Mechanism of supported Ru3Sn7 nanocluster-catalyzed selective hydrogenation of coconut oil to fatty alcohols

Article abstract of DOI:10.1039/c8cy00037a

As a promising hydrotreating catalyst, it was previously reported that Ru?OSn (Ru electronically interacts with Sn oxides) on RuSn/SiO₂ was the active site for fatty acid hydrogenation, but here in this work we found that Ru₃Sn₇ nanoclusters on RuSn/SiO₂ were responsible for the selective hydrogenation of diverse fatty acids and coconut oil to fatty alcohols. The XPS results indicated no interaction between Sn^δ+ and Ru⁰, suggesting that SnO_x may exist as isolated species. In contrast, the binding energy shifts of Ru⁰ and Sn⁰ in the XPS spectra demonstrated a strong interaction, as a result of the formation of Ru₃Sn₇ alloy nanoclusters. It was demonstrated that the highest yield of fatty alcohol was obtained with a Sn/Ru ratio of 7/3 (hydrogenation rate: 2.45 g g^-1 h^-1), and the careful selection of the Sn/Ru ratio and reduction temperatures greatly suppressed the formation of Sn and SnO₂ phases. The ratio of the stearyl alcohol formation rate to its consumption rate was 40.8 with Ru₃Sn₇/SiO₂ under the selected conditions. In catalysts with a Sn/Ru ratio higher than 7/3, the presence of additional SnO₂ catalyzes the formation of undesired esters at a rate of 0.31 g g^-1 h^-1. Excess SnO₂ would be reduced to Sn at temperatures higher than 600 °C, while Sn can catalyze ester by-product formation at a rate of 0.88 g g^-1 h^-1. The DFT calculations showed that CH₃COOH adsorbs on the Ru₃Sn₇ (111) surface via Sn-O interactions at the two top sites of adjacent surface Sn atoms, and such adsorption was mainly due to the electrostatic interactions between the molecule and the positively charged surface Sn atoms. The charge density difference (CDD) plots of co-adsorbed CH₃CO? and OH? intermediates indicated the bonding relationships between Sn-O and Ru-α-C, suggesting that the surface Sn atoms also took part in the catalytic reaction as an important surface sorption site as well as a Ru₃Sn₇ structure component, while Ru atoms bonded with α-C and hydrogenated the adsorbed intermediate species with the adsorbed H? to the final alcohol. A further indication that Ru₃Sn₇ was the active species in the bimetallic Ru-Sn catalyst was given by the much lower energy barrier for hydrogenation of acetic acid in Ru₃Sn₇ (111) (81.0 kJ mol^-1) compared to Ru (0001) (123.5 kJ mol^-1).

Full text of DOI:10.1039/c8cy00037a

View Article Online
View Journal
Catalysis
Science &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: Z. Luo, J. Kong,
Q. Bing, J. Liu and C. Zhao, Catal. Sci. Technol., 2018, DOI: 10.1039/C8CY00037A.
This is an Accepted Manuscript, which has been through the
Volume 6 Number 1 7 January 2016 Pages 1–308
Royal Society of Chemistry peer review process and has been
accepted for publication.
Catalysis
Science &
Technology
Accepted Manuscripts are published online shortly after
www.rsc.org/catalysis
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
ISSN 2044-4753
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
PAPER
Qingzhu Zhang et al.
Catalytic mechanism of C–F bond cleavage: insights from QM/MM
analysis of fluoroacetate dehalogenase
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
rsc.li/catalysis

Please do not adjust margins
Catalysis Science & Technology
Page 1 of 12
View Article Online
DOI: 10.1039/C8CY00037A
Catalysis Science & Technology
ARTICLE
Mechanism in Supported Ru3Sn7 Nanoclusters Catalyzed Selective
Hydrogenation of Coconut Oil to Fatty Alcohols
Zhicheng Luo1a, Qiming Bing1b, Jiechen Konga, Jing-yao Liub* and Chen Zhaoa*
Received 00th January 20xx,
Accepted 00th January 20xx
As a promising hydrotreating catalyst, it was previously reported that Ru…O=Sn (Ru electronically interacts with Sn oxides)
on RuSn/SiO2 was the active site for fatty acid hydrogenation, but here in this work we found that Ru3Sn7 nanoclusters on
RuSn/SiO2 were responsible for the selective hydrogenation of diverse fatty acids and coconut oil to fatty alcohols. The XPS
results indicated no interaction between Snδ+ and Ru0, suggesting that SnOx may exist as the isolated species. In contrary,
the binding energy shifts of Ru0 and Sn0 at XPS spectra demonstrated a strong interaction, as a result from the formation of
Ru3Sn7 alloy nanoclusters. It was demonsrated that the highest yield of fatty alcohol was obtained with a Sn/Ru ratio of 7/3
(hydrogenation rate: 2.45 g·g−1·h−1), and the careful selection of the Sn/Ru ratios and the reduction temperatures greatly
suppressed the formation of Sn and SnO2 phases. The ratio of stearyl alcohol formation rate to its consumption rate was
40.8 with Ru3Sn7/SiO2 at selected conditions. In catalysts with a Sn/Ru ratio higher than 7/3, the presence of additional
SnO2 catalyzes the formation of undesired ester with a rate of 0.31 g·g−1·h−1. Excess SnO2 would be reduced to Sn at
temperatures higher than 600 °C, while Sn can catalyze ester by-product formation with a rate of 0.88 g·g−1·h−1. The DFT
calculations showed that CH3COOH adsorbs on the Ru3Sn7 (111) surface via Sn-O interactions at the two top sites of
adjacent surface Sn atoms, and such adsortion was mainly due to the electrostatic interactions between the molecule and
the positively charged surface Sn atoms. The charge density difference (CDD) plots of co-adsorbed CH3CO* and OH*
intermediates indicated the bonding relationships of Sn-O and Ru-α-C, suggesting that the surface Sn atoms also took part
in the catalytic reaction as an important surface sorption site as well as a Ru3Sn7 structure component, while Ru atoms
bonded with α-C and hydrogenated the adsorbed intermediate species with the adsorbed H* to the final alcohol. Further
indication that Ru3Sn7 were the active species in the bimetallic Ru-Sn catalyst was given by the much lower energy barrier
for hydrogenation of acetic acid in Ru3Sn7 (111)(81.0 kJ·mol−1) compared to Ru (0001) (123.5 kJ·mol−1).
DOI: 10.1039/x0xx00000x
www.rsc.org/
1. Introduction
intermediates or alcohol products, respectively, to alkanes
over metals such as Ni [3-5], Pd [6, 7], or Pt [8], sites; and (3) the
The selective conversion of natural sources of oil to fatty dehydration of fatty alcohol products to olefins or ethers over
alcohols holds increasing interest as a sustainable route for acidic sites. The dissociation energy of the COO-R bond is
producing important intermediates used in the synthesis of particularly high at about 400 kJ·mol−1 [9], which indicates that
[1,
cosmetics, surfactants, and plasticizers
2]. The reaction the proper selection of a catalyst system with active sites that
pathway leading to the formation of fatty alcohols from can efficiently activate inert COO-R bonds in triglycerides
triglycerides is displayed in Fig. S1. In order to form fatty poses a particularly daunting challenge. In addition, the
alcohols from the hydrogenation of lipids with high selectivity catalyst will need to operate at a temperature high enough to
and yield, three side-reactions need to be minimized: (1) the ensure high activity while simultaneously avoiding unwanted
esterification between fatty acid intermediates and fatty side reactions such as dehydration of fatty alcohols.
