Producing of cinnamyl alcohol from cinnamaldehyde over supported gold nanocatalyst

Article abstract of DOI:10.1016/S1872-2067(20)63678-6

Chemoselective hydrogenation of unsaturated aldehyde to unsaturated alcohol has attracted growing interests in recent years due to its widespread applications in fine chemicals. However, the hydrogenation of the C=O bond was thermodynamically and kinetically unfavorable over the hydrogenation of the C=C bond. Thus, to obtain the unsaturated alcohol from the unsaturated aldehyde is very difficult in most of the catalytic systems. In this work, ZnAl-hydrotalcite-supported cysteine-capped Au₂₅ nanoclusters were used as the precatalysts for chemoselective hydrogenation of cinnamaldehyde to cinnamyl alcohol. The catalyst showed stable high selectivity (~ 95percent) at prolonged reaction time and complete conversion of the substrate. According to the results of the control experiments, the in-situ DRIFTS of the substrate under high pressure of hydrogen and the ²⁷Al MAS-NMR spectroscopy, we proposed that the difference of the preferential adsorption of the C=O bond to that of the C=C bond was derived from the nature of the support of the gold catalysts.

Full text of DOI:10.1016/S1872-2067(20)63678-6

Chinese Journal of Catalysis 42 (2021) 470–481
a v a i l a b l e a t w w w . s c i e n c e d i r e c t . c o m
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / c h n j c
Article
Producing of cinnamyl alcohol from cinnamaldehyde over
supported gold nanocatalyst
Yuan Tana,b, Xiaoyan Liua,*, Leilei Zhanga, Fei Liu a, Aiqin Wang a, Tao Zhanga,c
a CAS Key Laboratory of Science and Technology on Applied Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023,
Liaoning, China
b Zhejiang Normal University, Hangzhou Institute of Advanced Studies, Hangzhou 311231, Zhejiang, China
c University of Chinese Academy of Sciences, Beijing 100049, China
A R T I C L E I N F O
A B S T R A C T
Article history:
Chemoselective hydrogenation of unsaturated aldehyde to unsaturated alcohol has attracted grow‐
ing interests in recent years due to its widespread applications in fine chemicals. However, the hy‐
drogenation of the C=O bond was thermodynamically and kinetically unfavorable over the hydro‐
genation of the C=C bond. Thus, to obtain the unsaturated alcohol from the unsaturated aldehyde is
very difficult in most of the catalytic systems. In this work, ZnAl‐hydrotalcite‐supported cyste‐
ine‐capped Au25 nanoclusters were used as the precatalysts for chemoselective hydrogenation of
cinnamaldehyde to cinnamyl alcohol. The catalyst showed stable high selectivity (~ 95%) at pro‐
longed reaction time and complete conversion of the substrate. According to the results of the con‐
trol experiments, the in‐situ DRIFTS of the substrate under high pressure of hydrogen and the 27Al
MAS‐NMR spectroscopy, we proposed that the difference of the preferential adsorption of the C=O
bond to that of the C=C bond was derived from the nature of the support of the gold catalysts.
© 2021, Dalian Institute of Chemical Physics, Chinese Academy of Sciences.
Received 19 May 2020
Accepted 23 June 2020
Published 5 March 2021
Keywords:
Gold nanocluster
Support effect
Cinnamyl alcohol
Selective hydrogenation
Cinnamaldehyde
Published by Elsevier B.V. All rights reserved.
1. Introduction
search for green synthesis routes for synthesis of unsaturated
alcohols.
The selective hydrogenation of α,β‐unsaturated ketones and
aldehydes to the corresponding unsaturated alcohols was
among one of the most important industrial reactions [1,2]. The
unsaturated alcohols were considered as the key intermediates
in the synthesis of fine chemicals, fragrances, pharmaceuticals,
agrichemicals and cosmetics [3]. In traditional industrial pro‐
cess, the selective reductions of unsaturated ketones and alde‐
hydes could be achieved by using stoichiometric reducing agent
such as the sodium borohydride [2,4]. However, this method
involved harmful inorganic reagent even with high selectivity,
which was not environmentally friendly. Thus, it is desirable to
Nevertheless, it was difficult to obtain the unsaturated al‐
cohols from unsaturated ketones and aldehydes with high se‐
lectivity, due to the competition between the unsaturated
groups reacting with the hydrogen molecule. Cinnamaldehyde
(CAL) is a representative substrate containing the carbonyl and
vinyl groups simultaneously. As illustrated in Scheme 1, the
partial hydrogenation of CAL led to the formation of cinnamyl
alcohol (COL) and hydrocinnamaldehyde (HCAL), when the
C=O and the C=C groups were hydrogenated, respectively. Ac‐
tually, it was one of the most challenging subjects to obtain the
COL, since the hydrogenation of the C=O bond was thermody‐
* Corresponding author. Tel: +86‐411‐84379416; Fax: +86‐411‐84691570; E‐mail: xyliu2003@dicp.ac.cn
This work was supported by the National Natural Science Foundation of China (21776271, 21606227), the National Key Research & Development
Program of China (2016YFA0202801), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB17020100), Zhejiang Pro‐
vincial Natural Science Foundation of China (LQ20B030005), and CAS Interdisciplinary Innovation Team (BK2018001).
DOI: 10.1016/S1872‐2067(20)63678‐6 | http://www.sciencedirect.com/science/journal/18722067 | Chin. J. Catal., Vol. 42, No. 3, March 2021

Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
2.1. Chemicals
471
All chemicals were used as received and without any
purification. All glassware were washed with Aqua Regia
(VHCl:VHNO3 = 3:1) and rinsed with ethanol and ultrapure water.
Ultrapure water (18.2 MΩ) was used throughout this work.
Hydrogen tetrachloroaurate (III) hydrate (HAuCl4·3H2O)
was from Sinopharm Chemical Reagent (Beijing Co. Ltd.).
Cysteine (Cys) was purchased from Sigma‐Aldrich. Zinc nitrate
hydrate (Zn(NO3)2·6H2O), Magnesium nitrate hydrate
(Mg(NO3)2·6H2O), Nickel nitrate hydrate (Ni(NO3)2·6H2O),
Aluminium nitrate hydrate (Al(NO3)3·9H2O), were purchased
from Tianjin Damao Chemical Reagent Factory. Sodium
borohydride (NaBH4), sodium carbonate, sodium hydroxide
(NaOH), were purchased from Tianjin Tianli Chemical Reagent
Scheme 1. Reaction pathways of hydrogenation of cinnamaldehyde.
namically and kinetically unfavorable over the hydrogenation
of the C=C bond [1,2,5]. On the traditional precious metal cata‐
lysts (e.g., Pt, Pd, Ru, Rh etc.), producing HCAL and hydrocin‐
namyl alcohol (HCOL) were always easier than COL due to their
strong dissociation ability of the hydrogen and relatively easier
activation of the C=C bond [6–9].
Ltd.
The
cinnamaldehyde,
cinnamyl
alcohol,
hydrocinnamaldehyde and hydrocinnamyl alcohol used for
reaction and preparation of standard line were purchased from
Alfa Aesar. The solvent of isopropanol, ethyl alcohol, toluene,
ethyl acetate and tetrahydrofuran were from Kermel Chemical
Reagent Ltd.
Supported gold nanoparticles (NPs) have attracted intensive
attentions recently because of their high selectivity in various
redox reactions under moderate reaction conditions [10–12].
In previous works, the gold catalysts exhibited potential high
selectivity in the selective hydrogenation of unsaturated ke‐
tones and aldehydes, which is comparatively superior to other
supported noble metal catalysts [13,14]. The performances of
the gold catalysts were largely affected by the steric and elec‐
tronic effects [15–17] caused by the size and morphology of
gold particles [18–20] and the addictive elements [21–23].
What’s more, the activation of the C=C bond and the C=O bond
is also competitive over most of the supported gold catalysts
[19,24]. Thus, to get a stable high selectivity of unsaturated
alcohol on supported gold catalysts especially in high conver‐
sions confront with enormous challenges.
2.2. Preparation of the catalysts
The cysteine capped NCs were prepared according to our
previous work [25,26]. In a typical synthesis, aqueous solution
of HAuCl4 (110 mM, 5.00 mL) and Cysteine solution (5.5 mM,
150 mL) were successively added into 200 mL of ultrapure
water under gentle stirring (500 rpm). Then 30 mL of 1 M
NaOH solution was rapidly introduced into the dispersion
solution. After 10 min of stirring, five times excessive sodium
borohydride was quickly added to the above solution, and the
reaction was allowed to proceed under vigorous stirring for 3
h. After the addition of NaBH4, an obvious color change from
reddish brown to black brown was visible. When the reaction
was finished, the products were washed thoroughly with
ethanol‐water (V/V = 3:1) and separated by centrifugation.
Then the obtained Au25 nanoclusters were dissolved in water
and defined by the UV‐visible spectrum. The successful
synthesis of Au25 nanocluster was confirmed by the UV‐vis
spectrum, which showed four fingerprint peaks positioned at
440, 545, 670, and 780 nm, respectively (Fig. 1).
The ZnAl‐HT (Zn/Al atomic ratio = 3) support was prepared
as follows: Solution A was obtained by adding Zn(NO3)2·6H2O
(0.21 mol) and Al(NO3)3·9H2O (0.07 mol) to 200 g of ultrapure
water. Solution B was prepared by adding NaOH (0.438 mol)
and Na2CO3 (0.113 mol) to 200 g of ultrapure water. After the
dissolution of the powders, the solution A was slowly pumped
into the solution B (velocity: 3 mL/min) under constant stirring
in 70 °C water bath. And then additional NaOH solution was
added to maintain the pH value of the suspension at 10. The gel
aged at this temperature for 24 h. Then the mixture was filtered
and washed until the pH value of the filtrate became neutral.
The precipitate was dried in an oven at 80 °C overnight to
obtain the ZnAl‐HT. The MgAl‐HT and NiAl‐HT supports were
conducted by the same process, with the atomic ratio of
Mg(Ni)/Al of 3.
Previously, we designed
a gold catalyst by utilizing
ZnAl‐hydrotalcite (HT) supported Au25 nanoclusters (NCs) as
the precatalyst for chemoselective hydrogenation of nitrosty‐
rene to aminostryene [25]. Notably, it showed high selectivity
for hydrogenation of –NO2 group, while displaying almost inert
activity for C=C group. Considering the commonness, herein,
we applied this catalyst (Au25/ZnAl‐300) to the selective hy‐
drogenation of CAL to COL, which is much more competitive
between the –C=O and the –C=C than the –NO2 and the –C=C.
We found a stable high selectivity produced on this catalyst:
even in the prolonged reaction time, the selectivity of COL
could be preserved at above 95% in complete conversion of
CAL. This result was unprecedented through the previously
developed supported gold catalysts. The reaction route was
investigated based on the dynamic test, control experiment
(e.g., using styrene and the mixture of styrene and benzalde‐
hyde as the substrate, changing the solvent, etc.) and in‐situ
DRIFTS of CAL adsorption and reaction under high pressure of
hydrogen.
2. Experimental

472
Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
test. After the reaction, the products were condensed and
analyzed by GC‐MS (Varian, 450 GC, 320 MS) and GC (Agilent
7890 A) equipped with a FFAP capillary column. The control
experiments were conducted in the same stainless steel
autoclave under the similar reaction conditions. The substrates
were replaced by the styrene and/or the artificial mixture of
benzaldehyde and styrene.
0.6
0.5
0.4
0.3
0.2
0.1
0.0
440 nm
545 nm
2.4. Characterization
670 nm
780 nm
The actual Au loadings were measured with an inductively
coupled plasma atomic emission spectroscopy (ICP‐AES) on an
IRIS Intrepid II XSP instrument (Thermo Electron Corporation).
UV‐visible spectrum was recorded on a Cintra (GBC) apparatus
with water as a reference at room temperature. A continuous
mode was used in the wavelength range from 200 to 900 nm at
a scanning speed of 100 nm/min. The HAADF‐STEM was
recorded on a JEOL JEM‐2100F microscope equipped with
STEM dark‐field (DF) and Oxford detectors at 200 kV. The
signal of pure liquid cinnamyl aldehyde was recorded in the
Bruker Vertex 70V spectrometer using the attenuated total
reflection infrared (ATR‐IR). The spectroscopy was equipped
with a DLaTGS detector. The background spectrum was
recorded at room temperature and atmospheric pressure. The
liquid cinnamyl aldehyde was dropwise added onto the surface
of the diamond crystal and the signal of the substrate was
recorded when the liquid was evaporated. The in‐situ DRIFTS
of the cinnamyl aldehyde were recorded in the Bruker Vertex
70V spectrometer, equipped with a MCT detector in the range
of 400–4000 cm−1 with a resolution of 4 cm−1. The catalyst was
packed into the sample cell with the ZnSe window. Before the
test, the catalyst was pretreated at 120 °C for 30 min to remove
the absorbed water. The background spectrum was recorded
under atmospheric pressure in helium, after being cooled to
room temperature. Then, with the protection of helium, ~5 μL
of the cinnamyl aldehyde was introduced into the sample for
adsorption. Subsequently, 1.5 MPa of the hydrogen was
introduced into the cell using the pressure valve. After that, the
reaction temperature increased to the fixed temperature with
the heating rate of 10 °C/min. The spectra of the cinnamyl
aldehyde hydrogenation process were collected at 25–180 °C.
For every fixed temperature, the spectrum was collected
successively for twenty minutes to get the steady state. The 27Al
solid‐state magic‐angle spinning (MAS) nuclear magnetic
resonance (NMR) experiments were performed on a Bruker
Avance III 600 spectrometer equipped with a 14.1 T wide‐bore
magnet. The resonance frequencies in this field strength was
156.4 MHz for 27Al. The 27Al spectra was recorded use a single
pulse consequence with spinning rate of 10 kHz. Chemical
shifts of 27Al was referenced to 1.0 M aqueous Al(NO3)3 solution
at 0 ppm.
400
500
600
700
800
900
Wavelength (nm)
Fig. 1. UV‐visible spectrum of cysteine capped Au25 nanocluster.
The HTs supported Au25 NCs was prepared by simple
impregnation. In a general synthesis, 30 mg of the Au25
nanoclusters was dispersed into 10 mL of ultrapure water.
Then 2.00 g of the Zn(Mg/Ni)Al‐HTs supports was added into
the above suspension under vigorous stirring. After 1 h, the
product was washed with ultrapure water and collected by
centrifugation. The residue was then dried in an oven at 80 °C
for 12 h and calcined in muffle at 300 °C for 2 h with the
heating rate of 5 °C/min. The obtained sample was defined as:
Au25/ZnAl‐300, Au25/MgAl‐300 and Au25/NiAl‐300, with the
loadings of Au were 1.1 wt%, 1.0 wt%, and 1.0 wt%,
respectively.
The ZnAl‐HT supported gold nanoparticles (NPs) were
prepared by deposition‐precipitation (DP) method. In a typical
synthesis, 5.2 mL of the aqueous solution of HAuCl4 (19.12
gAu/L) was added into the suspension solution under vigorous
stirring. Then 2.00 g ZnAl‐HT powders were introduced. After
that, 1 M NaOH solution was used to adjust the pH value to 10.
The reaction was conducted at room temperature for 12 hours.
Then the solid was obtained after filtrating, washing and
drying. Before the catalytic test, the solids were calcined at 300
°C for 2 h, with the heating rate of 5 °C/min. The final loadings
of Au were determined by the inductively coupled plasma
atomic emission spectroscopy (ICP‐AES), it was 1.0%. The
obtained sample was defined as: 1.0% Au/ZnAl‐300 (DP).
2.3. Catalytic performance evaluation
The liquid‐phase hydrogenation of cinnamaldehyde to
cinnamyl alcohol was carried out in a stainless steel autoclave
equipped with a pressure gauge and magnetic stirring. Before
reaction, a mixture of cinnamaldehyde (0.05 M), n‐decane
(0.025 M), solvent (5 mL) was put into the vessel. Then, certain
amounts of catalysts were introduced into the autoclave. After
being sealed, the autoclave was flushed with hydrogen for six
times and then pressurized at 1.5 MPa. To initiate the reaction,
the reactor was heated to 130 °C in an oil bath without stirring
until the temperature reached to the specified value. That is,
after the temperature reached to 130 °C, the stirring was
started. The stirring rate was kept at 30 rpm till the end of the
3. Results and discussion
3.1. Catalytic performance
To evaluate the catalytic performances of the supported

Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
473
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.4
1.2
1.0
0.8
0.6
0.4
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
350 360 370 380 390 400 410 420 430 440
H2 Pressure (MPa)
Temrepature (K)
Fig. 2. The reaction rates changed as a function of the temperatures (a) and the pressures of H2 (b).
Au25 nanoclusters catalysts for selective hydrogenation of CAL
to COL, the influence of reaction temperature and H2‐pressures
were firstly studied. As displayed in Fig. 2, the reaction rates
increased with the enhancement of the temperature and the
hydrogen partial pressure. That meant high temperature and
high hydrogen pressure were beneficial for this reaction.
Except for the above two factors, the solvent effect on the
catalytic performances was also investigated. As seen from the
results in Fig. 3, the reactions conducted in six different sol‐
vents rendered to different activities and selectivities under the
same conditions. The best result was displayed in isopropanol
and ethanol. When the two substances were used as the sol‐
vent, the selectivity of COL came close to 95% at relatively
higher conversions. While the reactions were employed in oth‐
er solvents such as acetate and toluene, the selectivities were
far below that in isopropanol and ethanol. Previous analysis of
the impact of solvent suggested that both the isopropanol and
ethanol were good solvents and hydrogen donors to produce
the hydrogen in the reaction process [27,28]. That seems to be
good for the selective hydrogenation reaction.
catalysts were conducted under the same conditions. The
results were displayed in Table 1. Obviously, when using the
ZnAl‐HT supported Au25 NCs as the precatalyst for
chemoselective hydrogenation of CAL to COL, it exhibited the
best performance. At this moment, the reaction conditions
were performed at 130 °C and 15 atm H2. The conversion and
selectivity on the Au25/ZnAl‐300 catalyst were 98.3% and
95.4%, respectively. In comparison, the Au25/MgAl‐300 and
Au25/NiAl‐300 catalysts were synthesized with the same
method as that of the Au25/ZnAl‐300. Both the catalysts
displayed different catalytic performances, as shown in Table 1,
entries 2 and 3. Over the Au25/MgAl‐300 catalyst, the
conversion of CAL and the selectivity of COL were 99.2% and
39.5%, respectively. Meanwhile, 13.6% of HCAL and 38.4% of
HCOL were produced as the by‐products. Over the
Au25/NiAl‐300 catalysts, the results were entirely different. The
CAL could not be converted to the COL. Instead, 86.7% of the
HCOL was detected as the only product by chromatography
analysis, which might be caused by the easy activation of
hydrogen over the Ni‐based supports [29].
To compare the catalytic performances of supported gold
catalysts, different experiments over the supported gold
So as to examine the effect of the gold precursor on the
catalytic performance, the gold NPs synthesized by the (deposi‐
100
80
60
40
20
0
100
80
60
40
20
0
Table 1
Catalytic results for the hydrogenation of CAL over the gold catalysts in
this work.
Sel. (%)
COL HCAL HCOL
Size
(nm)
Conv.
(%)
Entry
Catalyst
1
2
1.1%Au25/ZnAl‐300
1.0%Au25/MgAl‐300
1.0%Au25/NiAl‐300
1.7 b 98.3 95.4
2.2 b
3.2 b 100
2.6 b
3.3
1.8
4.1
1.4
—
0
4.5
99.2 39.5 13.6 38.4
3
4
0
0
86.7
3.6
1.0%Au/ZnAl‐300(DP) a
61.5 85.9 3.7
75.7 50.0 17.3 16.8
c
5
6
7
8
9
10
1.0%Au/ZrO2
0.8%Au/Fe2O3
1.5%Au/TiO2
1.1%Au25/ZnAl‐HT
ZnAl‐300
1.1%Au25/ZnAl‐300e
c
93.0 45.4 4.9
51.9 14.6
3.6
0
0
0
0
d
0
26.8 43.6 1.6
9.6 23.9 3.1
1.7
28.4 35.2
0
Reactions conditions: 130 °C, H2 pressure 15 atm, cinnamaldehyde 0.25
mmol, solvent 5 mL isopropanol, 5 h, catalyst 50 mg (Au 1.0 mol%).
a
Catalysts were prepared by deposition‐precipitation method; b The
particle sizes were the same as those in Ref. [26]; c Catalysts were pro‐
vided by Haruta Gold Incorporated Company; d Catalysts were provided
by World Gold Council (WGC); e H2 was replaced by the N2.
Fig. 3. The catalytic performances over the 1.0% Au25/ZnAl‐300 cata‐
lyst with different solvents. Reaction conditions: cinnamaldehyde 0.25
mmol, catalyst 50 mg (Au 1.0 mol%), H2 pressure 15 atm, 130 °C, 5 h.

