Insight into the Correlation between Cu Species Evolution and Ethanol Selectivity in the Direct Ethanol Synthesis from CO Hydrogenation

Article abstract of DOI:10.1002/cctc.201801888

Cu/SiO₂ catalyst was prepared by the ammonia evaporation method for the direct synthesis of ethanol from CO hydrogenation. The catalyst exhibited the initial ethanol selectivity as high as 40.0 wt %, which dramatically decreased from 40.0 to 9.6 wt % on the stream of 50 h. XRD, XPS, TEM and N₂O titration techniques were employed to elucidate the ethanol selectivity change and catalyst structure evolution during reaction process. The experiment and characterization results indicated that both Cu⁺/(Cu⁺+Cu⁰) value and copper crystallite size had great effects on the ethanol selectivity. During the initial 38 h, the ethanol selectivity obviously decreased from 40.0 to 18.2 wt %, and Cu⁺/(Cu⁺+Cu⁰) value on the catalyst surface rapidly dropped from 0.67 to 0.39, while the copper crystallite size remained almost unchanged. However, during the reaction period of 38–50 h, the Cu⁺/(Cu⁺+Cu⁰) value possessed no distinct change, but a further decrease in ethanol selectivity and a rapid aggregation in Cu particles were observed simultaneously. The present systematic investigation demonstrated that the decrease of Cu⁺/(Cu⁺+Cu⁰) value was the main factor for the loss of ethanol selectivity during the initial 38 h, whereas the rapid growth of Cu particles during the reaction period of 38–50 h were mainly contributed to the further decline of ethanol selectivity.

Full text of DOI:10.1002/cctc.201801888

Accepted Article
Title: Insight into the correlation between Cu species evolution and
ethanol selectivity in the direct ethanol synthesis from CO
hydrogenation
Authors: Xiaoli Li, Guohui Yang, Meng Zhang, Xiaofeng Gao,
Hongjuan Xie, Yunxing Bai, Yingquan Wu, Junxuan Pan, and
Yisheng Tan
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201801888
Link to VoR: http://dx.doi.org/10.1002/cctc.201801888
A Journal of
www.chemcatchem.org

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
Insight into the correlation between Cu species evolution and
ethanol selectivity in the direct ethanol synthesis from CO
hydrogenation
Xiao-Li Li,[a,b] Guo-Hui Yang,[a] Meng Zhang,[a,b] Xiao-Feng Gao,[a,b] Hong-Juan Xie,[a]
Yun-Xing Bai,[a, b] Ying-Quan Wu,[a] Jun-Xuan Pan,[a] and Yi-Sheng Tan[a,c]
*
Cu/SiO2 catalyst was prepared by the ammonia evaporation method
for the direct synthesis of ethanol from CO hydrogenation. The
catalyst exhibited the initial ethanol selectivity as high as 40.0 wt%,
which dramatically decreased from 40.0 to 9.6 wt% on the stream of
50 h. XRD, XPS, TEM and N2O titration techniques were employed
to elucidate the ethanol selectivity change and catalyst structure
evolution during reaction process. The experiment and
characterization results indicated that both Cu+/(Cu++Cu0) value and
copper crystallite size had great effects on the ethanol selectivity.
During the initial 38 h, the ethanol selectivity obviously decreased
from 40.0 to 18.2 wt%, and Cu+/(Cu++Cu0) value on the catalyst
surface rapidly dropped from 0.67 to 0.39, while the copper
crystallite size remained almost unchanged. However, during the
reaction period of 38-50 h, the Cu+/ (Cu++Cu0) value possessed no
distinct change, but a further decrease in ethanol selectivity and a
rapid aggregation in Cu particles were observed simultaneously. The
present systematic investigation demonstrated that the decrease of
Cu+/ (Cu++Cu0) value was the main factor for the loss of ethanol
selectivity during the initial 38 h, whereas the rapid growth of Cu
Typically, suitable catalysts for higher alcohols synthesis
from CO hydrogenation can be divided into the following types:
the modified Cu-based methanol synthesis catalysts [7, 8], the
modified Fischer-Tropsch synthesis catalysts [9], the Mo-based
catalysts [10-12] and the Rh-based catalysts [13]. Among these
catalysts, the Cu-based catalysts have been widely investigated
because of their high activity and low price [3, 14-17]. It was
found that the ethanol selectivity could achieve ~ 40 Cmol% in
the direct ethanol synthesis from CO hydrogenation over Cu-
based catalysts [7, 9-11, 14-17]. Unfortunately, the previous Cu-
based catalysts employed for direct synthesis ethanol from CO
hydrogenation generally suffered a rapid deactivation, especially
the decline in ethanol selectivity, which limited its further large-
scale application.