alcohol products over acid or base sites; (2) the
Great progress has been made in the field of
decarbonylation or hydrodeoxygenation of any aldehyde homogeneous catalysis to achieve high yields of fatty alcohols
from fatty esters. The catalysts include well-defined ruthenium
complexes bearing chelating 2-(diphenylphosphino)-1-
methylimidazole
(PN)
or
pincer-type
2,6-(di-tert-
11]. More
[10,
butylphosphinomethyl)pyridine (PNP) ligands
recently, non-noble-metal Fe pincer complexes have been
reported to hydrogenate esters [12-14]. However, due to the
high cost of the metal ligands used in these catalyst systems
together with problems with respect to recycling the
homogeneous catalysts, heterogeneous catalysts have proven
to be more attractive choices for industrial applications. The
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 2 of 12
View Article Online
DOI: 10.1039/C8CY00037A
ARTICLE
Catalysis Science & Technology
well-known Cu-Cr based Adkins catalyst exhibits high efficiency conversion. Subsequently the bimetallic RuPt catalyst was
for the hydrogenation of fatty esters to alcohols, but requires tested, with the conversion of 20.6%. Whereas single Ru and
very harsh reaction conditions (250–350 °C and ca. 10–20 MPa Sn catalysts led to very small conversions (3% and 1.6%,
H2 pressure). In addition, toxicity problems associated with respectively), we were gratified to find that there was a
chromium compounds have driven researchers to find an remarkable increase in the conversion up to 82.7% when the
efficient Cr-free catalyst. Thus bimetallic Cu-M catalysts such bimetallic RuSn/SiO2 catalyst was employed, and the target
[15]
[16, 17]
as Cu-Fe
and Cu-Zn
have been developed for the product yield reached 80.0%.
formation of fatty alcohols derived from natural sources at
high pressures. Recently, we reduced coconut oil to fatty
alcohols with nearly 100% selectivity over a hydrothermally
synthesized Cu/SiO2 catalyst (consisting of surface Cu2O·SiO2
and Cu0 surface species) in methanol in absence of extraneous
100
80
60
40
20
0
a
Stearyl Alcohol
Other Products
RuSn
[19-
hydrogen [18]. In addition, other catalysts with noble metals
29]
RuPt
such as Rh, Ru, Pt, and Pd have been used to reduce ester
Pt
derivatives. Among these, RuSn bimetallic catalysts are shown
to very promising for the hydrogenation of esters or acids to
alcohol [21,22]. It was recognized that the carbonyl of acid or
acid derivatives (esters) polarized by either Sn2+ or Sn4+ Lewis
acid sites can be hydrogenated by the H· radical transferred
from an adjacent Ru-H site in a facile manner. Therefore, the
interacted Ru…O=Sn species is generally considered as the
active sites.
Ru
Co
Pd
Ni
Sn
PdSn
PtSn
b
Acid Conversion on Ru
Alcohol Yield on Ru
Acid Conversion on Sn
Alcohol Yield on Sn
80
60
40
20
0
Acid Conversion on RuSn
Alcohol Yield on RuSn
However, in this contribution, we report that the active
sites for hydrogenation of fatty acids and fatty esters to fatty
alcohols are demonstrated to be the supported Ru3Sn7
nanoclusters rather than Ru…SnO2 (Ru electronically interacts
with Sn oxides). The nature of the catalytic active sites was
carefully explored by the detailed characterized and the
relation with targeted catalytic tests. By comparing the
formation energies and energy barriers of acid hydrogenation
on potential facets, density function theory calculations reveal
that the active sites are comprised of Ru3Sn7 nanoclusters. In
addition, the role of Sn on RuSn for adsorption and reaction is
explored by the charge density difference (CDD) plot.
0
30
60
90
120
Time (min)
c
100
95
90
85
80
Ru
RuSn
Sn
0
20 40 60 80 100 120
Time (min)
Fig. 1. (a) The product distributions for stearic acid hydrogenation over
various metal centers supported on SiO2. Reaction conditions: Stearic acid
(1.0 g), catalyst (0.2 g, 5 wt.%), dodecane (80 mL), 240 °C, 4 MPa H2, 120
min., and stirring at 700 rpm. (b) Stearic acid conversion and alcohols yield
over various metal centers supported on SiO2 as a function of time.
Reaction conditions: Stearic acid (1.0 g), catalyst (0.2 g), dodecane (80 mL),
240 ˚C, 4 MPa H2, stirring at 700 rpm. (c) The stability of stearyl alcohol over
various metal centers supported on SiO2 as a function of time. Reaction
conditions: stearyl alcohol (1.0 g), catalyst (0.2 g), dodecane (80 mL), 240°C,
4 MPa H2, stirring at 700 rpm.
2. Results and discussion
2.1 Catalyst screening for selective hydrogenation of stearic
acid to stearyl alcohol
The rate-determining step in the conversion of
triglycerides to fatty alcohols is the hydrogenation of fatty
[2]
acids. Therefore, a fatty acid (C18 stearic acid) was selected
as the substrate for catalyst screening. A neutral SiO2 support
and a moderate hydrogenation temperature (240 °C) were
used to minimize unwanted side-reactions such as dehydration
of the target alcohol while maintaining appropriate catalyst
activity. The noble metals Pt, Ru, and Pd, the non-noble metals
Ni, Co, and Sn, and the bimetallic catalysts Ru-Pt, Pt-Sn, Pd-Sn,
and Ru-Sn were incorporated into the SiO2 support with a 5 wt%
loading by incipient wetness impregnation. The XRD patterns
of these catalysts are listed in Fig. S2.
Encouraged with this improvement in conversion with a
noble metal-Sn bimetallic catalyst, PtSn and PdSn catalysts
were also tested for stearic acid hydrogenation, but only led to
poor substrate conversions (8.4%, and 7.5% conversion,
respectively). As can be seen from the XRD patterns (Fig. S2),
the bimetallic catalysts PtSn/SiO2, PdSn/SiO2, and RuSn/SiO2
exhibited new diffraction peaks corresponding to PtSn4 (JCPDS
#65-9537), Pd3Sn2 (JCPDS #04-0801), and Ru3Sn7 (JCPDS # 26-
0504), respectively, implying that the thermodynamically
stable alloy phase was formed after co-impregnation of Sn and
the corresponding noble metals. Those results indicate that
To evaluate the activity of these catalysts, the
hydrogenation of stearic acid was conducted in batch mode
using dodecane as solvent, at conditions of 240 °C and 4 MPa
H2. As shown in Fig.1a, Almost all of the single metal catalysts
(Pt, Ru, Ni, Co, and Sn) were inactive, except for Pd with a 30%
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 3 of 12
View Article Online
DOI: 10.1039/C8CY00037A
Catalysis Science & Technology
ARTICLE
there is a synergistic interaction between Ru and Sn that plays
a major role in improving the conversion of fatty acid
hydrogenation. These results thus far suggested that in the
highly active RuSn/SiO2 catalyst, RuSn may form a new phase
or that Sn in the form of the oxide (SnO2) could dramatically
modify the metallic Ru center.
The kinetics curves of stearic acid hydrogenation for the
tested catalysts, showed that RuSn achieved the highest
hydrogenation rate (2.02 g·g−1·h−1), while such hydrogenation
rates were 0 with Sn sites and 0.07 g·g−1·h−1 with Ru sites (Fig.
1b). The comparative kinetics curves on other metals (Pt, Ru,
Ni, Co, RuPt, PtSn, and PdSn) were displayed in Fig. S3. In
addition, the stability of the produced fatty alcohol with the
developed RuSn catalysts at selected conditions should also be
considered. As displayed in Fig. 1c, the conversion rate of Fig. 2. The XRD patterns of Ru/SiO2 and Ru-Sn/SiO2 samples with different
stearyl alcohol over RuSn/SiO2 was 0.06 g·g−1·h−1, while such ratios of Sn to Ru.
rates were 0.04 and 0.02 g·g−1·h−1 over Ru/SiO2 and Sn/SiO2,
respectively. Thus, the ratio of stearyl alcohol formation rate
The TPR-H2 profiles of calcined metallic catalysts help to
to consumption rate was 40.8 with RuSn/SiO2 at 240 °C, identify the surface species during the hydrogen reduction
indicating that the target product was relatively stable at process (Fig. 3). Therefore, we conducted TPR-H2 experiments
selected catalytic conditions, as shown in Table 1. Having on the hydrogenation of calcined single metal catalysts (Sn,
ascertained the optimal catalyst for the model reaction Ru), and the bimetallic Ru-Sn catalyst with Sn/Ru ratios varying
involving stearic acid hydrogenation, we next sought to further from 2.6/3 to 15.3/3 (i.e. corresponding to Sn/Ru wt.% ratios
understand and explore the nature of the active site through ranging from 1.5/1.5 to 9/1.5). In the 5 wt.% monometallic Sn
catalytic tests and a detailed characterization of the RuSn/SiO2 catalyst, SnO2 was reduced at around 645°C, while in the single
catalyst.