474
Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
tion‐precipitation) DP method supported on the ZnAl‐HT were
prepared to make a comparison. The catalyst was denoted as
the Au/ZnAl‐300(DP). As shown in Table 1 entry 4, the
performance of the Au/ZnAl‐300(DP) catalyst was inferior to
that of the Au25/ZnAl‐300 catalyst, with the conversion and
selectivity decreased from 98.3% to 61.5% and from 95.4% to
85.9%, respectively. Furthermore, the Au/ZrO2, Au/Fe2O3 and
Au/TiO2 catalysts purchased from Haruta Gold Incorporated
Company and World Gold Council were also applied in this
reaction for comparison. To be noted, these results were also
not as good as that on the Au25/ZnAl‐300 catalyst. As shown in
Table 1, entry 5–7, the conversions of CAL under the same
reaction conditions over the Au/ZrO2, Au/Fe2O3 and Au/TiO2
catalysts were 75.7%, 93% and 51.9%, respectively. The
corresponding selectivity of COL was 50.0%, 45.4% and 14.6%,
respectively, which fell far below that of our Au25/ZnAl‐300
catalyst.
Fig. 4. Time courses of the yield of cinnamaldehyde and cinnamyl alco‐
hol over the Au25/ZnAl‐300 catalyst. Reactions conditions: 130 °C, H2 15
atm, isopropanol as the solvent, cinnamaldehyde 0.25 mmol.
The above results suggested that different supports and the
gold precursor would lead to distinctive performances. Thus,
the ZnAl‐HT supported Au25 NCs as the precatalyst was active
and selective for producing of COL from CAL. However, the
uncalcined Au25/ZnAl‐HT sample was much less active than the
heat‐treated sample (Table 1, entry 8), in which the conversion
of CAL was five times lower than that on the Au25/ZnAl‐300
catalyst. The reason might be attributed to the inhibition of
substrates by the thiol ligands on the surface [25]. In fact, the
thiol ligands were considered as a double‐edged sword. On one
hand, the thiol ligands maintained the structural integrity of
Au25 nanocluster, on the other hand, it disabled the gold sites
from adsorbing the reactants. Thus, the intact catalyst was not
active for hydrogenation of CAL. To remove the partial thiol
ligands through the heat treatment, the active sites were ex‐
posed and the reactants could be adsorbed and activated. Be‐
sides, previous analysis indicated that the isopropanol solvent
might produce the hydrogen in the reaction process [27], which
could contribute for the hydrogenation reaction. So, the ex‐
periment was conducted on blank ZnAl‐HT, which calcined at
300 °C before the catalytic test (Table 1, entry 9). It can be
found that the conversion of CAL on ZnAl‐300 support was only
9.6%, which was much lower than that on the Au25/ZnAl‐300
catalyst. It implied that the gold particles on the ZnAl‐300 sup‐
port played the major role in dissociating the hydrogen. What’s
more, as the H2 was replaced by the N2, the conversion of CAL
was just 28.4%, which further proved the source of hydrogen
was derived from the dissociated hydrogen species stem from
gold.
by‐product and HCAL could hardly be detected. It was worth
noting that the concentration of COL could be maintained at a
high level in complete transformation of CAL. Even increasing
the reaction time from 5 to 15 h, the yields of the by‐products
remained below 5% and the selectivity of COL preserved above
95%. That is, the unsaturated alcohol can be obtained at high
and stable selectivity on the Au25/ZnAl‐300 catalyst. This phe‐
nomenon was quite different with the previous work. Because
it is widely accepted that the C=C bond and the C=O bond are
competitively activated on the catalyst [2,5,19]. The stable high
selectivity at high conversion was also superior to the previ‐
ously reported gold catalysts [19,30–38], which being com‐
pared in Table 2.
3.2. Discussion about the high selectivity over the
Au25/ZnAl‐300
According to the previous literature, the catalytic perfor‐
mances of supported gold catalysts would be strongly depend‐
ent on the particle size of gold and the nature of the supports
[39,40]. In the present work, based on the above results (Table
1), the Au25/ZnAl‐300 catalyst exhibited better activity than the
Au/ZnAl‐300(DP) catalyst, which suggested the small gold par‐
ticles would be beneficial for obtaining the COL from CAL with
high conversions, since the supports of both catalysts were the
same ZnAl‐300. And it can be reasonably concluded that with
the decrease in the particle size, the low‐coordinated sites such
as edges and corners of gold will increase [40]. Thus, it provid‐
ed more active sites for dissociation of the hydrogen. However,
the selectivity of COL in these catalytic systems seems to be
largely dependent on the nature of the supports. Because the
catalytic performance over different supported gold catalysts
were within 5 nm, and the chemical valence of gold on the hy‐
drotalcites supported Au25 NCs didn’t show much differences
according to our previous work [26]. Thus, the distinctive cata‐
lytic performances might be strongly affected by the nature of
the support, which could influence the interfacial structure by
altering the interaction between gold and the supports or pro‐
vide the different adsorption sites for the substrates.
To study the products distribution of the reaction, we con‐
ducted the dynamic test of CAL to COL as a function of time
over the Au25/ZnAl‐300 catalyst. As seen from the Fig. 4, ini‐
tially, the desired product (COL) behaved a linear increase with
the sacrifice of CAL, which implied a first order reaction hap‐
pened at this stage. After about one hour, the transformation
rate slowed down as the product concentration increased
sharply. It meant the conversion of CAL might be hindered by
the high concentration of COL. Then, about five hours later, the
CAL was almost converted completely, and the selectivity of
COL attained 95.4%. At this time, little HCOL appeared as the

Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
475
Table 2
Catalytic results reported recently for hydrogenation of CAL to COL over the supported gold catalysts.
Reaction conditions
Catalyst
Conv. (%)
Sel. (%)
Ref.
Solvent
Au (mol%)
1.0
Tem. (°C)
130
PH2 (MPa)
t (h)
5
15
0.8
5
98.3
100
100
96
48
49
46
94
78
20
95.4
95.7
~92
40
Au25/ZnAl‐300
isopropanol
1.5
This work
Au/ZnO‐CP
Au/MgAlO
Au/meso‐CeO2
Au25/Fe2O3
Au25/TiO2
Au/DMF
Au/Mg2AlO
Au/5FeAl
Au/HDAE
Au/TiO2
isopropanol
ethanol
H2O/ethanol
1.0
2.4
1.0
110
120
100
2
1
1
[19]
[30]
[31]
0.5
97
100
100
94
85
67
88
47
91
89
toluene/ethanol
5.1
0
0.1
3
[32]
amide
ethanol
1.0
0.2
0.7
0.6
6.3
60
120
4
1
56
2
[33]
[34]
isopropanol
100
1
3
[35]
28
50
50
94
ethanol
ethanol
ethanol
60
60
100
0.1
0.1
8.5
—
—
2.8
[36]
[37]
[38]
Au/FeOOH
Au/Al2O3
14.8
0.1
To understand the good performance of the Au25/ZnAl‐300
catalyst and elaborate the specific property for obtaining the
desired product (COL) from CAL, the control experiments were
implemented under the same conditions on different gold
catalysts. Among them, the styrene and/or the artificial mixture
of styrene and benzaldehyde were used as the substrates. It has
been reported that the acidity or basicity of the support could
influence the catalytic performance for the selective hydro‐
genation reactions. Here, the gold catalysts with supports of
different acidity or basicity were chosen. The Au25/MgAl‐300
catalyst with strong basicity [26] and the Au/ZrO2 catalyst with
strong acidity [41] were selected to make a comparison. The
results were showed in Table 3. Over the Au25/ZnAl‐300
catalyst, only 3.8%–5.1% of the styrene could be converted
with or without the presence of benzaldehyde, while 77.7% of
the benzaldehyde could be transformed to the benzyl alcohol
with the presence of the styrene. It indicated that the
Au25/ZnAl‐300 catalyst was almost inert for the hydrogenation
of the vinyl group but active for the aldehyde group, which led
to the high selectivity to the benzyl alcohol. However, over the
Au25/MgAl‐300 catalyst, the conversion of styrene was 67.5%
when the styrene was used as the only substrate. Whereas with
the existence of benzaldehyde, the conversion of styrene
decreased to 58.2%. At the same time, the conversion of
benzaldehyde was 95.1%. The results suggested that the
aldehyde group and the vinyl group were competitively
hydrogenated on the Au25/MgAl‐300, and the C=O bond were
preferentially adsorbed on the catalyst instead of the C=C bond.
Over the Au/ZrO2 catalyst, the situation was different. When
the styrene and the benzaldehyde presented at the same time,
both of the two substrates could be converted to the
corresponding hydrogenated products with high conversion.
Furthermore, the conversion of styrene was even higher than
that of the single substrate presented. That meant the presence
of aldehyde group might promote the hydrogenation of vinyl
group on the Au/ZrO2 catalyst. From the results, the
Au25/ZnAl‐300 catalyst with moderate basicity and acidity
[26,42] showed the highest selectivity, which indicated the
typical basic or acidic supports might be detrimental to the
selectivity for the hydrogenation of CAL to COL in this work.
In order to further study the effect of the support on the ad‐
sorption and reaction of the substrates, the in‐situ DRIFTS of
CAL was measured. As indicated in Fig. 5, the substrate could
be adsorbed on the surface of the three different catalysts,
which could be evidenced by the shift (18–22 cm–1) with that of
the liquid substrate (1670 cm–1). The contributions of the IR
bands were listed in Table 4. Notably, when high pressure hy‐
drogen (1.5 MPa) was introduced, the appearance of the peak
at 3666 cm–1 over the Au25/NiAl‐300 catalyst suggested that
the –OH group was formed (Fig. 6(a)) [43–45]. Meanwhile, at
the temperature higher than 90 °C, the absorption band at‐
tributed to the –CHO (1688 and 2742 cm–1) [43,46–51] and the
–C=C (1626 cm–1) [44] disappeared. Concomitantly, the bands
due to the –CH2 (2931 and 2858 cm–1) [43,44,48] appeared,
which displayed the CAL was transformed to the HCOL. The
intensity of the absorption bands of the –OH and –CH2 remain
high when the temperatures ranged from 120 to 150 °C, which
reflected the Au25/NiAl‐300 catalyst was not selective for this
reaction.
Table 3
Control experiments with styrene and benzaldehyde as the substrates
over Au catalysts.
Conv. (%)
Styrene a Styrene b Benzaldehyde b
Size
(nm)
Catalyst
1.1% Au25/ZnAl‐300
1.0% Au25/MgAl‐300
1.0% Au/ZrO2
1.7
2.2
3.1
3.8
67.5
71.4
5.1
58.2
97.2
77.7
95.1
100
Reactions conditions: 130 °C, H2 pressure 15 atm, cinnamaldehyde 0.25
mmol, solvent 5 mL isopropanol, 5 h, catalyst 50 mg (Au 1.0 mol%).
a Substrate 0.25 mmol styrene; b Substrate 0.25 mmol styrene and 0.25
mmol benzaldehyde.
Over the Au25/MgAl‐300 catalyst, similar phenomena were
happened: the intensity of the vibration band ascribed to the

