Anda di halaman 1dari 43

Catalysis Reviews

Science and Engineering

ISSN: 0161-4940 (Print) 1520-5703 (Online) Journal homepage: http://www.tandfonline.com/loi/lctr20

CO2 Reforming of CH4

M. C. J. BRADFORD & M. A. VANNICE

To cite this article: M. C. J. BRADFORD & M. A. VANNICE (1999) CO2 Reforming of CH4 ,
Catalysis Reviews, 41:1, 1-42, DOI: 10.1081/CR-100101948

To link to this article: https://doi.org/10.1081/CR-100101948

Published online: 03 Feb 2007.

Submit your article to this journal

Article views: 4818

View related articles

Citing articles: 806 View citing articles

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=lctr20
CATAL. REV.—SCI. ENG., 41(1), 1 – 42 (1999)

CO2 Reforming of CH4


M. C. J. BRADFORD* and M. A. VANNICE

Department of Chemical Engineering


Pennsylvania State University
University Park, Pennsylvania 16802-4400

I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
II. REACTION CHEMISTRY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
III. CH4 ACTIVATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
IV. CO2 ACTIVATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
V. CARBON DEPOSITION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
VI. KINETICS OF CO2 REFORMING OF CH4 . . . . . . . . . . . . . . . . . . . . 13
VII. CATALYTIC MECHANISMS AND REACTION MODELING . . . . 25
VIII. SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

I. INTRODUCTION

Although technological practice should minimize environmental impact,


this is not always economically feasible. During the past decade, for example,
there has been increasing global concern over the rise of anthropogenic CO2
emissions into the Earth’s atmosphere, estimated to be around 2 ⫻ 1015 g of
carbon per annum [1]. Although the precise CO2 emission flux is uncertain,
there are several different indicators which raise the possibility that anthro-

*Present address: Exxon Chemical Co., 4500 Bayway Drive, Baytown, TX 77520-9728.

Copyright 䉷 1999 by Marcel Dekker, Inc. www.dekker.com


2 BRADFORD AND VANNICE

pogenic greenhouse gas emissions are causing a global problem. They include
a correlation of mean global temperature rise with atmospheric CH4 and CO2
concentrations [2], drastic changes in the annual temperature cycle of the
Northern Hemisphere [3], and increasing volatility of global weather patterns
[4]. Consequently, there has been increased interest in a better understanding
of CH4 and CO2 removal, disposal, and utilization as well as the influence of
these gases in the atmosphere. Today as well as historically, however, the
principal interest has been the study of the reaction between CO2 and CH4 to
produce synthesis gas (i.e., CO ⫹ H2), which can be used in chemical energy
transmission systems [5–8] or utilized in the Fischer–Tropsch reaction to
produce liquids [9,10]). The latter approach would be especially useful at
remote natural gas fields containing large amounts of CO2 and at petroleum
fields where natural gas is now being flared because liquids can be transported
less expensively than gases. Reforming with CO2, rather than H2O, is attrac-
tive because it can be employed in areas where water is not available and it
yields syngas with lower H2/CO ratios, which is a preferable feedstock for
the Fischer–Tropsch synthesis of long-chain hydrocarbons [11]. In other sit-
uations, a combination of CO2 and steam reforming may be advantageous
[12]. Indeed, a limited number of industrial processes are already in operation
which utilize the CO2 CH4 reforming reaction [13–16]. Nevertheless, it is
very doubtful whether transformation of CH4 and CO2 into other chemicals
could have a significant impact on the concentration of these gases in the
atmosphere.
CH4 and CO2 are relatively inexpensive due to their natural abundance;
hence, conversion of these two molecules to higher-value compounds is of
interest. The formation of acetic acid from these two gases over Pd and Cu
catalysts has been reported [17], and the oxidative coupling of CH4 to produce
C2 hydrocarbons using CO2 as the oxidant has been observed, although yields
are very low [18–20]; regardless, the preeminent reaction to convert these
two reactants is CO2 reforming of CH4, which was thoroughly explored by
Fischer and Tropsch in 1928 [22] and was investigated as early as 1888 [23].
To initiate our discussion of this topic—CO4 reforming of CH4 —the
chemical reactions most important to this process are briefly reviewed. As-
pects related to the activation of CH4 and CO2 will be covered next, followed
by a discussion of the problem of carbon deposition and deactivation. The
kinetics associated with CO2 reforming of CH4, obtained from the literature,
are then examined and analyzed. Finally, the status of reaction mechanisms
and kinetic modeling is evaluated and summarized.

II. REACTION CHEMISTRY

The reaction equilibrium for the production of synthesis gas from CH4
and CO2,
CH4 ⫹ CO2 ` 2H2 ⫹ 2CO (⌬H⬚ = 59.1 kcal/mol CH4) (1)
CO2 REFORMING OF CH4 3

FIG. 1. (A) Equilibrium conversions of CO2 (䡩) and CH4 (●) and (B) product
ratios of H2 /CO (▫) and H2O/CO (䡲) for simultaneous CO2 CH4 reforming and
RWGS reactions as a function of temperature. Reaction conditions: Ptot = 1 atm;
CH4 /CO2 /He = 1/1/1.8.

is typically influenced by the simultaneous occurrence of the reverse water–


gas shift (RWGS) reaction

CO2 ⫹ H2 ` CO ⫹ H2O (⌬H⬚ = ⫹9.8 kcal/mol) (2)

which results in H2 /CO ratios less than unity. The equilibrium conversions
of CH4 and CO2 for this set of reaction equilibria for a feed stream at 1 atm
total pressure (CH4 /CO2 /He = 1/1/1.8), calculated using the program ARL
SOLGASMIX [24], are shown in Fig. 1. The CO2 conversion is always
greater than that of CH4, due to the RWGS equilibrium. In practice, this is
advantageous for producing gas streams with a H2 /CO ratio of 1 or lower.
For example, to produce alkanes directly from synthesis gas, the following
reaction stoichiometry is observed [11]:
4 BRADFORD AND VANNICE

nCO ⫹ (2n ⫹ 1)H2 → CnH2n⫹2 ⫹ nH2O (3)


and the required H2 /CO feed ratio is (2n ⫹ 1)/n, which is never less than 2.
However, if the water–gas shift (WGS) reaction
H2O ⫹ CO → CO2 ⫹ H2 (4)
occurs simultaneously to completion, the overall reaction stoichiometry for
the production of alkanes is
2nCO ⫹ (n ⫹ 1)H2 → CnH2n⫹2 ⫹ nCO2 (5)
and the required H2 /CO feed ratio is (n ⫹ 1)/2n, which is less than or equal
to unity. In the presence of the WGS equilibrium, the steam reforming of
CH4, i.e.,
CH4 ⫹ H2O ` 3H2 ⫹ CO (⌬H⬚ = ⫹49.3 kcal/mol CH4) (6)
yields H2 /CO ratios greater than or equal to 3. The partial oxidation of CH4,
i.e.,
CH4 ⫹ –O
1
2 2 ` 2H2 ⫹ CO (⌬H⬚ = ⫺8.5 kcal/mol CH4) (7)

yields a H2 /CO ratio of 2. Consequently, if H2 /CO ratios less than or equal


to unity are desired (e.g., for the production of long-chain alkanes), then CO2
reforming of CH4 is preferable.
During the past decade, many research groups have investigated the
concept of ‘‘mixed’’ reforming; that is, the simultaneous reaction of CH4 with
CO2 [reaction (1)], H2O [reaction (6)], and O2 [reaction (7)], which has been
considered because it has several practical advantages over CH4, reforming
with either CO2 or H2O alone [12,15,16,25]. First, H2 /CO ratios ranging from
about 1 to 3 can be produced via adjustment of the CO2 /H2O/O2 feed ratio.
Second, as CH4 partial oxidation is mildly exothermic, the addition of O2 can
minimize the energy requirements for either CH4 CO2 reforming or
CH4 H2O reforming. Third, the formation of solid carbon, C(s), during CO2
reforming of CH4 may occur either via CH4 decomposition,
CH4 ` 2H2 ⫹ C(s) (⌬H⬚ = ⫹17.9 kcal/mol C(s)) (8)
or CO disproportionation (i.e., the Boudouard reaction),
2CO ` CO2 ⫹ C(s) (⌬H⬚ = ⫺41.2 kcal/mol C(s)) (9)
as discussed later. Addition of either H2O or O2 to the feed inhibits carbon
formation either via gasification,
H2O ⫹ C(s) ` CO ⫹ H2 (⌬H⬚ = ⫹31.4 kcal/mol C(s)) (10)
or oxidation,
xO2 ⫹ C(s) ` 2(1 ⫺ x)CO ⫹ (2x ⫺ 1)CO2
(⫺94.1 ⱖ ⌬H⬚ ⱖ ⫺26.4 kcal/mol C(s)) (11)
CO2 REFORMING OF CH4 5

Nevertheless, the focus of this review is CO2 reforming of CH4 in the absence
of either H2O or O2 in the feed.

III. CH4 ACTIVATION

The predominant mechanism for CH4 adsorption and dissociation on


transition metal surfaces has been claimed to be both direct [26–28] and
precusor mediated [29]. Consequently, it has been the topic of a debate which
may now have been resolved by the results of Seets et al., which show that
CH4 dissociation over Ir(110) gradually shifts from a precursor-mediated
mechanism at low temperatures to a direct dissociative mechanism at high
temperatures [29]. Ceyer et al. studied the interaction of CH4 with a Ni(111)
surface and suggested that in order to dissociate, CH4 must be distorted from
its tetrahedral shape to form a trigonal pyramidal structure, after which tun-
neling of a H atom through the activation barrier occurs [30]. Conversely,
van Santen and Neurock claim that the activation barrier for CH4 dissociation
on Ni does not involve molecular distortion and that it is dependent only on
the tunneling of a H atom through the activation barrier for H abstraction
[31]. Luntz and Harris utilized this argument to illustrate that the tunneling
of a H atom through the activation barrier during H abstraction is dominated
by a dynamic coupling to the vibrations of the metal lattice, giving rise to
thermally assisted tunneling [28]. Nevertheless, neither a quantitative model
nor a general consensus exists concerning the mechanism for CH4 adsorption
and dissociation on transition metal surfaces.
Simple extended Hückel molecular orbital (EHMO) calculations with
CH4 show that there are four filled bonding MOs and four empty antibonding
MOs [32]. Thus, it may be expected intuitively that electron donation from
the highest occupied MO of the metal, HOMOM, should dominate dissociative
CH4 adsorption. However, Trevor et al. have investigated the gas-phase re-
action of CH4 with unsupported Pt clusters and have shown that as the size
of the cluster increases, the cluster ionization potential decreases (and hence
the HOMOM energy level increases) while activity decreases; consequently,
they reasoned that charge donation from the HOMO of CH4 to the lowest
unfilled MO (LUMO) of the Pt clusters was important [33]. Kuijpers et al.
have shown that CH4 decomposition preferentially occurs on small Ni crys-
tallites [34], whereas Beebe et al. have demonstrated the structure sensitivity
of CH4 dissociation on Ni surfaces, with activity decreasing in the order
Ni(110) > Ni(100) > Ni(111) [35]. This structure sensitivity may be a con-
sequence of a change in either the metal work function (and hence the ion-
ization potential and HOMOM ) [36] or the geometric site distribution on dif-
ferent metal surface structures. Indeed, theoretical investigations of CHx
species (1 ⱕ x ⱕ 3) adsorbed on Pt(111) [37], Rh(111) [38], Ti(0001),
Cr(110), and Co(0001) [39] indicate that CHx fragments are preferentially
located at a site on the metal surface which completes its tetravalency; that
6 BRADFORD AND VANNICE

is, the stepwise decomposition of CH4 into CHx fragments on a metal surface
requires the concomitant occupation of higher coordination sites:
CH4 ⫹ 2M ` CH3 ⫺ M ⫹ H ⫺ M (12)
CH3 ⫺ M ⫹ 2M ` CH2 ⫺ M2 ⫹ H ⫺ M (13)
CH2 ⫺ M2 ⫹ 2M ` CH ⫺ M3 ⫹ H ⫺ M (14)
CH ⫺ M3 ⫹ 2M ` C ⫺ M4 ⫹ H ⫺ M (15)
where M␩ is an ensemble of ␩ surface metal atoms. This requirement dem-
onstrates the high probability of structure sensitivity in the formation of sur-
face carbon because of the large ensemble of metal atoms needed. Note that
the stoichiometry of reaction (15) is valid provided that the metal surface in
question has fourfold sites, such as Pt(100).
CHx species on transition metal surfaces have been detected using both
transient CH4 /D2 and CD4 /H2 exchange [40,41] and deuteration of preadsor-
bed CH4 [42]. However, simple model calculations based on –CHx bond
strengths on metal surfaces show that the distribution of CHx species observed
with the latter technique may be controlled by thermodynamic equilibrium
and not kinetics [43]; consequently, transient methods may be a better choice
to resolve the actual CHx distributions present on working metal surfaces. In
this regard, Osaki et al. have utilized pulse surface reaction rate analysis
(PSRA) to determine in situ the CHx intermediates formed during CO2 re-
forming of CH4 over Co/Al2O3 [44] and supported Ni [45], and the results
are in Table 1. Erkelens and Wösten used magnetic measurements to show
that between four and five surface bonds are formed per molecule of CH4
chemisorbed on a Ni/SiO2 catalyst at 298 K [46]. This suggests that x is 1
or less for Ni/SiO2, consistent with the value of x = 1 obtained by Osaki et
al. [45]. Matsumoto observed different CHx species during the adsorption of
hydrocarbons on different Ni catalysts and determined that CHx intermediates

