Fe-promotion of supported Rh catalysts for direct conversion of syngas to ethanol

Article abstract of DOI:10.1016/j.jcat.2008.10.013

The influences of support (silica or titania) and loading of Fe promoter on the activity and selectivity of Rh-based catalysts for the direct synthesis of ethanol from syngas were explored. The reaction was performed in a fixed-bed reactor system typicall

Full text of DOI:10.1016/j.jcat.2008.10.013

Journal of Catalysis 261 (2009) 9–16
Contents lists available at ScienceDirect
Journal of Catalysis
www.elsevier.com/locate/jcat
Fe-promotion of supported Rh catalysts for direct conversion of syngas to ethanol
∗
Mohammad A. Haider, Makarand R. Gogate, Robert J. Davis
Department of Chemical Engineering, University of Virginia, 102 Engineers’ Way, P.O. box 400741, Charlottesville, VA 22904-4741, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The influences of support (silica or titania) and loading of Fe promoter on the activity and selectivity
of Rh-based catalysts for the direct synthesis of ethanol from syngas were explored. The reaction was
Received 1 September 2008
Revised 16 October 2008
Accepted 18 October 2008
Available online 13 November 2008
performed in a fixed-bed reactor system typically operating at 543 K, 20 atm, WHSV of 8000 cm3
−1 −1
g
h
cat
and H2:CO ratio of 1:1. Characterization by H2 chemisorption and electron microscopy indicated that
rhodium was very highly dispersed on the supports and was in direct contact with the Fe promoter.
Although little ethanol was produced over 2 wt% Rh on silica, a similar loading of Rh on titania was
active for this reaction. Promotion of 2 wt% Rh/SiO2 by 1 wt% Fe produced a catalyst that exhibited a
22% selectivity to ethanol, with methane being the primary side-product. Addition of Fe to 2 wt% Rh/
titania also improved the selectivity to ethanol with the highest selectivity being 37% for a sample with
5 wt% Fe. The effects of temperature, pressure and H :CO ratio on the performance of 2 wt% Rh/TiO
Keywords:
CO hydrogenation
Ethanol
Acetaldehyde
Rh/TiO2 catalyst
Rh–Fe/TiO2 catalyst
FTIR spectroscopy
CO adsorption
CO desorption
2
2
and 2 wt% Rh–2.5 wt% Fe/TiO2 were also studied. Although the influence of pressure and H2:CO ratio
was moderate, higher temperatures clearly increased methane production at the expense of ethanol and
methanol. Adsorption and thermal desorption of CO in Ar or H2 were also studied by DRIFTS spectroscopy
on 2 wt% Rh/TiO2 and 2 wt% Rh–2.5 wt% Fe/TiO2. The gem-dicarbonyl species that was the primary
species on these catalysts at room temperature after exposure to CO was more thermally stable on the
Fe-promoted catalyst.
©
2008 Elsevier Inc. All rights reserved.
1
. Introduction
synthesis from syngas is that supported Rh has a great potential
for the reaction [15–21], but suitable supports and promoters are
needed to enhance the reactivity of Rh and the current high cost
of Rh may hinder its commercial utilization.
One possible reaction scheme for the direct conversion of H2
and CO to ethanol is shown in Fig. 1 [22]. Although the exact in-
termediates on the metal surface are subject to debate, Fig. 1 is
useful to describe the basic features of the reaction network. Based
on this scheme, the four specific functions that a catalyst should
perform include:
A promising alternative to petroleum-derived fuels and chem-
icals is the utilization of coal, natural gas, and biomass to make
syngas (CO and H2) for the production of alcohols and hydrocar-
bons. Indeed, large-scale facilities for the production of methanol
and Fischer–Tropsch hydrocarbons continue to be built around the
globe. Although the conversion of syngas to ethanol and higher al-
cohols is also desirable, highly active and selective catalysts still
need to be developed.
Previous studies clearly show that ethanol can be produced
from syngas over a wide variety of heterogeneous catalysts in-
cluding supported Co [1,2], Ni [3], Cu [4–8], Pd [9], and MoS2
1. Dissociation of the adsorbed CO to form adsorbed carbon (Ca)
and oxygen (Oa).
[10–12] and two reviews on the catalysts for the direct syn-
2
3
4
. Hydrogenation of the adsorbed carbon to form an adsorbed
methyl species (CH3a).
. Insertion of non-dissociated CO into the methyl species to form
an adsorbed acyl species (CH3COa).
. Hydrogenation of the adsorbed acyl species to form the ethanol
product.
thesis of ethanol from syngas have recently appeared [13,14].
The catalysts for the production of ethanol and other light al-
cohols from syngas can be classified broadly into 4 categories:
(a) Rh-based catalysts, (b) modified high-temperature and low-
temperature methanol synthesis catalysts based on ZnO/Cr2O3 and
Cu/ZnO/Al2O3, respectively, (c) modified Fischer–Tropsch catalysts
based on Co, Fe, and Ru, and (d) modified unsulfided and sul-
fided Mo-based catalysts. A growing consensus regarding ethanol
Rhodium is claimed to be the best catalytic metal for this reac-
tion because of its ability to perform all of these functions [15].
0
Although large ensembles of Rh atoms adsorb CO dissociatively,
single Rh0 and Rh atoms adsorb CO non-dissociatively and act as
+
Corresponding author. Fax: +1 434 982 2658.
E-mail address: rjd4f@virginia.edu (R.J. Davis).
*
CO insertion sites [23].
0
021-9517/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2008.10.013

10
M.A. Haider et al. / Journal of Catalysis 261 (2009) 9–16
(Waterbury, CT)) was used as the precursor of rhodium for all cat-
alysts.
2
.1.1. Synthesis of unpromoted Rh catalysts supported on silica and
titania
A 2 wt% Rh/SiO2 catalyst was prepared by dissolving 0.129 g
Rh(NO3)3·2H2O in 6 mL of distilled deionized water and adding
this solution dropwise with proper kneading to 2 g of silica (Cabot
Cab-O-Sil M-5, pre-washed with HNO3) to the point of incipient
wetness. The paste was dried overnight in air at 413 K and subse-
quently calcined in air at 723 K for 4 h.
Titania-supported Rh catalysts were prepared in a similar man-
ner as described above, except titania (Degussa P-25) was used as
the support instead of silica, and the Rh concentration in the solu-
tion was adjusted to give the desired loading.
