Hydrogenation of but-2-enal over supported Au/ZnO catalysts

Article abstract of DOI:10.1039/b103880j

The hydrogenation of but-2-enal over supported Au catalysts is discussed, together with a detailed characterisation study using X-ray diffraction, infrared spectroscopy and transmission electron microscopy. Au/ZnO catalysts are found to be selective for t

Full text of DOI:10.1039/b103880j

View Article Online / Journal Homepage / Table of Contents for this issue
Hydrogenation of but-2-enal over supported Au/ZnO catalysts
Jillian E. Bailie,a Halim A. Abdullah,b James A. Anderson,b Colin H. Rochester,b
Neville V. Richardson,c Nicholas Hodge,ad Jian-Guo Zhang,d Andy Burrows,d
Christopher J. Kielyd and Graham J. Hutchings*a
a Department of Chemistry, Cardi† University, Cardi†, UK CF10 3T B.
E-mail: hutch=cardi†.ac.uk
b Department of Chemistry, University of Dundee, Dundee, UK DD1 4HN
c School of Chemistry, St. Andrews University, Fife, UK KY 16 9AJ
d Department of Materials Science and Engineering, University of L iverpool, L iverpool,
UK L 69 3BX
Received 30th April 2001, Accepted 24th July 2001
First published as an Advance Article on the web 31st August 2001
The hydrogenation of but-2-enal over supported Au catalysts is discussed, together with a detailed
characterisation study using X-ray di†raction, infrared spectroscopy and transmission electron microscopy.
Au/ZnO catalysts are found to be selective for the formation of the unsaturated alcohol, but-2-en-1-ol rather
than the saturated aldehyde, butanal. In general, the addition of thiophene is found to enhance the yield of the
unsaturated alcohol. Detailed transmission electron microscopy and infrared spectroscopy studies show that
thiophene modiÐcation of Au/ZnO catalysts does not a†ect the Au-particle size or morphology; rather,
thiophene undergoes irreversible dissociative adsorption giving a surface in which the Au sites are
electronically promoted by sulfur. It is observed that thiophene modiÐcation does not give any marked e†ect
on catalyst performance for the catalysts that contain large Au-particles (P10 nm) and, hence, it is considered
that the sulfur promotion observed is associated with smaller Au nanoparticles. The highest but-2-en-1-ol
selectivities (D80%) are observed for 5 wt.% Au/ZnO catalysts reduced at 400 ¡C prior to reaction. It is
proposed that the origin of high selectivity is associated with large Au particles (10È20 nm in diameter) that are
present in this catalyst.
clearly that gold surfaces can activate hydrogen. Subsequent
Introduction
studies conÐrmed that gold can catalyse the hydrogenation of
linear alkenes,15h18 cyclohexene,19,20 alkynes,16,21 benzene22
and also acetone to isopropanol.23 Sermon et al.16 found that
Au/c-Al O catalysts containing less than 5% Au were inac-
The use of gold as a heterogeneous catalyst has received sig-
niÐcant attention in recent years.1 For a considerable time,
gold had been considered to be a metal of limited application
as a catalyst and, indeed, had been used as an inert diluent in
some catalytic systems.2 The renewed interest in gold as a het-
erogeneous catalyst comes from the observation that gold is a
catalyst of choice for two reactions. First, it was initially
predicted3 and subsequently demonstrated4,5 that gold would
be the most active catalyst for the hydrochlorination of
ethyne. Subsequently, Haruta and co-workers6,7 have shown
that supported gold catalysts were highly e†ective for the oxi-
dation of carbon monoxide at ambient conditions and even at
temperatures as low as [76 ¡C. The observation of catalytic
activity for oxidation by molecular oxygen at such low tem-
peratures has prompted a massive body of recent research1
and selective oxidation reactions have also now been reported,
2
3
tive for the hydrogenation of pent-1-ene, but very di†erent
results were obtained using Au/SiO catalysts. When hydro-
2
gen was present in massive excess, low concentrations of Au
supported on SiO were active for pent-1-ene hydrogenation
2
and the maximum activity was observed at 0.04 wt.% Au.
This observation indicates that hydrogenation reactions may
be sensitive to both Au particle size and the nature of the
support, as has been observed with oxidation reactions.6,7
In a prior communication, we have shown24 that supported
catalysts are selective for the hydrogenation of but-2-enal to
form the unsaturated alcohol but-2-en-1-ol. Furthermore, we
demonstrated that this selectivity could be enhanced by pre-
treatment of the supported gold catalysts with thiophene. In
this paper, we extend these earlier studies and show that the
particle size of Au is an important factor in the control of this
reaction selectivity. The observation of a promotional e†ect of
low levels of sulfur was based on our earlier studies using
Cu/c-Al O catalysts.25h27
e.g. Au/TiO for the oxidation of propene to methyl oxirane.8
2
In addition, Au/a-Fe O , is a more active catalyst for the
2
3
water-gas shift reaction at lower temperature than the
CuO/ZnO catalysts that are used commercially.9
Although gold is now being extensively investigated for
both selective and total oxidation reactions, there has been
much less attention given to the activity and selectivity of
selected gold catalysts for hydrogenation reactions. Couper
and Eley10,11 demonstrated that gold surfaces could convert
para-hydrogen to ortho-hydrogen. A number of studies also
showed that both gold foil12,13 and supported gold
catalysts14h16 are active for H ÈD exchange reactions at tem-
2
3
Experimental
Catalyst preparation
A series of ZnO supported Au catalysts with Au metal load-
ings of 0.25, 0.5, 1. 2, 5 and 10 wt.%, were prepared by copre-
cipitation according to the following procedure. A mixed
2
2
peratures above 200 ¡C. These early studies demonstrated
DOI: 10.1039/b103880j
Phys. Chem. Chem. Phys., 2001, 3, 4113È4121
4113
This journal is ( The Owner Societies 2001

View Article Online
solution of Zn(NO ) ÉH O (Aldrich, 99.999%) and
was started either immediately after reduction and cooling, if
3
2
2
HAuCl ÉH O (Strem Chemicals, 99.9% Au), which contained
necessary, to the reaction temperature or 15 min after injec-
tion of thiophene.
