Selective reduction of acetic acid to ethanol over novel Cu 2In/Al2O3 catalyst

Article abstract of DOI:10.1016/j.apcata.2013.05.042

Volatile fatty acids (VFAs) can be produced efficiently by simple thermochemical or biological biomass degradation. For the processing of these organic acids in hydrogen atmosphere, the consecutive reactions of acetic acid (AA) hydroconversion were studied in details looking for conditions of selective ethanol production over a novel and advantageous bimetallic composite applying indium as co-catalyst. The reactions were investigated in vapor phase at 240-380°C, 7-21 bar hydrogen and 0.5-3.5 bar acetic acid partial pressures in a fixed bed flow-through reactor using supported copper catalysts. In ₂O₃ admission can significantly increase AA hydroconversion activity of copper catalysts supported on various oxides and the yield of the produced ethanol. Efficient hydrogenating catalysts, containing finely dispersed metal particles were obtained by in situ reduction with H ₂ at 450 °C. In the catalysts modified with In₂O ₃ additive, formation of an intermetallic compound (Cu₂In) was strikingly observed resulting in a different, more advantageous catalytic behavior as of pure copper particles supported on different oxide supports. On comparing a commercial, conventionally used catalysts (Adkins: 72 wt% CuCr ₂O₄ + 28 wt% CuO) with the bimetallic alumina supported composite (Cu₂In/Al₂O₃) the new catalyst proved to be much more active and selective for producing ethanol. A schematic representation of reactions involved in the hydroconversion of acetic acid was explored and verified. The activity dependence on the reactant partial pressures denotes rate-controlling surface reaction according to Langmuir-Hinshelwood mechanism.

Full text of DOI:10.1016/j.apcata.2013.05.042

Applied Catalysis A: General 464–465 (2013) 313–321
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Selective reduction of acetic acid to ethanol over novel Cu2In/Al2O3
catalyst
a
b
György Onyestyáka, Szabolcs Harnosa,∗, Szilvia Klébert , Magdalena Stolcová ,
ˇ
Alexander Kaszonyib, Dénes Kallóa
a Institute of Materials and Environmental Chemistry, Research Centre for Natural Sciences, Hungarian Academy of Sciences, Pusztaszeri u. 59–67,
Budapest H-1025, Hungary
b Department of Organic Technology, Slovak University of Technology, Radlinského 9, Bratislava SK-81237, Slovak Republic
a r t i c l e i n f o
a b s t r a c t
Article history:
Volatile fatty acids (VFAs) can be produced efficiently by simple thermochemical or biological biomass
degradation. For the processing of these organic acids in hydrogen atmosphere, the consecutive reactions
of acetic acid (AA) hydroconversion were studied in details looking for conditions of selective ethanol
production over a novel and advantageous bimetallic composite applying indium as co-catalyst. The
reactions were investigated in vapor phase at 240–380 ◦C, 7–21 bar hydrogen and 0.5–3.5 bar acetic acid
partial pressures in a fixed bed flow-through reactor using supported copper catalysts. In2O3 admission
can significantly increase AA hydroconversion activity of copper catalysts supported on various oxides
and the yield of the produced ethanol. Efficient hydrogenating catalysts, containing finely dispersed
metal particles were obtained by in situ reduction with H2 at 450 ◦C. In the catalysts modified with
In2O3 additive, formation of an intermetallic compound (Cu2In) was strikingly observed resulting in a
different, more advantageous catalytic behavior as of pure copper particles supported on different oxide
supports. On comparing a commercial, conventionally used catalysts (Adkins: 72 wt% CuCr2O4 + 28 wt%
CuO) with the bimetallic alumina supported composite (Cu2In/Al2O3) the new catalyst proved to be much
more active and selective for producing ethanol. A schematic representation of reactions involved in the
hydroconversion of acetic acid was explored and verified. The activity dependence on the reactant partial
pressures denotes rate-controlling surface reaction according to Langmuir–Hinshelwood mechanism.
© 2013 Elsevier B.V. All rights reserved.