476
Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
completely different from the other two catalysts: all of the
1670
1623
(a) 0.5
peaks ascribed to the C=O bond [47–51], the C=C bond [44] and
the –CH asymmetric and symmetrical stretching vibration
[43,46,48] in aldehyde decreased simultaneously with the in‐
crease of the temperature (Fig. 6(c)), while there were no new
2813
3062
3028
2742
3082
1688
(b)
0.1
–1
signal appeared at 3666, 2928 or 2856 cm , such results sug‐
1626
1627
gested that the adsorption of COL or other intermediate prod‐
ucts were very weak and they were probably easily desorbed
from the surface of the Au25/ZnAl‐300 catalyst. That might be
the reason of high selectivity of cinnamyl alcohol on the
Au25/ZnAl‐300 catalyst.
1692
1689
0.05
(c)
(d)
To further clarify the high selectivity of COL from CAL over
the Au25/ZnAl‐300 catalyst, the 27Al solid‐state magic‐angle
spinning (MAS) nuclear magnetic resonance (NMR)
experiments were conducted to provide some evidences. In the
27Al MAS‐NMR spectroscopy (Fig. 7), there were two
characteristic peaks in ZnAl‐HT and MgAl‐HT supported Au25
nanoclusters catalysts at around 10 and 72 parts per million
(ppm), which represented Al3+ ions in octahedral (AlO) and
tetrahedral (AlT) coordination, respectively [55]. Besides, the
27Al MAS‐NMR peak at around 47 ppm chemical shift, which
could be assigned to Al3+ ions in pentahedral coordination (AlP)
[56], only appeared in the Au25/ZnAl‐300 catalyst rather in
Au25/MgAl‐300 and Au25/ZnAl‐200 catalysts. It suggested that
the coordinately unsaturated pentacoordinate Al3+ species only
existed in Au25/ZnAl‐300, that might supply the oxygen
vacancies for the preferential adsorption of the C=O bond.
Thus, the Au25/ZnAl‐300 catalyst exhibited the comparable
higher selectivity of COL than the Au25/MgAl‐300 and other
supported gold catalysts. The results of oxygen vacancy created
by ZnAl‐layered hydrotalcite were also agree with the previous
literatures [57,58].
0.1
1627
2992
4000
3500
3000
2500
2000
1500
1000
Wavenumbers (cm)
Fig. 5. ATR‐IR spectra of the liquid cinnamyl aldehyde (a) and in‐situ
DRIFTS of the adsorbed CAL over the catalysts: (b) Au25/NiAl‐300; (c)
Au25/MgAl‐300; (d) Au25/ZnAl‐300 at 25 °C.
–CH2 [43,44,48] increased with the decrease of that of the C=C
(1628 cm–1) [44] (Fig. 6(b)). But it appeared at higher temper‐
ature than that over the Au25/NiAl‐300 (120 vs. 150 °C). Be‐
sides, there were no free –OH detected on the Au25/MgAl‐300
catalyst as evidenced by the absence of the absorption band at
3666 cm–1, although the vibrational band of C=O decreased
with the rising temperature. This result indicated that the
Au25/MgAl‐300 catalyst was less active than the Au25/NiAl‐300
catalyst.
However, the situation over the Au25/ZnAl‐300 catalyst was
Table 4
The attribution of peaks in FTIR spectra of the hydrogenation process of CAL over the Au25/NiAl‐300 catalyst.
Entry
1
Peak position (cm–1)
3666
Group
–OH
Intensity
strong
Vibration type
Free –OH
Ref.
[43–45]
3082
3061
2
=CH–
weak
‐CH stretching vibration in aromatic or olefin
[43,49]
3
3025
weak
‐CH stretching vibration in benzene ring
[43,46]
2928
2856
‐CH asymmetric and symmetrical stretching vibration in
saturated alkane
4
5
–CH2
strong
strong
[43,44,48]
[43,46,48]
2812
2741
‐CH asymmetric and symmetrical stretching vibration in
aldehyde
–CHO
6
7
1688
C=O
C=C
strong
medium
C=O stretching vibration in unsaturated aldehyde
C=C stretching vibration in conjugate alkene
[47–51]
[44]
1626
1604
1576
1494
8
weak
skeleton stretching vibration in benzene ring
[52,53]
1452
1392
‐CH asymmetric and symmetrical bending vibration in
olefin
9
=CH‐
=CH‐
weak
weak
[53]
[53]
1293
1250
10
‐CH in‐plane bending vibration in olefin
11
12
1072
976
weak
weak
‐CH in‐plane bending vibration in benzene ring
‐CH out‐of‐plane bending vibration in olefin
[43,46]
[52,54]
=CH‐

Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
477
comparison with the activation of the C=O bond. Besides, a
unique atop adsorption mode of C=O group were favored on
silver atoms. Thus, it led to a high selectivity towards
unsaturated alcohol over the Ag catalyst. The situation on Au
catalyst was different, whose performance was highly
dependent on the nature of the supports [16,39]. As illustrated
in Yang’s work, the gold particles loaded on silica exhibited
poor selectivity to unsaturated alcohol. But on some reductive
supports, the gold catalysts displayed an enhanced selectivity,
such as the Au/TiO2 [17,36], Au/CeO2 [31] and Au/ZnO [19,22]
catalysts. That is, the structure of gold could be modified by
interacting with the support. Thus, a high density of charge on
gold surface might decrease the binding energy of the C=C bond
by increasing the repulsive four‐electron interaction [59,60].
For our Au25/ZnAl‐300 catalyst, the gold nanoclusters had a
strong interaction with the ZnAl‐HT supports by epitaxial
interaction and the Au‐S‐Zn bond at the interfaces, as we
proved previously [25]. Then, the geometry and electronic
structure of the catalyst could be regulated and optimized for
this reaction. What’s more, the Au25/ZnAl‐300 catalyst with
moderate basicity [26] could get an increased selectivity of
COL, which might be caused by the positive effect of bases
towards the activation of the C=O bond by means of their lone
electron pair interacted with the metal surface [2]. Thus, the
hydrogenation of the C=C bond was hindered as the
hydrogenation of C=O bond was preferentially favored.
To evaluate the active sites over the Au25/ZnAl‐300 catalyst,
the geometric construction and electronic structure of gold
particles as well as the interaction between gold and the
support were very important. Based on our previous work [25],
the thiol ligands on the surface of Au NCs were removed after
calcination at 300 °C. The gold atoms on the surface were
reduced to metallic state. In this work, we found the uncalcined
sample was much less active than the heat‐treated sample
(Table 1, entry 8), in which the conversion of CAL was five
times lower than that on the Au25/ZnAl‐300 catalyst. The
reason might be attributed to the inhibition of substrates by the
thiol ligands on the surface. When the experiment conducted
on blank ZnAl‐HT support (Table 1, entry 9), the conversion of
CAL was much lower than that on the Au25/ZnAl‐300 catalyst,
1688
(a)
1626
1604
1576
2742
2813
25 o
C
1250
1293
1392
1452
1494
1072 976
60oC
90 oC
2928
2856
120 o
150 o
C
3666
C
4000
3750
3500
3250
3000
2750
2500
2250
2000
1750
1500
1250
1000
Wavenumber (cm
2740
2812
1691
1628
(b)
25 o
C
60 o
C
C
90 o
120 o
150 o
180 o
C
C
2858
2931
C
4000
3750
3500
3250
3000
2750
2500
2250
2000
1750
1500
1250
1000
Wavenumber (cm