4
2
the calculated amounts of the Zn and Au salts required to give
the desired Au loading on ZnO, was heated to 80 ¡C. Na CO
2
3
solution (1 mol l~1) was then added, with continuous stirring,
until pH 9 was reached. The precipitate was then aged for 20
min prior to vacuum Ðltration, washed with hot deionised
Results and discussion
E†ect of [Au] on catalyst performance of Au/ZnO
H O (1 l), dried overnight at 110 ¡C and calcined at 400 ¡C for
2
4
h.
In our earlier communication24 we showed that 5 wt.% Au
supported on ZnO or ZrO was active for the hydrogenation
of but-2-enal and that, for catalysts prepared using coprecipi-
tation, high selectivities could be observed for the unsaturated
alcohol, but-2-en-1-ol. However, the 5 wt.% Au/ZnO catalyst
A ZnO supported catalyst containing 5 wt.% Au metal was
prepared by impregnation. The required amount of ZnO
support material was added to a solution containing a calcu-
2
lated amount of HAuCl to give a 5 wt.% loading of Au0. The
4
mixture was allowed to dry by evaporation to a thick paste at
was more active and selective than the 5 wt.% ZrO catalyst.
2
80 ¡C (ca. 5 h) with continuous stirring, dried at 110 ¡C for 16
In view of this, the hydrogenation of but-2-enal was investi-
h and then calcined at 300 ¡C for 4 h. The ZnO sample was
prepared by precipitation of a basic carbonate by the reaction
of zinc nitrate with sodium carbonate, followed by drying and
calcination at 400 ¡C for 4 h. Prior to use, catalyst samples
were pelleted and sieved to give agglomerates 0.6È1.0 mm in
diameter. Following this, they were reduced in situ in Ñowing
gated over Au/ZnO catalysts containing 1, 2, 5 and 10 wt.%
Au prepared by coprecipitation and the results are given in
Table 1. For all the catalysts, conversion tends to decrease
with increasing time-on-stream. However, the steady state
conversion attained after ca. 60 min, generally increases with
increasing [Au] with the best results being achieved with the 5
wt.% catalyst. In addition, the selectivity to the unsaturated
alcohol but-2-en-l-ol also increased with time-on-stream for
all catalysts and the highest selectivity was observed for 1
wt.% Au. However, the yield of but-2-en-1-ol at steady state
was observed to be 3.9, 6.4, 7.2 and 7.1 ] 10~4 mol g
catalysts~1 h~1 for the 1, 2, 5 and 10 wt.% catalysts respec-
tively. This indicates that, for the selective hydrogenation of
the C2O group in but-2-enal, catalysts containing higher con-
centrations of Au give the best results. By-product formation
is also increased with increasing Au concentration and, in par-
ticular, for 10 wt.% Au/ZuO the formation of total hydro-
genation and hydrogenolysis products becomes quite marked.
In view of this catalysts containing 5 wt.% Au give a reason-
able compromise between enhanced formation of the unsatu-
rated alcohol and reduced by-product formation. In contrast,
a 5 wt.% Au/ZnO catalyst prepared by impregnation, and
subjected to the same reduction and reactions conditions as
the co-precipitated catalyst, was found to be totally inactive
for this hydrogenation reaction.
H (60 ml min~1) by heating from ambient temperature to the
2
required reduction temperature at a rate of 5 ¡C min~1 and
then maintaining that temperature for 1 h. The reaction was
then started or thiophene modiÐcation was performed prior to
reaction.
Catalyst characterisation
X-ray powder di†raction. X-ray di†raction data was
obtained using an Enraf Nonius FR590 di†ractometer
working at 40 kV and 30 mA using Cu-Ka radiation. This
instrument was equipped with a static detector with a measur-
ing arc of 120¡.
Infrared spectroscopy. Compressed self-supporting discs of
calcined catalyst precursor were mounted in a vacuum infra-
red cell Ðtted with Ñuorite optical window and were then
reduced within the cell. Spectra were recorded at 4 cm~1
resolution using a Perkin Elmer Spectrum 2000 Fourier
Transform Infra-red Spectrometer. Unless otherwise stated,
discs were at ambient temperature in the spectrometer beam
The e†ect of reaction temperature on the hydrogenation of
but-2-enal was investigated for 2 and 5 wt.% Au/ZnO cata-
lysts and the results are given in Table 1. Again the conversion
of but-2-enal was found to decrease with time-on-stream at
the lower temperature of 150 ¡C as was observed at 250 ¡C
and, as expected, the steady conversion increases with increas-
ing temperature. The selectivity to but-2-en-1-ol increases with
time-on-stream and higher selectivities are observed at 250 ¡C
than at 150 ¡C.
(ca. 25 ¡C) during the adsorption and desorption of CO and
while spectra were recorded.
Transmission electron microscopy. Samples suitable for
transmission electron microscopy analysis were prepared by
dispersing the catalyst powder onto a lacy carbon Ðlm sup-
ported on a copper mesh grid. Transmission electron micros-
copy observations were made in a JEOL 2000 EX high
resolution electron microscope operating at 200 kV. Chemical
microanalysis was performed in a VG HB601 STEM oper-
ating at 100 kV with a 1 nm probe size, equipped with an
Oxford Instruments.Link EDX analysis system (RTS/FLSZ).
