...--r a .-..o« «L .1 I n 0., y'- .. P.. u .v . v ....r.. . — 4- .. . . a - 4. ., _ e'_ 7‘ .- n. ‘ q ' J ~. p " O-A‘ 1... -' Vi ‘V‘. ___r~~ -. '0 L .‘ P'a- r .' ‘— u v a ' - .l - ‘J v ‘. ‘. _I“. 2 in steam include the facts that it is simple to carry out, water is cheap and plentiful, and the effluent gases can be converted into many other products traditionally derived from petroleum. Gasification of coal is not currently used on a wide scale because extreme conditions are needed to achieve reaction rates that are reasonably fast. Hydrogen is known to strongly inhibit steam gasification rate. The process by which hydrogen inhibits gasification is not yet well characterized, but if it were there would be great potential for minimizing this phenomenon. The main objective of this investigation is to characterize the concentration and stability of hydrogen on coal char and Saran char during steam gasification and relate this to hydrogen inhibition. Gasification rate under varying conditions will be incorporated into linearized rate expressions that are based on the possible mode(s) of hydrogen inhibition. Regression parameters and rate constants calculated for the different rate expressions will be compared to determine which mode(s) of hydrogen inhibition are correct. A ~a- v - ‘ 7‘. ‘7 . .~' 4 u. H.- . .__ \c o ‘h. o n v a; I'l‘b p e ’- . I *. '1 ‘- '5‘. _ n v v ’u 3 1.2. Literature Review 1.2.1. Hydrogen Chemisorption on Carbon Hydrogen that has chemisorbed onto a carbon surface is very stable. and is generally accepted as dissociative in nature. Prolonged outgassing at 1300 K will not remove all dissociatively adsorbed hydrogen [1-8]; temperatures approaching 1800 K are required [1]. Dissociatively adsorbed hydrogen can saturate a graphite surface at 1373 K and 3 millitorr hydrogen [5], and has an equilibrium constant of 253 atm‘“2 at 973 K [6]. This constant, however, can be strongly affected by impurities in the carbon which may act as hydrogen dissociation sites. Dissociatively bound hydrogen (C(H)) will form a peak starting at about 1200 K during temperature programmed desorption (TPD). while another peak can be observed at about 900-1100 K due to associatively bound hydrogen (C(H)2) following exposure of carbon samples to gases containing hydrogen. 1.2.2. Hydrogasification of Chars 1.2 2.1. General Reaction Phenomena Understanding methane formation by direct attack of carbon by molecular hydrogen is important because it occurs during steam . ..Q.‘ -4 _ . '& .... a~“. ., _‘ . -.- , - . , ~r f“ . .,a.-»- - ,.. - o- a .. .- v. , u4ro~np. . .. . y.... ’ r . .._ I C -- 'Q. U. c -.' '1 ‘V F‘ A» u i ‘ 3" . ~ 5» 4...‘ ‘1"\ ~ ‘ a ‘ 7 -.~ .r .‘- .. .. ‘4 . a - ‘- .- ‘ _h 5“ 9“. ‘ .. a. b .|a VA- . 5‘ H l .I ‘§ \. - -L . ,- o r. . \ ‘v . n - 1‘. C, . k. 0‘- 4 gasification of chars when hydrogen partial pressures are high [9]. Hydrogasification is also the first major step in the HYDROCARB process. which results in very pure fuel grade carbon black [10.11]. Kinetic studies by Blackwood et al. [12-16] have shown that hydrogasification is first order in hydrogen partial pressure. and is not a strong function of char type [12.17.18]. Several researchers performing mechanistic studies have suggested successive dissociative hydrogen Chemisorption onto adjacent carbon atoms [19.20], with the cleavage of the bond between adjacent carbons being the rate limiting step [21]. Several others have suggested associative hydrogen Chemisorption of two hydrogen molecules onto the same carbon atom [8.22.23]. with the cleavage of carbon-carbon bonds also being rate limiting [22]. The rate of this reaction decreases rapidly with conversion when char samples are uncatalyzed [13,15,23-29]. Several two-stage reactions have been proposed to explain this phenomenon. most of which include initial rapid methanation of highly reactive surface carbon followed by slow reaction of the highly aromatic char base structure. Blackwood et al. [17] identify the reactive carbon, or "secondary" carbon, as amorphous and already partially enriched with hydrogen, and/or adjacent to oxygen functional groups [30]. Heating rate. final temperature, and o .-‘ I r u f ' 7 2. o O 0U ' o - f ‘- . JV... . --, . a . ._ _ F'fA‘-.‘ .. \’ v-“ I - .(Ar- ‘ ._ ‘ i h c "~..'v -‘ ' . d .. -‘ ”.1 '. f s '\‘ 5 purge gas composition during char preparation, as well as storage methods. largely determine the nature of these reactive surface carbons. 1.2.2.2. Role of Oxygen It is generally believed that the major source of active sites in all uncatalyzed char gasification reactions comes from the desorption of oxygen functional groups from the char surface. These groups desorb in the form of carbon monoxide and carbon dioxide when samples are heated to reaction temperatures [15.31-38], and in the form of water during hydrogasification [15]. Hydrogasification rate has been shown to be a strong function of the oxygen content of various chars [12.15.39]. and initial rate a strong function of oxygen surface concentration [36,40]. Hydrogasification rate can be increased by an order of magnitude by addition of 0.1% oxygen to the reactant gas [41]. 1.2.2.3. Structural Effects Initial gasification of highly reactive carbon is not the only reason why uncatalyzed rate decreases so dramatically with conversion. Figure 1 shows the various configurations of carbon atoms on the graphite basal plane. A strong preference for reaction of hydrogen with edge carbon atoms, as opposed to basal plane carbon atoms. has been shown [42]. Further probing into the reactivity of edge carbon atoms by Edge Carbon [___ N Zagll Atom Configuration "Armchair" Edge Carbon Atom Configuration Carbonyl ne ' o C "Zig Semi - Quino Figure 1'. Various carbon atom configurations and oxygen functional groups of the graphite basal plane. .." a .. ~ -. y . (,' r4...' up -., ~ ‘ — 1‘ ~ . v ‘ .“A-’. .. .¢_ - ’ u -. ._.' ‘ .- " .o "H.-- "o- 7 etch pit analysis of graphite by Yang et al. [6.21.43.44] shows that hydrogen binds more strongly to the "zig-zag" edge carbon atoms and removes the "armchair" edge carbon atoms more easily in hydrogasification and steam gasification, while oxygen and carbon dioxide show no edge preference in gasification [6 43]. Therefore, as hydrogasification or gasification in steam proceeds, the more highly reactive armchair edge carbon atoms are consumed. leaving the more stable zig-zag edge carbon atoms. 1.2.3. Steam Gasification of Chars 1.2.3.1. General Reaction Phenomena Steam gasification of chars consists of the reaction of steam with carbon to form a mixture of carbon monoxide and hydrogen, or "synthesis gas". which can then be converted into a spectrum of products [45]. Overall Reaction: H20+CF —59—R—+H2 +CO (1) Huttinger et al. [8,46,47] state that the uninhibited forward reaction can be broken down into the following two major steps. Oxygen Exchange: H20 + C; —k;—> H2 + 0(0) (2) Gasi fi cation: mop—“14004.0; (3) Crindicates a free carbon site. which is a surface carbon atom that is not saturated with chemical bonds. Carbon dioxide and methane are also r.‘-~ rnir 8 formed to a much lesser extent during steam gasification. Carbon dioxide is formed in the gas phase by the shift reaction. not at the carbon surface [8]. Shift Reaction: H20+CO<—K—§“—>H2 +C02 (4) Under normal conditions methane is formed at the carbon surface; it is neither homogeneously formed nor decomposed [8]. There are three major mechanisms by which steam gasification is inhibited [8.46.47]. Reverse Oxygen Exchange: H2+C(O)—kL>H20+CF (5) “Associative" Hydrogen Ads.: H2+CF<—&—>C(H)2 (6) Dissociative Hydrogen Ads.: %H2+CF<—K‘1—>C(H) (7) Selection of any one of these inhibition reactions when developing a rate expression gives an equation that has been supported by several researchers [8.46] k1CTPw = (8) 1+ (k1 Ik2)P\~ + r(k)l=i:'2 Basic Rate Expression: rho Dissociative hydrogen adsorption gives a value of 0.5 for n in Equation 8, which has been found to be the case for low temperatures, low hydrogen pressures and subatmospheric steam pressures [6,43,48,49]. Reverse oxygen exchange and “associative” hydrogen adsorption both give “ch - r .4479« . '4. v a! - 19¢ - '1 no. .'r‘f '0. v n, ‘ a 'v- 9 values of 1 for n in the basic rate expression, which was reported in early studies [50-54]. 1.2.3.2. Role of Oxygen A mechanism that has been cited as universal to all carbon gasification has been recently proposed by Chen, Yang. Kapteijn, and Moulijn [55,56]. At least two different types of oxygen surface complexes were identified by Kapteijn and Moulijn [56] by studying the exponential decay of CO curves following transient step changes in feed gas. The first type of complexes are semi-quinone and carbonyl, which are fairly stable at reaction temperature and can be seen in Figure 1. Molecular orbital calculations by Chen and Yang [55] have shown that the lowest energy conformation for the other type of complex is an off-plane oxygen atom bound to a carbon atom which is adjacent to the semi-quinone or carbonyl complex. The off-plane oxygen atom is bonded to a carbon atom that is in the “caved-in" or "sheltered" position on the zig-zag edge of a graphitic basal plane (see Figure 2). and lowers the bond energy of the adjacent "exposed" carbon atom by about 30%. This mechanism, shown in Figure 3. is said by both sets of workers to be universal to the reaction of char to all oxygen containing reactant gases, however neither group studied the effect of hydrogen on this off- plane oxygen atom. a... NWEIWEUINWV m... SHELTERED CARBON§}—‘ [EXPOSED CARBONS] ZlG-ZAG EDGE 4 ARMCHAIR EDGE GRAPHITIC BASAL PLANE CARBONS vvvvvivvv ll 1 ZIG—ZAG EDGE EHELTERED CARBONQ EXPOSED CARBONS AR MCHAIR EDGE Figure 2: Edge carbon atom configurations of the graphite basal plane. OF 4” MC 00 11 1) Oxygen adsorption to form semi-quinone groups on exposed edge carbons: 2) Off-plane oxygen adsorption onto sheltered edge carbons: 3) Desorption of semi-quinone surface groups to form carbonyl surface groups: 4) Adsorption of off-plane oxygen onto carbons that anchor carbonyl surface groups: 5) Desorption of carbonyl groups to leave semi-quinone surface groups: 0 O O O O O O 0 00000000 0 0 O 0 0 0 0 II II II II II II II C C C C C C C =0 Figure 3: Universal gasification mechanism proposed by Chen, Yang, Kapteijn, and Moulijn [56]. ..- , .- o o,.. a 4 . . I. \ -‘~ Iv‘ v 12 1.2.3.3. Hydrogen Inhibition The presence of hydrogen greatly reduces the rate of gasification in steam as well as hydrogasification rate [50—54]. Steam gasification of chars rate drops by an order of magnitude with the addition of only 1 ppm hydrogen [49]. Gasification with carbon dioxide is also inhibited to this degree by hydrogen at low pressures [5 6 43]. During char gasification in steam, carbon dioxide and methane formation rates are decreased as well as carbon monoxide and hydrogen formation rates due to hydrogen inhibition [8]. “Associative" hydrogen adsorption has been found by Hermann and Huttinger [45,47] to contribute to inhibition in steam gasification of chars at higher pressures. Their TPD studies show a hydrogen desorption peak at 900-1100 K following gasification, indicative of C(H)2 surface groups. Much larger peaks were found above 1273 K, proving that dissociative hydrogen inhibition still dominates. These investigators. and others [24.27.57]. have reported reaction rates approaching zero at carbon conversions as low as 40% in steam/hydrogen mixtures. Gasification rate has also been shown by Huttinger and Merdes [8] to be greatly reduced after exposure of carbon to hydrogen in sequential steam/hydrogen/steam reactions. The rate is reported not to return to 'elfl'f" "r. '0‘. 13 its previous value after the second gasification in steam is initiated. suggesting irreversible blockage of active sites on char surfaces. 1.2 4. Isotopic Studies Relatively few gasification studies have been performed using isotopes. even though isotope effects in chemical reactions have been discussed in several texts [58 59]. Gasification rates of graphite in ibO are reported to be twice as high as in DA). Yates and McKee concluded that breakage of the HO-H bond is involved in the rate limiting transition state complex [60]. while Mims and Pabst concluded that the difference in rate is due to a shift in the oxygen exchange equilibrium constant [61]. Very small isotopic effects were found in the H2anuiIk gasification of graphite at 1473 K and 20 torr hydrogen [62]. H/D exchange has been shown to take place readily over carbon at 673 K [63]. Transient kinetic methods. isotopic studies using 13C and 18O, and TPD have been successfully combined by Kapteijn et al. [56,64] to clarify mechanisms in the CIb gasification of carbon. Similar methods have been used to identify reaction intermediates and intermediate rate constants in the catalytic conversion of CO/Hz [65.66]. The experimental systems used are quite similar to the one proposed in this investigation. Low system transient responses (approaching one second) I! . 'A Q .4 "r-. - v . ‘nok- ' C n. -. 0» o. 0" 14 were stressed in all investigations, as well as the use of mass spectrometry for rapid and continuous sample analysis. 1.3. Previous Research in Our Laboratory The doctoral candidate has performed a kinetic study of the role of oxygen in hydrogasification of Saran char and coal char in partial fulfillment of the requirements for the Master of Science degree. During hydrogasification. the only oxygen present is that which is initially associated with the char sample. Even though steam gasification of chars has been studied much more extensively by other researchers, hydrogasification was chosen in order to focus on surface oxygen groups more closely. Pertinent results of previous research are summarized in Figures 4-6. It can be seen in Figures 4 and 5 that hydrogasification rate decreases rapidly with carbon conversion, which is observed by other researchers [13.15.23—26]. Oxidation via partial combustion increases hydrogasification rate for 4—5% carbon conversion for intermittently oxidized chars. as well as heat pretreated chars. It does not increase the reaction rate when used as a pretreatment for as-received chars. 15 case :38 Co». 00230 9.9 5260:6393: ”v 059”. $682.00 c0900 8.0 93 30 3.0 oowd BEE. O ggogOOO G <00 mflmw—QMXXXX X DMD U . ooooo om0 <0 0 x a . Q go.— o o 0 x . m 0 wow 0 I 4 ooh, 880930 B . _x0 I SO x Eam x55 0 10.4 Sam .5». < <0 . 080 Sam 0 0 fl [OP-UIW/(dLSYHO 30] 8108 IU DISUI C20: 33:” v New...) m..:...._:_ < #4 A» \U..\: 15 0N0 .meo cmumm .8 002:0 9.8 53858023: ”v 050E 868250 20050 9.0 9.0 00.0 00.0 .— _ p _ _ p h — p _ p — p p b 0.0 0 BE. . 0RHEAHVOAH0000 00 QQOOmwO §XXX¥¥ BU . 00 < x a O 0 <0 0 to.— o o x o r m 0 Wed 0 v < 40.0 0080920 0 . _x0 I SO x . 5:0 ”RAN O 40.0 cam NE. < <0 . 080 $00 0 0 a [36.Ulw/(aLsYHo 33] 31.03:] iuoisul 16 .meo :80 Lou— mmzno Em... 5085800ch ”0 230E coEmEoo c0360 0N0 0m.0 9.0 000 00.0 00.0 . . . _ _ _ . _ . . . . _ . . . . _ . . . . 0.0 g. x .. no x coo-co O» l . . 0 OD . 0» 0N 0» 40.0 > . _x0 I 030 x .825 No > o -3. 088930 I . 080 080 O . oh 135 ~ U!w/(dLS)’Ho 30] 81,05 iuoisul .moum oomtsm .99. :0 0003 mice 5., 00230 9.8 53858093: ”0 930E c0_w._0>c00 CODENO «C0000Q om mm om 9 S m o 124:......._...._...i_....o+mo.o H O o o . m .. mmom w o H w 7 . m. D P . a > O Aimee m. . w . w p n I , . $3 «0. .mzocflmw . m. 2.20.80 b > . ( imoa 18 Figure 6 shows a hydrogasification rate curve based on total surface area that appears to reach a steady state at 15-20% carbon conversion. This set of observations supports a hydrogasification reaction that is composed of three stages, in which each stage dominates during different ranges of carbon conversion. During the first stage. hydrogen reacts rapidly with the small amount of secondary carbon formed during char pyrolysis. This carbon is amorphous and tar-like; it contains a relatively high concentration of heteroatoms compared to the bulk of the char carbon. The second stage involves hydrogasification of base char carbon via active sites formed by functional group desorption, and reaction of carbons located primarily on the armchair edges of the base char. Active sites formed during the first reaction stage may propagate to the base char and contribute to hydrogasification during the second stage. Rate during the third stage is low because the carbon atoms that react are primarily those on the relatively unreactive zig-zag edges of the aromatic planes. Reaction rate is now roughly proportional to the char total surface area because there are very few active sites or armchair edges left, making the char surface relatively homogeneous, and the only source of functional groups is oxygen trapped in the bulk char. Intermittent oxidative treatments fix oxygen functional groups on the char surface, increasing the number of active sites and therefore __.. .,.o .._. ..'_ u 1 ¢. . _- . - . '. 4.. - — c '. "r. 19 increasing hydrogasification rate by a factor of 2-3. These extra active sites are consumed rapidly, so the rate returns back to its base level over the course of 4-5% carbon conversion. The oxidative pretreatment was not effective in increasing reaction rate because fixing oxygen functional groups on the char surface is balanced by the removal of the highly reactive secondary carbon. 1.4. Research Objectives The main objective of the proposed research is to characterize the concentration, stability. and reactivity of hydrogen adsorbed on carbon surfaces at gasification conditions. and to relate this information quantitatively to the extent of hydrogen inhibition observed during gasification of chars. Though this problem has been studied for nearly a century, there is still unresolved conflict between well established researchers as to which mode of hydrogen inhibition dominates char gasification in steam. Mims and Pabst [61] state that product inhibition is due to reverse oxygen exchange. not hydrogen Chemisorption. Huttinger and Merdes [8] claim that inhibition caused by hydrogen Chemisorption is much stronger than reverse oxygen exchange. For the first time, the concentration of surface hydrogen will be included explicitly in linearized rate expressions that have been ..-‘ 4 .ur - .0». o o... - >.,_ 20 derived from several possible reaction mechanisms. The rate expression that most closely matches collected data should reveal the actual mechanism(s) responsible for hydrogen inhibition. Chars of Saran and coal will be gasified in mixtures of HALOb/Ar and mixtures of DNLNh/Ar. ranging in composition from 40%/0%/60% to 0%/100%/0%. The first of the two specific objectives are to use TPD to determine the concentration. stability, and reactivity of hydrogen adsorbed onto char surfaces before and after gasification at a fixed temperature while varying reactant gas pressure. composition, and char conversions. The second specific objective is to use this information to determine rate constants and reaction mechanisms by matching rate data. adsorbed hydrogen concentration, and reactant gas partial pressures to linearized rate expressions. 1.4.1. Mechanism Identification Comparison of actual experimental data with rate expressions developed from Equations 2-9 should reveal the relationship between adsorbed hydrogen and gasification rate at different gasification conditions. and identify to what extent the various proposed mechanisms contribute to hydrogen inhibition. Direct active site measurement has been done for steam gasification of chars by other researchers, however temperature programmed desorption was only done to 1373 K, and desorbed .. ~- ,,- ,\ ,,.. .- _. ‘ n ‘ A . ,.,..r9f ,. .. . .'. at - v " .. uv-- .0.-- r -- .qu '9. o--. — I " .... 0. ‘ ‘ D \ - — v. 0' ‘ .4. 0'..- . __ V» ... _ r~.. "' u . u ' J 'v’l . r:- .I .,1 b 'l 0 p. .-.‘ ’ n .u . ’b- 21 species concentration was not incorporated into rate expressions [45]. TPD will be performed to-1773 K in this investigation, therefore the concentration of C(H) surface groups will be measured and considered in the development of rate expressions for char gasification in steam/hydrogen mixtures. There are three major types of surface carbon that contribute to the total number of active sites. Active Site Balance: [CT]: [CF]+ [C(O)]+ [C(H)] (9) A "free" surface carbon. denoted by CF. is a carbon atom that is not saturated with chemical bonds. Carbons denoted by the symbol C(O) are bound to oxygen. Carbons that are bound to hydrogen are denoted C(H) in the active site balance. however this symbol may refer to either dissociatively bound C(H) groups or “associatively” bound C(FDz groups. There are several linearized rate expressions that can be derived based on the three possible modes of inhibition. If reverse oxygen exchange is solely responsible for inhibition, then Equation 8 is correct with a value of n equal to 1, and the surface concentration of C(H) is not included in the expression. In this case. there is no correlation between adsorbed hydrogen and reaction rate. 22 If dissociative hydrogen adsorption is solely responsible for inhibition. then Equation 8 is correct with a value of n equal to 0.5. The linearized form is as follows (See Appendix A-1 for derivation): C(H)—C(H)i=(_1_) 1 ,1 (10> rcot-Too k1 Evy- I‘2— In this equation, instantaneous rate at any conversion has been subtracted from initial rate. A plot of the left side, which can be determined experimentally. versus 1/Fm will be linear. In this case. the value of the left side is not a function of hydrogen partial pressure. therefore the same plot should result for all hydrogen partial pressures. If reverse oxygen exchange and dissociative hydrogen adsorption are both responsible for inhibition. the rate equation becomes slightly more complicated than Equation 8. The linearized form is as follows (See Appendix A-2 for derivation): C(H)-C(H): =(Lijfie., 1 (11) rco.i""co k1k2 F’w I; Instantaneous rate has also been subtracted from initial rate in this equation. A plot of the left side versus Pmflm should be linear if the assumptions leading to this rate expression are correct. There are several other possible rate equations that would result from other combinations of the proposed mechanisms. Inclusion of 23 “associative” hydrogen adsorption as an inhibition mechanism, or inclusion of non-rapid oxygen exchange will result in more complicated rate expressions that must be analyzed by linear regression. Using the above methods and rate expressions. the unsteady state behavior of char gasification over the entire range of conversion will be able to be accounted for for the first time. 1.4.2. Isotope Effects There are several ways in which rate may differ between char gasification in FbO/Hz and gasification iriihO/Dz. Effects caused by differences in adsorption [58], as well as quantum-mechanical tunneling [67], should be negligible at elevated temperatures. If oxygen exchange is reversible and rapid, there will be an isotope effect on the equilibrium constant for this step. If gasification rate is proportional to the surface C(O) concentration, as it is in all proposed mechanisms, then the ratio of gasification rates should be the same as the ratio of the oxygen exchange equilibrium constants, which is about 1.3 at 1023 K [61]. Another way in which isotope effects may manifest themselves in char gasification is in the breakage of the bond(s) involved in the rate limiting step. Effects here are a result of differences between the masses of atoms or groups of atoms surrounding the bond in question, and 24 are more pronounced given a greater relative mass difference. A primary isotope effect will result if hydrogen or deuterium is on one or both ends of the bond. The rate ratio is calculated by taking the ratio of the partition functions. and the difference in zero—point energies between hydrogen and deuterium. T- H H H H H H _ 0 Tee _ Qgthngglecanns exp_ E0 E0 (12) QrotherlecQtrans RT Since char gasification is a surface reaction, the only terms that will contribute significantly to the rate ratio should be the vibrational partition functions and the zero-point energy differences [58]. Vibrational partition functions are calculated with the following formulas [68]: 1 Qfibzw (13) 1 k The ratio of vibrational frequencies is 1.41, however the ratio of vibrational partition functions is what contributes to the overall rate ratio. The contribution, which is a function of temperature. is 1.3 at 1073 K. This value is based on a bond stretching force constant of 5.75 N/cm which was derived from spectral data [69.70]. Zero-point energies are calculated with the following formula [58]: 25 Eo=— (15) The contribution to the rate ratio of the zero-point energy term is about 2.4 at 1073 K. This is also based on a bond stretching force constant of 5.75 N/cm. A secondary isotope effect will result if hydrogen or deuterium is bonded to another atom that is involved in the rate limiting bond breakage. In the case of a CH2CN‘(Ik group, the difference in mass is quite small compared to the total mass of each group 14/16, therefore a secondary isotope effect would have little influence on the rate ratio. There are many reasons why observed rate ratios may only give very limited information about rate limiting reaction mechanisms. First. distinguishing between the relative contributions of various isotope effects is very difficult. Second, there may be differences between the zero-point energies of adsorbed and gaseous species. altering the contribution from this term. Third, there may be small contributions from other terms in the rate ratio equation or other phenomena that are unknown in the char gasification system. Because of these reasons. information gained by isotopic studies should be qualitative at best. Chapter 2 EXPERIMENTAL Detailed descriptions of starting materials, equipment, and experimental methods are given in this chapter. Equipment used in previous studies [71,72] has undergone several modifications and additions, including addition of a gas metering and blending system, a small packed bed reactor that fits inside the existing pressure vessel. a greatly modified reactor flange. a low dead volume steam trap. a high temperature ceramic reactor, and a mass spectrometer. 