2
.1.2. Synthesis of Fe-promoted Rh catalysts supported on silica and
titania
As an example, a 2 wt% Rh–1 wt% Fe/TiO2 catalyst was prepared
Fig. 1. One possible reaction scheme for direct conversion of H2 and CO to form
ethanol (adapted from [22]).
by dissolving 0.315 g Rh(NO3)3·2H2O and 0.36 g Fe(NO3)3·9H2O in
Burch and Petch showed that Rh particles on a high purity silica
support produced very small amounts of ethanol from a glass-
lined reactor, with major products being hydrocarbons and a small
amount of acetaldehyde [15]. However, addition of Fe to the cata-
lyst significantly improved the product distribution toward ethanol.
This study was a warning to the catalysis community since volatile
iron carbonyls can deposit onto an Rh catalyst during CO reactions
taking place inside a steel reactor. The presence of iron was pro-
posed to stabilize an adsorbed acetyl species (CH3CHOa), thus fa-
voring subsequent hydrogenation to ethanol instead of desorption
of acetaldehyde [15]. The promotion of Rh is certainly not limited
to Fe. In fact, a wide variety of supports as well as promoters can
be used to affect the size, nature and the number of active sites for
ethanol formation. Although supports such as ceria [18,24], vana-
dia [19], niobia [25], zirconia [26], and promoters such as V, Nb,
Ta [27], Sm [28], and Mo [29] have a positive influence on the re-
action, Mn also appears to be a favorable promoter [20,21,30–33].
Examples of other prior art on the catalytic synthesis of hydro-
carbons and oxygen-containing compounds from CO and H2 over
supported Rh catalysts include the works of Ichikawa et al. [34,35],
Arakawa et al. [36,37], Prins et al. [38,39], and Niemantsverdriet et
al. [40,41].
Chuang et al. have summarized the current understanding of
ethanol synthesis on supported Rh [23]. The main products from
syngas reactions appear to be methane and C2 oxygenates, which
suggests chain growth pathways that form higher hydrocarbons
are not important in the reaction network. Oxophilic promoters
are claimed to enhance both the CO dissociation step needed to
form C atoms on the surface as well as the CO insertion step that
accounts for C2-oxygenates on the surface. More importantly, evi-
dence is provided to support the idea that CO insertion occurs on
a single Rh atom and that a positively-charged Rh atom is more
active than a neutral one. Thus, a proper balance of CO dissocia-
tion, multiple hydrogenation steps, and CO insertion is required to
maximize alcohol formation while minimizing methane formation.
The objective of this paper is to clarify the important roles of
support (silica or titania) and Fe promoter on the activity of Rh
for the direct synthesis of ethanol from syngas. Diffuse reflectance
infrared Fourier transform spectroscopy (DRIFTS) was also used to
probe the interaction of CO with the Rh catalysts.
7.5 mL of distilled deionized water and adding this solution drop-
wise with proper kneading to 5 g TiO2 (Degussa P-25) to the point
of incipient wetness. The paste was dried overnight in air at 413 K
and subsequently calcined in air at 723 K for 4 h.
Iron-promoted Rh catalysts on silica or titania with various
loadings of Fe were prepared in a similar manner with appropriate
concentrations of metal precursors.
2.2. Catalyst characterization
The metal loadings of the catalysts were determined by Gal-
braith Laboratories (Knoxville, TN) using ICP analysis. Additionally,
the Rh dispersion was evaluated by H2 chemisorption performed
using a Micromeritics ASAP 2020 automated adsorption system at
3
2
08 K. Calcined catalysts were reduced in flowing H2 at 473 K for
.5 h and evacuated at that temperature for 2 h prior to cooling
in vacuo to 308 K. An H/Rhsurf stoichiometry equal to unity was
assumed in the analysis.
Electron microscopy was performed on an FEI TITAN field-
emission transmission electron microscope equipped with an EDAX
Genesis EDS system operated at 200 kV. Catalyst samples in the
powdered form were sprinkled onto a copper grid with lacy carbon
support film. Images were recorded with a slow scan CCD camera.
To analyze individual particles, the electron beam was focused on
the middle of the target particle with a diameter slightly lesser
than the diameter of the target particle. The EDS spectra were an-
alyzed using FEI TIA software.
2
.3. Reactions of CO and H2
The catalysts were evaluated in
a fixed-bed micro-reactor
system (Autoclave Engineers’ BTRS Jr., Erie, PA). Calcined cata-
lysts were first pressed into pellets, crushed, and size-separated
(
+40/−80 mesh). Approximately 0.15 g catalyst was intimately
mixed with 2.5 g of silicon carbide (Universal Photonics, Inc.,
Hicksville, NY) and loaded into the reactor. The catalyst was then
3
−1
reduced in flowing dihydrogen (20 cm (STP) min ) at 573 K for
h under atmospheric pressure. Following reduction, the catalysts
were tested at nominally identical conditions of 543 K, 20 atm to-
2
3
−1
tal pressure, syngas (H2 + CO) flow of 20 cm (STP) min , H2:CO
3
−1 −1
ratio of 1:1, and a WHSV of 8000 cm g h
cat
. The catalyst bed
2. Experimental methods
temperature was monitored by a thermocouple inserted within the
catalyst bed. All gases (CO (GT&S) and H2 (GT&S)) were UHP grade
2.1. Catalyst synthesis
(99.999%). Additionally, CO was purified by passing it through a
The supported Rh catalysts were synthesized by incipient wet-
ness impregnation. Granular rhodium nitrate (Pfaltz and Bauer
silica trap immersed in a dewar containing a dry ice–acetone mix-
ture before introduction into the reactor. The results reported here

M.A. Haider et al. / Journal of Catalysis 261 (2009) 9–16
11
Table 1
Conversion of syngas over silica- and titania-supported Rh catalysts.
