Effect of component interaction on the activity of Co/CuZnO for CO hydrogenation

Article abstract of DOI:10.1016/j.jcat.2011.09.033

Co/CuZnO is known as a base metal catalyst active for C₂₊ oxygenate synthesis. This study probed the interactions of the different components of Co/CuZnO catalysts on CO hydrogenation using Fischer-Tropsch synthesis (250 °C, H₂/CO = 2) and SSITKA. Only combination of all three metal components produced a catalyst with relatively high C₂₊ oxygenate selectivity, but with much lower activity compared to that for Co/Al₂O₃. In situ reaction characterizations, albeit at somewhat different conditions than alcohol synthesis, helped explain interaction of the components. SSITKA, under methanation conditions, indicated that the most striking feature for the combination of Co with ZnO and/or Cu was a much decreased amount of reaction intermediates. Ethane hydrogenolysis results suggested that the different components for these catalysts were in close contact and few or no large ensembles (n ≥ 12) of Co atoms existed, confirming that ZnO and/or Cu covered/blocked a substantial number of active sites on Co for CO hydrogenation.

Full text of DOI:10.1016/j.jcat.2011.09.033

Journal of Catalysis 285 (2012) 208–215
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Effect of component interaction on the activity of Co/CuZnO for CO hydrogenation
a
a
a
b
a,
⇑
Xunhua Mo , Yu-Tung Tsai , Jia Gao , Dongsen Mao , James G. Goodwin Jr.
a
Department of Chemical and Biomolecular Engineering, Clemson University, Clemson, SC 29634, USA
Research Institute of Applied Catalysis, School of Chemical and Environmental Engineering, Shanghai Institute of Technology, Shanghai 200235, China
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 May 2011
Revised 22 September 2011
Accepted 24 September 2011
Available online 9 November 2011
Co/CuZnO is known as a base metal catalyst active for C2+ oxygenate synthesis. This study probed the
interactions of the different components of Co/CuZnO catalysts on CO hydrogenation using Fischer–Trop-
sch synthesis (250 °C, H
a catalyst with relatively high C2+ oxygenate selectivity, but with much lower activity compared to that
for Co/Al . In situ reaction characterizations, albeit at somewhat different conditions than alcohol syn-
2
/CO = 2) and SSITKA. Only combination of all three metal components produced
2 3
O
thesis, helped explain interaction of the components. SSITKA, under methanation conditions, indicated
that the most striking feature for the combination of Co with ZnO and/or Cu was a much decreased
amount of reaction intermediates. Ethane hydrogenolysis results suggested that the different compo-
nents for these catalysts were in close contact and few or no large ensembles (n P 12) of Co atoms
existed, confirming that ZnO and/or Cu covered/blocked a substantial number of active sites on Co for
CO hydrogenation.
Keywords:
Co/CuZnO
Co
Cu
ZnO
CO hydrogenation
Ethanol
Ó 2011 Elsevier Inc. All rights reserved.
Higher oxygenates synthesis
Ethane hydrogenolysis
SSITKA
1
. Introduction
mance with selectivities for C2+ oxygenates as high as 80% at
2
20–350 °C and 50–150 atm [22–26]. The fourth-generation IFP
In order to alleviate the demand for fuel from imported crude
catalysts were claimed to retain excellent activities after 1000 h
TOS, and the main components included Cu, Co, Al, Zn, A (an alkali
or alkaline-earth metal), and M (a group VIII noble metal). Zn may
be replaced by Cd or Mn; Al by Cr, Mn, or Ti. Preferred wt.% (with
respect to total metal) for each metal component except for A has
been given as: Cu (15–45%); Co (9–20%); Al (7–25%), Zn (15–50%);
and M (0.02–0.8%) [13]. These IFP catalysts can be prepared either
by co-precipitation (followed by alkalinization using impregna-
tion) or by decomposition of a homogeneous precursor mixture.
It has been reported that preparation procedure has a determining
effect on the final selectivity for alcohols, with the decomposition
method leading to a better catalytic performance [23]. Unfortu-
nately, scaling up the decomposition procedure for industrial
application is not straight forward due to the difficulty in control-
ling the decomposition of the glassy intermediate generated during
the process [23].
oil and reduce the harmful impact on the environment, there is a
renewed interest in the direct synthesis of oxygenates, especially
higher alcohols via syngas derived from coal, natural gas, or bio-
mass [1–3]. As summarized in several reviews [3–14], four types
of catalysts have been studied to carry out CO hydrogenation for
the formation of higher oxygenates: Rh-based catalysts, Mo-based
catalysts, modified methanol synthesis catalysts, and modified
Fischer–Tropsch synthesis (FTS) catalysts. Rh-based catalysts have
been studied extensively, in part due to unique CO adsorption
behavior [15], and have been found to have relatively high selectiv-
ities for the synthesis of C2+ oxygenates. However, they have low
overall activities, and the high cost of Rh makes Rh-based catalysts
less competitive for industrial application. The latest progress in
Rh-based catalysts can be found in recent publications and refer-
ences herein [16–21].
Among the base metal–based catalysts for higher oxygenate
synthesis from syngas, modified FTS catalysts developed at Institut
Francais du Petrole (IFP) in the 1980s based on a combination of Co
To the best of our knowledge, the excellent pioneering work of
IFP has not been commercialized, probably due to the fluctuation
of oil prices and the complexities of the catalysts (i.e., multiple
components and difficulty in preparation), and the catalytic pro-
cess for the direct synthesis of alcohols from syngas. In addition,
mechanistic studies to understand the correlation between the
physicochemical properties of the IFP catalysts and their catalytic
performance have also been surprisingly limited. This lack of fun-
(
a hydrocarbon synthesis catalyst) and CuZnO (a methanol synthe-
sis catalyst) have been reported to have the best overall perfor-
⇑
Corresponding author.
E-mail address: jgoodwi@clemson.edu (J.G. Goodwin Jr.).
0
021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.09.033

X. Mo et al. / Journal of Catalysis 285 (2012) 208–215
209
damental and systematic studies of Co/CuZnO-based catalysts has
certainly not helped in the development of catalytic processes to
convert syngas to higher oxygenates with high selectivity.
TPR was carried out using a quartz reactor. The effluent gas was
analyzed by a Pfeiffer mass spectrometer, and an Omega tempera-
ture logger was used to monitor the temperature of the sample.