26 27 Experimental methods that allow characterization of surface hydrogen include the combination of steady-state gasification with transient step changes in reactant gas composition and temperature programmed desorption (TPD). Rate measurement characterizes char reactivity, while transient monitoring is performed immediately following gasification to characterize species that are loosely bound to char surfaces at reaction conditions. TPD to 1773 K is then performed in a separate reactor to characterize species that are more stable on char surfaces. 2.1. Starting Materials 2.1.1. Chars Samples used in this study are chars of Dow Saran powder (MA 127) and Illinois #6 coal (PSOC 1493), which have also been used in previous studies [73,74]. Chars are prepared by heating the starting materials in ultra high purity nitrogen at 10 K/min to 1173 K for one hour in a quartz tube furnace. Chars are then crushed and sieved to -60+100 mesh particles. In order to facilitate the desired measurement of adsorbed species following gasification, chars are further pretreated by heating at 5 K/min to 1773 K in argon for 6 hours in order to anneal and clean sample surfaces. Char properties are given in Table 1. 28 Table 1: Ultimate analysis and surface area of chars (wt% dry basis) Component Saran Coal Annealed Annealed (yt%) Char[75] Char[75] SaranChar[76] CoalChar[76] Carbon 96.4 75.3 97.76 87.19 Hydrogen 0.5 0.5 0.019 < 0.001 Nitrogen 1.0 1.3 < 0.5 < 0.5 Sulfur 0.4 3.6 0.0023 1.36 Chlorine na na < 0.0010 0.0062 Ash 0.1 17.3 0.46 15.00 Oxygen 1.3 2.0 1.78 < 0.1 (diff.) TSA (HF/g) 1330 440 800 15 (C02 at (C02 at (N2 at 77 K) (N2 at 77 K) 298 K) 298 K) 2.1.2. Reactant Gases All gases used are ultra high purity grade (99.999%) except for deuterium (Scott, 99.7% D). which contains 99.4% Cb and 0.6% HD. Two argon gases are used: one is doped with 1.0% krypton for characterization of system transients during gasification and is used as the diluent along with hydrogen and steam. while the other is used as an inert only. All gases are further purified by flow through R&D Separations Model OT500-2 Oxygen/Moisture Traps to remove water and reduce oxygen to less than 10 ppb. HPLC grade H23 and Sigma 083 (99.9 atom% 0), used for producing the reactant gas stream, are outgassed at .- 9"" ' ’\ .b n O o r 'c. s,“ “an s . .rr; . a _ ‘r-c. - "7 '4‘.) .d Tn. (n 29 373 K for 30 minutes and stored under argon to minimize dissolved oxygen in the steam. 2.2. Experimental Apparatus The required experimental apparatus consists of four major components: gas flow control and metering equipment, a high pressure gasification reactor (1300 K maximum temperature). a high temperature TPD/annealing reactor (0.4 MPa maximum pressure). and a mass spectrometer/vacuum system (gas analysis). A schematic of the overall system is given in Figure 7. 2.2.1. High Pressure Reactor The high pressure reactor consists of a horizontally mounted 51 mm 00 x 19 mm ID Haynes Superalloy pressure vessel capable of operation at 1300 K and 6.6 MPa, with a flange closure on one end. It is externally heated with a Lindberg 1400 watt electric furnace, driven by an Omega programmable temperature controller. It has been used and described in previous studies [71,72], but has undergone several modifications described in the following sections. 2.2.1.1. Internal Microreactor A small packed bed reactor was designed and constructed to fit inside the main pressure vessel to maximize mass transfer at the 30 wBSmaa< .mEoetoaxm __Eo>0 K 059”. mo<2¢3u X89 _ mQZFEu. a. 0.. www mobnaomm mmemmmL¥UMonel fittings. The microreactor. shown in Figure 8. is quartz lined, has quartz frits on both ends. and has quartz wool gaskets to prevent sample contact with metal surfaces and to prevent small entrained particles from reaching the sample bed. This provides an 8 mm diameter x 31 mm length sample chamber that can hold up to 300 mg of -60+100 mesh Saran char powder, or 800 mg of -60+100 mesh coal char powder. 2.2.1.2. Flange Modifications Figure 9 shows the main pressure vessel with the modified flange design. The flange sealing groove has been altered from the original design to hold a 41 mm 00 x 32 mm ID x 3 mm deep Variseal internal face seal with Turcite polymer compound and a 301 Stainless Steel spring to withstand an operating temperature of 573 K. The flange has also been fitted with several more taps for feeding purge and reactant gases to the inner microreactor, to facilitate inert gas purging and pressurizing of the outer vessel. and to allow for optimal thermocouple placement. The flange closure contains ports for reactant gas inlet and outlet. 32 8809922 .mESE .0085 9389a :91 ”w 059“. vllmll 1" .3: ummmm> mmammmmm Eu; E150 — .D._ EEMN Uw._>m I L mumDOoozmmTF .0.0 SE 0.9 33 20385002 omcwE _owmo> 9.3on :9: ”m oSmE 4...... ...mmmm>¥lt7 mg 2._<> mm .x \VIY mmnwmmmm J «0853055 1 20mm 1055010.! Hm I! Ch 2520.0 £89. . z _ . C /7 EOFQ044<¢mn5w mmZ> ngmem 222 34 thermocouple probes, and purge. The gas inlet and outlet tubing. as well as the sample holder, have been designed to have the lowest possible dead volume so that transients in gas composition can be followed. A Conoflow ABP Series backpressure regulator has also been added for immediate pressure reduction downstream of the reactor, which also minimizes effluent gas residence time between the reactor and detection equipment. During operation. reactant gases have residence times of one to two seconds in the sample bed. 2.2.2. Flow Control and Mixing System The mass flow control and mixing system, which can be seen in Figure 7, allows rapid switching and mixing of up to four gas streams at flow rates of 0—300 ml(STP)/min per stream. Pressure equalization between active and vented gases is crucial so that backflow of gases does not occur upon rapid switching of the two streams. Unused streams are exhausted during switching, because mass flow controllers may take long periods of time to equilibrate. For introduction of steam into the manifold system. a Series 1350 Bio-Rad Laboratories HPLC pump is used to inject water into the reactant gas stream at flow rates as low as 0.6 cc/hr (10 ml(STP)steam/min). The HPLC grade water is boiled to drive off dissolved oxygen, then stored under argon to prevent dissolution of more oxygen. All lines upstream of the reactor are traced with heating 35 tape and heated up to 470 K for steam formation and prevention of condensation. Steam can not be introduced into the vacuum chamber for detection with the mass spectrometer, so it is condensed immediately upon exiting the reactor in a cooled 1.6 mm 00 copper tube and separated in a 3 ml glass trap to minimize dead space and backmixing. This item is described in detail in section 2.2.2.2. 2.2.2.1. Gas Blending/Rapid Switching System Also shown in Figure 7 is an externally heated low dead volume gas manifold system which has been added and built to allow for rapid switching and accurate blending of purge and reactant gases. Gas flow is controlled with four Porter 201-FSVB Mass Flow Controllers capable of 0-300 ml(STP)/min and a Porter PCIM4 Four Channel Interface Module. Reactant gas pressure upstream of the reactor is measured with an Omega DP 2000 Strain Gage for minimization of internal volume. A Valco Instruments UN Series 2-Position 4-Port valve is used to rapidly switch between reactant and purge gases. and is also heated. Pressure is controlled with two Veriflo Series ABP-1 Back Pressure Regulators, also heated. one on the reactor line and one on the vent line. Aluminum pegs have been placed inside all sections of the flow control system that contain space greater than 0.125" diameter, including the mass flow 36 controller fittings, strain gage. reactor fittings. and backpressure regulators as a last step to minimize internal volume. 2.2.2.2. Low Dead Volume Steam Trap The capacity to use steam as a reactant gas has been added to the high pressure reactor, however it cannot be introduced into the vacuum chamber for detection as are the other effluent species. The assembly that allows for rapid steam condensation with minimum internal volume is shown in Figure 10. Product gases from the high pressure reactor pass through a backpressure regulator to reduce pressure to roughly atmospheric. The steam is condensed and trapped out of the effluent gas immediately as it exits the regulator. Condensation is achieved by directing flow through a 1.6 mm 00 x 750 mm length copper tube that is coiled inside an ice bath, while separation is achieved just below the bath where the copper tube empties into a 10 mm 00 x 45 mm length glass trap. The trap is very small in order to reduce dead volume between the reactor and detection system. Water level in the trap is controlled by opening a Nupro fine metering valve Just enough to allow a drip rate equal to that of the condensation rate. Such separation is reported in the literature by Saber et al. [77]. REACTOR EFFLUENT ........................................ - RING SUPPORT PLUG VALVE APPROXIMATE SCALE '—'—V—F'V_|'_V—V—V—V—' 0 6 10 contlmeters 37 TO MASS SPECTROMETER \/ ICE BATH W.- 600ml NALGENE BEAKER IIIIIIIIIIIIIIHIIHIH Iii ! , III F LIQUID WATER TO DRAIN IEIA ., ~ a [I] III .___, Ill 1 I: ! VI : ‘1.6mm O.D. COPPER CONDENSING COIL ‘0‘ SWAGELOK {s2 QUICK-CONNECT <9 o FITTINGS LOW DEAD VOLUME PYREX PHASE SEPARATION CHAMBER (3cc) _ PLUG VALVE 66cm TYGON TUBING _L :I- =I!Il§1lil '"IIII'IIIIIIIIII'I'" NUPRO FINE METERING VALVE WITH VERNIER HANDLE |||III|||||||I Figure 10: Low Dead Volume Steam Trap 38 2.2.3. High Temperature Reactor The high temperature reactor, shown in Figure 11, has been designed and constructed for this work specifically for TPD to high temperatures. Alumina tubes are set in an annular arrangement to facilitate influent gas preheating and low dead volume. Samples are held in a packed bed by a stationary porous alumina frit on one end and thin layers of alumina beads on the other. The beads prevent the char from spilling into the annular space between the thermocouple well and the middle tube because they are too large to fit in the annular space themselves. but are small enough to prevent the char from spilling past them. Hand—tightened fittings have been used for the parts that must be disconnected each time a sample is loaded or unloaded to facilitate rapid sample transfer. The reactor can be operated at pressures slightly above atmospheric and fits inside a Mellen Series 3 8400 watt split design 1800 K tube furnace equipped with a Eurotherm programmable temperature controller to facilitate linear temperature ramping. 2.2.4. Mass Spectrometer An Ametek Dycor M100M Quadrupole Mass Spectrometer is used to analyze product gases from gasification and TPD. It is mounted on a vacuum chamber that is capable of achieving pressures down to 10‘8 torr. 39 Lopomom 2:880 SBanEo... :9: “Z. 059". DJOEZ<2 20m". m<0 PZwDIEZJ mozFE m tE >m mIm-u I mmm2> .m- E III EuoEONV wothu. 2020 nl. mozFE xodussw flII (2.23% «539:. -.I- I'll I I mm? 5.23.2 I mmE. 5.23.2 ”5555:.“ mmom. Hfiwi II- ..o _ 2&2 .D. _ EEM I 2.2.: All .0. O EEON .......u..a...... .E 2.... F 40 is equipped with an electron multiplier for analysis of species concentrations down to 100 ppm. and is interfaced with a personal computer for data collection and manipulation. Product gases from both reactors as well as calibration gases pass by one end of a one meter quartz fine capillary tube which continuously withdraws sample gases to the vacuum chamber, and achieves the final pressure reduction of 10*- 10"5 torr. A list of all mass spectrometer controller settings is given in Appendix B. A Marvac Scientific A20 rotary vane vacuum pump was connected to the tubing system near the fine capillary draw point in order to vary the pressure there. and therefore vary the pressure inside the vacuum chamber. A Nupro fine metering valve was installed between the capillary draw point and the rotary vane vacuum pump to precisely control vacuum chamber pressure. The valve was sized in order to be able to vary the capillary feed pressure above or below atmospheric pressure. 2.3. Experimental Techniques Figure 12 illustrates the basic experimental technique used for most experiments, which start with heating the sample to reaction temperature under an inert purge gas in the high pressure reactor. Once 41 Temperature .920 .6 5:90me nmEEEmoa 93932.9 #cmzcmmnzm new 5:85me .8 902503 .mEmEtmaxm ”NF 9:9“. 9:: - - .2 m 1 ,._ - f . / .. ow: Ema—hm . . m u. s e x .. qu-I-l-IQ .0 J m N: .. u u u’.IIIIIIIIIII-I-‘.. ./II .c09< ONT. 3930 mde cosmoEmmO " 093a uonenueouoo segoeds 42 temperature has stabilized. the system is pressurized. Gasification of char samples to characterize rate behavior begins when the reactor feed is switched from argon to a mixture of steam, hydrogen, and argon using a Valco 4-port valve. The reaction is then interrupted after a predetermined length of time by a step change in reactor feed from reactant gas to argon. and transient desorption of active species is monitored. The system transient response is caused primarily by convective backmixing in the volume between the 4-port valve and the detector. and is determined by monitoring the decay of 1.0 vol% krypton in the reactant gas argon. The system response is subtracted from the decay of transient species to get actual desorption behavior. Figure 13 shows the system transient response to a switch from 5% hydrogen in argon (1% krypton) to pure argon during a blank run through the high pressure reactor at 3.1 MPa heated to 1123 K. The mass spectrometer detects a change in the effluent gas composition 15 seconds after the step change in feed gas composition. and takes another 25 seconds to reach steady state. A high scan rate was used for this experiment, which caused the krypton concentration to be resolved to only :0.2%. This high value is a result of the trade—off between scan rate and detector resolution. 43 % Krypton in Effluent Stream 0.0 wd me 2535: v md m mN N m... md o o§.oo.t.m.:. 7:... - V O wmcmco H -4444 4 4444 Em h I a. H .. 4 .H .. 4 4 H . O I. - 4 4 H I c295. Q IL . c393... O w m m C C .I . wm we .wucoomm om a 2. $2: 2 EVEN... was Soc :2 vans nmfimc 05:6 3:68:80 mg 30% E mmcmco 99m ”9 9:9“. manly; 1013993 u! uefimp/IH % 44 The experimental response time is very close to the theoretical purge time of 32 seconds. The reactant gases spend most of the time in the high-pressure section of the apparatus between the 4-way switching valve and the backpressure regulator. The internal volume of this section is about 5 cc. dividing this by a gas flowrate of 10 sccm (3.1 MPa) gives 30 seconds. The internal volume between the backpressure regulator and the detector is also about 5 cc. dividing this by a gas flowrate of 180 sccm (steam condensed) gives 2 seconds. Once desorption of transients is complete, the char is subjected either to further reaction, or is transferred to the high temperature reactor for TPD to 1773 K to analyze residual adsorbed hydrogen. Mass spectroscopic analysis of effluent gas species. as well as calibration gases. is done throughout the course of all experiments. Software for data deconvolution and manipulation is presented in Appendix C. 2.3.1. Gasification 2.3.1.1. Gasification Conditions Temperature. pressure. hydrogen/steam ratio, and time of reaction are the four parameters that can be varied for a given gasification in the high pressure reactor. Temperature is fixed at 1123 K, since it produces detectable effluent species concentrations and no significant 45 mass transport resistances (see Arrhenius plot analysis in section 3.1.). Reactions are conducted at 0.3, 1.0. and 3.1 MPa, with flow rates up to 300 ml(STP)/min and sample sizes of 300-800 mg. Steam fraction is fixed at 40%, and the hydrogen/steam ratio is varied by changing the relative amounts of argon and hydrogen added. Some reactions are conducted with no steam to clarify or extend initial results. The use of Odbflh mixtures as reactant gases allows distinction between hydrogen fixed on char surfaces during gasification, and hydrogen initially present on and in the bulk char. This distinction is critical because measurement of hydrogen fixed on char surfaces during gasification is the primary focus of this investigation. Since reaction rate and adsorption rate differ when conducting gasification lrlIbO/Hz or OANTk. the isotopic effect is measured experimentally. Gaseous species containing H and D are not allowed to contact each other to avoid the possibility of H—O exchange masking important results; however, specific H—O exchange experiments are performed to gain further insight into reaction phenomena. 2.3 1.2. Gasification Procedure The first step in all experiments is to open the vacuum chamber to the capillary inlet line and flood the vacuum chamber from 1x10"8 torr 46 when sealed to 3x10* torr with argon. Once pressure stabilizes. the mass spectrometer is turned on so the background has adequate time to stabilize. As the vacuum chamber purges. samples are weighed in the quartz microreactor liner on a Mettler AE100 Analytical Balance (0.1 mg). then loaded into the microreactor. The sample chamber is then sealed with a quartz frit and quartz wool gasket. which are held in place by a fitting inserted into the open end of the microreactor. The microreactor is then purged with argon, loaded into the high pressure reactor, and purged with argon again along with the main pressure vessel and bypass lines. As the reactor system is purged. a mass spectra of background levels of various key species is recorded and subtracted from a mass spectra of the calibration gas to obtain the true response of C0. C02. and CH4. The heating tape and furnace are switched on after a reactor purge of 20 minutes and stabilize at the reaction temperature after about 1.5 hours. At this time the reactor is pressurized with argon. the bypass line is pressurized with reactant gas, and the steam trap condensing coil vessel is filled with crushed ice. Once the entire system stabilizes. continuous mass spectra collection is initiated. After five or more scans a step change in reactor feed from argon to reactant gas is made to initiate gasification by switching the 4-port valve (see 47 Figure 7) that interchanges the flow paths of these two streams. Pressure. temperature. and reactant gas flowrate are all monitored and require small periodic adjustments during the course of experiments; some experiments require intentional alteration of reactant gas composition. After a predetermined length of time, the 4-port valve position is switched to create a step change in reactor feed composition from reactant gas to argon and the transients are continuously monitored. Once the reactor effluent composition stabilizes, mass spectral scans are discontinued and the vacuum chamber is isolated unless the experiment duration is longer than eight hours and requires another calibration. The reactor and bypass lines are then depressurized. the furnace and heating tapes are shut off, and the system is allowed to cool overnight under an argon purge of 2 ml(STP)/min. After the furnace has cooled overnight. samples are removed and weighed as quickly as possible. 2.3.2. Temperature Programmed Desorption 2.3.2.1. TPD Conditions Temperature programmed desorption (TPD) is performed in the high temperature reactor by linearly heating char samples at 5 K/min to 1773 K in 30 ml(STP)/min argon and holding for 30 minutes while monitoring '5'- ..;.or-~n; .4 - Iu-v' ' V" 9}.» ""3”, _e -' .u. I. I" ‘1'" pm (- Jun.A ‘ F. " \ l" 1‘ -b J». r» V ‘R '~ A y" ‘§ I ' :“f‘AIn . ’ i‘ i i j...' '7. "I »‘ IA 1 I ‘ ”1J1 VU‘A“\“ ’5' . 48 the evolution of species from the char surface. This method of analysis has been chosen over others such as XPS and AES because it is sensitive to hydrogen and is easily integrated into our experimental system. Heating rates and argon flow rates are adjusted depending upon mass spectrometer sensitivity. 2.3.2.2. TPD Procedure After samples are removed from the gasification reactor and weighed, they are loaded into the TPD reactor. Exposure of samples to air during sample transfer is minimized and does not affect adsorbed hydrogen on the char surface. Exposure to air can result in Chemisorption of oxygen to samples, since it is reported that oxygen physisorption reaches equilibrium in 15 minutes at 348 K and 0.1 MPa [78]. To check this. TPD profiles of samples that were transferred between reactors were compared to TPD profiles of samples that were not by Zhang [79]. and no differences were observed over a half hour time span. As in gasification, the first steps in performing TPD are to open the vacuum chamber to the capillary inlet line, flood the vacuum chamber with argon, and turn on the mass spectrometer. During this time samples are loaded into the ceramic reactor, followed by a small amount of 20 mesh Alcoa Chemicals tabular alumina. This is done to prevent sample 49 spillage into the annular portion of the reactor since the alumina is too big to enter this area. but small enough to prevent char from passing through it. The reactor is then assembled. sealed by tightening elastomer o-ring fittings, loaded into the high temperature furnace. and purged with argon. As the reactor system purges. calibration of C0. C02 and CH4 is done. The furnace is switched on after calibration and a reactor purge of 30 minutes. and the mass spectra of desorption species are continuously monitored as the temperature increases to 1773 K over the course of about 2.5 hours. After a hold time of 30 minutes. the furnace is shut off, mass spectral scans are discontinued, the vacuum chamber is isolated, and the system is allowed to cool overnight. 2.3.3. Gas Detection and Calculation of Effluent Rates Gas composition is determined by collecting raw mass spectra of effluent species and calibration gases, subtracting background spectra from these values, and deconvoluting the corrected values to account for molecular fragmentation and peak overlap between species. Raw mass spectra consists of partial pressures of the various molecular masses in the vacuum chamber; however all species can produce different partial pressures for the same influent mole fraction. A percentage of each 50 species will double ionize or fragment. causing further complications in mass spectra. Extensive calibration and data deconvolution is done to account for these phenomena of mass spectrometry. Calibration gases are used to determine the relationship between partial pressures of various species inside the vacuum chamber and the actual concentrations of key species. Background pressure in the vacuum chamber is fairly low at about 10‘8 torr, mainly from water and hydrogen. The only reactor effluent species that is not sent to the mass spectrometer for detection is steam, which is condensed and collected in a small trap immediately upon exit of the reactor. Computer software developed by Zhang [79] is used to deconvolute the peak overlap between species, and is shown in Appendix C. 2 3.3 1. Mass Spectrometer Calibration Calibration of the mass spectrometer is performed by scanning a blend of two AGA Certified Standard Multicomponent Gas Mixtures. The first contains 2.00% CD, 2.03% CI». 2.00% CH4. and balance argon, while the second contains 2.05% CD. 2.03% C02. 2.01% CH4. and balance hydrogen. Five scans of a blend of pure argon and pure hydrogen (containing no key species of interest) are taken to obtain background levels. averaged, and then subtracted from the average of five scans of the same blend ratio of calibration gases to obtain actual peak values. 51 The mole fractions of key species in the calibration gas are then divided by the corrected mass spectrometer peak values to obtain actual responses. An extensive calibration of the mass spectrometer was done prior to the core gasification reactions in order to ensure as much accuracy and consistency as possible throughout the course of this investigation. Detector response can be a function of inlet pressure, time, carrier gas composition. and species concentration. Molecular species fragment to varying degrees upon detection by mass spectrometer, causing overlap between the spectra of different species. These phenomena have been investigated and are detailed in the following sections. 2.3.3.1.1. Variation of Pressure at Capillary Inlet Mass spectra of both purge Ar (AGA UHP. 99.999%) and carrier Ar (Matheson UHP. 0.9910% Kr) were taken at various capillary inlet pressures in order to find the pressure at which detector response is the highest and to help identify the source of possible contaminant oxygen in the system. Figure 14 shows the pressure inside the vacuum chamber as read by the ion gauge and the mass spectrometer as a function of pressure at the upstream end of the inlet capillary tube. The ion gauge gives an overall pressure from all species present in the vacuum chamber. while the mass spectrometer pressure is a sum total of the 52 ovm .wmm omega comcm 53> 93on 6.5 3238 Co :ozocé m mm .mnEmco E3309 E 9385 ”3 2:9“. Ammxv 93¢.me ~25 meQmo 09. our ow ov Ste 93on .mumEobooaw $22 .1. mcfimwm 9:305 325 co. I.I o+mod m-mo.N m-m_o.v m-m_o.® m-m_o.w Ymoé TwNé meé (um) Jeqweuo wnnoeA u! eJnsseJd 53 partial pressures of all individual species. The ion gauge shows an increasing vacuum chamber pressure for increasing inlet pressure, while the mass spectrometer maximum response is at an atmospheric inlet pressure. The ion gauge gives values that are about an order of magnitude greater than the mass spectrometer. This is because the numerical value of the mass spectrometer total pressure is arbitrarily set and can be calibrated. Figures 15 and 16 show mass spectrometer background partial pressures of species as a function of total vacuum chamber pressure for pure argon and Ar/1% Kr purge gases. They all go through maxima between 2x10'5 and 5x105 torr. with lighter species having maxima shifted toward lower pressure and heavier species shifted toward higher pressure. Mass 32 (Ch) behaves quite similarly to the other species and has roughly 1/4 the partial pressure of Mass 28 (N2 + small amount CO) over the course of all inlet pressures tested. This indicates that the source of oxygen is one or several small leaks in the vacuum system, not contamination in the purge and carrier gases. The average maximum response for the various species occurs at an inlet pressure that is roughly atmospheric. so it was for this reason and for ease of experimentation that an atmospheric inlet pressure was chosen. 54 .Qommmdm coma n__._: n mmm moSn: 93me .mnEmco E::om> .m> mommmE w:o:m> Lo 9:306 _mEmQ uczoemxomm ”me 239“. CS: @930 :2 >3 .mnEmco E::om> E 9399a :33 Ymve TMNe Ymoé m-m_o.w m-mo.m m-wo.v mfloN o+m_o.o II I I — I - I — I I I _ I I I — I I I — I I I — I I I o+mo m-m_N m-m_v m-mo mmw 80me 2 6... 85 w III oh: (110;) .IeIequnoeds ssew Aq quweuo uInnoeA uI eInssaId Iegued 55 .Eofibx fie: coma a1: n mmm 093% 2335 88:20 Ezsom> .m> mommmE m:o_.m> .6 mcsmmma _mEma _ocsoaxomm ”or 2:9“. :5: mango co. 3 59:20 E::om> E 93me :30... Ymve Ymme Ymoé m-m_o.w m-m_o.m mmoé m-mo.m o+mo.o .._......._..._..._..._...o+m_o ormm mm? 36 845 3?»- 8 -1- mm; 83%....1- 6853-6... . $5va mNIOI Soul-I . (110),)JGIGUJOJ103dS ssew Aq quuIeuQ wnnoeA uI aInssaId [awed 56 2.3.3.1.2. Linearity of Response The relationship between key component concentration and mass spectrometer response was determined by changing the ratio of the Ar/1% Kr carrier gas to the calibration gas (2.00% CO, 2.03% COL 2100% CHM and balance Ar). Figure 17 shows the partial pressure of key components as detected by the mass spectrometer inside the vacuum chamber as a function of feed partial pressure. Varying the ratio between the two gases allows the key component concentrations to vary from 0.2% at 10% calibration gas to 2.0% at 100% calibration gas. The response of all key species is linear with concentration, which indicates that only one representative calibration needs to be made to determine the response over the detectable range of key component concentrations. 2.3.3.1.3. Response as a Function of Time An investigation of mass spectrometer response over a five hour period was done to ensure that factors such as significant background changes, detector drift, and unknown phenomena do not significantly change system responses over time. A five hour interval was chosen because this is about an hour longer than an average experiment. Figure 18 shows two sets of peak heights of each of the three key components of the calibration gas mixture. The peak heights for all species remain almost constant over the five hours. even though Figure 18 shows the two 57 .593 m_ mam .mEmo co=m5=mo - cosmbcwocoo 36QO >9. 53> mmcoamm. 55598QO $9: *0 3:99.: ”t 059“. 31x. :2 .I. oocflmnv mmw cosm5=mo 28ch co? ow oo 04 om o I II. I I I I I — I I I I — I I I I *I|III‘|II . /.|..I..I*..I:I:I:I:I.}:I: .+mo.o wIMON w-mo.v m-wo.® .......6 I. 35.x 3...... \\\I\ oo—. x vm mmmS. LI. H . =1... . ‘1‘ 3.3m: _..}i I. Wmo F X. mm 3.22 IOI . .4. 6382-1. I. 53 XX 3 $22 _...:0 H fiwvé (.uoI) Iaiewonoeds ssew liq Jeqweuo uInnoeA u! eInsseId legued 58 .5325 9:: So: m>= .05 @20QO 2:99.03 BEE new mEmcanoo >9. Co £96: xmma 53:530me 89: E 3:20 ”m? 059“. 29m>> cmmoegf or x we mmmS. or x N 93.2 «Co 4084 V? mmm: OO $00M mm 822 3.6 $84 mr mmm: .23 .9; m H _mEE I- I l l I l I I l I I I l I | I I l I l I I I I o+wo.o mImo.m fiwoe hImmé \IIMON Aflufi .N (1101) IaIaqunoeds ssew Kq quweuo wnnoeA uI aInsseId named 59 nuxjor background peak heights, It and FOO. change significantly. MLIltiple calibrations must be done for long experiments. but most are of stiort enough duration so that one representative calibration is SLJfficient. 2.3.3.1.4. Response as a Function of Carrier Gas Composition It is important that the argon to hydrogen ratio in the calibration carrier gas matches the argon to hydrogen ratio in the actual reactant gas. because the response of key species is a function of carrier gas composition. Figure 19 shows the vacuum chamber total [Dressure and partial pressures as a function of carrier gas composition rwanging from pure Ar to pure Dz. Neither species nor the sum total give rwasponses that are linear with carrier gas composition, which prompted aruother similar experiment to analyze the key component responses as a furmtion of carrier gas composition. Figure 20 shows that the key component responses are not linear with carrier gas composition. with the largest response occurring at 100% [D as the carrier gas. It was for this reason that a second Calibration gas withltzonly as the carrier was used. For ease of Calculation both calibration gases have the same key component Composition; blending them in the right proportions to match the 60 .coEmanoo mam .oEmo .m> .mnEmco Ezsom> E wwwmm 8E8 V6 $5365 :2th “av 939“. 209.4 8558 E25560 EoEon. 9:39.“. .98 III 59.4 . . O I E32930 _ -1 - mImv 9mm ®-ww mIM _. (1101) JGJGLUOJIOGdS ssew liq quweug u1nnoe/\ uI sseId named 61 59:05 E3309 E wEmcanoo >9. .6 0050005 _0_tmn_ now 0.59“. oow _:09< m_ mam 8:05:00 - cEEmOQEoo mmm EEmo .m> €3thme 025.08 000 5:05:00 E0200 0.4 om ow om 44 000.2 _:..: mm 822 IO: w? 0005. :.I m: 0022 _-1 . 3 0005. _:... wIwo.m Wmmé ”mm? (1101) 191eu1011oads ssew Aq Jeqweuo uInnoeA u! emsseJd 1191qu 62 expected reactor effluent carrier/reactant gas composition will give the appropriate responses for each experiment. 2.3.3.1.5. Fragmentation Investigation A fragmentation study of C02 was done so that the contribution to the mass 28 peak from fragmented C02 can be subtracted so that a correct CO concentration can be calculated. Figure 21 shows partial pressure of selected atomic masses as a function of partial pressure C02 and varying carrier gas compositions. The response of the primary CO2. peak (mass 44) is linear with C02 partial pressure. as is the mass 28 peak for the fragment species. 2.3.3.1.6. CD4/CH4 Response Ratios A comparative study of the mass spectrometer responses of C04 and CH4 has been done to investigate the feasibility of using CHI-containing gas as a calibration gas for the isotopic investigations which require analysis of CD4-containing effluent streams. Response ratios in pure Argon and 33% Ar/67% 02 have been calculated to be 0.6590 and 0.3420 respectively. and are defined as follows: CH4 peak height CH res onse = 4 p % OH. in carrier gas (16) CD4 peak height % CD. in carrier gas CD4 response = (17) 63 .N00 E080: H0_E 0:0 :oEmanoo 00m :0E0o .0> 3:02:90: 00 0:0 N00 .6 0050005 _0_t0n_ #N 059: €09< 005208 E20500 0:00:00 00 _. 00 CV ON 0 TI- . I annuuuu: 83 mm 80.2 .80 4.0.0 + 3 805. «00 48.0 If 63 mm 80.2 .80 4.00 _.0 . . 3. 8m: .80 $00 _ .1 . . Two... Wmmé nImoN (1101) 131au1011oeds ssew Aq JGQUJBLK) u1nnoe/\ u! emsseId |egued 64 00.. response ( 18) CH. response CD. I CH4 response ratio = Tlfie response ratios are clearly dependent upon carrier gas composition. 2.3.3.2. Mass Spectrometer Data Deconvolution Peak overlap between species is caused by a number of factors. including fragmentation, double ionization, and isotopes. Deconvolution is accomplished by using mass 84 as the primary peak for Kr, mass 44 for COL nmss 40 for Ar, mass 28 for C0 (after subtraction of contribution from C02 fragmentation). mass 15 for CH4 (to avoid confusion with fragmentation of species containing oxygen). mass 4 for Dz.Inass 3 for HD, and mass 2 for H2 (after subtraction of contribution from Dz fragmentation). The Registry of Mass Spectral Data [80] contains the mass spectra of most of the species involved in the gasification reactions of this study. Gasification and desorption rates are determined by combining knowledge of the relative amounts of effluent species. and knowledge of the total gas flowrate. Relative amounts of various species in the reactor effluent streams are obtained from deconvolution of mass spectral data. The total gas flowrate is calculated from addition of the separate influent stream flowrates minus water, which is possible because the mass flow control system is quite accurate and the 65 cxantribution to total effluent flowrate due to gasified char is negligible. Programs were developed by Zhang [79] in Basic to convert the raw rnass spectral data into spreadsheet form, deconvolute the mass peaks. and convert the peaks into species concentrations. and are shown in Appendix C. The first program, named “back.bas”. averages the calibration background scans and converts them into a matrix form. The second program. named “cal_h.bas". averages the hydrogen calibration scans, subtracts the background peaks calculated from running “back.bas”. then divides the fraction of hydrogen in the calibration gas by the corrected peak height to get the mass spectrometer hydrogen response. The third program, named “cal_c_m.bas”, performs identical tasks based on scans of the C0, C02. and C10 containing calibration gas with the added task of subtracting the contribution from fragmented C02 to the CO peak. The next set of programs processes the mass spectrometer data taken during gasification experiments, initial sample weight. total gas flowrate. and calibration information. and calculates effluent rates and carbon conversion. The fourth program, named “backt.bas". organizes the mass spectrometer output data into matrix form. The fifth program. named “main~dem.bas". reads the matrix organized in “back bas” and the 66 calibration information calculated in “cal_h.bas" and "cal_c_m bas". and calculates the evolution rates of various species as a function of time. The sample weight and total gas flowrate are input to this program directly, while background scans are taken during the initial scans of an experiment and subtracted out accordingly. The fifth program, named “main-ins.bas”. reads all of the information from “main-dem.bas". performs the actual peak deconvolution, and calculates the corrected effluent rates based on carbon conversion. Total carbon conversion is also calculated and is usually about 2% less than that calculated based on weighing the sample before and after an experiment. This discrepancy is because of the loss due to a small amount of sample kept in the core reactor quartz liner by static electrical charge. 2.3.4. Char Characterization The three methods of characterizing char samples chosen for this investigation include mercury intrusion, nitrogen adsorption. and X—ray (ii‘ffraction. Information such as pore size distribution and volume can I36! gained down to the mesopore level (2—50 nm) with mercury intrusion ark: down to the micropore level (<2 nm) with nitrogen adsorption. Cllfiystal structure information such as unit cell dimensions and ordered _o._. 0.0 :0300E00m E0000 :020 :0:0w 00.00::0 .6 #20 02:05:54. ”mm 059“. O: 0:30:0QE0._.> vad 4-mm.m #1de med Ymmd 4Im_4.w I I I I — II I I I — H I I I _ I I I I — I I I I NV- _0E\_00¥ mdm n 0m _0E\_0o¥ 9%.. .u 0m. x mnow Xmm: I0- an: . _oE\_0ov. Emm u 0m I m- 6:508. 0.00 u 00 00 Iol . 410 Ill . (u1u1¥35/ |ou1u1)u1 e193 u011n10/(3 72 calculation of the system at these conditions yields a value of 0.0072. which is more than an order of magnitude lower that the transition range (If 0.1 to 3 [83]. Details of the bulk modulus calculations can be found 111 Appendix D. 3.2. Char Gasification in Steam 3.2.1. Evolution of Char Surface Area Figure 23 shows total (BET) surface area as measured by nitrogen adsorption for annealed Saran and coal char following H20/H2 (gasification at 1.0 MPa as a function of conversion for varying reactant SJas compositions. The annealed Saran char has an initial surface area (If 800 nF/g and increases linearly to about 1500 nF/g at 30% conversion. Sthface area remains fairly constant over the rest of the char (ZCN1version, and is independent of reactant gas composition. The annealed CKDéal char has a surface area that is about an order of magnitude less tIfiéan that of the Saran. (1» adsorption was used to measure char surface at‘eas in previous studies [37 73.84] and produced similar results. Mercury intrusion was also performed on annealed Saran char to CI'iaracterize porosity at the macropore (>50 nm) and mesopore (2-50 nm) ‘Ievels. The macropore/mesopore surface area is 0.12 nF/g, along with an average pore diameter of 47.5 um. 73 .005. o... .0 E0000 E 50005000 .000 :0390000 :000E: >0 20:0 00.00::0001—0 0003.0 .000... ”mm 0.59“. :o_0:0>:00 :0900 0:00:00 8 E 8 B 9. om cm 2 o I I I I — I I I 1 — I I I q — I I II 1 d I I + q - I I 1 I - I I I I - I I II I 0 0 0 120.80 .0: 4.0 0 .20 5200 .0: 4.8 I .20 850 .m... 4.0 > .20 :98 .NI .8 0 0:0 :0..0w 00609:: 4 _I—L-VI > 1 . .. 6 I. . l I I I 0 H 0 004 000 00m F 000 _. 000m (fi/zw) u011d1ospv ZN Aq 391v eoepns 1910] 74 3.2.2. Rate Dependence on Pressure. Composition, and Conversion Steam gasification rate of chars at all conditions tested is reported in this section. Annealed Saran char is used for most of the experiments, with annealed coal char and unannealed chars used for a few representative runs. Unannealed materials were chosen in order to relate findings in this investigation with previous research in our 1 aboratory (Section 1.3). Steam gasification rate is reported as the sum of CO and CO2 formation rates because C02 is formed in the gas phase from C0 by the shift reaction [8]. shown in Equation 4. Effluent compositions for two reactions are compared to the equilibrium Composition of the shift reaction in Figure 24. The gasifications were performed under conditions of minimum and maximum hydrogen concentration 110 ensure that the two extremes of effluent composition were tested. At 13I1 :0:_000E00 E0050 8:000: 0:0 AN00+NIHONI+000 :0:000: 5:0 0:: :0 :0:_0an00 E:::___:0m. ”4m 0590 20 0:30:00E0... 000m 000m 000 _. 000 _. 000 0 I I I T — I I J I — II J I _ I I I I — I I q I O O . 20:0:00 £009.55 :0: 0:_0> 02030.00 I I N_._ 4.8 O NI 4.: 0 0 002:.” :0 5:35.00 800:0 0:0 :000 00.00::< . S. (9)!) 9501 76 iri the reactor effluent it would prove that it is formed on the char SI1rface. Effluent compositions of the same two reactions are also compared 'tC) the equilibrium composition of the methanation reaction, which takes pilace in the gas phase. Methanation Reaction: 3H2 +CO<——K—M£—>CH4 +H20 ‘ (21) I4L1ttinger and Merdes [8] performed char gasification under similar c:c>nditions and concluded that methane formation was on the char surface ()rily, not in the gas phase. Figure 25 shows the effluent composition of (brie reaction to be close to the equilibrium curve, but the composition ()1: the other reaction is significantly above it. Both effluent C:c3mpositions should lie on the equilibrium curve if methanation Cicnminates. Both points lie above the curve and indicate greatly Cii ffering compositions. indicating that methanation does not control C>L1tlet concentration of key species. :3 -2.2.1. Annealed Char Steam Gasification :3 -2.2.1.1. Annealed Saran Char Steam Gasification Several runs of Annealed Saran char gasified under identical (:onditions are shown if Figure 26. This set of conditions. 1.0 MPa and 40/0/60 beflh/Ar. was chosen for analysis because there were more 77 0.300050: .0> :0:_000E00 E0050 .2000: 0:0 A410+ONIHNIm+000 :0:000: :0:0:0::0E 0:: 00 5:000:50 E:::___:0m_ ”mm 0590 00 0.300083 ooow com? 000.. E0080 E:::___:0m :2 020) 09030.00 N1.4.000 0 «1.4.0 O 0:2; 0 8:00:80 E005 0:0 :000 00.00::< 0F 0N (9)1) 9601 78 ON ..<\NI\0NI 865v 5 0:2 3 a 8:85.000 Emma .06 5200 00.00::0 u_o 0:3 _0:0>00 So: 000: :o_S_o>0 NOO+OO “mm 059“. :o_0:0>:oo :0900 0:00:01 or NF w v 0 IIIIJ4IIIIIIIIlllllllllllllL mod _..o 3.0 Nd mmd md (Ulw.06/ Ioww) emu uounIOAa Zoo+oo 79 experiments performed under this set of conditions than any other. Error anaTysis for these runs is shown at 2% conversion in Figure 27. A conversion of 2% was chosen for anaTysis because it represents steady state gasification, and it has the greatest number of vaTues to anaTyze at steady state. The average rate at 1.0 MPa and 40/0/60 Hthb/Ar is 1.43 mmoT/gC*min with a standard deviation of 25%. This error is substantiaT, however char gasification rate varies weTT outside of this range at other sets of conditions used in this investigation. The gasification rate curves. which were caTcuTated from mass spectraT data. are accurate representations of actuaT reaction rate because the weight Toss caTcuTated from integration of the rate curves matches cToseTy with measured weight Toss. CaTcuTated weight Toss is consistentTy 1-3% Tess than measured weight Toss, mostTy due to a smaTT amount of sampTe retention in the microreactor due to static eTectricity. Rate data at aTT conditions tested are presented in Figures 28-33. AnneaTed Saran char was gasified under mixtures of Ar, F00, and H2 at pressures of 0.3. 1.0. and 3.1 MPa. Reactant gases for aTT experiments contain 40% FbO and varying ratios of'Fh/Ar, with the exception of two experiments presented in Figure 29. in which reactant gas contained no 80 .CO_m._0>COO COD:60 okN Hm wcz: _m._0>®w 50:59.. 85:8 2 was. 3 :0 838580 50% :0:0 :0:0m 00.00::0 E9: 000: :0_5_o>0 NOO+OO KN 0:09”. 00:52 E0E_:0axw 0005 @505 VNON Noon moon 035 965 wmon .--o\wm~.nnm.mm---- -- -- -.-----.mwM:m>o.-.--- ....... 500300 0:00:05 o+mod Ymoé Ymod «Wm—Ne m-mm._‘ mumoN (inofi/Ioww) emu uonnIOAa Zoo+oo 81 .005. In 00 50005000 E0000 :0:0 :0:0m 00.00::0 80:: 000: 503.90 NOO+OO ”mm 0:39”: :o_0:0>:oo :0900 E00:0n_ N m m v m N _. o I I I _ I I I — I I I — I q I - I I I - I I I - I I I - I I I NI$O® .1 «1.084 IIIIIIIIIIIII O muofi » <<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>D>>Fm 0.00.00.00.01: 50.0 _..o (umofi/Ioww) eiea uounIOAa Zoo+oo . 015. _..m 0.0 50005000 E0000 :0:0 :0:0w 00_00::0 E0: 000: 50290 0:0 ”am 0:39“. :o_0:0>:00 :0900 E00:0n_ N m m v m N F o 82 I — I I I — I I I — I I I — I I I T— I I I — I I I - I I I voo-o > >>>’>>>>>>>>>>>’>>> >>>>>>’>>>H xxxxxxxxxxxxxxx . xxxxx 44440 («a I” So I +I+III+I+JIIT IJEIIF +++++++ +++ jam. +++ NI nxbm I NI $m V NI $09. + NI $00 ‘ NI £00 . NI £00 .:< $9. X v.0 (inafi/Ioww) emu uonnlma VH0 83 .0n=>_ o._. :0 :00005000 E0000 :0:0 :0:00 00_00::0 E0: 00: 50290 NOO+OO ”on 050:“. Co_m._m>:00 CODING “c00:0n_ N 0 m v m N _. 0 I I I — I I I — III‘I 1— T! I‘ll - I|I H — I I I — I I I — I ITJI $08 I m: gov 4 «1.00 > I O lllllll l I NI $0 I' 4 4 4 >>>>>>>>>>>>>>>D>>> 5000.0 50.0 3.0 v.0 (in‘ofi/Ioww) eiea uonnIOAa Zoo+oo 84 .002 0.? :0 :00005000 E0000 :0:0 :0:00 00_00::0 E0: 0.0: 5:290 <10 #0 050:0 :0_0:0>:oo 50:00 E00:00 0 : 0 0 v m N F c . ..._. ._..-_..._...—...I..._... P000 0 O 0 o o o o o 0 00000. IIIII 5.0 NI 0000 NI o\oo¢ NI AXE NI o\oo C>4I (U!w..ofi/Ioww) eiea uonnIOAa VH0 llllll _..0 85 00.2 0.0 :0 :00005000 E0000 :0:0 :0:00 00_00::0 E0: 00: :o_E_o>0 NOO+OO “N0 0590 CO_w._0>COO COP—mo «Imp—0n— 0 N 0 m v m N F . NI $0 > NI nx00 . >>>>>>>>>>>>>>b>b>> _.000 5.0 V0 (U!UJ:.05/|0UJUJ) emu uonnIOAa Zoo+oo . 00.2 0.0 :0 50005000 E00:0 :0:0 :0:00 00_00::0 E0: 0:0: 5:290 010 ”mm 0:390 :o_0:0>:oo :0200 ::00:00 86 h m m V m N —. o I I I . I I I — I I I - I I I — I I I — 1I I — I I I fiI + I F0000 000 o o whoo>>o>>>>. p» tooooorffififlflu>n I lu_‘0.0 (UTwafi/Ioww) emu uonnIOAa VH0 N100 b «10.0 0 lllll v.0 ---‘¢ .4qr. Eric. 8!. .,H ;,.‘ fl 5"." I.-‘ :Vy ‘ (l) 30 I 1.. ‘ '0“ v .A . II .. ‘q ‘ I- ‘ I r 1 4 I) (n 87 F00. Deuterated reactant gases were not chosen for the buTk of the gasification reactions for reasons given in Section 3.2.3. Reaction rates at aTT tested pressures share severaT trends. the most obvious being a significant decrease in CO+C02 formation rate and significant increase in CH4 formation rate with increasing Fb partiaT pressure. A decrease in reaction rate for both CO+C02ENK1(}h formation over the first 1% conversion is aTso easiTy noted. and is much more pronounced under higher F0 partiaT pressures. There is added difficuTty in observing this phenomenon at higher reaction rates because of the Tack of detector resoTution with conversion. In the extreme case of a totaT pressure of 1.0 MPa and 60% H2 in the reactant gas, the CO+C02 formation rate approaches zero. This has been observed by Weeda et aT. [85] at higher H2 partiaT pressures. Reaction rate stabiTizes after the first 1% conversion for aTT conditions tested. and appears to remain at a constant vaTue or increase sTightTy at higher conversions. Gasification at 3.1 MPa and 1.0 MPa are compared in Figures 34 and 35 to show the effect of totaT pressure on species evoTution rates. CO+COzikwhwtion rate does not change with totaT pressure when no hydrogen is present in the reactant gas. but increases with totaT pressure when hydrogen is present. This is extremeTy pronounced in the case of 60% H2 in the feed gas where CO+COz'flonnation rate decTines to 88 000005000 E00:0 :0:0 :0:00 00_00::0 E0: 0:0: :o_::_o>0 N00+00 ”vm 0:300 :0_0:0>:o0 :00:00 ::00:00 : 0 0 v m N F o I I I — 1 I I — I 1 - I I I — I I I 1I I I q I I I — I 1 I Fooo.o u 0 0100000020.: 0 n. m: 31.080002; I _u o 310.00.00.23 4 .0 5:0 7m: 31.000002; 4 a 3100:0020: D an. m. 31.000002; > III-IIII-flmQ m. 31.080023 0 4 444444444 0 31.03002; 0 8.0 % >>>>>>>>>>>>>>>>>>> m >>>>>>>>>>>>>>>>>>>>>>>>>>>>>> w . 0. o o o ooooooQQGOAooooooQ I: W m w 89 000005000 E00:0 :000 :0:00 00_00::0 E0: 0:0: 503.90 010 ”mm 0500 :0_0:0>:00 :00:00 E0200 0 v m N _. 0 D > D > > 501038.»; 0 @8083». >>>>>>>>>>>>>>>>>>>v>>>>>flmbd>< a: $80 mas. o a: 0:00: 00.2 F. a: .08: mas. o a... $9.: 00.2 F. a... 0:0: mas. o a... $00 «as. F a: #9 mas. o F a: .09 was. . _. m _. m _. m P 0 n. 444444443M00 « InIIIIIIIIII 4 II4 D > O C _.000 _.00 . _..0 (inofi/Ioww) emu uounIOAa VH0 . Q :r“ 1 55V Ig-O‘P“h .- T . g.-.vi‘~ . _ I- r r..I\n- T‘ ‘4‘.- 1» .5“ J I ~ I. ~' _ Ir.f JIx Fl... .r‘- ‘C '~UTL‘ :3 A N0 alt U“; k. ‘ In 90 zero at a totaT pressure of 1.0 MPa, but maintains a Tow but easiTy detectabTe vaTue at a totaT pressure of 3.1 MPa. CH4 formation rate increases with increasing totaT pressure at aTT compositions tested, and shows a more pronounced increase at higher hydrogen partiaT pressures. Figure 36 shows the CO+C£0 evoTution rate of an experiment at 1.0 MPa in which the steam partiaT pressure in the reactant gas was kept constant at 40%. but the baTance of the reactant gas was cycTed between 60% Argon and 60% H2. IncTuded on this graph as a reference is a CO+COz evoTution rate curve for a char gasification at 1.0 MPa with 40/0/60 H20/H2/Ar feed. The CO+C02 evoTution rate. rapidTy decTines to nearTy zero upon a step change to 60% H2 in the feed. which is consistent with resuTts presented previousTy. After switching back to a feed containing no Hz, the CO+CIO evoTution rate rapidTy increaSes back to what it woqu have been if there had been no step change in feed gas composition. This shows that beyond the first 1% conversion hydrogen inhibition is compTeteTy reversibTe. Though the concTusion drawn was different, work done by Huttinger and Merdes [8] shows a simiTar resuTt. Quadratic poTynomiaTs have been fit to both anneaTed Saran and anneaTed coaT char surface areas as a function of conversion. Software incTuded in Stanford Graphics (version 3.0) was used to caTcuTate the equation for this curve based on a Teast squares fit, which can be seen 91 .:00_000E00 000 E0:000: 00_0>0 0:2, 00 .20.: :0 :00005000 E00:0 :000 :0:00 00_00::0 E0: 0:0: :00:_0>0 N00+00 ”mm 0500 :0_0:0>:00 :00:00 ::00:00 om 0N 0N 0F 2 m o I I I I 1 I I I 1 I I I I I I I J I IJI I I I o D _ J _ DD W W NW. .. m W 08000: .NI $0 0 m n m . + . $0831.00 0. . I . m D. - u u " Z . u n u 3 . . I I . A m m if? m. . YV . u. D" .CCCHD " m . m .160... m m. . 0 .. D u " IV“ T I N.0 w DDD m m N... .08 . ofl . u u u . am At t I” . I. .. N... 0:0 1 01°80 Y. N... .00 m m N... #0 . m. 0.0 _r P.I : 4 0A A I . V” ‘0! (I) (A! ’:‘;C 'd—I if ‘e , -. 'r I U. U ‘r~- . 4 .. ‘ ~ 1 ' vs 1 ‘n 4:. h ' I l‘ 1"" Fr. .. ‘I 92 in Figures 37 and 38. For anneaTed Saran char the TSA curve fit is as foTTows: TSA (NF/96) = 756 + 30.1x - 0.26x2 (22) Percent carbon conversion is represented by x. The TSA for anneaTed coaT char is as foTTows: TSA (HF/g6) = 8.15 + 16 6x - 0.38x2 (23) Figures 39 and 40 compare CO+COzeMKi(}0 formation rates from anneaTed Saran char gasification reported on a weight basis to formation rates reported on a surface area basis for a reactant gas composition of 40/0/60 HiflWb/Ar and 1.0 MPa totaT pressure. An increase in surface area is mainTy responsibTe for the increase in CO+CCO evoTution rate for the first 40% conversion. The rate is constant at 0.02 mmoT/nF*min except for the first 3% conversion, which is thought to be an experimentaT anomaTy as it is inconsistent with aTT other experiments. Past 40% conversion. the surface area remains fairTy constant but the (Xh{Ik evoTution rate stiTT increases, indicating an increase in reactivity of the char surface by a factor of about two. A simiTar trend can be seen with CH4 formation in Figure 40. but the first 40% aTso shows a mde increase in reactivity of the char on a surface area basis. OveraTT. the char surface reactivity for CH4 formation increases by a factor of about four. Figures 41 and 42 show the resuTts of an 93 00 .E00:0 :_ :00005000 0:050:00 00:0 000:.50 _0:0: :000 :0:00 00_00::0 0: :0 00:00:00 :000_ _0_E0:>_00 000:0030 Km 0:300 :0_0:0>:00 :00:00 E00:00 on 00 00 00 00 ON 0 _. 0 I—IIII-TIIIH—IIfiI—IIII—IIIW—J 1II—IIIJI o L 000 I 000 000 00N_. 000 _. (fi/zw) uoudJospv ZN liq 931V aoeyns |e1o J. 000 _. . CC“ 94 0N .E00:0 :_ :00005000 0:050:00 00:0 000t:0 _0:0: :000 _000 00_00::0 0: :..: 00:00.00 :000_ _0_E0:>_00 0000030 ”00 050.0 :0_0:0>:00 :00:00 ::00:00 ON 0:. 0 _. m CV 00 ON: 00: 00N (B/zw) uondJospv ZN Aq 991V eoeyns |e1oi 95 0.000 00:0 000::30 :5. 0:0 £0.02, :2. :00 0:0: ”:<:NI\ONI 00:0:00 :_ 00.2 0.: :0 :00005000 E00:0 :000 :0:00 00_00::0 E0: 0:0: :00:_0>0 N00+00 ”00 0:300 :0_0:0>:00 :00:00 ::00:00 00 E 00 00 av 00 ON 0F 0 0+0:..2:12....._...._...J...j...._.... 0 .me . m m NmF : 00 0 I0 H Ilenflocoo N0 7m: m . 0000 m. N N...u._N I ”-»”FFFFFFFFFFF. mu 0. ... I . m R . D I v.0 a m . DD . m .n N00 1 w, W _. .’. 0 w w. H I . w 2 . D .. 0.0 N m N-m.v .. b . m m . N 0000 00:< F . m. C . 0.0005025 0 m. Nuwm 0.0 96 CH4 Evolution Rate (mmol/m2*min) 2me 00.0 momtzm :00 000 £902, :00 .00 0:2 :<\NI\ONI 00\0\0v E was. 0.? 00 000005000 E0000 00:0 00.0w 00_00::0 E0: 90. 0000.90 3.5 ”0v 200E C0_m._0>c00 COP—m0 «£00.50..— 00 E 00 om 0v om om S o O+M0.0 1. . . _ 1 . . . _ . . . q _ . . . . _ . . . . _ . . . u — u u . . _ . . . . o+mo H IVL “ COO. . Ymom : 000 DD . . O. F F: - . Fflbbbbbb . N m: H DFFN All . $3 I . n >§O . . o 1 EN . ’"‘”‘ . $3 I . H >0 . . >0 . I 30 «(mod .r b”. . H 0. £00m 002 F . . » 28m. 2925 o . $00 Numv (Ulwofi/Ioww) 6183 uonnIOAa VH0 97 CO+C02 Evolution Rate (mmol/m2*min) .0_000 0000 000000 :0: _000 0.0903 :0: 000 0000 ”._<\NI\ONI 0992‘ 0_ 005. 0.? 0.0 0000000000 E090 0000 000w 00.00000 0000 00.00 0000.05 NOO+OO ”Sq 050E 00_000>000 000000 0000000 3 NF 2 0 0 v N o 0+m0.0..._1.._..._......_..._... . 0_00m000< F . . 200ng03 O . 0-00.0 - - . 4 0-00.0 - - . 0000.. 0-03 I n 000000 - . O F F . o n > > > > > . > . . . O O 0.03 I v0.0 mod Ned mwd (Ulwofi/Ioww) 9er uounIOAa Zoo+oo 98 CH4 Evolution Rate (mmol/m2*min) 0.000 0000 000t00 0.0: 000 000.02, 0.0: 000 0000 0,000.0sz 00550 0. 0n..>. 0.? 00 000000.000 E0000 0000 0000m 00.00000 E00 0000 000:.0>0 010 ”N0 050.“. 00.0.9000 000.00 0000000 0 3 N0 or w o v N ______ W 0.00m00.>>>>> >>>d n D > AIII. b“ w > o m . FCF» . . C O H o+mo.o v.m_o.m mfloé ”Mm... m-m_o.N mflmd m-m_o.m (in.ofi/Ioww) 9er uonnlma VH0 99 experiment that was performed under identical conditions, but to a much . lower conversion. The rate behavior over the first 3% conversion is consistent with the rest of the experiments performed in this investigation. Figure 43 compares CO+C02 and CH4 formation rates from annealed Saran char gasification reported on a weight basis to formation rates reported on a surface area basis for a reactant gas composition of 40/60/0 HALWt/Ar and 1.0 MPa total pressure. This experiment had to be performed for 24 hours because the reaction rate under these conditions is much lower than that of the previously described experiment. hence the missing rate data between 3-8% conversion representing the overnight period. The low reaction rate is also the reason why there is so much more scatter in these data compared to the previously described experiment. The CO+COzitwwmtion rate appears to abruptly increase at the start of data collection after the first range of missing data. but this is an artifact of opening the inlet valve to the vacuum chamber that the mass spectrometer is attached to. There is essentially no CO+COzikNwmtion under these conditions, which is consistent with other experiments in this investigation. The increase in CH4 formation is partially compensated for by the increase in char surface area. but 100 0.000 00.0 000000 0.0: 000 000.02, 0.0: .00 0000 0000159.. 05050 0. 00.2 0.0 00 000000.000 E0000 .000 00.00 00.00000 E00 0000 000:.0>m. ”00 050.“. 0.000 W .1000 .m M 0-00.0 .0. m /m\ - . 0 0000 a R .m 0-00.0 M 0 0V. 0-00.0 m-mo.m C0_m._m>COO COP—m0 acmohml 00 0 . . . . . . . d . 0+000 .- 0-00.0 0 fl . 1 0.00? .I. 1| I ENE E N-Mm._. n D >>>> . b b» . . 9? > > 0-00 0 I l h p . $0200.10 . 020 000.60 . ... 000.020 :00 0-00 0 ” 000.020 «00.00 . 0-00.0 (uinB/loww) 9193 uounIOAg 101 reactivity per unit surface area still increases slightly over the first 20% conversion. 3.2.2.1.2. Annealed Coal Char Steam Gasification Annealed coal char steam gasification rate is shown in Figure 44. The CO+CCb and the CH4 formation rates are about half that of annealed Saran char on a weight basis. Rate increases gradually over the first 10% and appears to be stable up to 20%. Pressure fluctuations at the initiation of the reaction have made analysis of possible transient behavior difficult. Figures 45 and 46 compare CO+C02eMKi(}h formation rates from annealed coal char gasification reported on a weight basis to formation rates reported on a surface area basis for a reactant gas composition of 40/0/60 HXNWb/Ar and 1.0 MPa total pressure. Formation rates of all species on a surface area basis are actually several times higher for coal than for Saran, because of the catalytic effect of ash. Unlike rate curves for formation of species on a weight basis. rate curves for formation of species on a surface area basis clearly show initial transient behavior. The initial decrease in CO+COz-as well as Clh formation persists to almost 10% conversion, compared to annealed Saran char which displays initial transient behavior over the first 1% conversion. This is most likely due to a higher content of dangling and 102 ..<\~I\ONI 856v 5 £2 04 “m. cosmoEwmm Emfim Eco _moo 028:5 E0... 39 co=:_o>m mmm Haven. ”Eu 2:9“. co_m.m>coo coEmo Emohwa CH4 Evolution Rate (mmol/gC*min) om 2 S m o o+mo.o i . 1 . _ . . . . _ . . . _ _ . . . . o r v10 F Ymon r NOO+OO O . 8.0 . .r > > > $3 > O . + } > . . I 00.0 . ‘ll' F . . > 1 $3 I > > > O O . O O O 0 0| I . . o D O I 23 $3 I» n u . mflme Nrd (in..ofi/Ioww) eiea uounlma Zoo+oo 103 Emma 85 momtzm :5 new £995 :2: ”ERIE? 855w 5 $2 3 a cozmosmmm 58% 5% _moo _um_mmccm E0: 99 co_S_o>m NOO+OO ”9‘ 23mm co_m.m>coo concmo E851 om B 9 m o u q u u — J d u d — u u — d u o um, . O m m fl .- l I I I I I I I No.0 3 V . I _m m I a M . 36 m... M w I m n . mfigé : a 0 . .1 11 m . - O I 8d m 0 . I H No w 2 l I O O / m 0 0 mod w m a C C C C V m... C 4. O O 0 M. to md “In“ Ii 104 2.0.2 mew. 83.3 “_:: Em 29m; :5 ”ERIE? 855v s was. 3 a cozmoEmmm Emma 5:0 .80 3.32m Ea: 9m. c2565 v10 Nov 939". 5652.00 5950 £00ch om 2 2 m o o+mo ..+._.iq._w. . . _ ..+. o+mo.o r ml .. Mm. TQ Q Q Q Q Q Q Q Q Q Q Q . Vo . O O m . 4 m , m mmm .I c_E..NE\_oEEV O F » FFF . mm . EEEQBEEV F F I Ymoo .m u - > ’I Q n m p . d . F 4 I H O. C l . F ’ > + o 3:... Jr 33 (inafi/Ioww) emu U0!1n|0/\3 VH0 105 amorphous carbons in annealed coal char. and perhaps the ash in the coal can catalyze the rapid reaction of carbons that are semi-ordered that would not be subject to this kind of reaction in the low ash Saran. Both CO+C02 and CH4 formation rates remain constant up to 20% conversion after the initial transient behavior. 3 - 2.2.2. Unannealed Char Steam Gasification Unannealed Saran char was gasified in H20 at a lower temperature than the annealed material with and without the presence of H2 in the reactant gas. Reactivity profiles at 1000 K and 3.1 MPa are shown in Fi gure 47 and share most of the same trends with annealed Saran char. 1 ncluding a decrease in CO+C02 formation and an increase in CH4 formation with increasing H2 partial pressure. Gasification rate declines over the first 3% conversion and then stabilizes with no H2 present in the reactant gas. but when H2 is present the rate continues t0 decline over at least 18% conversion. Unannealed coal char was also gasified in H20 under the same (3C3'1cjitions as the unannealed Saran char, and shares the same trends as W91 1 . Figure 48 shows that the only differences in reaction rate behavior between unannealed coal and Saran chars are that the effect of hb’drogen inhibition is less pronounced with coal char. and it takes up 106 ON .mn=>_ fin ncm v.83 “m cosmoEmmm Emmi Eco cmcmm nmfimccmcz Ea: 9m. c2565 mmm finned Kw 23E co_m._m>coo coemo Emocmn. 9. N_. w v 0.00.00.00.00... 0 0 0 0 0 0 0 0 0 0 0 0 a: $8310 > a: $3 $8 4 a: «Dog «8+8 - a: $9 «8+8 0 l lllllll I F .1 .Jllllll l llllll l Bed 5.0 _..o (ugwggfilloww) 9193 uouniona 107 or .mn=>_ fim 9.6 x 003 gm cosmoEmmm Emma .20 _moo 868295 So: 29 co_S_o>m mam 6:605 “94 9:9“. CO_mL®>COU Cop—m0 “COOL—0Q o_. w o v N I I. IIIIIIIIIII I E» > FFFFFFFFFFF a: $8310 «I «cog «8+8 «1 $9410 a: #3 «8+8 04IF 4444444444 4 4 I I .....’.}~. . . F 4 d _.ood _+o.o Pd (ugwgofifloww) 9193 uonn|0A3 108 to 5% conversion for the coal char gasification rate to stabilize under conditions of no H2 in the reactant gas. 3 . 2.3. Isotopic Studies Annealed Saran char was gasified in 40%/60% DzO/Ar at 3.1 MPa for a Kinetic comparison between gasification rates of annealed Saran char in deuterated and non-deuterated reactant gases. Figure 49 shows that the rate in 020 is about three fourths of the rate in H20 and follows the same behavior, which is consistent with results of Mims and Pabst [61]. Initial rate declines over the first 1% conversion, followed by a gradual increase over the rest of conversion. The difference in gasification rate between the two isotopes indicates that the breaking Of hydrogen bonds is definitely involved in rate limitation. 3 . 3. Adsorbed Hydrogen Concentration 3 - 3.1. Transient Hydrogen Desorption Verifying and quantifying the existence of transient species on Sal"nple surfaces during gasification is crucial to development of an Understanding of reaction phenomenon. H2 (or Dz) immediately desorbing 1rr‘Om the newly quenched surface of a gasified char would be indicative of weakly and perhaps "associatively" bound hydrogen, which would have t0 be included in active site balances and mechanistic models of the 109 .comeq oaoo w_ mocflmn mmm ESQ—m9 ”was. fin E 838530. Own. 5 ONI .mco cmamw nofioccm E0: 29 co=:_o>m NOO+OO ”av 9.sz co_w.m>coo coEmo Emocwn. o« S «P w v o - u I — - l u — u I u — I I u — - I I o 08 $04 9 . 81$? 0 H I mod . .z: 9099999 9 . to 99 u 9000 O O O. O. O . 00000000 . 0 0 0 I 2.0 O O O . o o o o o 0 o . Nd (U!UJ+05/|°UJUJ) emu uonnIOAEI Zoo+oo 110 c:har gasification reaction. Further. the system transient response must [36? subtracted from any transient species desorption to obtain an actual species transient response. Figure 50 shows high-resolution scans of the end of a Di) ggaasification, where there has been a step change from reactant gas to {DLJrge gas. Both thellzand the Kr show abrupt declines in effluent concentration after about 30 seconds. The [)2 transient response curve rneatches the Kr curve exactly, indicating that there is no loosely bound [3 (Hi the char surface that is lost with the removal of De partial F>rkessure. This finding shows that all of the surface Dz Um~lb) can be quantified with TPD: high-resolution scans for transient surface D2 (or H2) do not need to be done at the end of each char gasification. 3 .3.2. Adsorbed Hydrogen: Unit Weight Basis Adsorbed hydrogen concentration was measured on annealed chars following H20/H2 gasification by TPD in flowing Argon at 5 K/min to 1773 K and holding for 1 hour. Figure 51 shows two typical hydrogen TPD profiles for annealed Saran char following H20/H2 gasification to different extents of char conversion. There is a large and distinct peak for each TPD from 1100-1600 K that indicates dissociatively bound hydrogen. but no peaks at lower temperatures which would indicate “associatively” bound hydrogen. .3552. :.N “m Lsq $2: 9 Agricuo 09 ov E0: cosmoEmmm 5:0 cmcmw vo_mmccm 95% 5:809:00 mam noon. E mmcmco amaw ”om 9:9“. 111 % Krypton in Reactor Effluent Eevmep m N F o o ....:_...._.1+..o .. 00 n «.o .I mmcmco 1. mod n 36 n H 1. 3d to .. . 1” 8d 3 .1 H n 1” mod ma .1 a .m fi ‘ ouooooln _‘ o o r o o 00 . w P :03?! o 41“” ‘44? Nwd . E28500 4 4 :..-.00 . . u n O. U . . 8 m . . S . E o (ugwgofi/pww) 9193 uonnlma ZQ 112 .x mm: am cozmoEmmm :..mb 9.26:8 cmzo cmcmm 323:5 E0: NI .6 man cozeommn nmEEmcmoa mSfiEQEmH ”Fm 239”. O: 939888. 8m: 82 8«+ com 08 com . . _ . o+mo d d a an fl «.9 a A Q m 00.00 P... «-m« m a e 4 a 4 .. «-mm w . m. 4 44 . a/ 4 . O. “ -l Numv W. . m\E._.mvNI_oEE v._. .co_m.m>coo .898 $3 ‘ . mxathNIBEE 0.0 62902.8 598 $4 0 N..wm 113 “Associatively” bound hydrogen will adsorb onto chars if they are cooled in the presence of hydrogen after gasification. Figure 52 shows TPD profiles for annealed Saran char after hydrogasification and subsequent cooling in argon or hydrogen. The char quenched in argon shows only dissociatively adsorbed hydrogen, while the char quenched in hydrogen produces another TPD peak centered upon 900 K, indicating 0.02 mmolH2(STP)/gC of “associatively" bound hydrogen. This hydrogen is not stable on char surfaces during gasification because the TPD peak center 1' s 223 K lower than the experimental gasification temperature of 1123 K. Figure 53 shows adsorbed hydrogen concentration on annealed chars at several conditions. Surprisingly. the adsorbed hydrogen Concentration appears to be independent of reactant gas composition and pressure, spanning the range from 0% H2 at 0.3 MPa to 100% Hz at 3.1 Mpa. The value starts out very low at 0.02 mmolH2(STP)/g for unreacted Chars. increases rapidly over the first 1-1.5% conversion to 1.2 rTiniolH2(STP)/g for Saran char. and increases gradually over the rest of Conversion to about 4.0 mmolH2(STP)/g. The initial adsorbed hydrogen Concentration is close to the value given by ultimate analysis of 0.019 wt%, (0.09 mmolH2(STP)/g). The average adsorbed hydrogen concentration over the bulk of conversion is about 2.5 mmolH2(STP)/g, which matches the quantity given by the ultimate analysis for unannealed Saran char of 114 .895 .o 502?»: E 9508 .28 was. no om 53853093: 5% 5:0 cmcmw vofioccm 8.. ".5an on... c323: ”mm 939.... 0: 2:6chth com? com? oow_. com com com 3 . _ u . u . . — . . . 4 4 o+wo.0 F 0 3. F . > O O» . . 7M > o o» .H 33 m. >>o > H m. coop. > m . ! O _u z n 883 3x83 1 «ms; m I > . w > H a u u . «-mme m H OH o»n H m .4. :_ 0208 O .U «mom m. «I 5.8.08 > .. ( . «-mm.« 115 .x mm: :0 E000 5 00:00:08 0532.0: 20:0 00_00::0 E0: 0030:0038 c0020? 000.0094 ”mm 059“. ow om 02202.00 009.00 E0201 0N I I — a: $8: 850 a: $80 9200 a: s8 _moo a... $80 5:00 a: 400$ :98 a: $00 5200 a: $8 $30 0Frnuersion in shown in Figure 54, along with the product gas evolution rates. A reactant gas composition of 40/60/0 H20/H2/Ar was used to [nccxjuce the most pronounced rate decline with conversion, and to ensure 'theat the initial rapid hydrogen adsorption is not limited by the avaailability of hydrogen. Figure 54 shows that the initial transient befliavior of gasification rate and hydrogen adsorption take place over ttme same range of carbon conversion. The reason why deuterated reactant SJases were not to be used for the bulk of the gasification experiments 155 because the initial hydrogen concentration on the unreacted annealed Saran char is so low. The original purpose for using 020 and D2 as reactant gases was to be able to distinguish between hydrogen that was already present on the char and that which was fixed on the char during the course of reaction, but the temperature chosen for pretreatment is high enough to remove essentially all of the hydrogen in the Saran char. therefore this distinction does not need to be made. 117 Adsorbed H2 Concentration (mmol/g) ..<\«_._\o«: 00004 000 was. 0.: ”000.908 so. 00 003005000 E00:0 .050 00.0.0. 00_00::0 E0: cos—0:000:00 00090? 0000000 0:0 00. 02360 00300091 ”40 059”. C0_mcm>c00 COP—NO “£00.01 v.0 no.0 mo mNd No 920 _..0 mod 0 o I I I J +1I I — II I I - I I I I — I I I I — I I I I — I 1 I I fl I I I I who 0.0 .. a . +0.0 i I . . 1 . Q 0 C Q 4 « « . 0.0 .. . H 1 «0.0 0.0 I H . . I l no.0 N _. I NI 00900.0( IUI . . VIC 4 . . «0050 0 . m._. vod (ungfi/pww) 0193 uounlma see) ionpmd 118 3.3.3. Adsorbed Hydrogen: Unit Area Basis Figure 55 shows adsorbed hydrogen concentration on annealed chars eat all conditions on a surface area basis. Initial rapid adsorption (aver the first 1% conversion is followed by a steady, gradual increase (over the rest of the char conversion. The increase in surface area with <:onversion partially offsets the increase in adsorbed hydrogen concentration. Adsorbed hydrogen concentration on a surface area basis roughly doubles from 1% to 80% conversion, which indicates an increase in the ratio of exposed edge carbon atoms to those of the graphitic basal planes. Since edge carbons have a much higher potential for becoming gasification sites, this is consistent with the conclusion cirawn of an increased surface reactivity from gasification rate curves [Dresented on a surface area basis. Adsorbed hydrogen concentration on glasified annealed coal char on a surface area basis is about twice that c>f gasified annealed Saran char. which is consistent with char gasification behavior. 119 .v. 0N: 00 E000 5 02005000 00050.6: 000:0 00_00::0 E0: 00_00::00000 c0090.»; 000000004 ”mm 059“. 02002.00 00000 E00000 00 00 04 0« F 0 . . _ . . I _ . . . _ . . I a - I I ‘10 a: 0000: 0200 + +Im a: $000 0900 x cm I. a... $00 .000 o F u - a... 00000 00.00 I m I . a: A003 0900 4 O > la" 0 I F000 8100000200 > O + n - a... $00 0900 O 0 .fl > X1 0 >0 I m o I - I m I «00.0 0 O m - 0 I O - O - 1. 000.0 (aw/loww) uouenueouoo ZH pquospv Chapter 4 MODEL FITTING - LINEAR REGRESSION ANALYSIS The most surprising result of this investigation is that adsorbed flydrogen concentration is independent of reactant gas composition and exxtent of reaction past the first 1% char conversion. Because of this. ttie linearized rate expressions originally developed, which include e>nversion. It was also assumed that higher hydrogen partial pressures 120 121 vujuld lead to higher adsorbed hydrogen concentrations, also causing a ckecline in rate. However, none of these assumptions proved to be true. Vfliat was found was that. after a very short period of transient lDehavior, the adsorbed hydrogen concentration increased only slightly vvith conversion. as did gasification rate. Application of the original linearized rate expressions (Equations 10 and 11) to the data leads to the left hand side being negative, which cannot work because rate constants and partial pressures must all be positive. Further, the lack of resolution of data over the first 1% conversion prevents these rate expressions from being applied. where the adsorbed hydrogen and gasification rate appear to behave more like what was originally assumed. Instead of using the original equations with adsorbed hydrogen Cxancentration expressed explicitly, basic rate expressions were Ckeveloped from Equations 1-8 which contain a term that includes the tcatal number of active sites in a lumped parameter. Hydrogen surface Ccnncentration. as well as surface oxides and surface “free" sites, are Eill expressed implicitly. The equations have been linearized and used lri regression of the rate data for several postulated hydrogen lrihibition mechanisms. Rate data at all conditions tested were regressed as a function of carbon conversion to see how well the various 122 rncxjels describe actual rate behavior at different stages of char gasification in steam. Rate data was then normalized with char surface area for the best icepresentative models and regressed over the entire span of tested (:onversion and conditions. The regression coefficients. F statistics. rate constants, and the error on rate constant terms were then compared to determine which inhibition model gives the best fit and therefore helps to identify the most dominant hydrogen inhibition mechanism. Methane formation rate data has also been regressed to develop a better understanding of the mechanism of methane formation. 4.1. Char Gasification in Steam The basic rate expression for gasification of chars in Esteam/hydrogen mixtures (Equation 8). as well as other similar EExpressions. have been linearized so that all gasification data can be fi t to them to determine which expression best describes the rate beehavior. These expressions have been derived from the elementary reaaction steps given in Equations 2-7. excluding the shift reaction SIiven in Equation 4. Also included in the derivation of all expressions is the active site balance, given in Equation 9. Common to all EXpressions are the two elementary steps of steam gasification: oxygen exchange and C0 desorption. The mode of hydrogen inhibition is what 123 ciifferentiates the various expressions. Rate expressions include one or nuare of the elementary reaction steps of reverse oxygen exchange. “associative” hydrogen adsorption, and dissociative hydrogen adsorption. 'To further investigate possible reaction mechanisms. oxygen exchange and l"€V€FS€ oxygen exchange are assumed to be rapid enough to be at «equilibrium in some rate expressions, which alters their form by removing a term. 4.1.1. Reverse Oxygen Exchange The first hydrogen inhibition elementary step reaction investigated was reverse oxygen exchange, which involves no hydrogen adsorption on the carbon surface; the C(H) term is not included in the active site balance. The resulting rate expressions include the partial FDressure of hydrogen in the denominator raised to the first power (n=1). 41.1.1.1. Reverse Oxygen Exchange Equation 24 is the rate expression for HibWt gasification of Ctiars including reverse oxygen exchange as the inhibitory elementary Step. k1CTR~ = (24 ) 1+ (k1 /k2)R~ + (k.1 11033.2 rco 'Hne key assumption made in the derivation of this expression is that the cxancentration of the intermediate surface species C(0) is constant. 124 vdnich is typical of Langmiur-Hinshellwood expressions. The numerator t£3rm includes contributions from the reactant gas partial pressure. fkarward reaction rate constant, and total active site concentration. 'The first term in the denominator is unity and represents the oxygen (exchange elementary step, while the second term in the denominator represents C0 desorption from the char surface. The last term in the denominator represents hydrogen inhibition by reverse oxygen exchange. and is proportional to hydrogen partial pressure raised to the first power (n=1). Equation 25 is the linearized version of the previous rate expression. -‘-=l ‘ H ‘ 114-H ‘ ll-“ilpil rco k2CT k1CT F’w k1CT k2 Pw FTigure 56 shows the regression coefficient of determination (r2) as Cxalculated using the LINEST function in Microsoft Excel 5.0. Rate data at: all gasification conditions tested were regressed simultaneously at 0 . 0.5. 1. 2, 5. 10. and 20% conversion. The first 2% conversion has an P2 value of about 0.7, while the rest of the conversion range shows an P2 value above 0.9. An F test was performed on the data to make sure Tine good fit above 1% conversion did not occur by chance. Figure 57 Sfinows the F statistic and the F critical values for 95% and 99% Ccnfiidence. It is easily seen that the F statistic is near or below the .Awomv 00000000 00908 00.0>0. 3.0.. 00:00. 0000:0000 =0 0.0 .000 00.0w 0200000 :0 000005000 E00:0 E0: 000 0.0. :0 00.00.00. .000: .0: 00.05000 00:00:00 ”00 0.39.... 00.0.0000 000.00 0000.00 ON or NF w v 125 N.o v.9 mo mo IIJ‘JIJIIIIJIIIII o (wagomeoo uogsseJBeJ) ZJ 126 .Amomv 00000000 00908 00.0>0. 7:50.. 00:00: 00000000 :0 :0 .000 00.0w 00_00::0 :0 000005000 E00:0 E0: 000 0:0. :0 02000.00. .000: .0: 00:_0> 00.0-“. Km 0.39“. 02002.00 000.00 000.00 8000008 0.00 000000000 .0000 ...... .000. T... 0 0 lllLllllllllllllllll. 0 IO oo_. 09. CON omN SGn|BA 18913:] 127 critical values at and below 2% conversion. but is well above them for the rest of carbon conversion. 4.1.1.2. Rapid Equilibrium Reverse Oxygen Exchange Equation 26 is another rate expression for Hibflb gasification of chars including reverse oxygen exchange as the inhibitory elementary step. RASTPW 26 (k1/kzlpw +(k_1/kDPHz ( ) RX): The key assumption made in the derivation of this expression is that reverse oxygen exchange and oxygen exchange are in rapid equilibrium making desorption of the C(O) surface complex rate limiting. also typical of Langmiur-Hinshellwood expressions. The numerator term includes contributions from the reactant gas partial pressure. forward reaction rate constant, and total active site concentration. Unlike the previous rate expression, the unity term in the denominator is no longer present. This is because the first term in Equation 26, which represents CO desorption from the char surface. is now the dominant reaction term. The second term in the denominator represents hydrogen inhibition, and as with the previous expression it is proportional to the hydrogen partial pressure raised to the first power (n=1). 128 Equation 27 is the linearized version of the previous rate expression. ii 1 H 1 lhlf’fll rco szT k1C-r k2 Pw Figure 58 shows the regression coefficient of determination (r2) as calculated by the same method outlined for the previous expression. The results are slightly worse. but almost the same as those of non-rapid reverse oxygen exchange. included in this figure for comparison. The first 2% conversion has an r2 value of about 0.7. while the rest of the conversion range shows an r2 value above 0.9. An F test was performed on these data as well. Figure 59 shows the F statistic and the F critical values for 95% and 99% confidence. As with the previous rate expression it is easily seen that the F statistic is near or below the critical values at and below 2% conversion. but is well above them for the rest of carbon conversion. 4.1.2. “Associative” Hydrogen Adsorption The second hydrogen inhibition elementary step reaction investigated was “associative" hydrogen adsorption, which involves a diatomic hydrogen molecule adsorbing onto one surface carbon atom to form two discrete C-H bonds on the same carbon atom. noted as C(H)2. Hydrogen blocks water molecules from reacting with formerly unsaturated 129 .Amomiv 00000008 00968 00.0>0. 0E0. 505 70.0.. 000 EOE _.u..0.. 000000 00050000 :0 .0 .000 00.00 020050 .0 000005000 0.000 0.0.. 0000 000. .0 00.00060. .000: .0. E0_0E000 00000000 ”mm 0.390 00_0.0>000 000.00 oxo om 0? _ NF 0 v o . . . _ 4 . . _ . . . _ . . . _ i . . . o .00. n. « .mg I” 0.0 (iuagoweoo UO!SSGJBGJ) ZJ 130 0000-0. 000000x0 000.08 00.0>0. 00.0. 0.05 .u..0.. 000 A000. 3.0.. 00.00. 00200000 :0 .0 .000 0000 0200000 .0 02.005000 0.00.0 .00.. 0.00 0.0. .0 02000.00. .000: .0. 00:_0> .00 ._.-u_ ”00 0.:0_n_ 00.0.0500 000.00 000.00 om or N: w v 1 I I u u I I I I I I I I I I I 'IIIIIIIILII'llI'IILIIIIIII'ILII'II-III 'I'lll' h I Ir 4 000-0 .08 .000 II 000-0 _...8 $00 . - - - 000 0.000 4 000 .08 .000 II 000 .08 $00 ......... 4 .000. . u c O 4 llllllllllllllllll 00.. CON oom oov com senleA 1331-3 131 surface carbon atoms. thereby inhibiting oxygen exchange by decreasing the surface concentration of free sites, noted as CF. The active site balance. seen in equation 9. contains a C(H)2 term instead of a C(H) term. 4.1.2.1. “Associative” Hydrogen Adsorption Only Equation 28 is the rate expression for H20/H2 gasification of Chars including “associative" hydrogen adsorption as the inhibitory elementary step. k1CTPw 28 1+(k1/k2)Pw +(k3/k-3)PH2 ( ) rco = ‘Fhe key assumptions made in the derivation of this expression are that izhe concentration of the intermediate surface species C(0) and C(H)2 are c:onstant, which is typical of Langmiur-Hinshellwood expressions. The 'form and terms are all identical to those of hydrogen inhibition by I“everse oxygen exchange except the last term in the denominator. The ‘last.term in the denominator represents hydrogen inhibition by "éassociative" hydrogen adsorption and is proportional to the hydrogen Darfijal pressure raised to the first power (n=1). which is also the case VVlth hydrogen inhibition by reverse oxygen exchange. Equation 29 is the linearized version of the previous rate expression. 132 _L=[ 1 ]+[ 1 ](._1_J+[ 1 {Eli} (29) rco kZCT k1CT Pw k1CT k—a Pw Figures 60 and 61 can be referred to for the regression coefficient of determination (r2) and F statistic. The form of this expression is identical to that of hydrogen inhibition by reverse oxygen exchange. so the conclusions drawn from these figures are also identical: a poor fit from 0-2% conversion. and a good fit above 2% conversion. 4.1.2.2. “Associative” Hydrogen Adsorption and Reverse Oxygen Exchange Equation 30 is the rate expression for H0040 gasification of chars including "associative” hydrogen adsorption and reverse oxygen exchange as the inhibitory elementary steps. “‘0pr (30) 1+ (k1 /k2)Pw + {(1.1 Ik2)+ (k3 /k_3)}PH2 + {(k.1 lk2Xk3 lk-3)}Pfi2 rco = The key assumptions made in the derivation of this expression are the same as those made for “associative” hydrogen adsorption only. The form and terms are a combination of those for "associative" hydrogen adsorption only and reverse oxygen exchange only except the last term in the denominator. The second to last denominator term is proportional to the hydrogen partial pressure raised to the first power (n=1). and represents a simple addition of the separate contributions of 133 .A_.0_00000.. Pu..0.. 00.00. 00200000 :0 .0 .000 00.00 020050 .0 02.00.0000 800.0 0.0.. 0.00 0.0. .0 02000.00. .000: .0. .00_00.000 02.20000 ”00 0.:0E 020.0>000 000.00 .0800 ow m. N. w w l 1 [$1 I l l I l I L41 I l l l o O N. o “I o ‘0. o °°. o (wagomeoo UO!SSGJBGJ) ZJ 134 .._.0_00000.. Fu..0.. .00.00. 00200000 :0 .0 .000 00.00 020050 .0 02.00.0000 0.00.0 E00 0.00 0.0. .0 02000.00. .000: .0. 00:_0> .00 .7... ”.0 0.30.... 020.0>000 000.00 000.00 cm 0.. N.. w v 0 _fii...j o G O O I 00 .0 o 0 n. 00.. I” 00. 000000000 0000 I H 000000000 $00 ...... C A 000 21$ .1. O O . 00w sanieA 1391-5 135 “associative" hydrogen adsorption and reverse oxygen exchange. The last denominator term is the product of the two separate inhibition terms. and is therefore proportional to the hydrogen partial pressure squared (n=2). Equation 31 is the linearized version of the previous rate expression. é z [02:7 ] + [010. 13:?) + [k123T]{[::—2:|+|:Ek3_3:|}[%] +l..;.liell:+:lT%J Figures 62 and 63 show the regression coefficient of determination (r2). the F statistic and the F critical values for 95% and 99% confidence as calculated by the same method outlined for the previous expressions. Results for reverse oxygen exchange/"associative" hydrogen adsorption only have been included for comparison. The results are nearly identical to those of reverse oxygen exchange only and reversible “associative” hydrogen adsorption only, with only a very slight improvement over either one. This slight improvement caused by the addition of the term in which hydrogen partial pressure is raised to the second power is due to the fact that the model has another parameter by which it can regress the data and not by an improvement in the reaction description by including both elementary steps. 136 ..<_._< 000 000. N 0:0 3.0.. 0:0 2.2 .o 000. 3.0.. 00.00. 0000.008 :0 .0 .000 00.0w 0200000 .0 02.005000 0.00.0 0.0.. 0.00 0.0. .0 02000.00. .000: .0. .00_0_..000 00_.0_0..00 ”mm 0.09“. 020.0>000 000.00 000.00 cm or N? w v o I u I — I I I — I I I — I I I _- . .. .. .00 ._.-n. ”mo 0.09“. 020.0>000 000.00 .000.0n_ ON or N_. w v o _mwmhi. 1..I.-..I...u.1..|.......fl.m..l. _....a.ul...l:.._.1:....u|:fl.l. I...I...m..l..Lh....u...m:l:m.n...3.1.3.1......I..__.I:I:1.1:I:T m ....... I. on I” 8. 000 000.00 .000.00 ON 0. N. w v o ......l....3..._....0 2.10.0.0": O H . 2150000. F": O ..No .1. O 0.0 . Q md (wepupoo UO!SSQJ53.I) ZJ 141 .. .00 .7“. ”mm 0.30.0 020.0>000 000.00 000.00 1. om 1” 8. L 00. 000000000 #00 Ill H 000000000 2.00 ...... 4 210. 0.0.1. ” O 1. SN _.<_._< .o 000. . u c O y 0mm sameA 1391-5 op“ A I, , - I A r” r u v ‘ ‘tzrc 'A": v! 13 l. I) (f? '1‘ 142 the critical values at and below 5% conversion. but is above them for the rest of carbon conversion. 4.1.3.2. Dissociative Hydrogen Adsorption and Reverse Oxygen Exchange Equation 34 is the rate expression for Hibdb gasification of chars including dissociative hydrogen adsorption and reverse oxygen exchange as the inhibitory elementary steps. k1CTPw (34) 1+ (k1/k2)Pw + (k_1/k2)PH2 + (k4 Ik.4){P.l’22 + (k—1 ”031521 rco = The key assumptions made in the derivation of this expression are the same as those made for dissociative hydrogen adsorption only. The form and terms are a combination of those of dissociative hydrogen adsorption or reverse oxygen exchange. plus another term in the denominator. The third denominator term is proportional to the hydrogen partial pressure raised to the first power (n=1). and represents reverse oxygen exchange. The fourth denominator term is proportional to the hydrogen partial pressure raised to the one half power (n=0.5), and represents dissociative hydrogen adsorption. The last term is the product of the two separate inhibition terms. and is therefore proportional to the hydrogen partial pressure raised to the three halves power (n=1.5). .Vv-f o. ‘q‘ “J -.l dd b .+F“ ‘I INC-- ‘0... b. (I) . Hi“!- ‘ho‘ “.'x. 0..“ ~ lx'l| :u ' a “r .cr ‘- p. s. V; Pul‘ ii."- w fly... '» 143 Equation 36 is the linearized version of the previous rate expression. *1 ‘ H ‘ 163-H 11511-521 rco k2CT k1CT Pw k1CT k2 F’w .[ 1 llk—‘li? .[ 1 1.0 .0 P342 (35. k1CT k-4 PW k1CT k2 k_4 PW Figures 66 and 67 show the regression coefficient of determination (r2). the F statistic and the F critical values for 95% and 99% confidence as calculated by the same method outlined for the previous expressions. with results for reverse oxygen exchange/“associative” hydrogen adsorption and dissociative hydrogen adsorption included for comparison. The results show a clear improvement upon inclusion of reverse oxygen exchange with dissociative hydrogen adsorption as the inhibition elementary steps. The results are nearly identical to those of reverse oxygen exchange or “associative” hydrogen adsorption. This is because there is a term with hydrogen partial pressure raised to the first power, which appears to create a much better fit between the data and models than a term with hydrogen partial pressure raised to the one half power. 144 .A000 000.00 .0020... ON or NF 0 v o I I — I I I — I I I — I I I — l- I I :<:0 0:0 000. m... 0:0 . .00 u c ” . . .00 .7“. Km 0.:0_u_ 020.0>000 000.00 .000.00 ON 9 N. w v IIrILIIrILIIrILIIrILIIrILIIrILIIrIKlllrllllrllllr o <10 0:0 000 .08 .000 ll <10 0:0 000 .08 $00 . - - - 210 9.0 00% 0F .F .00 n F_ F <1< .o 000 .08 $00 | <1< .o 000 .08 $00 ......... O .<1< .o 000. F u c O O llllllllllllllllllll cm 00.. on. com omN :l sanleA 139 .1.- (I) l)- IJI' 'fin [l 0 lat 146 4.2. Comparison of Hydrogen Inhibition Models 4.2.1. Linear Regression Parameters The models which incorporate hydrogen inhibition by reverse oxygen exchange or “associative" hydrogen adsorption (“n"=1) correlate much better with actual data than models based on dissociative hydrogen adsorption (“n =O.5) for steam gasification of annealed Saran char. A very slight improvement to the models can be made by combining modes of hydrogen inhibition in a single expression, but this is due to the fact that there are more parameters with which the data can be regressed to fit the model. as seen in Table 2. Other rate expressions were investigated involving methane formation as well as a combination of hydrogen adsorption and rapid equilibrium reverse oxygen exchange, but Table 2: Number of parameters in various kinetic models of H20/H2 gasification of annealed Saran char. Number of Model Parameters "n” = ? 2 1 for Rapid-ROE (no constant term) 3 1 for ROE, 1 for AHA, 0.5 for DHA 4 1 and 2 for ROE + AHA 5 0.5, 1, and 1.5 for ROE + DHA ‘0 P“ Jr- .5 0-! 9... Tel 147 they yielded unnecessarily complicated rate expressions similar to those given in Sections 4 1 2.2. and 4.1.3.2. None of the models fit the data well up to 1% conversion because the reaction is not at steady state over this range. Hydrogen is rapidly adsorbing onto the carbon surface, therefore the concentration of various surface intermediates is not constant. Constant surface group concentrations are one of the key assumptions made in the development of the mechanistic models. It is not surprising to find the dissociative hydrogen inhibition model fitting kinetic data well at 10% and 20% carbon conversion because there are significantly fewer degrees of freedom for the models at the higher conversions. Table 3 shows degrees of freedom as a function of Table 3: Degrees of freedom for linear regression of various kinetic models of H20/H2 gasification of annealed Saran char as a function of conversion. Percent Carbon 2 Model 3 Model 4 Model 5 Model Conversion Parameters Parameters Parameters Parameters O 30 29 28 27 0.5 27 26 25 24 1 27 26 25 24 2 22 21 20 19 5 17 16 15 14 10 10 9 8 7 20 7 6 5 4 _. I'D ‘U -~r 0" _- ‘- .. .~-If-1 . . ‘ 4 e ,- ~ “_J - '4-‘ c" ‘zffia‘l '04 5' _' 148 conversion for linear regression of the various kinetic models. Carbon conversions lower than 10% have enough degrees of freedom to show a significant difference between the way models which include reverse oxygen exchange as an inhibitory mechanism describe kinetic data and those that do not include reverse oxygen exchange. At 10% and 20% carbon conversion the number of degrees of freedom approach the number of model parameters. making it much more likely that a good fit between the model and data will occur by chance as seen by decreasing F test values at these conversions. A further attempt was made to identify the actual inhibitory mechanism by comparison of the regression results of two different forms of the linearized rate expressions for both reverse oxygen exchange/“associative" hydrogen adsorption and dissociative hydrogen adsorption. Equations 36—38 show the alternate forms of the linearized rate expressions. The difference between these expressions and the ones presented previously (Equations 24. 28, and 32) is that each side of the new expressions has been multiplied through by the water partial pressure. 2w:1.2..liril.:.i+i.2.—iieiea 211320 In 27.1 nei‘ 149 ['00 1(ch 1 k1CT k1CT k._3 1 rco 1(ch 1 k1CT k1CT k-4 1 Table 4 shows the regression results for both linearized forms of the previous expressions. Rate data from 2—20% conversion has been normalized to the total surface area. so it is relatively constant and independent of conversion. These parameters therefore represent char conversion over the entire 2—20% conversion range. Form 2 of the n=0.5 model shows an improvement over Form 1 for both the regression coefficient of determination and the F statistic. but neither is as good as the n=1 models. The r2 value for both forms of the n=1 model are the same, while the F statistic for Form 2 shows slight improvement over that of Form 1. Inspection of the rate constants shows that Form 1 is better because it yields a positive value for the constant 1/kih. A positive value for the group 1/kflh also allows a positive value for the group kJ/kd to be calculated. 4.2.2. Calculated Rate Constants Hydrogen inhibition by rapid equilibrium reverse oxygen exchange gives almost as good a fit to the data as the model derived for reverse Oxygen exchange. The difference between the linearized rate expressions Table 4: Regression results for linearized rate expressions. 150 Linear Regression Parameters H20/H2 Gasification of Char r.2 F stat. F crit. (99% confidence) l/kZCT (gC*min/mmol) 1/ szr error szr (mmol/gC*min) 1/k1Cr (gC*min*MPa/mmol) 1/k1Cr error ler (mmol/gC*min*MPa) k-1/k1k2Cr (gC*min/mmol) k.1/k1k2Cr error kl/k-l (unitless) k3/ k-3k1Cr (gC*mi n/mmol ) ka/k-aleT error ka/k-a (l/MPa) k4/k-4k1Cr (gC*mi n*MPa“2/mmol) k4/ k-4k1Cr error k4/k-4 (1/MPa1’2) n=1 Form 1 n=1 Form 2 n=0.5 Form 1 0.66 59 4.1 2 .9 0 OCDF—J 8 b—Ir\)\l boo 149 14 -21 221 13 -184 . A " '. Art'— ~I- J» ’ ~r‘ f‘ i I -Jc— “ w: :r' 7“ V n .15.- ! u—r . “4.. .. .. .0....6 ’ — I b“. 'A. “A "’ 0d;— s....| V» -!-‘ ‘ 3 ‘1 (Tx .r'JIA“ ‘ 3"» (t‘t v (I) (f) ‘A I ’9 i’d‘» 151 (Equations 25 and 27) is the term (1/th)(1/FM). which appears in the reverse oxygen exchange expression. Inspection of the rate constant group (1/th) in Table 4 shows that it has a very low value. lower than the error calculated for this parameter. This indicates that reverse oxygen exchange is at rapid equilibrium. because this term is effectively zero. Comparison of the equilibrium constants for reverse oxygen exchange and "associative" hydrogen adsorption helps to distinguish between the two mechanisms as inhibitory reaction steps. which is the most difficult distinction to make because the forms of rate expressions for these two mechanisms is identical. The equilibrium rate constant for reverse oxygen exchange (kJ/kJ) is 0.029, which indicates a low fractional coverage of C(O) surface groups compared to the free surface sites. CF. The equilibrium rate constant for “associative” hydrogen adsorption (ka/ka) is 425, which indicates a high fractional coverage of (X102 surface groups compared to Cr. Table 5 shows that the rate constant groups found in this investigation are closely matched to those found by other workers. TPD studies by Zhang [79] have shown that the surface coverage of (X102 groups is negligible at reaction conditions. much lower than the coverage of C(O) groups. 0.02 nmmflFb/gC of “associatively" bound le5 A. . . fl .Hd liv 3 . 1 . .01.... A 0 o v I I ( (\. r a I U A: I» A .- 1. l i . o .,\J P a. . . . .. ,. .o .1. _..». K I. .K r .H :1. 0.44 i v .IlIU 1 duri 152 Table 5: Comparison of rate constant groups. Regressed Huttinger and Weeda and Rate Constant Group Data Merdes [8] Kapteijn[86] kflh (mmol/gC*min) 0.22 6.4 (0.33)* kmh (mmol/gC*min*MPa) 2.6 10 (0.52)* 2.3 kJ/kd (unitless) 0.029 0.025 ka/ka (l/MPa) 425 68.3 * rate normalized to 1123 K via E0 = 56.1 Kcal/mol hydrogen was seen at 900 K from the TPD of an annealed Saran char that was cooled in hydrogen after gasification. This range is below the experimental gasification temperature of 1123 K. therefore “associatively" adsorbed hydrogen should not be stable on char surfaces at gasification conditions and therefore have a very low fractional coverage. TPD studies have shown some surface coverage of C(O) groups at gasification conditions. CO desorption peaks up to 0.07 mmol/g for gasification in steam and less than 0.04 mmol/g for gasification in steam/hydrogen are produced after gasification at 1123 K. with peak maxima at 1200 K. Larger CO desorption peaks were also produced at 1200 K during TPD for gasification at 1000 K; 0.21 mmol/gC after gasification in steam and less than 0.04 mmol/gC for gasification in steam/hydrogen. This indicates a small but significant surface coverage of C(O) groups 4.. Q v-J- arrf“' “*3 (I) H ) ‘1 *_.l (I) .'r-t'-. 5‘ 153 that are stable at gasification conditions, which is consistent with the equilibrium constant calculated when reverse oxygen exchange is the inhibitory mechanism. 4.2.3. Theoretical Rate Curves Theoretical steady-state char gasification rate curves based on the models and calculated rate constants for reverse oxygen exchange (n=1). “associative" hydrogen adsorption (n=1). and dissociative hydrogen adsorption (n=0.5) are compared to actual rate curves in Figures 68-73. Both models fit well for char gasifications performed at 0.3 MPa, but only experiments with 0 and 5% H2 in the feed gas were performed at this pressure. The n=1 model produces a much better overall fit than the n=0.5 model. which is supported by linear regression analysis and by comparison of calculated rate constants. The n=1 model is off by a factor of 2 at 1.0 and 3.1 MPa and 0% hydrogen in the feed gas. but the fit improves with increasing hydrogen partial pressure. The n=0.5 model is very close at 1.0 and 3.1 MPa and 0% hydrogen in the feed gas, but as hydrogen partial pressure increases. the error grows to an order of magnitude. 154 .39: L n :.. _ucm an=2 fim “m 5:858... Emma .20 cemw uofioccm E0... 99. co_S_o>o NOO+OO ”mm 239“. 5.20250 209.8 .589. n m m v m N w o canoe. N... $8 ....... .. 2282. N... 3m _ - . -- 3%. N... #8 I 3%. N... gm > canoe. N: $9. I 26.85. N... $0 5%. N: #9. 4 5%. NI #0 O . >..».».».»:>:>..>:>..>..>..>.».».».».».».».>..>..>..>..>..>:>.».».»:>m coco-0000000.“ llllll I _.od (Ulw.05/Ioww) eiea U0!ln|0/\El Zoo+oo 155 .39: L. n :.. van was. o9 6 5.50539 63$ .20 52.3 uofioccm E0: 9m... co_S_o>o NOO+OO ”mm 239”. 5.90250 coemo Eweon. m v m N _. couoEv NI axoom 853 N... #8 I canoe. N: $8 5%. N... #8 4 canoe. NI «on .. . .. 333 N1 osm > :89... N1 «so I... 5%. N1 «so O 444. _I Inc-oil >>>>>>>>>>>>>>>>>> «an J IIIILI I IIIIII I I 39.0 5.0 _..o (U!w..ofi/I0ww) emu uonnIOAa Zoo+oo 156 .28.: i u :.. new mn=2 md Hm cosmoEmmm Emowm .20 Snow nflmoccm E0: 23 co_S_o>o NOO+OO ”on 059“. 20.90250 coemo Emeon. N o n v m N F o .........._..._..._........... .685. N... gm -. H 3633...»... > . 26.85. N... «so I m .m.mu.N..§o O I“ MN: ..>..».».>.Hfl->.>.}.>.>.>.v.}flfl.u IIIIJ LI Bed 56 _..o (in.ofi/I0ww) emu U0!lnl0/\3 zoo+oo 157 .80.: ..m.o u :.. cam was. In E. cosmoEmmm Emma .20 cmfiw 8.8ch E0... 9.2 5:290 NOO+OO ”E 959“. Co_m..0>COO COP—m0 vamp—mn— n m m v m N w o d u — _ u - — - u n — - d d - u u u — - u d — u I u — 1 d‘ I cocoEv NI $8 ......... 2395 NI gym ..... ‘ i .83. N1 $8 l .28. N1 «cm > :89... N1 $9. II .685. N1 go I . 3%. N... $9. 1 5%. N1 $0 0 . PPPPPPPPP>>>>??P??P>:>..>..>..>:>:>:>.m.?bm o o a u u u u a 0.0.6....” IIIIII l 50.0 5.0 _..o (in.06/Ioww) eiea uonnIOAEI Zoo+oo 158 ._onoE ..m.o n :.. new mn:>_ 0.? .m cozmoEwmm Emma .20 cmfim cofioccm E9... 3.2 8.565 NCO-.60 ”N5 939”. c0_m._0>c00 COD-_mo “Emu-5n— » N o m .V .N N F o . ..._..._...q..._.q._..._..._...FOOD _- . m 288.5. $8 ...... :..-.855: $9---- 4 I. m 36.3% $8 I 3%. N... gm b 4 44‘. o 2285. N: #8 canoe. NI «.0 II . a. .smENI $3 4 333% «so 0 S... m m. m. >>>>>>>>>>>b>>>>>>§~ We 9 -------------- u m.- |||l¢.ll|d.ll 1“ to w 0 O O O O O O O O 0 w. . W . 0. . m. U m. 159 ._ouoE ..m.o n :.. new was. ad 5 cosmoEmmm E86 .20 swam uofioccm E0: 99 :o_S_o>o NOO+OO ”mN 9:9“. 20.90280 .5950 “swoon. m N o m w m N N o ..._...q..ld..._i.._.q.1...fi... _.OOd canoe. NI osm- ..... . 3%. N... «on > . .685. N1 ON... I... m .98. N... £6 0 .m 5.0 l ».>.>.»:>->b.>>r.>:>:>:rwr.>.»>-w Into .GINIIIII_- I (in-ofi/Ioww) emu uounlma Zoo+oo IIIIJI I (I. p l 160 4.3. Methane Formation Only Expressions for methane formation excluding steam gasification were derived to determine the mechanism by which direct hydrogasification takes place. Methane formation by successive “associative" hydrogen adsorption gives a rate expression as follows. 2 rcm = ( “ch”? (39) k-3 /k3)+ (1+ k5 /k-)PH2 Methane formation by successive dissociative hydrogen adsorption gives a much more complicated expression. with a general form as follows. k-C-P-i- f1 (k)+ r. (k) "22 + coon-2 + f-(k)P-iz’2 (40) rem = Figure 74 shows the natural log of methane formation rate plotted against the natural log of hydrogen partial pressure. The best fit line to the data has a slope of 0.42, which is fairly close to 0.6. The only way that the models can describe this relationship is if the dissociative hydrogen adsorption expression is used with a dominant last term in the denominator. therefore we can conclude that direct hydrogasification takes place by successive dissociative adsorption. This is consistent with the findings of Zielke and Gorin [19] and Cao and Back [20]. 161 00700. 0:050:00 =0 #0 .050 00.0.0- 00_00ccm “_0 003005000 E00? 5 05000.0 _0_t00 0000.22 :0 000. 00:08.8 00050:. No 0800:0000 ”E 050E Ami—2V 05000.0 Eran. c0mo.0>I c_ _- o T N- m- w- m- I I I I - I I I I‘— I I I I + I I I I - I I I I — I I I I “I AN: 03:. x va0 + 03V- 129 1.0.5 _. (U!w-ofi/Ioww) emu uonnIOAa VHo UI Dir friars by @0965 Cf \fl'. . 14;: An IthU‘ P‘ 'Hth] Vii the . Chapter 5 CHAR PROPERTIES DURING GASIFICATION 5.1. Hydrogen Adsorption Direct measurement of adsorbed hydrogen over the entire range of char conversion has shown that hydrogen inhibits steam gasification of chars by two distinct mechanisms which remain active over different ranges of conversion. The form of hydrogen inhibition that first affects chars during gasification is initial rapid dissociative hydrogen adsorption over the first 1% conversion. Figure 54 shows this along with the initial rapid decline in steam gasification rate over the same 162 163 conversion range. This adsorption decreases rate by covering active sites. termed “free” carbon sites. on the char surface with stable hydrogen atoms. This happens regardless of reactant gas composition and is irreversible at reaction conditions. The second form of hydrogen inhibition becomes dominant once the dissociatively adsorbed hydrogen is in place. The small number of remaining unblocked active sites (unsaturated surface carbon atoms) are subject to the dynamic equilibrium of oxygen exchange/reverse oxygen exchange with water. Most of the surface oxide complexes are stripped of their oxygen atoms by gaseous hydrogen before they can desorb in the form of C0. Active sites do not become blocked with “associatively" adsorbed hydrogen because it is not stable at reaction temperatures. Other workers that have identified dissociative hydrogen adsorption as the only mode of inhibition have performed char gasifications under conditions of very low reaction rate where this mode may dominate over reverse oxygen exchange [6,43,48,49]. 5.1.1. Initial Rapid Adsorption Hydrogen rapidly adsorbs on char surfaces at the initiation of steam gasification. as seen in Figure 53 and 55. even with no gaseous hYdrogen present in the reactor feed. The source of hydrogen must be that which is liberated by C0+C02ithmtion as well as surface oxide Orr.- II A N -r- "u +0 a »~\ 3» .K.» 164 formation. so a comparison has been made between the initial hydrogen generation rate and the initial hydrogen adsorption rate. Hydrogen adsorbs on annealed char surfaces at a rate of 0.30 nmmflFb/gC*min over the first 0.5% conversion, but is liberated at a rate of 0.18 nmmflFb/gC*min due to C0+C02 formation under conditions of 0% H2 in the reactant gas. The difference between the two values is 0.12 anflFb/gC*min. and over the course of 0.5% conversion gives a total T value of 0.60 mmolFb/gC. The C0 desorption peak upon outgassing the char sample after several percent conversion accounts for 0.07 anflFb/gC, which is about one tenth of the difference. 0 The most likely explanation for this is that the quantity of surface oxide groups during the initial transient phase of gasification attains a value that is several times greater than that present after several percent conversion. The char surface is not initially blocked by strongly bound hydrogen. so a relatively high concentration of surface oxides can form. liberating hydrogen in the process. These oxides. along with transient char structural properties. contribute to the initial high gasification rate. After the initial transient phase the Quantity of adsorbed oxygen declines to a fraction of its former value due to desorption and displacement by strongly bound hydrogen. 165 5.1.2. Gradual Adsorption over Time Hydrogen continues to build up very gradually on char surfaces after initial rapid adsorption upon exposure to hydrogen containing reactant gases, as seen in Figures 53 and 56. A final surface concentration of 2.6x10‘3 mmolemf corresponds to 25% coverage. Unlike the first 1% conversion, however. the gasification rate also increases very gradually along with surface hydrogen concentration. Even though these changes are very gradual compared to the initial changes in char properties. they indicate that changes in the char morphology also play an important role in determining reactivity. 5.2. Char Structure Char structure is very important in determining rate behavior because most chars have significant pore structure down to the micropore level, and have surfaces which are heterogeneous and constantly being renewed with reaction. Varying configurations of carbon atoms. as well as heteroatoms. can all affect gasification rate. These properties can also change in relative importance with conversion because of the (nonstant removal of surface carbon atoms throughout the course of gaSification. 166 5.2.1. Total Surface Area and Pore Structure The annealed Saran char used in this investigation is extremely porous. with a large fraction extending below 2 nm into the micropore range. The first indication of this is the long time required to perform nitrogen adsorption on the annealed chars. The time it takes for nitrogen to intrude into pores is inversely proportional to the pore diameter, and eight hours are required to complete a nitrogen BET on the annealed char samples. TSA calculated by nitrogen adsorption is very high, about 1500 nf/g, which is also indicative of a microporous material. Figures 37 and 38 show TSA gradually increasing with conversion, indicating that char micropore structure opens up and allows gaseous species greater access to the internal surface. Char reactivity appears to increase even on a unit area basis at higher conversion as seen in Figures 39. 40. and 43, but this may be partly due to the fact that the new char surface area opening up is fresh and more reactive than that which has been exposed to reactant gases for some time. Further evidence of char microporosity is seen by comparing the N2 BET surface to that which is calculated by mercury intrusion. which only measures down to the mesopore (2-50 nm) level. Mercury intrusion gives a much lower value of 0.12 nF/g which is four orders of magnitude lower .‘a b v.3- ‘P is..." 167 than 1500 nF/g. The value calculated by nitrogen adsorption is a significant fraction of the theoretical maximum of 2600 nF/g. but this is due in part to pore condensation. Taking this into account. the surface area created by micropores is still at least two to three orders of magnitude greater than those created by mesopores and macropores. The average pore diameter calculated by mercury intrusion is 47.5 pm. but this is due mainly to the interstitial voids between particles. The chars were ground and sieved to -60+100 mesh, which corresponds to 150- 250 um. 5.2.2. Domain Sizes via H/C Atom Ratio Values for adsorbed hydrogen concentration and total surface area were combined to calculate an average ideal graphitic crystallite size for annealed Saran char beyond 1% conversion. At steady state gasification, the adsorbed hydrogen concentration is approximately 4 anflFb/gC, which corresponds to an H/C ratio of 0.1. Assuming an ideal graphitic hexagonal crystallite with zig-zag configured edge carbon atoms. dissociatively adsorbed hydrogen. no hydrogen adsorbed on the basal planes. and a C-C bond length of 0.1421 nm [87] a graphitic basal plane width of 3.8 nm was calculated. The number of graphitic layers per crystallite was calculated based on the above plate width. a total 'v-'. Ian'- .4,‘ I‘v.‘ 168 surface area of approximately 10001ng, and an interlayer spacing of 0.3364 nm [88] to be three. The theoretical hydrogen fractional coverage of these crystallites compares closely to that which was calculated based on the experimental hydrogen surface coverage of 2.5x10'3 mmolHymf. Assuming an ideal crystallite oriented so its basal planes are horizontal, the ratio of the area of the sides to the area of the top and bottom corresponds to the hydrogen fractional coverage. A value of 35% is calculated for the ideal crystallite. which is close to 25% calculated from experimental data. 5.2.3. X-ray Analysis An attempt to confirm the previously stated theoretical values of crystallite dimensions with X-ray spectroscopy using both angle scattering (Bragg's Law) for unit cell dimensions and peak broadening (Scherrer Formula) for ordered domain size shows that there is almost no detectable ordered structure in annealed Saran char. Reference graphites tested included and a 325 mesh and -10+60 mesh Alpha Graphite and a 360 mesh Ultra "F” Graphite, all of which showed strong peaks corresponding to an interlayer spacing of 0.336 nm and ordered domain sizes corresponding to thicknesses of 18-40 nm, or 50-120 basal planes. 6... u".( yvd Ar“n D .A .‘flp: 0&1 J p.-. - iv. (.17 a l l[:- :- Ti- u. (T! ) 169 X-ray spectra for all carbons tested are shown in Figures 75-78. As can be seen in Figure 75, the annealed Saran char shows no detectable peaks corresponding to spacing between graphitic basal planes. The finer mesh graphites show relatively small peaks corresponding an ordered spacing of 0.426 nm, which is the spacing between the zig-zag faces along the basal planes. This peak was detectable in annealed Saran char. but was nearly beyond the resolution of the detector, as seen in Figure 79. Analysis of materials for ordered domain size resulted in 20-40 nm for the reference materials, while the extremely faint peak in Saran char corresponded to a domain size of 3.2 nm, which is fairly close to the theoretical calculated value of 3.8 nm for ideal graphitic crystallite basal plane widths. The lack of prominent X-ray diffraction peaks in Saran char annealed at 1773 K is consistent with small highly disordered graphitic crystallites, which are similar in dimension to those calculated for ideal graphitic crystallites based on an H/C ratio of 0.1 and TSA of 1000 nF/g. Several investigators have shown an abundance of defects in small crystallites. such as increased interlayer spacing in networks below 5 nm [89] and turbostratic structure [90]. The latter type of defect is misalignment of adjacent basal planes, which contributes to the lack of lines in powder photographs. These investigators also make 170 00 0:5005 =“_.. 0:5 200:. com. .6 80.6000 .0.-x .3 0.50.... 0.0.-:- N om ov om ON or o . a10.-x 0N 0.30.... ov mums-r N om {II I WWI I 00.0000 .0>0_.0E_ 05:00.0 60:000.... 00.0 o+mo v+w F v+wN v+mm v+mv Mgsueiul |eufigs Kai-x 172 05500.0 050_< 500:. 8+3- ..0 82.0000 .0.-x KN 0.20.0 90.: N or 05.0000 .0>0_.0.:. 25500.0 5.0.5005. 00.0 _ o+m_o m+m_. m+wN m+mm Mgsueim leufigs AeJ-x 173 .050 50.0w 00_00::0 500:. 00:00- .0 E26000 .0.-x ”E 0.30.“. 0.05% N om ov om ON or cow CON com oov Ausueiul leufigs AeJ-x 174 km .050 :0.0w 00_00::0 500E 00:00- .0 E26000 .0.-x .0 .50_> 00550.2 ”0N 0.00.“. 0.0.-:- N 00 0v mV 5» NM ('0 CO jIIIIIIIJIJIIIIIIIIIJIIIIIIII or ON on 0v om om Misueim |eu5gs ABJ-X 175 the point that X-ray methods become less precise with decreasing domain size. as this broadens peaks. Graphitic crystallite size of dimensions similar to those obtained in this investigation were formed by annealing a cellulose carbon at 5 K/min to 1573 K, yielding a thickness of 4-5 layers and basal plane width of 2.1 nm [90]. 5.2.4. Char Morphology Annealed Saran char has a very high surface area. is highly microporous. contains graphitic structure over small domains. but also contains a fair amount of disordered structure. Initial rapid char gasification rate can be partly attributed to reaction of a small amount of highly disordered “dangling” or "glassy" carbon atoms which form as a result of char preparation. These amorphous carbon atoms are much more likely than graphitic carbon atoms to be bound to heteroatoms and have strained C-C bonds. Figure 80 shows CH4 formation rate during hydrogasification of untreated (1023 K). outgassed (1273 K under vacuum). and annealed (1773 K) Saran char. It appears as if the rate curves are approaching each other as carbon conversion increases. Untreated Saran char has the highest reactivity because it has the highest initial concentration of amorphous carbon atoms and heteroatoms. while the outgassed char has a lower concentration and the annealed char has the lowest. The amorphous 176 ..0E..00v. mnn0m 0.> 00000 0. 005.020: .0 .0 0.20.00E0. .00.). v.0 .0 :0..00.-..0000.0>5 .050 5000 E0... 0.0. 5:230 VIO now 0.30.“. :0.0.0>:00 505.00 200.00 ON 0w N—. m V O . . . . J . . . . . q . . . . . . 1 , _.od 669:: 8.85... o o o o 096 o 000000.:O> 900000960 ’ 7H 0.2.20 0 . 3 u m. 1. to m- m. -_ u H . e . m... H .0; g: 1. N w iS‘O’OOO >> 6“ 00 00 >. o. C C - m C C - 1W. 0 0 . .ow 177 carbon atoms and heteroatoms react rapidly. leaving behind a more ordered, homogeneous. and inert graphitic structure. The graphitic base structure is also critical to reactivity because carbon atoms of the basal planes are practically inert while edge carbon atoms are much more reactive. The ratio of edge to basal plane carbon atoms may be partly responsible for the apparent increase in char gasification rate in steam on a unit area basis for conversions above 50%. This ratio could easily increase if the edges of the basal planes react away simultaneously and at a relatively constant rate, but would decrease if individual basal planes reacted away sequentially. The configuration of the edge carbon atoms can also change with conversion. A graphitic basal plane with edge carbon atoms in both zig— zag and armchair configurations can be seen by referring back to Figure 2. Reactant gases which contain hydrogen have been shown by other workers to preferentially react with edge carbon atoms of the armchair configuration, leaving behind the less reactive zig-zag edges. This phenomenon lends a great deal of insight into the steam gasification mechanism. and is addressed further is the following sections. 5.2.5. Char Active Sites An active site in char gasification is a surface carbon atom that is not saturated with chemical bonds and has the potential to react with 178 one or more gaseous or migrating surface species. If the unsaturated surface carbon atom reacts in such a way as to be removed from the char matrix in gaseous form and leave one or more adjacent unsaturated surface carbon atoms, it has successfully propagated. If the unsaturated surface carbon atom reacts in such a way as to bind to one or more species but remain on the char matrix. it has been blocked. If it becomes bound to a relatively unstable functional group, the group may desorb and re-expose the active site. If it becomes bound to a stable group, such as dissociatively adsorbed hydrogen. the propagation of the active site terminates as it is blocked. Edge carbon atoms have the potential to become active sites in the case of a graphitic char. but the vast majority of these are strongly bound to hydrogen during steam gasification. Those that are not bound to dissociatively adsorbed hydrogen are bound to other less stable functional groups or are unsaturated. Some functional groups may destabilize adjacent carbon atoms and increase their chances of becoming active sites, most notably the off-plane oxygen functional groups of the recently proposed “universal” char gasification mechanism [55,56]. 5.2.5.1. Etch Pit Analysis Of critical importance to understanding how hydrogen affects edge carbon atoms during gasification is a series of experiments on etch pit II! ("In / l‘l\ 0.- VM \ hi I»)! 179 conformation performed by Yang and Duan [43] and Yang and Yang [6]. They used etch decoration/transmission electron microscopy on graphite samples after gasification at 923-1023 K to identify the shape and orientation of etch pits on the basal planes of graphite samples. Reaction with 02(N"(Xb produced round etch pits on graphite surfaces. seen in Figure 81. which indicate a combination of zig-zag and armchair edge configurations. Hexagonal etch pits with zig-zag edges only, also pictured in Figure 81, were produced by reaction with HA),100 then C02. C02 then H20, C02 plus H2, and H20 plus H2. They conclude that hydrogen Chemisorption is preferred on the zig-zag edges and therefore responsible for the anisotropy of reactivity toward the two principal edge configurations. This is clear evidence that the presence of hydrogen changes the behavior of active sites on the char surface, or at least limits behavior to mechanisms which preserve the zig-zag edge configuration and consume the armchair edge configuration. Of particular interest are two experiments performed with successive exposure of graphite to HA) and then C02.'Hu3(Xk gasification rate at 1023 K was greatly reduced after 100 gasification. while no gasification with CIh at 923 K was observed after reaction with F00. Hexagonal etch pits with zig-zag edges were detected following both experiments. 180 . . . . . . . . . . . . 0:030:030' Round Etch Pit with “030303030 .Q..... Zig-Zag and Armchair .Q.Q.Q. 0 0.0. . . .0.0.0 ... . Edge Carbon Atom Configurations . . . . . ..... 0 0 0 . . 0'0. 0 0 0'0 0'0 0'0 0'0 0: :0 . . . . . . 0 0: :0 0 . .. Hexagonal Etch Pit Q. . . .. . . .. . . .. Wlth ZIg-Zag Edge .. . . 0. Carbon Atom .0 : :: :0 Configuration 0: 8. 0 -: :- 0 0'0 .' '. .030 03030 0.0.0 . . . . . . . . . ....... '0'0'0'0 0 0 0 0 0.0.0.0.0. 0 0 0 0 0 0 0 0 0 0 0 0'0'0'0'0'0 0:0:0:030:030:0:0.0:030:0g080g03030 I.I.I.I.I.O.I.O.I.O.I.I...I.I...O.I I...I.I.I.I.I.I.I.I.I.I.I.I.I.I.O.I IIIOIIOIIIOIIOOIOI Figure 81: Round and hexagonal etch pits of the graphite basal plane. 181 These experiments show how strongly hydrogen is bound at gasification conditions on zig-zag edge carbon atoms. The preserved zig-zag edge configurations must be caused by residual surface hydrogen since CIb gasification alone produces round etch pits. This hydrogen is stable enough at 923 K to completely inhibit gasification. indicating complete blockage of active sites. At 1023 K.CIb gasification takes place at a greatly reduced rate. but in such a way as to preserve the zig-zag edge configuration. Since no hydrogen is present in the reactant gas, the hydrogen must be migrating from one zig-zag edge to the next zig-zag edge exposed by gasification. 5 2.5.2. Active Site Propagation The work sited above shows that the presence of hydrogen during char gasification changes the behavior of active sites from those which show no preference for leaving zig-zag or armchair edge configurations to those which show a strong preference for leaving zig-zag edge configurations. The most reasonable explanation for this is that under the presence of hydrogen, most steady-state gasification takes place via active site propagation along zig-zag edges. Reactant gases that produce round etch pits show no preference for leaving zig-zag or armchair edge carbon atoms under conditions of no hydrogen. Graphitic basal plane edges are unsaturated or covered “I f 182 primarily with oxygen functional groups which are less stable than hydrogen, and therefore more likely to react and initiate active sites. If some active sites do propagate, they do so in a chaotic sequence. Reactant gases produce hexagonal etch pits with zig-zag edges in the presence of hydrogen because there is a much more ordered sequence of desorption of edge carbon atoms. The lack of armchair edge carbon atom configurations indicates that they are more reactive and less prone to hydrogen inhibition. while zig-zag edges are preserved. If active sites were generated on the zig-zag edges they would become pitted and rounded out. especially if the active sites only propagated to a few adjacent carbon atoms before termination. Evidence of the difficulty of initiating active sites on hydrogen-blocked zig-zag edges was seen by Yang and Duan [43] and Yang and Yang [6] in the greatly reduced C02 gasification rates following F00 gasification. Active sites are generated at defects in the zig-zag edges such as armchair-like arrangements, crystallite corners, crystallite defects. and unstable functional groups. These active sites propagate along the zig-zag edges and remove the outer row of carbon atoms. preserving the edge intact. Figure 82 shows two possible modes of active site propagation along a zig-zag edge. They are both two-step mechanisms which alternate between the removal of a cyclic carbon atom (step a) and 183 -zag edge. Figure 82: Active site propagation along graphite zig (‘1) F4‘ 'l 184 a dangling carbon atom (step b). It is more likely that mechanism 1 is correct because it entails sequential desorption of the edge carbon atoms, while mechanism 2 does not and requires transfer of the active site between non-adjacent carbon atoms. Figure 83 shows two possible modes of active site propagation along the armchair edge carbon atom configuration. Mechanism 3 alternates between two configurations which consist of removing a cyclic carbon atom and removing a dangling carbon atom. The major problem with this mechanism is that two adjacent carbon atoms in the "sheltered” position must be skipped for every two adjacent carbon atoms in the “exposed” position that desorb. This is an even more extreme example of transfer of the active site between non—adjacent carbon atoms. Mechanism 4 entails sequential desorption, but is fairly complexas it must alternate between four different configurations. 5.2.5.3. Active Site Behavior with Conversion Char surfaces are covered with active sites and easily gasified carbon atoms at the start of gasification. Those that react rapidly include loosely bound secondary carbon atoms that may be glassy. dangling, or saturated with a greater concentration of heteroatoms. Desorption of the secondary carbon atoms. as well as low-stability functional groups, generates active sites (unsaturated carbon atoms) on 185 4a) Figure 83: Active site propagation along graphite armchair edge. 186 the char edge surfaces. The annealing pretreatment process also generates active sites by causing strongly bound hydrogen to desorb from edge carbon atoms. leaving them unsaturated. Over the course of the first 1% conversion for annealed Saran char, 10% for annealed coal char. and 5-15% for unannealed chars. the high concentration of active sites decreases rapidly. The "secondary” carbon atoms are rapidly consumed, and hydrogen rapidly and strongly adsorbs to most unsaturated surface carbon atoms. Also decreasing is the ratio of the more reactive armchair to the less reactive zig-zag edge carbon atom configurations. Figure 84 shows how an etch pit or graphitic crystallite can change in shape from rounded features displaying zig-zag and armchair edge configurations to hexagonal features with zig-zag edges. It is over the initial transient conversion range that the active site termination rate is much greater than the active site generation rate. causing a significant decrease in char gasification rate. Dissociative hydrogen adsorption is responsible for the blockage and termination of nearly all active sites over this range. Other examples of active site termination include “collision” or “canceling out" of two active sites together, complete gasification of an entire graphitic basal plane, or excessive steric hindrance from adjacent basal planes. 187 Figure 84: Round etch pit and crystallite converting to hexagonal features with zig-zag edges. 188 In the last case the active site may just be “stalled out” and resume propagation after adjacent layers have gasified. After the first 1-15% conversion (depending upon char type) the “secondary" carbon atoms have reacted away. very few armchair edge carbon atom configurations are left. and dissociatively adsorbed hydrogen is in place. The active site termination rate is low and now roughly equals the generation rate. and the char shows a fairly constant reactivity with conversion. Most gasification takes place via active site propagation along zig-zag edges. Active sites are in a dynamic equilibrium with species that weakly adsorb, such as oxygen functional groups. but the effect of reverse oxygen exchange usually removes the functional groups before they can desorb in the form of CO and leave another unsaturated surface carbon atom. The abundance of defects in annealed Saran char, as well as most other chars. ensures that there will always be surface carbon atoms that are more prone to active site initiation than those of the zig-zag edge. 5.2.5.4. Comparison to the “Universal” Mechanism The char gasification mechanism identified as universal by Chen. Yang. Kapteijn, and Moulijn [55] and Kapteijn, and Moulijn [56] represents an extensive and thorough body of work, however it is very unlikely that this mechanism plays a role in char gasification when HQ. EU 10 I a“ ‘l- 189 gaseous hydrogen is present. The effects of reverse oxygen exchange. identified in this investigation as the major mode of hydrogen inhibition in steady-state steam gasification, should be even greater for off-plane oxygen atoms because they only form a single bond to the adjacent carbon atom. Also. the universal gasification mechanism (see Figure 3) is shown with all edge carbon atoms saturated with oxygen, but in steam gasification nearly all edge carbon atoms are saturated with hydrogen. Chen et al. do state that off—plane oxygen atoms are not abundant in C02 and H20 char gasification compared to 02 char gasification. but in the presence of hydrogen they are probably in such low concentration that they do not contribute significantly to char gasification rate. 5.3. Rate Enhancement The practical application of the knowledge of actual mechanisms by which steam/hydrogen gasification of chars is inhibited by hydrogen is to enhance the reaction rate. Hydrogen presents problems for achieving this goal because small quantities bind rapidly and strongly to char surfaces. covering and blocking the majority of surface carbon atoms that are at the edges of the graphitic basal planes. Gaseous hydrogen will always be present in steam gasification of chars because one 190 hydrogen molecule is liberated for every water molecule that reacts with a surface carbon atom. 5.3.1. Partial Combustion Intermittent treatments of annealed Saran char with molecular oxygen at 700 K were done by Zhang [79] to test the duration over which fixing oxygen functional groups on char surfaces enhances steam gasification rate. Mild rate enhancements were produced during gasification at 998 K, which could be due to contributions from fixation of surface oxygen functional groups as well as changes in char morphology. Rate enhancement was not achieved for char gasifications at 1123 K given the same oxidative treatments. .The reason for the differences in rate enhancement is because a much greater concentration of oxygen functional groups fixed during intermittent oxidation remains on char surfaces during gasification at 998 K. These groups are not stable at the temperature chosen for the majority of gasifications performed in this investigation (1123 K). so they do not remain on the char surface and do not enhance rate. 191 5.3.2. Catalysis Catalysis with alkali salts and transition metals is the most common method of significantly increasing char gasification rate. Gasification rates can be 2-3 orders of magnitude higher in catalyzed systems. The effect of hydrogen inhibition is greatly reduced because of the dissociation site/spillover action of catalysts, increasing gasification rate by increasing the number of active sites, not by lowering the activation energy [61]. Sulfur poisoning. not hydrogen, is the main form of inhibition in catalyzed char gasification. Sulfur is present as both a heteroatom in coal and as a mineral component in the coal ash. Charring can volatilize most of the sulfur initially present in the carbon matrix. but not the sulfur in the ash. Most model chars have very low concentrations of sulfur and very little ash, so poisoning is not a problem. Coal, on the other hand, always contains sulfur and ash. Charring to volatilize sulfur in the carbon matrix is not a problem, but getting rid of the ash is expensive and time-consuming. Complete demineralization of coal samples during previous experiments in our laboratory required heated baths of hydrofluoric, hydrochloric. and nitric acid [74]. Ash particles can catalyze steam gasification is some systems as is seen by comparison of rates based on surface areas of 192 annealed Saran and coal char, but this enhancement is fairly mild compared to catalysis with alkali or transition metals. Chapter 6 CONCLUSIONS/RECOMMENDATIONS 6.1. Char Gasification Mechanism Identification Char gasification rate behavior and properties of annealed Saran and coal char during gasification in steam were studied in order to identify the mode(s) of hydrogen inhibition at various stages of char conversion. Langmiur-Hinshellwood type linearized rate expressions based on the three principal modes of hydrogen inhibition were regressed 193 194 with rate data collected at 1123 K and varying reactant gas compositions and pressures. None of the three principal modes of hydrogen inhibition give rate expressions that fit the data from 0—1% conversion. where transient rate and adsorbed hydrogen behavior is observed. Beyond this range gasification rate and char properties are nearly constant. The expressions derived for reverse oxygen exchange and “associative” hydrogen adsorption, which are identical in form, fit the data much better than the expressions derived for dissociative hydrogen adsorption. Further comparison of the linearized rate expressions for reverse oxygen exchange and "associative" hydrogen adsorption shows that the equilibrium constant for reverse oxygen exchange indicates low fractional coverage of intermediate surface oxides. while the equilibrium constant for “associative” hydrogen adsorption indicates high fractional coverage of “associatively" bound hydrogen. Transient desorption and temperature programmed desorption of chars following gasification show low concentrations of surface oxides and no “associatively” bound hydrogen, therefore reverse oxygen exchange is the active mechanism by which hydrogen inhibits char gasification in steam above 1% conversion. 195 Temperature programmed desorption has been used to identify the major mode by which hydrogen inhibits gasification in steam over the initial transient range of char conversion. Dissociatively bound hydrogen covers the annealed char surface very rapidly at the onset of gasification, going from nearly zero to nearly saturated over the first 1% conversion. This adsorption behavior was found to be independent of hydrogen partial pressure. Gasification rate decreases significantly over the first 1% conversion. and is more pronounced under higher hydrogen partial pressures. It is concluded that dissociative hydrogen adsorption is the dominant mode of hydrogen inhibition over the initial stage of char conversion, however changes in char morphology also play an important role. 6.2. Char Structure During Gasification Char structure significantly affects rate behavior because most chars, including those used in this investigation. have a significant micropore network as well as several different carbon atom configurations. These properties can change in relative importance with conversion because of the constant removal of surface carbon atoms throughout the course of gasification. Initial “dangling” and “glassy” carbon is the result of charring, and contributes to initial high gasification rate. On the more ordered graphitic features, gasification 196 ratxe'is affected by the ratio of edge to basal plane carbon atoms, as vwell as the ratio of zig-zag to armchair edge configured carbon atoms. The largest features that display ordered structure in annealed Sharan char were found by TSA, H/C atom ratio, and X-ray scattering to be gucaphitic crystallites about 3.5 nm wide and three to five basal plane 'layers thick. These crystallites are quite small and filled with (defects, which are more prone to become active sites than crystallite lattice carbon atoms. A gradual increase in the ratio of reactive edge (:arbon atoms to inert basal plane carbon atoms can explain the very (gradual increase in gasification rate and adsorbed hydrogen cxancentration past 50% conversion. A more abrupt increase in the ratio (If inert zig-zag edge carbon atoms to the more reactive armchair edge (:arbon atoms can contribute to the initial rapid decline in gasification rate over the first 1% char conversion. Edge carbon atoms have the potential to become active sites. but 'the vast majority of these are strongly bound to hydrogen during gasification in steam. Experiments on etch pit conformation performed by Yang and Duan [43] and Yang and Yang [6] identify the shape and orientation of etch pits on the basal planes of graphite samples to be round in the case of gasification with non-hydrogen containing reactant gases, and hexagonal with zig-zag edges in the case of gasificatjon with 197 hydrogen containing reactant gases. They conclude that hydrogen (:hemisorption is preferred on the zig-zag edges and therefore responsible f0r the anisotropy of reactivity toward the two principal edge configurations. The presence of hydrogen limits behavior of active sites to Inechanisms which preserve the zig-zag edge configuration and consume the armchair edge configuration. The most reasonable explanation for this is that under the presence of hydrogen. most gasification takes place via active site propagation along zig-zag edges. Zig-zag edges alone are produced in the presence of hydrogen because there is a much more ordered sequence of desorption of edge carbon atoms. If active sites were generated on the zig-zag edges. the edges would become pitted and rounded out. Most active sites are generated on armchair edge. amorphous. and dangling carbon atoms. which is why these features are consumed rapidly with conversion. A much smaller number of active sites are generated at defects in the zig-zag edges such as armchair-like arrangements, crystallite corners. crystallite defects, and unstable functional groups. These active sites propagate down the zig-zag edges and remove the outer row of carbons. preserving the edge intact, It is very unlikely that the char gasification mechanism identified as universal by Chen. Yang, Kapteijn, and Moulijn [55] and 198 Kapteijn, and Moulijn [56] plays a role in char gasification when gaseous hydrogen is present. The effects of reverse oxygen exchange. identified in this investigation as the major mode of hydrogen inhibition in steady-state steam gasification, should be even greater for the "off-plane" oxygen atoms because they only form single bonds to the adjacent carbon atoms. Also, the universal gasification mechanism (see Figure 3) is shown with all edge carbon atoms saturated with oxygen. but in steam gasification nearly all edge carbon atoms are saturated with hydrogen. 6.3. Recommendations Recommendations for further research in this investigation of the mode(s) of hydrogen inhibition in char gasification by steam include transient desorption of oxygen functional groups and molecular modeling of the proposed active site behavior in this investigation. Transient desorption of oxygen functional groups following gasification in steam gives the concentration of metastable groups which are most active in char gasification. The concentration of these metastable groups should be directly proportional to char reactivity. Relating the concentration of these groups and char reactivity to hydrogen partial pressure and adsorbed hydrogen concentration should lend a great deal of insight into the mode(s) of hydrogen inhibition. 199 Molecular modeling of active site propagation in steam gasification of chars is much more important than molecular modeling of active site formation. The majority of carbon atoms should desorb via active site propagation. as concluded in this investigation. Reaction schemes proposed by most other workers [21 45.55.56.91] are based on the removal of a carbon atom from a straight. defect-free zig-zag edge. This is a form of active site formation, which should also be able to take place readily on amorphous surface carbon atoms. defect sites, and carbon atoms that are adjacent to functional groups. A few other workers have proposed reaction schemes which include active site propagation [8.52]. but dissociatively adsorbed hydrogen on the zig-zag edges has been neglected. APPENDICES Appendix A: Linearized Rate Expressions with Explicit Adsorbed Hydrogen Concentration 201 202 Appendix A-l: Dissociative Hydrogen Adsorption Linearized Rate Expression with Explicit Adsorbed Hydrogen Concentration 1) Hzo+c-—5‘—>H.+C(O) 2) C(0)—k2—-00+c- 4) y-H-+c-—'S:_.C(H) -4) C(H)—k3—>V2H2+C.= Site Balance: [Ci] = [CF] + [0(0)] + [C(H)] Pseudo Steady State: mo. = 0 Derivation of Rate Expression: r‘co = k2[C(0)1 mo. = 0 = k1[H201[Cr1 - k2[C(0)] r‘co = k1[H20][CF] r‘co = k1[H20][Cr - C(O) - C(H)] r‘co = k1[H20][CT - (r‘co/kz) - C(H)] l"co{1 + (k1/k2)[H20]} = k.[c. - C(H)][HzO] tc. - C(H)]/rco = {1 + (k1/k2)[H20]}/k1[H20] [c- - C(H).] - [C- - C(H)]/rco..- - m. = {1 + (k1/k2)[H20]}/k1[H20] E00). = [LJL+1 rCOJ—rco k1 Pw k2 203 Appendix A-2: Dissociative Hydrogen Adsorption and Rapid Reverse Oxygen Exchange Linearized Rate Expression with Explicit Adsorbed Hydrogen Concentration 1) H20+Cr—kl—>H2+C(O) -1) H2+C(O)—!(—"—->HzO+C.