1.5 wt.% Ru catalyst, RuO2 was reduced at around 150 °C. In
the 1.5 wt.% Ru-1.5 wt.% Sn catalyst, the RuO2 reduction peak
Table 1. Calculated reaction rates over different catalysts.
became broader, indicating a shift in the reduction
temperature of RuO2 that is closer to SnO2 due to interactions
between RuO2 and SnO2. In the 1.5 wt.% Ru-4.5 wt.% Sn
bimetallic RuSn catalyst (corresponding to a Sn/Ru atomic ratio
of 7.7/3, higher than the Sn/Ru ratio in the Ru3Sn7 nanocluster),
the SnO2 reduction peak at around 645 °C disappeared, and a
new broad peak for SnO2 reduction appeared at 515 °C. The
lowered reduction temperature is probably due to weak
interactions between isolated SnO2 and RuO2 species on the
SiO2 support surface. Moreover, an additional peak emerged
at around the reduction temperature ranges of 350-460 °C,
probably corresponding to the reduction of SnO2 species that is
close to RuO2. Hence, 460 °C is the minimum temperature
requirement for fully reduced Sn to incorporate with Ru atoms,
thereby forming Ru3Sn7 nanoclusters.
Catalyst
Ru/SiO2 Sn/SiO2
RuSn/SiO2
1.98
Alcohol formation rate
(g·g-1·h-1)
0.07
0
Alcohol consumption 0.04
rate (g·g-1·h-1)
0.02
0.06
2.2 Characterization of RuxSny/SiO2 samples
In order to modify the Sn/Ru ratio in the RuSn/SiO2
catalysts, the Ru content was kept constant at 1.5 wt.%, while
the Sn content varied from 0 to 9.0 wt.%. The XRD spectra of
the resulting catalysts were recorded in order to observe
phase transformations as the Sn/Ru ratio was varied (Fig. 2).
The monometallic Ru/SiO2 catalyst displayed the typical main
peaks observed in hexagonal crystalline Ru (JCPDS #65-1863),
with (100), (002), (101), and (102) reflection planes present.
When 1.5 wt.% Sn was incorporated, the peak intensity
corresponding to Ru became weaker, while a new peak
corresponding to Ru3Sn7 (JCPDS # 26-0504) appeared. As the
Sn content was increased to 4.5 wt.%, the Ru characteristic
peak disappeared and the peak of Ru3Sn7 was observed. The
peak intensity corresponding to cubic Ru3Sn7 gradually
increased up to 9 wt.%. It should be noted that an additional
SnO2 species (JCPDS # 41-1445) was observed in the XRD
patterns as higher amounts of Sn were incorporated into the
catalyst.
To prove our hypothesis regarding the temperature
requirements for the reduction of SnO2 to form Ru3Sn7
nanoclusters, we obtained the XRD patterns of the catalyst
samples after reducing calcined 1.5 wt.% Ru-4.5 wt.% Sn/SiO2
under a continuously flowing H2 atmosphere at 250 °C, 460°C,
and 600 °C. As can be seen from the XRD patterns (Fig. S4), the
SnO2 peak (JCPDS
# 41-1445) was still detected after
conducting the hydrogenation at 250 °C, showing that SnO2
reduction had not taken place. It should be observed that in
XRD experiments the Ru0 peak is not observable when the Ru
particles are too small. When the reduction temperature was
increased to 460 °C, the characteristic XRD peaks
corresponding to Ru3Sn7 (JCPDS
# 26-0504) appeared,
demonstrating that the reduction peak at 400 °C close to the
RuO2 peak (150 °C) in the TPR-H2 profile corresponded to the
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3

Please do not adjust margins
Catalysis Science & Technology
Page 4 of 12
View Article Online
DOI: 10.1039/C8CY00037A
ARTICLE
Catalysis Science & Technology
reduction temperature of SnO2 species. At higher energy (BE) of Ru 3d5/2 and Sn 3d5/2 in the reduced RuSn/SiO2
temperatures no additional XRD peaks are observed; in are summarized in Table S1 and Fig. 5. The reduced Ru/SiO2
addition, the intensity of the Ru3Sn7 peak increases as catalyst showed a Ru 3d5/2 peak at 280.2 eV, corresponding to
increasing amounts of reduced Ru and Sn species associate to the Ru0 species, as shown in Fig. 5a. The reduced Sn/SiO2
form the nanoclusters.
catalyst demonstrated a wide Sn 3d5/2 peak centered at 486.9
eV and a weak shoulder at 484.9 eV, which were oxidized tin
species (Snδ+), and reduced tin species (Sn0), respectively, as
displayed in Fig. 5b. However, the XPS technique cannot
distinguish Sn2+ and Sn4+ species due to the similar BE values.
On the reduced Ru-Sn/SiO2 catalysts with increased Sn
loadings (0, 1.5, 3.0, to 4.5 wt%), the BE of Ru0 3d5/2 was
dramatically decreased from 280.2, 279.8, 279.7 to 279.4 eV,
suggesting the state of Ru was largely altered, perhaps due to
the formation of Ru3Sn7 alloy. Till the addition of 4.5 to 9.0 wt%
Sn, the BE values of Ru became almost stable at ca. 279.2 eV,
attributed to the equilibrium of Ru and Sn atoms in the Ru3Sn7
alloy. While, the BE of Sn0 3d5/2 increased from 484.9 to 485.3
with the increase of Ru/Sn ratio. The decreased BE of Ru0 and
the increased BE of Sn0 indicated an electron transfer from Ru0
Fig. 3. The temperature-programmed hydrogen reduction of calcined to Sn0, further suggesting the formation of RuSn alloys.
Ru/SiO2, Sn/SiO2, and RuSn/SiO2 samples with varying Ru/Sn ratios.
However, the BE of Snδ+ kept unchanged at 486.9 eV on those
RuSn samples, suggesting the isolated species and no
Subsequently, FT-IR spectra of Ru/SiO2 and RuxSny/SiO2 interactions between Snδ+ and Ru0. Hence, these results
samples with adsorbed CO were taken to provide further directly demonstrate the interaction between Ru0 and Sn0
insights on the nature of the surface composition (Fig. 4). The rather than between Ru0 and Snδ+. In addition, the surface
IR spectrum of the Ru/SiO2 sample showed three main peaks: a atomic ratios of Sn0/Snδ+ on Ru-Sn/SiO2 reached the maximum
low intensity band at 2134 cm−1, a very intense one at 2078 value at 0.8 with a Sn content of 4.5 wt% (see Table S1),
cm−1, and a broad unsymmetrical band at 2016 cm−1, assigned demonstrating the most abundant Ru3Sn7 clusters at such
[30,
to the Ruδ+-(CO)x, Ruδ+-(CO)2, and Ru0-CO absorption bands
Ru/Sn ratio. With the increasing Sn contents (higher than 4.5
31],respectively. As Sn was added to Ru (Sn/Ru ratio < 7/3) thus wt%), isolated SnO2 began to form. Such SnO2 species had no
forming Ru-Sn bimetallic catalysts, the intensity of the CO electron transfer with Ru0, suggested from the stable BE values
adsorption peak sharply decreased, indicating that Ru-Sn of Snδ+ XPS at 486.9 eV. The results gained from XPS spectra
alloyed nanoclusters began to form and that such nanoclusters are highly consistent with the XRD, TPR-H2, IR-CO results in the
inhibited CO adsorption on the surface. When the Sn/Ru ratio former part.
exceeded 7/3, and the main species present were Ru3Sn7 and
a
b
SnO2, CO adsorption on the Ru-Sn catalysts completely
disappeared.