Received 4 March 2013
Received in revised form 30 April 2013
Accepted 31 May 2013
Available online xxx
Keywords:
Acetic acid reduction
Copper catalyst
Indium co-catalyst
Ethanol
Ethyl acetate
1. Introduction
economy of the biomass-to-fuel conversion by preserving the pos-
sible highest fraction of C–C and C–H bonds in constituents of the
A major 21st century goal of governments, academia and indus-
try is the efficient and economic utilization of biomass resources
for production of fuels and chemicals [1–6]. The key trend today
is to move away from food crops to nonfood ones. For this tran-
sition, it is vital that low-cost processing technologies should be
developed for conversion of low-cost biomass. The future success
of “biorefineries” will require fundamental understanding of the
types of processes best suited for converting the various biomass-
derived constituents into chemical moieties. Hence, the evolution
of a broad range of conversion technologies including various cat-
alytic processes is needed.
organic biomass. Instead of the very simple, but seriously disad-
vantageous thermochemical routes (as pyrolysis), a novel way, a
favorable biochemical destruction (branching from the biogas pro-
cess: i.e., stopped before the slow methanogenesis step proceeding
the rapid anaerobic acidogenic digestion stage) is suggested uti-
lizing volatile fatty acids (VFAs) for production of biofuels and
biochemicals [7–9]. It should be remarked that species of anaerobic
bacteria, including members of the genus Clostridium, can convert
directly sugar building units to AA from less expensive feeds, with-
out converting to conventional ethanol as an intermediate and the
formation of which is accompanied by CO2 evolution. The overall
chemical reaction conducted by these bacteria may be represented
as:
Direct and indirect methods are accessible for the produc-
tion of bio-fuel from biomass. The indirect methods are based on
Fischer–Tropsch (FT) technology. The direct methods improve the
C6H12O6 → 3 CH3COOH
∗
The microorganisms anaerobically digest the biomass and con-
vert it into mixture of volatile (short-chain) carboxylic acids (VFAs).
Corresponding author. Tel.: +3614381100.
E-mail address: harnos.szabolcs@ttk.mta.hu (S. Harnos).
0926-860X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2013.05.042

314
G. Onyestyák et al. / Applied Catalysis A: General 464–465 (2013) 313–321
Fig. 1. A schematic figure of the high-pressure fixed bed flow reactor system.
The carbon atom of carboxylic group is in a relatively high oxida-
tion state. Removal of oxygen atoms by catalytic hydrogenation is
extensively applied in industry. Catalysts like copper or nickel are
promoted e.g., by chromium or molybdenum. Recently, we discov-
ered novel highly efficient supported bimetallic catalysts, which
are more active and selective in the hydrodeoxygenation (HDO)
of octanoic acid to octanol at moderate pressure and tempera-
ture than conventional contacts [10–13]. Proper catalysts contain
indium as co-catalyst and another metal of high hydrogenation
activity, such as, copper or nickel, on various supports. Appearance
of Cu2In or Ni2In intermetallic compounds as detectable phases
result in significant increase of desired hydrogenation activity, i.e.,
the stepwise hydrogenation of carboxylic acids is stopped at alco-
hol formation and the hydrodecarbonylation of carboxylic acids
(loss of one carbon atom) is inhibited. Moreover, in presence of
mobile indium atoms or of indium containing nano clusters, mono-
or bimolecular alcohol dehydration capability of the support is sup-
pressed.
tested however, the catalytic properties were found to be controlled
principally by modification of the metallic phases.
Recently, the highly active Ni2In/Al2O3 composite catalyst has
been already tested on comparing the hydroconversion of octanoic
(C8) and acetic (C2) acid under similar conditions. The difference of
the length of alkyl-chains resulted in significantly different prod-
uct distributions [21]. The reduction of NiO formed from supported
nickel salt can be observed near below 450 ◦C, at the routine tem-
perature of pretreatment and became complete only far above
450 ◦C depending on the loading [19]. This means that NiO and
active metallic phases are present and their partition may influence
the catalytic properties. Therefore, a detailed study seemed to be
promising with the less active Cu/Al2O3 catalysts, which are com-
pletely reduced at the routine temperature of pretreatment being
near to the reaction temperature. Because of practical reasons, AA
hydroconversion to ethanol was studied over catalysts containing
copper in metallic form.
The selective hydrogenation of short chain carboxylic acids to
alcohol in a continuous flow system working in vapor phase under
mild conditions seems to be still a problematic step. Only a few
studies of such reactions have been carried out. The patent liter-
ature describes numerous catalysts for acetic acid hydrogenation
comprising of one or more noble metals of Group VIII dispersed
on Group III or IV metal oxides. Rachmady and Vannice published
at the beginning of this century a series of papers [14–18] related
to platinum catalysts supported on TiO2, SiO2, Al2O3 and Fe2O3
and the results are compared with that obtained without support
[14]. Acetic acid was the model to test the carboxylic acid adsorp-
tion and its conversion kinetics because of molecular simplicity and
wide range commercial application. Decarbonylation and decom-
position reactions dominated on Pt powder, Pt/SiO2 and Pt/Al2O3
similarly to Ni/Al2O3 catalysts shown in our recent studies [19,20].