The as-prepared sample (0.3 g) was pretreated at 300 °C in He for
1 h prior to a TPR measurement. During a TPR experiment, 5 v/v
% H /Ar was used at 30 mL/min and the temperature was ramped
2
from room temperature (RT) to 750 °C at 10 °C/min, while the
The objective of this study was to probe the interaction of dif-
ferent components in model Co/CuZnO catalysts deduced from
industrial formulations patented by IFP. Besides ex situ character-
ization techniques, including BET surface area, porosity analysis,
XRD, and TPR, the intrinsic natures of the different catalysts were
also probed by SSITKA (steady-state isotopic transient kinetic anal-
ysis) and ethane hydrogenolysis. SSITKA has been suggested to be
the most powerful in situ technique to provide definitive informa-
tion about the amount of active intermediates and their activities
under typical reaction conditions [27,28]. Ethane hydrogenolysis
is a structure-sensitive reaction that can be used to characterize
the decoration of metal surfaces due to the fact that blocking of
just one surface metal atom in a multi-atom site renders that cat-
alytic site inactive [29–35]. Thus, these two methodologies can
provide more insight into the interaction of different components
in the model Co/CuZnO catalysts. In order to better understand
the effect of the various components on the catalyst activity, espe-
cially as relates to Co, SSITKA measurements of CO hydrogenation
were carried out under methanation conditions.
effluent gas was analyzed.
2.3. Reaction
CO hydrogenation (under FTS conditions) and ethane hydrogen-
olysis were performed in a fixed-bed differential reactor as de-
scribed elsewhere [36]. The catalyst was loaded between quartz
wool plugs in the middle of the reactor. A thermocouple close to
the catalyst bed was used to monitor and control the bed temper-
ature. Molecular sieve traps (Alltech) were used to remove H
2
O in
He, H , and CO (99.999%, National Welders). CO was further puri-
2
fied by a CO purifier (Swagelok). Prior to reaction, the catalyst
was reduced in situ in hydrogen (flow rate = 30 mL/min) at 1 atm
by ramping the temperature from RT to 300 °C at a ramp rate of
5 °C/min and then holding at 300 °C for 1 h, followed by cooling
down to reaction temperature. Preliminary analysis suggested that
varying reduction temperature (300–500 °C) or reduction duration
time (1–15 h) did not result in noticeable changes in catalytic
behavior of the catalysts in this work. Thus, 300 °C and 1 h were
used for the catalyst reduction procedure for all the reactions.
For CO hydrogenation (FTS), a catalyst sample (0.05 g for CoZnO
2
. Experimental
2.1. Catalyst preparation
Zn(NO
Alfa Aesar), and Co(NO
without further purification. CuZnO, CoCu, and CoZnO were pre-
pared by co-precipitation. In general, the desired amounts of aque-
ous nitrate solutions were mixed and then precipitated using
3
)
2
Á6H
2
O (99.998%, Alfa Aesar), Cu(NO
3
)
2
Á3H
2
O (99.5%,
and Co/Al
of an inert (
the gas flow was switched to H
and a total pressure of 1.8 atm. The choice of the pressure
1.8 atm) was based on our previous experience with SSITKA, since
2
O
3
, and 0.3 g for the other catalysts) was diluted with 3 g
-alumina). The reaction was carried out at 250 °C, as
(30 mL/min) + CO (15 mL/min)
3
)
2
2
Á6H O (synthetic, Aldrich) were used
a
2
(
2 3
Na CO aqueous solution at room temperature. The resulting mix-
ture was left in a fume hood overnight, and then the precipitate
was filtered, washed six times using boiled water, and dried for
operating at 1.8 atm leads to a smooth switch of feeds without dis-
turbing reaction. It is well known that reaction conditions play a
significant role in governing the catalytic performance. However,
the main focus of this study was not to provide direct data for
industrial application, but, instead, to probe the nature of active
sites using in situ techniques under the reaction conditions as close
to commercial application as reasonably possible.
1
2 h at 120 °C, followed by calcination in static air 350 °C for 4 h.
Co/CuZnO was prepared by the impregnation to incipient wetness
of as-prepared CuZnO with an aqueous solution of Co(NO
Á6H
1.5 mL solution / 1 g CuZnO), followed by drying for 12 h at 120 °C
3
)
2
2
O
(
and calcination in static air at 350 °C for 4 h. The final compositions
of all the catalysts in this study were determined by elemental
analysis, and some variance in molar ratio can be seen among
the catalysts due to the difficulty in obtaining the exact metal com-
positions of the catalysts using the precipitation method. It was
also found from elemental analysis that each catalyst contained
For ethane hydrogenolysis, 0.6 g (0.1 g for CoZnO and Co/Al
of the catalyst was diluted with 3 g of an inert -alumina. After
reduction, the catalyst bed was cooled to 260 °C and the reaction
started by flowing a gas mixture containing 30 mL/min H
0 mL/min He, and 1.5 mL/min C
A Varian 3380 GC was used to analyze the effluent gas on line.
2 3
O )
a
2
,
4
2 6
H .
ꢀ
0.2 wt.% Na. For comparison purposes, a Co (10 wt.%) catalyst
Hydrocarbons and oxygenates were separated by a Restek RT-
QPLOT column of I.D. 0.53 mm and length 30 m and were analyzed
by an FID. CO and other inorganic gases were analyzed by a TCD
supported on Al (Alfa-Aesar, -phase/ -phase, 99.98%) was also
2
O
3
c
a
prepared using the incipient wetness impregnation method. The
experimental error of catalyst preparation was <10% based on ele-
mental analysis and activity tests.
after separation with a Restek HayeSepÒ
Q
column of I.D.
3
.18 mm and length 1.83 m. The identification and calibration of
gas products were accomplished using standard gases and liquids.
2
.2. Catalyst characterization
The selectivity of a certain product was calculated based on carbon
P
efficiency using the formula n
i
C
i
/
i i i i
n C , where n and C are the car-
The metal compositions of the catalysts were determined by the
bon number and molar concentration of the ith product, respec-
tively. CO conversion in CO hydrogenation and ethane conversion
in ethane hydrogenolysis were all kept below ca. 5%. No mass or
heat transport limitations were observed for the gas-phase reac-
tions carried out under the specified reaction conditions.
Galbraith Laboratory in Knoxville, Tennessee (USA).
A Micromeritics ASAP 2020 was used to obtain BET surface area,
pore volume, and average pore size at À196 °C using N
Prior to N adsorption, the catalyst samples were degassed under
vacuum at 150 °C for 4 h.
Powder X-ray diffraction patterns were collected in a Scintag XDS
000 h/h powder X-ray diffractometer (XRD) equipped with Cu K 1/
2 (k = 1.540592 Å and 1.544390 Å, respectively) radiation with a
2
adsorption.
2
2.4. SSITKA
2
K
a
a
A steady-state isotopic transient kinetic analysis (SSITKA) sys-
tem as described elsewhere [37,38] was used to carry out isotopic
analysis in order to examine how activity for hydrocarbon synthe-
sis changed due to interaction of Co with the other components.
step size of 0.03° in the 2h range of 5–65°. In this study, XRD work
was carried out on as-prepared samples, and the effect of reduction
on the structure of the catalysts will be probed in a follow-up study.