Ru0
Sn0
Snδ+
C1s
-9.0 Sn
4.0Sn
-7.5 Sn
-9.0 Sn
-7.5 Sn
-6.0 Sn
-4.5 Sn
-6.0 Sn
-4.5 Sn
-3.0 Sn
-1.5 Sn
-0 Sn
-3.0 Sn
-1.5 Sn
276 278 280 282 284 286 288
480 485 490 495 500
Binding energy (eV)
Binding energy (eV)
Fig. 5. XPS spectra of RuSn/SiO2 samples with varying Ru/Sn ratios. (a) Ru
3d5/2; (b) Sn 3d5/2. The Ru content was constant at 1.5 wt.%, while the Sn
Fig. 4. The infrared spectra of adsorbed CO on RuSn/SiO2 samples with content varied from 0 to 9.5 wt.%.
varying Ru/Sn ratios.
Based on the results from the XRD, TPR-H2, IR-CO and XPS
The XPS technique is conducted to obtain valuable experiments of the Ru, Sn, and Ru-Sn catalysts, some general
information of surface species on the catalysts. The binding conclusions can be reached. After reduction of a catalyst with
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 5 of 12
View Article Online
DOI: 10.1039/C8CY00037A
Catalysis Science & Technology
ARTICLE
a Sn/Ru ratio lower than 7/3, the main species present are Ru0 as shown in Fig 6c. With the monometallic 1.5 wt.% Ru/SiO2
and Ru3Sn7 nanoclusters, while further increasing the Sn/Ru catalyst, a conversion of only 1.6% was achieved, indicating
ratio leads to Ru3Sn7 and SnO2 being mainly formed. SiO2 that single Ru0 metallic centres are nearly inactive under the
supported single metal Ru and Sn catalysts have substrate selected conditions. In the bimetallic catalyst, as the Sn
conversion of only 5% and 1.6% respectively, and thus single content increased from 1.5 wt.% to 3.0 wt.%, the yield for fatty
Ru and Sn metal centres can be excluded as significant active alcohol formation sharply increased from 8.6% to 64.5%,
sites. Therefore, the most active catalytic sites should respectively. Within these Sn content ranges, the main
comprise Ru-Sn bimetallic centres, with the Ru-SnO2, Ru3Sn7, catalytic species in the reduced samples are Ru and Ru3Sn7.
and Ru3Sn7-SnO2 species being likely candidates as the most The highest yield of fatty alcohol (99.2%) was achieved when
significant active sites.
the Sn wt.% was 4.1%, corresponding to the formation of
Ru3Sn7 species. With greater Sn contents (4.5 to 9.0 wt.% Sn,
i.e. Sn/Ru ratios greater than 7/3), the yield of fatty alcohol
decreased (from 98.2% to 79.0%), and the yield of the ester
2.3 Understanding the nature of active sites on RuSn/SiO2
catalyst for hydrogenation of fatty acids
To further elucidate the active sites which are responsible side-product increased (from 0.5% to 18.3%) as a result of
for fatty acid hydrogenation, we first used SiO2 and the additional SnO2 formation as shown from the TPR-H2
SnO2/SiO2 catalyst to investigate the effectiveness of the experiment. These results strongly suggest that the most
support and the catalyst towards the esterification side active catalytic sites for alcohol formation are Ru3Sn7
reaction between the target product fatty alcohol and the nanoclusters, while SnO2 species are responsible for catalyzing
reactant fatty acid. The kinetics for the reaction was recorded the esterification of fatty alcohols and acids. Therefore, the
in dodecane at 240°C in the presence of H2 at 4 MPa pressure optimal catalyst for fatty alcohol formation is achieved with a
(Fig. 6a and 6b). Whereas SiO2 was almost completely inert Sn/Ru ratio of 7/3. In catalysts with a Sn/Ru atomic ratio < 7/3,
towards ester formation, only attaining 2% conversion after the alcohol yields increase with larger amounts of Sn up to a
120 min, SnO2/SiO2 led to a relatively fast reaction with a rate ratio of 7/3, while a Sn/Ru atomic ratio > 7/3 led to the SnO2
of 0.31 g·g−1·h−1, and after 120 min ester was formed with 12.3% catalyzed formation of ester by-product.
conversion, implying that silica supported SnO2 can promote
the formation of the ester side product.
Apart from the effect of the Sn/Ru ratios, which
determines the type of catalytic species present, the reduction
and the oxidation state of the Ru and Sn species has a
significant effect on the product distribution obtained from
fatty acid hydrogenation (Fig. S4). For example, as shown in Fig.
7a, with a reduced Sn/SiO2 catalyst the esterifcation of stearic
acid and stearyl alcohol took place with a rate of 0.88 g·g−1·h−1,
which was much greater than that with the SnO2/SiO2 catalyst
(Fig. 6b). With the calcined 1.5 wt.% Ru-4.5 wt.% Sn/SiO2
catalyst (corresponding to a Sn/Ru ratio slightly higher than
7/3) at 25°C, the target alcohol was obtained in 17.3% yield
(Fig. 7b). In this case, the main species are inferred to be
RuO·SnO2 and SnO2 which are inactive for substrate
hydrogenation, but which would be reduced in-situ by the
hydrogen present in the catalytic reduction to metallic Ru0 at
240 °C. With the same catalyst, at 250°C stearyl alcohol was
obtained in only 47% yield. At this temperature, the main
surface species were Ru·SnO2 and SnO2. The maximum yield of
stearyl alcohol was nearly quantitative (98.6%), and was
achieved at a reduction temperature of 460 °C. This result
further implied that the catalytic species was Ru3Sn7 (the main
species present after reduction at 460°C), but not Ru…O=Sn
(Ru electronically interacts with Sn oxides). At 600 °C, the
Fig. 6. Product distributions of conversion of stearic acid and stearyl alcohol presence of trace quantities of Sn in addition to the main
on the catalysts (a) SiO2, and (b) 5 wt.% SnO2/SiO2. Reaction conditions: Ru3Sn7 species led to a lower fatty alcohol yield (87.8%) and a
stearic acid (0.5 g), stearyl alcohol (0.5 g), catalyst (0.2 g), dodecane (80 mL), higher yield of ester by-product (6.7%). These results show
240°C, 4 MPa H2, 0-120 min, stirring at 700 rpm. (c) Impact of Sn/Ru ratios how the composition of the RuSn catalyst changes with varying
towards product distributions from stearic acid conversion. Reaction Sn/Ru ratios (Fig. 6c) and reduction temperatures (Fig. 7b).
conditions: Stearic acid (1.0 g), RuSn/SiO2 (0.2 g, Ru loading: 1.5 wt%, Sn With the monometallic Ru catalyst and reduction at 460°C, Ru
loading: 0-9 wt%), dodecane (80 mL), 240 °C, 4 MPa H2, 120 min, stirring at nanoclusters were dispersed on the SiO2 support surface. At
700 rpm.
the same temperature, when the tin content was increased to
Secondly, the Sn/Ru atomic ratio was varied and was 1.5 or 3.0 wt.% (corresponding to a Sn/Ru ratio < 7/3), SnO2
found to have a dramatic effect on the product distributions, species were reduced to Sn, which then associated with free
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5

Please do not adjust margins
Catalysis Science & Technology
Page 6 of 12
View Article Online
DOI: 10.1039/C8CY00037A
ARTICLE
Catalysis Science & Technology
Ru0 sites to form Ru3Sn7 nanoclusters. Therefore, when the thought to poison the active site responsible for hydrogen
Sn/Ru ratio is lower than the coordination number, isolated activation, Pouilloux et al. pointed out the existence two tin
SnO2 species are not present. When the Sn wt.% matches the oxide species (SnOx and SnOy), where the SnOx species (x<y)
Sn/Ru ratio of 7/3, Ru3Sn7 nanoclusters are exclusively formed would interact closely with ruthenium without significant
on the surface of SiO2. As the Sn content further increases formation of RunSnm alloy [34]. These SnOx species interact with
(4.5–9.0 wt.%), unreduced SnO2 species were present along the C=O bond of the ester to promote transfer of activated
with the Ru3Sn7 nanoclusters.
hydrogen bound to ruthenium to receive an acetal of tin. Such
acetal is subsequently converted to an aldehyde, which is then
fast hydrogenated into the final alcohol. In addition, excessive
SnOy can block free Ru0 sites, suppressing the hydrogenation
capability (see Scheme 1a).
