Hydrogenation of ketones over bifunctional Pt-heteropoly acid catalyst in the gas phase

Article abstract of DOI:10.1016/j.apcata.2014.10.032

Gas-phase hydrogenation of a wide range of ketones to alkanes, including hydrogenation of aliphatic ketones and acetophenone, was investigated using bifunctional metal-acid catalysis. The catalysts were comprised of a metal (Pt, Ru, Ni, and Cu) supported on acidic caesium salt of tungstophosphoric heteropoly acid Cs_2.5H_0.5PW₁₂O₄₀ (CsPW). The reaction occurred via a sequence of steps involving hydrogenation of ketone to alcohol on metal sites followed by dehydration of alcohol to alkene on acid sites and finally hydrogenation of alkene to alkane on metal sites. Catalyst activity decreased in the order: Pt > Ru >> Ni > Cu. Pt/CsPW showed the highest catalytic activity, giving almost 100% alkane yield at 100 °C and 1 bar pressure. Evidence is provided that the reaction with Pt/CsPW at 100 °C is limited by ketone-to-alcohol hydrogenation, whereas at lower temperatures (≤60 °C) by alcohol dehydration yielding alcohol as themain product. The catalyst comprised of a physical mixture of Pt/C + CsPW was found to be highly efficientas well, which indicates that the reaction is not limited by migration of intermediates between metal andacid sites in the bifunctional catalyst.

Full text of DOI:10.1016/j.apcata.2014.10.032

G Model
APCATA-15064; No. of Pages6
ARTICLE IN PRESS
Applied Catalysis A: General xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Hydrogenation of ketones over bifunctional Pt-heteropoly acid
catalyst in the gas phase
K. Alharbi, E.F. Kozhevnikova, I.V. Kozhevnikov∗
Department of Chemistry, University of Liverpool, Liverpool, L69 7ZD, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Gas-phase hydrogenation of a wide range of ketones to alkanes, including hydrogenation of aliphatic
ketones and acetophenone, was investigated using bifunctional metal–acid catalysis. The catalysts were
comprised of a metal (Pt, Ru, Ni, and Cu) supported on acidic caesium salt of tungstophosphoric heteropoly
acid Cs2.5H0.5PW12O40 (CsPW). The reaction occurred via a sequence of steps involving hydrogenation of
ketone to alcohol on metal sites followed by dehydration of alcohol to alkene on acid sites and finally
hydrogenation of alkene to alkane on metal sites. Catalyst activity decreased in the order: Pt > Ru » Ni > Cu.
Pt/CsPW showed the highest catalytic activity, giving almost 100% alkane yield at 100 ◦C and 1 bar pres-
sure. Evidence is provided that the reaction with Pt/CsPW at 100 ◦C is limited by ketone-to-alcohol
hydrogenation, whereas at lower temperatures (≤60 ◦C) by alcohol dehydration yielding alcohol as the
main product. The catalyst comprised of a physical mixture of Pt/C + CsPW was found to be highly efficient
as well, which indicates that the reaction is not limited by migration of intermediates between metal and
acid sites in the bifunctional catalyst.
Received 2 September 2014
Received in revised form 14 October 2014
Accepted 18 October 2014
Available online xxx
Keywords:
Ketone
Hydrogenation
Deoxygenation
Bifunctional catalysis
Platinum
Heteropoly acid
© 2014 Published by Elsevier B.V.
1. Introduction
secondary alcohol to alkene on acid sites and finally hydrogenation
of alkene to alkane on metal sites (Scheme 1).
Oxygen-containing organic compounds such as ketones, car-
boxylic acids, alcohols, and phenols, etc., readily available from
natural resources, are attractive as renewable raw materials for
the production of value-added chemicals and biofuels [1,2]. For
fuel applications, they require reduction in oxygen content to
increase their caloric value. Much current research is focussed
on deoxygenation of organic oxygenates using heterogeneous
catalysis, in particular for the upgrading of biomass-derived oxy-
genates obtained from fermentation, hydrolysis, and fast pyrolysis
of biomass [3–7].
Biomass-derived ketones can be further upgraded by aldol con-
densation and hydrogenation to produce alkanes that fall in the
gasoline/diesel range. The hydrogenation of ketones on supported
metal catalysts (e.g., Pt/C and Pd/C) to form alcohols is feasi-
ble and well documented [8]; however, further hydrogenation to
alkanes is rather difficult to achieve on such catalysts [9,10]. The
ketone-to-alkanehydrogenationcan be achieved much easier using
bifunctional metal–acid catalysts ([9–11] and references therein).
This process occurs via a sequence of steps involving hydrogena-
tion of ketone to alcohol on metal sites followed by dehydration of
Previously, this group have reported that platinum on acidic
supports, namely Pt on zeolite HZSM-5 [9] and acidic caesium salt of
tungstophosphoric heteropoly acid Cs2.5H0.5PW12O40 (CsPW) [10],
is active bifunctional catalyst for hydrogenation of methyl isobutyl
ketone (MIBK) to the corresponding alkane 2-methylpentane (MP)
in the gas phase. 0.5%Pt/CsPW, possessing very strong Brønsted
acidity in addition to Pt metal sites, has been found to be partic-
ularly efficient catalyst giving 100% yield of MP at 100 ◦C and 1 bar
pressure without alkane isomerisation [10]. Recently, Mizuno et al.
[11] have applied this catalyst for hydrogenation of ketones, phe-
nols, and ethers in the liquid phase at 120 ◦C and 5 bar H2 pressure.