Pt/Fe2O3 working at much higher temperature has particularly high
selectivity for formation of acetaldehyde, i.e., to the less reduced
intermediate. Pt/TiO2 catalysts were the most active and selective
in acetic acid hydrogenation to ethanol [14]. It has been strongly
evidenced that both metal (Pt) and oxide support are involved in
conversions. Thus, the interaction of acetic acid with the oxide sup-
port plays a major role in determining the hydroconversion kinetics
and the principal role of Pt in this case is to provide mobile, acti-
vated hydrogen atoms. In present work, various supports are also
2. Experimental
The catalytic hydrogenation of AA (96%, from Reanal) (model
compound with the shortest aliphatic chain) was studied in a
high-pressure fixed bed flow-through reactor at 21 bar total pres-
sure, 240–380 ◦C, 7–21 bar hydrogen and 0.5–3.5 bar acetic acid
partial pressures, i.e., under similar conditions as in Refs. [10,21].
The Fischer esterification of AA and ethanol (absolute, from Molar
Chemical) and ethylacetate (>99%, from Merck) hydrogenation
were also studied over the investigated catalyst samples to follow
the secondary reactions. The experimental setup is illustrated in
Fig. 1. In general, the catalysts were pretreated in hydrogen flow
in situ in the reactor at 450 ◦C and 21 bar for 1 h in order to obtain
active metallic surfaces for carboxylic acid reduction. The reaction
was allowed to run one hour at each condition to attain steady
state. Before sampling, the effluent during the second hour was
collected, depressurized and cooled to room temperature. The liq-
uid product mixture at ambient conditions was analyzed by gas
chromatograph (Shimadzu 2010) for AA products equipped with a
Restek Rt-U-BOND capillary column, and a flame ionization detec-
tor. The gaseous reactor effluent was analyzed for CO2, CO, CH4 and
light hydrocarbons using an on-line gas chromatograph (HP-5890)
having Carboxen 1006 PLOT capillary column and thermal conduc-
tivity detector (TCD). The results are represented as stacked area

G. Onyestyák et al. / Applied Catalysis A: General 464–465 (2013) 313–321
315
Table 1
Specific surface areas of the supports and the particle diameters of copper formed on the catalysts after reductive pretreatment.
Al2O3 (Akzo)
199
TiO2 (Degussa)
TiO2 (Aldrich)
SiO2 (Cabot)
200
SiO2 (Grace)
291
SiO2 (Sylobead)
BET (m2/g)
ØCu0 (nm)
55
43
12
123
563
165
20
34
30
graphs where the distance between two neighboring curves repre-
sents the concentration of a given product at a given temperature
in mole percent.
mean crystallite size of the copper particles was estimated by
the Scherrer equation. Nitrogen physisorption measurements
were carried out at −196 ◦C using Quantochrome Autosorb 1C
instrument for determining the specific surface area. Specific
surface area of the supports and particle diameters of copper
formed after reductive pretreatment are collected in Table 1.
Very similar surface and diameter values are obtained after load-
ing.
Activated ␥-Al2O3 (Ketjen CK 300, Akzo-Chemie, BET:
199 m2/g), SiO2 (CAB-O-SIL M-5, Cabot Corp., BET: 200 m2/g;
Grace, BET: 291 m2/g; Grace Sylobead, BET: 563 m2/g) and TiO2
(Aeroxide P25, Degussa, BET: 55 m2/g (anatase); Aldrich, BET:
12 m2/g (rutile)), were impregnated with NH4OH solution (Reanal)
of Cu (acetate)2 (Aldrich), dried, and calcined at 550 ◦C in air
stream. The sample designation, e.g., 9 Cu/Al2O3 formula repre-
sents a catalyst preparation containing 9 wt% Cu. For comparison,
an Adkins catalyst (consisting of 72 wt% CuCr2O4 and 28 wt%
CuO) was applied. Composite catalysts were prepared by adding
indium(III) oxide (Aldrich) to the samples in various amounts and
grinding the mixture in agate mortar.
3. Results and discussion
Ni2In/Al2O3 catalyst was proven more active in both octanoic
[10,11] and acetic acid hydroconversion [21] than Cu2In/Al2O3
since the conversion and alcohol yield curves are shifted by
20–30 ◦C to higher temperatures for Cu2In/Al2O3. The product
distributions are similar with some small, however, not neg-
ligible differences for the two bimetallic catalysts. Aside from
X-ray patterns were recorded by a Philips PW 1810/1870
diffractometer with monochromatic Cu K␣ radiation at elevated
temperatures using a high-temperature XRD cell (HT–XRD). The
Fig. 2. XRD patterns of 9 wt% Cu on (a) Al2O3 (Akzo-Chemie), (b) TiO2 (Degussa), (c) TiO2 (Aldrich), (d) SiO2 (Cabot) and (e) SiO2 (Grace) supports doped with 10 wt% of In2O3.