Based on our results, we suggest that non-alloyed
Ru…Sn=O species are not the active sites for acid
hydrogenation. First, the highest fatty alcohol yield was
obtained when the predominant surface species were Ru3Sn7
nanoclusters (Fig. 6c). If Ru…Sn=O was the active catalyst, the
highest activity should be reached when the Sn content is
higher than 4.1%, which corresponds to catalysts where free
tin and tin oxides are present. Secondly, the XRD patterns and
TPR-H2 profiles demonstrate that in hydrogenation conducted
with the 1.5 wt.% Ru-4.5 wt.% Sn/SiO2 catalyst the main
species present at 250 °C, 460 °C, and 600 °C were Ru-SnO2,
Ru3Sn7 and trace amounts of SnO2, and Ru3Sn7-Sn respectively.
However, the highest yield was achieved at a reduction
temperature of 460 °C, where Ru3Sn7 species are
predominantly present. Third, XPS results did not indicate
interaction between Snδ+ and Ru0 due to the unchanged BE
vaules of Snδ+, suggesting that SnOx may exist as the isolated
species. In contrary, the BE shifts on Ru0 and Sn0 in XPS spectra
indicated a strong interaction due to the formation of Ru3Sn7
nanoclusters. Hence, Ru3Sn7 nanoclusters, rather than
Ru…Sn=O species are concluded to be the principal catalytic
sites. We deduce that, in our case, the oxygen in the carbonyl
group prefers to be adsorbed on Sn atoms of Ru3Sn7/SiO2
clusters, which would be easily attacked by a H· bound to Ru.
This species would be sequentially hydrogenated to an
Fig. 7. (a) The product distributions reacted with stearic acid and stearyl aldehyde intermediate, and eventually to the target fatty
alcohol on 5 wt.% Sn/SiO2 as a function of time. Reaction conditions: stearic alcohol product (see Scheme 1b).
acid (0.5 g), stearyl alcohol (0.5 g), Sn/SiO2 (0.2 g), dodecane (80 mL), 240°C,
4 MPa H2, stirring at 700 rpm. (b) Impact of the reduction temperatures of
Ru-Sn/SiO2 samples towards product distributions from stearic acid
conversion. Reaction conditions: stearic acid (1.0 g), 1.5%Ru 4.5%Sn/SiO2
(0.2 g), dodecane (80 mL), 240°C, 4 MPa H2, 120 min., stirring at 700 rpm.
Our results strongly indicate that the active sites for
hydrogenation are Ru3Sn7 nanoclusters, while Sn or SnO2
species present when the Sn wt.% ˃ 4.1 can both catalyze the
formation of ester side product. However, it was previously
reported that non-alloyed Ru…O=Sn species (… represents that
Ru electronically interacts with Sn oxides) dispersed on the
catalyst support surface, rather than
a Ru-Sn alloy as
suggested by our findings was the active site for acid
hydrogenation. The carbonyl of acid or acid derivatives (esters) Scheme 1. Mechanism of the hydrogenation of ester into alcohol over RuSn
[32, 33]
can be polarized by either Sn2+ or Sn4+ Lewis acid sites
,
catalyst (a) presented by Pouilloux, (b) presented by ours. Color scheme:
and the activated carbonyl carbon can then be hydrogenated yellow: Ru; cyan: Sn.
by a H· radical transferred from an adjacent Ru-H site. While
the electronic interaction of tin with an adjacent ruthenium is
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 7 of 12
View Article Online
DOI: 10.1039/C8CY00037A
Catalysis Science & Technology
ARTICLE
Subsequently, a detailed characterization was performed myristic acid, palmitic acid, and stearic acid) was carried out
on the most active RuSn/SiO2 catalyst, with a Ru/Sn atomic with Ru3Sn7/SiO2 and the optimized reduction conditions
ratio of
7 to 3. The optimized catalyst achieved the (dodecane as solvent, 240°C, 4 MPa H2), with the results in
hydrogenation of stearic acid to the corresponding fatty Table 2. The selectivity for fatty alcohols and the conversion
alcohol in 99.2% yield (see Fig. 6c). The BET surface area of Ru- were nearly 100%, with hydrogenation rates varying from 2.47
Sn/SiO2 was 177 m2·g−1 (Fig. 8a), while the XRD patterns g·g−1·h−1 (palmitic acid) to 4.85 g·g−1·h−1 (myristic acid). For the
showed that the main catalytic sites were Ru3Sn7 nanoclusters hydrogenation of C3 propionic acid and C4 butyric acid, lower
(Fig. 8b). The TEM image of the Ru3Sn7 nanoclusters on the conversions of 84.2% and 81.3% together with selectivity of
amorphous SiO2 support showed that the particles were small 83.5% and 80.4%, respectively were attained. However, the
at around 3.9 ± 1.3 nm (Fig. 8c). In addition, TEM-mapping of a apparently lower activity of the catalyst on C3 and C4 acids may
selected portion of the TEM image demonstrated that the be due to the lower solubility of these reactants in dodecane.
Ru3Sn7 nanoclusters were dispersed throughout the support
surface (Fig. 8d).
Table 2. Hydrogenation of C3-C18 acid and coconut oil to the
corresponding alcohols catalyzed by Ru3Sn7/SiO2
a
b
Selectivity (%)
Rate
(g·g-
1·h-1)
Conv.
(C%)
Substrate
Alcohol Alkane
Ester
0.21
Propionic
acid
84.2
83.5
0.50
2.10
c
d
Butyric acid
Lauric acid
81.3
99.2
99.3
99.0
100
80.4
98.5
99.6
99.7
99.0
96.6
0.38
0.63
0.36
0.23
0.55
3.03
0.47
0.86
0.09
0.12
0.49
0.38
2.03
2.59
4.85
2.47
2.45
0.79
Myristic acid
Palmitic acid
Stearic acid
Coconut oil
Fig. 8. Characterization of Ru-Sn/SiO2 catalyst by (a) N2 adsorption and
99.5
desorption isotherms, (b) XRD pattern, (c) TEM, and (d) TEM-Mapping.
Reaction conditions: substrate (1.0 g), Ru3Sn7/SiO2 catalyst (0.2 g),
dodecane (80 mL), 240 ˚C, 4 MPa H2, 120-240 min., stirring at 700 rpm.
2.4 The extension of the substrate and coconut oil
hydrogenation to fatty alcohols
Conversion
100
Alcohols Yield
Alkanes Yield
Ester Yield
Having ascertained the composition of the most active
catalyst (Ru3Sn7), we explored the effect of hydrogen pressure
(ranging from 2 to 4 MPa) towards product selectivity (Fig S5a).
We were gratified to find that on increasing the pressure from
2 to 4 MPa, the selectivity increased from 48.9% to 99.2%,
suggesting that relatively high hydrogen pressures favor the
equilibrium for octadecanal hydrogenation to the side of
reduced product, i.e., stearyl alcohol. As the H2 pressure
increased up to 5 MPa, the selectivity towards stearyl alcohol
was nearly constant at 100%. Then, we investigated the
influence of temperature (from 220-280 °C) on the
hydrogenation of stearic acid under identical conditions (Fig.
S5b). The reaction rates increased from 1.74 to 4.51 g·g−1·h−1
within the 220–280 °C tested temperature range, representing
in the apparent activation energy of 36 kJ·mol−1 (Fig. S5c).
Therefore, 240°C and 4 MPa H2 were selected as the optimal
temperature and hydrogen pressure for the best catalytic
performance.