Catalyst
Conversion
%)
%Selectivity
CH4
(
C2H6
C3H8
C4H10
CH3CHO
CH3OH
EtOH
PrOH
EthAc
2
2
% Rh/SiO2
0.75
1.52
51.3
48.0
19.5
5.5
24.2
16.0
5.1
–
0.0
14.0
0.0
0.6
0.0
0.7
0.0
–
0.0
2.7
% Rh/SiO2a
(
Burch and Petch [15])
2
2
% Rh–1% Fe/SiO2
% Rh–1% Fe/SiO2
2.70
4.45
35.5
38.0
9.8
3.1
9.1
2.8
2.0
–
2.8
0.9
13.5
11.0
21.8
39.0
3.7
–
1.9
1.2
(
Burch and Petch [15])
2
% Rh/TiO2
5.66
7.47
47.4
42.4
3.2
4.9
14.8
10.4
5.2
3.4
5.7
13.3
1.9
1.6
11.1
13.4
0.0
0.0
10.7
10.6
2% Rh–1% Fe/TiO2
3
−1
Note. Nominal conditions are T = 543 K, P = 20 atm, 0.150 g catalyst (40–80 mesh), 2.5 g SiC to dilute the bed, H2:CO 1:1, syngas flow = 20 cm (STP) min . Conversion
ꢀ
ꢀ
(
%) = ni Mi × 100/MCO and selectivity = ni Mi / ni Mi where ni is the number of carbon atoms in product i, Mi is the mole percent of product i measured, and MCO is
the mole percent of carbon monoxide in the feed.
a
The balance of selectivity (5.9%) is to acetic acid.
are generally obtained after 2 h on-stream. Two HP 5890 Series
II gas chromatographs were integrated downstream of the reac-
tor for the analysis of reactants and products. The first one was
equipped with a flame ionization detector and a 50 m-long HP-1
cross-linked methyl silicone gum capillary column to monitor hy-
drocarbons, alcohols, acetaldehyde, methyl acetate (not observed)
and ethyl acetate. The second one was equipped with a thermal
conductivity detector and a 6-ft long Alltech CTR packed column
and was used to monitor the possible formation of CO2.
amounts of ethanol were produced over this catalyst at our stan-
dard conditions. The only products at the conditions of 543 K,
3
−1
20 atm, H2:CO 1:1, a syngas flow of 20 cm min , and a space
3
−1 −1
were methane and light hydrocar-
velocity of 8000 cm g
h
cat
bons. Only trace amounts of ethanol, methanol, acetaldehyde or
ethyl acetate were detected over the unpromoted Rh/SiO2 cata-
lyst. Burch and Petch [15] have also shown that hydrocarbons are
generally the preferred reaction products on unpromoted rhodium
supported on silica, although acetaldehyde and ethanol can be mi-
nor co-products.
The conversion of CO was based on the fraction of CO that
formed carbon-containing products according to:
Next, a promoted catalyst similar to that used by Burch and
ꢁ
ꢂ
ꢃ
Petch (2 wt% Rh–1 wt% Fe/SiO ) [15] was prepared and tested in
our laboratory at conditions similar to those reported by these au-
2
%
Conversion =
ni · Mi/MCO · 100%,
thors (543 K, 20 atm, 0.150 g catalyst, 2.5 g SiC as the bed diluent,
where ni is the number of carbon atoms in product i, Mi is the
percentage of product i detected, and MCO is the percentage of
carbon monoxide in the syngas feed. This equation was valid since
differential conversion was maintained throughout this study.
The selectivity to product i is based on the total number of
carbon atoms in the product and is therefore defined as:
3
−1
syngas 20 cm min , and H2:CO ratio of 1:1). The results summa-
rized in Table 1 for the Burch and Petch-type catalyst (third entry)
compare well to those reported in their paper [15]. The main prod-
ucts were methane, light hydrocarbons (ethane and propane), and
light alcohols (methanol and ethanol).
Table 1 also shows a direct comparison of Rh supported on
silica to Rh supported on titania. The titania-supported catalyst
was not only more active for syngas conversion, but also produced
a variety of oxygenate products including methanol, ethanol, ac-
etaldehyde and ethyl acetate. Adding 1 wt% Fe to a 2 wt% Rh/TiO2
catalyst further improved the CO conversion and oxygenate selec-
tivity. Since both silica and titania alone were inactive for syngas
conversion under these conditions, the product formation observed
in this study is attributed solely to the transition metal compo-
nents (Rh/Fe).
The influence of silica and titania support on CO hydrogena-
tion over supported Rh was also investigated in detail by Katzer
et al. [42], albeit at different reaction conditions. The conditions
used in their work were 473 K, 10 atm, and 4% CO in dihydrogen.
Their results also indicate that titania is the more active support
for Rh. They reported a 200-fold variation in the product forma-
tion rate for titania compared to silica, which was attributed to an
increase in the number of active sites for the reaction instead of
the changes in the energetics of the reaction.
The effect of iron on the performance of supported rhodium
catalysts for CO hydrogenation reactions were also investigated in
a pioneering study by Bhasin et al. [43] and later by Guglielminotti
et al. [44]. Their results also agree with our findings that iron has
a profound effect on the selectivity to ethanol. The first set of
authors [43] investigated the performance of Rh and Rh–Fe/SiO2
catalysts using a Berty-type backmixed reactor at nominal con-
ditions of 573 K, 1000 psig, and H2:CO ratio of 1:1. A nominal
loading of 2.5 wt% Rh was used along with five different Fe load-
ings (0.00 wt%, 0.05 wt%, 0.10 wt%, 0.20 wt%, and 0.50 wt%). Their
results show that the performance of a 2.5% Rh–0.20% Fe/SiO2 cat-
alyst was superior to that of the unpromoted 2.5% Rh/SiO2 catalyst
ꢁ
ꢂ
ꢃ
Si = (ni · Mi)/
ni · Mi .
2
.4. DRIFTS studies
A Bio-Rad FTIR (FTS-60A) spectrometer outfitted with an MCT
detector and a Harrick DRIFTS cell were used for the IR studies. The
cell allowed collection of spectra in a controlled environment over
a range of temperatures (298–548 K). To obtain the spectra pre-
sented here, 100 scans were co-added at resolution of 2 cm . The
procedure for collection of the DRIFTS spectra was as follows. First,
0–40 mg of powdered catalyst was combined with 70 mg of pow-
dered KBr and the mixture was placed onto the sample holder. The
powder surface was carefully flattened to obtain high IR reflectiv-
−1
3
3
−1
)
ity. The sample was then heated under flowing Ar (50 cm min
to 573 K, reduced in flowing H2 at 573 K for 2 h, purged in flow-
ing Ar at 573 K, and then cooled to 298 K in flowing Ar. After
collecting a background spectrum, CO was admitted to the cell at
3
−1
for approximately 10 min. The gas-phase CO was
50 cm min
purged from the cell by flowing Ar and a spectrum of adsorbed
CO was collected after purging for 60 min. Thermal desorption ex-
periments (heating rate 10 K min ) were also performed in either
flowing Ar or H2.
−
1
3
. Results and discussion
3
.1. Influence of support
The first entry in Table 1 illustrates how unpromoted Rh/SiO2
is a very poor catalyst for syngas conversion. Indeed, only trace

12
M.A. Haider et al. / Journal of Catalysis 261 (2009) 9–16
in terms of selectivity to ethanol. The selectivity to ethanol for the
promoted catalyst was 29.2%, while that for the unpromoted cat-
alyst was only 17%. It also appears that the increase in selectivity
to ethanol is at the direct expense of methane selectivity, which
reduced to 42.1% for the promoted catalyst from a value of 52.0%
for the unpromoted catalyst. Similar results were also reported by
Guglielminotti et al. [44] who found that adding Fe to Rh/ZrO2 in-
creased the selectivity of methanol and ethanol. Their experiments
were carried out at 493 K, 1 atm, and a H2/CO ratio of 3.