It is generally thought that Cu dispersion and/or the
distribution of Cu0 and Cu+ species have great influence on the
product selectivity during the hydrogenation process over Cu-
based catalysts [18-26]. Tuning catalyst preparation parameters
(e.g. promoters [18-22], Cu loading [23-26], calcination
temperature [24]) of Cu-based catalysts is really efficient to
modify Cu dispersion and Cu+/(Cu0 + Cu+) value. According to
the previous literatures, it is plausible that both more dispersed
Cu species and higher Cu+/(Cu0+Cu+) value are effective to
improve the product selectivity. However, since the catalyst
preparation process is tedious, the above-mentioned catalyst
system is too complicated to obtain a clear and intrinsic
explanation on the evolution of catalyst structure during the
reaction process.
particles during the reaction period of 38-50
h were mainly
contributed to the further decline of ethanol selectivity.
Introduction
Ethanol is widely used as fuel, fuel additive, and even a
potential substitute for gasoline [1-2]. Besides, it is also an
important intermediate for value-added fine chemicals [3-4]. With
increasing concerns for environmental pollution and depletion of
nonrenewable petroleum resources, the direct synthesis of
ethanol from CO hydrogenation, since its high atom economy
and low waste emission, has attracted growing attention in a few
decades [4-6].
To rule out the influence of catalyst preparation parameters
on catalyst structure, comparing the catalyst structure at the
reduced and spent stages is an alternative approach to clarify
the structure-performance relationship of Cu-based catalysts for
hydrogenation process. Sun et al. [27] pointed out that the
severe sintering and the formation of CoxC species on the used
catalyst caused the decline of activity. Sahibzada et al. [28]
studied the loss of Cu/ZnO/Al2O3 activity for methanol synthesis
from CO hydrogenation. They demonstrated a similar conclusion
that the decrease in methanol selectivity was resulted from the
a
State Key Laboratory of Coal Conversion, Institute of Coal Chemistry,
Chinese Academy of Sciences, Taiyuan 030001, China
b University of Chinese Academy of Sciences, Beijing 100049, China
c National Engineering Research Center for Coal-Based Synthesis, Institute of
Coal Chemistry, Chinese Academy of Sciences, Taiyuan 030001, China
E-mail addresses: tan@sxicc.ac.cn; Tel. / fax: +086-351-4044287
This article is protected by copyright. All rights reserved.

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
sintering of Cu particles. Shen et al. [29] found that the Cu/SiO2
catalyst showed outstanding activity with a total selectivity of
95 % for ethanol and methanol during dimethyl oxalate (DMO)
hydrogenation. It was found that the average size of copper
particles kept unchanged in the spent catalyst, but the
Cu+/(Cu0+Cu+) ratio increased during the course of reaction. By
comparing the reduced and used catalysts, Zhao et al. [30]
reported that the activity loss of ethanol formation from DMO
over Cu/SiO2 catalysts was mainly attributed to the destruction
of suitable surface Cu+ and Cu0 distribution and to the decline of
Cu dispersion. T. Williams et al. [31] also found that the
imbalance of Cu+ and Cu0 species and crystallite aggregation
resulted in rapid decrease of ethanol selectivity during DMO
hydrogenation. Comparing the structure changes on the reduced
and spent catalysts provided a superficial understanding of the
relation between structure evolution and performance of Cu-
based catalysts. However, the main factor, which influenced the
performance (activity and/or selectivity) change, was still
equivocal and controversial.
investigated by the characterization techniques, including X-ray
diffraction (XRD), X-ray photoelectron spectroscopy (XPS),
Transmission electron microscope (TEM) and N2O titration.
Furthermore, the correlation between the structure evolution and
the decrease of ethanol selectivity during reaction process was
systematically discussed.
Results and Discussion
Performance of Cu/SiO2 catalyst
The performance of Cu/SiO2 catalyst for ethanol synthesis
via CO hydrogenation was evaluated and the results were
presented in Fig. 1. The total alcohols selectivity showed a
decrease trend (from 30.2 to 23.9 %) with reaction time.