1689
1628
(c)
2742
2814
25 o
100 o
110 o
120 o
130 o
C
C
C
C
C
140 o
C
4000 3750 3500 3250 3000 2750 2500 2250 2000 1750 1500 1250 1000
Wavenumber (cm1)
Fig. 6. In‐situ DRIFTS of the hydrogenation process of CAL with in‐
creasing temperature after introducing 1.5 MPa of hydrogen over the
catalysts of Au25/NiAl‐300 (a), Au25/MgAl‐300 (b), and Au25/ZnAl‐300
(c).
3.3. The active sites and reaction mechanism
For the hydrogenation of α,β‐unsaturated ketones and
aldehydes, it is widely accepted that the reaction pathway
occurs via the Horiuti‐Polyani mechanism through the
adsorbed species of di‐σC=O η2, di‐σC=C η2 and di‐π η2 (η4) [1,2].
The key factor that governs the selectivity was generally
considered to be the chemisorption modes of the substrate. For
example, the catalysts with poor capability for adsorption of
the di‐σC=C η2 and/or di‐π η2 (η4) would result in the high
selectivity of the unsaturated alcohol [1,5]. Previously, Yang et
al [24] conducted the theoretical calculations on the adsorption
of crotonaldehyde on the Pt19, Au19, and Ag19 clusters. They
found that the di‐σC=C η2 and di‐π η2 (η4) adsorption modes
coexisted on the Pt surface. Thus, both of the C=C bond and the
C=O bond were activated on Pt surface simultaneously.
Therefore, the selectivity of crotonyl alcohol on Pt catalyst was
very poor. On the contrary, the C=C activation was very weak
on Ag catalysts because of the low adsorption energy in
AlT
AlO
Au25/ZnAl-HT
Au25/ZnAl-200
10.1
72.3
46.8
Au25/ZnAl-300
Au25/MgAl-300
-100
-50
0
50
ppm
100
150
200
Fig. 7. The 27Al MAS‐NMR spectra of supported Au25 nanoclusters cata‐
lysts.

478
Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
Scheme 2. The proposed mechanism of selective hydrogenation of CAL
to COL over the Au25/ZnAl‐300 catalyst.
Fig. 9. The HAADF‐STEM images of Au25/ZnAl‐300 catalyst before and
after using three times in hydrogenation of cinnamaldehyde. Reaction
conditions: 130 °C, H2 pressure 15 atm, cinnamaldehyde 0.25 mmol,
solvent 5 mL isopropanol, catalyst 50 mg.
because the support itself could not dissociate the hydrogen.
The protonated hydrogen that partly contributed to the
transformation of CAL over the support derived from the
decomposition of isopropanol, as the acetone was detected in
the products. Nevertheless, when the H2 was replaced by the N2
(Table 1, entry 10), the performance was also far below that on
the Au25/ZnAl‐300 catalyst. So, it can be concluded that the
hydrogen might be mostly dissociated on the low coordinated
gold particles which had low coverage of protective ligands.
What’s more, the gold catalysts with different supports
exhibited totally different performances (Table 1, entries
2,3,5–7), which suggested the properties of supports were of
great importance. The ZnAl‐HT appeared to be the best one.
Besides, the gold nanoparticles supported on ZnAl‐HT and
prepared by traditional DP method showed an inferior catalytic
performance (Table 1, entry 4) than the thiol stabilized gold
(Table 1, entry 1), that indicated the residual sulfur might take
a part in the reaction. Furthermore, the oxygen vacancies cre‐
ated by the appearance of the Alp might positioned at he
gold‐support interfaces for the Au25/ZnAl‐300 catalyst, which
could preferentially adsorb the C=O bond. Based on the above
reasons, the key reaction sites that affect the selectivity of COL
under the sufficient H2 might be the gold‐support interfaces.
Further investigation is still needed to make it clear.
A reaction mechanism for the hydrogenation of CAL to COL
over the Au25/ZnAl‐300 catalyst was proposed based on the
above results (Scheme 2). Firstly, the CAL molecular was
absorbed at the interface between gold and the support. During
the process, the oxygen vacancy originated from the
unsaturated pentacoordinate Al3+ species acted as the Lewis
sites and resulted in the preferential adsorption of the C=O
bond [1,61]. Then, at the second step, the adsorbed species
accepted the hydrogen proton from dissociation of hydrogen
on the gold particles and the migration of hydrogen from the
solvent to form the intermediate species. With the
deprotonation process, the COL was formed on the surface of
gold and the support. After desorption, the catalyst was
recovered and the next cycle was ready to proceed.
We also conducted the recycle experiments on the
Au25/ZnAl‐300 catalyst. The result was showed in Fig. 8. After
reusing three times, the conversion of cinnamaldehyde
decreased from 94% to 43%, with the selectivity from 99% to
95%. We speculated that the deactivation of the catalyst was
attributed to the drain of the gold catalysts at the transfer
process and the partly agglomeration of the gold particles,
because the HAADF‐STEM results indicated that the mean
particle sizes of gold before and after reusing in hydrogenation
of cinnamaldehyde increased from 1.7 to 2.1 nm (Fig. 9). That
might result in the reduction of the catalytic performance.
Further characterizations about this are still needed to make it
clear.
100
80
60
40
20
0
100
80
60
40
20
0
cycle 1
cycle 2
cycle 3
Recycle times
Fig. 8. The conversion of CAL and the selectivity of COL with the recycle
times. Reactions conditions: 130 °C, H2 pressure 15 atm, cinnamalde‐
hyde 0.25 mmol, solvent 5 mL isopropanol, catalyst 50 mg. After each
test, the catalyst was washed with isopropanol and separated by cen‐
trifugation. Then, the new reactant mixture added to the reactor and
moved to a next run.

Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
479
117, 10417–10418.
4. Conclusions
[5] P. Mäki‐Arvela, J. Hájek, T. Salmi, D. Y. Murzin, Appl. Catal. A, 2005,
292, 1–49.
[6] J. M. Grosselin, C. Mercier, G. Allmang, F. Grass, Organometallics,
1991, 10, 2126–2133.
[7] S. J. Chen, L. Meng, B. X. Chen, W. Y. Chen, X. Z. Duan, X. Huang, B. S.
Zhang, H. B. Fu, Y. Wan, ACS Catal., 2017, 7, 2074–2087.
[8] M. A. Vannice, B. Sen, J. Catal., 1989, 115, 65–78.
[9] F. Jiang, J. Cai, B. Liu, Y. Xu, X. Liu, RSC Adv., 2016, 6, 75541–75551.
[10] Y. Tan, H. Liu, X. Y. Liu, A. Wang, C. Liu, T. Zhang, Chin. J. Catal.,
2018, 39, 929–936.
[11] H. Wei, X. Wei, X. Yang, G. Yin, A. Wang, X. Liu, Y. Huang, T. Zhang,
Chin. J. Catal., 2015, 36, 160–167.
[12] P. Johnston, N. Carthey, G. J. Hutchings, J. Am. Chem. Soc., 2015,
137, 14548–14557.
[13] H. Rojas, G. Diaz, J. J. Martinez, C. Castaneda, A. Gomez‐Cortes, J.
Arenas‐Alatorre, J. Mol. Catal. A, 2012, 363, 122–128.
[14] Z. Konuspayeva, G. Berhault, P. Afanasiev, T. S. Nguyen, S. Giorgio,
L. Piccolo, J. Mater. Chem. A, 2017, 5, 17360–17367.
[15] C. H. Hao, X. N. Guo, Y. T. Pan, S. Chen, Z. F. Jiao, H. Yang, X. Y. Guo, J.
Am. Chem. Soc., 2016, 138, 9361–9364.
In summary, the ZnAl‐hydrotalcite‐supported Au25 NCs as
the precatalyst was proved very selective for the liquid‐phase
hydrogenation of CAL to COL. It was active for the hydrogena‐
tion of the C=O bond but inactive for the activation of the C=C
bond. Thus, it exhibited a stable high selectivity for unsaturated
alcohol even in the prolonged reaction time. The control ex‐
periment and in‐situ DRIFTS of the substrate under high pres‐
sure of hydrogen suggested that the scarce hydrogenation of
the vinyl group and the weak adsorption of the products over
the Au25/ZnAl‐300 catalyst might be responsible for the high
selectivity for hydrogenation of carbonyl group. Besides, the
27Al MAS‐NMR spectra of Au25/ZnAl‐300 catalyst implied there
were coordinately unsaturated AlP species, which might supply
the oxygen vacancy for the preferential adsorption of the C=O
bond. The control of the particle size of gold and the choice of
ZnAl‐HT support as well as the isopropanol solvent determined
the good catalytic activity and high selectivity of the target
product. We believe this work could provide good references
for designing the gold catalysts for chemoselective hydrogena‐
tion.
[16] Y. F. Zhang, S. H. Zhang, X. L. Pan, M. Bao, J. H. Huang, W. J. Shen,
Catal. Lett., 2017, 147, 102–109.
[17] Y. X. Liu, Y. M. Luo, Z. J. Wei, Asian J. Chem., 2013, 25, 8617–8620.
[18] E. Castillejos, E.Gallegos‐Suarez, B. Bachiller‐Baeza, R. Bacsa, P.
Serp, A. Guerrero‐Ruiz, I. Rodriguez‐Ramos, Catal. Commun.,
2012, 22, 79–82.
References
[19] H. N. Chen, D. A. Cullen, J. Z. Larese, J. Phys. Chem. C, 2015, 119,
28885–28894.
[20] X. Zhang, Y. C. Guo, Z. C. Zhang, J. S. Gao, C. M. Xu, J. Catal., 2012,
292, 213–226.
[1] P. Claus, Top. Catal., 1998, 5, 51–62.
[2] P. Gallezot, D. Richard, Catal. Rev. Sci. Eng., 1998, 40, 81–126.
[3] K. Bauer, D. Garbe, H. Surburg, Common Fragrance and Flavor
Materials: Preparation, Properties and Uses, 2008.
[21] G. J. Hutchings, F. King, I. P. Okoye, C. H. Rochester, Appl. Catal. A,
[4] T. Ohkuma, H. Ooka, T. Ikariya, R. Noyori, J. Am. Chem. Soc., 1995,
Graphical Abstract
Chin. J. Catal., 2021, 42: 470–481 doi: 10.1016/S1872‐2067(20)63678‐6
Producing of cinnamyl alcohol from cinnamaldehyde over supported gold nanocatalyst
Yuan Tan, Xiaoyan Liu*, Leilei Zhang, Fei Liu, Aiqin Wang, Tao Zhang
Dalian Institute of Chemical Physics, Chinese Academy of Sciences; Zhejiang Normal University; University of Chinese Academy of Sciences
The Au25/ZnAl‐300 catalyst displayed a stable high selectivity to cinnamyl alcohol from the hydrogenation of cinnamaldehyde, which was
attributed to the easier adsorption and activation of the C=O bond than that of the C=C bond.