Table 3 facilitated the promotional role of the support titania on
the catalysis by Rh.
3.3. Influence of Fe loading on Rh–Fe/TiO2
Four different catalysts were prepared and tested to examine
the effect of Fe loading and the results are summarized in Table 4.
The first three entries in Table 4 present reactivity results for
three separate runs of a 2 wt% Rh–1 wt% Fe/TiO2 catalyst. The
results indicate good experimental reproducibility since the con-
version of CO in these three experiments was at 7.47%, 9.02%, and
8.23%.
Hanaoka et al. [45] examined the effect Rh loading on CO hy-
drogenation over Rh/SiO2 catalysts at 553 K, 50 atm, H2:CO:Ar =
3
−1
4
5:45:10, and a flow rate of 100 cm min . The Rh loading was
varied from 1 wt% to 30 wt%. Although increasing the Rh load-
ing improved the CO conversion, the selectivity to various products
was also altered as a function of Rh loading. They found that the
selectivity to methane increased monotonically with an increase in
Rh loading, at the expense of CH3OH and C2H5OH.
The well-known suppression of chemisorption achieved by
high-temperature reduction of metal particles on titania occurs by
the formation of TiOx overlayers on the active metal and is often
termed as strong metal support interaction, or SMSI [46]. Indeed,
Logan et al. [47] have directly observed an amorphous titania film
that encapsulates titania-supported Rh particles. To investigate the
possible effect of TiOx overlayers on the activity of promoted cata-
lysts, we evaluated the activity and selectivity of 2 wt% Rh–1 wt%
Fe/TiO2 after a high-temperature reduction (HTR) at 748 K and
compared the results to those obtained after our standard low-
temperature reduction (LTR) at 573 K.
The fourth entry in Table 4 summarizes the results from a cat-
alyst after a high-temperature reduction. The CO conversion and
product selectivity are within the experimental error of those as-
sociated with the catalyst reduced at low temperature (first three
entries of Table 4). Similar conclusions were reported by Katzer
et al. [42] who found that although the CO and H2 chemisorption
capacity changes as a function of the reduction temperature, the
rate of CO hydrogenation and the product selectivities for Rh/TiO2
catalysts reduced at 673 K were essentially the same as for Rh/TiO2
reduced at 473 K. Moreover, Van’t Blik et al. [38] also reported
that the reduction temperature did not influence to a great ex-
tent CO hydrogenation over Rh/Al2O3 and Rh/TiO2. Interestingly,
Frydman et al. [25] reported that the rates of CO hydrogena-
tion on monometallic Rh catalysts supported on Nb2O5 decreased
substantially upon high-temperature reduction, whereas the CO
hydrogenation rates for bimetallic catalysts comprised of Co–Rh did
3
.2. Role of Rh loading on TiO2
The effect of Rh loading on the activity and selectivity of
titania-supported catalysts is summarized in Table 2. The results
show that CO conversion over the catalysts increased with the
Rh loading on titania, as expected. The selectivities reported for
the low loaded catalyst are somewhat less reliable than the other
entries because of very low level of conversion. In general, the se-
lectivities to C1–C4 hydrocarbons and to oxygenates were affected
little by the Rh loading, in contrast to the results of Hanaoka et
al. [45] who reported increasing CH4 selectivity with increasing Rh
loadings.
Table 3 summarizes the results from H2 chemisorption and
elemental analysis. Although the low loaded Rh/TiO2 catalysts
(0.5 wt% and 1 wt%) exhibited 100% dispersion, higher loadings
of Rh resulted in lower dispersions. Nevertheless, the calculated
turnover frequency that includes both Rh loading and dispersion of
−1
Rh on titania was fairly constant at about 0.02 s
at our standard
conditions. The relatively high dispersion of all the catalysts in
Table 2
Effect of Rh loading on syngas conversion over Rh/TiO2 catalysts.
Table 3
Catalyst
Conver- %Selectivity
sion (%)
Results from elemental analysis and H2 chemisorption used to determine TOF for
Rh/TiO2 catalysts.
CH4 C2H6 C3H8 C4H10 CH3CHO CH3OH EtOH EthAc
0
1
2
5
.5% Rh/TiO2 1.38
60.3 0.0
50.6 4.1
47.4 3.2
42.5 4.1
14.7 5.3
12.4 3.8
14.8 5.2
16.9 6.3
3.7
3.5
5.7
6.8
5.6
3.8
1.9
1.5
10.4
12.2
11.1 10.7
10.4 11.5
0.0
9.7
Catalyst composition
nominal)
Rh loading
(wt%)
H/Rh
(%)
Turnover
frequency (s
% Rh/TiO2
% Rh/TiO2
% Rh/TiO2
3.80
5.66
7.43
a
−1
)
(
0.5% Rh/TiO2
0.44%
0.80%
1.60%
4.40%
∼100%
∼100%
88%
54%
46%
0.0149
0.0222
0.0206
0.0159
0.0038
1
2
% Rh/TiO2
% Rh/TiO2
Notes. Nominal conditions are T = 543 K, P = 20 atm, 0.150
g
catalyst (40–
3
−1
.
8
0 mesh), 2.5 g SiC to dilute the bed, H2:CO 1:1, syngas flow = 20 cm (STP) min
ꢀ
ꢀ
5% Rh/TiO2
Conversion (%) =
ni Mi × 100/MCO and selectivity = ni Mi /
ni Mi where ni is
b
2% Rh/SiO2
Not determined
the number of carbon atoms in product i, Mi is the mole percent of product i mea-
sured, and MCO is the mole percent of carbon monoxide in the feed.
a
Molecules of CO converted per Rh surface atom per second.
Nominal.
b
Table 4
Effect of Fe loading on Rh–Fe/TiO2 catalysts for syngas conversion.