TABLE 1
CHx Species Observed on
Ni and CO During CO2
Reforming of CH4
Catalyst x
Ni/MgO 2.7
Ni/ZnO 2.5
Ni/Al2O3 2.4
Ni/TiO2 1.9
Ni/SiO2 1.0
Co/Al2O3 0.75
Source: Data from Refs.
44 and 45.
CO2 REFORMING OF CH4 7

with lower values of x were more likely to form carbonaceous deposits [47].
Consequently, Takayasu et al. have suggested that hydrogen atom spillover
onto the support during CH4 reforming helps to minimize inactive carbon
deposition due to CH4 decomposition, by shifting the equilibrium of reactions
(12)–(15), thereby stabilizing CHx intermediates on the support surface [48].
The decomposition of CH4 on supported transition metals has been stud-
ied extensively and, in general, yields surface carbon, H2, C2H6, and C2H4
[49–54]. The formation of C2 hydrocarbons implies the coupling of CHx
species; however, it has been reported recently that CH4 decomposition on
both SiO2- and Al2O3-supported transition metals yields CO at temperatures
exceeding 600 K via an interaction with the support, thus allowing the pos-
sibility of syngas reactions [55]. Turnover frequency (TOF) values for CO
formation from CH4 at 723 K calculated from their data are listed in Table
2. To facilitate comparison of the TOFs obtained by Ferreira–Aparicio et al.
[55] with those reported during CO2 reforming of CH4, their data were ex-
trapolated to a CH4 partial pressure of 380 Torr assuming a first-order reaction
in CH4; consequently, these TOFs are upper-limit estimates. With the excep-
tion of Ir/Al2O3, the TOFs for CO production during CO2 reforming of CH4
are higher than those observed during CH4 decomposition; however, the dif-
ferences are less than an order of magnitude. Thus, it seems plausible that
during CO2 CH4 reforming, CHx species formed on a transition metal sur-
face may react with either oxygen species or hydroxyl groups on the support
to yield CO and H2.
Although the precise mechanism by which dissociative adsorption of

TABLE 2
TOFs for CH4 Decomposition on SiO2- and Al2O3-Supported Metals Compared
with TOFs for CO2 Reforming of CH4 (at 723 K)
CH4 decompositiona CO2 CH4b
TOFCO (s⫺1) TOFCO (s⫺1)
Catalyst PCH4 = 53 Torr PCH4 = 380 Torrc PCH4 = 380 Torr
Ru/Al2O3 0.018 0.13 0.18
Ru/SiO2 0.021 0.15 —
Rh/Al2O3 0.009 0.064 0.18
Rh/SiO2 0.005 0.036 0.15
Ir/Al2O3 0.100 0.72 0.02
Ir/SiO2 0.011 0.079 —
Pt/Al2O3 0.013 0.093 0.18
Ni/Al2O3 0.016 0.11 0.72
Co/Al2O3 0.054 0.39 —
a
Data from Ref. 55.
b
Data from Table 6.
c
Extrapolated assuming that TOFCO = kPCH4.
8 BRADFORD AND VANNICE

CH4 occurs on a metal surface is unknown, it is reasonable to expect that


CH4 decomposition is dependent on the HOMO, lowest unoccupied molecular
orbit (LUMO), and geometric structure of the metal surface, and it is very
likely to be a structure-sensitive reaction. Dissociation of CH4 under CO2 
CH4 reforming conditions yields a distribution of CHx species that depends
on both the metal and the support. In addition, the interaction of CHx species
with surface oxygen or hydroxyl groups on the support surface may lead to
both CO formation and inactive carbon deposition.

IV. CO2 ACTIVATION

It is generally accepted that CO2 chemisorption and dissociation on a


transition metal surface is dominated by electron transfer and requires the
formation of an anionic CO⫺ 2 precursor [56,57], and CO2 activation is reported
to be structure sensitive [56,57]. For example, Segner et al. performed scat-
tering experiments related to CO2 adsorption on Pt(111) and showed that CO2
undergoes, with equal probability, either direct inelastic scattering or trapping
and desorption (with an estimated heat of adsorption of 5 kcal/mol), i.e., there
was no detectable dissociation, and they suggested that CO2 dissociation
is promoted at defect sites [58]. For reference, a listing of single-crystal sur-
faces which do and do not dissociatively chemisorb CO2 is provided in Table
3. The reaction of CO2 with H2 has also been reported to be structure sen-
sitive. Nikolic et al. used voltametry and in situ Fourier transform infrared
(FTIR) spectroscopy to show that the electrochemical reduction of CO2 and
H2 is negligible on a Pt(111) surface [59]. Rodes et al. later confirmed that
electrochemical reduction of CO2 with H2 on Pt(111) to yield CO and H2O
via the reverse water–gas shift (RWGS) reaction is much less than that on
either Pt(100) or Pt(110), and they provided evidence that CO2 reduction is
enhanced by the presence of surface defects [60]. Román-Martinez et al.

TABLE 3
Structure Sensitivity of CO2 Adsorption
Metal Dissociative chemisorption Nondissociative adsorption
Fe (111), (100) (110)
Ni (110), (100)a (111), (100)a
Cu — (110), (100)
Rh (533), (711) (111), (100)
Pd — (111), (100)
Ag — (110)
Re (0001) —
Pt — (111)
a
Dissociation on Ni(100) has been reported both to occur [56] and not to occur [57].
Source: Data from Refs. 56 and 57.
CO2 REFORMING OF CH4 9

recently reported that CO2 hydrogenation over carbon-supported Pt is struc-


ture sensitive and concluded that corner atoms are the active sites [61].
Although the general observation appears to be that both CO2 dissoci-
ation and reduction are structure sensitive, the TOFs for the RWGS reaction
over MgO-supported metal catalysts are much greater than those for CO2
reforming of CH4 [49]. Although most kinetic investigations of CO2 CH4
reforming have shown that the RWGS reaction operates very close to ther-
modynamic equilibrium [48,62–65], some authors maintain that the RWGS
reaction is not quasi-equilibrated [66–68]. Nevertheless, in most kinetic
studies of H2OCH4 reforming, the WGS reaction is assumed to be at
thermodynamic equilibrium to facilitate analysis [69]. In practice, the quasi-
equilibrium of the RWGS reaction results in H2 /CO ratios becoming a func-
tion of feed conversion [64], as illustrated by the thermodynamic calculations
in Fig. 1 and recently reported results from our laboratory [70,71], an example
of which is provided in Fig. 2. The latter confirms that the RWGS reaction
is near equilibrium over a wide range of temperatures.

V. CARBON DEPOSITION

Numerous authors, including Reitmeier et al. [72], White et al. [73],


Sacco et al. [74], and Gadalla and Bower [11], have presented calculations
which predict the thermodynamic potential of graphic carbon deposition as a
function of operating conditions for gas mixtures containing CH4, CO2, H2,
and H2O. Conclusions drawn from these calculations typically suggest op-

FIG. 2. Extent of the reverse water– gas shift equilibrium as a function of


temperature for Ni/TiO2 (䉭), Ni/C (䊱), Ni/SiO2 (●), and Ni/MgO (䡩). Reaction con-
ditions: CH4 /CO2 /He = 1/1/1.8; P ⬇ 740 Torr. (From Ref. 70.)
10 BRADFORD AND VANNICE

eration at high temperatures, ⬃1000 K, and with CO2 /CH4 ratios far above
unity to avoid regions where there is a thermodynamic potential for carbon
formation; however, from an industrial standpoint, it may be desirable to
operate at lower temperatures with CO2 /CH4 ratios near unity. This necessi-
tates the use of a reforming catalyst which incorporates a kinetic inhibition
of carbon formation under conditions where deposition is thermodynamically
favorable.
As mentioned earlier, the origin of inactive carbon during dry reforming
may occur via either CH4 decomposition [reaction (8)] or CO disproportion-
ation [reaction (9)]. CO disproportionation is exothermic; thus, the equilib-
rium constant decreases with increasing temperature. Conversely, CH4 de-
composition is endothermic; thus, the equilibrium constant increases with
increasing temperature. The calculations of Reitmeier et al. [72] illustrate
that for any reaction mixture of H2, CO, H2O, CO2, and CH4 at thermody-
namic equilibrium, the extent of graphitic carbon deposition during reforming
decreases at higher reaction temperatures, in agreement with experimental
observations reported in the literature [68]. This result would suggest that the
main contributor to carbon deposition is CO disproportionation. Other evi-
dence indicates that CO disproportionation is primarily responsible for the
formation of inactive carbon deposits during CH4 reforming. The carbon
formed during the reaction is often in the form of filamentous whiskers [49],
and Rodriguez, in a literature review about carbon nanofiber growth, reports
that the rate-determining step for the formation of filamentous whisker carbon
is the diffusion of carbon through a metal particle [75]. The driving force for
this diffusion process is considered to be heat generated by exothermic sur-
face processes, such as CO adsorption and disproportionation.
From a compilation of surface science studies, it is possible to speculate
on the mechanism of CO dissociation and subsequent carbon fiber growth on
metal crystallites. Numerous studies suggest that CO dissociation on transition
metal surfaces is initiated via adsorption at a multiply coordinated site [76–
78]. This adsorbed CO species then proceeds through a bent transition state
lying essentially parallel to the surface prior to dissociation [31,76,79]. After
dissociation on Ni, for example, subsequent CO adsorption on the C/Ni sur-
face induces migration of carbon to subsurface Ni layers [80]. These adsorbed
carbon atoms induce local reconstructing of the Ni surface, thereby length-
ening nearby NiNi bonds and permitting a deeper carbon penetration into
the Ni lattice [81]. At this point, carbon diffusion through the metal lattice
may occur until the carbon atoms deposit in eventual graphitic layers on the
back side of the metal crystallite.
Tavares et al. reported interesting microscopy results of carbon deposits
formed from CO/CO2 and CH4 /H2 mixtures, although they did not provide
the data [82]. With the former mixture, the metal crystallites were rough and
the carbon deposits were of an encapsulating type, whereas in the latter mix-
ture, metal crystallites were well faceted and carbon deposits were ordered
and filamentous. These results are consistent with other reports of carbon
CO2 REFORMING OF CH4 11

morphology on Ni-containing catalysts formed via CH4 decomposition [83].