Obviously, the primary products over Cu/SiO2 catalyst were
methanol and ethanol (~98.9 wt%) during the whole reaction
process, while the selectivity of ethanol sharply decreased from
40.0 to 9.6 wt%, simultaneously, methanol selectivity
dramatically increased from 58.9 to 90.0 wt%. Here, in order to
investigate the relation between the structure evolution and loss
of ethanol selectivity during reaction, the Cu/SiO2 catalyst at
different stages was studied by XRD, XPS, TEM and N2O
titration.
Based on the above literatures, we are convinced that
investigating the correlation between Cu species evolution and
performance over a simplified catalytic system is vital to explore
the key factor on ethanol selectivity change during the reaction
process. In heterogeneous catalysis, metals or metal oxides are
traditionally dispersed as nanoscale particles on a support with a
large surface to maximize the number of exposed active sites
[32]. As reported [33-36], copper species are much easier to be
highly dispersed on Al2O3 or SiO2 support with large surface in
comparison to the TiO2 support with a small surface. Despite
that Al2O3 support has been used for widespread applications,
the relatively high quantity of acid sites on Al2O3 surface may
catalyze unwanted isomerization or oligomerization reactions
according to many studies [28, 32, 36, 38]. Therefore, in the
present study, an inert SiO2 support with a large surface was
considered as a suitable choice [20-23, 39].
In this work, we prepared a simple Cu/SiO2 catalyst using
the ammonia evaporation method and evaluated its performance
for ethanol synthesis from CO hydrogenation. The ethanol
selectivity at initial 14.5 h over the catalyst achieved to 40.0 wt%
but sharply decreased to 9.6 wt% when the reaction time was
enhanced to 50 h. In the whole reaction process, the major
alcohol products were methanol and ethanol (above 98.9 wt%).
To elucidate the reason for the decline of ethanol selectivity, the
evolution of Cu/SiO2 catalyst structure at different stages was
Fig. 1. Performance of Cu/SiO2 catalyst for CO hydrogenation as a function of
reaction time
Reaction conditions: T = 300 oC, P = 3 MPa, GHSV = 5000 h-1, H2/CO = 2.5.
Catalyst characterization
Fig. 2 showed the XRD patterns of CS-calcined, CS-
reduced, CS-reaction-14.5 h, CS-reaction-28 h, CS-reaction-38
h and CS-reaction-50 h, in which the diffraction peak at around
22o came from amorphous silica [34]. The XRD pattern of the
CS-calcined sample showed two strong diffraction peaks at 2θ of
This article is protected by copyright. All rights reserved.

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
35.6 and 38.7o along with two weak peaks at 48.8 and 61.5o,
which were attributed to the typical peaks of CuO (JCPDS no.
05-0661) [40-42]. For CS-reduced sample, the peaks of CuO
phase almost disappeared, accompanied by the observable
Cu2O peaks at 36.6 and 42.4o (JCPDS no. 034-1354). It
indicated that most of CuO phase could be reduced to Cu+
species after reduced in the fixed-bed reactor, consistent with
the previous report by Zhao et al. [30]. It was found that the
peaks of Cu2O were dramatically weakened with the
enhancement of reaction time until they were vanished when the
reaction was evaluated to 50 h. Meanwhile, three weak
diffraction peaks at 2θ of 43.3o, 50.4o and 74.1o for Cu (JCPDS
no. 04-0836) emerged in CS-reaction-14.5 h sample [19]. The
intensities of Cu diffraction peaks remained almost unchanged in
the first 38 h but displayed an obvious growing with a further
increase of reaction time. The significant sharpening of Cu
diffraction peaks indicated that Cu particles possessed a severe
aggregation during the reaction process, which is probably
because that the reaction temperature (300 oC) in this work is
Table 1 Crystallite sizes of CuO, Cu2O and Cu over CS-calcined, CS-reduced,
CS-reaction-14.5 h, CS-reaction-28 h, CS-reaction-38 h and CS-reaction-50 h
catalysts
Crystallite sizes (nm)[a]
Samples
CuO
Cu2O
-
Cu
-
CS-calcined
CS-reduced
14.4
5.0
18.1
9.7
6.5
5.9
-
-
CS-reaction-14.5 h
CS-reaction-28 h
CS-reaction-38 h
CS-reaction-50 h
-
-
-
-
13.7
14.9
15.8
27.3
[a] CuO, Cu2O and Cu crystallite sizes calculated by the Scherrer formula
o
higher than the Hüttig temperature of Cu (134 C) [43] and the
CO hydrogenation to ethanol is an exothermic reaction. From
the calculation results based on the Scherrer equation (Table 1),
the Cu crystallite size showed a quite slight growth before 38 h
reaction and markedly increased from 15.8 to 27.3 nm when the
reaction time increased from 38 to 50 h. These results clearly
revealed that the CuO phase could transform into Cu2O during
the reduction process and the Cu2O species would be further
gradually reduced into metallic Cu during the CO hydrogenation
process [26, 30].