Catalyst
Conversion
%)
%Selectivity
CH4
(
C2H6
C3H8
C4H10
CH3CHO
CH3OH
EtOH
EthAc
2% Rh–1% Fe/TiO2
2% Rh–1% Fe/TiO2
2% Rh–1% Fe/TiO2
2% Rh–1% Fe/TiO2 (HTR)
2% Rh–2.5% Fe/TiO2
2% Rh–5% Fe/TiO2
2% Rh–10% Fe/TiO2
7.47
9.02
8.23
8.65
9.28
6.20
5.68
42.4
43.6
43.0
44.1
37.9
35.3
37.3
4.9
3.5
3.2
2.5
2.7
2.9
5.2
10.4
10.8
9.7
8.0
4.3
3.3
4.2
3.4
3.9
3.5
2.6
1.1
0.8
0.9
13.3
13.3
13.5
13.7
11.2
7.0
1.6
0.0
0.0
13.4
14.4
15.2
17.8
31.0
37.2
33.2
10.6
10.6
11.9
11.4
9.0
0.0
2.8
4.8
14.9
8.6
2.1
2.2
3
−1
Note. Nominal conditions are T = 543 K, P = 20 atm, 0.150 g catalyst (40–80 mesh), 2.5 g SiC to dilute the bed, H2:CO 1:1, syngas flow = 20 cm (STP) min . Conver-
ꢀ
ꢀ
sion (%) =
ni Mi × 100/MCO and selectivity = ni Mi / ni Mi where ni is the number of carbon atoms in product i, Mi is the mole percent of product i measured, and MCO
is the mole percent of carbon monoxide in the feed.

M.A. Haider et al. / Journal of Catalysis 261 (2009) 9–16
13
Table 5
Results from elemental analysis and H2 chemisorption used to determine TOF for
Rh–Fe/TiO2 catalysts.
Catalyst composition
nominal)
Rh and Fe loadings
(wt%)
H/Rh
(%)
Turnover
frequency (s
a
−1
)
(
2% Rh–1% Fe/TiO2
2% Rh–2.5% Fe/TiO2
2% Rh–5% Fe/TiO2
2% Rh–10% Fe/TiO2
2% Rh–1% Fe/SiO2
1.81% Rh–0.83% Fe
2.03% Rh–2.27% Fe
1.62% Rh–3.97% Fe
1.91% Rh–8.97% Fe
Not determined
49%
46%
38%
20%
0.0472
0.0511
0.0515
0.0767
0.0257
b
26%
a
Molecules of CO converted per Rh surface atom per second.
Nominal.
b
not vary appreciably as a function of reduction temperature. The
authors concluded that the combination of the two metals some-
how suppressed the negative influence of the high temperature
reduction. This phenomenon appears to be a function of the type
of catalyst, support, and other physicochemical properties of the
catalyst.
The next three entries in Table 4 illustrate the importance of Fe
loading on the catalyst. The results show that addition of 2.5 wt%
Fe improved the CO conversion while simultaneously increasing
the ethanol selectivity from 13.4% to 31.0%. Addition of 5 wt% Fe
reduced the conversion of CO, but improved the ethanol selectiv-
ity to 37.2%. Addition of 10 wt% Fe continued to reduce the CO
conversion on the catalyst, without an additional gain in ethanol
selectivity. Increasing the Fe content of the Rh/TiO2 catalysts also
reduced the ethyl acetate and acetaldehyde selectivity. Iron has
been claimed to be an effective catalyst for the hydrogenation of
acetaldehyde to ethanol in presence of Rh [15].
Fig. 2. Electron micrograph of a 2 wt% Rh–2.5 wt% Fe/TiO2 catalyst.
These results from Fe-promotion are in general agreement with
those of Burch and Hayes [17], who studied CO hydrogenation over
2
% Rh/γ -Al2O3 promoted by Fe (up to 10 wt%) at 543 K, 10 atm,
3
−1 −1
, and H2:CO ratio of 1:1. Their study,
WHSV of 8000 cm g
h
cat
however, revealed a more substantial enhancement of ethanol se-
lectivity (from ca. 2% over an unpromoted Rh/γ -Al2O3 to 48%
over a Fe-promoted catalyst). Furthermore, the increase in ethanol
selectivity was at the expense of methane selectivity, which de-
creased from ca. 48% over an unpromoted catalyst to about 23%
over the Fe-promoted one.
The enhanced selectivity of catalysts toward ethanol formation
with an increasing amount of promoter has been attributed to
the close interaction between the metal and the promoter, i.e.,
an increasing interfacial contact between the metal and the pro-
moter [17]. Thus, high loadings of Fe promoter will suppress H2
chemisorption by a simple covering of the active Rh surface. Ta-
ble 5 illustrates the observed decrease of exposed Rh with increas-
ing Fe loading on Rh/TiO2 catalysts. Table 5 also illustrates how Fe
promotes the turnover frequency for CO conversion. Evidently, the
interfacial sites are substantially more active than the unpromoted
Rh/TiO2 catalysts.
Fig. 3. EDS spectra of particles identified in the electron micrograph in Fig. 2 (spec-
tra are offset for clarity and are truncated for Ti and Cu).
Table 6
Effects of temperature and pressure on the reactions of syngas over a 2 wt% Rh/TiO2
catalyst.
Condition
1
2
3
4
5
T (K)
P (atm)
503
20
20
528
20
20
543
20
20
543
13.6
20
543
34
20
Since the results from H2 chemisorption and catalytic reac-
tion indicated a substantial interaction between Fe and Rh, we
performed electron microscopy on a representative sample. The
electron micrograph of 2 wt% Rh–2.5 wt% Fe/TiO2 in Fig. 2 shows
3
−1
)
H2 + CO (cm min
H2/CO
1:1
1:1
1:1
1:1
1:1
Conversion (%)
0.83
3.11
5.66
7.08
6.49
∼
2 nm bimetallic particles of Rh and Fe dispersed on titania. The
Selectivities (%)
CH4
C2H6
C3H8
C4H10
36.5
0.0
11.5
5.5
41.8
4.1
13.6
4.9
47.4
3.2
14.8
5.2
48.1
4.2
15.5
5.1
47.9
4.2
13.0
3.7
corresponding EDS spectra of the four individual particles marked
in Fig. 2 are presented in Fig. 3. In each case, both Rh and Fe are
present in the particle.
CH3CHO
CH3OH
C2H5OH
Ethyl acetate
CO2
5.6
4.5
4.4
13.5
13.1
0.0
5.7
1.9
11.1
10.7
0.0
4.2
2.1
11.3
9.3
0.0
4.4
2.2
11.4
13.2
0.0
10.8
16.2
13.9
0.0
3
ratio
.4. Influence of process variables: temperature, pressure, and H2:CO
ꢀ
ꢀ
The effects of temperature and pressure on the conversion and
selectivity of an unpromoted 2 wt% Rh/TiO2 and promoted 2 wt%
Rh–2.5 wt% Fe/TiO2 are summarized in Tables 6 and 7, respectively.