It has been suggested that carbon atoms, formed by CH4 decomposition on
the (100) and (110) Ni surfaces, diffuse across the Ni particle surface and by
deposition on the (111) surfaces form ordered graphite layers aligned parallel
to the metal–carbon interface [83]. The driving force for carbon diffusion in
this process has been assumed to be a carbon concentration gradient. To try
to clarify the issue of carbon deposition during CO2 reforming of CH4, Swaan
et al. [64] and Tsipouriari et al. [84] have independently used isotope labeling
and temperature programmed oxidation (TPO) to reveal that carbon deposits
originate from both CH4 and CO2, although primarily from CO2, suggesting
that CH4 decomposition is not the dominant carbon formation mechanism.
Carbon suppression on Ni catalysts by sulfur passivation has been com-
mercialized as the SPARG process [16]. The effectiveness of sulfur passi-
vation is attributed to control of ensemble size on the metal surface; that is,
catalytic promotion via selective poisoning works in practice because the
ensemble size necessary for carbon formation is larger than the ensemble size
for CH4 reforming [85], and sulfur preferentially eliminates larger ensembles,
thus inhibiting carbon deposition more than reforming [86]. Several recent
investigations of supported Rh [51,84], Pt [53,71], and Ni [87] catalysts have
shown that carbon deposition can be greatly suppressed when TiO2 is used
as a catalyst support. This is clearly illustrated by the results in Fig. 3 and
Table 4, which were obtained recently for a family of supported Pt catalysts.
The reduction in the rate of C deposition is evident, especially when ZrO2 is
used. A similar reduction in carbon deposition can be achieved by dispersing
TiO2 on the surface of ultrahigh purity (UHP) Pt powder and giving these
particles the same pretreatment (i.e., 1 h at 773 K under flowing H2); con-
sequently, a crystallite size effect does not appear to be an appropriate ex-

FIG. 3. TPH (temperature-programmed hydrogenation) spectra for Pt/TiO2,


Pt/ZrO2, Pt/Cr2O3, and Pt/SiO2 after reaction; conditions: H2 /He = 1/1, P ⬇ 740 Torr,
␤ = 10 ⫾ 1 K min⫺1. (From Ref. 71.)
12 BRADFORD AND VANNICE

TABLE 4
Carbon Deposition on Pt Catalysts Under Reaction Conditions T = 723 K,
P = 1 atm, CH4 /CO2 /He = 1/1/1.8
Time on stream Carbon removed by TPH
Catalyst (h) (C/Ptsurf)
0.79% Pt/SiO2 3 41
0.75% Pt/Cr2O3 3 28
0.82% Pt/TiO2 5 15 (3)a
0.31% Pt/ZrO2 73 9
a
Value in parentheses based on Ptsurf after reduction at 473 K.

planation [88]. The reason for this may well be analogous to that of sulfur
passivation (i.e., TiOx species decorate the metal surfaces and preferentially
eliminate large ensembles of metal atoms necessary for carbon deposition).
It has been suggested that carbon deposition is suppressed when the
metal is supported on a metal oxide with strong Lewis basicity [89–91]; i.e.,
increasing Lewis basicity of the support increases the ability of the catalyst
to chemisorb CO2 [92]. Increasing the concentration of adsorbed CO2 is sug-
gested to reduce carbon formation via CO disproportionation [reaction 10]
by shifting the equilibrium concentrations. However, Zhang and Verykios
reported that addition of a basic CaO promoter to Ni/␥-Al2O3 increased both
catalyst stability and carbon deposition [89]. In addition, the x-ray photo-
electron spectroscopy (XPS) results of Tang et al. also illustrated that the
addition of either MgO or CaO to Ni/␣-Al2O3 greatly increased both catalyst
basicity and carbon deposition during CO2 reforming of CH4 [93].
Alternatively, it is plausible that carbon deposition is more closely re-
lated to the catalyst structure. For example, Chen and Ren studied CO2 re-
forming of CH4 over Ni/Al2O3 and showed that carbon deposition is markedly
suppressed if NiAl2O4 is formed during the pretreatment procedure [94]. The
difference in apparent activation energies for the reduction of NiO (4.3 kcal/
mol) and NiAl2O4 (32 kcal/mol) is indicative of a relative strengthening of
the NiO bond in NiAl2O4 [95]. This strong interaction results in the for-
mation of primarily small Ni crystallites on the catalyst surface, which are
relatively stable toward sintering and carbon formation [96]. Prior to the study
of this spinel structure, Gadalla and Sommer had reported the improved per-
formance of solid solutions of NiO–MgO, which included extended activity
maintenance at high conversion and the absence of carbon deposition [67].
Since then, investigations of CO2 CH4 reforming over Ni/MgO catalysts by
Fujimoto and co-workers [97], Bradford and Vannice [70,87], and Rucken-
stein and Hu [98] have provided additional evidence that NiO–MgO solid
solutions can stabilize small Ni crystallites and enhance catalyst lifetime by
decreasing carbon formation.
Support acidity may also be important in regard to metal crystallite
CO2 REFORMING OF CH4 13

structure. Masai et al. studied Pd, Pt, and Rh dispersed on a number of sup-
ports and, based on NH3 measurements, reported that metal dispersion was a
strong function of the Lewis acidity of the support [99]. This relationship
was possibly due to the preference of metal atoms to reside at Lewis acid
sites on the support. Further evidence of this type of interaction comes from
the investigation of Roberts and Gorte, who studied Pt films on ZrO2 [100]
and ZnO [101] and provided evidence to indicate that the interaction of Pt
with Zrn⫹ and Znn⫹ cations influences both Pt morphology and sintering char-
acteristics. In addition, the relative resistance of Pt particles toward sintering
when supported on either TiO2 or Al2O3, as compared to either SiO2 or carbon
[102], may be attributed to Pt interactions with Lewis acid sites, although
subtleties exist in the Pt–TiO2 system, such as the formation of well-faceted,
quasi-two-dimensional structures [102]. The structure sensitivity of carbon
deposition on Ni crystallites during CO2 CH4 reforming has been addressed
recently, and on the basis of their transmission electron microscopy (TEM),
thermal gravimetric analysis (TGA), TPO, and TPH experiments, Kroll et al.
concluded that faceted and flat particles produce little or no filamentous car-
bon, whereas small, spherical particles produce encapsulating carbon [103].
Carbon formation during CO2 reforming of CH4 also depends on the
choice of metal. In general, it has been found that Ru, Rh, and Ir supported
on Eu2O3 [104], MgO [49,105], and Al2O3 [50,54,106] exhibit much less
carbon formation than supported Ni, Pd, and Pt. In addition, although Co
catalysts produce large amounts of carbon during reaction [107], as predicted
by the results of Sacco et al. [74], W wire exhibits very little carbon forma-
tion [63]. In contrast to bimetallic Pt–Au/SiO2 catalysts [43], Pt–Sn/SiO2 and
Pt–Sn/ZrO2 exhibit less carbon deposition during CO2 CH4 reforming than
the respective monometallic Pt catalyst analogs [108]. The reason for this
behavior is possibly due to Pt–Sn alloy formation and remains under inves-
tigation [108].
In summary, it appears that both CH4 decomposition and CO dispro-
portionation can contribute to the formation of inactive carbon deposits during
CO2 reforming of CH4, with the relative contribution of each depending on
the reaction conditions. Available data indicate that carbon formation is de-
pendent on several parameters, such as the metal, metal crystallite structure,
metal–support interactions, support acidity, and, possibly, the support
basicity.

VI. KINETICS OF CO2 REFORMING OF CH4

As shown in Table 5 for completeness, the reforming of CH4 with CO2


has been extensively studied using catalysts composed of transition metal
carbides and sulfides, unsupported metals, and supported Group VIII metals,
with the exception of Os [98,107–158,160–162,165,169]. However, the frac-
tion of these papers providing any fundamental kinetic parameters is not large,
14 BRADFORD AND VANNICE

TABLE 5
Catalyst Systems Investigated for CO2 Reforming of CH4
Metal Support References
Cu SiO2 110
Stuttgarter Masse 22
TiO2 109
Fe Al2O3 22, 111, 112
SiO2 109
TiO2 109
Co Al2O3 22, 42, 111, 112
SiO2 107
MgO 107
MgO/SiO2 107
C 107, 113
MgO/C 107, 113
ZSM-5 114
TiO2 109
Ni — 115
Al2O3 6, 11, 22, 45, 48, 62, 64, 89, 92 – 94,
96, 98, 111, 112, 116 – 128
SiO2 21, 45, 48, 70, 87, 92, 93, 98, 111,
112, 124, 128 – 134
Al2O3 – CaO 11, 89, 93, 123, 126
Al2O3 – CaO – MgO 93
Al2O3 – CaO – TiO2 66
Al2O3 – CeO2 126
Al2O3 – La2O3 119, 126, 128, 135
Al2O3 – MgO 48, 67, 93, 126, 128
Mg Al2O4 11, 49, 66
Al2O3 – SiO2 64, 136
BaO 137
C 70, 87
CaO 25, 89, 118, 123, 137, 138
CeO2 139, 140
MgO 45, 48, 49, 64, 67, 70, 87, 98, 103,
125, 128, 137, 141
MgO/CaO 91
MgO/SiO2 48, 124
MgO/SiO2 /Al2O3 136
MgO/Re2O3 /Al2O3 142
MgCO3 22
La2O3 64, 118, 119, 123, 134, 143
Nb2O5 43
SiC 144
SrO 137
TiO2 45, 64, 70, 87, 98b, 109, 125, 128,
145
ZnO 45
CO2 REFORMING OF CH4 15

TABLE 5 Continued
Metal Support References
ZrO2 64, 122, 146
Zeolites 92, 114, 147
Ru Al2O3 43, 62, 68, 116, 144, 148 – 152
C 43
CeO2 149
Eu2O3 104
La2O3 149
MgO 49, 105, 141, 149, 153
SiC 144
SiO2 144, 150, 154
Stainless steel 6
TiO2 109, 149
ZrO2 65
Rh Al2O3 51, 62, 68, 84, 103, 112, 114, 116,
129, 144, 148, 149, 152, 155,
156, 157, 160, 169
Al2O3 /SiO2 156
CeO2 151
Eu2O3 84, 149
La2O3 104
MgO 149, 157
SiC 49, 51, 84, 105, 141, 149, 153, 157
SiO2 144
TiO2 51, 84, 109, 114, 155, 157, 158, 159
TiO2 /SiO2 51, 84, 109, 149, 151, 155, 157
V2O5 /SiO2 151
ZrO2 158
ZrO2 /Y2O3 65
ZrO2 /SiO2 84, 151, 157, 160
Zeolites 151
114
Pd Al2O3 53, 62, 64, 99, 116, 129, 148, 151,
161
MgO 49, 53, 99, 105, 141, 153
MgO/SiO2 99
SiO2 53, 99, 109
TiO2 53, 99, 109
Zeolites 92, 99
ZrO2 65
Ir Al2O3 54, 62, 65, 116, 148, 149, 151, 152
CeO2 149
Eu2O3 104
La2O3 149
MgO 49, 54, 105, 149, 153
SiO2 54
TiO2 54, 109, 149
ZrO2 65 (continued )
16 BRADFORD AND VANNICE

TABLE 5 Continued
Metal Support References
Pt Al2O3 9, 62, 65, 99, 122, 126, 129, 148,
151, 160, 161, 162, 163
CeO2 164
Cr2O3 71
MgO 49, 105, 141, 153
NaY zeolite 92
SiC 144
SiO2 71, 108, 109
TiO2 9, 65, 71, 109, 162
ZrO2 9, 65, 71, 108, 122, 162, 165
Pt wire — 166
Pt powder — 88
Re Al2O3 152
W — 63
␣-WC — 167, 168
WS2 — 131
Mo Stuttgarter Masse 22
␤-Mo5C — 167, 168
MoS2 — 131
Ni– Cu Stuttgarter Masse 22
Ni– Rh CeO2 139, 140
Ni– Pt CeO2 139, 140
Ni– Rh-Pt CeO2 139, 140
Pt– Au SiO2 43
Pt– Sn SiO2 108
ZrO2 108

as indicated by Table 6, in which either reported values or those which could


be calculated from the data are tabulated for turnover frequency (TOF), ap-
parent activation energy, and partial pressure dependencies.
A wide range of activation energies, Ei, has been reported (i.e., values
for CH4 disappearance vary from 7 to 86 kcal/mol). In practice, a heteroge-
neous catalyst may have a distribution of catalytic active sites with different
intrinsic activation barriers for the chemical reaction of interest. Conse-
quently, it is interesting that the distribution of apparent activation energies
reported for CO2 reforming of CH4 in Table 6 shows that 14 ⫾ 1 kcal/mol
is the most frequently observed, a value which coincides with activation en-
ergies of 13.3 ⫾ 1.5 and 12.6 ⫾ 1.2 kcal/mol for CH4 dissociation on Ni(110)
and Ni(111), respectively [35]. An analysis of the reported values allows the
following observations. First, if the data reported by Erdöhelyi et al. for Pd/
MgO are excluded [53], the data in Table 6 illustrate that EH2 ⱖ ECO. In
addition, if the data for SiO2-supported catalysts are not considered, CO2
CO2 REFORMING OF CH4 17

inhibits H2 formation, as shown by the negative ␤ values in Table 6. These


two observations suggest that the apparent E values for CO and H2 are
strongly influenced by the reaction of CO2 with H2, presumably via the RWGS
reaction [reaction (2)]. Second, apparent Ei values can be strongly influenced
by the space velocity (SV), as shown in Fig. 4 for Pt/MgO catalysts. This
result shows that as the SV increases, CH4 and CO2 conversion decreases
along with the residence time, and the influence of the first part of the reverse
reaction (i.e., CO hydrogenation to CH4) becomes less significant. This prob-
lem, due to low equilibrium conversions at lower temperatures, can also affect
net rates and is discussed in greater detail later. Third, although Claridge et
al. did not themselves notice the phenomenon, their kinetic data obtained
over Al2O3-supported metals may illustrate a compensation effect [152]; that
is, their data can be correlated by the following relationship between the
apparent activation energy, E, and the preexponential factor, A:
E
ln A = ␣ ⫹ (16)
R␪
where ␣ is a constant, R is the gas constant, and ␪ is the isokinetic temper-
ature. With their data, it was found that ␣ ⬇ 0 and ␪ = 940 ⫾ 100 K so that
⫺1
ln A (s ) ⬇ 0.54E (kcal/mol) (17)
Such relationships are not uncommon [171], and a compensation effect has
also been observed for the reverse reaction, CO hydrogenation to CH4 [172].
The data in Table 6 show that the turnover frequency (TOF) for
CO2 CH4 reforming is dependent on the metal, as expected. Activation of
both CH4 and CO2 on a metal surface involves interaction with both the
HOMO and the LUMO of the metal, as discussed previously. An empirical
measure of the electronic structure of a metal is the %–d character [173];
consequently, it is of interest to determine if a correlation exists between
turnover frequency and %–d character. The kinetic data of Rostrup-Nielsen
and Bak Hansen [49] for both CO2 CH4 and H2OCH4 reforming over
MgO-supported metals are plotted as a function of metal %–d character in
Fig. 5. The TOF values for both reforming reactions over MgO-supported
metals do correlate with the %–d character, show a minimum in activity for
Pd and Pt, and provide evidence to support the conjecture by Rostrup-Nielsen
and Bak Hansen that the kinetics for both CO2 and H2O reforming of CH4
are similar [49]. More recent results have been reported for CO2 reforming
of CH4 over both SiO2-supported and TiO2-supported metals [109]. The for-
mer family, which uses a noninteracting support, produces a volcano-shaped
plot, whereas the latter family, which uses a support known to form TiOx
species which can migrate onto the metal surface, gives a correlation much
more similar to that shown in Fig. 5. The latter correlations are shown in
Figs. 6a and 6b. Consequently, there is strong evidence that the support util-
ized can have a significant effect on the overall catalytic behavior; this will
18
TABLE 6
Kinetic Parameters for CO2 Reforming of CH4; TOFs Extrapolated to Reaction Conditions of Ptot = 1 atm, CH4 /CO2 = 1, T = 723 K
Unless Otherwise Noted
TOFi (s⫺1) = A exp(⫺Ei /RT )P␣CH4P␤CO2
TOFi (s⫺1) Ei (kcal/mol) ␣ ␤
Catalyst CH4 CO2 CO H2 CH4 CO2 CO H2 CH4 CO2 CO H2 CH4 CO2 CO H2 Ref.
Co/Al2O3 — — — — — — 14 — — — — — — — — — 44
Co/C 0.06 0.10 0.16 0.10 11 9 9 11 — — — — — — — — 107a
Co/MgO/C — — — — 14 11 16 — — — — — — — — 107
Co/SiO2 0.16 0.26 0.42 0.22 10 8 9 11 — — — — — — — — 107a
Co/MgO/SiO2 — — — — 12 9 10 13 — — — — — — — — 107
Co/TiO2 0.05 — 0.15 — — — — — — — — — — — — — 43, 109a

Ni/Al2O3 — — 0.72 — — — 17 — — — — — — — — — 89, 143a


0.35 — — — 24 — — — ⫺0.3 — — — 0.2 — — — 90
— — — — 10 8 14 — — — — — — — — — 122
— — — — — — 20 — — — — — — — — — 45
— — — — 22 — — — 0.6 — — — 0.3 — — — 111

BRADFORD AND VANNICE


— — — — — — — — 0.55 — — — 0.35 — — — 170
— — — — 14 ⫾ 4 — — — 1.0 — — — 0.0 — — — 124
Ni/C — — 0.75 — 20 22 24 32 0.50 0.50 0.39 0.93 0.69 0.35 0.34 ⫺0.35 43, 70, 87
Ni/CaO/Al2O3 0.07 — — — 19 — — — 0.4 — — — ⫺0.6 — — — 90
— — — — 8 — — — — — — — — — — — 89
Ni/ZrO2 — — — — 14 10 12 — — — — — — — — — 122
Ni/SiO2 — — — — 13 — — — 0.03 — — — 0.5 — — — 129
0.61 — — — 13 — — — — — — — — — — — 131
— — — — — — 20 — — — — — — — — — 45
— — 0.52 — 23 21 20 27 0.44 0.27 0.18 0.49 0.15 0.64 0.44 0.11 70, 87
Ni/MgOSiO2 — — — — 10 — — — 0.8 — — — 0.0 — — — 124
CO2 REFORMING OF CH4
Ni/MgO — — 0.06 — 22 21 21 35 0.72 0.60 0.64 1.22 0.01 0.27 0.15 ⫺0.54 70, 87
0.46 — — — 28 — — — 1.0 — — — 0.0 — — — 49
— — — — — — 20 — — — — — — — — — 45
— — — — — — 15 — — — — — — — — — 141
Ni/CrOx /MgO — — — — 23 20 21 35 0.2 0.49 0.47 0.64 0.14 0.11 0.12 ⫺0.72 43
Ni/Nb2O5 — — 3.7 — — — 36 — — — — — — — — — 43a
Ni/ZnO — — — — — — 20 — — — — — — — — — 45
Ni/TiO2 — — — — — — 20 — — — — — — — — — 45
— — 2.7 — 26 21 23 32 0.40 0.43 0.44 0.90 0.18 0.41 0.30 ⫺0.15 70, 87
Ni/La2O3 — — 9.2 — — — 15 ⫾ 4 — — — — — — — — — 118, 143a
Ni foil — — — — 7 — — — 1.0 — — — — — — — 115

Ru/Al2O3 0.08 0.10 0.18 — 22 23 — — — — — — — — — — 50


0.63 — — — 13 — — — — — — — — — — — 152
1.5 3.0 4.5 1.5 26 18 20 30 — — — — — — — — 43a
Ru/C 0.10 0.21 0.31 0.10 26 21 22 36 — — — — — — — — 43a
Ru/Eu2O3 — — — — 19 ⫾ 1 — — — — — — — — — — — 104b
Ru/MgO 0.59 — — — 30 — — — 1.0 — — — 0.0 — — — 49
— — — — 12 — — — — — — — — — — — 105
— — — — — — 16 — — — — — — — — — 141
Ru/TiO2 4.3 7.2 11.5 5.7 18 17 18 23 — — — — — — — — 43, 109a

Rh/Al2O3 0.06 0.11 0.17 — 16 13 — — — — — — — — — — 50


— — 0.19 0.07 — — 16 18 — — 0.22 0.14 — — 0.43 ⫺0.6 51
— — — — 14 — — — 0.05 — — — 0.45 — — — 129
— — — — 21 — — — — — — — — — — — 155
0.17 — — — 10 — — — — — — — — — — — 152
0.08 — — — 22 — — — — — — — — — — — 151
— — — — — — — — 0.65 — — — 0.0 — — — 170
Rh/NaY — — — — — — — — 1.0 — — — 0.0 — — — 114
(continued )

19
20
TABLE 6 Continued
TOFi (s ) = A exp(⫺Ei /RT )P␣CH4P␤CO2
⫺1

TOFi (s⫺1) Ei (kcal/mol) ␣ ␤


Catalyst CH4 CO2 CO H2 CH4 CO2 CO H2 CH4 CO2 CO H2 CH4 CO2 CO H2 Ref.
Rh/SiO2 — — 0.15 0.08 — — 17 19 — — 0.22 0.38 — — 0.23 0.31 51
— — — — 20 — — — — — ⫺0.6 — — — 1.0 — 155
0.12 0.14 0.26 0.22 20 23 22 30 — — — — — — — — 43, 158a
Rh/VOx /SiO2 2.1 3.4 5.5 2.9 18 18 18 23 — — — — — — — — 43, 158a
Rh/MgO — — 0.34 0.18 — — 20 23 — — 0.33 0.16 — — 0.38 ⫺1.0 51
0.26 — — — 28 — — — — — — — — — — — 49a
— — — — 13 — — — — — — — — — — — 105
— — — — — — 16 — — — — — — — — — 141
Rh/TiO2 — — 0.21 0.11 — — 12 16 — — 0.29 0.24 — — 0.34 ⫺0.89 51
— — — — 15 — — — — — — — — — — — 155
15 25 40 20 21 18 19 26 — — — — — — — — 43, 109a
Pd/MgO — — 0.01 0.01 27 — 52 31 — — — — — — — — 53
— — — — 17 — — — — — — — — — — — 105
— — — — — — 18 — — — — — — — — — 141
0.03 — — — 27 — — — — — — — — — — — 49a
43, 109a

BRADFORD AND VANNICE


Pd/SiO2 0.23 0.34 0.57 0.35 28 23 24 31 — — — — — — — —
— — 0.04 0.03 34 — 34 34 — — 0.15 0.32 — — 0.12 0.24 53
Pd/Al2O3 — — 0.11 0.04 22 — 15 30 — — — — — — — — 53
0.07 0.10 0.17 — 20 22 — — — — — — — — — — 50
Pd/TiO2 1.1 2.3 3.4 1.0 18 15 16 25 — — — — — — — — 43, 109a
— — 0.79 0.19 22 — 15 35 — — — — — — — — 53
Ir/Al2O3 — — — — 21 14 16 23 — — — — — — — — 62
0.01 0.01 0.02 — 27 31 17 24 — — 0.5 0.78 — — 0.69 ⫺0.51 54
0.001 — — — 39 — — — — — — — — — — — 152
Ir/Eu2O3 — — — — 40 ⫾ 4 — — — — — — — — — — — 104b
CO2 REFORMING OF CH4
Ir/MgO 0.03 — — — 38 — — — — — — — — — — — 49a
— — — — 25 — — — — — — — — — — — 105
— — 0.11 — — — 34 52 — — — — — — — — 54
Ir/SiO2 — — 0.04 — — — 42 60 — — — — — — — — 54
Ir/TiO2 11 22 33 11 21 18 20 28 — — — — — — — — 43, 109a
— — 0.32 — — — 18 24 — — — — — — — — 54

Pt/Al2O3 0.06 0.12 0.18 — 15 12 — — — — — — — — — — 50


— — — — 14 13 — — — — — — — — — — 122
Pt/Cr2O3 0.42 1.08 1.5 0.18 16 15 16 19 — — — — — — — — 71c
Pt/ZrO2 — — — — 15 14 15 — — — — — — — — — 122
0.52 0.98 1.5 0.58 24 20 22 34 0.30 0.30 0.30 0.50 0.21 0.21 0.20 ⫺0.37 71
Pt/SiO2 0.45 0.85 1.3 0.50 15 19 17 — — — — — — — — — 71, 109c
Pt/TiO2 2.3 4.4 6.7 2.5 23 19 20 32 0.28 0.31 0.30 0.59 0.17 0.28 0.24 ⫺0.42 71, 109
Pt/MgO 0.03 — — — 35 — — — — — — — — — — — 49a
— — — — 33 — — — — — — — — — — — 105
— — — — — — 17 — — — — — — — — — 141
Pt wire — — — — 73 — — — — — — — — — — — 166
Pt powder — — 0.1 — 18 18 18 22 — — — — — — — — 88

Re/Al2O3 10⫺7 — — — 86 — — — — — — — — — — — 152


MoS2 0.01 — — — 18 — — — — — — — — — — — 131
WS2 0.001 — — — 26 — — — — — — — — — — — 131
a
TOFs calculated using TOF = kP 0.5 0.25
CH4PCO2.
b
TOFi calculated assuming 100% metal dispersion, which was not measured.
c
TOFs calculated using reaction orders for Pt/ZrO2.

21
22 BRADFORD AND VANNICE

FIG. 4. The influence of space velocity (SV) on apparent activation energy


for CO2 reforming of CH4 over Pt/MgO catalysts. (Data from Refs. 49, 105, and
141.)

be addressed in more detail later when reaction modeling is discussed. For


Rh crystallites of about 2.5 nm in diameter, Zhang et al. reported the follow-
ing influence of support on the initial TOF at 923 K: ZrO2 > TiO2 ⱖ Al2O3
> La2O3 ⬇ SiO2 > MgO [157]. Although this trend for supported Rh is
consistent with the results at 873 K reported by Nakamura et al. (i.e., TiO2
> Al2O3 > SiO2 [155]), it is inconsistent with the results at 773 K of Erdöhelyi
et al. (MgO > TiO2 ⬇ Al2O3 > SiO2 [51]), Basini and Sanfilippo at 1023 K

FIG. 5. Correlation of CH4 reforming turnover frequencies with % –d char-


acter for MgO-supported metals: CO2 reforming (䡩), H2O reforming (●). Reaction
conditions: T = 823 K, Ptot = 1 atm, H2O/CH4 /H2 = CO2 /CH4 /H2 = 10/2.5/1. WHSV
= 432,000 – 2,160,000 cm3 g⫺1 h⫺1. (Data from Ref. 49.)
CO2 REFORMING OF CH4 23

FIG. 6. Correlation of TOF for CO formation during CO2 reforming of CH4


with the metal % –d character. Reaction conditions: T = 723 K, Ptot = 740 Torr, CH4/
CO2 /He = 1/1/1.8, WHSV = 24,000 – 300,000 cm3 g⫺1 h⫺1. (a) SiO2-supported metals;
(b) TiO2-supported metals with TOFCO values calculated using metal surface atoms
determined after reduction either 473 K (䡩) or 773 K (●). (The curves shown are
meant only as a guide). (From Ref. 109.)

(Al2O3 > La2O3 > CeO2 > MgO > TiO2 [149]), and those of Mark and Maier,
who claim that the TOF from 873 and 973 K is independent of the support
(i.e., ZrO2 ⬇ TiO2 ⬇ Al2O3 ⬇ SiO2 [151]). A much smaller apparent E value
for Rh/MgO as compared to Rh on other supports could help reconcile the
temperature dependence of the observed differences; however, an inspection
of experimental E values in Table 6 shows that this is not the case. These
contradictory results have not yet been rationalized; however, one can surmise
that the degree of deactivation and the influence of the reverse reaction, es-
pecially the latter, are playing major roles in causing these differences; thus,
these differences may be attributable to a combination of physical phenomena
and experimental error.
Mark and Maier have shown that the activity of Rh/Al2O3 depends on
24 BRADFORD AND VANNICE

the space velocity at SVs lower than about 29,000 cm3 g⫺1 h⫺1, which they
attribute to external diffusion resistance [151] but may also be due in part to
proximity to equilibrium conversion. Our studies have also found the same
influence of SV on TOF for very active catalysts which can give conversions
approaching equilibrium conversions. An example of this is demonstrated in
Fig. 7 using VOx-promoted Rh/SiO2 catalysts, which are an order of magni-
tude more active than a typical Rh/SiO2 catalyst [43,158]. Consequently, as
Erdöhelyi et al. used a SV of only 6000 h⫺1 [51] and Basini and Sanfilippo
used SVs from 3000 to 11,000 h⫺1 [149], it seems quite possible that their
results are affected by the reverse (methanation) reaction. In addition, the
TOF calculations of Nakamura et al. are clearly inconsistent with the con-
versions, dispersions, and activities per gram catalyst reported in their paper
[155]; hence, the validity of their conclusions is questionable. Consequently,
the conflicting results of Mark and Maier [151] and Zhang et al. [157] are
likely to be due to differing extents of deactivation (i.e., carbon deposition),
although a Rh crystallite size effect cannot be excluded. For supported Pd
catalysts at 773 K, Erdöhelyi et al. [53] and Masai et al. [99] agree that on
a TOF basis, TiO2 > Al2O3 > MgO, SiO2, although they disagree on the
relative activities of MgO and SiO2. For supported Ni catalysts, Osaki et al.
conclude that on a TOF basis, TiO2 > Al2O3 ⬇ SiO2 ⬇ MgO [125], in agree-
ment with the results of Ruckenstein and Hu for Ni loadings near 13% [98].
Finally, the data of Zhang and Verykios illustrate that the TOF of Ni/La2O3
is much greater than that of Ni/Al2O3 [89,143].
The preceding analysis of results reported in the literature for CO2 re-
forming of CH4 over dispersed transition metals supports intuitive suspicions

FIG. 7. TOF for CO formation as a function of WHSV on Rh/SiO2 reduced


at 973 K (䡩), Rh/SiO2 reduced at 1173 K (●), Rh/VOx /SiO2 reduced at 973 K (䉭),
and Rh/VOx /SiO2 reduced at 1173 K (䊱). Reaction conditions: T = 723 K, P = 0.1
MPa, CH4 /CO2 /He = 1/1/1.8. (From Ref. 158.)
CO2 REFORMING OF CH4 25

that reaction kinetics can depend on both the metal and the support. Earlier
in this article, the role of the metal and the support in activating CH4 and
CO2 and in carbon deposition was discussed, and these analyses provide a
background for the subsequent discussion of proposed reaction mechanisms
and kinetic models for CO2 reforming of CH4.

VII. CATALYTIC MECHANISMS AND REACTION MODELING

Although the first proposal of a reaction sequence occurred 30 years


ago, most of our current insight into the reaction mechanism for CO2 reform-
ing of CH4 has been obtained during the past 4 or 5 years. The earliest
proposal of a reaction mechanism was probably in 1967 by Bodrov and
Apel’baum [115], who originally used the following model to describe results
of steam reforming of CH4 over a Ni foil [174]:
CH4 ⫹ * → CH2* ⫹ H2 (18)

`
CO2 ⫹ * O CO ⫹ O* (19)

O* ⫹ H ` O HO⫹*
2 2 (20)

CH * ⫹ H O`
2 O CO* ⫹ 2H
2 2 (21)

CO* ⫹` O CO ⫹ * (22)

In this proposed sequence of reactions, some of which are not elementary, *


denotes an active site on the catalyst surface, → denotes a slower irreversible
reaction, and Ò denotes a quasi-equilibrated reaction. The conceptual phi-
losophy of the mechanism suggested recently by Nakamura et al. [155] is
identical to this sequence. First, CH4 dissociatively adsorbs at an active site
in a rate-determining step to yield CH2 species. Subsequently, CO2 is con-
verted to H2O by the RWGS reaction. Finally, H2O reacts with CH2 species
to produce H2 and CO. Bodrov and Apel’baum reasoned that H2O reforming
of CH4 was more rapid than that of CO2 reforming because direct addition
of H2O to the feed was more efficient than producing it indirectly via reactions
(19) and (20) [115].
Erdöhelyi et al. [51,53] and Rostrup-Nielsen and Bak Hansen [49] have
recently proposed slight modifications to the Bodrov–Apel’baum mechanism.
The first modification allows for a distribution of CHx species to exist on the
catalyst surface, as discussed earlier; that is, step (18) is replaced by
CH4 ⫹ 2* → CH3* ⫹ H* (23)

CH3* ⫹ (3 ⫺ x)* ` CHx* ⫹ (3 ⫺ x)H* (24)

The second modification, to the RWGS reaction, stems from FTIR evidence
that adsorbed H atoms promote CO2 dissociation [51,53]:
26 BRADFORD AND VANNICE

CO2 ⫹ H* ` CO ⫹ HO* (25)

2HO* ` H2O ⫹ O* ⫹ * (26)

A final change suggests that an adsorbed O atom, rather than gas-phase H2O,
reacts with adsorbed CHx species; that is,
CHx* ⫹ O* ⫹ (x ⫺ 2)* ` CO ⫹ xH* (27)

2H* ` H2 ⫹ 2* (28)

Finally, similar to step (18), Osaki et al. have proposed that under reaction
conditions, the direct dissociative adsorption of CH4 yields gas-phase H2 [45];
that is

CH4 ⫹ * → CHx* ⫹ 冉 冊
4⫺x
2
H2 (29)

Several recent investigations have probed the rate-limiting steps asso-


ciated with CO2 reforming of CH4. Wang and Au used CD4 to resolve a kinetic
isotope effect using Ni/SiO2 [175] and Rh/SiO2 [176] catalysts and concluded
that CH4 dissociation is rate limiting and CO2 dissociation occurs prior to the
surface reaction of CHx fragments. Zhang and Verykios reported that the CH4/
CD4 isotope effect over Ni/La2O3 and Ni/␥-Al2O3 decreases with increasing
temperature [177], as predicted by ab initio density functional calculations
[178] and as shown experimentally on single-crystal surfaces [28]. In addi-
tion, the observed isotope effect on Ni/La2O3 was much greater than that over
Ni/␥-Al2O3 [177], a result possibly due to differences in the Ni crystallite
structure. Molecular beam studies have shown that although a strong CH4 /
CD4 isotope effect exists on both the Ni(111) [30] and Ni(100) surfaces [35],
there is essentially no isotope effect with Ni(110) [35]. The observed isotopic
scrambling between CH4 and CD4 during CO2 reforming of CH4 over Ni/
SiO2 provides additional evidence that CH4 decomposition is reversible [133].
In addition, Osaki et al. [45] employed PSRA and concluded that two reaction
steps are responsible for H2 production (i.e., dissociative CH4 adsorption to
form CHx species and the subsequent surface reaction of CHx ). Consequently,
they also concluded that the surface reaction of CHx and O is a rate-
determining step.
Solymosi and co-workers have suggested that adsorbed O atoms pro-
mote CH4 dissociation [50,54]; however, bond-order conservation (BOC)
Morse potential calculations have indicated that the activation barriers for O-
assisted CH4 dissociation are higher than those for unassisted CH4 dissociation
on Ni(111) and Ni(100) by 7 and 9 kcal/mol, respectively [179]. Conse-
quently, it is unlikely that adsorbed O atoms can promote CH4 dissociation.
Although Walter et al. have provided evidence that surface OH groups
rather than O atoms preferentially react with surface CHx species [180], that
is,
CO2 REFORMING OF CH4 27

CHx* ⫹ 2OH* → CO* ⫹ 冉 冊


x⫹2
2
H2 ⫹ O* ⫹ * (30)

most authors have preferred to claim that adsorbed O atoms are the key
intermediate. For example, in a study of CH4 partial oxidation and mixed
CO2 /H2OCH4 reforming, Qin et al. correlated CO formation rates with the
metal–oxygen bond strength of adsorbed O atoms, which was estimated by
the heat of formation of the most stable oxide per mole of metal [153];
however, these authors did not use specific activities in their correlation, thus
the validity of their correlation is very questionable. Regardless, even then,
such a correlation would not prove that adsorbed O atoms are the key reaction
intermediate because the binding energies for O and OH on transition metal
surfaces are directly proportional [181]. In addition, Shustorovich and Bell
have calculated the activation barriers for the two elementary steps C* ⫹
OH* → CO* ⫹ H* and C* ⫹ O* → CO* ⫹ * on Pt(111) to be 0 and 6
kcal/mol, respectively [182]. This provides support for the assumption that
adsorbed CHx fragments preferentially react with OH groups rather than O
atoms.
An alternate, more simplistic mechanism for CO2 reforming of CH4 has
been suggested by Mark and Maier [169] and Lercher et al. [183]:
CH4 ⫹ * → C* ⫹ 2H2 (31)

CO2 ⫹ 2* ` CO* ⫹ O* (32)

C* ⫹ O* ` CO* ⫹ * (33)

2CO* ` 2CO ⫹ 2* (34)

This sequence of steps was derived from a series of pulsed adsorption ex-
periments in which CH4 was shown to decompose stoichiometrically to car-
bon and H2, and CO2 was found to react stoichiometrically with surface car-
bon to yield CO [169,184]. However, pulsed CH4 adsorption experiments can
yield primarily atomic carbon, the most thermodynamically stable species
[43], whereas transient in situ experiments can identify CHx species as re-
active intermediates during CO2 reforming of CH4, as mentioned previously.
Furthermore, Osaki et al. have shown that the TOFs for CO2 reacting with
surface carbon resulting from CH4 decomposition are much lower than TOFs
for CO2 reforming of CH4 over supported Ni catalysts [125]. These results
do not support the alternate mechanism proposed by Mark and Maier [169]
and Lercher et al. [183].
By the use of partial pressure data such as those listed in Table 6, some
investigators have derived rate expressions to describe the reaction kinetics,
as shown in Table 7. The earliest report of a Langmuir-type rate expression
for CO2 reforming of CH4 was presented by Lewis et al. almost half a century
ago for a Cu/SiO2 catalyst [110]. However, they did not provide the reaction
28 BRADFORD AND VANNICE

TABLE 7
Proposed Rate Expressions for CO2 Reforming of CH4
Model Catalyst Ref.
kPCH4(PCO2 ⫹ PH2O)
r= Cu/SiO2 110
[1 ⫹ 24(PCO2 ⫹ PH2O) ⫹ 8PH2]2
kPCH4
r= Ni foil 115, 174
1 ⫹ a(PH2O /PH2) ⫹ bPCO
kKCO2KCH4PCO2PCH4
r= Rh/Al2O3 68
(1 ⫹ KCO2PCO2 ⫹ KCH4PCH4)2
kR[PCH4 ⫺ (P H2 2P CO
2
/KRPCO2)]
r= Ir/Al2O3 151
1 ⫹ (P CO /KR,CPCO2)
2

k兹K1K2PCO2PCH4 Ni/Al2O3
r= Ni/CaO – Al2O3 90, 185
(1 ⫹ 兹K1PCO2 ⫹ 兹K2PCH4) 2

Ni/SiO2
2
aPCH4P CO2 Ni/Al2O3
r= 89
(PCO2 ⫹ bP CO2 ⫹ cPCH4)2
2
Ni/CaO – Al2O3

mechanism from which their rate expression was derived, they performed
their experiments within a very limited experimental regime, and they as-
sumed that all parameters were temperature independent, with the exception
of the rate constant. The latter assumption implies that the heats of adsorption
of CO2, H2, and H2O are zero and is inconsistent with the reported heat of
adsorption of CO2 on Cu/SiO2 [184]. Bodrov and Apel’baum fit their data to
an expression which they originally derived to explain the kinetics of steam
reforming of CH4 [115,174], as mentioned previously. Richardson and Pari-
patyadar provided an expression based on a Langmuir–Hinshelwood ap-
proach involving redox mechanisms [68]. However, they also did not present
the reaction mechanism from which they derived their model, and the heats
and entropies of adsorption obtained from their adsorption parameters do not
satisfy the adsorption guidelines proposed by Boudart et al. and Vannice et
al. [186,187] because entropy changes are too small. Although other inves-
tigators have presented rate expressions for CO2 reforming of CH4
[89,90,151,185], they did not provide values for the adsorption and kinetic
parameters; thus, it was not possible to evaluate the validity of their models.
One of the most recent quantitative effort to model reaction kinetics for
this reaction is that of our group. After a thorough analysis of the data pre-
sented in Table 6, the numerous reaction models in Table 7, and the kinetic
studies previously discussed, the following reaction sequence has been pro-
posed for CO2 reforming of CH4 over supported Ni and Pt catalysts [70,71]:
CO2 REFORMING OF CH4 29