E†ect of reduction temperature on catalyst performance of
5 wt.% Au/ZnO
The hydrogenation of but-2-enal was investigated over 5 wt.%
Au/ZnO that had been reduced at 250È400 ¡C prior to reac-
tion and the results are given in Table 2. The conversion
decreases with increasing time-on-stream for all catalysts, but
at steady state the highest conversion was observed with cata-
lysts reduced at 300 ¡C. In general, the selectivity to the
unsaturated alcohol, but-2-en-1-ol increased with increasing
reduction temperature and, for the catalyst reduced at 400 ¡C,
ca. 80% selectivity to but-2-en-1-ol was observed. The forma-
tion of by-products from dimerisation, total hydrogenation
and hydrogenolysis also decreased with increasing reduction
temperature which accounts, in part, for the improved selec-
tivity to the selective hydrogenation product. It should be
noted that, in previous studies for the hydrogenation of but-2-
enal, which have recently been reviewed by Claus,28 the
highest selectivities for but-2-en-l-ol are in the range 60È70%,
at conversions comparable to those observed in this study. In
Catalyst testing
Reactions were carried out in a continuous Ñow, Ðxed-bed,
micro-reactor (9 mm i.d.) at atmospheric pressure using
on-line GC analysis (Varian 3400 GC, equipped with an FID
and DBwax megabore column, 0.53 mm ] 30 m) for separa-
tion and analysis of the reactant and products. The catalyst
(
200 mg) was reduced in situ in a Ñow of hydrogen (60 ml
min~1) at a speciÐc temperature for 1 h prior to reaction.
But-2-enal was fed into the system by means of a calibrated
syringe pump into a Ñow of hydrogen (weight hourly space
velocity \ 0.7 h~1 and H : but-2-enal ratio \ 14 : 1). ModiÐ-
2
cation of the catalyst was achieved via direct injection of either
0.5 or 1 ll of thiophene into a Ñow of hydrogen with the cata-
lyst sample maintained at 250 ¡C after reduction. The reaction
4114
Phys. Chem. Chem. Phys., 2001, 3, 4113È4121

View Article Online
Table 1 E†ect of Au loading on the performance of Au/ZnO catalysts for but-2-enal hydrogenation
Selectivity (%)a
[Au]
Reaction
temp./¡C
Time/
min
Conversion
(%)
(
wt.%)
BÈAL
B2OL
BÈOL
2EH
2E2H
Others
1b
2c
2b
250
150
250
10
7.2
4.4
3.5
3.3
2.4
45.1
36.8
39.1
37.6
39.9
41.5
57.7
60.9
62.4
60.1
È
È
È
È
È
È
È
È
È
È
È
È
È
È
È
13.4
5.5
È
È
È
3
6
0
0
1
1
20
80
10
7.3
2.9
2.2
1.4
1.3
50
25
6.3
È
È
È
È
16.6
È
È
È
È
È
È
È
È
È
2.2
1.6
1.8
4.4
5.2
2
3
4
6
0
0
0
0
63.2
65.4
75.1
68.8
35.4
32.8
20.5
26.1
10
6.7
8.3
4.7
4.8
5.6
5.2
4.9
6.7
81.8
86.2
82.3
71.0
72.8
65.8
56.3
56.0
2.3
7.3
8.1
2.5
È
È
È
È
È
È
1.5
È
È
È
È
È
È
È
1.6
È
È
È
È
È
È
È
4.7
3.9
2.7
2.2
2.3
1.9
1.5
1.5
2
3
4
6
0
0
5
0
15.0
26.8
24.9
32.2
42.2
42.5
1
1
1
05
20
80
5
c
150
10
6.0
2.5
0.8
1.0
0.4
0.2
20.8
32.6
37.9
49.5
56.7
67.2
44.4
61.3
54.7
48.1
43.3
32.8
33.7
2.5
È
È
È
È
È
È
È
È
È
È
È
È
È
È
È
1.2
3.9
7.4
2.4
È
2
3
6
0
0
0
1
1
25
80
È
È
5
1
b
250
250
10
41.4
8.7
8.9
7.6
7.8
43.8
37.8
35.1
35.9
34.1
29.7
47.8
51.7
51.0
54.1
10.7
È
È
È
È
4.6
È
È
È
È
2.1
0.9
È
1.1
1.1
9.1
13.5
13.2
12.0
10.7
3
6
0
0
1
1
20
80
0b
10
11.9
7.4
8.3
8.1
10.6
38.2
40.2
41.2
39.0
39.0
21.3
45.4
45.6
46.6
46.4
11.5
È
È
È
È
È
È
È
È
È
È
È
È
È
È
29.0
14.4
13.2
14.4
14.6
3
6
0
0
1
1
20
50
a BÈAl: butanal, B2OL: but-2-en-1-ol, BÈOL: butan-1-ol, 2EH: 2-ethyl hexanal, 2E2H: 2-ethyl-2-hexanal, others: butane and C
3 molecules.
b Catalyst reduced at 250 ¡C prior to reaction at 250 ¡C. c Catalyst reduced at 150 ¡C prior to reaction at 150 ¡C.