Note. Conversion (%) =
ni Mi × 100/MCO and selectivity = ni Mi /
ni Mi where ni
is the number of carbon atoms in product i, M is the mole percent of product i
i
measured, and MCO is the mole percent of carbon monoxide in the feed.

14
M.A. Haider et al. / Journal of Catalysis 261 (2009) 9–16
Table 7
Effects of temperature and pressure on the reactions of syngas over a 2 wt% Rh–
.5 wt% Fe/TiO2 catalyst.
2
Condition
1
2
3
4
5
T (K)
P (atm)
527
20
20
543
20
20
567
20
20
543
14
20
543
28
20
3
−1
)
H2 + CO (cm min
H2/CO
1:1
1:1
1:1
1:1
1:1
Conversion (%)
4.55
9.3
17.7
6.7
8.8
Selectivities (%)
CH4
C2H6
30.5
2.3
37.9
2.7
46.7
3.8
36.2
2.4
37.5
2.6
C3H8
3.8
4.3
5.8
3.8
4.0
C4H10
0.5
1.1
0.9
1.0
1.0
CH3CHO
CH3OH
C2H5OH
Ethyl acetate
CO2
12.6
4.0
34.6
11.6
0.0
11.2
2.8
31.0
9.0
10.9
0.0
23.7
8.2
10.1
3.3
35.7
7.5
11.1
2.8
30.4
10.6
0.0
Fig. 4. Deactivation of 2 wt% Rh/TiO2 and 2 wt% Rh–2.5 wt% Fe/TiO2 during CO hy-
drogenation at 543 K and 20 atm.
0.0
0.0
0.0
ꢀ
ꢀ
Note. Conversion (%) =
ni Mi × 100/MCO and selectivity = ni Mi /
ni Mi where ni
is the number of carbon atoms in product i, Mi is the mole percent of product i
measured, and MCO is the mole percent of carbon monoxide in the feed.
Table 8
Influence of H2:CO ratio on the performance of a 2 wt% Rh–1 wt% Fe/TiO2 catalyst.
Conditions
1
2
3
T (K)
P (atm)
543
20
20
543
20
20
543
20
20
3
−1
)
H2 + CO (cm min
H2/CO
1:1
2:1
3:1
Conversion (%)
8.23
21.4
33.9
Selectivities (%)
CH4
C2H6
C3H8
C4H10
CH3CHO
CH3OH
C2H5OH
Ethyl acetate
CO2
43.0
3.2
9.7
3.5
13.5
0.0
15.2
11.9
0.0
51.4
4.9
9.5
1.6
7.2
1.9
16.2
7.3
0.0
58.3
4.7
7.3
0.9
5.4
2.0
15.4
6.0
0.0
Fig. 5. Selectivity to methane and ethanol over 2 wt% Rh/TiO2 during the deactiva-
tion observed in Fig. 4.
ꢀ
ꢀ
Note. Conversion (%) =
ni Mi × 100/MCO and selectivity = ni Mi /
ni Mi where ni
is the number of carbon atoms in product i, Mi is the mole percent of product i
measured, and MCO is the mole percent of carbon monoxide in the feed.
Although the conversion of CO increased with temperature,
the selectivity to undesired methane increased at the expense of
ethanol and methanol. The influence of pressure on both cata-
lyst activity and selectivity was fairly insignificant compared to the
influence of temperature. The results suggest that the hydrogena-
tion of the (CHx)ad species becomes dominant at high temperature
[
48] and more generally, the activation energy for the formation of
Fig. 6. Selectivity to methane and ethanol over 2 wt% Rh–2.5 wt% Fe/TiO2 during
the deactivation observed in Fig. 4.
methane appears to be higher than that for ethanol synthesis [48].
The effects of temperature, pressure, and H2:CO feed ratio on
the performance of a 6% Rh–1.5% Mn/SiO2 catalyst for CO hydro-
genation in a microchannel reactor were also investigated by Hu
et al. [48]. As expected, the CO conversion increased quite signifi-
cantly with temperature over the range of 538 K to 573 K. As seen
in our study, the selectivity to methane also increased with tem-
perature at the expense of selectivity to ethanol. In general, the
mild influence of pressure and the increase in methane production
with increasing temperature at the expense of ethanol are consis-
tent with the results for the Fe-promoted Rh/TiO2 system reported
here.
increasing H2 in the feed stream, the selectivity to methane also
increased at the expense of the oxygenates, namely, acetaldehyde
and ethyl acetate.
It should be noted that CO2 was never observed as a reaction
product. Evidently, the extent of the water–gas shift reaction was
negligible under our reaction conditions.
3.5. Stability of unpromoted and Fe-promoted Rh/TiO2 catalysts
Table 8 summarizes the influence of H2:CO ratio over the range
of 1 to 3 on the conversion and selectivity of syngas conversion
reactions over 2 wt% Rh–1 wt% Fe/TiO2 at 543 K and 20 atm total
pressure. Although the CO conversion increased substantially with
Figs. 4, 5, and 6 illustrate how the CO conversion and prod-
uct selectivities change with time over an 80-h time interval. For
both the unpromoted and promoted Rh/TiO2, the catalysts deac-
tivated between 30 and 50% of their initial activity over the first

M.A. Haider et al. / Journal of Catalysis 261 (2009) 9–16
15
Fig. 8. DRIFTS spectra of adsorbed CO on a 2% Rh/TiO2: (a) after purging in Ar at
room temperature for 60 min, (b) after heating in H2 to 373 K, (c) after heating in
H2 to 448 K, (d) after heating in H2 to 548 K.
Fig. 7. DRIFTS spectra of adsorbed CO on a 2% Rh/TiO2: (a) after purging in Ar at
room temperature for 60 min, (b) after heating in Ar to 373 K, (c) after heating in
Ar to 448 K, (d) after heating in Ar to 548 K.
30 h, but were fairly stable thereafter. The selectivity to undesir-
able methane and desirable ethanol were not significantly affected
by deactivation.
3
.6. Diffuse reflectance infrared Fourier transform spectroscopy
(DRIFTS) studies
In an attempt to relate the structure of the catalyst to its func-
tion in CO hydrogenation, we probed the surfaces of an unpro-
moted and a promoted Rh/TiO2 catalyst by IR spectroscopy of ad-
sorbed CO. Adsorption of CO on supported Rh at room temperature
occurs in three distinct modes, gem-dicarbonyl with characteristic
−
1
−1
and ca. 2030 cm , linear CO
absorption bands at ca. 2100 cm
with absorption band at ca. 2070 cm , and bridged CO with a
−1
−1
broad absorption band at ca. 1840 cm . These assignments were
first made in a pioneering work by Yang and Garland [49] and
later confirmed by several studies [42,50–54]. In addition to these
three modes, additional bands for linear CO on oxidized rhodium
Fig. 9. DRIFTS spectra of adsorbed CO on a 2% Rh–2.5% Fe/TiO2: (a) after purging
in Ar at room temperature for 60 min, (b) after heating in Ar to 373 K, (c) after
heating in Ar to 448 K, (d) after heating in Ar to 548 K.
2
+
−1
3+ −1
at 2145 cm ), were reported by
(
Rh
at 2135 cm
and Rh
Trautmann and Baerns [54]. It is now well-accepted that the gem-
+
dicarbonyl species arise from singly-charged Rh species, while
linear and bridged CO arise from Rh clusters.
0
The adsorption of CO at room temperature can lead to an ox-
0
+
idative disruption of Rh clusters into Rh sites. Primet [55] sub-
scribes to the view that CO adsorption on small particles of Rh
(<1 nm) is dissociative at room temperature. The chemisorbed
oxygen (from CO dissociation) presumably leads to the forma-
+
tion of Rh sites, which then forms the gem-dicarbonyl species
upon further adsorption of CO. However, this suggestion was coun-
tered by Solymosi and Pasztor [51,56] who found no evidence of
CO dissociation on supported Rh particles at temperatures less
+
than 473 K. They proposed instead that the Rh is most proba-
0
bly formed via an oxidation of the Rh clusters by an OH group of
the support.
Fig. 10. DRIFTS spectra of adsorbed CO on a 2% Rh–2.5% Fe/TiO2: (a) after purging
in Ar at room temperature for 60 min, (b) after heating in H2 to 373 K, (c) after
heating in H2 to 448 K, (d) after heating in H2 to 548 K.
Figs. 7 and 8 show the IR spectra of CO adsorbed on 2 wt%
Rh/TiO2 as a function of temperature in the absence (Fig. 7) and
the presence (Fig. 8) of hydrogen. Major spectral changes were ob-
served upon heating in Ar (Fig. 7). The gem dicarbonyl species
was stable after heating to 373 K, as illustrated by the strong
the gem-dicarbonyl band was strongly attenuated by heating to
only 373 K and second, no features attributable to adsorbed CO
were observed at 548 K. The accelerated conversion of the gem-
dicarbonyl by dihydrogen agrees with the previous work [53] on a
0.5% Rh/TiO2.
Figs. 9 and 10 present the IR spectra from similar thermal de-
sorption experiments on 2 wt% Rh–2.5 wt% Fe/TiO2. Although the
spectral features are generally quite similar to the unpromoted sys-
−1
band at 2098 cm . At 448 K, the gem-dicarbonyl band was sig-
nificantly attenuated while the intensity of the linear CO band at
−
1
2060 cm
increased. At 548 K, the band due to linear CO shifted
−
1
to 2040 cm , which has been seen previously [54]. Although the
general behavior of the spectra for desorption of CO in dihydro-
gen was the same (Fig. 8), there were two major differences. First,

16
M.A. Haider et al. / Journal of Catalysis 261 (2009) 9–16
tem, there are some differences that warrant discussion. First, the
gem-dicarbonyl species was more stable on the Fe-promoted cat-
[11] J. Iranmahboob, H. Toghiani, D.O. Hill, Appl. Catal. A 247 (2003) 207.
12] G.-Z. Bian, L. Fan, Y.-L. Fu, K. Fujimoto, Appl. Catal. A 170 (1998) 255.
[13] J.J. Spivey, A. Egbebi, Chem. Soc. Rev. 36 (2007) 1514.
[
−1
alyst since its spectral signature at ca. 2094 cm
was prominent
[14] V. Subramani, S.K. Gangwal, Energy Fuels 22 (2008) 814.
[15] R. Burch, M.I. Petch, Appl. Catal. A 88 (1992) 39.
[16] R. Burch, M.I. Petch, Appl. Catal. A 88 (1992) 77.
at 448 K in both Ar (Fig. 9) and H2 (Fig. 10). Second, the desorp-
tion spectrum in presence of dihydrogen (Fig. 10) shows a band at
−1
2039 cm
due to linear CO at the highest desorption tempera-
[17] R. Burch, M.J. Hayes, J. Catal. 165 (1997) 249.
ture of 548 K. This contrasts the results from a 2% Rh/TiO2 catalyst
where no surface carbonyl was observed at 548 K in H2 (Fig. 8).
Taken together, these observations suggest that the Fe promoter
stabilizes the adsorbed CO on both the charged and neutral Rh
moieties.
[18] C. Mazzocchia, P. Gronchi, A. Kaddouri, E. Tempesti, L. Zanderighi, A. Kienne-
mann, J. Mol. Catal. A: Chem. 165 (2001) 219.
[19] P. Gronchi, E. Tempesti, C. Mazzocchia, Appl. Catal. A 120 (1994) 115.
[
20] M. Ojeda, M.L. Granados, S. Rojas, P. Terreros, F.J. Garcia-Garcia, J.L.G. Fierro,
Appl. Catal. A 261 (2004) 47.
[21] H. Ma, Z. Yuan, Y. Wang, X. Bao, Surf. Interface Anal. 32 (2001) 224.
[
[
[
[
[
22] M. Bowker, Catal. Today 15 (1992) 77.
23] S.S.C. Chuang, R.W. Stevens Jr., R. Khatri, Top. Catal. 32 (2005) 225.
24] Y. Wang, J. Li, W. Mi, React. Kinet. Catal. Lett. 76 (2002) 141.
25] A. Frydman, D.G. Castner, C.T. Campbell, M. Schmal, J. Catal. 188 (1999) 1.
26] A. Trunschke, H. Ewald, H. Meissner, S. Marengo, S. Martinengo, F. Pinna, L. Zan-
derighi, J. Mol. Catal. 74 (1992) 365.
4
. Conclusions
The direct synthesis of ethanol and other oxygenates from CO
and H2 over Rh catalysts depended on the nature of the support
and the presence of the Fe promoter. Silica-supported Rh was in-
effective at converting CO under the standard conditions used here
while titania-supported Rh produced some ethanol and other oxy-
genates, with methane being the major undesirable hydrocarbon
product. Promotion of Rh/silica and Rh/titania with Fe substan-
tially improved the activity of the catalysts and their selectivity
to ethanol and other oxygenates. Results from electron microscopy
and H2 chemisorption confirmed the direct interaction of Fe with
Rh. Results from thermal desorption of CO from unpromoted and
promoted Rh/TiO2 indicated that Fe stabilized the gem-dicarbonyl
formed upon CO adsorption on Rh and retained CO on the Rh at
elevated temperatures.
[27] T. Beutel, O.S. Alekseev, Y.A. Ryndin, V.A. Likholobov, H. Knozinger, J. Catal. 169
(1997) 132.
[28] H.Y. Luo, W. Zhang, H.W. Zhou, S.Y. Huang, P.Z. Lin, Y.J. Ding, L.W. Lin, Appl.
Catal. A 214 (2001) 161.
[
29] E. Guglielminotti, E. Giamello, F. Pinna, G. Strukul, S. Martinengo, L. Zanderighi,
J. Catal. 146 (1994) 422.
[30] Y. Wang, H. Luo, D. Liang, X. Bao, J. Catal. 196 (2000) 46.
[31] A.S. Lisitsyn, S.A. Stevenson, H. Knozinger, J. Mol. Catal. 63 (1990) 201.
[
[
32] H. Trevino, G.-D. Lei, W.M.H. Sachtler, J. Catal. 154 (1995) 245.
33] P.-Z. Lin, D.-B. Liang, H.-Y. Luo, C.-H. Xu, H.-W. Zhou, S.-Y. Huang, L.-W. Lin,
Appl. Catal. A 131 (1995) 207.
[34] M. Ichikawa, Bull. Chem. Soc. Jpn. 51 (1978) 2273.
[
35] A. Fukuoka, T. Kimura, M. Ichikawa, J. Chem. Soc. Chem. Commun. (1988) 428.
[36] H. Arakawa, K. Takeuchi, T. Matsuzaki, Y. Sugi, Chem. Lett. 13 (1984) 1607.
[
37] H. Arakawa, T. Fukushima, M. Ichikawa, S. Natsushita, K. Takeuchi, T. Matsuzaki,
Y. Sugi, Chem. Lett. 14 (1985) 881.
Acknowledgments
[
[
[
38] H.F.J. Van’t Blik, J.C. Vis, T. Huizinga, R. Prins, Appl. Catal. 19 (1985) 405.
39] B.J. Kip, F.W.A. Dirne, J. van Grondelle, R. Prins, Appl. Catal. 25 (1986) 43.
40] D.C. Koningsberger, C.P.J.H. Borgmans, A.M.J. van Elderen, B.J. Kip, J.W. Nie-
mantsverdriet, J. Chem. Soc. Chem. Commun. (1987) 892.
[41] J.W. Niemantsverdriet, S.P.A. Louwers, J. van Grondelle, A.M. van der Kraan,
F.W.H. Kampers, D.C. Koningsberger, in: Proc. 9th Int. Congr. Catal., Calgary, Al-
berta, vol. 2, 1988, p. 674.
The authors acknowledge the seed funding from University of
Virginia and partial support from the Chemical Sciences, Geo-
sciences and Biosciences Division, Office of Basic Energy Sciences,
Office of Science, U.S. Department of Energy, Grant No. DE-FG02-
95ER14549. The authors also acknowledge Dr. Mitsu Murayama for
[
42] J.R. Katzer, A.W. Sleight, P. Gajardo, J.B. Michel, E.F. Gleason, S. McMillan, Fara-
day Discuss. Chem. Soc. 72 (1981) 121.
acquisition of electron micrographs and analysis by EDS.
[
[
43] M.M. Bhasin, W.J. Bartley, P.C. Ellgen, T.P. Wilson, J. Catal. 54 (1978) 120.
44] E. Guglielminotti, F. Pinna, M. Rigoni, G. Strukul, L. Zanderighi, J. Mol. Catal.
A 103 (1995) 105.
References
[
45] T. Hanaoka, H. Arakawa, T. Matsuzaki, Y. Sugi, K. Kanno, Y. Abe, Catal. Today 58
[
1] X.Q. Qiu, N. Tsubaki, K. Fujimoto, Q.M. Zhu, Fuel Process. Technol. 85 (2004)
193.
2] T. Matsuzaki, K. Takeuchi, T. Hanaoka, H. Arawaka, Y. Sugi, Appl. Catal. A 105
1993) 159.
(
2000) 271.
46] S.J. Tauster, S.C. Fung, R.L. Garten, J. Am. Chem. Soc. 100 (1978) 170.
47] A.D. Logan, E.J. Braunschweig, A.K. Datye, D.J. Smith, Langmuir 4 (1988) 827.
48] J. Hu, Y. Wang, C. Cao, D.C. Elliott, D.J. Stevens, J.F. White, Catal. Today 120
2007) 90.
1
[
[
[
[
(
[
[
3] E.B. Pereira, G.-A. Martin, Appl. Catal. A 103 (1993) 291.
4] L. Majocchi, L. Lietti, A. Beretta, P. Forzatti, E. Micheli, L. Tagliabue, Appl. Catal.
A 166 (1998) 393.
(
[49] A.C. Yang, C.W. Garland, J. Phys. Chem. 61 (1957) 1504.
[
50] J.T. Yates, T.M. Duncan, S.D. Worley, R.W. Vaughan, J. Chem. Phys. 70 (1979)
[
[
[
5] M.J.L. Gines, E. Iglesia, J. Catal. 176 (1998) 155.
6] M. Xu, E. Iglesia, J. Catal. 188 (1999) 125.
7] A.B. Stiles, F. Chen, J.B. Harrison, X. Hu, D.A. Storm, H.X. Yang, Ind. Eng. Chem.
Res. 30 (1991) 811.
1219.
[51] F. Solymosi, M. Pasztor, J. Phys. Chem. 89 (1985) 4789.
[52] P. Gibson, P.C. van Berge, S. Afr. J. Chem. 44 (1991) 1.
[53] Z.L. Zhang, A. Kladi, X.E. Verykios, J. Mol. Catal. 89 (1994) 229.
[54] S. Trautmann, M. Baerns, J. Catal. 150 (1994) 335.
[
[
8] D.J. Elliott, F. Pennella, J. Catal. 114 (1988) 90.
9] D. He, Y. Ding, H. Luo, C. Li, J. Mol. Catal. A: Chem. 208 (2004) 267.
[
10] J. Iranmahboob, H. Toghiani, D.O. Hill, F. Nadim, Fuel Process. Technol. 79
[55] M. Primet, J. Chem. Soc. Faraday Trans. 1 74 (1978) 2570.
[56] F. Solymosi, M. Pasztor, J. Phys. Chem. 90 (1986) 5312.
(2002) 71.