2
10
X. Mo et al. / Journal of Catalysis 285 (2012) 208–215
The reaction conditions were the same as the standard FTS reac-
tion, except that the ratio of H :CO (containing 5% Ar) was 20:1 (to-
tal flow rate = 45 mL/min, CO:H :He = 1.5:30:13.5) and the
reaction conditions were 280 °C and 1.8 atm. The purpose of add-
ing a small amount of Ar to 12CO was to use Ar as an inert tracer
to determine gas-phase hold-up time. The use of a higher temper-
impregnation method, since our preliminary study found that Co/
CuZnO prepared this way resulted in similar activity, but higher
alcohol selectivity compared to a catalyst having the same compo-
sition but prepared using co-precipitation of all three metal com-
ponents. The loading ratios of the different metals were based on
the optimum ratios for the synthesis of alcohols reported in the lit-
erature [46,47], and no effort using statistical experimental design/
analysis was made to investigate catalyst synthesis in this study.
Due to the difficulty in obtaining the same ratio of metals as in
the starting materials using the precipitation method, some vari-
ance in molar ratio based on elemental analysis can be seen among
the catalysts. In addition, it was found from elemental analysis that
each catalyst contained ꢀ0.2 wt.% Na, which was residue from the
2
2
ature and a higher H
tion in CO hydrogenation to have CH
2
partial pressure shifts the product distribu-
as the primary product,
4
which simplifies the mass spectrometric (MS) analysis of the isoto-
pic transients. It has been, in general, accepted that the formation
of CH species is the first step for the formation of CH and any C2+
x 4
products (hydrocarbons or oxygenates), although the pathway for
CO dissociation is still not very clear. Therefore, the formation of
C2+ oxygenates share first step with CH
4
, and exploring the active
2 3
precipitation agent Na CO . It has been reported that alkali metals
sites under the methanation conditions also provides useful infor-
can have a large impact on the catalytic activity and selectivity of
Co-based and Cu-based catalysts for CO hydrogenation [48–52].
However, since all the catalysts in this study had similar amounts
of residual Na and the Na content was relatively low (it has been
reported that Na content 60.2 wt% does not lead to any significant
effects on the physicochemical properties of Co/CuZnO type cata-
lysts [51]), the effect of Na was not probed in detail in this study.
From Table 1, the BET surface area and pore volume of the cat-
alysts studied in this work followed the same trends with Co/
12
mation. Isotopic transient measurements started as CO (5% Ar)/
1
3
He was switched to CO/He after reaction reached steady state.
The same pressure for both 12CO and CO-feed streams was main-
tained by the back-pressure regulators in order to minimize per-
turbation of surface reaction during switching of the isotopically
labeled feed streams. The effluent gas was analyzed online by GC
for reaction rate and product distribution, and by a Pfeiffer mass
spectrometer (MS) with a high-speed acquisition system for isoto-
pic transients.
13
2
Al O
3
ꢀ CoZnO > CuZnO ꢀ Co/CuZnO > CoCu. Although Co/Al
2 3
O
Note that this study is part of a more extended investigation
using a variety of techniques under different conditions to gain
clues of the function of each component. The focus of this manu-
script is the effect of different components on the activity of Co/
CuZnO at the catalytic site level. The effect on the selectivity will
be addressed in detail in a follow-up study (manuscript is ac-
cepted) using multi-product SSITKA.
and CoZnO had at least 2X higher BET surface areas and pore vol-
umes than those for the other catalysts, the average pore sizes for
all the catalysts were around 22–30 nm. Since the BET surface area
measurements were carried out on as-prepared samples, the sur-
face areas of the catalysts under the reaction conditions after
in situ reduction may no longer follow exactly the trend as ob-
served for the as-prepared catalysts.
The structure and crystallite size of each calcined catalyst were
examined using powder XRD. It is worth pointing out that XRD can
only identify crystalline phases and cannot detect any X-ray amor-
phous phases that may also exist side by side with the detected
crystalline phases in these catalysts. As shown in Fig. 1, CuO (ICDD
card No. 74-1021) and ZnO (ICDD card No. 89-0511) structures
could be identified for CuZnO, but the peak intensities for CuO
were much stronger than those for ZnO. This observation is consis-
tent with the TEM results of Kasatkin et al. [53] in that ZnO and
3
. Results and discussion
3.1. Catalyst characterization
Table 1 lists the catalyst composition, preparation method, BET
surface area, and average pore volume and size for each catalyst
used for comparison in this study. It is well known that different
preparation methods can affect metal dispersion, reducibility,
and activity for CO hydrogenation catalysts [39–45]. The IFP cata-
lysts showing the best performance were prepared using a decom-
position method [23]; however, as discussed in the introduction,
this method has been reported to have problems with uncontrolled
decomposition of precursors [23], making the procedure not easily
adopted on an industrial scale. Thus, in this study, co-precipitation
was used to prepare CuZnO, CoCu, and CoZnO. Co/CuZnO was pre-
pared in this work by the addition of Co to CuZnO using the
2 3
Al O tend to show a lower degree of crystallinity than Cu. For
CoCu, only the diffraction peaks corresponding to CuO could be ob-
served, and for CoZnO, only the diffraction peaks corresponding to
Co
file of Co/CuZnO indicated the existence of CuO with relatively high
intensity and relatively weak peaks belonging to Co . The aver-
age Co crystallite sizes calculated using the Scherrer equation
for CoZnO and Co/CuZnO were 3.8 nm and 9.8 nm, respectively.
As indicated in our previous study [54], 10 wt.% Co loaded on
3 4
O (ICDD card No. 42-1467) were distinguishable. The XRD pro-
3 4
O
3 4
O
c-
Table 1
Composition, preparation method, BET surface area, pore volume, and average pore size of the catalysts studied.
Catalyst
Co wt.% a
Molar ratio a
Metal loading method b
BET surface area
Pore volume
(cm /g)
Average pore size
(nm)
2
d
3
e
e
(
m /g)
Co/Al
CuZnO
CoCu
CoZnO
Co/CuZnO
2
O
3
10
–
18.8
30.4
16.5
Impregnation
Co-precipitation
Co-precipitation
Co-precipitation
102.0
29.8
12.6
96.4
21.1
0.60
0.13
0.33
0.57
0.12
25.8
30.5
31.8
29.7
22.1
Cu:Zn = 2.5:1
Co:Cu = 1.0:2.5
Co:Zn = 1.1:1
c
Co:Cu:Zn = 1.1:2.5:1
Impregnation/co-precipitation
a
b
c
Based on elemental analysis.
All catalysts were calcined at 350 °C in static air.
CuZnO was first prepared by co-precipitation, followed by calcination at 350 °C; then, Co was added to CuZnO by the impregnation method, followed by calcination again
at 350 °C.
d
Max error = ±5%.
Max error = ±10%.
e

X. Mo et al. / Journal of Catalysis 285 (2012) 208–215
211
CuO. Apparently, the addition of Co to CuZnO in this study resulted
in the growth of CuO crystallites, most likely due to the sequential
preparation method used.