80
60
40
20
0
0
1
2
Time (h)
3
4
Fig. 9. Kinetics of coconut oil conversion over Ru3Sn7/SiO2 catalyst. Reaction
conditions: coconut oil (1.0 g), Ru3Sn7/SiO2 (0.2 g, Ru loading: 1.5 wt%, Sn
loading: 4.1 wt%), dodecane (80 mL), 240°C, 4 MPa H2, stirring at 700 rpm.
Finally, the optimized catalyst and reaction conditions
were employed for the hydrogenation of coconut oil (Fig. 9),
which took place with nearly 100% conversion and selectivities
of 96.6% C8-C18 fatty alcohol, 3% C8-C18 alkane and 0.4%
heavier ester in 240 min. The gained fatty alcohol formation
rate was 0.79 g·g-1·h-1. The fatty alcohol composition closely
Having developed a catalyst system for the efficient
reduction of stearic acid to the corresponding fatty alcohol,
the reduction of a series of C12-C18 fatty acids (lauric acid,
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7

Please do not adjust margins
Catalysis Science & Technology
Page 8 of 12
View Article Online
DOI: 10.1039/C8CY00037A
ARTICLE
Catalysis Science & Technology
mirrored the fatty acid composition of raw coconut oil (for a
First-principles density functional theory (DFT)
detailed composition analysis, see Fig. S6), with the major calculations were carried out to gain deeper insights into the
components being C12 lauryl alcohol (49.7%) and C14 myristyl structure of the active catalyst and the mechanism for the
alcohol (18.0%). The alkanes that were obtained originated catalytic hydrogenation reaction with the RuSn bimetallic
from hydrodeoxygenation of fatty alcohols.
catalyst (see SI for further computational details). First, to
In addition, the stability on Ru3Sn7/SiO2 catalyst for elucidate the reason for the formation of Ru3Sn7 species as the
coconut oil hydrogenation was tested by four consecutive runs, major alloy in the RuSn catalyst, even when the Sn content
which attained stable activities and selectivity (Fig. S7). The used during preparation of the catalyst led to a Ru/Sn ratio
reused catalysts after four runs were subsequently different from 3/7 (see Fig. 6c), we calculated the formation
characterized by TEM, XRD, and XPS measurements. The TEM energy of the three possible RuSn alloy (Ru1Sn2, Ru2Sn3 and
[35, 36]
images and XRD patterns (Fig. S8 and Fig. S9) showed that Ru3Sn7) from which crystal structure data are available
.
Ru3Sn7/SiO2 was only slightly sintered with the particle size of The DFT calculations showed that the Ru3Sn7 alloys had the
ca. 5 nm after four consecutive runs. The comparable XPS lowest formation energy at −21.2 kJ·mol−1·atom−1, compared
spectra before and after recycling runs demonstrated that the with −15.4 kJ·mol−1·atom−1 and −18.3 kJ·mol−1·atom−1 for the
Ru and Sn states on Ru3Sn7/SiO2 catalyst were almost not Ru1Sn2, and Ru2Sn3 alloys respectively, thus leading Ru3Sn7 to
changed (Fig. S10). A highly durable of Ru3Sn7/SiO2 catalyst is be preferentially formed in the RuSn bimetallic catalyst, which
thus demonstrated.
is entirely consistent with our experimental observations. The
negative formation energies of the three alloys also indicate
that the Ru and Sn metal components preferentially combine
to form RuSn alloys rather than remain as separate metals
during the preparation of RuSn catalyst.
2.5 DFT calculations
Fig. 10. Potential energy diagrams and optimized geometries of all reaction intermediates and transition states for different steps (a–d) in the catalytic
hydrogenation reaction of acetic acid to ethanol on Ru3Sn7 (111) surface. Energies (kJ·mol−1) include zero-point-energy correction. Only the upper two
layers of Ru3Sn7 (111) surface are shown in the diagrams. Color scheme: yellow: Ru; cyan: Sn, grey: C, red: O, and white: H.
Table 3. Calculated reaction energies (
△
interaction energies between coadsorbates for different steps in the the RuSn alloy, the mechanism of the hydrogenation reaction
H), energy barriers (Ea) and
To understand the source of the high reaction activity of
hydrogenation of acetic acid to ethanol on Ru3Sn7 (111) surface.
on Ru3Sn7 (111) surface, which is the major crystal surface
determined by the experiment in this work, was studied by
DFT calculations and compared with that on Ru(0001) (the
surface structures are shown in Fig. S11).37 To simplify the
calculations, a model reaction based on the simpler but
analogous reactant acetic acid (CH3COOH) was studied. The
intermediates present during the hydrogenation of CH3COOH
to ethanol (CH3CH2OH) on the Ru3Sn7 (111) surface are shown
in Fig. 10. In the first step, CH3COOH adsorbs on the Ru3Sn7
△
H
Ea
Einter
Step Reaction
(kJ/mol)
-101.3
-21.3
(kJ/mol)
(kJ/mol)
CH3COOH + * → CH3COOH*
/
/
1
2
3
4
5
CH3COOH* → CH3CO* + OH*
CH3CO* + H* → CH3CHO*
CH3CHO* + H* → CH3CH2O*
81.0
76.2
21.3
74.3
/
-87.8
12.5
-1.9
-2.9
-16.4
CH3CH2O* + H* → CH3CH2OH* -36.7
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 9 of 12
View Article Online
DOI: 10.1039/C8CY00037A
Catalysis Science & Technology
ARTICLE
(111) surface via Sn-O interactions at the two top sites of
adjacent surface Sn atoms, with an adsorption energy of
To explore the better catalytic capability of Ru3Sn7 (111)
−101.3 kJ·mol−1 and the average Sn-O distance of 3.62 Å. surface, the Bader electron analysis calculations on the stable
According to the Bader electron analysis results, on the Ru3Sn7 adsorption structures of different reaction intermediates were
(111) surface, Ru atoms possess negative charge of – 0.64 |e| carried out (Table 4 and Fig. 11). As shown in Table 4, the
in average, while Sn atoms have positive charge of +0.34 |e| in calculation results suggest that, in the whole reaction, the Ru
average. Thus, the adsorption of CH3COOH molecule to the site maintains negative charges ranging from -0.30 |e| to -0.57
Ru3Sn7 (111) surface is mainly due to the electrostatic |e| while the Sn site keeps positive charges ranging from +0.34
interactions between the molecule and the positively charged |e| to +0.67 |e|. Considering the electrostatic interactions
surface Sn atoms. The reduction of CH3COOH* begins with the between the reaction intermediates and the surface active
breaking of the carbonyl carbon-hydroxyl oxygen (C-O) bond, sites on the Ru3Sn7 (111) surface, the negatively charged Ru
with an energy barrier of 81.0 kJ·mol−1, yielding adsorbed site can help facilitating the adsorption of H* species on Ru
CH3CO* and OH* intermediates. The CH3CO* species adsorbs atom, reducing the adsorption strength of CH3CH2O* species
on the Ru3Sn7 (111) surface with the Sn-O and Ru-C bond and thus lowing the barrier of the hydrogenation of CH3CH2O*,
lengths of 2.30 and 1.99 Å, respectively, while the OH* species while the positively charged Sn site is propitious to the
adsorbs at the bridge site between two surface-Sn atoms with adsorption of CH3COOH molecule and helps stabilizing the O-
an average Sn-O bond length of 2.32 Å. Then, the intermediate containing intermediates. The significant electron transfer can
Sn-bound CH3CO* species undergoes a three-step successive also be confirmed from the charge density difference (CDD)
hydrogenation process to form the final ethanol product plots of different adsorption intermediates (CH3CO*, CH3CHO*,
involving two hydrogenated intermediates of CH3CHO* and CH3CH2O* and CH3CH2OH*) on Ru3Sn7 (111) surface as shown
CH3CH2O*. Our calculations of the adsorption of H* species on in Fig. 11. For example, the CDD plot of co-adsorbed CH3CO*
Ru3Sn7 (111) surface suggest that H* prefers to adsorb on the and OH* species (Fig. 11a) shows the electrons transfer from
Ru atom than the Sn atom (Figure S12), indicating that the the Sn atoms to the O atoms of the adsorbates and from the
hydrogenation reaction is more likely to happen on the Ru Ru atom to the
α
-C atom of CH3CO* species. These results
atom site. The first step involves hydrogen transfer to the acyl indicate the bonding relationships of Sn-O and Ru-
α
-C and
radical intermediate (CH3CO…H) followed by transfer to the suggest that the surface Sn atoms also take part in the catalytic
corresponding aldehyde (CH3CHO…H), yielding an ethoxy reaction by acting as a kind of surface catalytic sites rather
radical which abstracts
a third equivalent of hydrogen than just being a structure component. The Ru center bonded
(CH3CH2O…H). The energy barriers for these three steps are with α-C of CH3CO* and hydrogenated of such intermediate
76.2 kJ·mol−1, 21.3 kJ·mol−1, and 74.3 kJ·mol−1 (Table 3), with the dissociated H* to final alcohol with three steps of H*
respectively. The resulting CH3CH2OH is adsorbed above a addition.
layer of Ru atoms.