480
Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
1992, 83, L7–L13.
[41] F. Menegazzo, F. Pinna, M. Signoretto, V. Trevisan, F. Boccuzzi, A.
Chiorino, M. Manzoli, Appl. Catal. A, 2009, 356, 31–35.
[42] L. Liu, H. Li, Y. Tan, X. Chen, R. Lin, C. Huang, S. Wang, X. Wang, X. Y.
Liu, M. Zhao, W. Yang, Y. Ding, Catalysts, 2020, 10, 107.
[43] F. A. Al‐Bayati, M. J. Mohammed, Pharm. Biol., 2009, 47, 61–66.
[44] P. Calvini, A. Gorassini, Restaurator, 2002, 23, 48–66.
[45] J. L. Brown, T. Chen, H. D. Embree, G. F. Payne, Ind. Eng. Chem. Res.,
2002, 41, 5058–5064.
[46] H. Y. Chen, H. B. Ji, AIChE J., 2010, 56, 466–476.
[47] F. Zhao, S. Fujita, J. Sun, Y. Ikushima, M. Arai, Chem. Commun.,
2004, 2326–2327.
[48] H. S. Mansur, C. M. Sadahira, A. N. Souza, A. A. P. Mansur, Mater. Sci.
Eng. C, 2008, 28, 539–548.
[22] J. E. Bailie, G. J. Hutchings. Chem. Commun., 1999, 2151–2152.
[23] J. E. Bailie, H. A. Abdullah, J. A. Anderson, C. H. Rochester, N. V.
Richardson, N. Hodge, J. G. Zhang, A. Burrows, C. J. Kiely, G. J.
Hutchings, Phys. Chem. Chem. Phys., 2001, 3, 4113–4121.
[24] X. Yang, A. Wang, X. Wang, T. Zhang, K. Han, J. Li, J. Phys. Chem. C,
2009, 113, 20918–20926.
[25] Y. Tan, X. Y. Liu, L. L. Zhang, A. Q. Wang, L. Li, X. L. Pan, S. Miao, M.
Haruta, H. S. Wei, H. Wang, F. J. Wang, X. D. Wang, T. Zhang, Angew.
Chem. Int. Ed., 2017, 56, 2709–2713.
[26] Y. Tan, X. Y. Liu, L. Li, L. L. Kang, A. Q. Wang, T. Zhang, J. Catal.,
2018, 364, 174–182.
[27] H. P. Reddy Kannapu, C. A. Mullen, Y. Elkasabi, A. A. Boateng. Fuel
Process. Technol., 2015, 137, 220–228.
[49] E. Erasmus, S. Afr. J. Chem., 2013, 66, 216–220.
[50] K. Kervinen, M. Allmendinger, M. Leskelä, T. Repo, B. Rieger, Phys.
Chem. Chem. Phys., 2003, 5, 4450–4454.
[28] D. Wang, C. Deraedt, J. Ruiz, D. Astruc. J. Mol. Catal. A, 2015, 400,
14–21.
[29] H. Lu, H. Yin, Y. Liu, T. Jiang, L. Yu, Catal. Commun., 2008, 10,
313–316.
[51] C. Tengroth, U. Gasslander, F. O. Andersson, S. P. Jacobsson, Pharm.
Dev. Technol., 2005, 10, 405–412.
[30] Z. M. Tian, X. Xiang, L. S. Xie, F. Li, Ind. Eng. Chem. Res., 2013, 52,
288–296.
[52] J. N. Xin, Q. Xu, H. Zhang, X. C. Yang, L. H. Lü, Spectrosc. Spect. Anal.,
2008, 28, 784–787.
[31] M. M. Wang, L. He, Y. M. Liu, Y. Cao, H. Y. He, K. N. Fan, Green Chem.,
2011, 13, 602–607.
[53] H. Miyata, T. Ohno, F. Hatayama, J. Chem. Soc., Faraday Trans.,
1995, 91, 3505–3510.
[32] Y. Zhu, H. F. Qian, B. A. Drake, R. C. Jin, Angew. Chem. Int. Ed., 2010,
49, 1295–1298.
[54] C. P. Marshall, E. J. Javaux, A. H. Knoll, M. R. Walter, Precambrian
Res., 2005, 138, 208–224.
[33] P. G. N. Mertens, P. Vandezande, X. Ye, H. Poelman, I. F. J. Van‐
kelecom, D. E. De Vos, Appl. Catal. A, 2009, 355, 176–183.
[34] K. J. You, C. T. Chang, B. J. Liaw, C. T. Huang, Y. Z. Chen, Appl. Catal.
A, 2009, 361, 65–71.
[35] J. Lenz, B. C. Campo, M. Alvarez, M. A. Volpe, J. Catal., 2009, 267,
50–56.
[55] J. Z. Hu, S. Xu, J. H. Kwak, M. Y. Hu, C. Wan, Z. Zhao, J. Szanyi, X. Bao,
X. Han, Y. Wang, C. H.F. Peden, J. Catal., 2016, 336, 85–93.
[56] J. H. Kwak, J. Hu, D. Mei, C. W. Yi, D. H. Kim, C. H. F. Peden, L. F.
Allard, J. Szanyi, Science, 2009, 325, 1670–1673.
[57] A. A. Ali Ahmed, Z. A. Talib, M. Z. bin Hussein, A. Zakaria, J. Alloys
Compd., 2012, 539, 154–160.
[36] C. Milone, M. C. Trapani, S. Galvagno, Appl. Catal. A, 2008, 337,
163–167.
[37] C. Milone, C. Crisafulli, R. Ingoglia, L. Schipilliti, S. Galvagno, Catal.
Today, 2007, 122, 341–351.
[38] E. Bus, R. Prins, J. A. van Bokhoven, Catal. Commun., 2007, 8,
1397–1402.
[39] Sh. K. Shaikhutdinov, R. Meyer, M. Naschitzki, M. Bäumer, H. J.
Freund, Catal. Lett., 2003, 86, 211–219.
[58] Y. Zhao, G. Chen, T. Bian, C. Zhou, G. I. N. Waterhouse, L. Z. Wu, C. H.
Tung, L. J. Smith, D. O'Hare, T. Zhang, Adv. Mater., 2015, 27,
7824–7831.
[59] P. Claus, A. Brückner, C. Mohr, H. Hofmeister, J. Am. Chem. Soc.,
2000, 122, 11430–11439.
[60] C. Milone, R. Ingoglia, L. Schipilliti, C. Crisafulli, G. Neri, S. Galvagno,
J. Catal., 2005, 236, 80–90
[61] G. Li, H. Abroshan, Y. Chen, R. Jin, H. J. Kim, J. Am. Chem. Soc., 2015,
137, 14295−14304.
[40] M. Haruta, Angew. Chem. Int. Ed., 2014, 53, 52–56.
纳米金催化肉桂醛选择加氢制肉桂醇
谭 媛a,b, 刘晓艳a,*, 张磊磊a, 刘 菲a, 王爱琴a, 张 涛a,c
a中国科学院大连化学物理研究所, 中国科学院航天催化材料重点实验室, 辽宁大连116023
b浙江师范大学, 杭州高等研究院, 浙江杭州311231
c中国科学院大学, 北京100049
摘要: ,-不饱和醛/酮选择加氢生成不饱和醇是化学工业中一类重要反应, 在精细化工生产中具有广泛应用, 近年来吸引
了研究者的广泛关注. 该类反应因涉及不饱和官能团和碳氧双键的选择加氢而颇具挑战性: 以肉桂醛选择加氢生成肉桂
醇反应为例, 肉桂醛分子中同时含有共轭的C=C双键和C=O双键, 从热力学角度上看, C=O双键键能比C=C双键键能大, 因
而碳碳双键比碳氧双键更容易被活化从而加氢得到饱和醛; 从动力学角度上看, C=C双键也比C=O双键更容易加氢. 对于
传统的铂族贵金属催化剂, 其应用于该类反应时往往存在选择性低, 容易深度加氢等问题. 负载型金催化剂此前被报道在
该类反应中表现出高选择性, 然而在反应物接近完全转化时, 目标产物也容易发生过度加氢生成饱和醇.
前期的研究结果发现用锌铝水滑石作载体, 硫醇稳定的金原子团簇(Au25)作为金的前驱体制备负载型金催化剂时, 其
在不饱和芳香硝基化合物的选择加氢反应中表现出很高的选择性. 考虑到在肉桂醛分子中C=O双键的加氢相比于C=C双
键更加困难, 因此, 本工作尝试将上述催化剂应用于以肉桂醛为代表的不饱和醛/酮选择加氢反应中. 考察了反应温度、氢
气压力以及溶剂效应对反应活性的影响, 结果发现升高温度或提高压力都能明显提升反应速率, 然而不同的溶剂对催化性

Yuan Tan et al. / Chinese Journal of Catalysis 42 (2021) 470–481
481
能影响很大, 当以具备氢转移能力的异丙醇和乙醇作为反应溶剂时, 催化活性和选择性最优, 在反应温度为130 oC, 氢气压
力为15 atm, 异丙醇为溶剂时反应5 h, 肉桂醛的转化率和肉桂醇的选择性可以达到98.3%和95.4%, 并且延长反应时间至15
h, 目标产物也不会发生过度加氢生成苯丙醇, 其选择性可以维持在95%以上.
为了研究该催化剂高活性和高选择性的原因, 制备了不同粒径大小和不同载体负载的金催化剂, 结果发现相比于其它
负载型金催化剂, 以锌铝水滑石负载的Au25团簇作为催化剂前体制得的催化剂在肉桂醛选择加氢制肉桂醇反应中表现出
最优的活性和选择性. 对照实验和原位漫反射红外光谱测试表明上述催化剂对碳碳双键的加氢表现为惰性, 对目标产物
的吸附也相对较弱. 27Al固体核磁共振结果表明配位不饱和的五配位Alp物种可能为C=O双键的优先吸附提供所需的氧空
位, 这可能是该催化剂具有较高选择性的原因. 综上, 推测小尺寸的金颗粒具有较多低配位的金原子, 可以活化氢气, 而反
应物和产物的吸脱附性质与载体密切相关, 在以锌铝水滑石为前驱体制备的金催化剂表面, C=C双键吸附较弱, C=O双键
优先吸附, 产物较容易脱附, 不容易发生过度加氢反应, 因此该催化剂在肉桂醛选择加氢反应中表现出高活性和高选择性.
上述工作可以为设计制备高选择性的负载型金催化剂提供参考.
关键词: 金团簇; 载体效应; 肉桂醇; 选择加氢; 肉桂醛
收稿日期: 2020-05-19. 接受日期: 2020-06-23. 出版日期: 2021-03-05.
*通讯联系人. 电话: (0411)84379416; 传真: (0411)84691570; 电子信箱: xyliu2003@dicp.ac.cn
基金来源: 国家自然科学基金(21776271, 21606227); 国家重点研发计划“纳米科技”重点专项(2016YFA0202801); 中国科学院战
略性先导科技专项(XDB17020100); 浙江省自然科学基金(LQ20B030005); 中国科学院创新交叉团队项目(BK2018001).
本文的电子版全文由Elsevier出版社在ScienceDirect上出版(http://www.sciencedirect.com/science/journal/18722067).