Fig. 2 presents the TPR profiles of the calcined catalysts. Since
the reduction of ZnO below 700 °C is negligible [55], all the reduc-
tion peaks observed in this study in the temperature range of 30–
*
*
Co/CuZnO
o
o
o
*
* o
*
o
7
50 °C can be ascribed to the reduction of CuO and/or Co
3 4
O . As
shown in Fig. 2, several features are worth noting:
o
CoZnO
o
o
o
o
(
(
(
1) A major peak at 200 °C along with a shoulder peak at 220 °C
can be observed in the TPR profile of CuZnO. These peaks
correspond to the reduction of CuO to metallic Cu with the
major peak corresponding to the one-step reduction of
*
*
*
*
CoCu
0
CuO to Cu and the shoulder peak corresponding to the
Cu to Cu0 reduction of a small fraction of Cu species [56–
58]. It can also be seen that the reduction temperatures for
CuZnO were much lower than that for bulk CuO, which is
around 400 °C [59]. This observation is consistent with the
literature [57] that ZnO promotes the reduction of CuO by
enhancing the dispersion of CuO.
+
+
CuZnO
*
+
+*
+
*+
*
2) The TPR profile of CoZnO consists of two peaks in the tem-
perature range of 200–300 °C and 300–450 °C, correspond-
ing to the reduction of Co
respectively [60,61]. Compared to the TPR profile of Co/
Al , it is apparent that Zn also promoted the reduction
of Co by shifting reduction peaks to lower temperatures,
3 4
O to CoO and CoO to Co,
x
Co/Al O
2
3
x
2 3
O
3 4
O
in agreement with the literature [62]. In addition, the forma-
tion of hard-to-reduce species due to the strong interaction
1
0
20
30
40
-theta
50
60
2
of Co and the support, such as Co–Al
Al -supported Co, was not observed for CoZnO. This is evi-
denced by the full reduction of Co by 500 °C.
x y
O in the case of
2 3
O
Fig. 1. Powder XRD profiles for the different calcined catalysts. (o: Co
ZnO; x: -Al ).
3
O
4
; Ã: CuO; +:
c
2 3
O
3) For the TPR profiles for CoCu and Co/CuZnO, only one major
peak and one shoulder peak can be observed, indicating that
the reduction peak(s) of CuO overlap with those of Co
these two catalysts. It has been reported by Dalmon et al.
51] that Cu can enhance the reducibility of Co more signif-
icantly than Zn, which is in agreement with this work, as evi-
denced by the much lower reduction temperature compared
to that of CoZnO. The higher mean reduction temperature for
Co/CuZnO compared to that for CoCu may be an indication
that ZnO interferes with the reduction of Cu and the reduc-
tion promotion of Co caused by Cu.
3 4
O for
Al
Co-related phases, suggesting high dispersion of Co. The weak dif-
fraction peaks around 37° and 46° for Co/Al correspond to the
-Al phase. The average CuO crystallite sizes were relatively lar-
2 3
O did not lead to any discernable diffraction peaks ascribed to
[
2 3
O
c
2 3
O
ger with Co/CuZnO having the largest particle size of 14.6 nm for
CuZnO
(4) The lower reduction temperature required for all the cata-
lysts containing Cu indicates that Cu acts as a strong reduc-
tion promoter for Co. There is no evidence, however, that it
affects the reducibility of ZnO, at least up to 600 °C.
Co/CuZnO
CoCu
3
.2. Catalytic activities for CO hydrogenation (FTS)
Table 2 lists the activities and selectivities of the catalysts for
2
CO hydrogenation (FTS: H /CO = 2) at 250 °C and 1.8 atm. The data
were collected at 15 h TOS after steady state had been reached. The
catalytic activity under similar reaction conditions for an Rh-based
catalyst from a previous study in our laboratory has also been
listed for comparison purposes. In this study, CoZnO exhibited
CoZnO
the highest CO hydrogenation rate (8.75 lmol/gcat/s) based on
the bulk weight of catalyst. However, the activity of CoZnO based
on Co content alone (see column 3 in Table 2) was lower than that
2 3
of Co/Al O . The combination of Cu and Co led to a much decreased
2
Co/Al O3
activity, which is surprising since Courty et al. [23] reported that
the addition of Cu to Co altered the selectivity but had little effect
on the activity. Three facts probably account for these seemingly
contradictory results. First, and probably most importantly, the
presence of large amounts of Cu (as in the CoCu catalyst in this
study) leading to lower activities for Co catalysts is due to the fact
that Cu and Co do not alloy. Because of their thermodynamic prop-
0
100
200
300
Temp ( C)
400
500
600
o
Fig. 2. TPR profiles for the different catalysts.

2
12
X. Mo et al. / Journal of Catalysis 285 (2012) 208–215
Table 2
Steady-state catalytic properties of catalysts for CO hydrogenation.
a
Catalyst
CO conv. (%)
Rateb
mol/gcat/s)
Rateb
mol/gCo/s)
Selectivityc (C atom%)
a Values
(
l
(l
CH
4
C
2+HCd
MeOH
Acetaldehyde
EtOH
Other C2+ oxy.e
HC
Oxy
*
Co/Al
2
O
3
2.84
0.10
3.92
0.30
0.10
0.53
6.35
0.038
8.75
0.11
0.038
0.20
63.5
–
28.8
0.59
0.23
–
49.1
1.8
38.0
27.1
16.9
18.7
48.5
–
60.1
67.7
21.9
33.2
1.2
98.2
0.3
1.3
14.8
5.7
0.2
–
0.3
0.5
4.3
4.4
0.8
–
0.8
2.4
26.9
34.6
0.4
–
0.5
1.0
15.2
3.4
0.52
0
0.65
0.52
0.35
–
0.20
0
0.17
0.21
0.22
–
CuZnO
CoZnO
CoCu
Co/CuZnO
Rh–La–Fe/V f [17]
a
Catalyst: 0.3 g (0.05 g for Co/Al
Data taken at TOS = 15 h.
2
O
3
and CoCu); Inert : 3 g
a
-alumina; Reduction at 300 °C; Reaction conditions: T = 250 °C, P = 1.8 atm; Flow rate = 45 mL/min (H
2
/CO = 2);
b
Error = ±5% of all the values measured.
Carbon selectivity = carbon efficiency = n
Hydrocarbons with 2 or more carbons.
Oxygenates with 2 or more carbons.
/P
n C .
i i
c
d
e
f
i
C
i
Reduction at 500 °C and reaction at 230 °C. Other pretreatment/reaction conditions were the same as those for Co/Cu-based catalysts.
value for oxygenates purposely excludes MeOH, which is made by a very different mechanism and is not related to chain growth.
*
The
a
erties, Cu atoms when present in the vicinity of Co particles may sit
on the Co surface, blocking active FTS sites of Co as observed by
other researchers [24,63]. This effect will be addressed in more de-
tail later using SSITKA. Second, reaction conditions used in the dif-
ferent studies were quite different, and it has been reported that
reaction conditions can have a significant effect on the activity of
Co/CuZnO-type catalysts [51,64,65]. Third, the existence of other
components, especially alkali species, can significantly affect the
catalytic behavior. With regard to selectivity, similar hydrocarbon
with Co was to shift some hydrocarbon synthesis to C2+ oxygenate
synthesis.