The rate-determining step in the reduction process on the
Ru3Sn7 (111) catalyst surface is the first dehydroxylation step
of CH3COOH*. In contrast, in the same reaction carried out on
a Ru (0001) surface, the third hydrogenation step is the rate-
determining step. In addition, the Ru3Sn7 alloy has a lower
energy barrier for the rate-determining step (81.0 kJ·mol−1 on
Ru3Sn7 (111) vs 123.5 kJ·mol−1 on Ru (0001)). Our calculations
suggest that the Ru3Sn7 (111) surface has a higher catalytic
activity for the hydrogenation reaction than pure Ru, which is
consistent with our experimental results.
Table 4. Calculated Bader charges of Ru and Sn atom at the
adsorption sites on Ru3Sn7 (111) surface with different
adsorbed reaction intermediates.
Fig. 11. The charge density difference plots of different adsorbed reaction
Adsorbates
Pure Ru3Sn7 (111) Surface
CH3COOH*
Charge of Ru Charge of Sn
intermediates on Ru3Sn7 (111) surface: (a) CH3CO*+OH*, (b) CH3CHO*, (c)
CH3CH2O* and (d) CH3CH2OH*. The blue isosurface corresponds to charge
gains and orange isosurface corresponds to the charge lost. Color scheme:
yellow: Ru; cyan: Sn, grey: C, red: O, and white: H.
- 0.64
- 0.57
- 0.43
- 0.42
- 0.38
- 0.37
- 0.30
- 0.32
- 0.47
+ 0.34
+ 0.40
+ 0.67
+ 0.64
+ 0.41
+ 0.39
+ 0.36
+ 0.36
+ 0.34
CH3CO*+OH*
CH3CO*+H*
CH3CHO*
CH3CHO*+H*
CH3CH2O*
3. Experimental Section
CH3CH2O*+H*
CH3CH2OH*
Synthesis of the supported metal catalysts
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9

Please do not adjust margins
Catalysis Science & Technology
Page 10 of 12
View Article Online
DOI: 10.1039/C8CY00037A
ARTICLE
Catalysis Science & Technology
The metal catalysts supported on SiO2 were prepared by
4. Conclusions
incipient wetness impregnation. The metal precursor was
firstly dissolved in H2O to afford a solution. Then aqueous was
slowly dropped onto SiO2, and kept continuous stirring at
ambient temperature overnight. Afterwards, the as-formed
catalysts were filtered, dried at 80 °C overnight. Finally, it was
calcined in flowing air (flowing rate: 100 mL·min-1) at 350-
460 °C for 4 h, and reduced in flowing hydrogen (flowing rate:
100 mL·min-1) at 250-600 °C for 4 h.
With respect to the promising hydrotreating catalyst, the
earlier reports considered that Ru…O=Sn (Ru electronically
interacts with Sn oxides) was the active site for fatty acid and
fatty ester hydrogenation, but here in this contribution we
found that Ru3Sn7 nanoclusters as the active catalytic species.
The interaction between Snδ+ and Ru0 species were not found
because of the unchanged BE vaules of Snδ+ in XPS spectra,
suggesting that SnOx may exist as the isolated species.
However, the formation of Ru3Sn7 alloy nanoclusters led to an
obvious BE shift at Ru0 and Sn0 XPS spectra. Our catalytic tests
confirmed that the optimal Sn/Ru ratio during preparation of
the catalyst achieving the highest fatty alcohol yield was 7/3.
The most abundant Ru3Sn7 nanoclusters reached the highest
rate of 2.45 g·g−1·h−1, while avoiding ester formation at a rate
of 0.31 g·g−1·h−1 on catalysts formed with a Sn/Ru ratio ˃ 7/3
due to the presence of SnO2/SiO2 species. To control the
nature of the active species during preparation of the catalyst,
the higher reduction temperature at around 460°C suppressed
Sn/SiO2 formation, while Sn/SiO2 would catalyze the formation
of by-product ester at a rate of 0.88 g·g−1·h−1. Thus, supported
Ru-Sn bimetallic catalyst consisting of Ru3Sn7 nanoclusters as
the active sites, was highly active to hydrogenate of C12-C18
fatty acids and coconut oil to the corresponding fatty alcohols
in near quantitative conversion in a non-polar solvent and
under moderate conditions (240°C, 4 MPa H2).
Catalyst characterization
The specific surface area and pore size distribution were
performed by N2 adsorption at 77 K on a BELSORP-MAX
instrument after activating the samples under vacuum at 573 K
for 10 h. Powder X-ray diffraction (XRD) patterns were
measured on Rigaku Ultima IV X-ray diffractometer utilizing
Cu-Kα radiation (λ = 1.5405 Å) operated at 35 kV and 25 mA.
Scanning electron microscopy (SEM) equipped with energy
dispersive X-ray (EDX) analysis unit (Oxford, UK) was operated
on
a Hitachi S-4800 microscope to illuminate crystal
morphology and size. Transmission electron microscopy (TEM)
images were performed on a FEI Tecnai G2 F30 microscope
working at 300 kV.
X-ray photoelectron spectroscopy (XPS) were performed
with Al Kα (hν = 1486.6 eV) radiation on a Thermo Scientific K-
Alpha spectrometer. Charging effects were corrected by using
the C 1s peak due to adventitious carbon with EB fixed at
284.8 eV.
Density functional theory (DFT) calculations showed the
rate-determining step in the reduction process on the Ru3Sn7
(111) catalyst surface was the first dehydroxylation step of
CH3COOH*, and by comparison, the third hydrogenation step
was the rate-determining step on a Ru (0001) surface. In
accordance with our experimental results, the Ru3Sn7 alloy
showed a lower energy barrier for the rate-determining step
(81.0 kJ·mol−1 on Ru3Sn7 (111) vs 123.5 kJ·mol−1 on Ru (0001)).
In addition, the CDD plot of co-adsorbed CH3CO* and OH*
species demonstrated the electrons transferred from the Sn
atoms to the O atoms of the adsorbates and from the Ru atom
to the α-C atom of CH3CO* species (bonding relationships of
Sn-O and Ru-C). This suggested that the surface Sn atoms took
part in the catalytic reaction as a surface adsorption sites as
well as a Ru3Sn7 structure component, while Ru bonded with
α-C of CH3CO* and hydrogenated of such intermediate with
the dissociated H* to alcohol products, leading to Ru3Sn7
nanoclusters as an excellent hydrogenation catalysts for
converting of esters and acids.
The temperature programmed reduction (TPR) analysis of
calcined catalysts was performed by using Micromeritics
tp5080 apparatus equipped with a thermal conductor detector
(TCD), using a 5% H2/He mixture (flowing rate: 20 mL·min−1)
and a heating rate of 5 K·min−1.
The IR spectra of adsorbed CO were measured on a
Nicolet Fourier transform infrared spectrometer (NEXUS 670).
The samples were activated in H2 at 673 K for 1 h, and then H2
was removed by Vacuum pump. After the temperature was
cooled to 313 K, the samples were exposed to CO until
adsorption saturation. Subsequently the samples were
evacuated at 313 K to remove the physically adsorbed CO.