Fig. 3. Cu 2p XPS spectra of catalysts: (a) CS-calcined (b) CS-reduced
(c) CS-reaction-14.5 h (d) CS-reaction-28 h (e) CS-reaction-38 h
(f) CS-reaction-50 h
XPS analysis was carried out to elucidate the chemical
states of Cu species. As displayed in Fig. 3, for the CS-calcined
sample, the Cu 2p3/2 peak at ~933.0 eV, Cu 2p1/2 peak at ~954
eV and the satellite peak at 942-944 eV could be observed,
indicating that copper existed in the oxidation state of +2 [44].
Considering the asymmetry of the Cu 2p3/2 envelope, the peak of
CS-calcined sample could be deconvoluted into two
contributions centered around 933.4 and 935.6 eV, which were
corresponding to CuO and copper phyllosilicate, respectively
[42]. After the sample was reduced, the peak at 933.4 eV of CuO
shifted toward lower BEs (932.7eV), revealing that the CuO
phase was reduced to a low valence state of +1 or 0. However,
the peak at 935.6 eV of copper phyllosilicate and the satellite
peak at 942-944 eV were still detectable in the CS-reduced
Fig. 2. XRD patterns of catalysts: CS-calcined, CS-reduced, CS-reaction-14.5
h, CS-reaction-28 h, CS-reaction-38 h and CS-reaction-50 h
o
sample, confirming that a temperature of 300 C was insufficient
This article is protected by copyright. All rights reserved.

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
to reduce the CS-calcined sample due to the strong interaction
between silica support and Cu species. It was found that the
characteristic peaks of copper phyllosilicate (935.6 eV) and Cu
2p satellite peak (942-944 eV) apparently weakened along with
the reaction time and disappeared at 50 h. It suggested that
most of Cu2+ species were reduced to Cu+ or Cu0 when reaction
time was 50 h. These results revealed that the copper valence
transition occurred during CO hydrogenation.
Cu+/(Cu++Cu0) value dramatically decreased from 0.67 to 0.39
within first 38 h and tended to be stable after 38 h. These results
indicated that the chemical states of copper species in Cu/SiO2
catalyst underwent changes during reaction, which directly
affected the Cu+/(Cu++Cu0) value.
The TEM images of CS-calcined, CS-reduced, CS-reaction-
14.5 h, CS-reaction-28 h, CS-reaction-38 h and CS-reaction-50
h were given in Fig. 5. It was easily observed that the metal
nanoparticles were distributed uniformly on the silica surface for
the catalyst after calcined. As shown in Fig. 5, the particle
diameter distribution possessed the mean particle size of
approximately 5.2 nm on the CS-reduced sample. In addition,
the average crystallite size did not obviously increase (5.0 ~ 5.8
nm) before 38 h, while the particle size grew significantly from
5.8 to 8.9 nm when the reaction time increased from 38 to 50 h.
These results were in good agreement with XRD.
The Cu0 surface areas (SCu(0)) of CS-reaction-14.5 h (187.4
m2/g-Cu), CS-reaction-28 h (193.1 m2/g-Cu), CS-reaction-38 h
(186.9 m2/g-Cu) and CS-reaction-50 h (116.0 m2/g-Cu) catalysts
were measured using N2O titration. During the initial 38 h, the
SCu(0) value (~ 190 m2/g-Cu) remained almost unchanged.
Interestingly, Cu particle size did not obviously change before 38
h, as revealed by XRD and TEM results. Nevertheless, the SCu(0)
value clearly decreased from 186.9 to 116.0 m2/g-Cu when the
reaction time increased from 38 to 50 h, resulted from the
aggregation of Cu particles [3, 20, 26, 30, 43]. Consistently, the
above-discussed XRD and TEM results showed the similar
variation trend during the reaction period of 38-50 h. Thus, the
evolution of Cu particle size was systematically investigated
based on the aforementioned characterizations including XRD,
TEM and N2O titration.
Fig. 4. XAES spectra of catalysts: CS-calcined, CS-reduced, CS-reaction-14.5
h, CS-reaction-28 h, CS-reaction-38 h and CS-reaction-50 h
The amount of Cu+ and Cu0 could be determined via X-ray
auger electron spectra (XAES). In the XAES spectra of CS-
reduced sample (Fig. 4), the appearance of two obvious
overlapping peaks at around 570 and 574 eV was employed to
distinguish between Cu0 and Cu+ species [23, 45]. From the
deconvolution results (Fig. 4), the molar ratio of the surface
Cu+/(Cu0+Cu+) was significantly affected by reaction time. The
Fig. 5. TEM images of catalysts: (1) CS-calcined, (2) CS-reduced, (3) CS-reaction-14.5 h, (4) CS-reaction-28 h, (5) CS-reaction-38 h and (6) CS-reaction-50 h
This article is protected by copyright. All rights reserved.

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
Cu+/(Cu++Cu0) declined. When the Cu+/(Cu++Cu0) value was
relatively high, the high Cu+ species density around Cu0 species
could efficiently fix copper species and, hence, the particle size
remained unchanged. However, when Cu+/(Cu++Cu0) value was
below a certain value, due to the lower Hütting temperature of
Discussion of Cu species evolution
Fig. S1 showed the in situ XRD patterns of Cu/SiO2 catalyst
during the reduction and reaction process. In accordance with
the results displayed in Fig. 2, the XRD pattern of the CS-
calcined sample showed three strong diffraction peaks at 2θ of
o
Cu (134 C) [44, 46, 47] and the exothermic CO hydrogenation,
o
35.6, 38.7 and 48.8 , ascribed to the typical peaks of CuO
the low amount of Cu+ species in vicinity of metallic Cu could not
hamper the metallic Cu of migrating, consequently resulting in
the migration and aggregation of Cu particles.
(JCPDS no. 05-0661) [40-42]. During the reduction process, it is
easily observed that the intensities of CuO diffraction peaks
decreased gradually, accompanied by the observable Cu2O
peaks at 36.5 and 42.4o (JCPDS no. 034-1354). Moreover, after
reduction treatment, most of CuO phase could be reduced to
Cu+ species. When the catalyst was exposed under the H2/CO
mixture gas, the peaks of Cu2O were dramatically weakened
with the enhancement of operation time. Simultaneously, three
diffraction peaks at 2θ of 43.3o, 50.4o and 74.0o for Cu (JCPDS
no. 04-0836) emerged [19]. The intensities of Cu diffraction
peaks increased slightly in the initial reaction period but
displayed an obvious growth with time further increased, which
indicated that Cu particles possessed a severe aggregation.
These results were in good agreement with the discussion of the
conventional XRD.
Fig. 6. Schematic representation of the deactivation of Cu/SiO2 catalyst
Study on the correlation between structural
evolution and the change of ethanol selectivity
According to the XRD (including conventional and in situ
results) and XPS analysis, it was suggested that the Cu2O
species can be reduced into metallic Cu, thereby resulting in the
decrease of Cu+/(Cu++Cu0) value with reaction time. H. Oguchi
and his co-workers [46, 47] reported that Cu2O was more stable
than Cu metal. It revealed that, when the Cu+/(Cu++Cu0) value
was high, the high Cu+ species density around Cu0 species
could efficiently fix copper species and retard the particle growth.
However, when Cu+/(Cu++Cu0) value was below a certain value,
the metallic Cu migration was not able to be suppressed
because of the relatively low Cu+ species density. Thus, the
rapid aggregation of Cu particles occurred inevitably during the
reaction period of 38-50 h (XRD, TEM, N2O titration).