Table 2 E†ect of reduction temperature on the performance of 5 wt.% Au/ZnO for but-2-enal hydrogenation at 250 ¡C
Reductiona
Product selectivity (%)b
temperature
Time
/min
Conversion
(%)
/¡C
BÈAL
B2OL
BÈOL
2EH
2E2H
Others
250
300
350
400
20
13.1
8.7
8.9
7.6
7.8
32.5
37.8
35.1
35.9
34.1
50.2
47.8
51.7
51.0
54.1
È
È
È
È
È
2.8
È
È
È
È
1.4
0.9
È
1.1
1.1
13.1
13.5
13.2
12.0
10.7
3
6
0
0
1
1
20
80
20
13
28.6
27.7
27.9
25.0
25.2
60.1
57.1
56.0
63.9
62.9
È
È
È
È
È
È
È
È
È
È
È
11.3
10.7
11.0
8.7
3
6
0
0
12.3
10.8
13.5
11.8
4.5
5.1
2.4
2.4
1
1
30
80
9.5
20
6.7
3.4
4.5
4.4
3.7
26.1
29.5
28.4
28.6
30.6
63.8
57.7
61.2
60.8
57.5
È
È
È
È
È
È
È
È
È
È
È
È
È
È
È
10.1
12.8
10.4
10.6
11.9
3
7
0
0
1
1
20
95
20
6.1
7.7
6.3
4.7
5.3
18.9
16.4
17.4
17.6
16.1
76.5
80.2
79.2
79.6
80.7
È
È
È
È
È
È
È
È
È
È
È
È
È
È
È
4.6
3.4
3.4
2.8
3.2
3
6
0
0
1
1
20
85
a Catalysts calcined at 400 ¡C, 4 h; treated with H from ambient temperature to the speciÐed reduction temperature prior to cooling to the
2
reaction temperature of 250 ¡C. b BÈAL: butanal, B2OL: but-2-en-1-ol. BÈOL: butan-1-ol, 2EH: 2-ethyl hexanal, 2E2H: 2-ethyl-2-hexenal,
others: butane and C molecules.
3
Phys. Chem. Chem. Phys., 2001, 3, 4113È4121
4115

View Article Online
view of this, the 80% selectivities observed in this study rep-
resent a signiÐcant improvement.
E†ect of thiophene pre-treatment on Au/ZnO catalysts
In our previous communication, we showed that thiophene
modiÐcation of Au/ZnO and Au/ZrO catalysts could
2
enhance the selectivity and yield of but-2-en-l-ol. The e†ect
was maintained as the catalyst deactivated with increased
time-on-line. We have now investigated the hydrogenation of
but-2-enal over 5 wt.% Au/ZnO that had been reduced at
250 ¡C and pretreated with 0, 0.5 and 1.0 ll thiophene, rep-
resenting 0, 8.2 ] 10~5 and 1.64 ] 10~4 (mol S) (mol Au)~1
respectively, and the results are shown in Fig. 1. Pre-treatment
with thiophene had a marked e†ect on the initial conversion
observed at 10 min time-on-steam, with thiophene pre-treated
catalysts given lower conversion. However, the e†ect was less
pronounced when steady state conversion had been attained.
For the catalyst treated with 0.5 ll thiophene prior to reac-
tion, the yield of the unsaturated alcohol but-2-en-l-ol was sig-
niÐcantly enhanced when steady state conversion had been
attained, compared with both the untreated and the 1.0 ll
thiophene-treated catalysts. This indicates that the e†ect is
sensitive to the concentration of thiophene, which has pre-
viously been observed with thiophene modiÐcation of Cu/c-
Al O catalysts for the same reaction.25h27
Fig. 1 E†ect of thiophene-modiÐcation on but-2-enal conversion
and the yield of but-2-en-1-ol over 5 wt.% Au/ZnO. Key: (K) 0 ll
thiophene; (L) 0.5 ll thiophene; (|) 1.0 ll thiophene. Open symbols:
but-2-enal conversion (%); closed symbols: rate of but-2-en-1-ol for-
mation (10~4 mol g~1 h~1).
at 150 ¡C. However, in this case, the yield of but-2-en-l-ol
remains similar for both untreated and treated catalysts at
steady state conversion. By way of contrast, thiophene pre-
treatment of the 5 wt.% Au/ZnO catalyst, leads to an increase
in the rate of formation of but-2-en-ol (Fig. 2). Hence the e†ect
of thiophene pre-treatment is also sensitive to the concentra-
tion of Au present in the material.
The e†ect of thiophene pre-treatment on a 5 wt.% Au/ZnO
catalyst that had been reduced at 400 ¡C was investigated and
the results are also given in Table 3. Comparison with the
results for the untreated catalyst (Table 2) shows that thioph-
ene pre-treatment of this catalyst does not have a signiÐcant
e†ect, in contrast with the result for the same catalyst reduced
at 250 ¡C (Fig. 2).
2
3
The e†ect of thiophene pre-treatment was further investi-
gated using 2 and 5 wt.% Au/ZnO catalysts pre-treated with
0
1
.5 ll thiophene and the results are given in Table 3. At both
50 and 250 ¡C, thiophene pre-treatment leads to a lower con-
version being observed (compared data in Table 3 with that in
Table 1). For all the catalysts, thiophene pre-treatment leads
to an increase in selectivity to the unsaturated alcohol but-2-
en-l-ol, the e†ect being most pronounced for 2 wt.% Au/ZnO
Table 3 E†ect of reaction temperature on thiophene-pretreated catalysts on the hydrogenation of but-2-enal
Reaction
Product selectivity (%)a
[
Au]
temperature
Time
/min
Conversion
(%)
(
wt.%)
/¡C
BÈAL
B2OL
BÈOL
2EH
2E2H
Other
2
-Sb
-Sc
150
10
4.8
1.3
1.2
0.7
0.5
48.9
37.9
29.9
34.2
31.4
46.2
56.9
64.1
55.4
58.4
È
È
È
È
È
È
È
È
È
È
È
È
È
È
È
5
5.2
6
10.4
10.2
2
3
4
5
0
0
0
0
2
250
150
10
5.0
5.4
4.6
3.8
3.1
3.0
75.2
66.4
59.9
58.7
52.8
55.0
16.9
29.7
36.0
38.6
44.6
42.3
0.9
È
È
È
È
0.4
È
0.9
È
È
È
È
È
È
È
È
È
6.5
3.9
3.2
2.8
2.7
2.7
2
4
6
5
0
0
1
1
20
50
È
5
-Sb
10
5.8
2.4
2.0
1.0
0.5
0.2
25.1
27.2
31.7
42.1
48.6
66.9
65.4
71.5
66.2
53.7
51.4
33.1
7.0
È
È
È
È
È
È
È
È
È
È
È
È
È
È
È
È
2.5
2.4
2.1
È
È
È
2
3
6
0
0
0
1
1
20
80
È
5
-Sc
-Sd
250
250
10
27.9
9.3
11.1
10.9
44.5
32.3
29.6
26.8
20.0
56.6
60.6
65.6
15.8
È
È
1.5
È
È
È
1.6
È
È
È
16.6
11.1
8.8
3
6
0
0
180
È
7.8
5
10
4.8
5.6
4.9
4.8
4.8
5.5
6.1
29.4
16.9
15.8
16.4
15.2
13.3
11.9
55.1
80.3
81.3
81.1
83.3
85.1
86.9
È
È
È
È
È
È
È
3.6
tr
tr
È
È
È
È
È
È
È
È
È
È
È
11.9
2.8
2.1
2.5
1.6
1.6
1.2
2
3
6
9
0
0
0
0
1
1
70
80
a BÈAL: butanol; B2OL: but-2-en-1-ol; BÈOL: butan-1-ol; 2EH: 2-ethylhexanal; 2E2H: 2-ethyl-2-hexenal; other: butane and C products.
3
b Catalyst reduced at 150 ¡C, and treated with 0.5 ll thiophene. c Catalyst reduced at 250 ¡C and treated with 0.5 ll thiophene. d Catalyst reduced
at 400 ¡C, treated with 0.5 ll thiophene and cooled to 250 ¡C for reaction.
4116
Phys. Chem. Chem. Phys., 2001, 3, 4113È4121

View Article Online
bands became indistinguishable at higher coverages with a
combined maximum at 2045 cm~1 (Fig. 4A). Adsorbed CO
molecules giving rise to these bands were resistant to desorp-
tion by evacuation at 25 ¡C. Exposure of 5 wt.% Au/ZnO to
thiophene at 250 ¡C (the temperature of modiÐcation used in
catalysis experiments), followed by evacuation at 25 ¡C and
admission of CO, gave bands at 2108 and 2043 cm~1 (Fig.
4B). The former band disappeared and the latter band grew
and shifted to 2058 cm~1 with increasing CO pressure (Fig.
4
B). Furthermore, a weak band also appeared at 2015 cm~1
with increasing CO pressure. Evacuation of the sample at 298
K for 30 min was observed to reduced the 2058 cm~1 band
intensity by 65%.
Fig. 2 E†ect of thiophene-modiÐcation on the rate of but-2-en-1-ol
formation over wt.% Au/ZnO at 250 ¡C. (…) Au/ZnO; (=)
5
thiophene-modiÐed Au/ZnO.
The observed bands at 2040È2055 cm~1 are at unusually
low wavenumbers for CO adsorbed on Au,1 although bands
at 2060 and 2040 cm~1 have previously been reported for
Characterisation of supported Au catalysts
Au/SiO and Au/MgO respectively, which had been reduced
Powder X-ray di†raction. The powder X-ray di†raction pat-
terns of Au/ZnO samples prepared by coprecipitation as a
function of Au content are shown in Fig. 3. ReÑections due to
Au crystallites were observed only with the highest loading, 10
wt.% Au/ZnO (Fig. 5(e) below). However, the reÑections due
to ZnO broadened with increasing Au content indicating
smaller ZnO crystallites were present. Using the Scherrer
equation, the calculated ZnO particle size decreased from
2
in H at 350 ¡C.30 In earlier studies, a band at 2070È2080
2
cm~1 for Au/SiO reduced at 375 ¡C has been attributed to CO
adsorption on large Au particles.32 A band observed at 2038
cm~1 for Au/TiO was not assigned in the study by Bollinger
2
and Vannice.33 The proposed explanation of the bands at
2
040È2055 cm~1 was that they were due to CO on Au0 ter-
races. However, there is a problem with this interpretation in
that 5 wt.% Au/ZnO material only gave bands around 2045
cm~1, even though the Au particle sizes present would
demand the existence of a signiÐcant proportion of incom-
pletely coordinated edge sites.34 Bands in this region might
[
10 nm for 0.25 wt.% Au/ZnO to 8.5 nm for 10 wt.%
Au/ZnO.
Infrared spectroscopy. Similar spectroscopic characteristics
also be attributed to l vibrations of gem-dicarbonyl species
were observed for the adsorption of thiophene on reduced
ZnO and Au/ZnO. Hydrogen bonding interactions between
weakly adsorbed thiophene molecules and surface hydroxy
groups on ZnO led to decreases in adsorption intensity at
sym
which, for Au(CO) , are known to give a band at 2072
2
cm~1.35 This assignment has been rejected for Au/TiO
2
because of the complete absence of a band below 2000 cm~1
due to l
vibrations which give a band at 1936 cm~1 for
3680, 3618, 3540 and 3440 cm~1. These were observed on a
asym
Au(CO) .36 This band is also absent from the spectra of
broad group band envelope due to vibrations of hydroxy
groups which were perturbed by thiophene. Bands due to
hydroxy groups on ZnO have previously been reported to
exist at 3670, 3620, 3550 and 3440 cm~1.29 Desorption of
thiophene from ZnO or Au/ZnO by evacuation at 25 ¡C was
apparently complete after 30 min, there being no detectable
evidence for residual thiophene or surface products of chemi-
sorption. The hydroxy groups were e†ectively restored to their
initial state before exposure to thiophene by this simple evac-
uation treatment.
2
Au/ZnO catalysts. However, the absence of this band could be
explained of the metalÈsurface selection rule,37 requiring that
only vibrations giving dipole changes perpendicular to the
surface were infrared active, was operating. A shoulder at 1950
cm~1 has been observed for Au/MgO but was assigned to
bridging or multi-bounded CO.30 The assignment of at least a
component of the bands at 2040È2055 cm~1 to gem-
dicarbonyl species at incompletely coordinated Au sites would
be consistent with the expected population of edge sites34 on
2
2
È4 nm particles. However, the absence of a band at 1900È
000 cm~1 does not allow this assignment to be conÐrmed
Admission of CO to Au/ZnO at low pressure gave a band
at 2035 cm~1 with a shoulder at ca. 2045 cm~1. These two
and, therefore, assignment to linear CO still remains a plaus-
ible alternative.
Fig. 4 FTIR spectra of 5 wt.% Au/ZnO reduced at 250 ¡C and (A)
exposed to CO at (a) 2, (b) 6 and (c) 10 mm Hg and (d) evacuated for
30 min. (B) exposed to thiophene (2 mm Hg) at 250 ¡C, evacuated at
25 ¡C and exposed to CO at (a) 2, (b) 6 and (c) 10 mm Hg and (d)
evacuated for 30 min.
Fig. 3 Powder X-ray di†raction patterns of Au/ZnO catalysts pre-
pared by coprecipitation. (a) 0.25 wt.% Au, (b) 0.5 wt.% Au, (c) 1 wt.%
Au, (d) 5 wt.% Au (e) 10 wt.% Au.
Phys. Chem. Chem. Phys., 2001, 3, 4113È4121
4117

View Article Online
The modiÐcation of 5 wt.% AU/ZnO with thiophene at
50 ¡C (Fig. 4B) gave thiophene decomposition products on
faceted, untwinned, single crystals having a truncated cub-
2
octahedral morphology. The 10 wt.% Au/ZnO material (Fig.
5(c)) also showed the presence of a good dispersion of discrete
Au particles, but notably a bimodal size distribution was
apparent. Larger particles with diameters in the range 6È10
nm were commonly observed along with others that were
much smaller in size, typically 1È3 nm.
The e†ect of thiophene pre-treatment and reaction with
but-2-enal was investigated and typical micrographs for 2
wt.% Au/ZnO are given in Fig. 5(d) and (e). The unused cal-
cined sample that had been reduced in hydrogen at 150 ¡C
showed the presence of cub-octahedral particles of Au, that
are typically 2.5 nm in size (Fig. 5(a)). Pre-treatment with 0.5
ll thiophene did not markedly a†ect the Au particle size, mor-
phology, or distribution (Fig. 5(d)). Following reaction for the
hydrogenation of but-2-enal at 150 ¡C, the catalyst again
showed the presence of small, clean Au particles (Fig. 5(e))
suggesting that, under these reaction conditions, the Au parti-
cle size is not a†ected signiÐcantly by the hydrogenation reac-
tion.
the Au Surface which resulted in shifts of 8È13 cm~1 to higher
wavenumbers for subsequently adsorbed CO. These products
could not be detected and, therefore, identiÐed from spectra of
adsorbed thiophene alone. However, the thiophene modiÐ-
cation hardly a†ected the number of available sites for CO
adsorption, but nevertheless, had an e†ect on all the sites in
the sense of strengthening the CÈO bond in the adsorbed CO.
Electron transfer from metal surfaces to adsorbed sulfur
species have been reported previously.27,38,39
Transmission electron microscopy. Samples of 2, 5 and 10
wt.% Au/ZnO prepared by coprecipitation, were examined by
transmission electron microscopy (Fig. 5). The 2 wt.%
Au/ZnO (Fig. 5(a)) showed the presence of 2È4 nm diameter
Au nanoparticles typically exhibiting cub-octahedral or iso-
hedral morphology, as well as some even smaller particles.
The ZnO support particles, for all samples, were highly crys-
talline with 15È20 nm diameter. The 5 wt.% Au/ZnO sample
(
Fig. 5(b)) exhibited a very narrow size distribution of Au par-
The e†ect of reduction temperature on the 5 wt.% Au/ZnO
was also systematically examined by electron microscopy, as
shown in Fig. 6. The unreduced sample (Fig. 6(a)) showed the
ticles with diameters of 4È5 nm, which were homogeneously
distributed over the support. The Au particles were well
Fig. 5 Transmission electron micrographs for Au/ZnO catalysts with di†erent metal loadings. The fresh calcined catalysts are (a) 2 wt.% Au, (b)
wt.% Au, (c) 10 wt.% Au. Image (d) shows the 2 wt.% Au/ZnO catalyst modiÐed by thiophene, (e) shows 2 wt.% Au/ZnO unmodiÐed following
reaction at 150 ¡C.
5
4118
Phys. Chem. Chem. Phys., 2001, 3, 4113È4121

View Article Online
Fig. 6 Transmission electron micrographs of fresh calcined 5 wt.% Au/ZnO catalysts. (a) is the unreduced material, whereas the others are
reduced at (b) 250, (c) 300, (d) 350 and (e) 400 ¡C respectively.
presence of a uniform dispersion of 4 nm diameter Au par-
ticles. These particles were epitaxial with the substrate oxide
and preferentially expose M111N and M100N surface facets that
are characteristic of a cub-octahedral morphology. In many
cases, there is evidence of a thin disordered surface coating of
unknown composition on the Au particles. The sample
reduced at 250 ¡C (Fig. 6(b)) showed the presence of a homo-
geneous distribution of Au nanoparticles with M111N and
M100N surface facets again evident. Most particles were still
typically D4 nm in diameter but a signiÐcant number had
grown to around 6 nm. There was no evidence of a surface
overlayer on any of the Au particles and, in this case, the
particleÈsubstrate epitaxy is less convincing. In addition, there
is some evidence of sub-nanometric “ÑeckÏ-type contrast on
the ZnO support. STEM-energy dispersive X-ray (EDX)
analysis has been used to identify the origin of this contrast.
Fig. 7 shows these EDX spectra. The existence of weak signals
at D2.1 and 8 eV from the area of the “ÑeckÏ is clear evidence
that the contrast originates from a very small Au particle or
raft on the ZnO support. The sample reduced at 300 ¡C (Fig.
6(c)) has a much larger size distribution of Au particles, typi-
cally ranging in diameter from 3È8 nm. There are also a con-
siderable number of sub-1 nm diameter “ÑeckÏ-type Au
particles present. Furthermore, there is little evidence of any
Phys. Chem. Chem. Phys., 2001, 3, 4113È4121
4119

View Article Online
for the Au nanoparticles, with particles of 1È3 nm and 6È10
nm being present. In addition, as the reduction temperature
was increased from 250 to 400 ¡C, it was apparent that this
induced a bimodal distribution in the Au nanoparticles. The
relative proportion of particles in the range 4È5 nm decreased
with increasing reduction temperature giving a concomitant
increase in the proportions of smaller and larger particles. If
we consider the catalytic results of these two sets of samples
together, we note that the sample containing low Au concen-
trations (1.2 wt.% Au) also contain small particles similar to
those observed in the 10 wt.% Au/ZnO sample and the 5
wt.% Au/ZnO sample reduced at higher temperatures. In view
of this fact, we do not consider that the small Au nanoparti-
cles (diameter \2 nm) are associated with the high selec-
tivities to but-2-en-1-ol observed, in the absence of thiophene
modiÐcation, with the high [Au] and high reduction tem-
perature catalysts. Hence, we propose that the active sites for
selective hydrogenation of but-2-enal to but-2-en-1-ol are
associated with the presence of large Au particles. This is in
contrast to the active site in supported Au catalysts used for
the oxidation of CO under ambient conditions. For this latter
reaction, it is generally agreed that the most active catalysts
comprise 2È5 nm Au particles and that it is the periphery
between the Au particle and the support that provides the
active site for CO activation under mild conditions. Larger Au
particles are inactive for CO activation, whereas we consider
that these are both active and selective for the selective hydro-
genation of a,b-unsaturated aldehydes.
Fig. 7 Energy disperse X-ray spectra taken from various regions in
the 5 wt.% Au/ZnO catalyst. These were obtained with a 1 nm probe
positioned on (a) a region of clean ZnO support, (b) a 4 nm Au parti-
cle seen in proÐle and (c) a region showing nm scale “ÑeckÏ-like con-
trast on the ZnO support.
In contrast, the e†ect of thiophene is most marked for the
catalysts with relatively low [Au] (\5 wt.% Au) or prepared
using low reduction temperatures which contain a vast major-
ity of Au nanoparticles with diameters typically in the 2È5 nm
range. In e†ect, this provides evidence in support of the pro-
posal that large Au particles are the origin of the enhanced
selectivity to but-2-en-1-ol. The samples giving the highest
selectivity to but-2-en-1-ol (i.e. 5 wt.% Au/ZnO reduced at
epitaxial relationship between the Au particles and the
support oxide in this sample. The sample reduced at 350 ¡C
(
Fig. 6(d)) shows evidence of considerable sintering of the Au,
since particles 10È15 nm in diameter are not uncommon.
Many of the larger particles show multiple twinning.
However, Au particles in the range 4È6 nm in diameter are
still evident, as are the “ÑeckÏ-like sub-nanometric Au rafts.
The sample subjected to the highest reduction temperature of
400 ¡C) still contain an appreciable concentration of small Au-
400 ¡C (Fig. 6(e)), showed even more evidence for Au sintering
nanoparticles, but the thiophene treatment has a very limited
e†ect on the selectivity to but-2-en-1-ol. If the small Au par-
ticles were the origin of the high selectivity, it could reason-
ably be expected that this would have been enhanced further
on thiophene addition. As this is not observed, it can be con-
cluded that the e†ect is associated with the larger Au particles.
Transmission electron microscopy has shown that thiophene
modiÐcation does not a†ect the Au particle size and so the
origin of the e†ect of thiophene modiÐcation is not expected
to be morphological in nature. It is clear that thiophene modi-
Ðcation leads to a promotion in the rate of formation of but-2-
en-1-ol from the hydrogenation of but-2-enal for supported
Au catalysts. Furthermore, infrared spectroscopy studies indi-
cate that thiophene undergoes irreversible dissociative adsorp-
tion giving a surface on which the Au sites are still available
for CO adsorption, and that the e†ect of the sulfur is to act as
an electronic promoter as observed for Cu catalysts pre-
viously.27h29 It is possible that this is a general e†ect for this
triad of metals and results for Ag/ZnO catalysts will be
reported elsewhere.40
and particles up to 20 nm in diameter have been observed.
Consequently, a much smaller fraction of the Au particles are
now in the 4È6 nm and 1È3 nm size ranges. SigniÐcantly, there
is still some occasional evidence of the presence of sub-nm Au
particles via the “ÑeckÏ-type contrast on the ZnO supports.
Hence, from this detailed transmission electron microscopy
study, it is apparent that, as the reduction temperature is
increased from 250 to 400 ¡C, there is a signiÐcant e†ect on
the Au particle size distribution. As the reduction temperature
is increased, the 4È5 nm Au particles, characteristic of the
unreduced material, are transferred to give Au particles with
both larger and smaller diameters, i.e. creating an e†ective
approximate bimodal distribution.
Comments on the nature of the active site for selective
hydrogenation of but-2-enal to but-2-en-1-ol
There are two e†ects that require further comment, namely (a)
the e†ect of Au particle size on hydrogenation selectivity and
(
b) the e†ect of thiophene modiÐcation. The results of the
present study indicate that, as both the Au loading and cata-
lyst reduction temperature are increased, the selectivity for the
formation of the unsaturated alcohol is enhanced. In addition,
thiophene modiÐcation of catalysts reduced at 400 ¡C has no
signiÐcant e†ect and there is no further enhancement in the
yield of the unsaturated alcohols, whereas for catalysts
reduced at 250 ¡C, there is a pronounced enhancement in the
yield of unsaturated alcohol upon thiophene modiÐcation.
The transmission electron microscopy studies indicate that, as
the Au concentrations and catalyst reduction temperatures are
increased, the Au particle size also generally increases.
However, for the sample containing the highest [Au] investi-
gated (10 wt.% Au/ZnO), there is a clear bimodal distribution
Acknowledgements
We thank the EPSRC and Synetix for Ðnancial support, and
the Malaysian Government and Universiti Putra, Malaysia,
for a research studentship for HAA.
References
1
G. C. Bond and D. T. Thompson, Catal. Rev.-Sci. Eng., 1999, 41,
3
19.
2
F. Besenbacher, I. Chorkensor†, B. S. Clausen, B. Hammer,
A. M. Molenbroek, J. K. NÔrskov and I. Stensgaard, Science,
1998, 279, 1913.
3
G. J. Hutchings, J. Catal., 1985, 96, 292.
4120
Phys. Chem. Chem. Phys., 2001, 3, 4113È4121

View Article Online
4
5
6
B. Nkosi, N. J. Coville, G. J. Hutchings, M. D. Adams, J. Friedl
and F. E. Wagner, J. Catal., 1991, 128, 366.
21 D. V. SokolÏskii and A. Ualihanova, Russ. J. Phys. Chem., 1982,
56, 633.
B. Nkosi, M. D. Adams, N. J. Coville and G. J. Hutchings, J.
Catal., 1991, 128, 378.
22 D. A. Cadenhead and N. G. Masse, J. Phys. Chem., 1966, 70,
3558.
23 C. T. H. Stoddart and C. Kemball, J. Colloid Sci., 1956, 11, 532.
24 J. E. Bailie and G. J. Hutchings, Chem. Commun., 1999, 2151.
25 G. J. Hutchings, F. King, I. P. Okoye, M. B. Padley and C. H.
Rochester, J. Catal., 1994, 148, 453.
26 G. J. Hutchings, F. King, I. P. Okoye, M. B. Padley and C. H.
Rochester, J. Catal., 1994, 148, 464.
27 M. B. Padley, C. H. Rochester, G. J. Hutchings and F. King, J.
Catal., 1994, 148, 438.
M. Haruta, S. Tsubota, T. Kobayashi, H. Kageyama, M. J. Genet
and B. Delmon, J. Catal., 1993, 144, 175.
M. Haruta, Catal. T oday, 1997, 36, 153.
T. Hayashi, K. Tanaka and M. Haruta, J. Catal., 1998, 178, 566.
H. Sakurai, A. Ueda, T. Kobayashi and M. Haruta, Chem.
Commun., 1997, 271.
7
8
9
1
1
0
1
A. Couper and D. D. Eley, Discuss. Faraday Soc., 1950, 8, 8172.
A. Couper and D. D. Eley, Proc. R. Soc. L ondon, Ser. A, 1954,
211, 536.
28 P. Claus, T op. Catal., 1998, 5, 51.
1
2
R. J. Mikovsky, M. Boudart and H. S. Taylor, J. Am. Chem. Soc.,
29 K. Atherton, G. Newbold and J. A. Hockey, Discuss. Faraday
Soc., 1971, 52, 33.
1
954, 76, 3814.
1
1
3
4
H. Wise and K. M. Sancier, J. Catal., 1963, 2, 149.
30 J. Y. Lee and J. Schwank, J. Catal., 1986, 102, 207.
31 C. R. Guerra and J. H. Schulman, Surf. Sci., 1967, 7, 229.
32 D. J. C. Yates, J. Colloid Interface Sci., 1969, 29, 194.
33 M. A. Bollinger and M. A. Vannice, Appl. Catal. B, 1996, 8, 417.
34 R. P. Eischens, W. A. Pliskin and M. J. D. Low, J. Catal., 1962, 1,
180.
G. Locco, S. Enzo, G. Fagherazzi, L. Schiffini, I. W. Bassi, G.
Vlaic, S. Galvagno and G. Parravano, J. Phys. Chem., 1979, 83,
2
527.
1
5
6
A. Saito and M. Tanimoto, J. Chem. Soc., Chem. Commun., 1988,
8
32.
1
P. A. Sermon, G. C. Bond and P. B. Wells, J. Chem. Soc.,
35 D. McIntosh and G. A. Ozin, Inorg. Chem., 1977, 16, 51.
36 F. Boccuzzi, S. Tsubota and M. Haruta, J. Electron Spectrosc.
Relat. Phenom., 1993, 64/65, 241.
37 R. G. Greenler, J. Chem. Phys., 1966, 44, 310.
38 R. J. Madon and H. Shaw, Catal. Rev.-Sci. Eng., 1977, 15, 69.
39 J. Oudar, Catal. Rev.-Sci. Eng., 1980, 22, 171.
40 J. E. Bailie and G. J. Hutchings, submitted for publication.
Faraday T rans. 1, 1979, 75, 385.
1
1
7
8
G. C. Bond and P. A. Sermon, Gold Bull., 1973, 6, 102.
A. Saito and M. Tanimoto, J. Chem. Soc., Faraday T rans. 1,
1
988, 84, 4115.
J. Erkelens, C. Kemball and A. K. Galway, T rans. Faraday Soc.,
963, 59, 1181.
R. P. Chambers and M. Boudart, J. Catal., 1966, 5, 517.
1
9
0
1
2
Phys. Chem. Chem. Phys., 2001, 3, 4113È4121
4121