Afterwards, the IR spectra of adsorbed CO were recorded until
no further changes in the spectra were observed.
Catalytic tests
We performed the experiments for conversion of stearic
acid and coconut oil as follows: firstly we added the mixture of
stearic acid or coconut oil (1.0 g) and catalyst (0.2 g) and the
solvent of dodecane (80 mL) into a Parr autoclave (300 ml).
Then the autoclave was firstly purged with N2 to remove the
residual air, and in a next step the reaction gas H2 (4 MPa) was
introduced into the Parr reactor at ambient temperature. The
reaction was carried out at 240 °C at a stirring speed of 600
rpm. For the kinetic study, the liquid products were collected
by in situ sampling every 30 min., and then analyzed by a
Shimadzu GC coupled with GC-MS.
Acknowledgements
This research was supported by National Natural Science
Foundation of China (Grant Nos. 21573075, 21373098), the
National Key Research and Development Program of China
(Grant No. 2016YFB0701100), the Recruitment Program of
Global Young Experts in China, and the Shanghai Pujiang
Program (Grant No. 14PJ1403500). Part of the computational
time is supported by the Performance Computing Center of
Jilin Province.
10 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Catalysis Science & Technology
Page 11 of 12
View Article Online
DOI: 10.1039/C8CY00037A
Catalysis Science & Technology
ARTICLE
28 A. M. Silva, O. A. A. Santos, M. J. Mendes, E. Jordão, M. A.
Fraga, Appl. Catal. A. 2003, 241, 155-165.
29 M. A. Sánchez, Y. Pouilloux, V. A. Mazzieri, C. L. Pieck, Appl.
Catal. A. 2013, 467, 552-558.
30 S. Eckle, Y. Denkwitz, R. J. Behm, J. Catal. 2010, 269, 255-
268.
Notes and references
1
2
3
J. Pritchard, G. A. Filonenko, R. van Putten, E. J. M. Hensen ,
E. A. Pidko, Chem. Soc. Rev. 2015, 44, 3808-3833.
G. Chelucci, S. Baldino and W. Baratta, Acc. Chem. Res.
2015, 48, 363-379.
B. Ma, J. Hu, Y. Wang, C. Zhao,Green Chem. 2015, 17, 4610-
4617.
31 S. Y. Chin, C. T. Williams, M. D. Amiridis, J. Phy. C. B. 2006,
110, 871-882.
4
5
B. Ma, C. Zhao, Green Chem. 2015, 17, 1692-1701.
B. Ma, X. Yi, L. Chen, A. Zheng, C. Zhao, J. Mater. Chem. A
32 V. M. Deshpande, K. Ramnarayan, C. S. Nnarasimhan, J.
Catal. 1990, 121, 165-173.
2016,
J. Zhang, C. Zhao, ACS Catal. 2016,
J. Zhang, C. Zhao, Chem. Commun. 2015, 51, 17249-17252.
B. Ma, H. Cui, D. Wang, P. Wu, C. Zhao, Nanoscale 2017, 9,
5986-5995.
4, 11330-11341.
33 V. M. Deshpande, K. Ramnarayan, C. S. Nnarasimhan, J.
Catal. 1990, 121, 174-182.
34 Y. Pouilloux, F. Autin, C. Guimon, J. Barrault, J. Catal. 1998,
176, 215-224.
35 L. Di, S. Yao, S. Song, G. Wu, W. Dai, N. Guan, L. Li, Appl.
Catal. B, 2017, 201, 137–149.
36 J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996,
77, 3865-3868.
6
7
8
6, 4512-4525
9
Y. Luo, Comprehensive Handbook of Chemical Bond
Energies, CRC Press, London, 2007
10 D. Spasyuk, S. Smith, D. G. Gusev, Angew. Chem. Int. Ed.
2012, 51, 2772-2775.
37 O. Schwomma, H. Nowotny, A. Wittmann, Monatsh. Chem.
1964, 95, 1538-1543.
11 T. V. Stein, M. Meuresch, D. Limper, M. Schmitz, M.
Holscher, J. Coetzee, D. J. Cole-Hamilton, J. Klankermayer,
W. Leitner, J. Am. Chem. Soc. 2014, 136, 13217-13225.
12 P. Dupau, M. L. Tran Do, S. Gaillard, J. L. Renaud, Angew.
Chem. Int. Ed. 2014, 53, 13004-13006.
13 S. Werkmeister, K. Junge, B. Wendt, E. Alberico, H. Jiao, W.
Baumann, H. Junge, F. Gallou, M. Beller, Angew. Chem. Int.
Ed. 2014, 53, 8722-8726.
14 S. Chakraborty, H. Dai, P. Bhattacharya, N. T. Fairweather,
M. S. Gibson, J. A. Krause, H. Guan, J. Am. Chem. Soc. 2014,
136, 7869-7872.
15 Y. Hattori, K. Yamamoto, J. Kaita, M. Matsuda, S. J. Yamada,
Am. Oil. Chem. Soc. 2000, 77, 1283-1288.
16 E .K. Poels, D. S. Brands, Appl. Catal. A. 2000, 191, 83-96.
17 D. S. Brands, G. U-A-Sai, E. K. Poels, A. Bliek, J. Catal. 1999,
186, 169-180.
18 L. Wu, L. Li, B. Li, C. Zhao, Chem. Commun. 2017, 53, 6152-
6155.
19 T. Miyake, T. Makino, S. Taniguchi, H. Watanuki, T. Niki, S.
Shimizu, Y. Kojima, M. Sano, Appl. Catal. A. 2009, 364, 108-
112.
20 S. Koso, H. Watanabe, K. Okumura, Y. Nakagawa, K.
Tomishige, Appl. Catal. B-Enviro. 2012, 111, 27-37.
21 (a) M. C Chia, Y. J. Pagan-Torres, D. Hibbitts, Q. Tan, H. N.
Pham, A. K. Datye, M. Neurock, R. J. Davis, J. A. Dumesic, J.
Am. Chem. Soc. 2011, 133, 12675-12689; (b) J. Pritchard, G.
A. Filonenko, R. V. Putten, E. J. M. Hensen, E. A. Pidko,
Chem. Soc. Rev., 2015, 44, 3808—3833.
22 (a) M. J. Mendes, O. A. A. Santos, E. Jordão, A. M. Silva,
Appl. Catal. A. 2001, 217, 253-262; (b) Z. Z. Qing, L. Z. Hong,
L. Bin, G. S. Zhuo, Appl. Catal. A. 2006, 302, 208-214.
23 (a) F. C. A. Figueiredo, E. Jordão, W. A. Carvalho, Appl.
Catal. A. 2008, 351, 259-266; (b) S. A. S. Corradini, G. G.
Lenzi, M. K. Lenzi, C. M. F. Soares, O. A. A. Santos, J Non-
Cryst Solids., 2008, 354, 4865-4870.
24 (a) R. Luque, J. H. Clark, K. Yoshida, P. L. Gai, Chem.
Commun. 2009, 35, 5305-5307; (b) D. A. Echeverri, J. M.
Marin, G. M. Restrepo, L. A. Rios, Appl. Catal. A. 2009, 366
342-347.
,
25 (a) C. Huang, H. Zhang, Y. Zhao, S. Chen, Z. Liu, J. Colloid
Interface Sci. 2012, 386, 60-65; (b) M. A. Sánchez, G. C.
Torres, V. A. Mazzieri, C. L. Pieck, J. Chem. Technol.
Biotechnol., 2017, 92, 27-42.
26 M. A. Sánchez, V. A. Mazzieri, M. Oportus, P. Reyes, C. L.
Pieck, Catal. Today 2013, 213, 81-86.
27 R. D. Adams, E. M. Boswell, B. Captain, A. B. Hungria, P. A.
Midgley, R. Raja, J. M. Thomas, Angew. Chem. Int. Ed. 2007,
46, 8182-8185.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11

Please do not adjust margins
Catalysis Science & Technology
Page 12 of 12
View Article Online
DOI: 10.1039/C8CY00037A
ARTICLE
Catalysis Science & Technology
Table of Content:
12 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins