 

 

‘4
.- V-_._~...qu

THEmS
2

(ma

IlllllllllllllllllllllllllllllllllllllllI

31293 01712 6560

This is to certify that the

thesis entitled

Reactive Properties of Hydrogen Adsorbed
on Carbon Surfaces

presented by

Zhigang Zhang

has been accepted towards fulfillment
of the requirements for

M. S . degree in Chemical Engineering

 

 

\

Major professor

0-7639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

 

 

LIBRARY

Michigan State

University

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINE-S return on or before date due.

 

 

 

MTE DUE

DATE DUE

DATE DUE

 

HEB"! g. 7091

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1]” WWW.“

 

Reactive Properties of Hydrogen Adsorbed
on Carbon Surfaces

By

Zhigang Zhang

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements for the degree of

MASTER OF SCIENCE
Department of Chemical Engineering

1997

Reactive Properties of Hydrogen Adsorbed on Carbon Surfaces

Abstract

During carbon gasiﬁcation (Hz/steam/Argon), hydrogen quickly adsorbs on and
covers most of the surface active sites of carbon. It is well known that hydrogen seriously
inhibits further steam gasiﬁcation. In this thesis, adsorbed hydrogen is investigated
extensively, including its desorption properties (temperature program desorption proﬁle),
stability, and reactivity.

The adsorbed hydrogen is quite stable on the char surface. To desorb all
hydrogen, it is necessary to heat carbon to 1500°C. Molecular oxidation has no
preference to remove hydrogen. From D-H exchange, we have conﬁrmed that adsorbed
hydrogen is in a dynamic state during gasiﬁcation and continuously exchanges with
hydrogen in the gas phase.

From a kinetic model of the steam/H2 gasiﬁcation data, we conclude that
hydrogen inhibits steam gasiﬁcation of carbons by reverse oxygen exchange that depends
on hydrogen partial pressure.

Gasiﬁcation rate can be enhanced by molecular oxidation, but mechanism is very
complicated. Removal of adsorbed hydrogen, surface area change, and change in the
carbon surface structure all contribute to the rate enhancement; further, the gasiﬁcation

temperature and reactant composition both affect the mechanism of rate enhancement.

ACKNOWLEDGMENTS

The author wishes to thank Dr. Dennis J. Miller for his guidance, understanding
and ﬁnancial support throughout this project.

I would like to thank Michael Lussier for his cooperation and help during this
project.

I would like to thank professor Pinnavaia and professor Drazl for their severing
my committee members.

Finally the author owes a great debt to his wife for her unselﬁsh patience and

understanding during his studies and the preparation of this thesis.

TABLE OF CONTENTS

 

 

 

 

LIST OF TABLES ............................................................................................................................. vi
LIST OF FIGURES ......................................................................................................................... vii
CHAPTER 1. INTRODUCTION .............................................................................................. 1
1-1. Background ................................................................................................................................ 2

1-2. Literature Review ...................................................................................................................... 3

1-2-1. Hydrogen Adsorbed on the Carbon Surfaces ................................................................... 3

1-2-2. Oxygen Complexes on the Carbon Surfaces ................................................................... 6

1-2-3. Methane Formation in Gasiﬁcation ................................................................................. 7

1-2-4. CO Formation in Steam Gasiﬁcation ............................................................................. 10

1-2-5. Hydrogen Inhibition ...................................................................................................... 11

1-2-6. Isotopic Effect ............................................................................................................... 12

1-3. Research Objectives ................................................................................................................. 13

1-3-1. Characterization of Adsorbed Hydrogen ....................................................................... 13

1-3-2. Isotope Effects ............................................................................................................... 13

1-3-3. Rate Enhancement by Molecular Oxygen ..................................................................... 13

1-3-4. Hydrogen Stability on Carbon Surfaces ........................................................................ 14

1-3-5. Surface Oxygen Group Desorption ................................................................................ 14
CHAPTER 2. EXPERIMENTAL ........................................................................................... 15
2-1. Reactants .................................................................................................................................. 16

2-1-1. Carbon Sample .............................................................................................................. 16

2-1-2. Reactant Gases ............................................................................................................... 17

2-2. Equipment and Analytical Instruments .................................................................................... 18

2-2-1. High Pressure Reactor ................................................................................................... 18

2-2-2. Flow Control System ..................................................................................................... 19

2-2-3. Ceramic Reactor ............................................................................................................ 20

2-2-4. Mellen Split Tube Furnace System ................................................................................ 21

2-2-5. Mass Spectrometer ........................................................................................................ 22

2-3. Reactor Characterization .......................................................................................................... 23

2-3-1. High Pressure Reactor Internal Bypass Detection ......................................................... 23

A. Structure Analysis and Experiment Design .................................................................... 23

B. Experiments and Results ................................................................................................. 24

2-3-2. Ceramic Reactor Response ............................................................................................ 28

2-3-3. Fluid Dynamics Calculation .......................................................................................... 28

2-4. Reaction Conditions and Experimental Procedures ................................................................. 29

2-4-1. Reaction Conditions ...................................................................................................... 29

2-4-2. Experimental Procedures ............................................................................................... 30

A. Preparation of Mass Spectrometer ................................................................................. 30

B. Gasiycation ..................................................................................................................... 30

C. Temperature Programmed DesogptionﬂPDZ ................................................................. 3]

D. Surface Structure Characterization ................................................................................. 32

 

iii

2-5. Data Analysis ........................................................................................................................... 33

 

 

 

 

 

 

 

 

 

 

2-5-1. Mass Spectrometer Calibration ...................................................................................... 33
2-5-2. General Data Deconvolution ......................................................................................... 33

A. Time Shift in MASS Data ................................................................................................. 35

B. Species Deconvolution ..................................................................................................... 3 7

2-5-3. CD4 and D20 Deconvolution ......................................................................................... 38
2-6. Carbon Sample Characterization ............................................................................................. 40
2-6-1. Outgas of As-prepared Saran and Coal Char ................................................................ 40

A. Outgas of Saran Char ...................................................................................................... 40

B. Coal Char TPD Proﬁles .................................................................................................. 43

2-6-2. Saran Char Annealing .................................................................................................... 43

A. The Eﬂects of Heating Rate and Annealing Temperature on BET. .................................. 44

B. The Eﬂect oLAnnealing Time ........................................................................................... 46

C. Conclusions ..................................................................................................................... 48

2-6-3. Annealed Saran Char Exposure in Air ........................................................................... 49
CHAPTER 3. THEORETICAL CALCULATIONS ..................................................... 50
3-1. Shift Reaction Equilibrium ...................................................................................................... 51
3-1-1. Equilibrium Constant Changes with Temperature ......................................................... 55
3-1-2. Simulation of Steam Gasiﬁcation (with Shiﬁ Reaction) ............................................... 56
3-1-3. Comparing with Experiment Data ................................................................................. 59
3-2. Desorption Activation Energy of Hydrogen ............................................................................ 60
3—2-1. Average Desorption Activation Energy ......................................................................... 60
3-2-2. Peak Shape Simulation .................................................................................................. 63
3-2-3. Estimating the Desorption Activation Energy ............................................................... 65
3-2-4. Distribution of Desorption Activation Energies ........................................................... 66
3-3. Active Site Density Estimation ................................................................................................ 69
3-3-1. BET Surface Area and Adsorbed H2 ............................................................................. 69
3-3-2. Comparing with Hermann (1986)'s Data ....................................................................... 70
3-4. Reaction Activation Energy ..................................................................................................... 71
CHAPTER 4. CHAR GASIFICATION .................................................................................... 73
4-1 . Saran Char Gasiﬁcation ........................................................................................................... 74
4-1-1. As-prepared Saran Char Steam and Steam/H2 Gasiﬁcation .......................................... 74
4-1-2. Annealed Saran Char H2 /Steam Gasiﬁcation ............................................................... 76
4-1-3. Hydrogen TPD Proﬁles of Annealed Saran Char Following Gasiﬁcation .................... 78

A. _l_12 gasiZication ................................................................................................................. 78

B. _l_1 & Gasiﬁcation ........................................................................................................... 80

C. Steam Gasiﬁcation ........................................................................................................... 80

D. Adsorbed Hydrogen (Deuterium) Change with Conversion ............................................ 80

E. BE TArea Changes with Conversion ............................................................................... 82

4-1-4. The Desorption Proﬁles of CO and CO; after Gasiﬁcation ........................................... 82
4-1-5. Weakly Bound Hydrogen Detection .............................................................................. 84
4-1-6. Discussion of Saran Char Gasiﬁcation .......................................................................... 84

iv

4-2. Coal Char Gasiﬁcation ............................................................................................................. 87

 

 

 

4-2-1. As-received Coal Char Steam/Hz Gasiﬁcation .............................................................. 87

4-2-2. Annealed Coal Char Steam Gasiﬁcation ....................................................................... 89

4-2-3. Hydrogen TPD Proﬁles of Annealed Coal Char Following Gasiﬁcation ...................... 91

4-2-4. The TPD Proﬁles of CO and C02 after Coal Gasiﬁcation ............................................. 92

4-2-5. Discussion of Coal Gasiﬁcation .................................................................................... 92

4-3. Isotopic Effects ........................................................................................................................ 93

4-3-1. Annealed Saran Char Gasiﬁcation by D2 ....................................................................... 93

4-3-2. Annealed Saran Char Gasiﬁcation by D20 .................................................................... 95

4-4. Mechanism Derivation and Identiﬁcation ................................................................................ 96

4-4-1. Mechanism .................................................................................................................... 96

A. Reverse Oxygen Exchange Inhibition .............................................................................. 96

8. Hydrogen Inhibition by Formation ofa C (H) 2 Complex ................................................. 96

C. Hydrogen Inhibition by Formation of a C (H) Complex .................................................. 96

4-4-2. Kinetic Model ....................................................................................................................... 97

4-5. Conclusion ............................................................................................................................... 99
CHAPTER 5. HYDROGEN STABILITY ON CARBON SURFACES ........ 100
5-1. Stability of Adsorbed Hydrogen to Oxidation by 02 ............................................................. 101

5-2. D-H Exchange During Gasiﬁcation ....................................................................................... 106

5-3. H-D Exchange During Gasiﬁcation ....................................................................................... 110

5-4. Conclusions ............................................................................................................................ I 13
CHAPTER 6. RATE ENHANCEMENT ........................................................................... 114
6-1. Oxidation Effects ................................................................................................................... 116

6-2. H; gasiﬁcation Rate Enhancement ......................................................................................... 118

6-3. Steam Gasiﬁcation Rate Enhancement .................................................................................. 121

6-4. Hz/Steam Gasiﬁcation Rate Enhancement ............................................................................. 125

6-5. Conclusions ............................................................................................................................ 126
CHAPTER 7. CONCLUSIONS AND RECOMMENDATIONS ....................... 127
7-1. Conclusions ............................................................................................................................ 128

7-1-1. The Stability of deposited Hydrogen during Gasiﬁcation ........................................... 128

7-1-2. Gasiﬁcation and Hydrogen Inhibition ......................................................................... 128

7-1-3. Rate Enhancement by Molecular Oxygen Oxidation .................................................. 129

7-1-4. Isotopic Effects ............................................................................................................ 129

7-2. Recommendations .................................................................................................................. 130

LIST OF REFERENCES ............................................................................................................ 131

LIST OF TABLES

CHAPTER 2

Table 2-1 Analysis of Carbon Sample .......................................................................................................... 16
Table 2-2 Dynamics Calculation .................................................................................................................. 28
Table 2-3 Possible Fragmentation ................................................................................................................ 34
Table 2-4 Fragmentation Patterns in Argon Carrier ..................................................................................... 34
Table 2-5 The K2 Change with Carrier Gas .................................................................................................. 39
Table 2-6 Comparison of Element Analysis and TPD .................................................................................. 41
Table 2-7 Impurities from Exposure ............................................................................................................. 49
Table 2-8 The Peak Position changes with Exposure Time .......................................................................... 49
CHAPTER 3

Table 3-1 Thennodynamic Data ................................................................................................................... 55
Table 3-2 Heat Capacity Coefﬁcients ........................................................................................................... 55
Table 3-3 Active Sites Occupied by Oxygen ................................................................................................ 70
Table 3-4 Active Sites Occupied by Hydrogen ............................................................................................ 70
Table 3-5 Experimental Data ........................................................................................................................ 71
Table 3-6 Apparent Active Energy for Hz/Ar Gasiﬁcation ........................................................................... 71
Table 3-7 Apparent Active Energy for HZO/HZ/Ar Gasiﬁcation .................................................................. 72
CHAPTER 4

Table 4-1 Rate Constants from Experimental Data at 850 °C ...................................................................... 89
CHAPTER 5

Table 5-1 Experimental Data Summary ..................................................................................................... 101
Table 5-2 Material Balance ......................................................................................................................... 103
Table 5-3 Oxidation Effects ........................................................................................................................ 104
Table 5-4 D-H Exchange Eﬁ‘ects ................................................................................................................ 109
Table 5-5 H-D Exchange Effects ................................................................................................................ 112
CHAPTER 6

Table 6-1 Estimation of the Surface Site Density ....................................................................................... 116

vi

LIST OF FIGURES

CHAPTER 1

Figure 1-1 Zig-zag and arm-chair conﬁguration ........................................................................................... 5
Figure 1-2 Hexagonal etch pit on the basal plane formed by steam .............................................................. 5
Figure 1-3 Mechanism for methane formation .............................................................................................. 7
Figure 1-4 Methane formation mechanism of Huttinger ............................................................................... 9
CHAPTER 2

Figure 2-1 Overall system sketch ................................................................................................................ 18
Figure 2-2 High pressure reactor sketch ...................................................................................................... 19
Figure 2-3 Flow control system ................................................................................................................... 19
Figure 2-4 Ceramic reactor (not to scale) .................................................................................................... 20
Figure 2-5 Mellen split tube fumace ........................................................................................................... 21
Figure 2-6 Temperature proﬁle of Mellen furnace ...................................................................................... 21
Figure 2-7 MASS spectrometer ................................................................................................................... 22
Figure 2-8 High pressure reactor cell (not to scale) ..................................................................................... 23
Figure 2-9 Charred cherry pits oxidation (ceramic reactor) ........................................................................ 25
Figure 2-10 Charred cherry pits oxidation (high pressure reactor) .............................................................. 25
Figure 2-11 Oxygen by-pass in high pressure reactor (-60~+100 cherry pits) ............................................. 26
Figure 2-12 Oxygen by-pass in high pressure reactor(-30~+60 cherry pits) ................................................ 27
Figure 2-13 TPD system response (30 cc H2 pulse) ..................................................................................... 27
Figure 2-14 Experimental scheme ................................................................................................................ 30
Figure 2-15 Before time shift ....................................................................................................................... 35
Figure 2—16 After time shift correction ......................................................................................................... 36
Figure 2-17 After the time shift, artiﬁcial peak no longer exists .................................................................. 37
Figure 2-18 Outgas of as-prepared Saran Char (H2 proﬁle) ......................................................................... 41
Figure 2-19 Outgas of as-prepared Saran Char(CO proﬁle) ......................................................................... 41
Figure 2-20 Outgas of as-prepared Saran Char(COz proﬁle) ....................................................................... 42
Figure 2-21 Low heat rate proﬁle ................................................................................................................. 42
Figure 2-22 Coal char TPD proﬁle, heating rate 10 °C /min ........................................................................ 43
Figure 2-23 BET with pretreatment temperature (Heating rate 3°C/min) .................................................... 44
Figure 2-24 BET with pretreatment temperature (5°C/min heating rate) ..................................................... 45
Figure 2-25 BET area changes with high temperature annealing time ......................................................... 46
Figure 2-26 N2 proﬁle in BET for Saran char ............................................................................................... 47
Figure 2-27 N2 proﬁle in BET for annealed Saran char ............................................................................... 47
Figure 2-28 TPD proﬁle following 2 min. hour exposure ............................................................................ 49
Figure 2-29 TPD proﬁle following 0.5 hour exposure ................................................................................. 50
Figure 2-30 TPD proﬁle following 2 days exposure .................................................................................... 50
Figure 2-31 Uptakes of hydrogen changes with exposure time in air ........................................................... 51
Figure 2-32 Uptakes of CO and N2 changes with exposure time char in air ................................................ 51
Figure 2-33 Uptakes of C02 changes with exposure time in air ................................................................... 52
Figure 2-34 Uptakes of H20 changes with exposure time ............................................................................ 52

vii

CHAPTER 3

Figure 3-1 Equilibrium changes with temperature ...................................................................................... 56
Figure 3-2 Simulation of shift reaction(T=850°C, H2=120, H20=120ml/min) ........................................... 58
Figure 3-3 Simulation of shift reaction(T=850°C, H2=120, H20=120mlfmin) ........................................... 58
Figure 3-4 Simulation of shift reaction (T=850°C, H2=5, HZO=120mllmin) ............................................. 58
Figure 3-5 Simulation of shift reaction(T=850°C, H2=0, HZO=120ml/min) ............................................... 58
Figure 3-6 Simulation of shift reaction(T=725°C, H2=120, H20=120ml/min) ........................................... 58
Figure 3-7 Simulation of shift reaction(T=725°C, H2=12O, HZO=120ml/min) ........................................... 58
Figure 3-8 Simulation of shift reaction(T=725 °C, H2=5, HZO=120ml/min) .............................................. 59
Figure 3-9 Simulation of shift reaction(T=725 °C, H2=0, HZO=120ml/min) .............................................. 59
Figure 3-10 Comparison of experiment data with the equilibrium CO & C02 ............................................ 59
Figure 3-11 Desorption activation energies change with peak temperature ................................................. 61
Figure 3-12 Simulation of desorption(Tm=1000 °C and E=237.8KJ/mol) ................................................... 65
Figure 3-13 Activation energy depends on peak temperature Tm ................................................................. 66
Figure 3-14 Distributions of desorption activation energies for D2 TPD ..................................................... 68
Figure 3-15 Distributions of desorption activation energies for H2 TPD ..................................................... 68
Figure 3-16 Adsorbed Hydrogen vs. Conversion ......................................................................................... 69
CHAPTER 4

Figure 4-1 Gasiﬁcation of Saran char in 3.1 MPa H20(40%)/Ar (60%) at 725°C ...................................... 74
Figure 4-2 Saran char gasiﬁcation in 3.1 MPa HZO(40°/o)/Hz (60%) at 725 °C .......................................... 75
Figure 4-3 4 hr. Annealed Saran char H2 Gasiﬁcation at 850 °C and 3.1 MPa ........................................... 75
Figure 4—4 Annealed Saran char gasiﬁcation ............................................................................................... 77
Figure 4-5 Annealed Saran char 40%HzO/60%H2 gasiﬁcation ................................................................... 77
Figure 4-6 H; TPD proﬁle following 6 hr pure H2 gasiﬁcation of Saran char ............................................ 78
Figure 4-7 Comparison of H2 TPD proﬁles ................................................................................................ 79
Figure 4-8 H; TPD proﬁle following 6 hr 60%H2/40%Ar gasiﬁcation ....................................................... 79
Figure 4-9 H; TPD proﬁle following Steam 40%/Ar gasiﬁcation ............................................................... 80
Figure 4-10 Adsorbed hydrogen change with conversion ............................................................................ 81
Figure 4-11 BET changes with conversion ................................................................................................... 81
Figure 4-12 CO TPD proﬁles following 40% steam/60% argon gasiﬁcation .............................................. 83
Figure 4-13 Hz TPD proﬁles of annealed Saran char after gasiﬁcation ....................................................... 83
Figure 4-14 Coal char gasiﬁcation in 40% HZO/6O% Ar .............................................................................. 87
Figure 4-15 Coal char gasiﬁcation at 725 °C ............................................................................................... 88
Figure 4-16 Annealed coal char Gasiﬁcation at 850 °C ............................................................................... 88
Figure 4-17 N2 BET surface area with conversion ....................................................................................... 89
Figure 4-18 40%H20+60%Ar Gasiﬁcation at 850 °C and 1MPa ................................................................. 90
Figure 4-19 40%HZO+6O%Ar Gasiﬁcation at 850 °C and lMpa ................................................................. 90
Figure 4-20 Desorbed H2 changes with conversion ...................................................................................... 91
Figure 4-21 CO TPD proﬁles following steam gasiﬁcation ......................................................................... 92
Figure 4-22 Annealed Saran char D2(60%)/Ar gasiﬁcation ......................................................................... 94
Figure 4-23 Deposited D during D2 gasiﬁcation .......................................................................................... 94
Figure 4-24 D20 gasiﬁcation of annealed Saran char at 3.1 MPa and 850 °C ............................................ 95

viii

CHAPTER 5

Figure 5-1 CO and C02 proﬁle during the oxidation ................................................................................ 102
Figure 5-2 H2 TPD proﬁles change with oxidation time ........................................................................... 102
Figure 5-3 D2 TPD proﬁles change with different oxidation time ............................................................. 103
Figure 5-4 DZO proﬁle during the oxidation ............................................................................................. 105
Figure 5-5 H20 proﬁle during the oxidation ............................................................................................. 105
Figure 5-6 TPD of 6 hours D2+Ar and 0.8 hr H2+Ar gasiﬁcation ............................................................ 106
Figure 5-7 TPD of 6 hr. D2+Ar and 2.5 hours H2+Ar gasiﬁcation ........................................................... 107
Figure 5-8 CD4 and CH4 evolution during switch ..................................................................................... 107
Figure 5-9 D2 TPD proﬁles with H2 gasiﬁcation time .............................................................................. 109
Figure 5-10 TPD of 4 hours H2+Ar and 0.7 hr. D2+Ar gasiﬁcation ........................................................... l 10
Figure 5-11 TPD of 4 hours H2+Ar and 1.3 hr. D2+Ar gasiﬁcation ........................................................... 111
Figure 5-12 TPD of 4 hours H2+Ar and 2.5 hr. D2+Ar gasiﬁcation ........................................................ 111
Figure 5-13 Hz TPD proﬁles with D; gasiﬁcation times ............................................................................ 112
CHAPTER 6

Figure 6-1 Experimental scheme ............................................................................................................... I 15
Figure 6-2 CO and C02 desorption proﬁles .............................................................................................. 116
Figure 6-3 TPD proﬁles after Dz gasiﬁcation and oxidation .................................................................... 117
Figure 6-4 Rate enhancement by oxidation (H2 gasiﬁcation ..................................................................... 118
Figure 6-5 Armchair carbon maintain its structure during gasiﬁcation ..................................................... 119
Figure 6-6 Rate enhancement by oxidation ( 725 °C) ............................................................................... 121
Figure 6-7 Rate enhancement by oxidation.( 850 °C) ............................................................................... 123
Figure 6-8 Rate enhancement by oxidation ( 725 °C) ............................................................................... 125

CHAPTER 1
INTRODUCTION

l-l. Background

The production of gaseous fuel from coal has been practiced for hundreds of
years. In 1947, production reached a peak in the US, and then was gradually replaced by
natural gas and petroleum. During the 1970’s, it came to be realized that the world’s
supplies of inexpensive petroleum and natural gas are limited. The eventual depletion of
petroleum and natural gas reserves necessitates the development of alternative sources to
replace these inexpensive and clean fuels.

In US and most other countries in the world, one potentially alternative fuel
source is substitute(synthetic) natural gas (SNG) produced by gasiﬁcation of coal, of
which the world has abundant reserves. SNG is superior to coal as an energy source
because it is an easily transported fuel, burns much cleaner and more efﬁciency than coal,
and can be converted to oil and other chemicals without technical and economic
difﬁculty. For all possible gasiﬁcation processes, steam gasiﬁcation is a top choice,
because water is cheap and the gasiﬁcation process is simple to carry out. Coal
gasiﬁcation is not currently used on a wide scale because extreme conditions are needed
to achieve reasonable reaction rates. The extreme conditions lead to high production
cost. Another economic reason is the cheap petroleum and natural gas still are available,
so we do not have to use the troublesome coal resource.

The presence of hydrogen at carbon (or coal) surfaces strongly inﬂuences steam
gasiﬁcation rate. The process by which hydrogen inhibits gasiﬁcation is not yet well
characterized. If the inhibition were there, understanding its mechanism would have

potential to minimize this retardation.

1-2. Literature Review
1-2-1. Hydrogen Adsorbed on the Carbon Surfaces

Hydrogen chemisorbed on a carbon surface is very stable, and is generally
accepted as dissociative in nature. Numerous studies (for example Steinbergm)have
shown that prolonged outgassing at 1000 °C will not remove all dissociatively adsorbed
hydrogen; temperatures approaching 1500 °C are required (Redmondm ). Also, the
hydrogen adsorption rate is very rapid. Biederman et alm found that dissociatively
adsorbed hydrogen could saturate a graphite surface at 1100 °C and 3 millitorr H2. The
hydrogen adsorption equilibrium constant is high, up to 253 atm'o’s at 700°C LYﬂg and
Xangm). Furthermore, this constant can be strongly affected by impurities in the carbon
surface which may act as hydrogen dissociation sites.

Adsorbed hydrogen can be characterized by temperature programmed desorption
(TPD). In addition to a peak in starting around 900 °C in TPD proﬁle, which is indicative

of dissociatively bound hydrogen (C(H)), several investigators (Huttinger et all”) reported

 

a second peak around 600~800 °C following exposure to gases containing molecular

hydrogen. This peak is suggested to represent associatively bound hydrogen (C(H)2) on

the carbon surfaces.

 

;Steinberg M.. 1987 Int. Conf. on Coal Sci. 11, J.A. Moulijn ed.. 953 (1987).
:HRedmond J. P. and Walker, P..L Jr., J. Phys. Chem. 64. 1093 (1960)
:Biederman D. L. et a1 Carbon 14, 351 (1976).
zYang. R..T and Yang. K. L., C_a_rbon 23(5), 537 (1985).
5Huttinger, K..J and Merdes W..F, —C__arbon 30(6). 883 (1992).

 

 

 

4

Uncatalyzed carbon gasiﬁcation rate decreases dramatically with conversion. At
the beginning of the gasiﬁcation, the ratio of carbon edge atoms in zigzag and arm chair
conﬁguration may be undeﬁned ( Figure l-la). Because of the irreversible hydrogen
inhibition, the reaction rate is reduced to near zero when the edges of the carbon layers
are formed exclusively by carbon atoms in zigzag form (Figure l-lb) (Huttingerlsl).

Experiments and molecular orbital simulation showed that the hydrogen only
could chemisorb on edge sites of graphite, not on the basal plane. The gasiﬁcation of
"armchair" edge carbons is much easier in hydrogasiﬁcation and steam gasiﬁcation (Pan
and Yangml), while the gasiﬁcation of oxygen and carbon dioxide shows no edge

[61,Yang and Yangm). Therefore, as hydrogasiﬁcation and

preference (Yﬂg and Dﬂ
steam gasiﬁcation proceed, the more highly reactive armchair carbons will be consumed.
As shown in Figure 1-2, the edges of hexagons are formed by carbon edge atoms in
zigzag conﬁguration. The reason is that these carbon edge atoms are blocked by

formation of C(H) complexes, because their further reactivity with hydrogen is lower

than that of the carbon atoms in armchair conﬁguration (Pan and Yanglm). This

 

interpretation is strongly supported by the observation that gasiﬁcation inhibitors based
on halogen, boron, or phosphors compounds block only carbon edge atoms in zigzag
conﬁguration (1119mm).

Although an enormous number of studies have been done on hydrogen adsorption
and its role in gasiﬁcation, the exact structure or pathway of formation of this

associatively bound hydrogen is not fully understood.

 

6 Yang. RT. and Duan. R.Z., Carbon 23(3). 325 (1985).
7 E,J Hippo, N murdie and A. Hyjiazie, Carbon 27. 689 (1989)

 

 

a b

Figure 1-1 (a) Before reaction: undeﬁned ratio of carbon edge atoms
in zig-zag and arm-chair conﬁguration (b) After reaction:
prevailing carbon edge atoms in zig-zag conﬁguration

Zig-zag conﬁguration

Arm-chair
conﬁguration

 

Figure 1-2 Hexagonal etch pit on the basal plane formed by steam

1-2-2. Oxygen Complexes on the Carbon Surfaces

It is generally believed that the major source of active sites in all gasiﬁcation
reactions comes from the desorption of oxygen functional groups from the carbon
surface. These groups desorb in the form of carbon monoxide and carbon dioxide when

samples are heated to reaction temperatures (Treptau and Millerl 81), and in the form of

 

water during hydrogasiﬁcation (Blackwood [9]). Hydrogasiﬁcation rate has been shown to
be a strong function of the oxygen content of various Chars (Blackwoodm), and initial

rate is a strong function of oxygen surface concentration (Treptau and Miller [8]).

 

Hydrogasiﬁcation rate can be increased by an order of magnitude by addition of 0.1%

oxygen to the reactant gas (Cao and Back [10]). However, the temperature of C-HzO

 

reaction is higher than 700°C, so only surface oxygen groups with intermediate stability
at these temperature will contribute to the reaction. Highly stable groups may actually be
considered as poisons (Walkerll 11).

Recently, Pan and Yang “2] had detected a signiﬁcant amount of oxygen complex

 

on graphitic carbon (but not on nongraphite carbon) which desorbed as CO at
temperature higher than 1200°C. On the basis of this result and molecular orbital
calculations, they proposed an oxygen complex which was formed by bonding oxygen
atoms on the lattice carbon atoms that were saturated. It was shown that the caved in
carbon atoms on the zigzag edge of the graphite could form bonds with oxygen atoms

and that such bonding would signiﬁcantly weaken the carbon- carbon bonds, hence

 

° Treptau. M.H. and Miller. 11.1 .. Carbon 29, 531 (1991).

9 Blackwood. J. D. Aust. J. Chem. 12, 14 (1959).

‘° Cao. JR. and Back M.H.. Carbon 23. 141 (1985).

“ Walker. Carbon. Vol. 29, P41.1991

‘2 Pan.Z..J and Yang. R. T. Ind Eng. Chem Res 1992.31.2675

 

 

 

 

 

7

leading to gasiﬁcation. So, the understanding of active surface oxygen complexes is a key
to understanding the carbon gasiﬁcation reaction. Surface oxygen groups play a
dominant role in the steam gasiﬁcation reaction. Mostly recently, the molecular orbital
calculations gig and Yimgfm) showed that C-C bond energies (for desorption from
graphite) are lowered by approximately 30% by bonding of oxygen on the saturated
carbon atoms. The bonding of an oxygen atom to a saturated carbon atom (in caved-in
position) adjacent to the unsaturated edge carbon atoms that already bonded to an oxygen
atom forms a new type of complex. The molecular orbital calculation results also
showed that this type of complex is substantially more active than known complexes
(semiquinones and carbonyl). This more active complex may be a main contributor to the

carbon gasiﬁcation.

 

Figure 1-3 Mechanism for methane formation on zigzag and armchair faces by
successive H addition. C-C bond breakage upon the third H addition.
(Redrawn from Pan and Yang, 1990“”)

 

1-2-3. Methane Formation in Gasiﬁcation

There are two relatively new theories to account for methane formation. The ﬁrst

one attributes methane formation to direct attack of carbon by molecular hydrogen, which

 

‘3 Chen and Yang, Ind. ErLtLChem. Res..32.2835, 1993

 

8

is given by Pan and Yang 0990“”). The second one proposes methane formation only

 

occurs on dangling carbon atoms (Huttinger,1992[5]).

Pan and Yang’s mechanism is shown in Figure 1-3. Methane formation may

 

come from both zigzag and armchair edge faces. In both case, the surface C-C bond
cleavage is required for the third H addition. After that, dangling bonds remain and
methane formation is energetically very favorable. The simulation results of extended
Hucke molecular orbital (EHMO) showed that the ﬁrst H addition is relatively easier than
the second H addition on both edge faces of graphite. The simulation also indicates the
edge surface carbon atoms become saturated (hence free valence reaches a minimum) and
inactive for the chemisorption after two H addition. Consequently, the third H addition or
the C-C bond breakage step is the rate limiting step for methane formation. The zigzag
edge atoms are more reactive for hydrogen chemisorption for both ﬁrst and second H
addition. A most signiﬁcant result from this simulation is that there is a reversal in the
relative reactivity between the two edge planes upon the second H addition. After second
H addition, the cleavage of C-C bond is easier for armchair face. These results are
consistent with the experimental observations. It seem the formation of two CH3 groups
on the armchair face appears to be sterically unfavorable due to possible H-H interactions
from neighboring -CH3 groups. However, their EHMO calculation showed that the

formation of such a structure is entirely feasible with no H-H interactions.

O
. . I 0 g I
\C¢C\C¢C\C/C\C/C\C/C\C/ +H20 \C¢C\C/C\C/ \G .C/CW’
(L, (b b b b b ———* (b b b l b (E +CO+H2
/§C/ \C/ §C/ \C/ \C¢\ /\(l;/‘\(l3/ §c|/ \(nz/ \CI:¢\
| | | || |
l +1420
\C¢6\C- FCH HEH 'C/éw/ +HZO \Céc\C¢d\C' HCH C/ \C/
3113 Wabbit/immm
/\C/ \\.C/ §C/ \C/ \C¢\ /\C/ N/ \C/ \c/ \c/\

Figure 1-4 Methane formation mechanism of Huttingerls]

The second mechanismls] is totally derived from the gasiﬁcation of PVC coke.
After proving the homogeneous formation of methane is impossible (1000 °C and 10 bar)
by experiments, they claimed the methane must come from the carbon surface. But the
-CH2 complex is deﬁnitely not involved in the formation of methane and is a inhibitor to
methane formation. The present of steam is an absolute prerequisite. They used a old
model (Blackwood, 1958“”) (Figure 1-4) to rationalize their mechanism. The decisive
intermediates of methane formation are dangling carbon atoms which form methylene or
methyl groups, but no surface complexes. A pre-condition of the formation of dangling
carbon atoms, and thus methylene or methyl groups, is ring opening reactions, which are
only possible by the attack of steam.

Another possibility would be that methylene groups are directly formed as
intermediates. Finally, methane is formed by the reaction of methylene or methyl groups
with hydrogen. From this mechanismm], the hydrogen inhibition of methane formation
and hydrogen is also necessary to form methane.

It seems that these two theories are contrary at ﬁrst glance. But when we look into

the materials they used, it is not difﬁcult to understand the consistency of the two

lO

mechanisms. Pan and Yang's mechanism is mainly from graphite (gasiﬁcation and pits

 

etching) and Huttinger's mechanism is from the gasiﬁcation of disordered (turbostratic)
carbon (PVC coke). The ﬁrst mechanism may be always correct, but the rate is very slow
(etch decoration experiments need 10 hours to get a pit of 0.1 pm). Huttinger even could
not detect the methane formation using their coke (low surface area 1 m2/ g) in pure
hydrogen. The second mechanism is relative fast for amorphous carbon. For the really
possible gasiﬁcation processes (low to middle rank coal ), the carbon is most like

Huttinger's coke and not graphite. So, relatively, the second mechanism is more

important.

1-2-4. CO Formation in Steam Gasiﬁcation

Gasiﬁcation of carbon (coal, coke) with steam leads to carbon monoxide and

hydrogen as main products. The overall reaction is:

C+H20—>C0+H2 131-1
This reaction has been the basis for production of synthesis, reduction, town and
fuel gases for many decades before petroleum. In view of the great importance of this
reaction, it is surprising that the ideas about the mechanism and also the rate limiting step

are still speculative. Three models are proposed for this reaction, namely the oxygen

[16]

exchange model (Ergunnsl) and two hydrogen inhibition models (Long and Gadsby et

[18]

alml). Huttinger gave a summary of all three models:

Oxygen Exchange: C I + H20 —k—‘>C(O) + H2 E 1-2

 

“ Zielke. CW. and Gorin. 15., Ind. and Enq. Chem. 47(4), 820 (1955).

'5 Ergun. S and Mentser. M In Chemistry and Physics of Carbon (by P.L.Walker. Jr) Vol. 2 P203. 1966
‘6 Long. F.J.. and Sykes. K.W.. Proc. Royal Soc. London A193. 377 (1948).

'7 Gadsby. J.. Hinshelwood. C.N., and Sykes. K.W.. Proc. Royal Soc. London A187. 129 (1946).

‘° Huttinger. K.J.. ACS Div. Fuel Chem. Prepr. 34(1). 56 (1989).

 

 

ll

Gasiﬁcation: C(O) —i+ C0 + C, E 1.3

Associative hydrogen adsorption: Cf + H2 ——"3—> C(H)2 E 1-4
Dissociative hydrogen adsorption: Cf + -21-H2 —£—>C(H) E 1-5

where Cf represent a free carbon site.

All models yield formally identical surface CO formation rate equation:

k. PW
1+ 1:21;. + 1:,sz

 

CO formation rate rs = [51-6

Of these model , dissociative hydrogen adsorption gives a value of 0.5 for n in
Equation E l-6, which has been found to be correct for low hydrogen pressures and

subatrnospheric steam pressures (Yang and Yang [4]). Reverse oxygen exchange and

 

associative hydrogen adsorption both give values of n=1 .0 in this rate expression, which

was reported in early studies (Blackwood“gl)

 

1-2-5. Hydrogen Inhibition

All researcher agrees that the presence of hydrogen greatly reduces steam
gasiﬁcation rate. Steam gasiﬁcation rate drops by an order of magnitude with the addition
of only 1 PPM hydrogen (mum). Carbon dioxide gasiﬁcation is also inhibited to this
degree by hydrogen at low pressures (Egget a1 “’61). During steam gasiﬁcation, carbon
dioxide and methane formation rates are decreased as well as carbon monoxide and
hydrogen formation rates due to hydrogen inhibition [Huttinger 5]. Associative hydrogen
adsorption has been found to contribute to inhibition in steam gasiﬁcation at higher
pressures (Huttingernsl). TPD studies show a hydrogen desorption peak following

gasiﬁcation at 600 °C, indicative of C(H)2 surface groups. Much larger peaks were found

12

above 1000 °C , proving that dissociative hydrogen inhibition still dominates the carbon
surface. These investigators have reported reaction rates approaching zero at carbon
conversions as low as 40% in steam/hydrogen mixtures. Gasiﬁcation rate is also greatly
reduced after exposure of carbon to hydrogen in sequential steam/hydrogen/steam
reactions (Hutinger [5 1). The rate does not return to its previous value after the second

steam gasiﬁcation is initiated, suggesting irreversible blockage of active sites.

1-2-6. Isotopic Effect

Relatively few gasiﬁcation studies have been performed using isotopes, even
though isotope effects in chemical reactions is very important @1211). Gasiﬁcation
rates of graphite in H2O are reported to be twice as high as graphite in D20. Some
investigators concluded that breakage of the C-H bond is involved in the rate limiting
transition state complex mp”), while investigators of later studies concluded that the
difference in rate is due to a shift in the oxygen exchange equilibrium constant @1221).
Very small isotopic effects were found in the H2 and D2 gasiﬁcation of graphite at 1200

°C and 20 torr hydrogen pressure (Gulbransen [23]). I-I/D exchange has been shown to

take place readily over carbon at 300 °C (Ishikawa, Y. et al [24]).

 

 

‘9 Blackwood. DJ and McGrory. F.. Aust. J. Chem. 11. 16 (1958).

2° Montet, G.L.. and Myers. G.E.. Carbon 9. 673 (1971).

2‘ Yates. J. T. Jr. and McKee. o.w., J. Chem. Phys. 75. 2211, (1981).

22 Mims. CA and Pabst, J.K., ACS Div. Fuel Chem. Prepr. 25(3). 263 (1980).

’3 Gulbransen. E.A.. Nature 212, 1420 (1966).

2‘ lshikawa. Y. et al., Chem. and Phys. of Carbon 12. Walker, P.L.Jr. and Thrower, P.A.. ed.. 40 (1975).

 

 

 

 

13

1-3. Research Objectives

The main objective of this thesis is to characterize the concentration, stability, and
reactivity of hydrogen adsorbed on carbon surfaces after gasiﬁcation.

We use annealed Saran char as main carbon sample. Saran char is generally
considered to be amorphous, but it does contain small regions of ordered, graphitic

structure. From the book of Blackmanm'

, we know the coke would consist of small
crystals when the pitch is charred above 1000°C. Our X-ray experimentsm] conﬁrm that
the Saran char contains graphitic structure over small domains and also contains
disordered structure over larger domains. The small graphitic crystals in Saran char make

the sample have structures similar to graphite; the small size of these ordered structures

increases the gasiﬁcation rate to measurable levels.

1-3-1. Characterization of Adsorbed Hydrogen

We use high temperature TPD (temperature program desorption) to measure the
desorption of hydrogen following H2/steam gasiﬁcation. Using this information to
understand the nature of adsorbed hydrogen, attempt to characterize the hydrogen

inhibition mechanism.

1-3-2. Isotope Effects
Prior research has shown that char gasiﬁcation rate in H2O/H2 and D2O/D2 was
different. This phenomena probably is from different adsorption energies (Ozaki [2”). If

oxygen exchange is reversible and rapid, there will be an isotope effect on the equilibrium

 

25 Ozaki, A. Isotopic Studies gt ngrogeneous Catalysis, New York, Academic Press, Inc. (1977).

l4

gasiﬁcation experiments and evaluation D2 stability on the carbon surface to verify the

isotopic effect and try to get some details about this issue.

1-3-3. Rate Enhancement by Molecular Oxygen

If adsorbed hydrogen does inhibit the steam/H2 gasiﬁcation, it is reasonable to
believe removing the adsorbed hydrogen to recover the activity of carbon. Of course, we
already know that the reaction rate will decrease rapidly in the initial phase of
gasiﬁcation. So, we cannot expect the rate recovery to last very long. Also, oxidation by
molecular 02 will not only react with surface hydrogen, it also will react with surface
carbon. This may change the surface structure. The formation of new arm-chair edges

will increase the reaction rate too.

1-3-4. Hydrogen Stability on Carbon Surfaces
Understanding the stability of adsorbed hydrogen is also important for ﬁguring
out the hydrogen inhibition mechanism. H-D, D-H exchange and oxidation by 02 will be

used to exam the stability of adsorbed H2/D2.

1-3-5. Surface Oxygen Group Desorption

The oxygen group on carbon surface is already investigated by many researchers.
Huttinger[5] also attributed to methane formation to the existence of surface oxygen.
Determining the relative amount of oxygen and hydrogen on the carbon surface will help

to further understand the inhibition mechanism.

CHAPTER 2

EXPERIMENTAL

16

2-1. Reactants

2-l-l.Carbon Sample

In this study, samples used for gasiﬁcation are Saran char, coal char and charred
cherry pits. Saran char was prepared from Dow Saran resin by heating at 10°C/min to
900°C and holding 30 minutes in a nitrogen ﬂow of 400 cc/min. The average char yield is
25.50%. The foam-like char is crushed and sieved into -60~+100 mesh powder by a
ceramic pestle and mortar. Saran chars are further pretreated by heating at 5 °C /min. to
1500 °C in argon for 6 hours in order to anneal and clean the sample surface. Coal char is
from the previous research (PSOC 1493), and is also further heated at 5°C /min. to
1500°C in argon for 6 hours. Charred cherry pits are only used for by-pass detection in

the high pressure reactor, and were prepared in previous researchm]. The properties of
two chars are given in Table 2-1.

Table 2-1 Analysis of Carbon Sample

 

 

 

Component Saran Char Coal Char Annealed Saran Annealed Coal
Carbon (wt.) 96.4 75.3 97.76 87.19
Hydrogen (wt.) 0.5 0.5 0.019 < 0.001
Nitrogen (wt.) 1.0 1.3 < 0.5 < 0.5
Sulfur (wt.) 0.4 3.6 0.0023 1.36
Chlorine (wt.) NA NA < 0.0010 0.0062
Ash (wt.) 0.1 17.3 0.46 15.00
Oxygen (by diff.) 1.3 2.0 1.78 < 0.1
TSA (BET) mz/g 850 280 800 N2 BET 13.5 N2 BET
TSA (C02) mz/g 1330 440

 

 

2-1-2.Reactant Gases
All gases (Argon, Argon with 1% Krypton and Hydrogen) used are ultra high
purity grade (99.999%) except for D2, which contains 99.4% D2 and 0.6% HD. Two

kinds of argon gases are used. One is mixed with 1% krypton for characterization of

17

system transients during gasiﬁcation and is also used as the dilute ( product carrier for
mass analysis) along with hydrogen and steam. The other one is pure argon and is used as
an inert only. All gases are further puriﬁed by ﬂowing through R&D Separations Model
OT500-2 Oxygen/Moisture Traps to remove water and reduce oxygen to less than 10
PPM. HPLC grade H2O is outgassed at 100 °C (boiling) for 30 min to remove the

dissolved air and stored under argon or nitrogen to minimize dissolved oxygen contain in

the steam.

 

26 Michael G. Lussier etc. Carbon Vol. 32 N08, Ppl493-1598,1994

18

2-2. Equipment and Analytical Instruments

A high pressure reactor, gas ﬂow controllers and a mass spectrometer are the
main experimental apparatus. A high temperature reactor (1500 °C) is used for TPD
(temperature program desorption) and char annealing. A sketch of the overall system is in

[27]

Figure 2-1. All equipment is designed and installed by Lussier , the more descriptions

are given in his dissertation.

 

 

 

Mass Flow High pressure Mass
Controller Reactor Spectrometer

 

 

 

High temperature 486 PC
Reactor Computer

 

 

 

Figure 2-1 Overall system sketch

2-2-l.I-Iigh Pressure Reactor

The high pressure reactor (Figure 2-2) consists of a horizontally mounted 5.1 cm
OD x 1.9 cm ID Haynes Superalloy pressure vessel capable of operation at 1000 °C and
6.6 MP3, with a ﬂange closure on one end. It is externally heated with a Lindberg
electric furnace. The main pressure vessel houses a 19 mm OD x 11 mm ID x 57 mm
length Inconel 625 packed bed microreactor, which is capped at both ends. The
microreactor, shown in Figure 2-8, is quartz lined, has quartz frits on both ends, and has
ceramic ﬁber gaskets to prevent any sample contact with metal surfaces. This provides a

9 mm diameter x 31mm length internal sample chamber.

 

27 Michael G. Lussier, Ph.D. dissertation Michigan State University (1997)

Steel shell Microreactor Quartz lmer

 

 

 

Figure 2-2 High pressure reactor sketch
2-2-2.Flow Control System
Four mass ﬂow controllers and a four way valve allow rapid switching and
mixing of up to four gas streams at ﬂow rates of 0-300 cm3 (STP)/min per stream as
shown in Figure 2-3. 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 10 cm3 (STP)steam/min. Heating tapes are used to prevent steam

 
    

To reactor

H1 Mass
controller
To back pressure

controller
MK! Mass HPLC pump
controller

 

condensation.

Figure 2-3 Flow control system

20

2-2-3.Ceramic reactor

The ceramic reactor is an externally heated ﬁxed bed rector (Figure 2-4). Its main
structure consist of three ceramic tubes. The inner tube with one end sealed is used to put
in a thermocouple. The outer tube acts as a gas passage and container with one end
sealed. The middle tube is the sample bed with one end open and the other end capped by
a porous ceramic frit that can allow the gas to enter the reactor and seal the char sample at
the same time. Special stainless steel valves and polyﬂuortetraethylene and rubber rings
are used to seal the gaps between tubes. The reactor can be heated to 1500 °C by a
Mellen split tube furnace (see next section), which is equipped with a programmable
temperature controller to facilitate a linear temperature ramping. This reactor is designed
for TPD. Blowing air is used to cool down the part of the reactor outside the furnace to

protect the rubber seals.

Cooling Air
Thermocouple

  

Alumina Sample

  

 

 

 

 

 

 

 

 

Gas outletl U \ Frit

1‘ Gas inlet

Figure 2-4 Ceramic reactor (not to scale)

21
2-2-4.Mellen Split Tube Furnace System

 

 

 

Power
supply

 

 

 

 

Figure 2-5 Mellen split tube furnace
This system (Figure 2-5) contains a tubular furnace and programmable controller.
The designed temperature of the fumace is 1500°C. The fastest heating rate is about
5°C/min (between 20°C to 1500°C). At the lower temperature, the heating rate may be
over 5°C/min, but above 1350°C, it is difﬁcult to achieve a rate over 5°C/min. Only when
the furnace is well preheated ( for example, the fumace is over about 700°C ), can the
heating rate exceed 5°C/min at temperatures over 1350°C. The temperature proﬁle of this

furnace is quite ﬂat (Figure 2-6 is the temperature proﬁle at 1500°C).

1600
1400 w
1200.
1000 ;
800 -
600

400 ~
200‘

Temperature (C)

 

Distance from center (in)
Figure 2-6 Temperature proﬁle of Mellen furnace

22

2-2-5.Mass Spectrometer

An Ametek Dycor M100 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, and is interfaced with a personal
computer for data collection and storage. Product gases from the high pressure reactor
pass through a back pressure regulator to reduce pressure to atmosphere. A condensation
trap is used to separate steam from other gases, to avoid ﬂooding the vacuum chamber.
By a ﬁne quartz capillary tube, sample gases are continuously drawn to the vacuum

chamber. System sketch is given in Figure 2-7.

 

 

 

Mass Mass 486 PC
detector spectrometer computer
controller

 

 

 

 

    
    

Sample in

 
   
 

  

Molecular
PumP

Vacuum chamber

 

Figure 2-7 MASS spectrometer

23

2-3. Reactor Characterization

2-3—1. High Pressure Reactor Internal Bypass Detection

A. Structure Analysis and Experiment Design

 

Because of the structure of the high pressure reactor (Figure 2-8), by-pass is

 

 

 

 

unavoidable.
r Quartz liner
QVVQI‘VFJ {CA ,_;EME C_ELJ CL" L)" L/ 7123
Inﬂuem ~ ” " jEfﬂuent
——>-? ‘——>-
.444 4-4. .. .. . . .i\~r\r\.\~2.‘

' s..\_mwhx, Inconel Steel

Figure 2-8 High pressure reactor cell (not to scale)
The reactant gases have two possible ways to go:
Flow through the packed char bed
By-pass through the gap between the quartz liner and Inconel steel reactor.

The quartz is very fragile, so it is impossible to use force to avoid by-pass.

To detect the extent of by-pass, three experiments were designed. The ideas are

Use high activity carbon (charred cherry pits) to make sure 02 converts to CO2

quickly and completely.
Using low temperature (about 600 °C) to limit the C02 + C <—> CO reaction (at

that temperature, the equilibrium constant is 0.27).

24

In ceramic reactor, the internal by-pass is impossible. By comparing the
experimental results of extent of gasiﬁcation in the two reactors, we can calculate
the extent of by-pass.

Two different particle sizes (-60~+100 and -30~+60) are used to detect the

relative amount of bypass, which should change with particle size. The actual

experimental conditions are

0

O

606 °C

60 cc/min. 5% 02 (or 2.5%) in argon

1~2 hours gasiﬁcation.

-60~+100 and -30~+60 charred cherry pits

atmospheric pressure in ceramic reactor and high pressure reactor

B. Experiments and Results

 

Figure 2-9 gives the product gas proﬁles for -60~rl-100 charred cherry pit

oxidation in the ceramic reactor. From this ﬁgure, we can clearly see that the dominant

product is CO2 and no 02 can be found in the outlet. That means all reactant O2 is

converted to CO2 and CO in these reaction conditions.

Using the same sample and conditions, the product proﬁle from the high pressure

reactor is much different (Figure 2-10). Quite a lot of molecular 02 still exists in

products. That means that by-pass happens.

“ f ' FigureldﬂChmed

25

 

 

r —e— 002 cclmin
2 L —v— "02"cc/min
—><— CO" cclmin

 

 

Products ﬂow rate cclmin
0'1
1

 

 

 

0 20 4O 60 80 1 00 ‘l 20

. M w Timeminutes) . w. . _
* , Figure29Charred cherry emanation-(mane.r.;:‘_t.:reeor)~

2.5

 

 

—o— 002 cclmin output
—fB- CO cclmin output
—x— 02 cclmin output

  
  
    

2.0 —

 

 

1.5 r-

1.0 _

0.5 -

     

 

Products Flow rate Ml/min

 

00 :~ .. .. :: :: :1 :: :: ::
0 20 40 60 80 100

g (Time (minutes)

 

120 140 160

  

pusoaaarmrmghpressmireeoo » :

26

At this low temperature, the catalytic effect of the ceramic reactor wall is

negligible. The second reaction (002 + C <—> CO) is limited. So, we can estimate the by-

 

 

 

 

 

 

 

OutletO

ﬁ'ct' b th 1' f—l.

pass a mu y eraroo(1nlet02)
.9 3.5 50
E 30 - - - - ‘ ‘ ‘ “ ‘ ‘ ‘ "
8 - i— ' ' ' - - - I - I - I I “50 g
2 2.5 P _ _ _ I <3
920- ======='-- “403
g :2--- ‘5’
E 1.5 F ' ‘ . . A A — 30 1':
‘0 OE . ‘ ‘ i i A ‘ - . . A 8
{3'5 1- ‘ +02cc/minoutput 20 (I)

—o—ozoc/1(rpur) “ o

E 0.5 r u —B—by-pee;n92n 3‘
CL 0.0211991111ni1441L rrrirrrrerrr ILIJJG V1 10

 

O 0 4O 60 80 100 120 140 160

.,,A . “mammal _ I .
, , Figure‘z-IIOXYSenBy-vass inhigh pressurereactor-605100 icheﬂ‘yipits) . ‘ ‘ ‘

For -60~+100 charred cherry pits (Figure 2-11), ﬁ'om the amount of exiting 02 we
can calculate the bypass fraction, which is about 40%. Little increase with conversion can
be found. This phenomenon is caused by the ash particles collapsing, which increases the
char bed resistance. After the experiment, a layer of ash was found on the inlet side of the
sample.

For -30~+60 charred cherry pits (Figure 2-12), the bypass is about 30 %. It is
reasonable, because the ﬂow resistance in the char bed is smaller than for the -60~+100
char.

From the above experiments, we can conclude that the extent of gas by-pass in the

high pressure reactor is about 40% for -60~+100 char.

Figure2-12xygenbypassmhrghpfessurereactor(30~+600herrypt8) - 4

Instant Rate

4
3.5
3
2.5
2
1.5
1
0.5
0

Products Flow Rate ML/min

4.5E-8
4.0E-8
3.5E-8
3.0E-8
2.5E-8
2.0E-8
1.5E-8

27

 

I

 

F
a. 1

A v A '
v a 2 ’4 v

A
V

9 9

i—l gong-40-:nngconnannﬁn

- v v C " .- v v v v v v ' v v v - .

 

p —£r— 02 cclmin output
—9— 02 cclmin (input

-—<+— by-pass %

 

.
I l a I I A A -
. . A A a . n - A I
. n n

l l l l

. . 1
n n A

l

 

l.

 

 

\
uuuv

0 20

4O 60 80 100

Time (minutes)

120

VU

35
30

25
20

15
10

140 160

 

 

J l

 

 

l l l L

 

 

5

10 15
_. ..T®9;(Wut98),..,_

20

% ssed-Kg ueBAxo

28

2-3-2. Ceramic Reactor Response

The dead volume in the TPD reactor will broaden the TPD peak. To eliminate this
problem, the ceramic reactor was designed to achieve as small as possible a dead volume.
Figure 2-13 is a specially designed pulse experiment to check the effect of back mixing
on the TPD proﬁle, in which a H2 pulse was injected into the purge line. The two
injections produced two sharp step responses. The tail of the peak exists for less than 3
minutes. For a typical TPD experiment, the desorption time is more than 300 minutes. So,

we have conﬁrmed that back mixing can be neglected in our TPD experiments.

2-3-3. Fluid Dynamics Calculation

Because of the by-pass, it is necessary to check diffusion resistance. The internal
diameter of quartz liner is 9 mm. Char particle size is -60 ~ +100 (0.149~0.250 mm). The
sample average particle diameter is 0.2 mm. The ratio of particle diameter to quartz liner
is about 0.02. From Perry’s handbook, we can get bed porosity e=0.32. The parameters
calculated are given in Table 2-2.

Table 2-2 Dynamics Calculation

 

 

Pressure Gas ﬂow rate Cross area U crn/sec. UT cm/sec. UT cm/sec.
(MPa) cm3/min. cm2 (apparent) true velocity after by-pass
0.1 300 0.63585 7.863 24.57 14.74
1 300 0.63585 0.7863 2.45 1 .47
2 300 0.63585 0.3931 1.23 0.74
3 300 0.63585 0.2621 0.82 0.49

 

 

 

In most case we operate the reactor at 3.1 MPa. After 40% by-pass, the true gas
velocity of 0.49 cm/s is large enough compared with the very low gasiﬁcation rate. So,

we can conclude that the mass transfer resistance will not affect the gasiﬁcation reaction.

 

29

2-4. Reaction Conditions and Experimental Procedures
2-4—1. Reaction Conditions

For our gasiﬁcation process, temperature, pressure, H2 (D2)/steam/argon ratio, and
the reaction times are the changeable parameters.

0 For most experiments, the temperature was ﬁxed at 850 °C. At this temperature,
the effluent species concentrations are well within the detectable range, and mass
transport resistances are not signiﬁcant (see Section 2-3). For some highly active
samples (coal char, cherry pits char), a temperature of 725 °C also was used to
slow down the reaction rate.

0 The pressure is limited by the reactor pressure chamber, which can withstand up
to 6MPa. Most experiments were conducted at 0, 0.3, 1.0, or 3.1 MPa.

9 The reactant ﬂow rates are up to 300 cc(STP)/min, which is limited by the mass
ﬂow controller. This ﬂow rate is also required to eliminate the transport resistance
to get true kinetics.

O The sample sizes are limited by the size of the quartz liner. For Saran char,
maximum sample size is 0.35 to 0.45 g; for coal char and cherry pit char, it is
about 0.7 to 0.9g, depending on the sample density.

0 Steam ﬂow rate was ﬁxed at 120 cc/min(40 mole %) for most experiments.

 

30

2-4-2. Experimental Procedures

The experimental scheme is given in Figure 2-14.

 

 

 

    

E H2 Argon ,
m I —\ ' i
g, :’ H20 ’: 1
’ -' 1 -l

’5 ‘ :’ ‘1‘: . 0
E . . Temperature 1f 3
1... :‘i , 9)
g-o :; .--o
C a“. g
8 0
c =
c- 1 ‘1
o , 1‘ ‘

L____V,.___c. . _._ . - .-... -. .- . . _ ,__. _ _ w __.__.__}5}:L__ _.. _____ __\!-_._, ._‘

Purge Gasiﬁcatior. Purge Outgas

Figure 2-14 Experimental scheme
A. Preparation of Mass Spectrometer
Before gasiﬁcation and TPD, we need to prepare the mass spectrometer. After
opening the capillary inlet valve, the vacuum chamber was pressurized with purge argon
(from the bypass line). When the background of the mass spectrometer was stabilized, we
collected the background data (usually 5 scans), and then conducted the calibration for

H2(D2),CO, CO2, and CH4 (or CD4) to get the real time responses.

B. Gasiﬁcation [‘1

After the high pressure reactor was purged 20~30 minutes to expel residual air,
the heating tape and furnace power was turned on. At same time, the reactor was
pressurized with argon. The bypass line was pressurized with reactant gas a little later,
and the steam trap was ﬁlled with crushed ice (if the reactant including steam). Once the

pressure and temperature of the entire system were stabilized, mass spectrometer data

‘ Most 7000 series gasiﬁcation experiments were done by Lussier

31

collection was begun. After ﬁve or more scans, a step change in reactor feed from purge
argon to reactant gas was made to begin the gasiﬁcation by switching a 4-way valve that
interchanges the ﬂow paths of these two streams. Pressure and temperature were
monitored and required small periodic adjustments of regulator and temperature
controller during the course of experiments. After a planned length of time, the 4-way
valve position was switched back to create a step change in reactor feed composition
from reactant gas to purge argon. After the reactor efﬂuent composition (monitored by
mass spectrometer) had stabilized, the data collection was stopped and the vacuum
chamber was closed. The reactor and bypass lines were then depressurized and furnace
and heating tapes power was turned off. The reacted sample was allowed to cool down
for about 10 hours (usually overnight), and maintained in an argon purge of 2~5 cclmin

for the purpose of protection.

C. Temperature Programmed Desorption(TPD)

 

After samples were removed from the gasiﬁcation reactor and weighed, they were
loaded into the high temperature ceramic reactor quickly to limit the exposure time in air.
Short exposure of samples did not affect the accuracy of hydrogen measurement, as
shown in Section 2-6-3.

Aﬁer samples were loaded into the ceramic reactor, we added a small amount of
tabular alumina (-14~ +28 mesh) on the top of the sample to prevent sample spillage into
the annular portion of the reactor. The alumina particle size was chosen to ensure that the
alumina was too big to enter the annular area, but small enough so as not to allow char to

pass through it. The reactor was then assembled and sealed by tightening ﬁttings. After

 

32

loading the whole reactor into the high temperature furnace, the reactor was purged with
argon for 30 minutes. When the calibration (H2 or C0, C02, CH4) was ﬁnished, the
furnace was turned on and the mass data collection begun. A linearly heating rate of 5
and 10 °C/min were used in most TPD experiments. The highest temperature the furnace
can reach is 1500 °C. This temperature is necessary for hydrogen desorption. Lower ﬂow
rates broaden the TPD proﬁle, while higher ﬂow rates dilute H2 (CO and CO2)
concentrations in the efﬂuent gas. A ﬂow rate of 30 cm3 (STP)/min was used to balance
the two factors. The desorbing species were continuously monitored for the entire TPD
process. After the temperature reached 1500 °C, a one hour holding time was used for

further possible desorption. After that, the furnace power was shut off automatically by
the temperature program controller.

D. Surface Structure Characterization

 

A Micromeritics Pulse Chemisorb 2700 was used to measure the BET surface
area. The total surface areas of char were measured by nitrogen adsorption. Twenty to
ﬁfty mg of char were loaded into a quartz sample tube, which was then sealed to the
Chemisorb and heated to 150 °C for more than 30 min to drive off possible weakly
adsorbed species (mostly water, which makes the instrument unstable). After ﬁnishing
the calibration, a mixed gas, which contained 5%, 10% and 18.75% N2 in helium at a
total ﬂow rate of 16 cc/min, was passed over the sample. The efﬂuent gas composition
was measured for nitrogen concentration by detector, and the desorption and adsorption

proﬁles and peak area were recorded.

33
2-5. Data Analysis

2-5—1. Mass Spectrometer Calibration

Calibration of the mass spectrometer was performed by scanning two AGA
certiﬁed standard multi component gas mixtures. The ﬁrst contains 2.00% CO, 2.03%
CO2, 2.00% CH4, and balance argon, while the second contains 2.05% CO, 2.03% C02,
2.01% CH4, and balance hydrogen. Five scans of carrier gas containing no key species
of interest were taken to obtain background levels, averaged, and then subtracted from
the average of ﬁve scans of calibration gas to obtain actual peak values. The mole
fractions of key species in the calibration gas were then divided by the corrected mass
spectrometer peak values to obtain actual responses. Purge gas composition and
calibration carrier gas composition must both be identical and must match the reactant
gas composition to be used for gasiﬁcation because the response of key species is a
function of can'ier gas composition and is about three times higher in hydrogen than in

argon.

2-5-2. General Data Deconvolution

The most difﬁcult work in mass spectrometer use lies in data deconvolution.
Fragmentation, double ionization, and natural isotopes all may cause peak overlap in the
mass spectrum. In our system, only M84 (krypton), M44(CO2) and M40 (argon) are
simple peaks with no overlapping. The main purpose of our analysis is to detect the CO,
H2, CH4 and CO2. That mean only CO2 does not involve peak overlap. Table 2-3 gives

the possible fragmentation of these species.

 

34

Table 2-3 Possible Fragmentation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Channel

M2 H2

M3 HD

M4 D2

M14 N(N2) CH2(CH4) CD(CD4)
M15 CH3

M16 CH4 0 0(H20) O(CO2) O(CO) 0(02)
M17 CH3D CD2H OH 01F
M18 H2O CD3

M19 CD2H DHO

M20 CD4 D20

M28 C0 N2 CO’

M32 02

M44 C02

 

 

 

 

 

 

 

 

 

Table 2-4 Fragmentation Patterns in Argon Carrier“I

Channel CO2 CO CD. CH4 D20 H2O D2
M2

M3

M

 

 

‘ These experiments were done by Lussier.

35

A. Time Shift in MASS Data

All 12 channels of the mass spectrometer were scanned sequentially. That causes

another problem, time shift.

4.0E-7
3.5E-7 —
3.0E-7 —
2.5E-7 ~
ZOE-7 —
1.5E-7

 

  
 

 

 
 
 
  

—<->— M28
_a— M14*210
—+— Sub

 

 

 

 

5.0E-8 '
OOE‘I'OQMIlIIJ' 1m1111Lm111111l11L1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
. Tim¢<hf> w, ..
.~Figurc2-15-Bcf9rjetimeShifty. » .w . , , N

For a one second dwell time, it takes 2 minutes to scan all 12 Channels. That means
the time difference between channel 1 and channel 12 is up to 2 minutes. But in the data
ﬁle, we have to put all 12 channels into the same time array. Ifwe set channel 1 and 12 to
same mass, the curve will shift about 2 minute. Ifwe only use one channel to deconvolute
the data, it does not matter. But when we need to deconvolute the overlapped peak (C0
and N2 of M28), we have to subtract one channel ﬁ'om another. This problem appears and
will produce an artiﬁcial peak due to the time shift (Figure 2-15). To eliminate the time
shift effect, we need to shift the time axis for a special time for each channel. In that way,

we can get the exact overlapped curves (Figure 2-16).

36

4.0E-7
3.5E-7 —
3.0E-7 -
2.5E-7 *—
2.0E-7 *-
1.5E-7 *- '
—*— M28
1.0E-7 --><-— M14*21(J
5.0E-8 1

00E+0 Li1l11L11111111111l11111111111111
O 0.2 0.4 0.6 0.8 1 1.2 1.4

_ w , 4 M ,_ .. N . Time(hr).
' _ * . Figur¢216

 

    
 

 

 

 

 

 

  

shxt’tcorrectmn
But the subtraction still is impossible, because we do not have corresponding
points after the time shift (Figure 2-16). Before we can subtract spectra, we have to ﬁt the
data by a ﬁJnction.
First order ﬁtting seem does not work around the peaks (the data values change
rapidly. The second order ﬁtting (Y=a+bX+cX2) works much better.

From continuous three points, we can get the ﬁtting parameters:

 

b: (r: -y3)(x1’ -x2’)-(yr -y2)(x2’ ~15?)
(x2 ‘x3)(x12-x22)-(x1“x2)(x22-x32)
c:(y1—y2)—b(xr—x2) .
3‘12“”:

 

a = y1 —bxl —cx,2
The ﬁtted results are given in Figure 2-17. We can see the second order ﬁt is good
enough. From Y=a+bX+cX2, we can get corresponding points using the shifted time.
Now, we can subtract any channel. Figure 2-17 gives the results. We can see that the

artiﬁcial peak almost disappears.

 

37

 

   
   
     
 
   

 

4.0E-7

3.0E-7 L

2057 T «0.- shifted M14X210
+ M28
—x— M14'210

1.0E-7 —V— M28-shifted M14x210

 

 

 

 

 

0.0E+0 v L?— ‘7 v . 1:7 (7 v v ' -
0 0.3 0.6 0.9 1.2
Time(hour)

' . ’ Figure 12-1 7* After the time shift, artiﬁcial, peak neienig‘er' exists

B. Species Deconvolution

Methane: M=15 for CH4 (to avoid confusion with fragmentation of species containing

oxygen).

 

H _ (M15 — M15mgmd) x Flowrate x Reponse ( ml )

‘ Sample weight min. gram

H2/D2:M=4 for D2, M=3 for HD, and M=2 for H2 (after subtraction of contribution

from D2 fragmentation.

H (M1— M2 background) x Flowrate x Reponse [ ml ]
2 — Sample weight min. gram

 

C02: C02 is the easiest among all deconvolutions. After subtract the background, we

simply multiple the mass data by the response factor and carrier gas ﬂow rate.

co,

 

_ (M44 — M44bwk8mmd) x Flowrate x Reponse ( m] j
— Sample weight min. gram

C0: The M28 peak includes C0, fragmentation of C02 and N2. After time shift of C02
and M14, we can subtract contribution from C02 and N2 fragmentation, and then

subtract the background. At last we can get the real C0 ﬂow rate:

38

_ (M28— 0.15 x M44' -210x M14‘) x Flowratex Reponse [ ml )
Sample weight

CO

 

min. gram

Note: M44* and M14* are the data which are processed using the time shift technique to

avoid artiﬁcial peak.

2-5-3. CD4 and D20 Deconvolution
CD4 and D20 deconvolution is the most complex of all deconvolutions. The
principle is

=> P18 from H20, CD3, DO
:> P20 from D20, CD3
:> P17 from OH, CD2H, CDH3

From the fragmentation experiments, we know most P17 comes from OH. After
using the H20 fragmentation data, we can subtract the H20 part from P18, so that only
CD4 and D20 contributes to P18 and P20. We try to use the two dimensional matrix to
deconvolue the CD4. First we need to ﬁnd background for P17, P18, P20.

0 P20back: If there no D2 and D20 present, the P20 chamber pressure should be at the
background, it is 4.5x10'll torr.

O P17back: In D2 fragmentation experiments, after a long time purging the chamber, we
believe that no more 0H exists in the chamber. The P17 can thus be used as
background, again 4.5x 10’10 torr.

o P18back: In D20 fragmentation experiments, P18 is only from D0 (not from H2O). So

(P20-P20back)*KD2o=(Pl 8-P1 Shack).=P20=P l 8/KD2o-Pl 8/KD20+P20back
This is a linear equation and after ﬁtting the experiment data, we can get Pl 81min

1.8x10’ll torr.

39

Because of the slow displacement of H20 by D20 in the vacuum chamber, we
need to take into account the H20 fraction in P18. We have to subtract the H20 from P18
before deconvulotion. From H2O fragmentation experiments, we can get KH20
(fragmentation factor) by

(P18'P1 8back)l'lKHZO=(P17'1)1 7back)-
The value of K1120 is 0.163. At last, we can get the “real” P20 and P18 (only from

D20 and CH4) from

P20r=P20-P20 back ; P18r=P18-(P17-P1 7back)/KH2O

Now, we can deconvolute the CD4 and D20. If we set
X is fraction of M20 from D20, K1 is the M18 fragment of D20
Y is fraction of M20 from CD4, K2 is the M18 fragment of CD4
We have : X+Y=P20r and X*K1+Y*K2=P18r. From these two equation we can get:
X=[P20r-P18r/K2]/[1-k1/k2] and Y=[P20r-P18r/Kl]/[l-k2/k1]
The procedure of getting Kr and K2 is given as follows. In CD4 fragmentation
experiment, X=0, So K2=[P18-P18 back]/[P20—P20 back].

The value is given in Table 2-5.
Table 2-5 The K2 Change with Carrier Gas

 

Ix, 0.71 0.53 0.38
[Carrier gas(Ar+D2) 100%Ar 66%Ar 33%Ar ﬂ

 

In D20 fragmentation experiment, Y=0, So
KI=IP18'P18 backI/[PZO'PZOback]
K1 is 0.17 in argon carrier gas. Thus we can deconvolute CD4 and D20 in D20/D2

gasiﬁcation.

 

 

40

2-6. Carbon Sample Characterization

2-6-1. Outgas of As-prepared Saran and Coal Char

A. Outgas of Saran Char

 

The results of outgassing as-prepared Saran char are given in Figure 2-18 to 2-21.
Peaks for CO2 desorption in the range of 250°C ~1000°C, for C0 in the range of
500-900 °C, and for H2 from 700°C up to 1500 °C were observed.

We can ﬁnd from these ﬁgures that a temperature of 1500 °C is enough to
removes essentially all adsorbed species from the surface. If only removing surface
oxygen groups, 1050°C is enough. Above this temperature, all CO and CO2 peaks
disappear. For H2, a higher temperature, over 1400 °C, has to be used to get a ﬂat peak
tail.

To verify the above ﬁnding ,we determined the total amount of H2, C0, and CO2
desorbed from as prepared Saran char by integrating the desorption curves over the
course of outgassing. A comparison of the quantity desorbed with that calculated from
the char hydrogen and oxygen contents as determined by ultimate analysis is given in
Table 2-6. The agreement is surprisingly good, and indicates that heating to 1500°C
effectively removes essentially all hydrogen and oxygen from the Saran char. Only the
sum of CO and C02 can be compared, as both contain oxygen. This result is important in
light of our goal of measuring the quantity of hydrogen on char surface following

gasiﬁcation and correlating the quantity to gasiﬁcation rate.

41
Table 2-6 Comparison of Element Analysis and TPD

from CO from
H2 from Outgas 62 cc/ 0.5x21 cc/ 6.5 cc/

W % from 0.55% 0.75% 0.46%
Ultimate 0.5% 1.3%

 

1.6
1.4 -
1.2 —
1.0 —
0.8

0.65
0.4 — + H2
0.2 —
0.0

 

 

cm3/min. g(initial)

 

 

 

 

 

 

0 300 600 900 1200 1500
, , Temperature(.°C degree) .. .
Franz-momma of as-preparediSaranéehar , '

0.50

 

   
   

 

0.40 -

 

—-V—— C021 CC/g

 

 

0.30 ~

0.20 -

Cm3lmin. g(initial)

0.10 m

 

 

 

0 300 600 900 1200 1500
_ ,, .. “Temperaturefcl _ _.
Figure»?1901113118ofaeépireparedisaranehm “ ' ’

cm3lmin. g(initial)

cclming

0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00

0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00

42

 

 

 

 

 

 

 

 

 

—V— C02 6.5 cclg
0 300 600 900 1200 1 500
Temperature (C degree)

 

 

 

 

 

0 300 600 900 1200 1 500

Temperature (°C)

  ' FigureZ-Zl Low rate TPD proﬁie, heating rate 2 °C/mrn 2' g V

purge argon—60an W=0~. 65 Sg (Exp-015)

43
It should be noted that the baseline of M28 is not ﬂat, probably because adsorbed

N2 from the reactor also contributes to the M28 signal. To get the real CO signal, we need

to subtract the N2 component using the M14 data.

B. Coal Char TPD Proﬁles

Coal char TPD proﬁle (Figure 2-22) is different from Saran char. For C0, two
clear peaks can be found. The amount desorbed is much greater than Saran char (10
times). This is an indication that coal char contains more oxygen than Saran char. The ﬁrst
peak position is at about the same position as Saran char. The second peak is situated at

1200-1400 °C, and seems like the caved-in oxygen proposed by Yang‘“.

_b
O)

 

  
    
   

 

—9— H2 59cc/g
—9— CO 52 CC/g
——1-— 002 9 00/9

.-‘
N
T

 

 

 

.0
rs

 

Products evolution rate NIL/min
o
co

 

.0
o

900 1200 1 500

, .5 temperature (°C.)- V . .w ,.

FrgurchzzCoalchmTPDproﬁlehea ratelOC/mm I

 

2-6-2. Saran Char Annealing

The pretreatment temperature of char has signiﬁcant effects on gasiﬁcation rate.
Many previous researchers have done lots of work on this issue. However, very few
researchers have used a temperature higher than 1200°C. Our purpose is to stabilize the

char structure by using high temperature annealing ( up to 1500°C and 10 hr. ).

44

Saran char pretreatment experiments were done at the following conditions:

0 Saran char 0.5-1 gram

9

Argon gas ﬂow rate 30-120 cc/min

0

Temperature range 900-1500 °C

0

Heating rate 3°C/min or 5°C/min

A. The Effects of Heating Rate and Annealing Temperature on BET

950
900 -
350 _ 823
800 —
750 —
700 Heating rate 3 Klmin. 664
650 _ -%Er-BET as

600 l l l l l
900 1 000 1 1 00 1200 1 300 1400 1 500

FigureZ23BETareachangeswrthpretreatmenttemperature(t1ngrate3C/mm)

 

775

 

BET Area (square meter)

 

 

 

 

950
900 e 866 350

850 _ \
800 — 79 >

750 - . . '
Heating rate 5 Clmrn.

700 *- —@— BET
650 —

600 l l l l l
900 1000 1 1 00 1200 1 300 1400 1 500

Fraum224BETareac eswrth pretreatmenttemperature(5C/nnn h

 

 

 

 

 

BET Area (square meter)

 

 

 

are)

   

45

The pretreatment experiments are performed at several diﬂ'erent temperatures and
two heating rates ( 3°C/min and 5°C/min ). At ﬁrst, the purpose of changing the rate was
to reduce experiment time; however, when comparing the experiment data we found the
heating rate has some effect on the BET surface area. Figures 2-23 and 2-24 give the
BET area changing with pretreatment temperature at two heating rates. Although both
ﬁgures have the same trend with pretreatment temperature, the change is not very
signiﬁcant. For 3°C /min heating rate, BET area decreases 28% after a 1450 °C heat
treatment. For 5°C heating rate, the change is only 14% after 1500 °C treatment. From
the N2 adsorption and desorption curves in BET measurements, we can see a trend that as
pretreatment temperature increases, the adsorption process becomes slower. This suggests
that the pore structure of Saran char must have changed. With the increase of
pretreatment temperature, the changes become more and more signiﬁcant. Unfortunately
the structure of Saran char is too weak to stand the high pressure in mercury porosirnetry
measurements, so we cannot directly get the distribution of micro pore diameters to

further verify the above deductions.

46
The BET surface area is shown as a function of annealing time in Figure 2-25 for

Saran char. With the increase in annealing time, the BET surface area decreases quickly,
then stabilizes after 6 hours. In N2 adsorption and desorption (Figure 2-26 and Figure 2-
27), we can see that the relative peak intensity and adsorption times before annealing and
after annealing are signiﬁcantly different. After annealing, the longer adsorption time and

lower peak intensity suggests that the number of micro pores must have increased and the

total surface area have decreased.

47

50

 

40 L desorption
30 —

20H

adso tion
10 - ”D

0 1/1—m L 1
0 1 2 3 4 5 6 7 8

p p _ ﬂ ., ‘ Adsorption (or Desorption )time (minutes) p ..
7Figure2-26N2proﬁlein BET for:fSaran CharPn2=0‘.05viatrnx(samplef\fveight=0.048g).

 

 

 

Desorption(adsorption)rate cclmin

50

 

40 i
30 — desorption

20—

adso tion
10 — ”D

0 m 1 1 1 __
0 2 4 6 8 10 12

r .. _. ,. __.,.Adsorption(orDesorption)time(minUteS)_. _ , .1 _
§ 1_rignrejz-zzm'z";amazement:'foranneaieoisaran cheriPh2=095anew-0.0483) _

 

 

 

Desorption(adsorption)rate cclmin

48
Annealing also changes the physical structure of Saran char. The difference is that

when the annealing reaches a certain time (in this experiment it is about 6 hours), the
BET surface area does not change further. From these phenomena, we can conclude that
high temperature annealing makes the char reach a relatively stable structure. It is highly

possible that annealing is a process of big pores collapsing and micro pores opening.

C. Conclusions

Both high temperature treating and annealing have effects on the Saran char's
physical structure. With pretreatment temperature increasing, the BET area decreases.
With the annealing time increase, the BET area decreases at ﬁrst and then becomes stable
(after 6 hours for Saran char). The annealing signiﬁcantly changes Saran char BET area
(~34%). High temperature treatment also is a process of further pyrolysis, which
contributes to part of the surface change. We can speculate that big pores collapse and
micro pores open during the annealing process.

After annealing, we can ﬁnd that the Saran char near the frit (gas inlet) is a little
darker than the normal one. It seems that the impurities in argon gas do react with the
char sample. However, from simple estimation, the largest possible amount gasiﬁed with
the active gas in argon is no more than 0.0013g (lOOcc/min, argon purity 99.999%, 20
hr.). Total weight loss is at least 0.1 g in the high temperature pretreatment experiment,

so, the weight loss is mainly due to pyrolysis.

2-6—3. Annealed Saran Char Exposure in Air

When outgassed Saran char is exposed in air, it will adsorb impurities from the air

continuously. The amount of adsorbed impurities depends on exposure time. For

49
2-6-3. Annealed Saran Char Exposure in Air

When outgassed Saran char is exposed in air, it will adsorb impurities Item the air
continuously. The amount of adsorbed impurities depends on exposure time. For
determining the extent of this effect, three experiments were designed and three different
exposure times were used. Table 2-7, Table 2-8 and Figure 28-34 gives the

experimental results.

Table 2-7 Impurities ﬁ'om Exposure

 

§nes.ie-ilr.ta!se..-...-§.ﬂ..z. .............. 1.991.192) ........ 2-99.2. ........
911122592.mien---“-..5..2.-.!-9.c./.s--.-----5..ll-ésﬂs.-..---i-4:.1..cs!s----
£§Pﬂ§99~§ﬂz.---.;--.5.12-92/18--..."1.2.9.339!!!"...-§-7:.2..°..°./8.-.-

xpose 2 days 5 55.4 cc/gr E 83.3 cc]; 5 68.7 cc/g_

 

Table 2-8 The Peak Position (°C) Changes with Exposure Time

 

 

 

 

 

 

 

 

 

 

 

 

.-..1.’sak.r>9.§it.i99 ....... H2€0(N2)<302H20 ......
Exrose2mm1280660 ..... '---.?.?9 ........ I}! 929915.--
EXPOROShrl210660 ..... 910N026“
Expose2days i1300 5 800 5 210&720 5 210
0.35 .
' o CO+N217.30clg . .
025 _ —v—— H22.1oclg O
a) . O
E 0.20; . Q
g 0.15
0.10
0.05
0.00
100 300 500 700 900 1100 1300 1500

Temperature( °C )
, Figure 2-28 TPD proﬁle of annealed Saran char following 2 min. hour exposure.

50

 

 

 

 

 

 

 

 

0.6
A COZ(7.2chg)
0.5 *- +— H2 4.3ch 000
o CO(N2) 26.3ch9 @9000” ‘o
C? 0.4 ~ Q Q.
.E 0' 0Q
E 0 3 — o" o
(‘0 O Q
g 0 2 _ 095 GO
. _, . Go
‘A -e
0.1 - ow AAA ‘
0 0 ; (3.0.0. .._egg-sagaozejs-33.",‘, , , , _. . . . . . . . . .. t. .. . .-..~. .1
0 300 600 900 1500

Temperature( °C)
Figure 2-29 TPD proﬁle of annealed Saran char following 0.5 hour exposure.

 

 

 

 

  

 

 

3'0 0 ‘ v C02(68.7cclg)
: 2 5 _ ox. o CO(N2) 83.3oclg
m ' ' Q. b --x~ H2 55.4ch9
'E a" "0“ O—

a 2.0 — .

a?

.5 1.5 — 0

g .

°° 1.0 —

E

0 0.5 —

0.0 ‘." I I i ' ‘ ' ‘ - ' .- "a!!!
0 300 600 900 1200 1500
Temperature(°C)

Figure 2-30 TPD proﬁle of annealed Saran char following 2 days exposure.

51

 

 
 
   

 

F —x— H2 2.1 cc/gz min.
— + H2 4.3ch 0.511
—e— H2 55.4cclg 2 day

  

 

CCImin.g
OOPPPr‘r‘r‘f-‘N

 

bbhmmowembb

 

0 ‘ 1200 150

 

 

 

3
'- —x— CO(N2)17.3cc/g)2m
2.5 ~ “ '1 _a— CO(N2)26.3ec/g 0.511
s - —e— CO(N2)83.3cc/92d
' kg

 

 

 

 

 

 

 

0 V 300 600 900 1200 1 500

Temperature (°C)

Figure2-32Uilifiis of CO or N2 change With exposure time char m air(anneaied Saran) ,

 

52

 

 

1,6 — ’ —x— 002 =4.1cclg 2m
—v— 002(720clg) 0.5 h
+ 002 (68.7cc/g) 2d

    

 

 

 

CCImin.g
o
00

 

 

 

O 300 600 900 1200 1 500

,. w. w. .1 Temperature(."Q ., . , . .
Y "F'iig‘ur‘efzrﬁiUptdkéwf CozchangesmthestureﬂmemalI(annealedSaranchar) i ’

 

2.5E-7

 

2.0E-7

1.5E-7

 

1.0E-7

5.0E-8

 

 

Raw data (chamber pressure)

 

ODE-+0

 

0 300 600 900 1200 1 500

iFi'g’ili‘IeIiZ-134UptakesOfH20ChﬁngesWith expdsur’estmOfahneeledisatahchatin air ,.

53

From the above experiments, it is clear that even very short exposure (of course,
the air will enter the reactor when exposing the sample to air, and the sample could
absorb impurities from this part of air), the changes in absorbed H2, CO and C02 are
detectable. The follow conclusions could be made.

0 Dissociatively adsorbed H2 (emitted above 1000 °C) probably comes from
decomposition of the water in air on the char surface. All exposures can increase the
quantity desorbed, but the 2 minutes and 0.5 hr. exposure is tolerable, since the M2
signal only is about three time of the background.

0 For M28, the 2 minutes and 0.5 hr. exposure are not much different from no
exposure, but the two days exposure will produce several peaks. One is at 200 °C and
the other are at about 1000 °C. After subtracting N2, the ﬁrst peak almost disappears.
So, the ﬁrst M28 peak must come from adsorbed N2. The second peak is from oxygen
adsorbed from air.

0 For C02 (M44), no obvious peak can be found after a short time exposure (2 minutes
and 0.5 hr.). But two days exposure does increase the M44 peak. The adsorbed C02
(or 02 ) partially contributes to the M28 peak.

. For H20 (M18), a short exposure to air will not induce an obvious peak at
200~250°C. The M18 peak (compared with no exposure experiment) is caused by air
entering into the reactor when exposing the sample. But, for the two days exposure,

the M18 (H20) peak is obvious.

CHAPTER 3

THEORETICAL ALCULATIONS

3-1. Shift Reaction Equilibrium

The composition of gas produced from carbon gasiﬁcation is affected by the

CO+H20 shift reaction. So, we need to investigate it ﬁrst.

55

3-1-1. Equilibrium Constant Changes with Temperature

 

CO+HZO<——>H2 +CO2
P P

Ke= ”2 C02 Ln(Ke)=-—A——(—}—
Puzo co RT

According to Van’t Hoff’ s equation:

T
LnKT = LnK298.15 + £9815

AH m 2(T) .7
RT

T

T
AHmtT) =AH,..(T.>+ (98,,deT

At 298.15 °K the thermal data are

(To=298.15K)

Table 3-1 Thermodynamic Data

 

 

 

[ CO H20 C02 H2 CO+H20=COz+H2 |
I AH (KJ/mol) 410.53 -24l.826 -393.51 0 41154 I
“AG (KJ/mol) 437.168 -228.582 -394.373 0 -28.623 |

 

 

 

 

 

 

From the data of Table 3—1, we can get Ke (298.15)= 103469 by using equation

E3-1. Heat capacity can be calculated by using C P = a + bT+ C T2 . The coefﬁcients are

listed in Table 3-2.

Table 3-2 Heat Capacity Coefﬁcients

 

 

 

 

 

 

a b c
CO 26.86 0.00697 -8.20E-07
C02 26 0.0435 -1 .48E-05
H2 29.07 -0.00084 2.01 E-06
H20 30.36 0.00961 1.18E-06
A a, b, c -2.15 0.026084 -1.32E-05

 

 

56

2 __ 2 3 - 3
Maﬁa) = AHO (T)+Aa(T—298.15)+ Ab(T 298.15 )+ Ac(T 298.15 )

 

7X" 2 3
Au T Ab Ac 1 0
LnKe(T)—-LnKe(7;))+ R LnT+2R(T-T)+6R (T—T)+R[—AHm(T)

1 1
+Aa7}, + AbT02 + AcT3]*[———]
T 73

This change in Ke with temperature is given in Figure 3-1.

 

 

12
10~
3-
i?
x 6.
E
4
2,-
0 L m

 

 

 

400 500 600 700 800 900 1000

Temperature (°C)
Figure 3-1 Equilibrium constant changes with temperature

3-1-2. Simulation of Steam Gasiﬁcation (with Shift Reaction)

We assume that only CO comes from carbon surface reaction and that C02 only
comes from the shift reaction, which is supported by other researchers (Hermann,
1986”“). The formation rate of CO is C1, and Y is the conversion of steam (H20) from
surface gasiﬁcation reaction, X is the formation rate of C02, and W0 is H20 ﬂow rate,

and H0 is hydrogen ﬂow rate.

 

 

C + H 20 (—-> C0 + H 2
Before gasiﬁcation W0 Ho
Aﬁer gasiﬁcation W0(1 -Y) WOY HO+W0Y
C0 + H20 (———) CO2 + H2
After shiﬁ reaction WOY -X w0(1-Y)-X X H0+W0Y+X

 

28 Gunter Hermann and Klaus J. Huttinger, carbon Vol. 24, No. 6. PP 705-713,]996

57

If we assume the shift reaction reaches equilibrium, then we can solve for the

reactant gas composition.

__PH,PCO2 _ (H0+W0Y+X)XX
Pﬂzol’co (WO(l—Y)—X)X(WOY-X)

 

 

Ke

Solving this equation we get

 

H. + cho + WOY i\/(H, + cho +W0Y)2 -4(Ke-1)KCWO(1- Y)Y
2(Ke -1)

 

X = E3-4

At 850 °C, Kc=0.882 and only one root is possible. At 725 °C, Kc=1.40 and there
are two positive roots, but only one is meaningful. Total outlet mole ﬂow (dry base, not
including water) is:

T = 2W0Y + H0 + X
The outlet hydrogen concentration is

H,%= x100%

 

WOY+H0+X
T

The CO concentration is

wov-x

C0% = x 100%

 

The C02 Concentration is C02 % = %x 100%

We use the following parameters to simulate the process:
W0=120 cc/min.=0.054 mol/min.
Ho=0~120 cc/min.=0~0.054 mol/min.
In a two hour steam gasiﬁcation, the maximum carbon conversion is about 10% at

our experimental conditions.

Th f ethecarbonconsum t'o t-M-ZRE 05 l/ '
ereor, p1nrae-12.01x120—. - mo min

The corresponding to the steam conversion is m = 0.0155 = 1.56%

40/ 22400

58

The corresponding to the steam conversion is M = 0.0155 = 1.56%

40 / 22400
So, in the following simulation we set the steam conversion Y<2%. For 850 °C
gasiﬁcation, the simulation results are given in Figure 3-2~3-5; For 725 °C gasiﬁcation,
the results are given in Figure 3-5~ Figure 3-9.

1.2

1.0 — ::- $91. /

0.8 —
0.6 r- /
0.4 —

02 P
00 1 1 l l l A [All 1 Lilli L1

0 0.5 l 1.5 2 0 0.5 l 1.5 2

.. . _.H20°°“"e'5i°_‘_‘. % _ __ .. _. _.._HZQ°°‘_‘V?'5i°“ %. . _
4 , ﬁgure 3-2 Hmec.‘H2‘=i-20,H20=120mvminx 55mm. 3—3‘Ta8509C.‘HF120.'H507=120m11min

 

 

CO&C02

 

 

 

 

 

 

 

 

 

35
30 —

25 -
20 M0096

 

15 ~
l0 —
5 _

CO&C02

llllTlTll

 

 

 

 

  

 

 

 

. m- 1 hm” i
0 0.5 l 1.5 2 0 0.5 l 1.5 2

,1 _.1.411209°nversi99% V _ w M ,.,..H29<=°nv¢rsi9n.% . .
res-4 r=ssocnz=sazo=uomumm ‘ 35T=350CH2=0320=129WM

 

 

   

 

 

1.2 3.0
l.0 2.5
0.8 2.0
0.6 1.5
0.4 1.0
0.2 0.5

0.0 0.0
0 0.5 l 1.5 2 0 0.5 l 1.5 2

A-....329.99“,."9F5i99..%.., . . . _,..,,.,.,.....HZQ¢°‘.‘V¢’..5199%, .. .
r nm34T=7256H2=120H20=120mem mma7r=7zscaz=izoazo=xzommn

CO&C02

 

 

 

 

   

 

 

59

 

 

 

 

 

 

 

 

 

 

20 35
18 —
l6 __ 30
8 12 P 20 _ ........ 00%
10 ~— — C0296
g 3 _ W 00% 15 _
o — 002%
6 " 10 r
4 ..
2 — 5 ”
0 l w--mm 1‘ _._.. 0 l __ l 2.... l .... am":
0 0.5 l 1.5 2 0 0.5 l 1.5 2
H2O conversion % H2O conversion %

Fiﬁitﬁ$ST=725,°”C;H2#5.H2'0=‘I:20mlfm1’ 'n'-‘ TQLFigure3-i9 T=725.°C.H2=0;1H20=120m11min"V

Apparently, if the reactant does not contain H2 and the shift reaction reaches

equilibrium, the main product should be CO2.

3—1-3. Comparing with Experiment Data

We still assume only CO comes from carbon surface. Dividing the C0+CO2
evolution rate by the total H2O input, we can get the instantaneous conversion (Y) of
steam. Therefore we can calculate the equilibrium CO&CO2 composition using E3 -4.
Figure 3-10 gives the comparison. We can see that the shift reaction is far from

equilibrium in our Saran char steam gasiﬁcation process.

 

 

 

 

 

 

 

 

 

0.30
E 0.25 — - - - Equilibrium-CO
3 Equilibrium—C02
E 0.20 _ --O-- EXP-CO
:1? 0.15 _ _a_ EXP-C02
6
3 0.10
'8
E 0.05 'q

0.00 '

1o 30 50 70 90 110

Time (minutes)
‘ I Figure 3-10 Comparison of expenment CO and CO2 With the
. , equilibrium CO & CO2 (E9635 steam gastﬁcatlon)

60

3-2. Desorption Activation Energy of Hydrogen

3-2-1. Average Desorption Activation Energy

The rate of desorption from a uniform surface can be written as:

N(t) = —%f = AnO "e43“RT E3-5

Where An is the rate constant (frequency factor)
E is the activation energy of desorption
0 is the surface coverage (mole/m2)
N(t) is desorption rate (mole/secmz)
n is the order of desorption reaction

The situation is simpliﬁed by experimentally imposing a uniform rate of heating
on the sample. T=T0+Bt and assume E is independent of surface coverage 0. For ﬁrst

order desorption n=1, after differentiating the above equation we can get:

dN(t) = A] Ege_E/RT + (A0)e-E/RT(_ E)(_ _Ez E3'6
dt dt R T

When the desorption rate reaches a maximum (peak position), we have:

M = 0 E3-7
dt

Combining equation E3-5, E3-6, and E3-7 we can obtain an equation due to Redheadm]:

E2 =§—'e’E’/RT"' E3-8
RTm ,6

 

Where Tm is the temperature at which the desorption rate reaches maximum, [3 is
the heating rate, A; is rate constant of desorption, and E is the desorption activation

energy for a uniform surface and uniform adsorbed species.

 

29 P. A. Redhead ,Vacuum, 12, 203 (1962)

61

This equation shows that T... is independent of coverage for ﬁrst order desorption
with constant B. So, the desorption activation energy can be found directly from a

measurement of Tm, provided a value of A1 is assumed.

The relation between E and Tm is very nearly linear. In the range of
10'3 > A‘— > 108 ,
16

E 3-8 can be approximated by following equation ( Redhead”):

E Ale
= 111——
RT

m

 

- 3.64 E3-9

In the range of 500~1200 °C, the approximation can be made more accurate by

using the following equation:

 

 

—E— = 111% — 3.1 133-10
arm [3

280 1
—e— 153-10

240 _ —i— E3—8

 

 

‘6 + E3—9(Redhead)
5. zoo _
x
L” 150 H
120 l J l l l I

 

 

500 600 700 800 900 1000 1100 1200

Temperature (°C)

62

A comparison is given in Figure 3-11. It can be seen that the lines of E3-8 and
E3-l0 almost overlap. The activation energy can be determined without assuming a value
of the rate constant A 1 by varying the heating rate B and plotting ln(Tm) against ln(B); E

can then be obtained from the relation:

_E_ + 2 = dlnﬂ
RT dlnTlm

m

E3-11

 

The rate constant A. can then be determined by substituting E into E3-8. If only

two heating rates are used, E can be calculated by the following equation:

£2- 5;:
ln( ) 2111(T)

ﬂ' ”' R 133-12

 

Unfortunately, for reasonable accuracy, B must be varied by at least two orders of
magnitude (Redheadlzgl). For our experimental conditions, this variation cannot be
reached due to the limitations of the furnace. We thus can not use this method reliably to
get the activation energy.

In the same way, we can get the similar equation for second order desorption

E = 26,.A2 8-0m, = 90A2 e—m... 153-13

RTE. ﬂ ﬂ

 

where 00 is the initial surface coverage, and 0m is the surface coverage at Tm. In
this case, Tm now depends on the surface coverage. 00 may be found from the area under

the curve of the desorption rate as a function of time. For variable initial surface coverage

63

00, plotting ln(6l0 7;?) against l/Tm gives a straight line of slope E/R. A2 is then found by
plugging E into the above equation (E3-l3).

Thus, the order of desorption can be determined from the behavior of the
maximum in desorption rate with coverage curves. A ﬁrst order desorption with ﬁxed
activation energy gives rise to a peak in the desorption rate curve that does not change in
temperature with coverage. If the temperature of the peak maximum decreases with
coverage, the desorption reaction will be second order desorption or ﬁrst order with an

activation energy dependent on coverage. These two cases can be distinguished by the

plot of ln(1907;f) vs. l/Tm . A straight line means second order desorption.

3-2-2. Peak Shape Simulation

The shape of the experimental curve of desorption rate as function of sample
temperature can be used to determine the order of desorption and whether the activation
energy is constant or alternatively a function of surface coverage. In fact, the most
accurate data we can get from desorption experiments is peak temperature Tm. For a

linear temperature heating rate, we integrate E3-5 to obtain

 

f 91 W
=ln— forn=1
A T’ -5 9200 9
‘13—" I3 RTdT =_I6;=< 1 2 1 > (using dT=0dt) E3-14
r, 9. =—-— forn=2
( 92 61 J

 

The integral may be evaluated by using the substitution u = — f? , integrating by

parts, and using the identity

64

IerI = f— x H(u) for |u|>>1
u

1! 2!
where H(u) =1+—+—7+ .....
u u

Because 0 = — >>1 (this is true for most case, for example, when E=200 KJ/mol

RT

and T=1000 K, u=24), using only the ﬁrst two terms in the H(u) expansion is enough for

reasonable accuracy. We can get

 

7‘2 -L f— if ~
1%" 19 ”61!: f2; {72,28 R12 _ T128 RI, } E3-15
T

By combining E3-l4, E3-15 and E3-5, we can obtain an equation to describe the

shape of the desorption rate curve. For ﬁrst order desorption:

2 £1;
—Ln(—]X-)=£[i—i]+(l) e “(T ’4) -1 E3-l6
N R T Tm T

M m

where Nm is desorption rate at Tm.
It can be seen that the desorption rate curve (Figure 3-12) is asymmetric about the

maximum at Tm.

In the same way, for second order desorption, we have:

N {11+(T)ze-;.1%-Hl} E,J,

—N—~cosh"2 -£[i—i] E3-8
N 2R T Tm

65

Thus, the desorption curve is symmetric about the maximum at Tm when |T-Tm| is
small. Figure 3-12 shows the theoretical desorption rate curve for ﬁrst order and second

order.

 

 
   

 

_ 1st order desorption
........ 2nd order desorption

0.8

 

 

 

0.6 »

N/Nm

 

 

600 800 1000 1200 1400 1600

Temperature (°C)
Figure 3-12 Simulation of desorption ( Tm=1000 °C and E=237.8KJ/mol)

3-2-3. Estimating the Desorption Activation Energy

The desorption energy B may be calculated ﬁom Tm with the usual assumption
that A=109 sec'l Redheadml). Figure 3—13 is the energy change with peak position Tm
for ﬁrst order desorption. The asymmetric shapes in most of our H2 TPD curves (see
Section 2-6 and Chapter 4) suggest that the hydrogen desorption from Saran char is more
like a ﬁrst order reaction if E does not change with coverage. For a peak position located
at 1100 °C, the average desorption activation energy is about 250 KJ/mol (60 Kcal/mol),

a quite strong chemisorption. The C-H bond energy ( Pan and Yang) is 376 KJ/mol for

 

zigzag face and 354 KJ/mol for armchair face. It seems that the estimated desorption

energy here is smaller. The Error probably come from the A value assignment.

66

280
260 _
240 _
220 2
200 2.
180 -
160 e

140 » 1 1 :
500 600 700 800 900 1000 1100 1200

Temperature (°C)

 

 

Activation Energy (E KJ/mol)

 

 

 

Figure 3-13 Activation energy dependence on peak temperature Tm

3-2-4. Distribution of Desorption Activation Energies

The difference between the experimental TPD proﬁle and the simulated TPD
proﬁle shows that either E changes with coverage or that the surface is not uniform. The
observation that the population of hydrogen does not decrease evenly over the entire TPD
spectrum also suggests that the char surface exhibits a distribution of desorption
activation energies. Du et a1. [30] has developed distributed activation energy models for
CO. Similarly, these models can be used to calculate the adsorbed hydrogen activation
energy distribution.

According to E, the relationship between instantaneous H2 desorption rate and
the distribution of desorption activation energies can be expressed as:

M

dE
= S E —— E -19
dt 1H1. ( 1d, 3

where E is the local desorption activation energy, as approximated by an instantaneous

step at energy E. S(E) is the desorption activation energy probability density function.

 

3° Du, Z., A.F Saroﬁm, and J. P. Longwell, Energy & fuels 4,296 (1990).)

 

67
d[H]/dt is the desorption rate of H2, and [H0] is the total initial amount of H2 on surface.

The parameters are related by:

E AIT
— = ln

-— 3.1 E3-10
RT

 

Since the TPD data give the instantaneous rate directly, dE/dt can be obtained

from the above equation:

iii: Rﬂ£Lnﬁ—2.l + LnT) E3-20
dz ,6

In the original equation E3-10, T is peak temperature. Here, it is desorption
temperature. That is because local activation energy is approximated by an instantaneous
step. We think there is a continuous energy distribution, so every local activation energy
corresponds to a desorption temperature T. If we use a linear heating ramp T=To+Bt, S(E)
can be calculated using the following equation (combining E3-l9, E3-20 and using
d[H]=0.5d[H2]):

05de /dt

 

S(E) = E3—21

[H0]R,B(Ln A — 2.1+ LnT]
,6
For simpliﬁcation, we also use A=109 as in above section, which is also used by
most researcher. Figure 3-14 presents the D2 desorption activation energies for the
following TPD of a Saran char deuterium gasiﬁcation. Figure 3-15 is the H2 desorption

energy distribution.

68

0.04

 

0.03 _

0.02

0.01 -

S(e) [(mmol/g)/(KJ/mol)]

 

 

0.00» ~~111ty
100 120 140 160 180 200 220 240 260 280 300

 

E (KJ/mol)

Figure 3-14 Distribution of desorption activation energies for D2 TPD
spectra Following 6 hours D2 gasiﬁcation (E9619)

0.05

 

0.04 _
0.03 --
0.02 r

0.01 ~—

S(e) [(mmol/g)/(KJ/mol)]

 

 

 

 

(100 'ﬁ* 1 2 22. 1 1 1
100 120 140 150 180 200 220 240 250 280

 

E (KJ/mol)

Figure 3-15 Distribution of desorption activation energies for H2 TPD spectra
following 4 hours H2 gasiﬁcation of annealed Saran char (E9624)

3-3. Active Site Density Estimation

3-3-1. BET Surface Area and Adsorbed H2

69
Figure 3-16 is the adsorbed hydrogen concentration on annealed Saran char

following H20/H2 gasiﬁcation (at 850 °C reactant gas composition: argon O~60°/o,H2O

 

 

 

 

 

o~40%,H2 0~100%).
g 0.09 +
g 0.03 - V
51 0.07 - o 40% H2,3.1MPa
3 0.06 .X A 60% H2,3.1MPa
: O + v 0% H2,1.0MPa
1% 0.05 P 0 $1 5% H2,1.0MPa
1, 0.04 g V x 40%H2,1.0MPa
f 003 _ [:13 + 60%H2,1.0MPa

- V o 0% H203 MPa

5 0.02 - m 5% 112,03 MPa
8 0.01 '-
2 000 1 1 1 1 1 1 1

 

 

 

O 1 O 20 30 40 50 60 70 80

, w V, V ConVﬂSionl‘Vo .
A .Figure3:16Adsorbed:hYdrOgen:vs..convericn A ,1 ; .1 . .

In the above ﬁgure, although there is some scattering, we ﬁnd that the quantity of
H2 per unit BET area is almost constant and does not change with carbon conversion. For
Saran char, this value is about 0.05 cm3(STP)/m2. Thus, we can estimate the active site
density at about 2.2 x 10‘ mol l m2 , which corresponding to 37.2 A2 per hydrogen atom.
That means on average, for every 37.2 A2 BET area there is an active site.

Table 3-3 gives the calculation of the fraction of active sites occupied by oxygen

(00 complex) in steam gasiﬁcation (with or without H2).

Table 3-3 Active Sites Occupied by Oxygen

70

 

BET (from

 

 

 

 

 

 

 

 

 

 

 

 

ll "0" cc/g site (mol O)/g . )‘ site (mo|)/ m2 % of total site # Exp. # I
COHVCl’SlOll
II 5.5 2.46E-O4 l 133 2.17E-07 9.7 9635 (H20) '1
u l l 4.91E-05 1025 4.79E-08 2.1 9636(H2O+H2)||
Table 3-4 Active Sites Occupied by Hydrogen
“"112" cc/g site (mol 2)/g Bria—mm conversion) (mol 119/ m1 % oftotal site # Exp. # I
1 52.08 2.33E-03 1133 2.05E-06 90-3 96350120) I
[I 39.31 1.75E-03 1025 1.71E-06 97.9 9636(H20+H2)|

 

 

 

 

 

 

3-3-2. Comparing with Hermann (1996)'s Data'zg]

The value 2.2><1045 mol (H2)/ m2 is much smaller when compared to

Hermann’sm] data [1.2><10‘4 (mol 2CO+H2)/m2]. The reason probably is the carbon

material used is much different. Hermann used more active coke (600°C from PVC) with

a BET area of only 1 mZ/g. (For our char, the BET area is over 800 times larger than

Hermann’s char).

 

‘ From regression H2 BET data. In low conversion (<40%), BET=810+18.26*conversion%

71
3-4. Reaction Activation Energy

According to the Arrhenus law:

K = Koe ”’ E3—22
K is reaction rate constant, B is reaction activation energy.
R = Koe R" f (6,.) E3-23

R is reaction rate and f (c,) is a function of reactant concentration.

If we know the reaction rates (R) at different temperature, we can calculate the

reaction activation energy use equation E3-22

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

R]
R In —
R2
E = —1—'—1—‘ E3-24
(—-— - —)
T. T.
Table 3-5 Experimental data
reactants Temperature Rate (cc/min.g)
CO 002 CH4 SUM rate
9635 H20+Ar 725 0.0610 0.0290 0.0074 0.0974
9637 H20+Ar 850 0.1900 0.7700 0.5000 1 .4600
9636 H20+H2+Ar 725 0.0570 0.0240 0.0200 0.1010
9638 H20+H2+Ar 850 0.1200 0.0500 0.0970 0.2670
Table 3-6 Apparent Active Energy for H20+Ar Gasiflcation
H20+Ar gasiﬁcation Tl=725 °C T2=850°C
CO CO; CH.. use total rate CO+COZ
R850/ R725 R850/ R725 R850lR725 R850/ R725 R850lR725
3.1148 26.5517 67.5676 14.9897 10.6667
(1/T2-IT1) -1.12E-04 -1.12E—04 -1.12E-04 -1.12E-04 -1.12E-04
LN(R850/R725) 1.1362 3.2791 4.2131 2.7074 2.3671
Ea cal/mol 20247 58435 75080 48247 42183
or Kcallmol 20.2 58.4 75.1 48.2 42.2

 

 

 

 

 

 

 

72

Table 3-7 Apparent Active Energy HzO+H2+Ar Gasification

 

 

 

H20+H2+Ar gasification TI=725°C, T2=850°C
CO CO; CH, use total rate CO+C02
R850lR725 R850lR725 R850lR725 R850lR725 R850lR725
2.1053 2.0833 4.8500 2.6436 2.0988
(1/T2-fl’1) -1.12E-04 -1.12E-04 -1.12E-04 -1.12E-04 -1.12E-04
LN(R850/R725) 0.7444 0.7340 1.5790 0.9721 0.7413
Ea cal/mol 13266 13080 28138 17324 13211
or Kcallmol 13.3 13.1 28.1 17.3 13.2

 

 

 

 

 

 

The calculated results are given in Table 3-5~3-7. Comparing Table 3-6 and

Table 3-7, The activation energy for steam gasiﬁcation is larger than the steam/Hz

gasiﬁcation.

CHAPTER 4

CHAR GASIFICATION

74
4-1. Saran Char Gasification

4-1-1. As-prepared Saran Char Steam and Steam/H; Gasification

For steam/argon gasiﬁcation (Figure 4-1), the initial gasiﬁcation rate is relatively
high. After about 2 % carbon conversion, the rate curve tends to level out, although a
slight decline still can be seen. Compared to CO+C02 formation, the CH4 evolution rate is
very limited.

The steam/Hz gasiﬁcation is given in Figure 4-2. Methane is the dominant
product. The rate curve tends to decrease quickly out to 14 % carbon conversion. The
hydrogen inhibition of CO formation is obvious; at 10% carbon conversion, the total
gasiﬁcation rate is only about one third of the gasiﬁcation rate of steam/argon. The CO:
evolution rate is lower than steam gasiﬁcation and also very limited compared to CO and

CH4, since H2 in the gas phase has shiﬁed CO; to CO, or the adsorbed hydrogen inhibits

 

 

 

 

 

 

the formation of C02
g 0.3
a —*7-— CO
§ 0.25 + 002
E 0.2 9 CH4
8
c 0.15
E
8 0.1
e; 0.05
. It" 0 NW. rue-1'”

 

 

0 2 4 6 8101214161820
Conversion%

 

__ . . _, ﬁcatlonofSarancharm31Mpa j f ~w
. : -*--g‘ii:4;55320(40%)/A1‘5(50°@28t27259.0(Exam);-;;, g!

Rate (% Carbon Converted/Min)

Rate (% Carbon Converted/Min)

75

0.16 2 v7

0.14 .7. C

0.12
0.1

0.08

0.06 \

0'04 — \‘VW \Nxs‘e’eeeeeeee

o 02 — WW WWW , 98W
° WV‘V’W-V-VV-WVVW

0 534.3,; A, 1A A A AA 1AA MA MAMAAJAAAAAIAAAAAA

0 2 4 6 8 10 12 14 16 18
Conversion%

Figure 4-2 Saran char gasification in 3.1 MPa
H20(40%)/Hz (60%) at 725 °C (EXP-X-8).

 

 

 

 

 

 

 

 

 

0.08 ”‘7 _,,

11 ’70 CH4
0'07 11, L27:
0.06 {1‘

1
33: 1 K.
- 1
0.03 I WWWWQﬁ
0.02 1 l
0.01 L
0 1.5 _ 1 1 1 1 1 1 1
0 1 2 3 4 5 6 7 8

Conversion%

Figure 4-3 4 hr. Annealed Saran char H2 gasification
at 850 °C and 3.1 MPa. (E7062)

76
4-1-2. Annealed Saran Char H2 ISteam Gasification

The only product in pure hydrogen gasiﬁcation (Figure 4-3) of annealed Saran
char is CH4. The initial rate is about twice the gasification rate at higher conversion.
Although a very slight increase above 4% conversion can be seen, the methane formation
rate tends to stabilize after 1% carbon conversion. Aﬁer 4% conversion, a slight upward
trend can be seen in the rate curve. We could say that the gasiﬁcation process reaches
such a state that the surface reaction reaches steady state and the surface area increases
slowly with carbon burn off.

For annealed Saran char steam/Hz gasiﬁcation (Figure 4-4), the total gasiﬁcation
rate is lower than that of as-prepared Saran char at the initial stage. But, after 2% carbon
conversion, the rate is about same as that of as-prepared Saran char. Methane formation
rate is very low compared to CO and C02 formation. A slight increase can be seen in rate
with conversion.

The inhibition of hydrogen on steam gasiﬁcation rate of annealed Saran char is
clearly seen in Figure 4-5. The steam/Hz gasiﬁcation product distribution is similar to
that of as-prepared Saran char, and CO+COz formation rate is about one-fourth the rate in
steam/argon gasification of annealed Saran char. Methane formation in H20/Hz
gasification is much higher than H20/Ar gasiﬁcation, but the rate is still lower than in
pure H2 gasiﬁcation. This phenomenon is contrary to Huttinger’sls] conclusion, as they

thought HzO could enhance methane formation rate.

 

5 Huttinger. K.J. and Merdes, w.1=., Carbon 30(6), 883 (1992).

77

0.18
0.16 —

0.14 — M
0.12
0.10 '
0.08 1
0.02 V

0.00wwﬁlu$411ﬂ1ﬁ1119141f3111101111?111?111B_Lj
0 2 4 6 8 10 12 14 16 18 20
Conversion%

Figure 4.4 Armealed Saran char gaStﬁeatton of40%H20/60%Ar . _ ‘ '
., 31850 °C and 3 71 MPa [2 hour] (E7030) - -

 

 

 

 

 

98

 

 

Rate (% Carbon Converted/Min)

 

 

 

 

 

 

 

 

£0.06-

a

30.05 +90

0: I +002

§004 —e——CH4

.. 1.

«10.02 ’

0

£0.01
0 .

a 0 1 ................ _ ......
“5 0 1 2 3 4 5 6 7 8

Conversion %

Figure 4;$ 4‘ ed Saran char 40%H20/60%H2gasrﬁcatton ' '

at 850°C and 3 1 MPa. ‘ [4? hour] (7058)

 

78
4-1-3. Hydrogen TPD Proﬁles of Annealed Saran Char Following Gasification

A. H2 gasiﬁcation

Figure 4—6 is a typical hydrogen temperature programmed desorption (TPD)
proﬁle following gasiﬁcation in pure H2 for 6 hours. It shows a maximum peak at about
1200 °C, with H2 evolution staring about 900 °C and continuing up to 1500 °C; this is
clear evidence that hydrogen is dissociatively chemisorbed on the char surface. The
hydrogen peak is broad and asymmetric, indicating a distribution of hydrogen adsorption
energies on the char surface, or that desorption is a ﬁrst order reaction (see Chapter 3).

The H2 desorption peak shape and location in TPD changes with conversion
following pure hydrogen gasiﬁcation as given in Figure 4-7. We can see the basic shape is
very similar, but the quantity of adsorbed hydrogen increases and the peak maximum shift

to higher temperature with increasing conversion.

 

   
    

 

 

 

 

 

 

 

1 1—
% 0'8 _— 42 cclg Hydrogen]
(0 ~ —*— H2
5, 0.6 :—
.E :
g 0.4 L—
0 t
O ..
£1 0.2 f
ObLJ llllllLJmlLL
0 300 600 900 1200 1 500

g g . f 1 Temperature (°C) 1. , _
" ....anure4~6HzTPD proﬁle followmgéhrpure H2 """" ‘ ; .‘  ,.
' 'iéfa”"’”j'ficatton‘ ofSaran char at 850 °C and 3 1MPa (7063)

 

79

 

 

 
  
  

 

 

 

 

1
94 6hr 13.8%,42.0 cclg
0.8 +4hr, 5.9%, 33.2 cclg
UP —~—2hr3.8%, 15.6 cclg
.g _
g0.6
I
0-2.’
01 .-‘—....,__,.“__,=_____,__m_"__‘:I,” , . . . . . ,
300 600 900 1200 1500

Temperature (°C)

Figure ‘4—7 Comparison of H2 TPD profile following pure H2 gasrﬁcauon of annealed
.- , . = Saran char at 850 °C and 3 1 MPa(E 7064, E7062 and 7063)

 

 

H2 CCImin.gram(int.)

 

 

0 300 600 900 1 200 1 500
g _ Temperature(F’C)
Figure 48 I12 TPD proﬁle followmg 6 hr 60%Hy40%Ar
. gasiﬁcation of Saran char(E7065)

char
1111‘

11111111

80
B. szAr Gasiﬁcation
After the addition of 40% argon to hydrogen during gasiﬁcation, the TPD proﬁle

following reaction (Figure 4-8) is very similar to that following pure hydrogen

gasiﬁcation. The peak maximum position is at a little lower temperature.

C. Stgm Gasiﬁcation

For steam gasiﬁcation (Figure 4—9), the TPD proﬁles are different in that their

shapes are more symmetric.

 

1.0

 

.0
00

.0
m

 

H2 cclmin
.0
A

 

 

\‘
X‘.::3A

O

 

   

. .- ‘ “.r-r» . z ; f h: ’ 1 1 1 1 1 l 1 1 1 1 1
300 600 900 1200 1500
. Temperature((f’c). ‘

.~.;. .g-’zefahnéaiédsarlaﬁ:chatfatJSSOYC and£190}51511941570358.70393040)7 3 t L i '
D. Adsorbed Hydroggn (Deuterium) Chgnge with Conversio_n

The quantity of hydrogen adsorbed as a function of carbon conversion is given in
Figure 3-10. For a wide variety of gasiﬁcation conditions, hydrogen adsorption on Saran
char during gasiﬁcation is characterized by very rapid uptake upon initial exposure to

reactant gas up to 0.5% conversion. At high conversions, the absolute amount of

hydrogen adsorbed increase steadily out to 70% char conversion, where the value of

81
nearly 100 cm3(STP)/g corresponds to H/C ratio in the char of 1:10. The quantity adsorbed

depends solely on the char conversion and is independent of reactant gas composition.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A 100
m
8 x X
v 80 .—
C x
8 x
0 60 —
i x X X X X H2(ads)
I 40 _ x x . )§ . DZ+H2(adS)
'0 X )2
a: ﬁx
.0 Q
5 20 $3“
to
'o
< 0 1 1 1 1 1 1
0 10 20 30 40 50 60 70
Conversion %
Figure 4-10 Adsorbed hydrogen change with conversion
1800
g 1600
<
8 1400
m
E 1200
‘0 E] Unreacted
[I] 1000 Q 0% H2
m Q 5% H2
i“ X 60% H2
600 I 1 1 1 1 1 I
0 10 20 30 40 50 60 70 80

Conversion %
Figure 4-11 BET area changes with conversion

82
E. BET Area Changes with Conversion

 

The surface area (BET) change with conversion is given in Figure 4-11. At
conversion less that 30%, the increase is near linear. After that, the trend tends to be ﬂat.
The maximum possible BET area is 1600 mZ/g.

4-1-4. The Desorption Profiles of CO and CO2 after Gasification

The quantity of CO and CO2 desorbed from annealed Saran char during TPD
following gasification was monitored in several experiments. When hydrogen was
present in the reactant gas during the gasiﬁcation at 850 °C, only very small quantities of
CO (less than 0.04 mmng) and no CO2 was observed. When reaction gas contained
only steam/argon, slightly greater quantities of CO (up to 0.07 mmng) were observed in
TPD (Figure 4-12). For all reactant gases, the CO desorption peak is symmetric with a
peak maximum at 1000 °C. The CO desorbed represents “stable” surface oxides on the
char at gasiﬁcation conditions.

The low temperature gasiﬁcation is a little different. When hydrogen is present in
the reactant gas, less than 0.04 mmol CO/g and no CO2 was observed during TPD.
Following gasification in steam/argon only, in contrast, 0.21 mmol CO/g was observed to
desorb during TPD. The shape and location of the CO TPD peak following gasiﬁcation at
725 °C was the same as following gasification at 850 °C, signifying that the nature of the

surface oxides on the char is the same at the two temperatures (Figure 4-12).

83

 

  
 

-é—— 725 ClSteaml

0.15 r o—- 850 Clsteam

 

 
 

 

 

 

0.10
0.05
0.00 ...
700 800 900 1 000 1 1 00 1200 1 300
Temperature (°C)

Figure 4-12 CO TPD profiles following 40% steam/60% argon
gasiﬁcation at 725 °C and 850 °C (E9635 and E9637) _

 

 

 

  

 

 

 

 

0.25
. + Reactant purge when cooling
x inert gas purge when cooling
E 0.2 —
S
g 0.15 —
é
o _
o 0.1
N /
xx _ . ,
0 WMQ‘PMx-Vgxlﬁl
0 300 600 900 1200 1500

Temperature (°C)
Figure 4-13 H2 TPD proﬁles of annealed Saran char after gasiﬁcation T=850 °C,
P=1 MPa, cooling in argon and cooling in hydrogen (E7069 and E9603).

84
4-1-5. Weakly Bound Hydrogen Detection

The so-call weakly bound hydrogen (associative hydrogen) can not be seen in our
regular TPD. The reason is the temperature we used for gasiﬁcation is over 700 °C, and
from the literature, weakly-bound hydrogen only exists below 650 °C. To verify this
phenomena, a special experiment was designed in which we cooled the gasiﬁed sample in
hydrogen instead of argon. Figure 4-13 is the result of this experiment. It is clear that
there is a small peak starting at 650 °C that is identiﬁable as weakly adsorbed H2; we

cannot ﬁnd this peak in any TPD with inert gas purging.

4-1-6. Discussion on Saran Char Gasiﬁcation

During gasification, carbon surface area tends to increase with conversion due to
the carbon burning off and pore opening, and the active sites will decrease with reaction.
These two factors contribute to the rate curve shape (apparent rate). The surface area
increase is a slow process. For as-prepared Saran char, the surface active sites are
abundant. During gasiﬁcation, the active sites decrease in number faster than the surface
area increases, so the rate curve has a decreasing trend.

For gasiﬁcation of annealed Saran char, the rate increases with carbon conversion.
After annealing, the microstructure of Saran char must be stabilized. The process of
losing active sites is slower than the surface increase, and we always see an increasing
trend in the annealed Saran char gasiﬁcation rate.

From Saran char gasiﬁcation, we found that the quantity of hydrogen adsorbed on
annealed Saran char is independent of reactant gas composition and, on a unit surface
area (BET) basis, is a single, constant value (2x10’3 mmol H2/m2) at conversions above

0.5%. Gasiﬁcation rate, on the other hand, is strongly dependent on the reactant

85

composition. It is thus clear that the large quantities of hydrogen adsorbed can not be
responsible for observed difference in gasiﬁcation rate at different composition.

As evidenced by the high temperature TPD peak, dissociative chemisorption
clearly takes place on the annealed char surface during gasiﬁcation. The presence of
strongly bound hydrogen certainly affects gasiﬁcation rate by blocking potentially
reactive sites for steam gasification, but because this hydrogen saturates the surface at all
conditions, it exerts a constant inhibition.

The weakly bound hydrogen, proposed to be associatively adsorbed as surface
methylene groups is present in a concentration of 0.025 mmol H2/g, about 1% of total
quantity of hydrogen adsorbed on the char. Because it is stable only at low temperature,
the associatively adsorbed hydrogen could not inhibit gasiﬁcation by blocking active
sites.

Hydrogen occupies the vast majority (over 95%) of adsorption sites on the char
surface, as evidenced by comparing the quantities of CO and H2 released during TPD.
Computational molecular modeling showed the hydrogen preferably adsorbs on both
zigzag and armchair sites on graphite surface; these ordered edges are thus the location of
the large quantities of hydrogen. The constant value of surface hydrogen concentration
per unit area (~2x10‘3 mmol H2/m2) indicates that the quantity of edge sites per unit area
remains constant over the course of conversion.

At conversion large than 1%, the char surface consists of ordered edges saturated
with hydrogen (3-5 mmol/g), as well as less ordered structure(defects and amorphous
material)[27]. So, we could say that hydrogen gasification at edge sites makes the primary

contribution to methane formation. Steam gasiﬁcation is proposed to take place primarily

86

on less-ordered structure, where surface oxides can steadily adsorb (0.2 mmol/ g) and
adsorbed hydrogen does not block C(O) formation. Steam gasiﬁcation likely does not
occur to a signiﬁcant extent at edge sites.

The fact that both quantities of hydrogen adsorbed and char reactivity per unit
area is unchanging over conversion from 5~40% is interesting. It seems that the relative
concentrations of ordered (edge) and amorphous sites on the char surface are constant
over this conversion range or physical changes in char morphology with conversion

coincidentally result in constant rate.

 

27 Michael G. Lussier, Ph.D. dissertation Michigan State University (1997)

37
4-2. Coal Char Gasiﬁcation

Limited numbers of coal char gasiﬁcation experiments were done for the purpose
of verifying the rate trends.
4—2-1. As-received Coal Char Steam/H2 Gasiﬁcation

Steam/Ar gasiﬁcation of Coal char is different with Saran char (Figure 4—14) in
that the dominant product is C02. The ash in coal char catalyzes the shiﬁ reaction. The

composition of reactor eﬁluent almost reaches the equilibrium composition of the shift

reaction (see Chapter 3.1).

 

 

 

 

 

 

E

2 0.14

E 0.12 —- +00

t g.

2 0.1— +002

8 +CH4X10

o 0.08 a ‘2

8006 \§‘555555853“A..l ‘ .
g ' ¢¢::. “-iAé-A$aa-
O 0.04 73993::3::=:e=:::.-.:
€50.02,

m I. Y 7 vv' 1:
g 0 ""‘VW—V—V~HWV—Y?vvvvvvvvvvv

 

 

0 1 2 3 4 5 6 7 8
Conversion %

Figure 4314 Coal Char gasrﬁcatron 1n 1 0 MPa 40% H20/60%
*' ‘ at 725 °C[2 hours] (EXP-X3) ' ,
After addition of H2 to steam gasiﬁcation (Figure 4-15), the rate proﬁle is much
different from steam/argon gasiﬁcation. The CO evolution rate is higher than CO2. This

further veriﬁes the shift reaction is rapid in this system. Below 4% conversion, the rate

decreases quickly, and then the curve levels out above 4% conversion.

Rate (% Carbon Converted/Min)

Rate (°/0 Carbon Converted/Min)

88

 

 

 

 

 

 

 

 

 

4 6 8 10 12 14 16

Conversion %

Figure 4—15 C0a1 char gasrﬁcation 1n 1 0 MPa- H2O(40%)/H2
, at 725 °C. 6 hours reaction time, Conversion 16. 3% (X 7)

 

 

 

 

 

 

 

 

0.06

0.05— I'l‘é‘iéia“iaassaa

0.04 — ‘

A.A —v—— CO
0.03 +002 ..
+ CH4X10

0.02 A '

001 P 2 -—e-‘-'—-.:.?4-2-2M”1 f i - 9 1 '1 . .7
11:.»- - "' ‘1

0 E 1 l l 1 l l 1“

0 6 8 10 12 14 16 16 20

Conversron % _

89

4-2-2. Annealed Coal Char Steam Gasiﬁcation

Using different gasiﬁcation times, three steam/argon gasiﬁcations were done on the
annealed coal char (PSOC 1493). The rate curve of gasiﬁcation for 6 hours is given in
Figure. 4-16. The dominant product is CO2, again different from annealed Saran char for
which the major product is CO.

After gasiﬁcation, the N2 BET surface area was measured and is given in
Figure 4-17. Compared to Saran char, the BET area is much lower. The BET areas
increase rapidly with conversion (up to 10%) and then levels out, similar to Saran char.
The low initial area is attributed to pore blockage during charring; carbon sample surface

area rapidly increases as the pore structure is opened by the carbon burn-off.

200

 

 

 

BET surface Area (square meter lg)

 

 

 

O l l l l
0 5 1 0 1 5 20 25
.. .. Conversion (%)
Figure 4—17 N2 BET ’jI-[fcc area With conversron after gasrﬁcatron

at T9850 so and :1 MPa (40%H2I+60%Ar)

_A
.N
O

 

10.0 +—

6.0 H
4.0 *-
2.0 *—

0.0 l l 1 J l

 

8.0 ~>____,,/>

 

l

 

Adsorbed H after reaction (cc/g)

Conversion%

8 1O 12 14 16 18

20 22

1 __ Fig-use#13:40%H20+60°AAxigasiﬁcation1112850 59C, P=1MPa [17

0.07

 

0.06 —G\

0.05 P

'2
T

0.03 F
0.02 —
g 0.01 L
000 L l l l 1

H2 [cc/square Meter ( BET )]

 

 

l

 

Adso

8 1O 12 14 16 18

Conversion%.

20 22

. .1} Figure42129240%H2051760%Ar}gaaﬁcatibn11172850.??(2,P§=§1Mpa

91
Figure 4-18 gives the quantity of adsorbed hydrogen from TPD analysis as a

function of conversion. Because of the lower surface area, less hydrogen is adsorbed on
coal than on Saran char (at the same conversion ). Similar to Saran char, the quantity of
H2 adsorbed per unit area (Figure 4-19) is essentially unchanged with conversion. The
quantity of hydrogen adsorbed per unit area, however, is approximately l.5~2 times that
on Saran char; perhaps because of the presence of ash or hetero-atoms that may enhance
adsorption. Ifwe assume each H is adsorbed on one active site, the site density is about
5x10'3 mmol/mz.
4—2—3. Hydrogen TPD Proﬁles of Annealed Coal Char Following Gasiﬁcation
Figure 4-20 is H2 proﬁle following steam gasiﬁcation of coal char. The peak
position is a little lower temperature than that for the Saran char (1150 °C). The peak

position does not change with conversion. The shape tends to be a symmetric curve.

 

 

 

 

 

 

 

0.25
—><— 13.8% 8.0 cclg
a, 0'20 ’ +20.8%10.7cclg
c' 968.7% 7.60le
'g 0.15 +— ,
E
N 0.10 —
I
0.05 —
0.00 “i=4 1 1 1
300 500 700 900 1100 1300 1500

Temperature (°C)
Figure 4-20 H2 TPD proﬁles ofsteam * - ,, ... . ,
, . : and I MPa Desorbed H2 changes With conversron (E9601 9602 9603) I f

 

[cation ofanneaied coal char at 850 °C f. ;

92
4-2-4.The TPD Proﬁles of CO and CO2 after Coal Gasiﬁcation

The TPD proﬁles of annealed coal char are totally different than those for annealed
Saran char. No CO2 peaks are observed in the TPD aﬁer steam gasiﬁcation, but complex
large CO peaks are obvious. Figure 4-21 gives the CO TPD proﬁles for three diﬂ‘erent
steam gasiﬁcation times. The complex proﬁles suggest that there are many different types

of oxygen groups on the Saran char aﬁer gasiﬁcation.

 

 

 

 

0.12 - <
c. —o— 20.8%: 9.3 cclg ” ..
E 0.1 _ _._ 13.8%: 4.7 cclg ,
3 + 8.7%: 9.2 cclg I 16,.
2 0.08 —~ \-
5 0 06 — i ‘
§ ‘ ‘1
9
Lu
0
0

 

 

 

400 600 800 1 000 1200 1400 1 600
Temperature (°C)

 

charat850°Cand3lMPaforthreeconversmn  

4-2-5.Discussion of Coal Gasiﬁcation

From the limited number of coal char gasiﬁcation experiments, we found the same
trends in gasiﬁcation rate proﬁles. But the ash in coal char accelerates the shift reaction,
and is also responsible for the large amount of oxygen groups on the coal char surface
during gasiﬁcation.

In summary, coal char gasiﬁcation is more complicated than Saran char due to
impurities present. Although we found some trends that are very similar to Saran char, the

steam/H2 gasiﬁcation mechanism may not be exactly the same.

93
4-3. Isotopic Effects

4-3-1. Annealed Saran Char Gasiﬁcation by D2

Compared to H2 gasiﬁcation, gasiﬁcation rate in D2is much lower as seen in
Figure 4-22. Comparing with Figure 4-3, we can ﬁnd the trend is also different. The D2
from the following TPD is also smaller than the hydrogen gasiﬁcation (Figure 4-23).

Gasiﬁcation of annealed Saran char in 3.1 MPa D2 at 850 °C for 6 hours led the
deposition of about 13.7 cc (STP)/g D2 onto the char surface (Figure-23), and about four
percent carbon conversion. This quantity of D2 is substantially less than the
corresponding quantity of H2 adsorbed in hydrogen; four hour gasiﬁcation in 3.1 MPa H2
results in about 32 cc(STP)/ g adsorbed. A duplicate run (F igure-23) conﬁrms the
reliability of the quantity of D2 adsorbed.

The reasons for low gasiﬁcation rate and low quantity of adsorbed D2 still are not
fully understood. From the adsorbed H2(D2) with conversion plot (Figure 4-10), it seems
the low D2 is still in the same range as H2. The difference in the diffusive coefﬁcient
between H and D could cause the difference of reactivity and accessibility to micro pore
structure. Then the lower carbon conversion causes the smaller amount of D2 deposition

Also, the upturn in D2 TPD near 1500 °C may suggest that not all D is desorbed

before 1500 °C; e.g., some remains on surface. (No similar upturn in H2 TPD). We tried
to use D-NMR experiments to measure adsorbed D on the char surface. But the amount
of D is too small and beyond the NMR detection limit. So, we cannot get any signiﬁcant

D peaks for the gasiﬁed char sample or the residual char sample (After TPD).

0.012
0.010
0.008
0.006

0.002
0.000

Rate (% Carbon Converted/min)

0.004 -,

94

 

  

 

 

lliLLLLilllLi_1_LllillllillllilllliI
0 0.5 1 1.5 2 2.5 3 3.5 4

Conversion %
Figure 4-22 Annealed Saran char D2(6O%)1Ar

 

gasiﬁcation at 850 °C and 3.1 MPa, time: 6 hr. (E9620)

 

 

 
  
   
 

 

Repeat D2 gasification
—*— D2 13.7 cclmin
~><— D2 11.9 cclmin

 

 

 

.1 ,,,,,,

lllllll
' 1:553:22; - : ------

300 600 900 1 200 1 500

 

Temperature (°C)

Figure 4-23 The deposited D after D2 gasiﬁcation on the char is less than

that of H2 gasiﬁcation( two TPD after 6 hr. D2 gasiﬁcation )

95
4-3-2. Annealed Saran Char Gasiﬁcation by D20

 

 
  
  
 

 

 

 

 

 

 

 

7:: 0.14

g ’M
g 0.12

g 01 +004 +C02

g + CO —8— CO+C02

o 0.08

c _ W
3 0.06 :, 22.22::3..W
8 0.04

g

‘“70.02 HAW Q 9 0
g 0 llliLllLillllJLllLJLlHlillllillLliJ_lL

 

0 2 4 6 8 1 0 12 14 1 6
Figure '4-124.D20 gasiﬁcation 0f anneaied Saranchar
* ~ Lat-341M921 “(17850.9(31131118? Zaheurjs. H .. ~, ‘
Gasiﬁcation rate of annealed Saran char in D20 is given in Figure 4-24. Compared
to H2O gasiﬁcation, the total gasiﬁcation rate in D20 is about 3/4 that of steam
gasiﬁcation (Figure 4-4). This also showed that the D isotope slows the carbon
gasiﬁcation rate.
The molecular weight of D20 is about 10% higher than H2O, which causes 25%
lower gasiﬁcation rate. Ifthe gasiﬁcation rate is limited by oxygen exchange, then the

isotope will not affect the steam gasiﬁcation at all. So, the isotopic effect must come ﬁ'om
the equilibrium of steam decomposition reaction (C + 020 <::> D2 + C(O) ).
Basically, the experimental results of isotopes are reasonable, because the heavy

molecule will lower the diffusion and adsorption ability and slow reaction rate. But form

our limited number of experiments, we cannot get ﬁirther details about this issue.

96
4-4. Mechanism Derivation and Identiﬁcation

4-4-1. Mechanism

A. Reverse Oxygen Exchange Inhibition

C,+H,o—“—'—>C(0)+H, (1)
C(O)L>CO+ C, (2)
C(0) + H, ——*——> C, + 11,0 (3)
C, = Cf + C(10) (site balance) (4)

Assume surface concentration of C(O) sites remains constant, so that amount produced

equals the amount consumed. The formation rate of CO is:

C, PWkl k2 C, PWkl
= k k (A)
k,+k_,P,,+k,PW 1+4], +—"Pw
k2 H k2
I_3_. Hydrogen Inhibition by Formation of a C(H)2 Complex

r = kzccw) =

 

 

C,+H,o—"'—>C(0)+H, (1)
C(0)——'S=—>C0+C, (2)
C, + H, —L2C(H), (3)
Ct = Cf + C C(11), + qu) (site balance) (4)

The C(O) balance(formation rate equal consumption rate) is from (1) and (2). Assume

that reaction (3) reaches equilibrium. In the same way, we get:
C,P k
r = k,Cm,, = k W lk (B)
3 l
1 + —— PH + —— Pw
k_3 k 2

 

_C_, Hydrogen Inhibition by Formation of a C(H) Complex

C,+H,O—"'—>C(O)+H, (1)
C(0)—5=——>Co + C, (2)
1
C, +511, ——*‘—>C(H) (3)
1

C(H)—Lacr'i'EHz (4)

97
C, = Cf + CCU” (site balance)(5)

Neglect C(H)2 , assume equation (3) is reversible and reach equilibrium easily. At steady
state, C(O) formation (1) rate equals consumption (2) rate. The CO formation rate is

r- 2 ('(0)" k k ( )
1+—‘—P3~5+—i Pw
k 4 k2

 

The generalized rate expression is thus
k 1 PW
1 + KIPw + K, P;

 

r=k,C(.,,,, = (D)

For oxygen exchange K2=k1/k2
For hydrogen inhibition by formation of a C(H)2 complex: K2=k3/k-3
For hydrogen inhibition by formation of a C(H) complex K2=k4/k.4
4-4-2. Kinetic Model

Results of all gasiﬁcation experiments in steam/H2 mixture were ﬁt to Equation
(D) to determine which mode of hydrogen inhibition best ﬁts the data and to calculate
rate constant for elementary reaction steps in the model. Gasiﬁcation rate (CH4 can be
neglected compared to CO and CO2) for each experiment was calculated on a per unit
BET surface area basis. The regression coefﬁcient (r2) for dissociative hydrogen
adsorption (n=0.5) was 0.67 and for reverse oxygen exchange or weak hydrogen
adsorption (n=1) was 0.957, clearly showing that dissociative hydrogen adsorption is not
responsible for inhibiting gasiﬁcation rate at our experimental conditions. Here we
already rule out the weak hydrogen adsorption from the weakly adsorbed hydrogen

detection. This result is supported by the fact that rate depends on hydrogen partial

98

pressure while quantity of hydrogen dissociatively adsorbed does not.(The regression

work of this section was done by Michaelml)

The rate constants and elementary rate constants are shown in Table 4-1. For

reverse oxygen exchange, the equilibrium constant (kl/1C1) is 0.028, signifying that

oxygen fractional coverage on active sites (C(O)/Cf) is relatively low when gas phase

hydrogen is present. The equilibrium constant for associative hydrogen adsorption is

large (k3/k.3 =K2=420 MPa'l ), indicating that if “associative” adsorption is the primary

mode of hydrogen inhibition, then hydrogen fractional coverage on the char surface

would have to be very high compared to oxygen.

Table 4-1 Rate Constants from experimental Data at 850 °C

 

Rate constants for Equation D with n=1

 

 

 

kf i 5.3x10'5 mmol C m'zsechPa'I =Crk1
K. 12 MPa’i
K2 1 420 MPa"

 

 

Elementary constants for Steam Gasiﬁcation

 

k l Ct=kf

5.3x10'5 mmol C m'zsec'TMPa'I

 

k3Ct=k(/K1

4.5xlO-6 mmol m"2 sec‘I

 

 

 

Elementary constants for Reverse Oxygen Exchange Steam Gasiﬁcation

 

- k1C1=er2/K1

1.9x10'3 mmol C m'zsec'IMPa']

 

 

 

Equilibrium Constant for “Associative” Hydrogen Adsorption

 

 

1K2

l 420 MPa'7

 

 

27 Michael G. Lussier, Ph.D. dissertation Michigan State University (1997)

 

99
4-5. Conclusions

Hydrogen inhibits the steam gasiﬁcation of carbons via different mechanisms at
different extents of conversion. At very low conversion, or in environments where the
quantity of hydrogen is limited, dissociative adsorption is observed to inhibit steam
gasiﬁcation at ordered edge sites on the char. At higher hydrogen pressures, the effect of
adsorbed hydrogen is still present but can not be observed because it saturates the edge
sites on the char surface. At these higher pressures, steam gasiﬁcation rate is observed to
be inhibited by reverse oxygen exchange which is dependent on hydrogen partial pressure
in the reactant gas. “Associative” hydrogen adsorption is rule out as an inhibition
pathway in high temperature gasiﬁcation (850 °C). The results obtained here help to
unify the ﬁndings of prior researchers who have conducted experiments at widely varying

conditions and attributed hydrogen inhibition to different mechanism.

CHAPTER 5

HYDROGEN STABILITY ON
CARBON SURFACES

101
5-1. Stability of Adsorbed Hydrogen to Oxidation by O2

To examine the stability of adsorbed D or H to molecular oxygen, a series of
experiments (gasiﬁcation then oxidation) were conducted (Table 1 ). First, annealed
Saran char was gasiﬁed in 3.1 MPa H2 or D2 at 850 °C for 4~6 hours to ensure the
surface was saturated with H or D. Then the reactor was cooled down to oxidation
temperature (450~480 °C) with argon purging, and oxidation was conducted in 5 or 10
mol % O2 in argon for l~3 hours. This temperature range of oxidation was chosen from
the char oxidation experiment (see Chapter 2). Such a temperature was used to ensure
that the oxidation was conducted in excess oxygen at low reaction rates and not all
carbon was oxidized when exposed to 02. At this temperature, the carbon bum-off (by
molecular oxygen) rate continuously decreased with time, and a lot of molecular oxygen

existed in the efﬂuent gases (see Figure 5-1).

Table 5-1 Experimental Data Summary

 

 

 

 

EXP. No 9621 9622 9623 9624 7062 9620
Gasiﬁcation agent 40% D2+Ar 40% D2+Ar H2 H2 H2 40% D2+Ar
Gasiﬁcation time (hour) 6 6 4 4 4 6
Apparent weight Lossl” 4.7% 5.3% 9.7% 8.3% 6.0% 4.0%
Oxidation time 1 hour 1.9 hour 2.5 hr 1.7 hr

Oxidation Temperature 45 5 °C 445°C 455 °C 476°C NA NA
CO produced during Oxidation cc/g’I 1,3 1,] 2,6 1,8

CO2 Produced during Oxidation cc/g'2| 15, 5 23,2 3 32 46.0

% of initial carbon converted 2.1% 3.0% 49% 6.2%

CO from TPD cc/g‘3r 6.9 8.2 1.] 17.3

co2 from TPD cc/gm 1.63 1.04 0.50 1.14

02 uptake(of int. weight)%[‘” 1.7 1.7 0.4 3.4

 

 

' From the residual sample weighting

2 From the analysis of efﬂuent gas composition during oxidation
3 From the analysis of efﬂuent gas composition during TPD

‘ From the amount of CO and CO2 during the TPD

 

Evolution rate cc[STP]/g.min

1.0

0.8

0.6

0.4

0.2

0.0 a

102

 

 

 

     

 

_ i —1— co 2.1 cclg
—e—— 002 45.9 cclg
-,-._.-_. ﬂ..- .-. ---._.___.-._--...-.-_ __- _ _--_ iii-3:-

 

0 20 40 60 80 100 120 140 160

, Reaction time (minutes)

.FisﬁreS-liCOand COz'prOﬁIesdurinsoxidatiohffr‘t‘léw, 96.20.0511111119624) * 4

H2 Desorption rate [cc(STP)/min.g]

0.40
0.35
0.30

0.25 L

0.20
0.15
0.10
0.05

 

 

 

—+— Carbon burn off 4.9% (H2=16.5 cclg)
””—9— Carbon burn off 6.2% (H2=9.4 cclg)
__—x— No oxidation (H2=31.4 cc/g) /

 

 

 

   

 

 

,\
\ ‘ s
- .
' :9.

 

-._.... ,,,,,.,« 1
300 500 700 900 1 100 1 300 1500

 

Temperature °C

S-ZHzTPDproﬁleatdrfferemondanonseventyfollowmg '

’ ~ H2gasiﬁcation‘andi oxidation(9623,9624JO6‘2);2* 5

103

 

 

 

 

20'25 —x—-— No oxidation D2=13.7 cclmin

E —+—— Carbon burn off 2.1% D2=8.0 cclg

:0: 02 ——9— Carbon burn off3.0°/o D2= 5.3 cclg

91

3.0.15

’5

g 0.1

80.05

8

300 500 700 900 1100 1300 1500

Temperature (°C)

Figure 5—3 D2 TPD proﬁle at different oxrdanon seventy followrng
6 hours D2 gasrﬁeanon at 850 °C and 3 1{ MPa (9620 9621 9622)

The effect of oxidation by 02 on adsorbed hydrogen is seen in Figure 5-2, and the

corresponding effect on adsorbed deuterium is given in Figure 5-3. Clearly, oxidation is

able to remove adsorbed H or D at least to some degree ﬁ'om the char surface, as the

amount of hydrogen (deuterium) remaining declines as the severity of oxidation increases.

Second, more weakly bound hydrogen (or deuterium) is clearly removed ﬁrst, as the

temperature of the peak maximum desorption rate of the residual hydrogen increases with

the severity of oxidation.

Table 5-2 Material Balance

 

 

 

 

 

 

 

 

 

 

 

EXP. No 9621 9622 9623 9624
Weight loss in gasiﬁcation 4.0% 4.0% 6.0% 6.0%
from product gas analysis

Weight loss in oxidation 2.1% 3.0% 4.9% 6.2%
ﬁom product gas analysis

0 uptake from TPD analysis 1.7% 1.7% 0.4% 3.4%
Actual recorded sample 4.7% 5.2% 9.7% 8.3%
weight loss

Difference 0.3% 0.1% 0.7% 0.5%

 

 

104
Table 5-3 Oxidation Effects

 

 

 

 

 

 

 

H2 gasiﬁcation D2 gasiﬁcation
EXP. No 9623 9624 9621 9622
Before Oxidation(H2) 31.4 cc/g 31.4 cc/g 13.7 cc/g 13.7 cc/g
Associated Carbon with H2 adsorbed 3.4% 3.4% 1.5% 1.5%
Oxide Carbon 4.9% 6.2% 2.1% 3.0 %
H2 left after oxidation 16.5 cc/g 9.4 cc/g 8.0 cc/g 5.3 cc/g
Associated Carbon with H2 adsorbed 1.8% 1.0% 0.86% 0.57%

 

 

 

 

 

 

 

From all H2 (and D2) TPD proﬁles, only one major peak is found, so it is
reasonable to assume all H is bound to carbon by single bonds. Aﬁer 4 hour hydrogen
gasiﬁcation, 31.4 cc/g H2 is adsorbed on the carbon sample, which will associate with
0.034g/g carbon (3.4% of the initial char sample). The formation of oxygen groups on
the char surface will offset the weight loss from carbon burn off (see Table 5-2)

Table 5-3 summarizes the oxidation effect. In all cases, we can see that the
amount of oxide carbon [carbon bound as C(O)] is greater than the amount of carbon that
is bound to H (or D) before oxidation, but quite a large amount of H (or D) still exists on
the carbon sample.

If we assume that oxidation only starts from the outer surface of carbon particles,
we can deduce that adsorbed hydrogen not only exists on the outer surface, it also exists
inside the pore structure. From D20 (or H2O) proﬁles of oxidation following D2 (or H2)
gasiﬁcation (Figure 5-4 and Figure 5-5), we can see that the H(or D) evolution (in the
form of H20 or D20) is fast at the beginning of oxidation and continuously decreases
during the oxidation process. That means the H (or D) on the outer surface can be easily
removed upon exposure to the molecular oxygen, but the H (or D) in the pore structure is

not that easily accessed. It is necessary to burn all surrounded carbon before removing

105

them. From all these, we would say only the hydrogen (or deuterium) that is accessible to

molecular oxygen can be removed.

 

   

 

 

 

 

 

 

0.10
’c‘ -
E 0.08 L —+—-0202.5 cclg)
32’ 006+
E
g 0.04—
E
g 0.02r-
LIJ
000 hi, illllillllillllilllhilLLLillllillllillll

O 10 20 30 40 50 60 70 80 90
Time (minutes)

Figure 5.41320 prdfile during oxidation (T=4’55°c, Poz=0.05atm)'(9621)

.0
w
o

 

   

Q
A
O)

 

—+— H2011.90c/g

 

0.12

0.08

0.04

Evolution rate cc[STP]/g.min

 

 

000 LllliLJlJillllillllillllilLJlill

II 111
V

O 20 4O 60 80 100 120 140 160
Time (minutes)

" Figure ssmowreéam aaasnamsvc.vp¢.=o~.gra:mx9'si24) ,

 

106
5-2. D-H Exchange During Gasiﬁcation

To ﬁirther explore the stability of surface H, we designed D-H exchange
gasiﬁcation experiments to study the exchange of adsorbed deuterium. After a long
gasiﬁcation (6 hours) in 3.1 MPa D2 at 850 °C, it is assumed that only D2 is present on the
carbon surface. Following gasiﬁcation, we switched the reactant gas to H2 and continued
the gasiﬁcation an additional length of time. Aﬁer gasiﬁcation, we conducted TPD to
measure adsorbed D and H. Before H2 gasiﬁcation, the quantity of D2 on carbon is 13.7
cc/g. After 0.8 hour H2 gasiﬁcation, 9.5 cc/g D2 exists on carbon sample from TPD
(Figure 5-6). So, a 0.8 hour gasiﬁcation in H2 is not enough to exchange all D off the char
surface. The identical peak position of D2 HD and H2 in TPD proﬁles is a evidence that

the exchange has no preference for different adsorption energy distribution if it does exist.

 

 

 

 

  

 

 

 

 

g 0.16

E 0-14 — Total16.5 cclg

E 0.12 _ —<>—H22.5cc/g

a, + HD 9.1 cclg

‘g‘ 0-10 — —+— 02 4.9 cclg

2 0.08 —

E 0.06 -

,§ 0.04 -

% 0.02 _ ,.

[1).] 0.00 #2- 2741'»"'-"V""‘1‘: 1 1 1 1 l 1 L1 1‘ f ﬁsts-0"."
500 700 900 1 100 1 300 1500

Temperature (°C)

Figure 5-6 TPD cf sample fmm 6 hours D2+Ar and 0. 8 hr: H2+Ar gasrﬁcation
* ‘ at P=_3 1 MPa and T=850 °C annealed Saran char (9608)

107

 

 

 

 

    

 

 

 

 

2’ 0.30

E Total 18.1 cclg

: 025 — —0— HD 2.1cc/g

a. ——4— DZXSO 0.1 cclg

('7) 0.20 —><—— H215.9cc/g

8

a) 0.15

E

c 0.10

.9

§ 0.05 2 . . ..-

O . ..- — -‘ ””2”,“ ‘ I.

> «Mt. "‘:"~‘=:.:-...._.. , . ..

“J 0.00 .. . . .. .. . 4. . . . .- ............. ”:55223352 :5;
500 700 900 1 100 1 300 1 500

Temperature (°C)

Figure 5-7 TPD of sampie from 6 hr. D2+Ar and 2. 5 hours H2+Ar
' gasiﬁCation at 3.1 MPa,T=850 °C, annealed Saran char (9610)

 

 

 

 

 

 

 

,2 2.55-9
% Switch here
'0 -
$2.05 9 WW2 2,. , W
7:71.559 — _ ,
4‘.“- 02 gasiﬁcation H2 gaSIﬂcatIon
2 1.059 ~ 2
0
'g —+— CH4
65.0E-10 ” ———x—— CD4
u’r
0 20 40 60 80 100 120

Time (minutes)
F Igure 5—8 CD4 and CH4 evolution during switch (6 hr. D2+Ar and 0. 8 heur H2+Ar
' gasiﬁcation P33 1 MPa, T~850 .°:C, annealed Saran char(9608) """"

108

Figure 5—7 gives the quantity of H2, HD, and D2 present on the char after six
hours D2 gasiﬁcation followed by 2.5 hr H2 gasiﬁcation. Clearly, most adsorbed
deuterium has been replaced by H on the char sample. The quantity of H2 adsorbed (15.6
cc/g) is very close to the quantity of D2 desorbed following D2 gasiﬁcation only (13 cc/g).
It is reasonable to conclude that sites on which D is adsorbed following D2 gasiﬁcation
are the only ones available for H adsorption during H2 gasiﬁcation.

From the transient proﬁle of CD4 (Figure 5-8), we can see the evolution is fast
after switch, so the outer surface D-H exchange is fast. But the D trapped in the pore
structure is much more difﬁcult to exchange and even 2.5 hr H2 gasiﬁcation can not
replace all deuterium.

The simple calculation is given in Table 5-4. When the amount of carbon gasiﬁed
by H2 approaches the amount of carbon that is bound to D, there is more than half the
residual D on the char sample (E9608). Even for 2.5 hours H2 gasiﬁcation, where the
3.8% carbon converted is over twice the amount of carbon bound to D, 1.1 cc/g D2 still
stays on the char. This suggests that H2 gasiﬁcation does not necessarily occur on the
sites which deuterium adsorbed on. It is well known that after adsorbing hydrogen, the
armchair carbon sites tend to be gasiﬁed ﬁrstly. So, the deuterium adsorbed on armchair
sites will be peeled off readily. On another hand, the deuterium adsorbed on zigzag

carbon can not be replaced until the surrounding carbon is burnt off( see Chapter 6 ).

109

Table 54 DH Exchange Effects

 

 

 

 

 

 

 

 

 

 

 

 

 

Exp. No 9620 9608 9610
H2 gasiﬁcation time 0 hr 0.8 hr 2.5 hr
Before switch cc D2/g 13.7 13.7 13.7
D2 left cc/g 13.7 9.5 1.1
Associated carbon % 1.5 1.0 0.1
Gasiﬁed carbon by H2 % 0 1.2 3.8
0 30 L —+— 4.9 cclg (0.83 hr H2 gasiﬁcation)
’ -—6— 0.1 cc/g (2.5 hr H2 gasiﬁcation) X10
025 _ —*— 13.7 cc/g (No H2 gasiﬁcation)
c
E 0.20 -
“’5 0.15 —
0.10 ~
0.05 ~
0. 00 Hg;+£’»~—A~4«b~/‘ﬁr"\aﬂfﬁtr—ﬁ-—ﬁ A ~ 2 egg-ricﬁ‘uiu‘ii'
700 900 1 100 1 300 1500
Temperature (°C)

Figure 59 D2 TPD proﬁles change with H2 gasrﬁcatton time followmg
- ' 6hrD2 gasiﬁcationat 850 °C (9608, 9610,9620)

Figure 5-9 is the summary data of the D-H exchange experiments. The amount of
residual D decreases with H2 gasiﬁcation time following the 6 hours D2 gasiﬁcation. The
most signiﬁcant phenomena are the peak position and shape. The amount of residual D
changes with time of following H2 gasiﬁcation, but the peak position and peak shape are
unchanged. That means that D-H exchange is different ﬁom molecular oxidation in that

this process does not take place with more weakly-bound hydrogen preferentially.

110
5-3. H-D Exchange During Gasiﬁcation

 

 

 

 

 

 

 

 

 

2’ 0.16

E Total 22.1 cclg

z. -—><—— H2=9.8 cclg

& 0.12 — —+— HD=10.1cc/g

23), —0— 'DZ=2.2 cclg

m 0.08 —- .

r

c

.2 0.04 —

3

">5 .., . . = =

LIJ0.00 ”3-“1 1.
700 900

Temperature (°C)

Figure 5—19 TPD of sample from 4 hours H2+Ar and 0. 7 hr D2+Ar gasiﬁcation

To further understand the mechanism of surface H exchange, we also designed
H-D exchange gasiﬁcation experiments to study the exchange of adsorbed hydrogen.
After a long gasiﬁcation (4 hour) in H2, 32 cc/g H2 is present on the carbon sample. We
then switched reactant gas to D2 and continued the gasiﬁcation an additional length of
time. After gasiﬁcation, we conducted TPD to measure adsorbed D and H. We can see
that a 0.7 hour gasiﬁcation in H2 can exchange half the adsorbed H2 (Figure 5-10). This is
faster than D-H exchange (comparing to Figure 5-6).

From Figure 5-11 and Figure 5-12, we can further identify that the fast exchange
of hydrogen by deuterium only occurs at the beginning. But after 2.5 hr. D2 gasiﬁcation,
fair amounts of hydrogen (5.9 cc/g) still exist on the carbon sample. This is another
signiﬁcant different from D-H exchange. It seems that adsorbed hydrogen could migrate
into the pore structure of the carbon sample and make it diﬂicult to access for the
following D2 gasiﬁcation. It is also possible the faster H-D exchange rate comes from the

high hydrogen gasiﬁcation rate (The adsorbed H is active than D).

111

0.14

 

 

 
  
   
  
     

 

 

 

 

 

 

a) Total 20.5 cclg
.E 0.12 —+— H2=7.3 cclg
g + HD=10.8 cclg
8 010 —e—— DZ=2.4 cclg
9, 0.08
l!
c 0.06
.9
§ 0.04
9
m 0.02
000 - " ‘ ' i “
700 800 900 1000 1100 1200 1300 1400 1500

Temperature (°C)
Figure 54-1 TPD of sample from 4 :i

 

at P3 1 MPa,T"850 °c annealed Saran ohar(9631)

 

 

 

 

 

 

 

 

0.14
U? 1 Total 21.5 cclg
E 0'12 —e— H2=1.8 cclg
B 0.10 _ ——+— HD=8.2 cclg
—><— 02=11.5 cclg
9 0.08
S
c 0.06
.2
§ 0.04
o
$10.02
o.oo ' : ‘ 1 ° ‘ ‘ ’ '
600 800 1 000 1 200 1 400

Temperature (°C)

[37 H2+Ar and l 3 hr D2+Arga31ﬁcation.’,f*"" .

Figure 5—12 TPD of sample ﬁom 4 hours H2+Ar and 2 5 hr D2+Ar gasrﬁcatton‘e}; 3;.

at P‘3 } MELT-"850 9C, annealed Saran of, ,i 2,

 

59632)

112

 

0.20

—~x—— 02 gasification 0.7hr
——+—— D2 gasification 1.3hr
—o—— D2 gasiﬁcation 2.5 hr

 

 

0.15 ~

 

0.10 r-

0.05

H2 Desorption cc[STP]Imin.g

 

 

 

0.00 « ‘ - 1 - . Y . . “I: ; ' ‘ : °~‘-.~‘ 221:1:2: 1-4;;
700 900 1100 1300 1500

V ._ Temperature (°C)
* flvaigure 5413 H2 TPD proﬁles change With D2 gasrﬁcation times foliowmg

 

4 hr H2 gasdication at 850 °C, 3 1 MPa (9630 9631 9632)

Table 5-5 H-D Exchange Eﬂ‘ects

 

 

 

 

D2 gasiﬁcation time 0 hr 0.7 hr 1.3 hr 2.5 hr
Before switch cc H2/g 31.4 31.4 31.4 31.4
H2 leﬁ cc/g 31.4 14.9 12.7 5.9
Associated carbon % 3.4 1.6 1.4 0.74
Gasiﬁed carbon by D2 % O 0.50 0.87 1.7

 

From the summary experimental data (Table 5-5), we can clearly see the H-D
exchange is much faster than corresponding D-H exchange. For 0.7 hour D2 gasiﬁcation,
only 0.5% carbon conversion results in exchange of half of the adsorbed hydrogen (which
bonds to 1.7% carbon). That means that the adsorbed H does exchange with D2 in gas
phase. clearly does happen in this case. The fast exchange also indicates that the bond
energy of CD is larger than C-H. So, adsorbed H is easily replaced by D2 from gas phase,

but H-D exchange is not such easy.

113

54. Conclusions

In conclusion, the adsorbed hydrogen (or Deuterium) is quite stable. The
molecular oxygen can remove the adsorbed hydrogen when the associated carbon is burnt
off. Carbon atoms and adsorbed hydrogen atoms have same opportunities to be oxided.
No preference is found in the oxidation process; weakly bound hydrogen (probably most
on outer surface) is removed ﬁrst. “Trapped” H (or D) is very difﬁcult to be removed
unless the surrounding carbon atoms are burnt off.

Adsorbed D(H) can be exchanged during the following H2 (D2) gasiﬁcation. The
exchange process has no preference for different bind energy of adsorbed D if it does
exist. H-D exchange is faster than D-H exchange and the exchange really take place, as
opposed to simple gasiﬁcation of carbon with adsorbed species (H or D) present. We also
found that some H could be trapped in the inner structure, and could not be exchanged (or

removed) until the surrounding carbon was gasiﬁed.

CHAPTER 6

RATE ENHANCEMENT USING
MOLECULAR OXIDATION

115

Introduction
From Chapter 4, we know that adsorbed hydrogen strongly inhibits steam/H2
gasiﬁcation. It is therefore reasonable to think about removing adsorbed hydrogen during
gasiﬁcation to recover and retain activity of the carbon. Reaction rates decrease rapidly
during the initial steam/H2 gasiﬁcation, due to rapid uptake of hydrogen. Probably, this
would occur following H2 removal during gasiﬁcation, so we would not expect the rate

recovery following H2 removal could last very long, unless hydrogen is continuously

removed from the surfaces.

The experimental scheme to investigate hydrogen removal during gasiﬁcation is
given in Figure 6-1. First, Saran char is gasiﬁed in H2/steam for 4 to 6 hours, and then
switched to purge argon. When the reactor temperature drops to 450~480 °C, the char
sample was oxidized in 10% (or 5%) O2 in argon for a certain time. After ﬁnishing the
oxidation, reactant gas is switched back to purge argon and the reactor as heated up
again. Once the reactor reached gasiﬁcation temperature, gases were switched to

H2/steam to conduct the gasiﬁcation again for 1~3 hours.

 

 

 

 

 

 

 

 

. - Tompontun —- 112/smut
..._PurgoArgon —02IArgon
u

s s = g
a 3 3 = "
o 1 ° 8
I ‘ICIE----1\g 8 ’.--§-‘ 0

I 5, g g I 85 ‘

I s ._ I 3%, ‘
' E o --<-d no a
o E o ‘

\

 

 

 

 

 

 

 

 

Figure 6-1 Experimental scheme

116
6-1. Oxidation Effects

Oxidation following gasiﬁcation will remove adsorbed hydrogen to some degree,
as seen in Chapter 5. After gasiﬁcation and oxidation, the following CO and CO2 TPD
proﬁles were taken (Figure 6-2 and 6-3). The shape of the proﬁles depends strongly on
oxidation conditions. For H2 gasiﬁcation and then oxidation, the CO peak is located at
900~950°C and the CO2 peak sits at about 700 °C (Figure 6-2); for D2 gasiﬁcation and

oxidation, the CO peak is located at 900°C and the CO2 peak at 600 °C (Figure 6-3).

 

 

  

 

 

 

 

 

0.25
—+—— (:0: 17.3 cc'g
—0— CO2:1.lcc:‘g
0‘20 b ——)<— H294 cc.."g
.E 0.15 —
E
8 0.10 —
0.05 —
0001.2- _ .. ... , .. .. .. .- _ _ ‘ a 2....:::::'.’
0 300 600 900 1200 1 500

Temperature (°C)
Figure 6-2 CO and CO2 desorption proﬁles from annealed Saran char
foliowmg gasrﬁcation in H2 for 4 hr at 850 °C and then oxidation :11

10% 02_' in Ar at 476 °C for ii. 7} hour(E9624)

Table 6-1 Estimation of The Surface Site Density

 

 

I EXP. No C0 C02 H2 (Or D2) Sum of sites Before oxidation I
E9624 17.3 1.1 9.4 28.9 ~34
E9621 6.9 1.6 8.0 18.1 ~13

 

 

F rom the two experimental results (Figure 6-2 and Figure 6-1), we can see that
most adsorbed oxygen is given off during the TPD in the form of CO. Table 6—1 gives the

simple estimation of the surface site density by adding C0, C02 and H2 or D2. From these,

l 17
we ﬁnd the adsorbed oxygen is using the same sites as adsorbed hydrogen. The CO peak

after 1300 °C is most likely from the ceramic TPD reactor, as it is seen during the blank

 

 

—+— co: 6.9 cclg
0.12 r —e— 002:1.6 cclg
—><—— 02: 8.0 cclg

 
 
   

 

 

 

 

 

300 500 700 900 1 100 1300 1 500
Temperature (°C) I . ,. _

Figure 6—3 TPD proﬁles of annealed Saran after 6 hours D2 gasrficatron ff i? 3 , I

1 hour omdatron (5% ()2 1n argon and 60 cclmin) at 455 °C (E9621) '

 

118
6-2. H2 Gasification Rate Enhancements

Results of H2 gasiﬁcation rate enhancement experiment is given in Figure 6-4.
We can see that the enhanced methane formation rate is almost twice as high as before
oxidation. The rate enhancements apparently can continue a long time. This phenomenon
is unexpected, since we only expect rate enhancement at initial stages of reaction. To

explain this, we need to look into the detailed structure of the carbon sample.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.4 %_1_1
.E' .l
E 1.2
g 1.0
CH4 rate enhancemen
8 0.8 ~~o~ Gasiﬁcationl
‘g -0- Gasiﬁwﬁonz
'5 0.6
>
L” 0.4
v
2:) 0-2 § mm
0.0 ' L 1 ‘1
O 50 1 00 150 200 250 300
Time (minutes)

5' 1.4 ~»——*
E
g 1.2 _ b Gasiﬁcation 2

_ —x— CO
8 1.0 ( ) 0 CO2
S 0.3 _. —e— CH4
0
II, 0.6 -
43 0.4 — ”GeeW-eee—ee—e
:3
.8 0.2 TXVXIXX
0': 0.0 “I:f:00-35-515'25‘4—246‘6—‘5186'“*§5'5"&5155_55f59<§7§

 

 

 

50 70 90 1 10 130 150 170
Time(minutes)
Figure 6-4 Annealed Saran char gasiﬁed in 3.1 MPa H2 at 850°C to 5.9% conversion
(Gasification 1), then oxidized at 450 °C with Po,=0.1 atm for 1 hour, then gasiﬁed again

in 3.1 MPa H2 (Gasiﬁcation 2). (a) Rate curves for H2 gasiﬁcation showing rate
enhancement; (b) Product proﬁles during gasification 2. (E9633)

119

First, the surface area of carbon samples increases with gasiﬁcation. Also,
according to many researchers, the zigzag carbon conﬁguration becomes more and more
dominant on the carbon surface as gasiﬁcation progress. Apparently, these two factors
work together to give the apparent constant methane (CH4) formation rate after the initial
stage of reaction. Our experiments (Chapter 4) showed that the increase in surface area is
very slow. To maintain relative constant gasiﬁcation rate during the hydrogen
gasiﬁcation, the rate of increase of the zigzag carbon conﬁguration must also be very
slow. Therefore, we can say that the changes in conﬁguration of atoms on the carbon
surface is a very slow process. In most case, after one layer gasiﬁcation, the armchair

conﬁguration is more likely to be maintained than the zigzag conﬁguration(Figure 6-5).

 

Figure 6-5 Armchair carbon maintain its structure during gasiﬁcation

In this ﬁrst step, the most active dangling edge carbon atoms are burnt off. Next,
the carbon atoms at the middle of the armchair positions are burnt off. Now, most of the
surface carbon is in the zigzag conﬁguration. Apparently, the top two carbon rings are
still more active than the others. The further burn-off of the four top carbon atoms will
recover the full armchair conﬁguration. Of course, this is only one possible explanation,

Lussierm] provide more detail in his dissertation. That is why the gasiﬁcation rate can be

120

maintained for a long time. That does not mean that the armchair conﬁguration will be
maintained forever. During the gasiﬁcation process, the zigzag conﬁguration always
tends to replace armchair conﬁguration.

Although this process is slow, the zigzag conﬁguration will dominate the carbon
surface after a long enough time. According to Y_ang[6]’s work, oxidation processes have
no preference for armchair or zigzag sites. After oxidation, the fraction of armchair sites
should increase and the ratio of zigzag sites should decrease.

The dimensions of ordered structure in Saran char is very small and small ordered
elements may be burnt off with gasiﬁcation( or oxidation) progress. This process will
open enclosed pore structure and increase the surface area. In the above experiments, the
BET surface area increases about 20% due to oxidation. After gasiﬁcation 1 the BET is
900 mz/g, and after oxidation the BET area is 1100 mzlg. Steam and H2 gasiﬁcation
prefer to occur on the armchair carbon, and hydrogen prefers to adsorb on zigzag carbon.
The increase of surface area and armchair ratio work together to make the apparent rate
enhancement possible. All these contribute to the long time constant methane formation
rate.

Yang’s theory may explain the initial high rate due to the formation of new
armchair carbon. But their results come from graphite, and cannot explain why the

enhancement can continue for a long time in our experiment.

 

6 Yang, RT. and Duan, R.Z., Carbon 23(3), 325 (1985).

121
6-3. Steam Gasification Rate Enhancements

0.07
0.06
0.05
0.04
0.03
0.02
0.01
0.00

 

 

volution ML/min.

   

 

 

Products evolution MUmln.

 

 

0 20 40 60 80 1001201401600- 0 20 40 60 80 100120140160

0.35

0.30

025 l b T ' 00(1)
- 0 00(2)

0.20

0.15

0.10 W
1 l

0.00 01111111

0 20 40 60 80100120140160

 

 

 

20 CH4

—<% 002

 

 

 

Products evolution MUmln.

 

 

 

 

Products evolution MUmln.

200 250 300
Time (minutes) Time (minutes)
Figure 6-6 Rate enhancement by oxidation.

Conversion: 5.0% (gasiﬁcationl), 2%(oxidation), 10.7%(gasification 2) [H2O gasiﬁcation
of annealed Saran char at 725 °C and 3.1 MPa; oxidation at 425 °C and 3.1 MPa for 1 hr]
a) CH4 formation; b) CO2 formation; C) CO formation; (1) overall reaction

Steam gasification is more complicated than pure hydrogen gasification. The
gasiﬁcation temperature also affects the rate enhancement results.

For the 725 °C, rate enhancement experiment is given in Figure 6-6. Methane
formation rate is almost four times higher as after than before oxidation, although the
rate is very low compared to the CO and CO2 formation rate. This methane rate
enhancement is much larger than in H2 gasification. Practically, this enhancement is
meaningless, since the enhanced methane rate still is too low to have any value, but it

useful for understanding the methane formation mechanism. In this experiment, very low

oxidation temperature (425 °C) was used to burn off about 2% of initial carbon. Thus,

122

we cannot believe that the surface area has changed signiﬁcantly. The methane rate
enhancement only can come from the uptake of oxygen groups on the carbon surface.
The adsorbed oxygen substitutes the adsorbed hydrogen, so the extension of hydrogen
inhibition decreases. According to Huttinger’sls] methane formation mechanism, the
oxygen group is necessary for methane formation and oxygen groups can exist on the
carbon surface at low gasiﬁcation temperatures.

Although many researchers thought that CO2 only comes from the shift reaction,
we do not agree. From the CO and CO2 evolution rate proﬁles, we can clearly see that
some CO2 does come from carbon surface reaction. Theoretically, at the same reaction
conditions, the extent of the shiﬁ reaction should not change. At the beginning of the
second gasiﬁcation, the formation rate of CO and CO2 changes with time (Figure 6-

6 b,c). We can only conclude that CO2 not only comes from the shift reaction, but also
from the surface carbon gasiﬁcation at 725 °C. After oxidation, a fair amount of oxygen
groups exist on the carbon surface. So, the dominant product is CO2, and the CO rate is
very low at the beginning of the gasiﬁcation. With reaction progress, the surface oxygen
groups are consumed, the CO2 formation rate decreases and the CO rate increases. Even
after 2.5 hours, the CO2 formation rate continues to decrease. At the same time, the CO
formation rate increases to the value before oxidation and is still increasing. Nevertheless
the decreasing CO2 formation rate is still much higher than the value before oxidation.
The surface area increase also contributes to the apparent rate enhancement. After the
total (17.7%) carbon conversion during gasiﬁcation and oxidation, the BET area

increases about 50% (from 800 to 1200 mZ/g).

 

5 Huttinger, K]. and Merdes, W.F., Carbon 30(6), 883 ( 1992).

123

total (17.7%) carbon conversion during gasiﬁcation and oxidation, the BET area increases

about 50% (from 800 to 1200 m2/g).

 

 

   

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.E .
g 0.06 'E 0-5
2 0.05 — :1 0.5 1—
c a 5 C
.3 0.04 — g 0.4 p
3 003 — “3 03 L
g ‘ + CH4 T; ‘ I—
cu m
0.02 3 '2 m
g 0.01 — U 0.1 _.
a ‘ llllllllllll llllllllllllllllll .8 llllllllllh lllllllILlJLllJLJJ
g 0.00 A 9 0.0 A
O. 0 50 100 200 250 300 350 °- 0 50 100 200 250 300 350
.E .
E .5
:1 E 0.6 ,
2 :2 0.5 _ ~o— coz —e— CH4
c 2 1 CO
.9 g 0.4 _
u .
g g 0.3 —
o 5 0.2 _ d K. I
‘33) *3 0.1 P— + I vTr
E 0.06 ”11111111111171“?
0. 0 50 100 200 250 300 350 °- 0 50 100 200 250 300 350
Time (minutes) Time (minutes)

Figure 6-7. Rate enhancement by oxidation.
Conversion: 7.0% (gasiﬁcation 1), 14.4% oxidation) and 18.6% (gasiﬁcation 2).
gasiﬁcation in 3.1 MPa steam at 850 °C; oxidation at 450 °C for 70 min[9637]
a) CH4 formation; b) CO2 formation; C) CO formation; d) overall reaction
Results of the 850 °C steam gasiﬁcation rate enhancement experiment (Figure

6-7) is much different from the results of 725 °C gasiﬁcation. The formation rate of CO
after oxidation is higher than before oxidation, and the rate seems quite stable. The CO2
formation rate is slightly enhanced, but returns to the value before oxidation after a while.
From Figure 6-2 and Figure 6-3, we know that most adsorbed surface oxygen groups can
not survive at the temperatures higher than 850 °C. The high reaction temperature will

lead to very little oxygen groups deposition on the carbon surface during oxidation. So,

enhanced methane rate and CO formation rate are limited at high temperature

124
surface area about 42% (from 950 to 1350m2/g), Which roughly agrees with the apparent

rate enhancement of methane and CO.

The initial higher CO2 formation rate probably comes from the limited adsorbed
oxygen and the active carbon atoms, which are formed during the oxidation. We can see
that the rate quickly decreases back to the value before oxidation. It seems that the
surface area increase did not affect the CO2 formation! It is highly possible that this
amount of CO2 is only from the shift reaction. Unfortunately, the surface reaction
combined with the shift reaction make this phenomenon more complicated. From these

limited experiments, we can not ﬁgure out the real reason.

125
6-4. H2/Steam Gasification Rate Enhancements

The steam/H2 gasiﬁcation at 725 °C is more interesting (Figure 6-8). The CO rate
change is similar to steam gasification at 725 °C. The CO rate is very low at the

beginning of the second gasiﬁcation, then increases. The enhanced CO2 rate continues for
I a limited time and then returns to the level before oxidation. This is further evidence of
hydrogen and oxygen in a competitive adsorption. Hydrogen is more easily to adsorbed
on the carbon surface. Because H2 exists in the gas phase, the surface oxygen groups are
quickly replaced by adsorbing hydrogen. The methane rate enhancement looks more like

pure H2 gasiﬁcation, as the rate is about twice as before oxidation.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

d r.‘

E 0'4 "€0.12

g 0 g

i ' a l -' °"' 4 °
° '5

'2” T l €0.06 .T__ “V,

0.1 .03 _ FHI
2 go la 1
30 3041111111111191ﬁﬁ 1111I
°- °- 40 90 140 240 290 340
€0.18 £03
2 b g, —e— 002 20.24 — + CO
€0.12 ._ g o coz
'§ .13; €0.10 __ _9_ CH4
$0.06 1 o $0.12 LL.— 5
g (1; N U) .06 - .+__ I ’39:.29v .‘
lg 02:31111111NH'11111L111 gowrﬁilg I'I- ”cm
(L 7 a. 7

 

40 90 140 240 290 340 40 90 140 240 290 340

Time (minutes) Time (minutes)
Figure 6-8. Rate enhancement by oxidation.
Conversion: 2.1% (gasiﬁcationl), 1.6% oxidation) and 8.1% (gasiﬁcation2). Gasiﬁcation
of annealed Saran char' in 3.1 MPa H2/H2O at 725 °C; oxidation at 450 °C for 60 min
a) CH4 formation; b) CO2 formation; C) CO formation; (1) overall reaction .

126
6-5. Conclusions

Rate enhancement by molecular oxygen itself is meaningless, since the main
products are CO2. But the limited number of rate enhancement experiments do give us
some detail information about the Saran char gasiﬁcation.

o The 725 °C and 850 °C steam/H2 gasiﬁcation is much different. At the low
temperature steam gasiﬁcation, surface oxygen groups are very important for
gasiﬁcation. So, the history of the char (which will decide the surface oxygen) will
strongly affect the gasiﬁcation rate. But at high temperature, the oxygen groups can
not survive. The oxidation process has very limited affect on the subsequent
following gasiﬁcation.

0 Comparing Figure 6-6 and Figure 6-7, we know that the enhanced gasiﬁcation rate
at low temperature (725°C) can match the rate of high temperature gasiﬁcation
(850°C). The methane formation rate also has same trend.

9 Oxidation by molecular 02 will not only react with surface hydrogen, it also will react
with surface carbon. This will change the surface structure. The formation of new
arm-chair edges, the increased surface area, and the uptake oxygen groups all will
contribute to the enhancement of the H2/steam gasiﬁcation rate. By these limited

experiments, it is impossible to give a universal mechanism.

CHAPTER 7

CONCLUSIONS AND
RECOMMENDATIONS

128

The results of these studies lead to some important conclusions about the
hydrogen deposited during carbon gasiﬁcation. However, there are still many unanswered
questions, and some recommendations are thus made within the framework of the present
study to resolve some of the relevant questions.
7-1.Conclusions
7-1-1.Hydrogen Stability after Char Gasification

The adsorbed hydrogen (or Deuterium) is quite stable on the char surface.
Molecular oxygen can remove adsorbed hydrogen with carbon burnt off. Carbon atoms
and adsorbed hydrogen atoms have the same chance to be oxidized, and no preference is
found in the oxidation process. Weakly bound hydrogen tends to be removed ﬁrst. Some
hydrogen may migrate into the pore structure, where even severe oxidation condition can
not remove it.

Adsorbed D (H) can be exchanged during the following H2(D2) gasiﬁcation. The
exchange process has no preference for adsorbed D(H) with different binding energies.
From H-D exchange, we have conﬁrmed that the adsorbed hydrogen is in a dynamic state

during gasiﬁcation and continuously exchanges with hydrogen in the gas phase.

7-1-2.Gasiﬁcation and Hydrogen Inhibition

Hydrogen inhibits the steam gasiﬁcation of carbons via different mechanisms at
different extents of conversion. At very low conversion, or in environments where the
quantity of hydrogen is limited, dissociative adsorption is observed to inhibit steam
gasiﬁcation at ordered edge sites on the char. At higher hydrogen partial pressures, the
effect of adsorbed hydrogen is still present but can not be observed because it saturates

the edge sites on the char surface; steam gasiﬁcation rate is observed to be inhibited by

129

reverse oxygen exchange which is dependent on hydrogen partial pressure in the reactant
gas. Because of the unstability of “associative” adsorbed hydrogen, it is ruled out as an
inhibition pathway in high temperature gasiﬁcation (850 °C).

7-1-3. Rate Enhancement by Molecular Oxygen Oxidation

At the low temperature steam gasiﬁcation, surface oxygen groups are important
for gasiﬁcation. So, adding oxygen groups by oxidation to the char surface will
signiﬁcantly enhance the subsequent gasiﬁcation rate. But at high temperature, the
oxygen groups can not survive, so the oxidation process has a very limited effect on
subsequent gasiﬁcation.

Oxidation by molecular 02 will not only react with surface hydrogen, it also will
react with surface carbon. This will change the surface structure. The formation of new
arm-chair edges, the increased surface area, and the uptake of oxygen groups all will
contribute to the enhancement of H2/ steam gasiﬁcation rate.

7-1-4.lsotopic Effect

The isotopic effect have been seen on the D-H and H-D exchange, D2 gasiﬁcation
and the D20 gasiﬁcation. H-D exchange is much faster than D-H exchange. At the same
reaction conditions, deuterium gasiﬁcation rate is about one fourth of H2 gasiﬁcation

rate;. the gasiﬁcation rate of D20 is about three fourth of the steam gasiﬁcation rate.

130

7-2.Recommendations

There are still some speculations involved in the interpretation of the stability of
adsorbed H(D). Experiments have shown that H is easily exchanged by D, but D is
exchanged by H only with difﬁculty. We know that the mass difference will affect
adsorption and diffusion ability, but the detail is unclear yet. For further work, molecular
orbit theory is needed to simulation the adsorption on carbon surface.

The rate enhancement is very meaningful to industrial application. We got some
preliminary results, but the real mechanism is unclear yet. For the three possible source of
enhancement: removing of adsorbed hydrogen, surface area change and change the
carbon surface structure, we do not know which is the dominant factor. Further
investigation is advisable by other surface analysis technique.

Hydrogen desorption energies distribution is another speculation. Armchair and
Zigzag are two possible sites for the adsorption of hydrogen, so at most two adsorption
energy exist. Because the TPD process is a combination of surface desorption reaction
and the hydrogen migration from bulk carbon to the surface, it is difﬁcult to determine
the energy distribution from the TPD proﬁle. For further identifying, details of
mechanism of desorption and adsorption of hydrogen on carbon, probably graphite (with
real surface adsorption) is need to measure the desorption of hydrogen from the carbon,

One can then use desorption theory to get the real desorption energy.

131

LIST OF REFERENCE

~

Steinberg, M., 1987 Int. Conf. on Coal Sci. 11, J. A. Moulijn, ed., 953 (1987).

Redmond, J. P. and Walker, P. L. Jr., J. Phys. Chem. 64, 1093 (1960).

Biederman, D.L. et al., _C_a£bo_r_1 14, 351 (1976).

Yang, R.T. and Yang, K.L., Car—bog 23(5), 537 (1985).

Huttinger, KJ. and Merdes, W.F., C_arb_o_1_1_ 30(6), 883 (1992).

Yang, RT and Duan, R.Z., C_arbﬂ 23(3), 325 (1985).

E,J Hippo, N murdie and A. Hyjiazie, C_arbﬂ 27, 689 (1989)

Treptau, M.H. and Miller, D.J., 9m 29, 531 (1991).

Blackwood, J.D. Aust. J. Chem. l2, 14 (1959).

10 Cao, .l.R., and Back M.H., §a_rb0323, 141 (1985).

11 Walker, gm, Vol. 29, P4l,l99l

12 Pan, Z.,J and Yang, R. T. Ind. Eng. Chem. Res. 1992,3l,2675

13 Chen and Yang, Ind. Eng Chem. Res.,32,2835, 1993

14 Zielke, CW. and Gorin, E., Ind. and Eng. Chem. 47(4), 820 (1955).

15 Ergun, S and Mentser, M In Chemistry and Physics of Carbon (by P.L.Walker. Jr) Vol. 2 P203. 1966
16 Long, F.J., and Sykes, K.W., Proc. Royal Soc. London A193, 377 (1948).

17 Gadsby, J., Hinshelwood, C.N., and Sykes, K.W., Proc. Royal Soc. London A187, 129 (1946).
18 Huttinger, K.J., ACS Div. Fuel Chem. Prepr. 34(1), 56 (1989).

19 Blackwood, DJ. and McGrory, F ., Aust. J. Chem. 11, 16(1958).

20 Montet, G.L., and Myers, G.E., C3329; 9, 673 (1971).

21 Yates, J.T.Jr. and McKee, D.W., J. Chem. Phys. 75, 221 1, ( 1981).

22 Mims, CA. and Pabst, J.K., ACS Div. Fuel Chem. Prepr. 25(3), 263 (1980).

23 Gulbransen, E.A., l_\l_at_u_re 212, 1420 (1966).

24 Ishikawa, Y. et al., Chem. and Phys. of Carbon 12, Walker, P.L.Jr. and Thrower, P.A., ed., 40 (1975).
25 Ozaki, A., Isotopic Studies of Heterogeneous Catalysis, New York, Academic Press, Inc. (1977
26 Michael G. Lussier etc. Car—born Vol. 32 N08, Ppl493-1598,l994

27 Michael G. Lussier, Ph.D. dissertation Michigan State University (1997)

28 Gunter Hermann and Klaus J. Huttinger, @093 Vol. 24, No. 6. PP 705-713,1996

29 P. A. Redhead ,v__acu_um_, 12, 203 (1962)

30 Du, Z., A.F Saroﬁm, and J. P. Longwell, Energy & fuels 4,296 (1990).)

31 Blackman, Modern Aspects of graphite Technology, p6,(l970).

 

VOQQO‘MDWN

 

 

 

 

 

 

 

 

nrcnran smrE UNIV. LIBRARIES
llll111ll11111111111111|111111111111
31293017126560