In the case of (f) SiO2 (Grace-Sylobead) InCl3 is applied containing the same amount of indium. The diffractograms were recorded at the indicated temperatures after 30 min
treatment at each temperature in H2 flow.

316
G. Onyestyák et al. / Applied Catalysis A: General 464–465 (2013) 313–321
Fig. 3. AA hydroconversion over 9 wt% Cu on (a) Al2O3 (Akzo-Chemie), (b) TiO2 (Degussa), (c) TiO2 (Aldrich), (d) SiO2 (Cabot), (e) SiO2 (Grace) and (f) SiO2 (Grace-Sylobead)
catalysts characterized by the distribution of the main products between 240 and 380 ◦C at 21 bar total pressure, represented by stacked area graphs. Each sample was
admixed with 10 wt% of In2O3 before the reductive pretreatment with exception (f) SiO2 (Grace-Sylobead) applied InCl3 containing the same amount of indium. The WHSV
of AA was 1.0 h−1. Partial pressures of AA and H2 were 2.1 and 18.9 bar.
the different hydrogenation activity of the main metals (Ni or
Cu) similarity of catalytically active alloy phases seems probable
(Ni2In and Cu2In).
In2O3 cannot be reduced at all. Exceptionally indium admission was
also carried out with InCl3 impregnation into the 9Cu/SiO2 (Grace
Sylobead), because the mixture of the oxides composing the cata-
lyst bed cannot be pressed to pellets without binder. Although the
presence of InCl3 phase being finely distributed cannot be detected
in Fig. 2f, the reduction to metallic indium takes place under similar
conditions as for In2O3 resulting in similar nice signals for forma-
tion the Cu2In phase as in Fig. 2d or e.
The catalytic activity and selectivity of the various supported
catalyst preparations are characterized by the product distribu-
tion as a function of reaction temperature in stacked area graphs
(see Fig. 3). Main valuable products formed from acetic acid are
acetaldehyde, ethanol and ethyl acetate. As by-products, acetone,
carbon dioxide (only when the samples do not contain copper) and
water (over each catalyst) were determined as well. All the supports
are of low activity for undesired catalytic reactions (e.g., Fig. 4a).
Reduction of In2O3 results in liquid indium metal on the surface
of support the catalytic activity of which is essentially lower and
the selectivity poorer than for the composite catalyst containing
both indium and copper (Fig. 4b and c). The activity and selectivity
of supported Cu can be improved by In2O3 addition (Fig. 5). After
reduction, the appearance of Cu2In phase is detectable beside cop-
per and excess indium. It seems thus likely that this new phase
should probably be responsible for the drastic changes of catalytic
properties. Similar effects of indium addition could be observed for
supported nickel, i.e., the activity for C–O and C=O hydrogenation
X-ray diffraction patterns evidence that in the novel indium con-
taining composite catalysts, activated at 450 ◦C in H2 flow before
the reaction, In2O3 is reduced (excepted when the support is
Aldrich TiO2 shown in Fig. 2c) and liquid indium metal should be
present under the reaction conditions being the reduction temper-
ature far above the melting point of indium (156.4 ◦C). Reduced
copper reacts immediately with indium and Cu2In particles are
formed as shown in Fig. 2. Cu nanoparticles appear on various sup-
ports when copper oxide is fully reduced up to 350 ◦C, and the small
copper particles do not change up to 650 ◦C without indium dop-
ing (not shown here). Reduction of admixed In2O3 (or InCl3) can
be observed at much higher temperature, i.e., between 350 and
450 ◦C (see Fig. 2) detected by disappearance of In2O3 diffraction
lines (+) and appearance of the Cu2In phase in the diffractogramms.
This transformation is completed up to 450 ◦C. Mainly Cu2In phase,
but Cu4In and Cu9In4 phases may also be present, however, in
much lower concentration. These phases were identified on ICDD
database [22], the corresponding ICDD numbers are Cu2In: 42-
1475, Cu4In: 42-1477 and Cu9In4: 42-1476. 10 wt % In2O3 admitted
to 9 wt % copper-containing samples corresponds to the Cu2In sto-
chiometric composition characteristic for the new intermetallic
compound detected on all the samples excepted when a low surface
area support as Aldrich TiO2 was applied (Fig. 2c). On this support,

G. Onyestyák et al. / Applied Catalysis A: General 464–465 (2013) 313–321
317
penetrate into the micropores to form metal nanoparticles dur-
ing reduction. Therefore, copper atoms form big metal particles
(165 nm) with very low active surface, which is reflected in
the extremely low activity. Near this disadvantageous situation,
the exceptional application of different indium source (InCl3)
further handicaps the catalyst creating a much lower activity
than with the pure, monometallic Cu-catalyst. Although reduc-
tion of InCl3 and the formation of Cu2In particles seem to be
undisturbed below the sublimation temperature of the salt (see
Fig. 2f) something is wrong from catalytic point of view. To
prove this effect, another but highly active Cu loaded silica
gel, Grace (BET: 291 m2/g) was tested without In2O3 as indium
source using InCl3. Instead of the nice hydroconversion shown
in Fig. 3e a similar low activity was observed than we can
see in Fig. 3f. Instead of InCl3 using the likewise water solu-
ble In(NO3) completely same results can be obtained than with
In2O3.
Regarding the comparison of the applied support of appropri-
ate morphology in Fig. 3a, b, d, e, the chemical composition of the
supports do not seem to be important, because the properties of
the supported bimetallic particles are dominant in the acetic acid
hydroconversion. Consequently the further studies can be carried
out using only the activated ␥-Al2O3 (Ketjen CK 300, Akzo-Chemie,
BET: 199 m2/g), commercial support.
The metal free alumina is able to ketonize acetic acid at higher
reaction temperature (Fig. 4a), however, this capability can be
partly suppressed by indium admission (Fig. 4b) and completely
eliminated by loading copper (see Fig. 4c). In hydrogen, copper on
alumina enables attack only of the C–O bond (see Fig. 4) contrary to
nickel [10,21] where the hydrogenolysis of C–C bond is dominant
resulting in the hydrodecarbonylation of carboxylic acids. Indium
doping can block this reaction route during hydrodeoxyganation
over more active hydrogenating metals (e.g., Pt or Ni) resulting in
drastic change of the product distribution, instead of chain shorten-
ing the step by step reduction toward alcohol dominates. Since only
the last reaction route is characteristic for copper catalysts indium
admission changes only the product distribution in this reaction
way and increases the reaction rates (compare Fig. 4c and Fig. 3a.
as well as Fig. 5).
The influences of indium doping can be seen in details in
Fig. 5 and a comparison can be found with a commercial Adkins
catalyst. Cu2In/alumina catalyst is significantly more active than
monometallic alumina supported Cu or chromium containing
industrial catalysts (see Fig. 5a) and the changes of the selectiv-
ity are different. In low temperature range, the high activity of the
composite catalyst is ascertained by the high yield of acetaldehyde
(see Fig. 5b) being the intermediate product of alcohol formation.
At higher reaction temperature, the indium doped sample gives the
highest yield of the most desired product ethanol (see Fig. 5c). On
comparing Figs. 4c and 5d ethyl acetate is formed and is efficiently
hydrogenated to ethanol.
The nascent ethanol reacts easily with the reactant acetic acid
in the Fischer esterification reaction even without catalyst contrary
to longer carboxylic acids (e.g., octanoic acid) [11] (see Fig. 6). The
boiling points of these two main products (ethanol and its ester) are
almost the same thus their separation is rather expensive. Using the
product mixture as fuel the separation is not necessary. In order
to obtain pure ethanol or ethyl acetate the reaction mechanism
is to be explored and reaction conditions optimized accordingly.
The acidic sites of the alumina support can effectively catalyze
ester formation. Ethyl acetate yield is lower when SiO2 is used
as support. However, this kind of advantage of silica is masked in
presence of metals active in ester hydrogenation. The yield of ethyl
acetate is restricted by the presence of hydrogenating metals: sim-
ilarly by indium or copper and more effectively by Cu2In alloy. The
genesis of the ester cannot be prevented and its hydrogenolysis
Fig. 4. AA hydroconversion over (a) Al2O3, (b) Al2O3 + 10% In2O3 and (c) 9Cu/Al2O3
catalysts characterized by the distribution of the main products, represented by
stacked area graphs. Reaction conditions as in Fig. 3.
increased while for undesired hydrogenolysis of C–C bond became
negligible.
The Cu2In active phase on Al2O3 and SiO2 supports of
200–300 m2/g specific surface (Fig. 3a, d, e) demonstrate practi-
cally equal activities and very similar selectivities independently
of the supports in a wide reaction temperature range (280–380 ◦C).
Specific surface area of the supports does not play any role in activ-
ity the Cu2In surface should be decisive. Cu2In on Aldrich TiO2 is
not detectable and on the highly microporous Grace–Sylobead SiO2
can be formed the largest metal crystals with relatively low surface
in the meso- or macro-pores of support (Table 1 (size of the mono-
and bimetallic particles was found to be nearly the same in many
cases)). This conclusion is supported by observations for Cu-zeolites
[23]. On applying Cu-zeolites as catalyst precursors containing cop-
per cations completely dispersed, the reduction results in complete
destruction of zeolite structure, i.e., in a similarly low surface sup-
port (shown in Fig. 3c), however, the active metal appears in high
dispersion and thus of high activity [12,24].
Grace–Sylobead SiO2 support (BET: 563 m2/g) is a highly micro-
porous material of high specific surface and the precursors cannot

318
G. Onyestyák et al. / Applied Catalysis A: General 464–465 (2013) 313–321
Fig. 5. Conversion (a) and the yield of acetaldehyde (b), ethanol (c) and ethyl acetate (d) over Adkins, 9Cu/Al2O3 and 9Cu/Al2O3 + 10 wt% In2O3 catalysts in AA hydroconversion
as a function of reaction temperature. Reaction conditions as in Fig. 3.
can be catalyzed by all the metals applied although with differ-
ent effectiveness. These differences are demonstrated in Fig. 7 for
alumina supported indium, copper, and Cu2In. The ethyl acetate
hydrolysis over Cu2In/alumina producing ethanol and acetic acid
seems to be hardly detectable since acetic acid formed is imme-
diately converted to ethanol. Thereafter diethyl ether and water
are produced by dehydration of ethanol proceeding on alumina
support. Ethanol can form via ester hydrolysis catalyzed by alu-
mina (as shown in Fig. 6 for the reverse reaction) and by multistep
hydrogenolysis of ester. Former way results in AA formation hardly
observable on the composite catalyst (Fig. 7c) which indicates that
hydrogenolysis activity surpasses the acidic activity of the support.
Ethanol formation is higher for supported Cu than for supported
In pointing to higher hydrogenolysis activity of Cu than that of
In. Thus a sequence of increasing activity In < Cu < Cu2In seems
probable.
complexity of the reaction network. On monometallic catalyst,
acetaldehyde can hardly be detected while on bimetallic sam-
ple acetaldehyde seems to be a primary product and ethanol
a secondary one. Acetaldehyde as an intermediate has a maxi-
mum concentration as a function of space time (Fig. 8b). Ester
formation rate increases both with aldehyde concentration and
ethanol concentration hence both Tishchenko and Fischer reac-
tions seem probable (Scheme 1). The kinetics results would be
captured as rate equations in which the Arrhenius parameters
(normalized pre-exponential factor, apparent activation energy)
could be compared with those obtained in other laboratories.
Fig. 9 demonstrates different activities of the investigated alu-
mina supported mono- and bimetallic copper catalysts obtained
from initial slopes of conversion curves similar as shown in
Fig. 8. Formation of the indium containing bimetallic catalytically
active metal phase results in approx. 1.5 times higher appar-
ent activation energy (108 kJ/mol instead of 70 kJ/mol determined
on the monometallic copper surface) reflecting a new quality
of the active alloyed surface resulting in a different reaction
mechanism. A similarly 1.5 times higher pre-exponential factor
reflects a high increase in the reaction rate of the step-by-step
reduction of acetic acid to ethanol. Although in this paper the
apparent activation energy was determined only for the over-
all hydrogenation reaction the values obtained are comparable
with Ref. [14] for the monometallic Pt catalyst supported by tita-
nia.
Fig. 8 shows the conversion curves determined on the mono-
and the bimetallic catalysts at 320 ◦C, which demonstrate the
Hydroconversion of acetic acid to ethanol and ethyl acetate
increases roughly linearly with increasing hydrogen partial pres-
sure over Cu2In/alumina composite catalyst (Fig. 10). Acetaldehyde
selectivity is high at low hydrogen partial pressures (where the
conversions are low). The hydrogenation activity is suppressed by
increased AA partial pressures (i.e., at higher AA coverages) (see
Fig. 11). The found concentration changes are in accordance with
earlier conclusions. As for the activity dependence on the reac-
tant partial pressures, it is in line with the Langmuir–Hinshelwood
kinetics and mechanism.
Above observations concerning the reaction routes in
acetic acid hydroprocessing are summarized in Scheme 1.
Fig. 6. Ethyl acetate yield in the reaction of acetic acid:ethanol (1:1 mol) mixture as
a function of reaction temperature. Reaction conditions as in Fig. 3.

G. Onyestyák et al. / Applied Catalysis A: General 464–465 (2013) 313–321
319
Fig. 8. Stacked area graphs of AA hydroconversion over (a) 9Cu/Al2O3 and (b)
9Cu/Al2O3 + 10 wt% In2O3 catalysts as a function of space time at 320 ◦C and 21 bar
total pressure.
the intermediate acetaldehyde in the Tishchenko reaction [25]
(reaction 4). Ethyl acetate can be hydrogenated directly to two
ethanol molecules most efficiently on the composite catalyst
(reaction 5). Monomolecular ethanol dehydration to ethylene
has not been detected under the reaction conditions applied.
Diethyl-ether can be formed in bimolecular dehydration at higher
ethanol coverage (reaction 6). From acetic acid after AA–anhydride
formation decarboxylation results in formation of acetone and
carbon dioxide (reaction 7) but this reaction is characteristic
only for some kind of supports at high reaction tempera-
ture.
Fig. 7. Stacked area graphs of ethyl acetate hydrogenolysis over (a) Al2O3 + 10 wt%
In2O3, (b) 9Cu/Al2O3 and (c) 9Cu/Al2O3 + 10 wt% In2O3 catalysts as a function of
reaction temperature at 21 bar total pressure. WHSV of ethyl acetate was 1 h−1. The
partial pressure of ethyl acetate and H2 were 2.1 and 18.9 bar.
CH3COOH
Comparison of the octanoic acid and the acetic acid hydro-
conversions over 9Cu/Al2O3 + 10% In2O3 catalyst significant
differences (different by-products), but important similarities
can also be observed (cf. Fig. 3a and Fig. 4d in Ref. [11]). In the
case of octanoic acid, formed alcohol is dehydrated to ether and
octene, while in the case of acetic acid intermediate aldehyde
can be detected in much higher concentration, ether yield is
much less, ethylene or ethane is not formed at all and ester
formation is significant. However, the courses of main trans-
formations, i.e., the selective hydrogenation of carboxylic acids
to alcohols are very similar. The conversions are practically the
same at the same temperature when the molar feed rates of
different carboxylic acids are equal. It means that the reactiv-
ity of carboxylic group hardly depends on the length of alkyl
chains.
CH3COOH
(CH3)2CO + CO2 + H2O
7.
H2
1.
H2O
CH3CHO
CH3COOC2H5 [Tishchenko]
2H2
CH3CHO
4.
H2
2.
5.
3.
[Fischer]
CH3COOC2H5
C2H5OH
All the reactions responsible for the formation of products
are shown in Scheme 1. Acetic acid can be hydroconverted
first to acetaldehyde and then to ethanol (reactions 1 and 2).
Depending on the reaction conditions, fractions of unconverted
reactant acetic acid and the product ethanol can form ethyl
acetate according to the Fischer esterification mechanism (reac-
tion 3). Ethyl acetate could however be formed directly from
H2O
CH3COOH
6.
H2O
C2H5OC2H5
Scheme 1. Schematic diagram of the reactions of AA hydroconversion on the basis
of detected products.

320
G. Onyestyák et al. / Applied Catalysis A: General 464–465 (2013) 313–321
4. Conclusions
Indium doping has been found to be efficient in the reduction
of acetic acid with hydrogen to ethyl alcohol. Hydroconversion of
acetic acid over alumina or silica supported Cu catalysts the ethanol
yield can be increased significantly by indium metal formed by
reduction of the admixed In2O3. This efficient promotion can be
correlated with appearance of a new alloy phase or intermetal-
lic compound (Cu2In), which can also be generated perfectly from
InCl3 as precursor, but in this case the composite catalysts are prac-
tically inactive.
On comparing hydroconversion of octanoic (C8) and acetic
(C2) acid over Cu2In/Al2O3 composite catalyst under similar con-
ditions the molar conversion rates are practically the same. It
suggests that the molecular reduction rate of the carboxylic group
to aldehyde group on the bimetallic surface is approximately equal
for lighter carboxylic acid than for heavier one. However, the
product compositions are different reflecting the differences in
the succeeding reaction steps: After hydroconversion of octanoic
acid ester formation cannot be detected, but alcohol dehydra-
tion is more significant, while for acetic acid ester formation
is considerable and acetaldehyde intermedier formation can be
detected.
Fig. 9. Comparison of the reaction rates over In-free and In doped 9Cu/Al2O3 cata-
lysts in Arrhenius plots for the hydroconversion of acetic acid related for the catalyst
mass at 21 bar total pressure.
The activity dependence on the reactant partial pres-
sures denotes the rate-controlling surface reaction in terms of
Langmuir–Hinshelwood kinetics.
Combining relatively simple pyrolytic (or biochemical) tech-
nologies resulting in production of acetic acid with the catalytic
transformations shown, bioethanol and chemicals, such as ethyl
acetate or acetaldehyde, can be obtained from biomass, e.g., from
lignocellulosic raw material more efficiently over a novel supported
bimetallic catalyst system.
Acknowledgments
The authors wish to express their appreciation to Mrs. Ágnes
Farkas Wellisch for her technical assistance. Thanks is due to
the Hungary-Slovakia Cross-border Co-operation Program (Project
registration number: HUSK/1101/1.2.1/0318) for supporting this
research.
Fig. 10. Stacked area graph of AA/H2/He hydroconversion over 9Cu/Al2O3 + 10%
In2O3 catalysts as a function of H2 partial pressure at 320 ◦C and 21 bar total pres-
sure. WHSV of AA was 1.0 h−1. Constant partial pressure of AA was 2.1 bar. The H2
partial pressure was changed by changing the H2/He ratio.
References
[1] G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (2006) 4044–4098.
[2] J.N. Cheda, G.W. Huber, J.A. Dumesic, Angew. Chem. Int. Ed. 46 (2007)
7164–7183.
[3] D. Kubicˇka, Collect. Czech. Chem. Commun. 73 (2008) 1015–1044.
[4] M. Stöcker, Angew. Chem. Int. Ed. 47 (2008) 9200–9211.
[5] M.F. Demirbas, Appl. Energy 86 (2009) S151–S161.
[6] S.N. Naik, V.V. Goud, P.K. Rout, A.K. Dalai, Renewable Sustainable Energy Rev.
14 (2010) 578–597.
[7] H.N. Chang, N.J. Kim, J. Kang, C.M. Jeong, Biotechnol. Bioprocess. Eng. 15 (2010)
1–10.
[8] M.T. Holtzapple, C.B. Granda, Appl. Biochem. Biotech. 156 (2009)
525–536.
[9] V. Pham, M.T. Holtzapple, M. El-Halwagi, J. Ind. Microbiol. Biotechnol. 37 (2010)
1157–1168.
[10] Gy. Onyestyák, Sz. Harnos, D. Kalló, Catal. Commun. 16 (2011)
184–188.
[11] Gy. Onyestyák, Sz. Harnos, D. Kalló, Catal. Commun. 26 (2012) 19–24.
[12] Sz. Harnos, Gy. Onyestyák, J. Valyon, Appl. Catal.,
439–440.
A
31 (2012)
[13] Gy. Onyestyák, Sz. Harnos, D. Kalló, in: H.G. Woo, H.T. Choi (Eds.), Indium: Prop-
erties, Technological Applications and Health Issues, Nova Science Publishers,
New York, 2013, pp. 53–80.
[14] W. Rachmady, M.A. Vannice, J. Catal. 192 (2000) 322–334.
[15] W. Rachmady, M.A. Vannice, J. Catal. 207 (2002) 317–330.
[16] W. Rachmady, M.A. Vannice, J. Catal. 208 (2002) 158–169.
[17] W. Rachmady, M.A. Vannice, J. Catal. 208 (2002) 170–179.
[18] W. Rachmady, M.A. Vannice, J. Catal. 209 (2002) 87–98.
[19] Sz. Harnos, Gy Onyestyák, D. Kalló, React. Kinet. Mech. Cat. 106 (2012)
99–111.
Fig. 11. Stacked area graph of AA/H2/He hydroconversion over 9Cu/Al2O3 + 10 wt%
In2O3 catalysts as a function of AA partial pressure at 320 ◦C and 21 bar total pressure.
WHSV of AA was 1.0 h−1. Constant partial pressure of H2 was 11.0 bar. The AA partial
pressure was changed by changing the AA/He ratio.

G. Onyestyák et al. / Applied Catalysis A: General 464–465 (2013) 313–321
321
[20] Gy. Onyestyák, Sz. Harnos, Á. Szegedi, D. Kalló, Fuel 102 (2012) 282–288.
[21] Gy. Onyestyák, Sz. Harnos, A. Kaszonyi, M. Stolcova, D. Kalló, Catal. Commun.
[23] Sz. Harnos, Gy. Onyestyák, D. Kalló, Microporous Mesoporous Mater. 167 (2013)
109–116.
[24] Sz. Harnos, Gy. Onyestyák, R. Barthos, M. Stolcova, A. Kaszonyi, J. Valyon, Cent.
ˇ
ˇ
27 (2012) 159–163.
[22] B.D. Cullity, in Elements of X-ray Diffraction, 2nd, Addison-Wesley Publ Co.,
Reading, MA, 1978.
Eur. J. Chem. 10 (2012) 1954–1962.
[25] M.P. Atkins, Top. Catal. 24 (2003) 185–186.