CH4 ⫹ * S
k1

k⫺1
CHx* ⫹ 冉 冊 4⫺x
2
H2 (35)

冋 册
K2
2 `
CO2 ⫹ * O CO2* (36)

K3

`
H2 ⫹ 2* O 2H* (37)

2 冋 `
K4
CO2* ⫹ H* O CO* ⫹ OH* 册 (38)

K5
`
OH* ⫹ H* O H2O ⫹ 2* (39)
K 6

CH * ⫹ OH*`
x O CH O* ⫹ H* x (40)
k7
CHxO* ⫹ → CO* ⫹ 冉冊 x
2
H2 (41)

3 冋 `
1/K6
CO* O CO ⫹ * 册 (42)

which corresponds to an overall reaction stoichiometry of


CH4 ⫹ 2CO2 → H2 ⫹ H2O ⫹ 3CO (43)
A discussion supporting this sequence of elementary steps is as follows. The
first step in this kinetic model, CH4 adsorption and dissociation to a CHx
fragment, is a simplification of steps (12)–(15) in which any geometric con-
straint associated with precisely defined adsorption sites is removed to sim-
plify the mathematical modeling, and it corresponds to reaction (29), except
that it is allowed to be reversible. The formation of CHx species is supported
by previous studies, and the choice of the particular species formed has little
influence on the derived rate expression [70,71]. Pulsed CH4 decomposition
experiments at elevated temperatures (973 K) have shown that pure CH4 in
an inert atmosphere decomposes stoichiometrically to C and 2H2 [151],
whereas the D2 –CH4 exchange reaction on Rh/MgO [105] and CO2 reforming
of CH4 /CD4 mixtures over Ni/SiO2 [132] have revealed that a distribution of
CHx fragments exists on the catalyst surface, as already mentioned. Although
Kroll et al. have suggested that Ni3C is the active phase during reaction [103],
Osaki et al. have determined that the reactivity of surface carbon (x = 0) on
supported Ni is lower than that of CHx species under reforming conditions
and concluded that it is only a minor reactive intermediate [125]. The re-
versible CH4 adsorption and dissociation step proposed in our model is sup-
ported by the effect of H2 addition to the feed gas, which increased the CH4
30 BRADFORD AND VANNICE

concentration and showed that the CH4 consumption rate was reversible
[70,71].
Bond-order conversion Morse potential (BOC–MP) calculations indi-
cate that the activation energies for O-assisted CH4 dissociation are higher
than those for unassisted CH4 dissociation on both Ni and Pt surfaces
[43,179]. Thus, step (35) in our proposed reaction sequence excludes pro-
motion by adsorbed O atoms. In regard to step (40), it has been proposed
that either hydroxyl groups [70,180] or O atoms [147,153,175] react with
surface CHx fragments. As discussed previously (see Section III), these O and
OH species may also originate from the support. Nevertheless, the reaction
of CHx with either adsorbed O or OH does not alter the mathematical form
of the derived rate expression; thus, exclusion of either species as the inter-
mediate oxidant is not possible on the basis of kinetic data alone. Further-
more, the correlation of Qin et al. [153] does not allow discrimination be-
tween the two intermediates, as mentioned earlier. However, the activation
barriers calculated for step (40) (with x = 0) by Shustorovich and Bell [182]
do provide further support for assuming that adsorbed CHx fragments pref-
erentially react with OH groups.
The isotopic studies discussed earlier have indicated that CH4 dissoci-
ation is a slow step [28,175,177], whereas Osaki et al. concluded from their
PSRA study that the surface reaction between CHx and O is rate determining
[125]. The choice of both reaction steps (35) and (41) as slow kinetic steps
in the proposed reaction sequence was influenced by the values of activation
energies reported for analogous homogeneous gas-phase reactions. The re-
actions between CHx and OH (as well as between C and OH) occur in the

FIG. 8. DRIFT spectra during CO2 adsorption on TiO2 and TiO2-supported


Ni, Pd, Pt, Ru, and Cu at 293 K, referenced to specrum of the reduced catalyst prior
to gas admission. Conditions: Ptot = 740 Torr, CO2 /Ar = 1/4. (From Ref. 109.)
CO2 REFORMING OF CH4 31

gas phase with no activation barrier; that is, they are facile, free-radical re-
actions, and activation barriers are also zero for OH formation [188]. Con-
versely, the activation barriers for CHxO (1 < x < 3) decomposition in the
gas phase range from about 17 to 81 kcal/mol [188]. Nevertheless, because
a surface reaction between CHx and O (or OH) is difficult to differentiate
from the decomposition of a subsequently formed CHxO intermediate on the
basis of isotope effects and PSRA alone, the results of Wang and Au [175],
Zhang and Verykios [177], and Osaki et al. [125] are not inconsistent with
the choice of CH4 dissociation and CHxO decomposition as slow kinetic steps
in our kinetic model.
Reaction steps (36)–(39) plus step (42) describe the RWGS reaction and
are assumed to be quasi-equilibrated to accommodate experimental results
that this reaction is near thermodynamic equilibrium (see Fig. 2). Rostrup-
Nielsen and Bak Hansen have shown that the TOF for the RWGS reaction

FIG. 9. DRIFT spectra during CO2 reforming of CH4 at 723 K on reduced


(A) Pt/TiO2, (B) Pt/ZrO2, (C) Pt/Cr2O3, and (D) Pt/SiO2. The reference spectrum was
that of the catalyst prior to gas admission. Reaction conditions: PCH4 = PCO2 = 80
Torr, balance Ar/He = 10/1. (From Ref. 71.)
32 BRADFORD AND VANNICE

is more than 20 times higher than that for CO2 reforming of CH4 over 0.9%
Pt/MgO [49]. Nondissociative adsorption of CO2 under reaction conditions is
assumed to occur on the support in the form of carbonates, as evidenced by
in situ DRIFT spectra such as those shown in Fig. 8. The strong bands at
1588 and 1436 cm⫺1 represent bicarbonate species, whereas the weaker bands
at 1675 and 1222 cm⫺1 are associated with bidentate carbonate species and
the bands at 1540, 1380, and 1363 cm⫺1 can be assigned to a surface formate
species [109]. Hydrogen atom-assisted carbonate dissociation has been dis-
cussed previously [70] and is further indicated by TPH (temperature-
programmed hydrogenation) results [71]. On Pt/TiO2, reaction steps (36)–
(38) may involve a type of redox cycle, as previously suggested by DRIFTS
and TPSR results [70,109]. Otsuka et al. have also provided evidence to

FIG. 10. Fit of the proposed kinetic model for CO2 reforming of CH4 as a
function of CH4 partial pressure (panel A) and CO2 partial pressure (panel B); (a)
10.1% Ni/MgO at (▫) 773 K, (䉭) 798 K, and (䡩) 823 K; (b) 0.82% Pt/TiO2 at (▫)
673 K, (䉭) 698 K, and (䡩) 723 K. [Data for (a) from Ref. 70; data for (b) from Ref.
71.]
CO2 REFORMING OF CH4 33

indicate that a redox reaction may occur during CO2 reforming of CH4 over
Pt/CeO2 [164].
If it is assumed that the most abundant reaction intermediate (mari) is
CHxO, then L = [*] ⫹ [CHx O*] and the following general expression can be
derived for the rate of CH4 consumption [43,70]:

k̂1PCH4PCO2
rCH4 = ˆ ¯ ˆ (44)
(4⫺x)/2
(k⫺1K/k7)PCOP H2 ⫹ [1 ⫹ (kˆ 1 /kˆ 7)PCH4]PCO2
where k̂i = ki L, ki is the rate constant for step i, L is total number of active
sites, and K̄ is a lumped equilibrium constant. Computer optimization of ki-
netic data obtained with Ni and Pt catalysts [70,71] failed to locate a global
minimum for x and indicated that the statistical fit of the data (as measured
by residual sum of squares) with Eq. (44) was essentially independent of x
(0 ⱕ x ⱕ 4). A value of x = 0 would physically imply that adsorbed C atoms
are the active CHx intermediate and adsorbed CO is the mari; however, in-

FIG. 10. Continued


34 BRADFORD AND VANNICE

TABLE 8
Optimized Parameter Values for Rate Expression (44)
Parameter
T
Catalyst (K) k̂1a,b k̂7c kˆ ⫺1K
¯b

10.1% Ni/MgO 773 0.031 23.1 0.17


798 0.054 20.5 0.15
823 0.085 33.6 0.16
1.2% Ni/TiO2 673 0.003 0.77 0
698 0.010 3.7 5.4
723 0.042 5.4 0.85
0.82% Pt/TiO2 673 0.099 3.4 37
698 0.12 7.4 24
723 0.27 10.4 9.1
a
Value is reported ⫾ 95% confidence interval.
b
Units of ␮mol g cat⫺1 Torr⫺1.
c
Units of ␮mol g cat⫺1 s⫺1.

clusion of adsorbed CO in the site balance did not improve the statistical fit
of the model, although adsorbed CO can be detected under reaction condi-
tions, as shown in Fig. 9 for Pt. As discussed previously, it is very possible
that a distribution of CHx species exists under reforming conditions; thus, for
simplicity, the partial pressure data for Ni and Pt catalysts were optimized to
Eq. (44) by assuming an arbitrary but reasonable value of x = 2 [70,71]. The
recent evidence for the identification of a CH2O surface species supports this
choice [109]. Two examples of the fit of this rate expression to experimental
data are shown in Figs. 10a and 10b for Ni/MgO and Pt/TiO2, respectively,
and values of the optimized model parameters are provided in Table 8. Plots
of the two model parameters kˆ 1 and kˆ 7 versus reciprocal temperature yield the
activation energy for step (35) in the forward direction, Ek̂1 , and the activation
energy for step (41), Ek̂7. The activation energies in Table 9 obtained with
this model can be compared with literature values for similar but not identical

TABLE 9
Activation Energies for Elementary Steps Obtained from Eq. (44)
Catalyst Step Parameter E (kcal/mol)
10.1% Ni/MgO 35 k̂1 26
41 k̂7 9.3
1.22% Ni/TiO2 35 k̂1 51
41 k̂7 38
0.82% Pt/TiO2 35 k̂1 19
41 k̂7 22
CO2 REFORMING OF CH4 35

reactions on Pt(111) [181,189,190]. Although these values are not widely


divergent, they are consistent with the proposal that the reaction is not oc-
curring solely on the Pt surface but primarily in the metal–support interfacial
region. Considering that CHx O species most likely are formed and subse-
quently decomposed in the metal–support interfacial region [70,71,109], this
suggestion is not unreasonable.

VIII. SUMMARY

This is an important reaction that has attracted significant attention dur-


ing this decade. Consequently, in contrast to prior brief reviews of CO2 re-
forming of CH4 published in 1995 [191] and 1996 [192], a complete com-
pilation of prior work as well as a thorough review of the reaction kinetics
has been provided here. The activation of both CH4 and CO2 on transition
metal surfaces may depend on both electronic and geometric factors such that
their dissociative adsorption may be structure sensitive. Empirical correlations
of the TOF for CO2 reforming of CH4 with the metal %–d character lend
support to the importance of electronic properties. Evidence has been pro-
vided to indicate that differences in activity for CO2/CH4 reforming over a
given metal dispersed on different supports may be due to metal–support
interactions and/or the participation of O or OH species from the support in
the metal–support interfacial region. Metal–support interactions, such as TiOx
formation on the metal surface, metal–carbon bonding, solid–solution for-
mation, and metal morphology changes, can also influence carbon deposition
during CO2 reforming CH4. Analysis of kinetic data from the literature indi-
cated that the numerous variations in catalyst activity and apparent activation
energy may frequently be due to the use of low space velocities which give
conversions in proximity to that controlled by thermodynamic equilibrium;
under these conditions, kinetic data for CO2 reforming of CH4 can be altered
by the reverse reaction (i.e., CO hydrogenation to methane). Although nu-
merous kinetic models have been proposed to describe the CO2/CH4 reform-
ing reaction quantitatively, the most consistent model currently available in-
vokes the following: the reversible dissociation of CH4 to yield CHx species
and H2, nondissociative adsorption of CO2 on the support, H-promoted CO2
dissociation in the metal–support interfacial region, the reaction of CHx spe-
cies with OH (or O) species to yield CHxO species in the metal–support
interfacial region, and CHx O decomposition in the metal–support interfacial
region to yield CO and H2.

ACKNOWLEDGMENTS

The authors would like to thank NEDO, a Japanese international joint


research program, for sponsoring the research requiring a review of this lit-
36 BRADFORD AND VANNICE

erature, the U.S. Department of Education for a GAANN Fellowship for


MCJB, and the NSF for an equipment grant (CTS-9311087) used to acquire
the DRIFTS system which provided some of the cited results.

REFERENCES

1. D. Dyrssen and D. R. Turner, in Carbon Dioxide Chemistry: Environmental


Issues (J. Paul and C.-M. Pradier, eds.), Athenaeum Press, Cambridge, 1994,
p. 317.
2. R. A. Houghton and G. M. Woodwell, Scientific Am., 260, 36 (1989).
3. D. Schneider, Scientific Am., 272(2), 13 (1995).
4. B. Hileman, Chem. Eng. News, 28 (April 14, 1997).
5. T. A. Chubb, Sol. Energy, 24, 341 (1980).
6. J. H. McCrary, G. E. McCrary, T. A. Chubb, J. J. Nemecek, and D. E. Sim-
mons, Sol. Energy, 29, 141 (1982).
7. J. D. Fish and D. C. Hawn, J. Sol. Energy Eng., 109, 215 (1987).
8. J. H. Edwards, K. T. Do, A. H. Maitra, S. Schuck, and W. Stein, Sol. Eng.,
1, 389 (1995).
9. J. R. H. Ross, A. N. J. van Keulen, M. E. S. Hegarty, and K. Seshan, Catal.
Today, 30, 193 (1996).
10. M. A. Vannice, Catal. Rev.— Sci. Eng., 14, 153 (1976).
11. A. M. Gadalla and B. Bower, Chem. Eng. Sci., 43, 3049 (1988).
12. J. R. Rostrup-Nielsen, J.-H. Bak Hansen, and L. M. Aparicio, Sekiya Gak-
kaishi 40, 366 (1997).
13. H. J. Töpfer, Gas Wasserfach, 117, 412 (1976).
14. S. Teuner, Hydrocarbon Process., 64, 106 (1985).
15. S. Teuner, Hydrocarbon Process., 66, 52 (1987).
16. N. R. Udengaard, J.-H. B. Hansen, D. C. Hanson, and J. A. Stal, Oil Gas J.,
90, 62 (1992).
17. M. Kurioka, K. Nakata, T. Jintoku, Y. Taniguchi, K. Taskaki, and Y. Fujiwara,
Chem. Lett., 244 (1995).
18. K. Asami, T. Fujita, K.-I. Kusakabe, Y. Nishiyama, and Y. Ohtsuka, Appl.
Catal. A: General, 126, 245 (1995).
19. K. Asami, T. Shikada, and K. Fujimoto, Bull. Chem. Soc. Jpn., 64, 266 (1991).
20. T. Nishiyama and K.-I. Aika, J. Catal., 122, 346 (1990).
21. T. Sodesawa, A. Dobashi, and F. Nozaki, React. Kinet. Catal. Lett., 12, 107
(1979).
22. F. Fischer and H. Tropsch, Brennstoff Chem., 3, 39 (1928).
23. J. Lang, Zeitschr. Physikal. Chem., 2, 161 (1888).
24. J. A. Peters, ARL Technical Report 88-008, The Pennsylvania State University,
1988.
25. V. R. Choudhary, A. M. Rajput, and B. Prabhakar, Catal. Lett., 32, 391 (1995).
26. B. Ølgaard Nielsen, A. C. Luntz, P. M. Holmblad, and I. Chorkendorff, Catal.
Lett., 32, 15 (1995).
27. A. C. Luntz and H. F. Winters, J. Chem. Phys., 101, 10980 (1994).
28. A. C. Luntz and J. Harris, Surf. Sci., 258, 397 (1991).
CO2 REFORMING OF CH4 37

29. D. C. Seets, M. C. Wheeler, and C. B. Mullins, Chem. Phys. Lett., 266, 431
(1997).
30. S. T. Ceyer, Q. Y. Yang, M. B. Lee, J. D. Beckerle, and A. D. Johnson, in
Methane Conversion (D. M. Bibby, C. D. Chang, R. F. Howe, and S. Yurak,
eds.), Elsevier, Amsterdam, 1988, p. 51.
31. R. A. van Santen and M. Neurock, Catal. Rev. Sci. Eng., 37, 557 (1995).
32. J. P. Lowe, Quantum Chemistry, 2nd ed., Academic Press, Boston, 1993.
33. D. J. Trevor, D. M. Cox, and A. Kaldor, J. Am. Chem. Soc., 112, 3742 (1990).
34. E. G. M. Kuijpers, A. K. Breedijk, W. J. J. van der Wal, and J. W. Geus, J.
Catal., 81, 429 (1983).
35. T. P. Beebe, Jr., D. W. Goodman, B. D. Kay, and J. T. Yates, Jr., J. Chem.
Phys., 87, 2305 (1987).
36. M. Kaminsky, Atomic & Ionic Impact Phenomena on Metal Surfaces,
Springer-Verlag, New York, 1965.
37. C. Minot, M. A. van Hove, and G. A. Somorjai, Surf. Sci., 127, 441 (1982).
38. A. D. E. Koster and R. A. van Santen, J. Catal., 127, 141 (1991).
39. C. Zheng, Y. Apeloig, and R. Hoffman, J. Am. Chem. Soc., 110, 749 (1988).
40. L. M. Aparicio, J. Catal., 165, 262 (1997).
41. A. Ozaki, Isotopic Studies in Heterogeneous Catalysis, Academic Press, New
York, 1977.
42. K. Fujimoto, personal communication (1995).
43. M. C. J. Bradford, Ph.D. thesis, The Pennsylvania State University (1997).
44. T. Osaki, H. Masuda, T. Horiuchi, and T. Mori, Catal. Lett., 34, 59 (1995).
45. T. Osaki, H. Masuda, and T. Mori, Catal. Lett., 29, 33 (1994).
46. J. Erkelens and W. J. Wösten, J. Catal., 54, 54 (1978).
47. H. Matsumoto, Shokubai, 16, 122 (1974); Shoukubai, 18, 71 (1976); Hyomen,
15, 226 (1977).
48. O. Takayasu, N. Hongo, and I. Matsuura, Stud. Surf. Sci. Catal., 77, 305
(1993).
49. J. R. Rostrup-Nielsen and J.-H. Bak Hansen, J. Catal., 144, 38 (1993).
50. F. Solymosi, Gy. Kustán, and A. Erdöhelyi, Catal. Lett., 11, 149 (1991).
51. A. Erdöhelyi, J. Cserényi, and F. Solymosi, J. Catal., 141, 287 (1993).
52. F. Solymosi and J. Cserényi, Catal. Today, 21, 561 (1994).
53. A. Erdöhelyi, J. Cserényi, E. Papp, and F. Solymosi, Appl. Catal. A: General,
108, 205 (1994).
54. A. Erdöhelyi, K. Fodor, and F. Solymosi, Stud. Surf. Sci. Catal., 107, 525
(1997).
55. P. Ferreira-Aparicio, I. Rodrı́guez-Ramos, and A. Guerrero-Ruiz, Appl. Catal.
A: General, 148, 343 (1997).
56. F. Solymosi, J. Mol. Catal., 65, 337 (1991).
57. J. Wambach and H.-J. Freund, in Carbon Dioxide Chemistry: Environmental
Issues (J. Paul and C.-M. Pradier, eds.), Athenaeum Press, Cambridge, 1994,
p. 31.
58. J. Segner, C. T. Campbell, G. Doyen, and G. Ertl, Surf. Sci., 138, 505 (1984).
59. B. Z. Nikolic, H. Huang, D. Gervasio, A. Lin, C. Fierro, R. R. Adzic, and E.
B. Yeager, J. Electroanal. Chem., 295, 415 (1990).
60. A. Rodes, E. Pastor, and T. Iwasita, Anal. Quim., 89, 458 (1993).
38 BRADFORD AND VANNICE

61. M. C. Román-Martinez, D. Cazorla-Amorós, C. Salinas-Martinez de Lecea,


and A. Linares-Solano, Langmuir, 12, 379 (1996).
62. P. D. F. Vernon, M. L. H. Green, A. K. Cheetham, and A. T. Ashcroft, Catal.
Today, 13, 417 (1992).
63. H. D. Gesser, N. R. Hunter, A. N. Shigapov, and V. Januati, Energy Fuels, 8,
1123 (1994).
64. H. M. Swaan, V. C. H. Kroll, G. A. Martin, and C. Mirodatos, Catal. Today,
21, 571 (1994).
65. A. N. J. van Keulen, Ph.D. dissertation, University of Twente (1996).
66. A. M. Gadalla and M. E. Sommer, Chem. Eng. Sci., 44, 2825 (1989).
67. A. M. Gadalla and M. E. Sommer, J. Am. Ceram. Soc., 72, 683 (1989).
68. J. T. Richardson and S. A. Paripatyadar, Appl. Catal., 61, 293 (1990).
69. J. R. Rostrup-Nielsen, in Catalysis: Science and Technology (J. R. Anderson
and M. Boudart, eds.), Springer-Verlag Berlin, 1984, p. 1.
70. M. C. J. Bradford and M. A. Vannice, Appl. Catal. A.: General, 142, 97
(1996).
71. M. C. J. Bradford and M. A. Vannice, J. Catal., 173, 157 (1998).
72. R. E. Reitmeier, K. Atwood, H. A. Bennet, Jr., and H. M. Baugh, Ind. Eng.
Chem., 40, 620 (1948).
73. G. A. White, T. R. Roszkowski, and D. W. Stanbridge, Hydrocarbon Process.,
54, 130 (1975).
74. A. Sacco, Jr., F. W. A. H. Geurts, G. A. Jablonski, S. Lee, and R. A. Gatell,
J. Catal., 119, 322 (1989).
75. N. M. Rodriguez, J. Mater. Res., 8, 3233 (1993).
76. G. Blyholder and M. Lawless, Surf. Sci., 290, 155 (1993).
77. F. Zaera, E. Kollin, and J. L. Gland, Chem. Phys. Lett., 121, 464 (1985).
78. S. Johnson and R. J. Madix, Surf. Sci., 108, 77 (1981).
79. X. Yin-Sheng and H. Xiao-Le, J. Mol. Catal., 33, 179 (1985).
80. E. O. F. Zdansky, A. Nilsson, and N. Mårtensson, Surf. Sci., 310, L83 (1994).
81. G. A. Somorjai, in Bonding Energetics in Organometallic Compounds (T. J.
Marks, ed.), ACS Symposium Series 428, American Chemical Society, Wash-
ington, DC, 1990, p. 218.
82. M. T. Tavares, I. Alstrup, C. A. Berriardo, and J. R. Rostrup-Nielsen, J. Catal.,
147, 525 (1994).
83. V. V. Chesnokov, V. I. Zaikovskii, R. A. Buyanov, V. V. Molchanov, and L.
M. Plyasova, Catal. Today, 24, 265 (1995).
84. V. A. Tsipopuriari, A. M. Efstathiou, Z. L. Zhang, and X. E. Verykios, Catal.
Today, 21, 579 (1994).
85. J. R. Rostrup-Nielsen, Stud. Surf. Sci. Catal., 68, 85 (1991).
86. J. R. Rostrup-Nielsen, J. Catal., 85, 31 (1984).
87. M. C. J. Bradford and M. A. Vannice, Appl. Catal. A.: General, 142, 73
(1996).
88. M. C. J. Bradford and M. A. Vannice, Catal. Lett., 48, 31 (1997).
89. Z. L. Zhang and X. E. Verykios, Catal. Today, 21, 589 (1994).
90. T. Horiuchi, K. Sakuma, T. Fukui, Y. Kubo, T. Osaki, and T. Mori, Appl.
Catal. A: General, 144, 111 (1996).
91. O. Yamazaki, T. Nozaki, K. Omata, and K. Fujimoto, Chem. Lett., 1953
(1992).
CO2 REFORMING OF CH4 39

92. G. J. Kim, D.-S. Cho, K.-H. Kim, and J.-H. Kim, Catal. Lett., 28, 41 (1994).
93. S.-B. Tang, F.-L. Qiu, and S.-J. Lu, Catal. Today, 24, 253 (1995).
94. Y.-G. Chen and J. Ren, Catal. Lett., 29, 39 (1994).
95. S. Sridhar, D. Sichen, and S. Seetharaman, Z. Metallkd., 85, 9 (1994).
96. A. Bhattacharyya and V. W. Chang, Stud. Surf. Sci. Catal., 88, 207 (1994).
97. T.-G. Chen, O. Yamazaki, K. Tomishige, and K. Fujimoto, Catal. Lett., 39,
91 (1996).
98. (a) Y. H. Hu and E. Ruckenstein, Catal. Lett., 36, 145 (1996); (b) E. Ruck-
enstein and Y. H. Hu, J. Catal., 162, 230 (1996); (c) Y. H. Yu and E. Ruck-
enstein, J. Catal., 163, 306 (1996); (d) E. Ruckenstein and Y. H. Hu, Appl.
Catal. A: General, 154, 185 (1997); (e) Y. H. Hu and E. Ruckenstein, Catal.
Lett., 43, 71 (1997); (f) Y. H. Hu and E. Ruckenstein, Langmuir, 13, 2055
(1997).
99. M. Masai, H. Kado, A. Miyake, S. Nishiyama, and S. Tsuruya, in Methane
Conversion (B. M. Biddy, C. D. Chang, R. F. Howe, and S. Yurchak, eds.),
Elsevier, Amsterdam, 1988, p. 67.
100. S. Roberts and R. J. Gorte, J. Phys. Chem., 95, 5600 (1991).
101. S. Roberts and R. J. Gorte, J. Phys. Chem., 93, 5337 (1990).
102. R. T. K. Baker, E. B. Prestridge, and R. L. Garten, J. Catal., 56, 390 (1979).
103. V. C. H. Kroll, H. M. Swaan, and C. Mirodatos, J. Catal., 161, 409 (1996).
104. J. S. H. Q. Perera, J. W. Couves, G. Sankar, and J. M. Thomas, Catal. Lett.,
11, 219 (1991).
105. D. Qin and J. Lapszewicz, Catal. Today, 21, 551 (1994).
106. A. T. Ashcroft, A. K. Cheetham, M. L. H. Green, and P. D. F. Vernon, Nature,
352, 225 (1991).
107. A. Guerrero-Ruiz, A. Sepúlveda-Escribano, and I. Rodrı́guez-Ramos, Catal.
Today, 21, 545 (1994).
108. S. M. Stagg and D. E. Resasco, Stud. Surf. Sci. Catal., 111, 543 (1997).
109. M. C. J. Bradford and M. A. Vannice, Catal. Today (in press).
110. W. K. Lewis, E. R. Gilliland, and W. A. Reed, Ind. Eng. Chem., 41, 1227
(1949).
111. O. Tokunaga and S. Ogasawara, React. Kinet. Catal. Lett., 39, 69 (1989).
112. O. Tokunaga, Y. Osada, and S. Ogasawara, Fuel, 68, 990 (1989).
113. A. Guerrero-Ruiz, I. Rodriguez-Ramos, and A. Sepúlveda-Escribano, J. Chem.
Soc., Chem. Commun., 487 (1993).
114. R. N. Bhat and W. M. H. Sachtler, Appl. Catal. A: General, 150, 279 (1997).
115. I. M. Bodrov and L. O. Apel’baum, Kinet. Catal., 8, 326 (1967).
116. A. T. Ashcroft, A. K. Cheetham, M. L. H. Green, and P. D. F. Vernon, Nature,
352, 225 (1991).
117. L. A. Rudnitskii, T. N. Solboleva, and A. M. Alekseev, React. Kinet. Catal.
Lett., 26, 149 (1984).
118. Z. Zhang and X. E. Verykios, J. Chem. Soc., Chem. Commun., 71 (1995).
119. R. Blom, I. M. Dahl, Å. Slagtern, B. Sortland, A. Spjelkavik, and E. Tangstad,
Catal. Today, 21, 535 (1994).
120. A. Bhattacharyya, W.-D. Chang, M. S. Kleefisch, and C. A. Udovich, U. S.
Patent 5,399,537 (1995).
121. A. Bhattacharyya, J. B. Hall, and C. Choi-Feng, ACS 210th National Meeting,
Div. Petr. Chem., 1995, p. 640.
40 BRADFORD AND VANNICE

122. K. Seshan, H. W. ten Barge, W. Hally, A. N. J. van Keulen, and J. R. H. Ross,


in Natural Gas Conversion II (H. E. Curry-Hyde and R. F. Howe, eds.),
Elsevier, Amsterdam, 1994, p. 285.
123. Z. Zhang, X. Verykios, S. M. MacDonald, and S. Affrossman, J. Phys. Chem.,
100, 744 (1996).
124. A. Takano, T. Tagawa, and S. Goto, J. Chem. Eng. Japan, 27, 727 (1994).
125. T. Osaki, T. Horiuchi, K. Suzuki, and T. Mori, J. Chem. Soc., Faraday Trans.,
92, 1627 (1996).
126. Z. Chang, Q. Wu, J. Li, and Q. Zhu, Catal. Today, 30, 147 (1996).
127. T. Horiuchi, K. Sakuma, T. Fukui, Y. Kubo, T. Osaki, and T. Mori, Appl.
Catal. A: General, 144, 111 (1996).
128. L. Yong, Y. Changchun, D. Zuejia, S. Shikong, and D. Cun, Cuihua Xuebao
(Chinese J. Catal.), 17, 212 (1996).
129. Y. Sakai, H. Saito, T. Sodesawa, and F. Nozaki, React. Kinet. Catal. Lett., 24,
253 (1984).
130. T. Sodesawa, A. Dobashi, and F. Nozaki, React. Kinet. Catal. Lett., 12, 107
(1979).
131. T. Osaki, T. Horiuchi, K. Suzuki, and T. Mori, Catal. Lett., 35, 39 (1995).
132. V. C. H. Kroll, P. Delichure, and C. Mirodatos, Kinet. Catal., 37, 698 (1996).
133. V. C. H. Kroll, H. M. Swann, S. Lacombe, and C. Mirodatos, J. Catal., 164,
387 (1997).
134. P. Gronchi, D. Fumagalli, R. Del Rosso, and P. Centola, J. Thermal Anal., 47,
227 (1996).
135. Å. Slagtern, V. Olsbye, R. Blom, I. M. Dahl, and M. Fjellvȧg, Appl. Catal.
A: General, 165, 379 (1997).
136. V. R. Choudhary, B. S. Uphade, and A. S. Mamman, Catal. Lett., 32, 387
(1995).
137. E. Ruckenstein and Y. H. Hu, Appl. Catal. A. General, 133, 149 (1995).
138. V. R. Choudhary and A. M. Rajput, Ind. Eng. Chem. Res., 35, 3934 (1996).
139. T. Inui, Stud. Surf. Sci. Catal., 77, 17 (1993).
140. T. Inui, K. Saigo, Y. Fujii, and K. Fujioka, Catal. Today, 26, 295 (1995).
141. O. Takayasu, E. Hirose, N. Matsuda, and I. Matruura, Chem. Express. 6, 447
(1991).
142. D. Z. Wang, Appl. Catal. B: Environmental, 6, N28 (1995).
143. Z. L. Zhang and X. E. Verykios, Appl. Catal. A: General, 138, 109 (1996).
144. R. Moene, Ph.D. dissertation, Delft University of Technology (1995).
145. T. Osaki, J. Chem. Soc., Faraday Trans., 93, 343 (1997).
146. W. Hally, J. H. Bitter, K. Seshan, J. A. Lercher, and J. R. H. Ross, Stud. Surf.
Sci. Catal., 88, 167 (1994).
147. J.-S. Chang, S.-E. Park, and H. Chon, Appl. Catal. A: General, 145, 111
(1996).
148. F. Solymosi, Gy. Kustán, and A. Erdöhelyi, Catal. Lett., 11, 149 (1991).
149. L. Basini and D. Sanfilippo, J. Catal., 157, 162 (1995).
150. A. A. Ponelis, Stud. Surf. Sci. Catal., 107, 555 (1997).
151. M. F. Mark and W. F. Maier, J. Catal., 164, 122 (1996).
152. J. B. Claridge, M. L. H. Green, and S. C. Tsang, Catal. Today, 12, 455 (1994).
153. D. Qin, J. Lapszewicz, and X. Jiang, J. Catal., 159, 140 (1996).
154. P. G. S. van Zyl, M.S. thesis, University of Pretoria (1996).
CO2 REFORMING OF CH4 41

155. J. Nakamura, K. Aikawa, K. Sato, and T. Uchijima, Catal. Lett., 25, 265
(1994).
156. K. Walter, O. V. Buyevskaya, D. Wolf, and M. Baerns, Catal. Lett., 29, 261
(1994).
157. Z. L. Zhang, V. A. Tsipouriari, A. M. Efstathiou, and X. E. Verykios, J. Catal.,
158, 51 (1996).
158. M. Sigl, M. C. J. Bradford, H. Knözinger, and M. A. Vannice, Topics Catal.,
to be published.
159. J. Raskó and F. Solymosi, Catal. Lett., 46, 153 (1997).
160. A. M. Efstathiou, A. Kladi, V. A. Tsipouriari, and X. E. Verykios, J. Catal.,
158, 64 (1996).
161. B. L. Gustafson and J. V. Walden, U.S. Patent 5,068,057 (1991).
162. J. H. Bitter, W. Hally, K. Seshan, J. G. van Ommen, and J. A. Lercher, Catal.
Today, 29, 349 (1996).
163. J. H. Bitter, K. Seshan, and J. A. Lercher, J. Catal., 171, 279 (1997).
164. K. Otsuka, T. Ushiyama, and I. Yamanaka, Chem. Lett., 1517 (1993).
165. A. N. J. van Keulen, K. Seshan, J. H. B. J. Hoebink, and J. R. H. Ross, J.
Catal., 166, 306 (1997).
166. Z. Yu, K. Choi, M. P. Rosynek, and J. H Lunsford, React. Kinet. Catal. Lett.,
51, 143 (1993).
167. A. P. E. York, J. B. Claridge, C. Marquez-Alvarez, A. J. Brungs, and M. L.
H. Green, ACS 213th National Meeting, Division of Fuel, 1997, p. 66.
168. A. P. E. York, J. B. Claridge, A. J. Brungs, S. C. Tsang, and M. L. H. Green,
Chem. Commun., 39 (1997).
169. M. F. Mark and W. F. Maier, Angew. Chem. Int. Ed. Engl., 33, 1657 (1994).
170. N. W. Cant, R. Dümpelmann, and A. M. Maitra, Stud. Surf. Sci. Catal., 107,
491 (1997).
171. G. A. Somorjai, Introduction to Surface Chemistry and Catalysis, John Wiley
& Sons, New York, 1994.
172. M. A. Vannice, J. Catal., 37, 462 (1975).
173. L. Pauling, Proc. Roy. Soc. A., 196, 343 (1949).
174. N. M. Bodrov, L. O. Apel’baum, and M. I. Temkin, Kinet. Catal., 5, 614
(1964).
175. H.-Y. Wang and C.-T. Au, Catal. Lett., 38, 77 (1996).
176. H.-Y. Wang and C. T. Au, Appl. Catal. A: General, 155, 239 (1997).
177. Z. Zhang and X. E. Verykios, Catal. Lett., 38, 175 (1996).
178. H. Burghgraef, A. P. J. Jansen, and R. A. van Santen, J. Chem. Phys., 101,
11012 (1994).
179. E. Shustorovich and A. T. Bell, Surf. Sci., 268, 397 (1992).
180. K. Walter, O. V. Buyevskaya, D. Wolf, and M. Baerns, Catal. Lett., 29, 261
(1994).
181. E. Shustorovich, Adv. Catal., 37, 101 (1990).
182. E. Shustorovich and A. T. Bell, Surf. Sci., 248, 359 (1991).
183. J. A. Lercher, J. H. Bitter, W. Hally, W. Niessen, and K. Seshan, Stud. Surf.
Sci. Catal., 101, 463 (1996).
184. D. B. Clarke, I. Suzuki, and A. T. Bell, J. Catal., 142, 27 (1993).
185. T. Osaki, T. Horiuchi, K. Suzuki, and T. Mori, Appl. Catal. A: General, 155,
229 (1997).
42 BRADFORD AND VANNICE

186. M. Boudart, D. E. Mears, and M. A. Vannice, Ind. Chim. Belg., 32, 281 (1967).
187. M. A. Vannice, S. H. Hyun, B. Kalpakci, and W. C. Liauh, J. Catal., 56, 358
(1979).
188. J. A. Miller and C. T. Bowman, Prog. Enegy Comb. Sci., 15, 287 (1989).
189. F. Zaera, Catal. Lett., 11, 95 (1991).
190. Y.-K. Sun and W. H. Weinberg, Surf. Sci., 227, L86 (1990).
191. J. H. Edwards and A. M. Maitra, Fuel Process. Technol., 42, 269 (1995).
192. S. Wang, G. Q. Lu, and G. J. Millar, Energy Fuels, 10, 896 (1996).

Anda mungkin juga menyukai