Here, we investigate the hydrogenation (hydrodeoxygenation)
of a variety of ketones including aliphatic ketones and acetophe-
none in the gas phase using bifunctional metal–acid catalysts
comprising Pt, Ru, Ni, and Cu supported on CsPW. It is demon-
strated that 0.5%Pt/CsPW is a highly efficient and versatile catalyst
for the ketone-to-alkane hydrogenation, and an insight into reac-
tion mechanism is gained.
2. Experimental
2.1. Chemicals and catalysts
∗
MIBK (99%), diisobutyl ketone (80%), acetophenone (≥98%), 2-
octanone (98%), cyclohexanone (≥99%), 3-pentanone (≥99%), and
Corresponding author. Tel.: +44 151 794 2938; fax: +44 151 794 3588.
E-mail address: kozhev@liverpool.ac.uk (I.V. Kozhevnikov).
http://dx.doi.org/10.1016/j.apcata.2014.10.032
0926-860X/© 2014 Published by Elsevier B.V.
Please cite this article in press as: K. Alharbi, et al., Appl. Catal. A: Gen. (2014), http://dx.doi.org/10.1016/j.apcata.2014.10.032

G Model
APCATA-15064; No. of Pages6
ARTICLE IN PRESS
K. Alharbi et al. / Applied Catalysis A: General xxx (2014) xxx–xxx
2.2. Techniques
2
The BET surface area and porosity of catalysts were determined
from nitrogen physisorption measured on a Micromeritics ASAP
2010 instrument at −196 ◦C. Before measurement, the samples
were evacuated at 250 ◦C for 2 h. Powder X-ray diffraction (XRD)
spectra of catalysts were recorded on a PANalytical Xpert diffrac-
Scheme 1. Ketone hydrogenation via bifunctional metal–acid catalysis.
˚
tometer with a CuK␣ radiation (ꢀ = 1.542 A). XRD patterns were
inorganic chemicals used for catalysts preparation were purchased
from Aldrich and 2-butanone (99%) and 2-hexanone (98%) from
Acros Organics. Carbon-supported platinum 10%Pt/C (7.1% Pt from
ICP analysis) was from Johnson Matthey. H2 and N2 gases (>99%)
were supplied by the British Oxygen Company.
Cs2.5H0.5PW12O40 (CsPW) was prepared according to the liter-
ature procedure [12] by adding dropwise an aqueous solution of
Cs2CO3 (0.47 M) to an aqueous solution of H3PW12O40 (0.75 M) at
40 ◦C with continuous stirring. The precipitate obtained was aged
in aqueous slurry for 24 h at room temperature. The slurry was
then slowly evaporated to dryness in a rotary evaporator at 45 ◦C
to afford the CsPW as a white powder. The catalyst was calcined
under vacuum at 150 ◦C/10−3 kPa for 1.5 h.
Bifunctional metal–acid catalysts were prepared by wet impreg-
nation of CsPW with an appropriate metal precursor (Pt(acac)2,
H2PtCl6, RuCl3, Ni(NO3)2, and Cu(NO3)2) followed by reduction
of metal ion to metal with H2. The metal loadings quoted were
confirmed by the ICP-AES elemental analysis; these were in good
agreement with the preparation stoichiometries since the prepa-
rations did not involve operations such as filtration and washing
which could cause metal loss.
0.5%Pt/CsPW was prepared by stirring CsPW powder with
0.02 M Pt(acac)2 solution in benzene at room temperature for 1 h,
followed by slow evaporation of benzene in a rotary evaporator
at room temperature [13]. The catalyst was calcined under vac-
uum at 150 ◦C/10−3 kPa and then reduced in an oven by a hydrogen
flow at 250 ◦C for 2 h. Two other modifications of this catalyst were
prepared by impregnation of CsPW with an aqueous solution of
H2PtCl6, followed by drying in a rotary evaporator at 45 ◦C and
the same calcination and reduction procedures. One, designated as
0.5%Pt/CsPW-I, was prepared by direct wet impregnation of CsPW
powder with 0.1 M aqueous solution of H2PtCl6 involving stirring
the aqueous slurry for 24 h at room temperature, followed by the
workup procedure. The other, designated as 0.5%Pt/CsPW-A, was
prepared by adding 0.1 M aqueous solution of H2PtCl6 to the freshly
precipitated aqueous CsPW slurry and ageing the mixture at room
temperature with stirring for 24 h, followed by the workup.
The physical mixture of 10%Pt/C and CsPW containing 0.5% of Pt
was prepared by grinding a 1:20 w/w mixture of the two compo-
nents.
attributed using the JCPDS database. ICP-AES elemental analysis
was carried out on a Spectro Ciros emission spectrometer.
Metal dispersion of Pt and Ru in the catalysts was mea-
sured in a flow system by hydrogen chemisorption using the
hydrogen–oxygen titration pulse method described previously
[14]. A catalyst sample (50 mg) reduced by hydrogen was pre-
exposed to air at room temperature for 1 h then placed in a glass
sample tube connected to a Micromeritics TPD/TPR 2900 instru-
ment and stabilised at a specified temperature under nitrogen flow.
The hydrogen–oxygen titration was carried out at room tempera-
ture for Pt catalysts and at 100 ◦C for Ru catalysts. Fifty microliter
pulses of pure H2 were injected in the N2 flow in 3 min inter-
vals until the catalyst was saturated with hydrogen. The metal
dispersion, D, defined as the fraction of metal (M) at the surface,
D = Ms/Mtotal, was calculated assuming the stoichiometry of H2
adsorption: MsO + 1.5 H2 → MsH + H2O [15,16]. The average diam-
eter of metal particles, d, was obtained from the empirical equation
d (nm) = 0.9/D [16]. For the commercial 10%Pt/C catalyst, Pt disper-
sion was also determined by pulse chemisorption of CO in He flow
at 50 ◦C (20 mg catalyst sample, 50 ␮L pulses of pure CO, adsorp-
tion stoichiometry Pts:CO = 1). For Cu catalysts, the metal particle
size was determined by XRD using the Scherrer equation, with line
broadening assessed as the full width at half maximum intensity
(FWHM). The dispersion of Cu particles, D, was calculated from the
equation d (nm) = 1.1/D [17].
2.3. Catalyst testing
The hydrogenation of ketones was carried out in the gas phase
in flowing H2. The catalysts were tested at 60–100 ◦C under atmo-
spheric pressure in a Pyrex fixed-bed down-flow reactor (9 mm
internal diameter) fitted with an on-line gas chromatograph (Var-
ian Star 3400 CX instrument with a 30 m × 0.25 mm HP INNOWAX
capillary column and a flame ionisation detector). The tempera-
ture in the reactor was controlled by a Eurotherm controller using
a thermocouple placed at the top of the catalyst bed. The gas
feed contained a variable amount ketone in H2 as a carrier gas.
The ketone was fed by passing H2 carrier gas flow controlled by
a Brooks mass flow controller through a stainless steel satura-
tor, which held the liquid ketone at appropriate temperature to
maintain the chosen reactant partial pressure. The downstream
gas lines and valves were heated to 180 ◦C to prevent substrate
and product condensation. The gas feed entered the reactor at the
top at a flow rate of 20–100 mL min−1. The reactor was packed
with 0.2 g catalyst powder of 45–180 ␮m particle size. In some
cases, to reduce conversion, a smaller amount of catalyst was used
as a homogeneous mixture with silica of a total weight of 0.2 g.
Prior to reaction, the catalysts were pre-treated in H2 for 1 h at the
reaction temperature unless stated otherwise. The dehydration of
2-methyl-4-pentanol was studied similarly, except using N2 as a
carrier gas instead of H2. Once reaction started, the downstream
gas flow was analysed by the on-line GC to obtain reactant conver-
sion and product selectivity. The selectivity was defined as moles
of product formed per one mole of reactant converted and quoted
in mole per cent. The mean absolute percentage error in conversion
and selectivity was ≤10% and the carbon balance was maintained
within 95%.
Two modifications of 5%Ru/CsPW, designated as 5%Ru/CsPW-
I and 5%Ru/CsPW-A, were prepared by impregnation of CsPW
with 0.1 M aqueous solution of RuCl3 similar to the preparation
of 0.5%Pt/CsPW-I and 0.5%Pt/CsPW-A.
10%Cu/CsPW catalyst, designated as 10%Cu/CsPW-I, was pre-
pared as described elsewhere [13] by stirring CsPW powder with an
aqueous solution of Cu(NO3)2·6H2O for 24 h at room temperature,
followed by drying in a rotary evaporator at 65 ◦C and calcination at
150 ◦C/10−3 kPa for 1.5 h. Finally, the sample was reduced in H2 flow
at 400 ◦C for 2 h. Another modification of this catalyst, designated
as 10%Cu/CsPW-A, was prepared by adding an aqueous solution
of Cu(NO3)2·6H2O to the freshly precipitated aqueous CsPW slurry
and ageing the mixture at room temperature with stirring for 24 h,
followed by the same workup.
Two modifications of 10%Ni/CsPW catalyst, designated as
10%Ni/CsPW-I and 10%Ni/CsPW-A, were prepared similarly to the
corresponding 10%Cu/CsPW catalysts using Ni(NO3)2·6H2O as a
precursor.
Please cite this article in press as: K. Alharbi, et al., Appl. Catal. A: Gen. (2014), http://dx.doi.org/10.1016/j.apcata.2014.10.032

G Model
APCATA-15064; No. of Pages6
ARTICLE IN PRESS
K. Alharbi et al. / Applied Catalysis A: General xxx (2014) xxx–xxx
3
Table 1
Catalyst characterisation.
a
Catalyst
SBET
Pore volumeb Pore sizec Dd
de (nm)
(m2 g−1
)
(cm3 g−1
)
(Å)
0.5%Pt/CsPW
0.5%Pt/CsPW-I
0.5%Pt/CsPW-A
5.0%Ru/CsPW-I
5.0%Ru/CsPW-A
10%Cu/CsPW-I
10%Cu/CsPW-A
10%Ni/CsPW-I
10%Ni/CsPW-A
10%Pt/C
128
108
144
103
116
35
35
93
74
801
0.100
0.091
0.085
0.088
0.084
0.044
0.023
0.090
0.049
0.58
30
34
24
34
29
51
27
39
26
29
0.46
0.10
0.11
0.048
0.054
0.019h
0.012h
2.0f
9.0f
8.2f
19f
f
e
17f
d
59g
88g
c
b
a
0.44i
2.0f
25
30
35
40
45
50
55
60
65
70
2 Theta (deg)
a
BET surface area.
Single point total pore volume.
Average BET pore diameter.
Metal dispersion in catalysts as determined from H2 chemisorption.
Metal particle diameter.
Values obtained from the equation d (nm) = 0.9/D [16].
From XRD.
b
Fig. 1. XRD patterns (CuK␣ radiation) for fresh catalysts: (a) 0.5%Pt/CsPW, (b)
10%Cu/CsPW-I, (c) 10%Cu/CsPW-A, (d) 10%Ni/CsPW-I, (e) 10%Ni/CsPW-A, and (f)
10%Pt/C.
c
d
e
f
g
h
i
of 10%Ni/CsPW-I and 10%Ni/CsPW-A did not reveal any Ni metal
phase (Fig. 1), which prevented the determination of Ni dispersion
in these catalysts. The absence of Ni phase may be explained by Ni
oxidation to NiO on exposure of these catalysts to air [19]. It should
be noted, however, that no XRD pattern of NiO was observed either
(Fig. 1), which may be explained by fine dispersion of the NiO in
our catalysts.
Values obtained from the equation D = 1.1/d (nm) [17].
Form CO chemisorption.
3. Results and discussion
3.1. Catalyst preparation and characterisation
For 10%Pt/C catalyst, Pt dispersion and particle size were deter-
mined from CO chemisorption (D = 0.44 0.07, d = 2.0 nm, average
of three measurements); these values are in agreement with many
literature reports. It should be noted that H2 titration overestimated
Pt dispersion for this catalyst to give a value of D = 1.2 0.1. This
may be explained by hydrogen spillover onto the carbon support,
which has been reported previously ([20] and references therein).
XRD was also unable to determine correctly the size of Pt parti-
cles. 10%Pt/C exhibited a clear pattern of the Pt fcc phase, with 39.8
(111), 46.3 (200), and 67.5 (220) reflections (Fig. 1), from which the
particle size was estimated to be d = 28 5 nm (average of three
measurements, D = 0.032). This implies that the XRD was biased
towards larger Pt particles.
The catalysts studied together with their characterisation data
are shown in Table 1. To examine the effect of catalyst prepa-
ration on catalyst activity, the preparation procedure was varied
regarding the use of different metal precursors and impregnation
conditions (see Section 2). The metal loading of Pt and Ru was
0.5% and 5%, respectively, and 10% for Cu and Ni catalysts due to
lower catalytic activity of Ru, Cu, and Ni compared to Pt. As can
be seen from Table 1, the catalyst preparation procedure had lit-
tle effect on the catalyst texture, whereas the metal loading had
significant effect on the catalyst surface area. The catalysts had sur-
face areas between 35 and 144 m2 g−1 and low porosities typical of
CsPW-based catalysts. 0.5% Pt and 5% Ru catalysts had the surface
areas above 100 m2 g−1, whereas 10% Ni and Cu catalysts had lower
surface areas, 74–93 m2 g−1 for Ni and 35 m2 g−1 for Cu catalysts.
Table 1 also shows metal dispersion, D, and metal particle diam-
eter, d, in the catalysts. For Pt/CsPW and Ru/CsW catalysts, these
were determined from H2 titration and for Cu/CsPW from XRD.
The results for 0.5%Pt/CsPW catalysts demonstrate a strong effect
of catalyst preparation on the metal dispersion. The use of Pt(acac)2
as a platinum source and benzene as a solvent for impregna-
tion gave a Pt dispersion D = 0.46 corresponding to a Pt particle
diameter d = 2.0 nm. This dispersion was much higher than that
obtained by impregnation with H2PtCl6 from aqueous solution, i.e.,
D = 0.10–0.11 (d = 8.2–9.0 nm) for the Pt/CsPW-I and Pt/CsPW-A cat-
alysts. The mode of impregnation from aqueous media, i.e., with or
without ageing of the PtII–CsPW aqueous slurry, practically did not
affect the Pt dispersion and particle size. The same was observed
for the Ru/CsPW-I and Ru/CsPW-A. On the other hand, Cu/CsPW-A
had larger Cu particles (88 nm), hence a lower Cu dispersion, than
Cu/CsPW-I (59 nm) (Table 1).
Previous H2-TPR, XRD, and FTIR studies have shown that CsPW
in Pt/CsPW and Pd/CsPW catalysts is resistant to reduction by
H2 below 600 ◦C, and the primary (Keggin) structure of CsPW is
retained in CsPW-supported Pt, Pd, and Cu catalysts after H2 treat-
ment at 400 ◦C [13].
3.2. Hydrogenation of MIBK over CsPW-supported metal catalysts
The hydrogenation of methyl isobutyl ketone (MIBK) was stud-
ied in more detail using CsPW-supported Pt, Ru, Ni, and Cu catalysts
in order to optimise catalyst preparation and gain an insight into
reaction mechanism. Representative results are shown in Table 2.
CsPW alone in the absence of metal exhibited very low activity
at 80–100 ◦C. As found previously [10], non-acidic metal catalysts,
such as Pt/C and Ru/C, are active in hydrogenation of MIBK to
alcohol, 2-methyl-4-pentanol (MP-ol), at 100 ◦C, but further hydro-
genation to alkane, 2-methypentane (MP) becomes feasible only at
temperatures as high as 300 ◦C. In contrast, bifunctional metal–acid
catalyst 0.5%Pt/CsPW, prepared from Pt(acac)2 in benzene solu-
tion, showed excellent activity in MIBK hydrogenation, giving 100%
MP yield at 100 ◦C and 1 bar pressure (Table 2). No MP isomeri-
sation was observed at this temperature. It can be seen that both
0.5%Pt/CsPW-I and 0.5%Pt/CsPW-A, prepared from H2PtCl6 in aque-
ous solution, were less active than the 0.5%Pt/CsPW catalyst (cf.
MIBK conversions at 60 and 80 ◦C). This may be explained by the
higher Pt dispersion in 0.5%Pt/CsPW (Table 1) and the presence of
chloride in 0.5%Pt/CsPW-I and 0.5%Pt/CsPW-A. Reaction selectivity
was greatly affected by the temperature. At 60 ◦C, MP-ol was the
Fig. 1 shows the results of powder XRD analysis of CsPW-
supported metal catalysts. The XRD of 0.5%Pt/CsPW displays only
the well-known pattern of crystalline CsPW [18]; it did not reveal
any Pt metal phase, which is probably due to the fine disper-
sion and low concentration of Pt in the catalyst. 10%Cu/CsPW-I
and 10%Cu/CsPW-A exhibited sharp XRD pattern of Cu metal, with
(111) and (200) reflections at 43.1 and 50.3◦, respectively, in agree-
ment with the standard values 43.3 and 50.4◦ (JCPDS, copper
04-0836). From these, a mean diameter of 59–88 nm for Cu par-
ticles was obtained using the Scherrer equation. The XRD spectra
Please cite this article in press as: K. Alharbi, et al., Appl. Catal. A: Gen. (2014), http://dx.doi.org/10.1016/j.apcata.2014.10.032

G Model
APCATA-15064; No. of Pages6
ARTICLE IN PRESS
K. Alharbi et al. / Applied Catalysis A: General xxx (2014) xxx–xxx
Table 3
4
Table 2
Hydrogenation of MIBK over bifunctional metal–acid catalysts.a
Hydrogenation of MIBK over Pt/CsPW and Pt/C + CsPW catalysts.a
Temperature Conversion
Selectivity (mol%)
Temperature Conversion
Selectivity (mol%)
Catalyst
Catalyst
(◦C)
(%)
(◦C)
(%)
MP
MP-ol
Otherb
MP
MP-ol
Otherb
CsPW
80
100
1
3
27
22
0
0
73
78
60
80
100
94
99
100
12
60
100
81
29
0
7
11
0
0.5%Pt/CsPW
60
80
100
76
97
100
13
42
100
80
50
0
7
8
0
0.5%Pt/CsPW
10%Pt/C + CsPWc
60
80
100
91
97
99
17
90
100
81
10
0
2
1
0
60
80
100
32
88
98
41
89
98
49
11
0
10
0
2
0.5%Pt/CsPW-I
0.5%Pt/CsPW-A
5%Ru/CsPW-I
5% Ru/CsPW-A
a
Reaction conditions: 0.2 g catalyst, 3.6% MIBK in H2 flow, 1 bar pressure,
20 mL min−1 flow rate, 2 h time on stream, catalyst pre-treatment at reaction tem-
perature for1 h in H2 flow.
C1–C5 cracking products, mainly propene and butenes, together with small
amount of C6+ condensation products.
60
80
100
59
94
100
16
29
100
84
63
0
0
8
0
b
c
Physical mixture of 0.02 g 10%Pt/C + 0.18 g CsPW (0.5% Pt content).
60
80
100
36
71
96
24
75
100
76
25
0
0
0
0
and MIBK conversion scaling almost linearly with the Pt loading in
the catalyst, are also in agreement with reaction occurring under
kinetic control.
60
80
100
31
66
90
8
29
98
92
67
2
0
4
0
Table 3 compares the performances of bifunctional catalyst
0.5%Pt/CsPW having metal and acid sites in a rather close proxim-
ity to each other and the corresponding 1:20 w/w physical mixture
of 10%Pt/C and CsPW containing 0.5% of Pt, in which metal and
acid sites are a longer distance apart. It can be seen that these
two catalysts give very similar MIBK conversions in the temper-
ature range of 60–100 ◦C. This indicates that the reaction is not
limited by migration of intermediates between metal and acid sites
in the bifunctional catalyst [22]; hence, the metal and acid sites are
10%Ni/CsPW-I
10%Ni/CsPW-A
10%Cu/CsPW-I
10%Cu/CsPW-A
350
350
350
350
24
19
9
93
95
93
97
0
0
2
0
7
5
5
3
4
a
Reaction conditions: 0.2 g catalyst, 3.6% MIBK in H2 flow, 1 bar pressure,
20 mL min−1 flow rate, 2 h time on stream, catalyst pre-treatment at 100 ◦C/1 h in
H2 flow.
C1–C5 cracking products, mainly propene and butenes, together with small
amount of C6+ condensation products.
b
100
80
main product, whereas at 100 ◦C, MP was formed in almost 100%
yield, in agreement with the previous report [10]. This suggests the
change of the rate-limiting step with increasing the temperature
(see below).
60
40
20
0
Conversion
2MP
MP-ol
5%Ru/CsPW-I and 5%Ru/CsPW-A exhibited close activities
in MIBK hydrogenation (Table 2). These catalysts matched
0.5%Pt/CsPW in activity and selectivity at 100 ◦C, but at a ten
times higher metal loading, in agreement with the previous report
[10], yet they were considerably less active than 0.5%Pt/CsPW
at lower temperatures 60–80 ◦C. Ni/CsPW and Cu/CsPW catalysts
were much less active, showing a moderate activity at 350 ◦C,
with Ni being more active than Cu. The mode of preparation of
these catalysts, i.e., with or without ageing of CsPW and metal
precursor aqueous slurry, had little effect on their performance. It
should be noted that since the non-metal doped CsPW support also
showed some catalytic activity (Table 2), the activity of Ni/CsPW
and Cu/CsPW could to some extent be attributed to the CsPW rather
than to the Ni and Cu. Therefore, the activity of the catalysts stud-
ied in terms of MIBK conversion per unit metal weight decreased
in the order: Pt > Ru » Ni > Cu. The 0.5%Pt/CsPW catalyst, prepared
from Pt(acac)2 as a platinum source in benzene solution, showed
the best performance in the MIBK-to-MP hydrogenation.
other
0
50
100
150
Time (min)
Fig. 2. MIBK hydrogenation over 0.5%Pt/CsPW (0.2 g catalyst, 60 ◦C, 3.6% MIBK in H2
flow, 1 bar pressure, 20 mL min−1 flow rate, catalyst pre-treatment at 100 ◦C/1 h in
H2 flow).
3
2.5
2
The time course shown in Fig. 2 demonstrates stable catalytic
activity of 0.5%Pt/CsPW in MIBK hydrogenation for 2.5 h on stream.
Previously, extended stability tests showed no catalyst deactivation
at least for 14 h on stream [10]. The reaction obeys the Arrhe-
nius equation with an activation energy Ea = 69 kJ mol−1 in the
temperature range 80–110 ◦C where MP is by far the main reac-
tion product (Fig. 3). The high activation energy indicates that
the reaction occurred under kinetic control. This is supported by
the Weisz–Prater analysis [21] of the reaction system. Assum-
ing spherical catalyst particles and Knudsen diffusion regime, the
Weisz–Prater criterion was calculated to be CWP = 1.2 × 10−2 < 1
indicating no internal diffusion limitations. Other results reported
previously [10], such as the close to zero reaction order in MIBK
1.5
1
0.5
0
0.00255 0.0026 0.00265 0.0027 0.00275 0.0028 0.00285
1/T (K-1)
Fig. 3. Arrhenius plot for MIBK hydrogenation over 0.5%Pt/CsPW (0.025 g catalyst
diluted with 0.175 g SiO2, 3.6% MIBK in H2 flow, 1 bar pressure, 100 mL min−1 flow
rate, MIBK conversion range X = 3–17%).
Please cite this article in press as: K. Alharbi, et al., Appl. Catal. A: Gen. (2014), http://dx.doi.org/10.1016/j.apcata.2014.10.032

G Model
APCATA-15064; No. of Pages6
ARTICLE IN PRESS
K. Alharbi et al. / Applied Catalysis A: General xxx (2014) xxx–xxx
5
activation energy that the MIBK-to-MP hydrogenation. Conse-
quently, at lower temperatures, ∼60 ◦C, MIBK-to-MP hydrogena-
tion appears to be limited by MP-ol dehydration, resulting in
formation of MP-ol as the main product. At higher temperatures,
∼100 ◦C, the MP-ol dehydration step, having higher activation
energy, becomes faster than the MIBK-to-MP-ol hydrogenation
step. The latter becomes the rate-limiting step, which results in
high MP selectivity. The last step in Scheme 1, i.e., 2-methylpentene
hydrogenation, appears to be fast, which is supported by the
absence of 2-methylpentenes in the products of MIBK hydrogena-
tion. Alkene hydrogenation is known to be significantly exothermic
and fast, especially on Pt catalysts ([24] and references therein). The
heat of hydrogenation of 4-methyl-1-pentene has been reported
tobe −121 kJ mol−1 [24]. Therefore, hydrogenation of ketones to
alkenes rather than to alkanes via metal–acid bifunctional pathway
on a single catalyst bed is not feasible. This could be better achieved
by a consecutive two-step process with ketone-to-alcohol hydro-
genation on a metal catalyst as the first step followed by alcohol
dehydration on an acid catalyst as the second step.
Scheme 2. Dehydration of 2-methyl-4-pentanol over CsPW.
0.030
0.025
0.020
0.015
0.010
0.005
0.000
0
0.5
1
1.5
2
2.5
3
P (kPa)
3.4. Hydrogenation of aliphatic ketones over Pt/CsPW
Fig. 4. Effect of MP-ol partial pressure on the rate of MP-ol dehydration over CsPW
(0.025 g catalyst diluted with 0.175 g SiO2, 80 ◦C, 1 bar pressure, 100 ml min−1 N2
flow rate, X = 8–20%).
Table 5 shows the results for hydrogenation of C3–C9 aliphatic
ketones over 0.5%Pt/CsPW. It can be seen that these ketones reacted
very similarly to MIBK. At 60 ◦C, alcohols were the main products,
except for the higher C8–C9 ketones 2-octanone and diisobutyl
ketone, which gave mainly the corresponding alkanes. The lat-
ter indicates that the corresponding C8–C9 alcohols were easier
to dehydrate over CsPW as compared to the alcohols related to
the lower ketones. At 100 ◦C, all ketones gave the correspond-
ing alkanes in high yields 87–100%. Therefore, 0.5%Pt/CsPW is a
versatile bifunctional catalyst for the gas-phase hydrogenation of
aliphatic ketones to alkanes.
not required to be located in close proximity for the reaction to
occur. It should be noted, however, that the 0.5%Pt/CsPW catalyst
produced more by-products (C1–C5 hydrocarbons and C6+ conden-
sation products) than the mixed 10%Pt/C + CsPW catalyst (Table 3).
3.3. Dehydration of 2-methyl-4-pentanol over CsPW
The dehydration of the secondary alcohol MP-ol is the second
step in MIBK hydrogenation through the bifunctional metal–acid-
catalysed pathway (Scheme 1). It was studied to obtain knowledge
about the rate-limiting step in the MIBK hydrogenation.
3.5. Hydrogenation of acetophenone over Pt/CsPW
Acetophenone is an example of aromatic ketone, and its hydro-
genation is more complicated than that of aliphatic ketones
MP-ol dehydration over CsPW was found to yield two 2-
methylpentene isomers with high selectivity (Scheme 2). MP-ol
conversion increased with increasing the temperature to reach
100% above 80 ◦C (Table 4). The reaction was found to be close
to zero order in MP-ol (0.15 order) in the MP-ol partial pressure
range of 1–3 kPa (Fig. 4). Previously, zero order in alcohol has been
observed for isopropanol dehydration over CsPW [23]. The activa-
tion energy of MP-ol dehydration was found to be Ea = 130 kJ mol−1
in the temperature range of 60–80 ◦C (determined under differen-
tial conditions at MP-ol conversion X = 0.8–12%).
Previously, it has been suggested that MIBK-to-MP hydro-
genation over Pt/CsPW at 100 ◦C is limited by the first step, i.e.,
hydrogenation of MIBK to MP-ol (Scheme 1). This is mainly based
on the fact that reaction rate scales with Pt loading, while MP
selectivity remains constant ∼100% [10]. The results on MP-ol
dehydration obtained here fully support this view. First, the MP-
ol dehydration is fast at 100 ◦C; and second, it has much higher
Table 5
Hydrogenation of aliphatic ketones over 0.5%Pt/CsPW.a
Ketone
Temperature Conversion
Selectivity (mol%)
(◦C)
(%)
Alkane
Alcohol
Otherb
Acetone
60
100
62
98
5
87
70
9
25
4
2-Butanone
3-Pentanone
2-Hexanone
MIBK
60
100
99
100
5
99
77
0
19
1
60
100
94
100
21
100
76
0
3
0
60
100
98
100
21
99
63
0
11
1
60
100
96
100
18
100
77
0
5
0
60
100
92
99
28
99
72
0
0
1
Cyclohexanone
2-Octanonec
Table 4
Dehydration of MP-ol over CsPW.a
60
100
41
95
68
98
27
0
5
2
Temperature
(◦C)
Conversion
(%)
Selectivity (mol%)
2-Methylpentene
Other
60
100
40
99
93
93
4
0
3
7
Diisobutyl ketone
40
60
80
100
3
43
99
79
97
100
100
21
3
0
a
Reaction conditions: 0.2 g catalyst, 2.0% ketone in H2 flow, 1 bar pressure,
20 mL min−1 flow rate, 3 h time on stream, catalyst pre-treatment at 100 ◦C/1 h in
H2 flow.
100
0
a
b
Reaction conditions: 0.2 g catalyst, 1.6% MP-ol in N2 flow, 1 bar pressure,
20 mL min−1 flow rate, 3 h time on stream, catalyst pre-treatment at reaction tem-
perature for 1 h in N2 flow.
C1–C5 cracking products together with small amount of ketone condensation
products.
c
5 h time on stream.
Please cite this article in press as: K. Alharbi, et al., Appl. Catal. A: Gen. (2014), http://dx.doi.org/10.1016/j.apcata.2014.10.032

G Model
APCATA-15064; No. of Pages6
ARTICLE IN PRESS
6
K. Alharbi et al. / Applied Catalysis A: General xxx (2014) xxx–xxx
Scheme 3. Hydrogenation of acetophenone via bifunctional metal–acid catalysis.
Table 6
sites. Catalyst activity has been shown to decrease in the order:
Hydrogenation of acetophenone.a
Pt > Ru » Ni > Cu. 0.5%Pt/CsPW has been demonstrated to be versatile
catalyst for the hydrogenation of aliphatic ketones, giving almost
100% alkane yield at 100 ◦C and 1 bar pressure. Evidence has been
provided that the reaction with Pt/CsPW at 100 ◦C is limited by
ketone-to-alcohol hydrogenation, whereas at lower temperatures
(≤60 ◦C) by alcohol dehydration resulting in alcohol formation as
the main product. The catalyst comprising of a physical mixture of
10%Pt/C + CsPW has been found to be highly efficient as well, which
indicates that the reaction is not limited by migration of interme-
diates between metal and acid sites in the bifunctional catalyst.
The mixed 10%Pt/C + CsPW shows better performance stability in
acetophenone hydrogenation compared to the Pt/CsPW catalyst,
which suffers from deactivation.
TOSb
(h)
Conversion
(%)
Selectivity (mol%)
Catalyst
Ethylcyclohexane Ethylbenzene
0.5%Pt/CsPW
0.5%Pt/CsPW
3
5
6
3
5
74
67
59
80
79
98
61
19
96
94
2
39
81
4
0.5%Pt/CsPW
10%Pt/C + CsPWc
10%Pt/C + CsPWc
6
Reaction conditions: 100 ◦C, 0.2 g catalyst, 0.5% acetophenone in H2 flow, 1 bar
pressure, 20 mL min−1 flow rate, 3 h time on stream, catalyst pre-treatment at
100 ◦C/1 h in H2 flow.
a
b
Time on stream.
c
Physical mixture of 0.02 g 10%Pt/C + 0.18 g CsPW (0.5% Pt content).
regarding the reaction selectivity and catalyst stability. The
hydrogenation of acetophenone through bifunctional metal–acid-
catalysed pathway can be represented by Scheme 3.
References
[1] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411–2502.
[2] E.L. Kunkes, D.A. Simonetti, R.M. West, J.C. Serrano-Ruiz, C.A. Gaertner, J.A.
Dumesic, Science 322 (2008) 417–421.
[3] M. Snare, I. Kubickova, P. Maki-Arvela, K. Eranen, D. Yu Murzin, Ind. Eng. Chem.
Res. 45 (2006) 5708–5715.
[4] H. Bernas, K. Eranen, I. Simakova, A.-R. Leino, K. Kordas, J. Myllyoja, P. Maki-
Arvela, T. Salmi, D. Yu Murzin, Fuel 89 (2010) 2033–2039.
[5] J.G. Immer, M.J. Kelly, H.H. Lamb, Appl. Catal. A 375 (2010) 134–139.
[6] P.T. Do, M. Chiappero, L.L. Lobban, D. Resasco, Catal. Lett. 130 (2009) 9L 18.
[7] M. Arend, T. Nonnen, W.F. Hoelderich, J. Fischer, J. Groos, Appl. Catal. A 399
(2011) 198–204.
[8] R.A. Augustine, Heterogeneous Catalysis for the Synthetic Chemist, Marcel
Dekker, Inc., N. Y, 1996.
[9] M.A. Alotaibi, E.F. Kozhevnikova, I.V. Kozhevnikov, J. Catal. 293 (2012)
141–144.
[10] M.A. Alotaibi, E.F. Kozhevnikova, I.V. Kozhevnikov, Chem. Commun. 48 (2012)
7194–7196.
[11] S. Itagaki, N. Matsuhashi, K. Taniguchi, K. Yamaguchi, N. Mizuno, Chem. Lett.
43 (2014) 1086–1088.
[12] Y. Izumi, M. Ono, M. Kitagawa, M. Yoshida, K. Urabe, Microporous Mater. 5
(1995) 255–262.
[13] M.A. Alotaibi, E.F. Kozhevnikova, I.V. Kozhevnikov, Appl. Catal. A 447–448
(2012) 32–40.
[14] F. Al-Wadaani, E.F. Kozhevnikova, I.V. Kozhevnikov, J. Catal. 257 (2008)
199–205.
[15] J.E. Benson, M. Boudart, J. Catal. 4 (1965) 704–710.
[16] J.E. Benson, H.S. Hwang, M. Boudart, J. Catal. 30 (1973) 146–153.
[17] A. Dandekar, M.A. Vannice, J. Catal. 178 (1998) 621–639.
[18] T. Okuhara, H. Watanabe, T. Nishimura, K. Inumaru, M. Misono, Chem. Mater.
12 (2000) 2230–2238.
Two bifunctional catalysts were tested at 100 ◦C: 0.5%Pt/CsPW
and 1:20 w/w mixture of 10%Pt/C and CsPW. The results are shown
in Table 6. Both catalysts gave ethylcyclohexane and ethylbenzene
as the main products. It should be noted that no 1-phenylethanol
and styrene was observed amongst the products, which can be
explained by their high reactivity. Thus, 1-phenylethanol is known
to be dehydrated much easier than aliphatic secondary alcohols
[25]. Initially, 0.5%Pt/CsPW gave 98% selectivity to ethylcyclohex-
ane at 74% acetophenone conversion (73% yield). However, this
catalyst suffered from deactivation, resulting in significant loss in
conversion and ethylcyclohexane selectivity over time on stream.
After 6 h, the conversion dropped to 59%, and ethylcyclohexane
selectivity to 19% in favour of ethylbenzene (81%). In contrast,
the 10%Pt/C + CsPW mixed catalyst showed very little deactiva-
tion with time, yielding 74–77% of ethylcyclohexane and only
3–5% of ethylbenzene over 5 h on stream (Table 6). The poor sta-
bility of 0.5%Pt/CsPW to deactivation may be due to blocking Pt
sites by adsorption of acetophenone and/or ethylbenzene on the
neighbouring strong proton sites of CsPW.
4. Conclusions
[19] M.L. Toebes, J.H. Bitter, A.J. van Dillen, K.P. de Jong, Catal. Today 76 (2002) 33–42.
[20] G. Bergeret, P. Gallezot, in: G. Ertl, H. Knözinger, F. Schüth, J. Weitkamp (Eds.),
Handbook of Heterogeneous Catalysis, 2, Wiley-VCH, New York, 2008, p. 745.
[21] P.B. Weisz, C.D. Prater, Adv. Catal. 6 (1954) 143–196.
[22] P.B. Weisz, Adv. Catal. 13 (1962) 137–190.
[23] G.C. Bond, S.J. Frodsham, P. Jubb, E.F. Kozhevnikova, I.V. Kozhevnikov, J. Catal.
293 (2012) 158–164.
[24] G.C. Bond, Metal-Catalysed Reactions of Hydrocarbons, Springer, N. Y, 2005
(Chapter 7).
In this work, we have investigated the gas-phase hydrogenation
of a wide range of ketones to alkanes, including hydrogenation of
aliphatic ketones and acetophenone, using bifunctional metal–acid
catalysis. The bifunctional catalysts comprise Pt, Ru, Ni, and Cu
metals supported on acidic caesium salt of tungstophosphoric het-
eropoly acid Cs2.5H0.5PW12O40 (CsPW). The reaction occurs via a
sequence of steps involving hydrogenation of ketone to alcohol
on metal sites followed by dehydration of alcohol to alkene on
acid sites and finally hydrogenation of alkene to alkane on metal
[25] C.K. Ingold, Structure and Mechanism in Organic Chemistry, 2nd ed., Bell, Lon-
don, 1969.
Please cite this article in press as: K. Alharbi, et al., Appl. Catal. A: Gen. (2014), http://dx.doi.org/10.1016/j.apcata.2014.10.032