= 2) C(0)—k2—>co+c- 4) mam-Low) -4) C(H)—£90040 Site Balance: [CT] = [Cr] + [C(O)] + [C(H)] Rapid Equilibrium: k. = ki/k.1 = [H2][C(O)]/[HzO][Cr] Derivation of Rate Expression: rco = k2[C(O)] [C(O)] = k1[HzO][Cr]/k-1[H21 r... = k2k1[H20][Cr]/k-1[H2] = k2k1[H20][Ci - C(O) - C(H)]/k-1[H2] m. = k2k1[HzO][Ci - (rm/k2) - C(H)]/k-1[H2] m. + rco{k1[HzO]/k-1[H2]} = 000010- - C(H)]/k-1[H2] r‘co{1 + k1[H20]/k-1[H2]} = k2k1[H20][Cr - C(H)]/k-1[H2] [(3. - C(H)]/rco = {1 + k1[H20]/k.1[H2]}k-1[H2]/k2k1[HzO] [c- - C(H).] - [0 - C(H)]/rco.. - r... = (k-1/k2k1)[H2]/[H20] + 1/0 w-[£)5+.2+i rCOJ—rco k1k2 F’w k2 Appendix 8: Mass Spectrometer Controller Settings RS-232 BAUD RATE STOP BITS DATA BITS PARITY PROTOCOL ECHO DELTA_D HOURS MINUTES DAY MONTH YEAR FIL PROT MULT VOLT DIST MODE CHANNEL TAB MASS TAB DNELL NO. SCANS TAB HI TAB L0 TAB CALIB AUTO ZERO COMPUTER 9600 7 1 NONE XON / XOFF OFF 0.50 1 44 3 7 94 1.0E-O3 -800 TAB 12 84 120 msec 0 1.0E+00 1.0E-15 1.0E+00 ON ELEC CUR EL ENERGY FOCUS ELEC MULT FIL RES FIL VOLTS FIL CUR FIL TOT SENS FREQUENCY RF TUNE EMISSION AMP CAL CAL MASS 1 L0 RES LO POS LO SENS CAL MASS 2 HI RES HI POS HI SENS QUAD HEAD TOT PRES 204 1.000E-03 -30.0 -20.0 OFF 0.73 2.2 3.1 ON 20.0 2.803E+06 6.984 1.0E-03 5.0E+03 1 3945 1.00 6.00 100 1000 -0.10 7.00 1 7.4E-06 Appendix C: Basic Programs for Deconvolution and Rate Calculation from Mass Spectrometer Data Program 1: Determination of Background Levels of Various Masses for Calibration Filename = back. has 10 DIM B(17,5).BA(17).F(30) 15 READ XXX,RAN$ 20 OPEN "I".#1.RAN$ 30 FOR N=1 TO XXX 40 INPUT#1.F(N) 50 NEXT N 60 FOR N=1 TO F(1) 70 INPUT#1,B(1.N).B(2.N).B(3.N).B(4,N).B(5.N).B(6.N).B(7.N). B(8,N).B(9.N).8(10,N).B(11,N).B(12.N).B(13.N).B(14.N). B(15,N) .B(16,N).B(17,N) 80 NEXT N 85 PRINT B(1,4) : IF B(1.4)=1234 THEN GOTO 90 86 LIST 200 90 CLOSE 100 FOR T=1 TO 17 110 8A(T)=(B(T.1)+8(T.2)+B(T,3)+B(T,4)+B(T.5))/5 :REM average 120 NEXT T 140 OPEN "O",#1."BACKAVEG.dat" 145 FOR T=1 TO 17 150 PRINT#1 .BA(T) 154 PRINT 8A(T) 155 NEXT 160 CLOSE 200 DATA 25 ."0006-1.tab" 205 206 Program 2: Hydrogen Calibration and Response Calculation Filename = cal_h bas 10 DIM H(17.5).HA#(17).8A#(17).P#(17).F(30) 15 READ XXX,RAW$.HPER 20 OPEN "I".#1.RAW$ 30 FOR N=1 TO XXX 40 INPUT#1.F(N) 50 NEXT N 60 FOR N=1 TO F(1) 70 INPUT#1.H(1,N).H(2.N).H(3.N).H(4,N).H(5,N).H(6.N).H(7.N). H(8.N).H(9.N),H(10,N).H(11,N).H(12,N).H(13,N).H(14,N), H(15.N).H(16.N).H(17,N) 80 NEXT N :CLOSE 95 PRINT H(1,4) :IF H(1,4)=1234 GOTO 100 97 LIST 300 100 OPEN "I",#1,"BACKAVEG.DAT" 110 FOR T=1 TO 17 120 INPUT#1.BA#(T) 125 NEXT :CLOSE 140 FOR T=6 TO 17 150 HA#(T)=(H(T.1)+H(T,2)+H(T,3)+H(T,4)+H(T.5))/5 :REM average 160 P#(T)=HA#(T)-BA#(T) :REM subtract background 170 NEXT T 180 REM deconvolution 210 HS#=HPER/(P#(6)-.034*P#(8)) :REM get the response 230 OPEN "O".#1."H_RESPON.DAT" 240 PRINT#1 .HS# :CLOSE 260 PRINT HS# 300 DATA 25."7011a.TAB" ,0 0126 207 Program 3: C0, C02, CH4 Calibration and Response Calculation Filename = cal_c_m.bas 10 DIM C(17,5).CA#(17).BA#(17).S#(17) 15 READ XXX,RAW$ 16 PRINT RAW$ 20 OPEN "I".#1,RAW$ 30 FOR N=1 TO XXX 40 INPUT#1.F :PRINT F:NEXT 60 FOR N=1 TO 5 7O INPUT#1.C(1.N).C(2.N).C(3.N).C(4.N).C(5.N).C(6.N).C(7,N) .C(8,N).C(9.N).C(10.N).C(11.N).C(12,N).C(13.N).C(14.N). C(15,N).C(16.N).C(17.N) 80 NEXT N :CLOSE 90 PRINT C(1.4) :IF C(1,4)=1234 GOTO 100 95 LIST 300 100 OPEN "I".#1."BACKAVEG.DAT" 110 FOR T=1 TO 17 120 INPUT#1,BA#(T) 130 NEXT :CLOSE 140 FOR T=6 TO 17 150 CA#(T)=(C(T,1)+C(T,2)+C(T,3)+C(T,4)+C(T.5))/5 :REM average 160 S#(T)=CA#(T)~BA#(T) :REM subtract background 170 NEXT T :REM deconvolution and get the response: 180 RCO#=.02/(S#(14)-.105*S#(16)) :REM get the response CO 190 RC02#=.0203/S#(16) :REM get the response C02 200 RCH4#=.02/S#(9) :REM get the response CH4 210 OPEN "O".#1."C_respon.dat" 220 PRINT#1 ,RCO#.RCO2#.RCH4# 230 CLOSE 235 PRINT " C0 C02 CH4" 240 PRINT RCO#.RC02#.RCH4# 300 DATA 25."7020_d tab" 208 Program 4: Organization of Mass Spectrometer Output Data into Matrix Format Filename = backt.bas 10 DIM V(17.200).S#(17.200).TIME(200).F(30).CO(200).C02(200) .CH4(200).H2(100) 20 READ XXX,RANDAT$.SAVEDAT$ ,K 30 OPEN "I".#1,RANDAT$ 40 IF EOF(1) THEN END 50 FOR N=1 TO XXX 60 INPUT#1.F(N) 70 PRINT F(N) 80 NEXT N 85 REM K=F(1) 90 FOR N=1 TO K 100 INPUT#1,V(1,N).V(2,N).V(3,N).V(4,N).V(5.N),V(6.N).V(7,N) .V(8,N).V(9.N).V(10.N).V(11.N).V(12.N),V(13.N).V(14,N).V(15,N).V(16.N),V (17.N) 102 PRINT N 110 NEXT N 120 CLOSE 130 PRINT V(1 3) ."To see if this value is 1234" 145 TI=60*V(2,1)+V(3.1)+V(4,1)/6O 160 FOR N=1 TO K 170 TIME(N)=V(2.N)*60+V(3 N)+V(4,N)/60 180 NEXT N 400 OPEN "O".#1.SAVEDAT$ 410 PRINT#1,"Time M2 M3 M4 M15 M16 M17 M18 M20 M28 M32 M44 M84 415 PRINT "Time M2 M3 " 420 FOR N=1 TO K 430 PRINT#I,INT(100*TIME(N))/100.V(6.N).V(7.N),V(8,N).V(9.N) -V(10.N).V(11.N).V(12.N).V(13,N).V(14,N),V(15,N),V(16.N). V(17,N) 440 PRINT INT(10*TIME(N))/10.V(6,N).V(7,N).N 450 NEXT 460 CLOSE 480 DATA 25."0003-1.tab",0003-1r.DAT" .93 209 Program 5: Deconvolution and Calculation of Evolution Rates of Various Species Filename = main-dem has 10 DIM V(13.300).S#(13.300).TIME(300).A0(13).CO(300). C02(300).CH4(300).H2(300) 20 3O 40 50 60 70 80 90 READ RANDAT$,SAVEDAT$.GASFL,SAMW,RH2 OPEN "I",#1,RAWDAT$ IF EOF(1) THEN END INPUT#1.A,8 FOR L=1 TO B INPUT#1.Q$ PRINT Q$ NEXT L 100 A=A-1 110 120 130 140 150 160 170 180 190 PRINT A,B FOR N=1 TO FOR J=1 TO INPUT#1.V(J,N) NEXT J PRINT N,V(1,N).V(2,N) NEXT N REM ********************************** FOR J=1 TO B A B 200 AO(J)=V(J.1) 210 220 230 240 250 260 270 280 290 300 310 320 325 330 340 FOR N=1 TO A IF V(J,N) < A0(J) THEN A0(J)=V(J.N) REM PRINT AO(J).J NEXT N NEXT J REM ********************************** CLOSE FOR J=1 TO 8 FOR N=1 TO A V(J.N)=V(J,N NEXT N NEXT J PRINT "below is the found background" FOR T=2 TO B PRINT A0(T) )-A0(J) 210 350 NEXT 360 INPUT "The backgroud is OK?",K$ 37 0 REM **~k*~k****~k**********~k*~k~k*~k~k**~k*~k**~k 380 OPEN "I".#1,"C_RESPON.DAT" 390 INPUT#1,RCO.RC02.RCH4 400 CLOSE 410 X=GASFL/SAMN 420 FOR N=1 TO A 430 H2(N)=(V(2,N)-.034*V(3.N))*RH2*X 440 CO(N)=(V(10.N)-.105*V(12,N))*RCO*X 450 C02(N)=V(12 N)*RC02*X 460 CH4(N)=V(5 N)*RCH4*X 470 NEXT 480 OPEN "0".#1.SAVEDAT$ 490 PRINT#1,"Time(min) H2 CH4 CO COZ/ml/min" 500 FOR N=1 TO A 510 PRINT#l ,V(1.N).H2(N).CH4(N).CO(N).C02(N) 520 PRINT V(1.N).C02(N).N 530 NEXT 540 CLOSE 550 DATA "7020raw.dat",7020dec.DAT",300,0.3735,300000 211 Program 6: Calculation of Evolution Rates Based on Carbon Conversion Filename = main-ins.bas 10 DIM TIME(200).CO(200).CO2(200).CH4(200).H2(200). X(200).Y(200) 20 DIM COX(200).C02X(200).CH4X(200),W(200).SUM1(200). SUM2(200).SUM3(200) 30 DIM WSUM(200).INSCO(200).INSCH4(200).INSC02(200) 40 READ RAWDATS,SAVEDAT$.GASFL,W0 50 OPEN "I".#1.RAWDAT$ 6O REM IF EOF(1) THEN END 70 INPUT#1 .A,B 80 INPUT#1 .Ql$.Q2$.Q3$.Q4$.05$ 81 PRINT 01$.02$.03$.Q4$.05$ 82 A=A~1 90 FOR N=1 TO A 100 INPUT#1 .TIME(N).H2(N).CH4(N).CO(N).C02(N) 110 PRINT TIME(N).CO(N).N 120 NEXT N :CLOSE 125 PRINT 130 INPUT "The Data Read[time(min) and CH4 C0 C02 cc/min g(initial)] is OK";T$ 140 FOR N=1 TO A 150 COX(N)=CO(N)*NO :CO2X(N)=C02(N)*NO :CH4X(N)=CH4(N)*NO 160 NEXT 165 SUM1(1)=0:SUM2(1)=0:SUM3(1)=0 170 FOR J=2 TO A 200 SUMl(J)=SUM1(J-1)+((CH4X(J)+CH4X(J-1))/2)*(TIME(J)-TIME(J-1)) 210 SUM2(J)=SUM2(J-1)+((COX(J)+COX(J-1))/2)*(TIME(J)-TIME(J-1)) 220 SUM3(J)=SUM3(J-1)+((C02X(J)+C02X(J-1))/2)*(TIME(J)-TIME(J-1)) 230 NSUM(J)=(SUM1(J)+SUM2(J)+SUM3(J))/(22.41*1000)*12.011 240 REM wsum(J) is total Carbon loss (gram) 250 PRINT USING "###.#### ":SUM1(J).SUM2(J).SUM3(J).NO—NSUM(J).J 260 NEXT J 263 PRINT "CH4 (ML) CO (ML) C02 (ML) left sample Data Point " 265 PRINT " " 267 PRINT 100*SUM1(A)/(SUM1(A)+SUM2(A)+SUM3(A)). " CH4 Carbon lose percent%" 212 268 PRINT 100*SUM2(A)/(SUM1(A)+SUM2(A)+SUM3(A)). " CO Carbon lose percent%" 269 PRINT 100*SUM3(A)/(SUM1(A)+SUM2(A)+SUM3(A)). " CO2 Carbon lose percent%" 270 INPUT "The Gas volume(ML) of CH4 C0 C02 is correct ?",T$ 280 FOR N=1 TO A 291 N(N)=NO-NSUM(N) 310 INSCO(N)=COX(N)/N(N) 320 INSC02(N)=C02X(N)/N(N) 330 INSCH4(N)=CH4X(N)/N(N) 340 NEXT N 350 OPEN "O".#1.SAVEDAT$ 360 PRINT#1,"Time(min) CH4(cc/min) CO(cc/min) C02(cc/min) conversion(%)" 370 FOR N=1 TO A 380 PRINT#l .TIME(N).INSCH4(N).INSCO(N).INSC02(N).100*NSUM(N)/w0 390 X(N)=INT(10000*(N0-W(N)))/10000: Y(N)=INT(1000000!*(N0- N(N))/N0)/10000 400 PRINT "N_Loss":X(N):"g ";Y(N) "%";TAB(35);N:"TH DATA POINT" 410 NEXT 420 CLOSE :P1=INT(10000*NSUM(A))/10000 :P2=INT(10000*(w0-NA))/10000 440 PRINT "Among the lose.CH4% C0% C02%" 441 Zl=SUM1(A)/(SUM1(A)+SUM2(A)+SUM3(A)) 442 22=SUM2(A)/(SUM1(A)+SUM2(A)+SUM3(A)) 443 Z3=SUM3(A)/(SUM1(A)+SUM2(A)+SUM3(A)) 445 PRINT Z1*Y(A).Z2*Y(A).Z3*Y(A) 450 PRINT "Calculated N-Loss is:";P1;"g ":100*NSUM(A)/NO:"% of sample weight" 510 DATA "7020de.dat"."7020-inS.DAT",180 ,0 3735 Appendix D: Annealed Saran Char Bulk Modulus 0.3 = L./rC/CDe (eq.4.74. Lee) 2.40 10“5 0G = (4.17x10”3) 4x = 0.00719 (1.33x10‘ )(O.00607) L=Dp/6 (Table 4.1. Lee) L = (250pm/6)(cm/10‘ .m) = 4.17x10‘3 cm [0 :(O.Qchonv.)( molC ){(0.390)(1000mm3/cm3)}( min ) 10090 *min 12.01190 “[(4mm)2(31mm)] 60 sec rC = 2.40x10”‘3 mOl/cm3*sec moI*K 8.314cm3 *MPa c = (O.4)(3.’MPa)[ )/1123K =1.33x10" mol /cm3 E1;=1+(r3:2—1)X+D:,k (eq.14.2, Lee) i— 1+((18/40)1/2 _1)(0.4)+ 1 _ S/Cm2 De — 2220 0.00607 — 0.00607 0,.12 =Dm.12fe(e/k) (eq.14.3, Lee) 09.12 = (2804X0.792) = 2220 cm2/ S 0... =Dkf,(elk) ‘ (eq.14.4, Lee) De,k = (0.00766X0792) = 0.00607 cmz/ s 213 214 1/2 1 1 Dm=0.0018583{[T3(M_+M—2):l fiPoffiLJ} (eq.14.5. Lee) 1 r 3(—1_ 1.)]1/2 ‘ 0m12 =0.0013533< [1123 ”+40 . f=2804 cmZ/s ' [(31.6)(3.36)2(0.745)] 012 =(o1+02)/2 (p.484, Lee) 012 = (3.30 + 3.418)/2 = 3.36 A Q12 =f(e12) (Table 142. Lee) (212 = 0.745 812 = 8182 (p.484, LEG) 812 = J(124k)(110k) = 117k fe(elk)=22 (eq.14.12, Lee) 16(3 / k) = 0.392 = 0.792 t)k =9.7x103F T/M (eq.14.7, Lee) Dk = 9.7x103(1x10-7),/1 123/18 = 0.00766 cmz/s 215 Appendix D Nomenclature: (D0 = generalized Thiele modulus for first order reaction (unitless) L = characteristic length of particles (cm) rc = intrinsic reaction rate (mol/nfi*sec) C = concentration of key species (mol/cm3) [k == effective binary diffusivity of L00 in Ar (an/s) Up = diameter of particle (cm) De12== effective molecular diffusivity of L00 in Ar (emf/s) Dek = Knudsen diffusivity of Hi)(cmf/s) ratio of molecular weights of L00 to Ar (unitless) mole fraction of'LbO (unitless) correction factor for porous medium (unitless) porosity of medium (unitless) tortuosity of medium (unitless) temperature (K) molecular weight (g/mol) pressure (atm) = Lennard-Jones parameter (A) = Lennard-Jones parameter (unitless) Lennard-Jones parameter (unitless) average pore radius (cm) -h X 3 (D II II n ll II so “OZ—17rd) u "5 ('0 5 II II Appendix E: Linearized Rate Expression Derivations for Various Rate Expressions 216 217 Appendix E-l: Linearized Rate Expression Derivation for Reverse Oxygen Exchange 1) Hzo+cF——"—‘—>HZ+C(0) -1) H2+C(0)i—->Hzo+cp 2) C(0)L>00+cs Site Balance: [C1] = [CF] + [C(O)] Pseudo Steady State: mo) = 0 Derivation of Rate Expression: rtan = 0 = kd}h0][CF] - k4[H2][C(O)] - k2[C(0)] 0 = k1[H20]{[Cr] - [C(O)]} - k-1LH2][C(O)] - k2[C(O)] 100120in = [0(0)]{kLLH201 + k-1[H2] + k2} [0(0)] = klszoucTJ/{klmzoi + k-1[H2] + k2} k2[C(0)] FCO k2k1[H20][C1]/{k2 + 100120] + k-1[H2]} FCO rm ___ k1cTPw 1+ (k1 /k2)Pw + (k.1 /k2)1=1.2 i=0;]+[k101][%]+[k101][%][%) 218 Appendix E-Z: Linearized Rate Expression Derivation for Rapid Equilibrium Reverse Oxygen Exchange 1) Hzo+cpi+Hz+cHzo+cF 2) C(O)—k—2>CO+CF Site Balance: [C1] = [Cr] + [C(O)] Rapid Equilibrium: K1 = k1/k-1 = [H2][C(O)]/[H20][Cr] Derivation of Rate Expression: [CF] k-1[H2][C(0)]/k1[H20] [CT] {(k-i/ki)[H2][C(0)]/[H20]} + [0(0)] [0(0)] = [oi/{1 + (k.1/k1)[H2]/[H20]} k2[C(0)] r‘CO k2[C1]/{1 + (k-1/k1)[H2]/[H20]} r‘CO r _ IqC¢R~ CO (k1/k2)Pw+(k_1/k2)PH2 ii ‘ H 1 Hit) R30 sz1 k1CT k2 F>W 219 Appendix E-3: Linearized Rate Expression Derivation for “Associative” Hydrogen Adsorption 1) L00+0F——k—‘—>Hz+0(O) 2) 0(0)—&—>co+ce 3) rewriscm): -3) C(H)2L>Hz+CF Site Balance: [C1] = [Cr] + [C(O)] + [C(H)2] Pseudo Steady State: rum = 0 , foo-()2 = 0 Derivation of Rate Expression: f0(H)2 = O = k3[H2][Cr] - k-3[C(H)2] [C(H)2] = (k3/k-3)[H2][CF] mm = 0 = k1[H20][C[] - k2[C(0)] k2[C(O)] k1[H20][Cr] [0] = (k2/k1)[C(0)]/[H20] [C(H)2] = (k3/k-3)(kz/k1)[H2][C(0)]/[H20] k2[C(0)] = k1[H20][CF] k2[C(O)] = 1001201001] — [0(0)] - [0002]} k2[C(O)] = k1[H20]{[Ci] - [C(O)] - (k3/k-3)(kz/k1)[H2][C(0)]/[H20]} k2[C(0)] = k1[H20][Ci] - k1[H20][C(0)] - (k3/k-3)k2[H2][C(0)] 220 “Associative” Hydrogen Adsorption Continued... k1[H20][C1] = [0(0)]{k2 + 10000] + (ks/k—3)k2[H2]} [0(0)] = k1[H20][C1]/{k2 + 10000] + (ks/k.3)k2[H2]} k2[C(0)] r‘CO k2k1[H20][C1]/{k2 + 10000] + (k3/k-3)k2[H2]} T‘CO _ kfith -1+(k1/k2)Pw + (k3 /k_3)PH2 -‘--L 1 H ‘)(-1-)+L ‘ HP”) rco kzc'r k1CT Pw k1CT k—a F’w I’co 221 Appendix E-4: Linearized Rate Expression Derivation for Reverse Oxygen Exchange and “Associative” Hydrogen Adsorption 1) H20+Cr—L>Hz+C(O) -1) H2+0(0)—‘L‘—>H20+0F 2) 0(0)—'—‘2—>0o-cs 3) Hz+C1=—k—3—>C(H)2 -3) C(H)2——k'3——>Hz+CF Site Balance: [C1] = [Cr] + [C(O)] + [C(H)2] Pseudo Steady State: (“0(0) = 0 , rem): = O Derivation of Rate Expression: fixmz = O = kd}b][Cr] - ka[C(H)2] [C(H)2] = (ks/k0)[tb][CF] rtn» = O = kH}bO][Cr] - k1[ib][C(O)] — k2[C(O)] kdlhOJLCF] = k1LLh][C(O)] + k2[C(O)] [0.] = [C(O)]{k.1[H2] + k2}/k1[HzO] [C(H)2] = (ks/k-3)[H2][C(O)]{k-1[H2] + k2}/k1[HzO] k1[H20][CF] = k-1[H2][C(0)] + k2[C(O)] kaH20]{[cTJ - [0(0)] - [0002]} = k-ILHzlLC(O)] + k2[C(O)] 222 Reverse Oxygen Exchange and “Associative” Hydrogen Adsorption Continued... k1[HzO]{[C1] - [0(0)] - (ks/k-3)[H2][C(O)]{k-1[H2] + k2}/k1[HzO]} = k-1[H2][C(0)] + k2[C(O)] k1[H20][Ci] = [C(0)]{k1[H20] + (k3/k-3)[H2]{k-1[H2] + k2} + k-1[H2] + k2} [0(0)] = k1[HzO][C1]/{k1[H20] + (ka/k-3)[H2]{k-1[H2] + k2} + MHZ] + k2} r00 = k2[C(O)] k2k1[H20][C1]/{k1[HzO] + (ka/k-3)[H2]{k-1[H2] + k2} + k-1[H2] + k2} r‘CO = r00 = k2k1[H20][Ci]/{k1[H20] + (k3/|<-3)k-1[H2]2 + (k3/k-3)k2[H2] + k-1[H2] + k2} _ k1CTPW rco — 1+ (k1 ”(2)13W + {(k.1 /k2)+ (k3 1k.3 )}P(.2 + {(k.1 Ik2Xk3 /1<_3)}P§2 E1? = [022. i + [100. iii-13] + ik10Tiii£in+ [If—:iiigivii 1.2.11:—:1:+:1%::)J 223 Appendix E-S: Linearized Rate Expression Derivation for Dissociative Hydrogen Adsorption 1) H20+C1=—k—‘+H2+C(O) 2) 0(0)—ki—>00+0. 4) %H2+CF——-)k‘ C(H) -4) C(H)—k—‘——>}él-i2+CF Site Balance: [C1] = [C1] + [C(O)] + [C(H)] Pseudo Steady State: mo) = O . rem) = 0 Derivation of Rate Expression: rem, = 0 = k4[H2]2[C1] - k.4[C(H)] [C(H)] = (k4/k-4)[H2]’[CF] F010) = 0 = k1[H20][CF] - k2[C(O)] k2[C(O)] = k1[H20][CF] [01] = (k2/k1)[C(O)]/[HzOJ [C(H)] = (ka/k-4)(k2/k1)[H2]2[C(O)]/[H20] k2[C(O)] k1[H20][CF] k2[C(O)] k1[HzO]{[C1] - [0(0)] - [000]} k2[C(O)] = k1[HzO]{[C1] - [0(0)] - (k4/k.4)(k2/k1)[H2]2[C(O)]/[HzO]} k2[C(O)] = k1[H20][C1] - k1[H20][C(0)] - (ka/k-4)k2[H2]2[C(0)] 224 Dissociative Hydrogen Adsorption Continued... k1[HzO][C1] = [0(0)]{k2 + (00120] + (k4/k-4)k2[H2]2} [0(0)] = k1[HzO][C1]/{k2 + (00120] + (k4/k-4)k2[H2]2} k2[C(O)] r00 k2k1[HzO][C1]/{k2 + k1[HzO] + (ka/k-4)k2[H2]2} Y‘CO = k1CTPw 1+(k1/k2)Pw+(k4/k-4) ‘4} ii 1 H 1 iii—H; Pig rco R201 k1CT Pw k1C'r k-4 Pw rco 225 Appendix E-6: Linearized Rate Expression Derivation for Reverse Oxygen Exchange and Dissociative Hydrogen Adsorption 1) H2O+0.—‘L>Hz+0(0) -1) H2+C(O)-—k“——)H20+C1= 2) C(O)—k2—>CO+CF 4) ysz+c.—‘5‘—>0(H) -4) C(H)—k'—‘—>}5H2+Cs Site Balance: [C1] = [C1] + [C(O)] + [C(H)] Pseudo Steady State: ram = O . [mm = O Derivation of Rate Expression: 1c...) = 0 = 100121001] - k-4[C(H)] [C(H)] = (kn/k4)[Lb]2[C1] r‘0(0) = 0 = k1[HzO][Cr] - k-1[H2][C(O)] - k2[C(O)] 10111201101] = k-1[H2][C(O)] + k2[C(O)] [01] = [C(O)]{k-1[H2] + k2}/k1[HzO] [C(H)] = (ka/k.4)[H2]2[C(O)]{k-1[H2] + k2}/k1[HzO] k1LH201101] = k-1[H2][C(O)] + k2[C(O)] k1[H20]{[C1] - [0(0)] - [000]} = k-1[H2][C(O)] + k2[C(O)] 226 Reverse Oxygen Exchange and Dissociative Hydrogen Adsorption Continued... k1[HzO]{[C1] - [0(0)] - (k4/k-4)[H2]2[C(O)]{k.1[H2] + k2}/k1[HzO]} = k-1[H2][C(0)] + k2[C(O)] k1[H20][C1] = [C(O)]{k1[HzO] + (ka/k-4)[H2]2{k-1[H2] + k2} + k-1[H2] + k2} [0(0)] = k1[H2O][C1]/{k1[HzO] + (ka/k-4)[H2]2{k-1[H2] + k2} + k-1[H2] + k2} k2[C(O)] rCO F00 k2k1[H20][Ci]/{k1[H20] + (k4/k-4)[H2]2{k-1[H2] + k2} + k-1[H2] + k2} F‘CO k2k1[HzO][C1]/{k1[HzO] + (ka/k.4)k-1[H2]3’2 + (k4/k.4)k2[H2]2 + k-1[H2]+k2} _ k1CTPw 1 + (k1/k2)PW + (k-1/k2)PH2 + (k4 /k_4){P}3|/22 + (k_1/k2)P|3$2} I'co kch k1CT F’w k1CT k2 Pw k1CT k-4 w k1C1' k2 k_4 Pw rco 227 Appendix E-7: Methane Formation Only “Associative” Hydrogen Adsorption 3) Hz+C1=—k—3—>C(H)2 -3) C(H)2—kL>Hz+CF 5) H2+0(H)2—£—>0H.+c. Site Balance: [C1] = [C1] + [C(H)2] Pseudo Steady State: rcr = O . [C(H)2 = O Derivation of Rate Expression: res = O = k3LH2][CF] - k-3[C(H)2] + kSLH2][C(H)2] k3[H2][CF] = k-3[C(H)2] + ks[H2][C(H)2] k3[H2]{[C1] - [C(H)2J} = k-3[C(H)2] + k5[H2][C(H)2] k3[H2][C1] = [C(H)2]{k-a + [(5012] + k3[H2]} ks[C(H)2][H2] I'CH4 rCH4 ks[H2]k3[H2][C1]/{k-a + k5[H2] + k3[H2]} rCH4 ks[H2]2[C1]/{(k.3/k3) + (1 + ks/k3)[H2]} ksCTPfiz (k_3 lk3)+ (1 + k5 ”(3)13H2 rem = 228 Appendix E-B: Methane Formation Only Dissociative Hydrogen Adsorption 4) tz+c.—i‘—+0(H) -4) C(H)—L‘Axwarcp 6) tz+ C(H)—'56—)C(H)2 -6) C(H)2—i(f—+XH2+C(H) 7) %H2+C(H)2——k7—>C(H)a -7) C(H)a—k—"—+XH2+C(H)2 8) xH2+C(H)a—L—>CH4+CF Site Balance: [C1] = [C1] + [C(H)] + [C(H)2] + [C(H)3] PSGUdO Steady State: for = r001) = rC(H)2 = [00113 = 0 keCTPfiz f1(k)+ 120013.19,2 + [3003.2 + 1. (@3352 r0114 = 10. 11. 12. 13. 14. 15. REFERENCES Redmond, 3.P. and Walker, P.L. 3r.. J. Phys. Chem. 64, 1093 (1960). Bansal, R C.. Vastola. F.3.. and Walker, P.L. 3r., Carbon 9, 185 (1971). Bansal, R.C.. Vastola. F.3.. and Walker, P.L. 3r., Carbon 12, 355 (1974). Tobias, H. and Soffer, A.. Carbon 23, 291 (1985). Biederman. D.L. et al.. Carbon 14, 351 (1976). Yang. R.T. and Yang, K.L., Carbon 23(5). 537 (1985). Matsumoto. S.. Sato, Y.. and Setaka. N.. Carbon 19. 232 (1981). Huttinger, K.3. and Merdes, W.F.. Carbon 30(6). 883 (1992). Mims. C.A. and Krajewski, 3.3.. J. Catal. 102, 140 (1986). Steinberg, M., 1987 Int. Conf. on Coal Sci. 11, 3.A. Moulijn, ed.. 953 (1987). Steinberg, M.. 1987 Int. Conf. on Coal Sci. 11, 3.A. Moulijn. ed.. 1064 (1987). Blackwood, 3.0., Aust. J. Applied $01., 199 (1962). Blackwood. 3.0. and McCarthy, 0.3., Aust. J. Chem. 19, 797 (1966). Blackwood, 3.0. Aust. J. Chem. 15, 397 (1962). Blackwood. 3.0. Aust. J. Chem. 12, 14 (1959). 229 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 230 Makino. M. and Toda, Y.. Fuel 60. 321 (1981). Blackwood, 3.0., McCarthy, 0.3.. and Cullis. B.D., Aust. J. Chem. 20, 2525 (1967). Blackwood, 3.0.. Cullis. 8.0.. and McCarthy, 0.3., Aust. J. Chem. 20, 1561 (1967). Zielke, C.H. and Gorin, E., Ind. and Eng. Chem. 47(4). 820 (1955). Cao, 3.R. and Back, M.H.. Carbon 30(6). 505 (1981). Pan. Z.3. and Yang, R.T.. J. Catal. 123, 206 (1990). Shaw, 3.1., Proc. Int. Conf. Coal 501., Dusseldorf, Germany, 209 (1981). Cao, 3.R. and Back. M.H.. Carbon 20, 505 (1982). 3untgen, H.. Carbon 19. 167 (1981). Lizzio. A.A., 3iang, H.. and Radovic, L.R., Carbon 28, 7 (1990). Huttinger, K.3. and Nill, 3.3., Carbon 28, 457 (1990). Muhlen, H.3., Van Heek, H.. and Juntgen H.. Fae] 64. 944 (1985). Gardner, N.. Samuels, E., and Wilkes. K.. ACS Div. In Chem. 131, 209 (1974). Cypres, R.. Ghodsi, M., and Feron, D., Thenmochemica Acta 81, 105 (1984). Blackwood. 3.0. and McCarthy, 0.3., Aust. J. Chem. 20. 2003 (1967). Carangelo. R.M., Solomon, P R.. and Gerson, 0.3., Fae] 66, 960 (1987). Hauge, R.H., et al., AC8 Div. Fae] Chem. 28(1). 35 (1983). Walker, P.L. 3r., Carbon 28, 261 (1990). 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 231 Phillips. R.F., Vastola. F.3., and Walker. P.L. 3r., Carbon 8, 197 (1970). Hart, P 3.. Vastola. F.3., and Walker, P.L. 3r., Carbon 5, 363 (1967). Treptau, M.H. and Miller, 0.3.. Carbon 29, 531 (1991). Toomajian, M., Lussier. M., and Miller, 0.3.. AC5 Div. FUel Chem. Prepr. 36(1). 139 (1991). Zoheidi, H. and Miller. 0.3., Carbon 25. 809 (1987). F] :! Saber. M.3., Falconer. 3.L.. and Brown, L.F., F097 65. 1356 (1986). 3 Hashimoto, K.. et al., Chem. Lett. 5. 555 (1984). Cao, 3.R.. and Back M.H.. Carbon 23, 141 (1985). H Wood. 8.3. and Sancier, K M., Cata. Rev. Sci. Eng. 26, 233 (1984). Yang, R.T. and Duan, R.Z., Carbon 23(3). 325 (1985). McKee. 0.w. in 'Chemistry and Physics of Carbon', Walker. P.L. Jr. and Thrower. P A.. ed. Marcel Dekker, New York, 61(1) (1981). Hermann. G. and Huttinger, K 3.. Carbon 24(6). 705 (1986). Huttinger, K 3.. ACS Div. Fuel Chem. Prepr. 34(1). 56 (1989). Huttinger, K.3.. Carbon 26(1). 79 (1988). Giberson, R.C. and Walker. 3.P.. Carbon 3, 521 (1966). Montet. G.L., and Myers. G.E., Carbon 9. 673 (1971). Gadsby, 3., Hinshelwood. C N.. and Sykes. K.w.. Proc. Royal Soc. London A187, 129 (1946). Long, F.3., and Sykes. K.M.. Proc. Royal Soc. London A193, 377 (1948). 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 232 Blackwood. 0.3. and McGrory. F.. Aust. J. Chem. 11, 16 (1958). Lewis. N.K.. Gilliland, E R.. and Hipkin, H.. Ind. Enq. Chem. 45, 1698 (1953). 30hnst0ne. H.F.. Chen. C.Y.. and Scott. D S., Ind. Enq. Chem. 44, 1564 (1952). Chen. S.G. et al.. Ind. Eng. Chem. Res. 32. 2835 (1993). Moulijn. 3.A. and Kapteijn, F.. Carbon. 33, 8 (1995). Goya], A.. Zabransky. R.F., and Rehmat, A.. Ind. Enq. Chem. Res. 28. 1767 (1989). Ozaki. A.. Isotopic Studies of Heterogeneous Catalysis, New York. Academic Press. Inc. (1977). Melander, L. and Sanders. w.H.. Reaction Rates of Isotopic Molecules, New York. 3ohn Wiley and Sons (1980). Yates. 3.T.3r. and McKee, 0.w.. J. Chem. Phys. 75, 2211. (1981). Mims. C.A. and Pabst. 3 K.. ACS Div. FUel Chem. Prepr. 25(3). 263 (1980). Gulbransen. E.A.. Nature 212. 1420 (1966). Ishikawa. Y. et al., Chem. and Phys. of Carbon 12, Walker, P.L.3r. and Thrower, P.A., ed.. 40 (1975). Kapteijn, F.. Meijer, R.. and Moulijn, 3.A., ACS Div. FUel Chem. Prepr. 36(3). 906 (1991). Biloen. P.. J..Mblec. Catal. 21, 17 (1983). Mims. C.A. et. al., Surf. Sci. 0f Catal. 15, 230 (1992). Laidler, K.3., Chemical Kinetics. New York. Harper Collins Inc.. (1987). 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. £30. E31. £32. £33. 233 Steinfeld. 3.1.. Francisco. 3.8.. and Hase, M.L.. New Jersey. Prentice Hall. (1989). Lide. D.R., CRC Handbook of Chemisthy and Physics 78, New York. CRC Press. Inc.. (1997). Huber. K.P. and Herzberg. G.. Mblecular Spectra andflMblecular Structure IV, Constants of Diatomic Mblecules. New York. Van Nostrand Reinhold. (1979). Zoheidi, H.. Ph D. Dissertation, Michigan State University. (1987). Treptau. M.H.. Ph D. Dissertation. Michigan State University. (1989). Toomajian. M.E . Lussier. M 0., and Miller, 0.3.. FUel 71, 1055 (1992). Lussier. M L.. MKS. Theses. Michigan State University, (1992). Commercial Testing and Engineering C0.. Lombard, IL (1990). Galbraith Laboratories. Inc . Knoxville, TN (1995). Saber, 3.M , Falconer, 3.L.. and Brown. L.F., J. Catal. 90, 65 (1984). Ismail, I.M.K. and Walker. P.L.3r., Carbon 27(4). 549 (1989). Zhang. Z.. MKS. Theses. Michigan State University. (1997). Einar. S., Abrahamsson. S.. and McLafferty, F.w., Registhy of Mass Spectral Data 1, (1974). Warren. 8 E., X-ray Diffraction. Addison—Wesley Publishing C0.. Reading, MA (1969). McLachlan, D.3r.. X-ray Chystal Structure. McGraw-Hill Book Company. Inc.. New York (1957). Lee. H.H. Heterogeneous Reactor Design, Butterworth (1985). 84. 85. 86. 87. 88. 89. 90. 91. 234 Toomajian. M.E.. 1151 Thesis. Michigan State University. (1991). Needa. M., Abcouwer, H H.. Kapteijn, M. and Moulijn, 3.A.. Fuel Proc. Technol.. 36, 235 (1993). Needa. M., Ph.D. Dissertation, University of Amsterdam. (1995). Lide. D.R.. CRC Handbook of Chem. And Phys.. 75 (1995). Takahashi et.al., The Surface Energies of Graphite, ref.#3, Abrahmson. 3.. 338. Ubbelhode. A.R. and Lewis, E.A.. Graphite and Its Chystal Compounds. Oxford University Press, London (1960). Riley. H.L.. Chemisthy and Industhy. 58. 394 (1939). Skokova. K. and Radovic, L.R., ACS Div. FUel Chem. Prepr., 41(1). 143 (1996). "‘1111111111111“ Q-II...-_...-I-I-I-I-I-I-I-I'