3.3. SSITKA
3.3.1. CO hydrogenation under methanation conditions
Table 3 summarizes the results of the SSITKA study under
methanation (H /CO = 20) conditions. As mentioned earlier, since
2
the current focus of our study is to investigate the effect of catalyst
distributions were observed for CoCu, CoZnO, and Co/Al
2
O
3
, show-
components on activity and in particular Co activity, carrying out
SSITKA under methanation conditions presents the most direct
way. It can be seen that the steady-state rate for methanation
was significantly larger for Co/Al O and significantly smaller for
ing mostly hydrocarbons. The selectivities for oxygenates for these
three catalysts were all <5%.
As expected, the main product for CuZnO was MeOH with a
selectivity of 98.2%. The activity of this catalyst was the lowest
among the catalysts in this study, which is consistent with our pre-
vious studies showing low activities for Cu and ZnO catalysts sup-
2
3
CuZnO than the rest of the catalysts. The steady-state rates for
methanation, however, followed the same trend as that for FTS
with H /CO = 2 on a Co basis (see Table 2): Co/Al O > CoZ-
2
2 3
ported on SiO
2
and Al
2
O
3
(<0.001
lmol/gcat/s for 0.5 wt.% Cu(ZnO)/
nO > CoCu > Co/CuZnO ꢁ CuZnO.
SiO
2
[66], ꢀ0.01
lmol/gcat/s for 10 wt.% Cu(ZnO)/Al
2
O
3
).
One obvious observation is that the relative activities for CoZnO
and CuZnO were much decreased compared to their relative activ-
ities under FTS conditions using Co/Al O as a ‘‘standard.’’ Since it
The Co/CuZnO catalyst exhibited the same activity (on a per g
catalyst basis) as that for CuZnO, but a quite different selectivity.
Much less hydrocarbons (than CoZnO and CoCu) and MeOH (than
CuZnO) but a significant increase in the selectivity for higher oxy-
genates were observed for Co/CuZnO. The overall performance of
this catalyst is close to the Rh-based catalysts from our precious
studies [17,36], except that the activity is much lower. It is obvious
that modifying Co with just Cu or ZnO only impacted Co activity.
Only the incorporation of all these three components led to a much
higher selectivity for C2+ oxygenates.
2
3
is well known that the activation energy for CO hydrogenation
should not differ significantly (observed activation energy is usu-
À1
ally between 70 and 105 kJ mol [68,69]) for different catalysts,
the dramatic change in relative catalytic activity for CoZnO and
CuZnO should not be caused just by the change in reaction temper-
ature (250 °C for FTS and 280 °C for methanation). Thus, the signif-
icantly decreased relative catalytic activity for these two catalysts
under methanation conditions must mainly be due to effects of the
different partial pressures of reactants. It is well known that the
rate of CO hydrogenation is also a strong function of the partial
In order to further understand the possible difference in reac-
tion pathways, chain growth probability factors for both hydrocar-
bons and oxygenates were also calculated based on an Anderson–
Schulz–Flory (ASF) product distribution. Since CO hydrogenation
pressures of H and CO [20,21,70,71]. Different catalysts can have
2
different partial pressure dependence depending on variations in
on CuZnO only produced MeOH and CH
ties were zero. As shown in Table 2, it is obvious that
for hydrocarbons for Co-containing catalysts were reasonably sim-
ilar with the exception of Co/CuZnO. Similar values for C2+ oxy-
4
, chain growth probabili-
the rate-limiting step, even when the same mechanism is operable.
Obviously, adding the complexity of producing hydrocarbons,
MeOH, and higher oxygenates increases the mechanistic complex-
ity and consequently the partial pressure variations. Apparently,
the combination of ZnO with Co was not as effective at higher par-
a
values
a
genates were also obtained for all the Co-containing catalysts,
suggesting that the use of these different components with Co does
tial pressures of H (methanation conditions) compared to at lower
2
not affect the higher oxygenate formation mechanism. However,
a
partial pressures of H (FTS conditions). This observation supports
2
values for hydrocarbons and oxygenates differed significantly, as
might be expected, given their related but slightly different mech-
anisms and active sites. This observation is in agreement with
the claim in the literature that one possible promoting function of
Zn is to supply atomic H to Cu metal through H spillover [65,72]. In
other words, the promoting effect of ZnO is not obvious under
those for a Cu/Co/ZnO/Al
2
O
3
/K catalyst [67] and Rh-based catalysts
methanation conditions, since the availability of H on catalyst sur-
2
[
16,21]. The HC was significantly lower for Co/CuZnO than for the
a
face is no longer a limiting fact due to the much higher partial pres-
other Co-containing catalysts, although still greater than that for
oxygenates. It appears that the effect of the combination of CuZnO
sure of H . Despite the much decreased relative activity compared
2
to regular CO hydrogenation for the ZnO-containing catalysts, the

X. Mo et al. / Journal of Catalysis 285 (2012) 208–215
213
Table 3
a
SSITKA results for CO hydrogenation on the catalysts studied.
c
d
Catalyst
CO conv.
%)
SS Rate b
mol C/gcat/s)
CH
(%)
4
selectivity
s
CH4 (s)
TOFITK
N
CH4
À1
(
(
l
(s
)
(
l
mol/gcat)
39.8
0.005
0.50
0.80
0.62
2
Co/Al O
3
6.9
10.47
0.001
0.13
0.30
0.08
98.1
67.2
93.7
99.6
92.9
3.7
7.9
4.2
2.7
8.3
0.27
0.13
0.24
0.37
0.12
CuZnO
CoCu
CoZnO
0.003
3.38
2.66
2.16
Co/CuZnO
a
b
c
Catalyst: 0.3 g (0.1 g for CoZnO), Inert :
Reaction conversion was less than 5% in all cases. Experimental error: ±10%.
TOF based on SSITKA, calculated as TOFITK = 1/
a
-alumina 3 g; Reaction at 280 °C; P = 1.8 atm; Flow rate = 45 mL/min (H
2
/CO/He = 30/1.5/13.5).
s .
CH4
d
i i i
N = Rate * s .
use of methanation conditions still permitted the determination of
the effect of different components on the surface reaction kinetic
parameters for CO hydrogenation, as will be discussed in more de-
tail in the following section.
2 3
Co. TOF for Co/Al O based on hydrogen chemisorption results
1
from our previous study [54] is 0.11 s , also much smaller than
À1
the TOF obtained from SSITKA (0.27 s ). This difference can be
mainly explained by the different methods used in determining ac-
tive sites. As demonstrated in previous findings of our research
2
group [37,75], TOF commonly calculated using H or CO chemi-
sorption is often an order of magnitude lower than that obtained
using SSITKA, due to an overestimation of the number of active
3
.3.2. Surface reaction kinetic parameters for CO hydrogenation on
different catalysts
Fig. 3 shows typical normalized transients after a switch from
2
reaction sites by both H and CO chemisorption. The results ob-
1
2
13
CO (5% Ar) to CO during CO hydrogenation under methanation
tained in this work once again suggest that the number of active
sites based on chemisorption is not reliable in an absolute sense.
With respect to the concentration of surface reaction intermedi-
conditions. SSITKA allows the determination of how different com-
binations of Co, Cu, and ZnO affect the average surface reaction res-
idence time of methane (sCH4, which is inversely related to the site
ates leading to CH
CH synthesis sites), Co/Al
mediates than CoZnO. CoZnO had slightly more CH
than CoCu and Co/CuZnO, but the difference was not significant
<50%). CuZnO had the lowest concentration of CH intermediates,
4
(and an approximation for the number of active
had about 50 times more CH inter-
intermediates
TOFITK [28]) and the concentration of active surface reaction inter-
mediates of methane (NCH4). The methods used to calculate the
average surface residence time and the concentration of active sur-
face intermediates have been described in detail elsewhere [28].
The TOFITK calculated based on the residence time of CH
cursors for different catalysts followed the trend of CoZnO > Co/
Al
ꢀ CoCu > CuZnO ꢀ Co/CuZnO, but the values were within a
factor of 3. The combination of ZnO with Co boosted the site activ-
ity of Co/Al by ca. 1/3; however, the combination of Co and Cu
seemed to have little effect on the activity of the Co sites (not sur-
prising, since Cu has essentially no activity for CO hydrogenation).
It can also be seen that the reaction sites on CuZnO and Co/CuZnO
were much less active. CuZnO is well known to be primarily a
MeOH synthesis catalyst.
4
O
2 3
4
4
(
4
4
pre-
about two orders of magnitude less than that for Co/CuZnO. This
observation also implies that there may be two completely differ-
ent sites for the formation of hydrocarbons and methanol, respec-
tively. Since under FTS conditions, the main product for CuZnO is
MeOH, it can be suggested that the formation of MeOH is dimin-
ished on CuZnO under methanation conditions, and these sites
are not active for the formation of CHx species. Once the reaction
2 3
O
2 3
O
4
conditions shifted to favor the formation of CH , the limitation in
reaction sites for the formation of CH was more obvious. Interest-
4
ingly, some earlier studies carried out by other research groups
found that the addition of Co to CuZnO led to much decreased
MeOH synthesis activity without significantly altering the forma-
tion of rates for other products (include hydrocarbons and oxygen-
ates) [45,76]. It was proposed by these authors that only a small
fraction of the measured Cu surface area is active for methanol for-
mation, and the role of Co was to block the pathway for methanol
formation. Another conclusion can be drawn here is that the addi-
2 3
It is noteworthy that the turnover rate for Co/Al O obtained
from SSITKA in this study after temperature, pressure, and reduc-
ibility correction was about one order of magnitude higher than
TOF data reported in the literature [73,74] for supported and bulk
1
0
0
0
0
0
.0
.8
.6
.4
.2
.0
1
2
CH
4
4
tion of Co to CuZnO led to a larger number of active CH synthesis
12
CO
sites based on the comparison of NCH4 for CuZnO and Co/CuZnO.
However, the increased amount of active methanation sites with
the addition of Co to CuZnO was still not comparable to that for
Ar
Co on Al
why Co/CuZnO had a much smaller methanation activity than
Co/Al
2 3
O , with even a smaller weight percentage of Co. This is
2 3
O .
3.4. Ethane hydrogenolysis
Ethane hydrogenolysis was chosen as a characterization reac-
tion to investigate the Co surface in the catalysts, since this reac-
tion has been reported to be very structure sensitive on Co and
to require site ensembles of ca. 12 metal atoms [29–31]. The main
purpose of using ethane hydrogenolysis in this study was to deter-
mine whether large Co ensembles exist after the addition of other
components; therefore, no extensive study, including theoretical
0
10
20
30
40
50
60
70
80
Relative time (s)
2 3
Fig. 3. Normalized isotopic transients for Co/Al O during CO hydrogenation under
methanation conditions.

2
14
X. Mo et al. / Journal of Catalysis 285 (2012) 208–215
Table 4
tained by in situ techniques, including SSIKA and ethane hydrogen-
olysis, is more meaningful.
Ethane hydrogenolysis results.a
4
CH formation rate (l
mol/gcat/s)b
Catalyst
Co/Al
Cu/Al
ZnO/Al
CuZnO
CoCu
Ethane conv. (%)
2
O
3
1.41
<0.01
<0.01
<0.01
0.46
0.16
0.0
0.0
4. Conclusions
2
O
3
2
O
3
Despite the fact that the Co/CuZnO-based catalysts have re-
ceived a great deal of attention in the past for the synthesis of high-
er alcohols, this study has explored for the first time at the site
level the effect of individual components on activity. Based on
0.0
0.009
0.004
0.0
CoZnO
Co/CuZnO
0.04
<0.01
a
2 3
BET analysis, XRD, and TPR for Co/Al O , CuZnO, CoCu, CoZnO,
Catalyst: 0.6 g (0.1 g for Co/Al
2
O
3
and CoZnO); Inert : 3 g a-alumina; Reduction
and Co/CuZnO, it was found that the different combinations of
Cu, Zn, and/or Co resulted in significant differences in surface area,
structure, crystallite size, and reducibility for the catalysts. These
catalysts also exhibited significantly different catalytic activities
and selectivities for CO hydrogenation under FTS conditions. ZnO
improved the dispersion of Co to some extent and might also en-
hance hydrogenation. Cu greatly enhanced reducibility of Co. Only
the combination of all these three metal components, however, led
to the formation of higher oxygenates with substantially higher
selectivity. There appears to be no correlation of surface area, par-
ticle size, and reducibility with activity and selectivity of the
catalysts.
In situ reaction characterizations using SSITKA at methanation
conditions and ethane hydrogenolysis were employed to develop
a better understanding of the possible interactions of the various
components. Although probably similar to what exists under CO
hydrogenation at alcohol synthesis conditions, some interactions/
catalyst structures may be affected somewhat at the different reac-
tion conditions. It was found from SSITKA under methanation con-
ditions that compared to Cu, the addition of ZnO to Co boosted the
site activity to some extent, while Cu seemed to have little effect
on the activity of Co. The most pronounced differences for these
catalysts were observed for the amounts of the reaction intermedi-
ates on the surface. The combination of Co with ZnO and/or Cu led
to much decreased surface concentrations of reaction intermedi-
at 300 °C; Reaction conditions: T = 260 °C, P = 1.8 atm; Flow rate = 71.5 mL/min (H
2
/
C
H
2 6
/He = 30/1.5/40); Data taken at TOS = 10 min. CH is the only product under the
4
reaction conditions.
b
Error = ±5% of all the values measured.
analysis for ethane hydrogenolysis, was carried out in this work.
Table 4 lists the activities of the different catalysts for ethane
hydrogenation at TOS = 10 min. Cu (20 wt.%) and ZnO (20 wt.%)
supported on Al
shown in Table 4, Cu/Al
no activity for ethane hydrogenolysis. For Co-containing catalysts,
only Co/Al showed comparable activity to other traditionally
O
2 3
were also tested for comparison purposes. As
2
O
3
and ZnO/Al as well as CuZnO had
2 3
O
2 3
O
supported Co catalysts [77]. All the other catalysts had very small
or negligible activity for ethane hydrogenolysis. These results sug-
gest that there were no or few relatively large ensembles (n P 12)
of Co surface atoms on Co/CuZnO, CoCu, or CoZnO. Thus, it is likely
that the different metal components in these Co/CuZnO-related
catalysts were in intimate contact with each other, and most of
the Co particles were decorated by ZnO or Cu or both.
3.5. Summary of the effect/interaction of the different components
Based on the results from physicochemical characterization,
reaction activities, and isotopic tracing, it is obvious that the signif-
icant differences in activity and selectivity cannot be simply as-
cribed to surface area, structure, crystallite size, or reducibility.
Higher surface area may contribute to the higher activity observed
4
ates for CH . The results for ethane hydrogenolysis indicated that
the different components for these multi-component catalysts
were in close contact, and few or no relatively large ensembles
(n P 12) of Co atoms existed on the surface in these catalysts.
for Co/Al
the activity difference under methanation conditions. The same
rationale can be applied to reducibility. Co/Al had the lowest
2 3
O and CoZnO under FTS conditions, but cannot explain
Thus, it can be suggested that Co/CuZnO achieves its high selectiv-
ity for oxygenates by blockage of a significant portion of the hydro-
carbon synthesis sites while retaining oxygenate synthesis sites,
but at the cost of the high activity normally expected for Co
catalysts.
2 3
O
reducibility at 300 °C, the reduction temperature used in this
study, but its catalytic activities were the highest per unit weight
of Co under both FTS and methanation conditions. Combining the
SSITKA and ethane hydrogenolysis results, the following sugges-
tions can be made:
Acknowledgments
This research was funded by the US Department of Energy
(
1) Cu helps in reduction of Co, but Cu does not improve Co
activity because it decorates the Co surface, blocking surface
reaction sites.
(
Award No. 68 DE-PS26-06NT42801). Dongsen Mao thanks the Sci-
ence and Technology Commission of Shanghai Municipality, China
08520513600), for financial support of his work at Clemson Uni-
(
(
2) Neither ZnO nor Cu by itself causes Co-containing catalysts
to produce a significant fraction of C2+ oxygenates.
3) ZnO also appears to decorate the Co surface, leading to fewer
surface reaction sites (both for methanation and ethane
hydrogenolysis).
versity as a visiting scholar.
(
References
[1] B. Cornils, Crit. Rep. Appl. Chem. 14 (1987) 1.
2] G.A. Mills, Fuel 73 (1994) 1243.
3] V. Subramani, S.K. Gangwal, Energy Fuels 22 (2008) 814.
4] S.A. Hedrick, S.S.C. Chuang, A. Pant, A.G. Dastidar, Catal. Today 55 (2000) 247.
[5] J.J. Spivey, A. Egbebi, Chem. Soc. Rev. 36 (2007) 1514.
[
[
[
(
4) Although the presence of both Cu and ZnO leads Co to selec-
tively make significantly more C2+ oxygenates, they also
appear to block significantly a large fraction of Co sites,
resulting in a diminished overall activity.
[6] K.G. Fang, D.B. Li, M.G. Lin, M.L. Xiang, W. Wei, Y.H. Sun, Catal Today 147
(
2009) 133.
[
[
7] G.J. Hutchings, South-African Tydskr. Chem. 39 (1986) 65.
8] K. Klier, A. Beretta, Q. Sun, O.C. Feeley, R.G. Herman, Catal. Today 36 (1997) 3.
It is worth noting that the surface morphology in this study was
elucidated mostly from SSITKA and ethane hydrogenolysis. Unfor-
tunately, electron microscopy is very limited for catalysts having
such loadings of metals as these, is an ex-situ method, and cannot
really distinguish active sites in any case. Thus, the information ob-
[9] M. Bowker, Catal. Today 15 (1992) 77.
10] R.G. Herman, Catal. Today 55 (2000) 233.
11] P. Forzatti, E. Tronconi, I. Pasquon, Catal. Rev. – Sci. Eng. 33 (1991) 109.
12] G. Van der Lee, V. Ponec, Catal. Rev. – Sci. Eng. 29 (1987) 183.
[13] X. Xu, E.B.M. Doesburg, J.J.F. Scholten, Catal. Today 2 (1987) 125.
[
[
[

X. Mo et al. / Journal of Catalysis 285 (2012) 208–215
215
[
[
14] S.S.C. Chuang, R.W. Stevens Jr., R. Khatri, Top. Catal. 32 (2005) 225.
15] J.P. Hindermann, G.J. Hutchings, A. Kiennemann, Catal. Rev. – Sci. Eng. 35
[48] N. Tien-Thao, M.H. Zahedi-Niaki, H. Alamdari, S. Kaliaguine, J. Catal. 245 (2007)
348.
(
1993) 1.
[49] N. Tien-Thao, M.H. Zahedi-Niaki, H. Alamdari, S. Kaliaguine, Appl. Catal. A-Gen.
326 (2007) 152.
[
[
[
[
16] J. Gao, X. Mo, J.G. Goodwin Jr., J. Catal. 275 (2010) 211.
17] X. Mo, J. Gao, N. Umnajkaseam, J.G. Goodwin Jr., J. Catal. 267 (2009) 167.
18] Y. Choi, P. Liu, J. Am. Chem. Soc. 131 (2009) 13054.
19] D. Mei, R. Rousseau, S.M. Kathmann, V.-A. Glezakou, M.H. Engelhard, W. Jiang,
C. Wang, M.A. Gerber, J.F. White, D.J. Stevens, J. Catal. 271 (2010) 325.
20] J. Gao, X. Mo, J.G. Goodwin Jr., J. Catal. 268 (2009) 142.
[50] A.M. Hilmen, M.T. Xu, M.J.L. Gines, E. Iglesia, Appl. Catal. A-Gen. 169 (1998)
355.
[51] J.A. Dalmon, P. Chaumette, C. Mirodatos, Catal. Today 15 (1992) 101.
[52] Ø. Borg, S. Eri, E.A. Blekkan, S. Storsæter, H. Wigum, E. Rytter, A. Holmen, J.
Catal. 248 (2007) 89.
[
[
21] N. Subramanian, J. Gao, X. Mo, J.G. Goodwin Jr., W. Torres, J.J. Spivey, J. Catal.
[53] I. Kasatkin, P. Kurr, B. Kniep, A. Trunschke, R. Schlögl, Angew. Chem. Int. Edit.
46 (2007) 7324.
2
72 (2010) 204.
[
22] P. Courty, P. Chaumette, C. Raimbault, P. Travers, Revue De L’INSTITUT France
Du Petrole 45 (1990) 561.
[54] Y.-T. Tsai, X. Mo, A. Campos, J.G. Goodwin Jr., J.J. Spivey, Appl. Catal. A-Gen. 396
(2011) 91.
[
[
23] P. Courty, D. Durand, E. Freund, A. Sugier, J. Mol. Catal. 17 (1982) 241.
24] R.M. Bailliardletournel, A.J.G. Cobo, C. Mirodatos, M. Primet, J.A. Dalmon, Catal.
Lett. 2 (1989) 149.
[55] M.S. Liang, W.K. Kang, K.C. Xie, J. Nat. Gas Chem. 18 (2009) 110.
[56] M. Richter, M.J.G. Fait, R. Eckelt, E. Schreier, M. Schneider, M.-M. Pohl, R. Fricke,
Appl. Catal. A-Gen. 73 (2007) 269.
[
[
[
[
[
[
[
[
[
[
25] A. Sugier, E. Freund, J.-F. Le Page, 1982, US 4346179.
26] A. Sugier, E. Freund, 1978, US 4122110.
27] S.L. Shannon, J.G. Goodwin, Chem. Rev. 95 (1995) 677.
28] J.G. Goodwin Jr., S. Kim, W.D. Rhodes, Catalysis 17 (2004) 320.
29] G.A. Martin, J. Catal. 60 (1979) 452.
30] G.A. Martin, Catal. Rev. Sci. Eng. 30 (1988) 519.
31] D.W. Goodman, Surf. Sci. 123 (1982) L679.
32] G.J. Haddad, B. Chen, J.G. Goodwin, J. Catal. 160 (1996) 43.
33] B. Chen, J.G. Goodwin Jr., J. Catal. 158 (1996) 511.
34] F.V. Pinto, A.S. Escobar, B.G. de Oliveira, Y.L. Lam, H.S. Cerqueira, B. Louis, J.P.
Tessonnier, D.S. Su, M.M. Pereira, Appl. Catal. A-Gen. 388 (2010) 15.
35] A.S. Escobar, M.M. Pereira, H.S. Cerqueira, Appl. Catal. A-Gen. 339 (2008) 61.
36] J. Gao, X. Mo, A.C. Chien, W. Torres, J.G. Goodwin Jr., J. Catal. 262 (2009) 119.
37] N. Lohitharn, J.G. Goodwin Jr., J. Catal. 257 (2008) 142.
38] J. Gao, X. Mo, J.G. Goodwin Jr., Catal. Today 160 (2011) 44.
39] R.T. Figueiredo, M.L. Granados, J.L.G. Fierro, L. Vigas, P.R. de la Piscina, N. Homs,
Appl. Catal. A-Gen. 170 (1998) 145.
40] V. Mahdavi, M.H. Peyrovi, Catal. Commun. 7 (2006) 542.
41] L. Liu, T.S. Zhao, Q.X. Ma, Y.F. Shen, J. Nat. Gas Chem. 18 (2009) 375.
42] A. Kaddouri, C. Mazzocchia, Catal. Lett. 131 (2009) 234.
43] N.N. Madikizela-Mnqanqeni, N.J. Coville, Appl. Catal. A-Gen. 272 (2004) 339.
44] N.N. Madikizela-Mnqanqeni, N.J. Coville, J. Mol. Catal. A-Chem. 225 (2005) 137.
45] R. Cao, W.X. Pan, G.L. Griffin, Langmuir 4 (1988) 1108.
46] F. Arena, K. Barbera, G. Italiano, G. Bonura, L. Spadaro, F. Frusteri, J. Catal. 249
[57] H.C. Yang, F.W. Chang, L.S. Roselin, J. Mol. Catal. A-Chem. 276 (2007) 184.
[58] S. Kieger, G. Delahay, B. Coq, B. Neveu, J. Catal. 183 (1999) 267.
[59] M. Gupta, J.J. Spivey, Catal. Today 147 (2009) 126.
[60] A.A. Khassin, T.M. Yurieva, G.N. Kustova, I.S. Itenberg, M.P. Demeshkina, T.A.
Krieger, L.M. Plyasova, G.K. Chermashentseva, V.N. Parmon, J. Mol. Catal. A-
Chem. 168 (2001) 193.
[61] A. Sirijaruphan, A. Horváth, J.G. Goodwin Jr., R. Oukaci, Catal. Lett. 91 (2003) 89.
[62] N.N. Madikizela, N.J. Coville, J. Mol. Catal. A: Chem. 181 (2002) 129.
[63] J.E. Baker, R. Burch, Y. Niu, Appl. Catal. 73 (1991) 135.
[64] M. Kurtz, J. Strunk, O. Hinrichsen, M. Muhler, K. Fink, B. Meyer, C. Wöll, Angew.
Chem. Int. Edit. 44 (2005) 2790.
[65] M.S. Spencer, Catal. Lett. 50 (1998) 37.
[66] X. Mo, J. Gao, J.G. Goodwin Jr., Catal. Today 147 (2009) 139.
[67] W.X. Pan, G.L. Griffin, J. Catal. 114 (1988) 447.
[68] G.A. Huff, C.N. Satterfield, Ind. Eng. Chem. Proc. Des. Dev. 23 (1984) 696.
[69] G.P. van der Laan, A.A.C.M. Beenackers, Catal. Rev. – Sci. Eng. 41 (1999) 255.
[70] B. Sarup, B.W. Wojciechowski, Can. J. Chem. Eng. 67 (1989) 62.
[71] I.A. Fisher, A.T. Bell, J. Catal. 162 (1996) 54.
[72] R. Burch, S.E. Golunski, M.S. Spencer, J. Chem. Soc., Faraday Trans. 86 (1990)
2683.
[
[
[
[
[
[
[
[
[
[
[
[
[73] E. Iglesia, Appl. Catal. A-Gen. 161 (1997) 59.
[74] F.H. Ribeiro, A.E.S. von Wittenau, C.H. Bartholomew, G.A. Somorjai, Catal. Rev.
– Sci. Eng. 39 (1997) 49.
[75] K. Sudsakorn, J.G. Goodwin Jr., A.A. Adeyiga, J. Catal. 213 (2003) 204.
[76] J. Wang, R. Oukaci, I. Wender, D.G. Blackmond, J. Catal. 153 (1995) 100.
[77] G.J. Haddad, J.G. Goodwin Jr., J. Catal. 157 (1995) 25.
(
2007) 185.
[
47] J.C. Slaa, J.G. Van Ommen, J.R.H. Ross, Catal. Today 15 (1992) 129.