In the entire process of CO hydrogenation, the primary
products over Cu/SiO2 catalyst were methanol and ethanol
(above 98.9%). During reaction, the selectivity of ethanol sharply
decreased from 40.0 to 9.6 wt%, while methanol selectivity
dramatically increased from 58.9 to 90.0 wt%. According to the
characterization results, the CuO phase was observed in CS-
calcined sample, and most of CuO phase (Cu+ species)
transformed into Cu2O after reduction treatment and then the
Cu+ species could be further reduced to Cu0 species gradually
with reaction time (Fig. 2, 3). The Cu+/(Cu++Cu0) value on the
catalyst surface significantly dropped from 0.67 to 0.39 during
the initial 38 h and tended to be stable after 38 h (Fig. 4).
Conversely, the copper crystallite size did not change obviously
before 38 h, but it grew rapidly with the further increase of
reaction time (Fig. 2, 5). Here, a plot of ethanol selectivity,
Cu+/(Cu++Cu0) value and copper crystallite size versus reaction
time was presented in Fig. 7. It was obvious that both the
Cu+/(Cu++Cu0) value and copper crystallite size influenced the
ethanol selectivity. To further understand the correlation between
Based on the characterizations above, the proposed Cu
species evolution over Cu/SiO2 catalyst was displayed in Fig. 6.
As described in Fig. 6, the copper species were highly dispersed
on the silica surface. Cu2O phase (Cu+ species) was the main
copper species and high Cu+/(Cu++Cu0) value was detected in
the CS-reduced sample. After reduced, feed gas (CO + H2) was
introduced into the reactor. Cu2O (Cu+ species) gradually
reduced into metallic Cu with reaction time and thus the value of
This article is protected by copyright. All rights reserved.

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
Cu species evolution and ethanol selectivity change during the
reaction, Fig. 7 can be roughly divided into two stages: 0-38 and
38-50 h.
into ethanol (The results of in situ Fourier transform infrared
convinced our assumption and the relevant results and
discussion were put in supporting information.). When
Cu+/(Cu++Cu0) value decreased, as shown in Fig. 8, the
decrease in the amounts of Cu+ sites restrained the C-O
cleavage in CHxO, and the amounts of CHx species was bound
to decline, which inevitably resulted in a decrease in ethanol
selectivity. Meanwhile, the content of CHxO species increased
on the catalyst surface by limiting the C-O cleavage in CHxO.
Then, the CHxO species basically underwent a step-wise
hydrogenation to methanol. Thus, methanol selectivity
significantly improved with decreasing in the Cu+/(Cu++Cu0)
value (Fig. 1).
Fig. 7. Ethanol selectivity, Cu+/(Cu++Cu0) and Crystallite size as a function of
reaction time
During the initial 38 h, the ethanol selectivity decreased
from 40.0 to 18.2 wt%. The value of Cu+/(Cu++Cu0) declined
sharply, while the copper crystallite size kept almost unchanged.
This result revealed that the Cu+/(Cu++Cu0) value was closely
related to ethanol formation before 38 h. Ma et al. reported [23,
30, 44] that the Cu0 species boosted the H2 decomposition and
Cu+ species facilitated the C-O cleavage in CHxO by polarizing
the C=O bond to increase the content of CHx species, thus
favoring the formation of ethanol. The authors also pointed out
that the synergy effect of Cu+-Cu0 species required a suitable
amount of the two sites in catalyzing methyl acetate to ethanol.
In addition, in our previous studies on K-Cu/Al2O3 catalysts [37],
we found that the synergy of Cu+ and Cu0 species played an
important role in ethanol synthesis via CO hydrogenation.
Hence, the cooperation of Cu+ and Cu0 sites on the catalyst
surface should be considered as a key cause for the loss of
ethanol selectivity during the initial 38 h. A possible mechanism
of ethanol synthesis from CO hydrogenation was proposed in
this work (Fig. 8). As displayed in Fig. 8, the Cu0 species on the
Cu/SiO2 catalyst surface boosted the H2 decomposition and
provided H species [13, 23, 44, 48]. The CHxO species, formed
by CO hydrogenation, has been considered as the key reactive
intermediates involved in ethanol formation [49] and Cu+ species
as electrophilic or Lewis acidic sites were proposed to act
toward accelerating the C-O cleavage in CHxO to give CHx
species [13, 23]. Then, the CHxCO species formed by CO
insertion into CHx and terminated by H species, thereby turning
Fig. 8. Schematic representation of methanol and ethanol from CO
hydrogenation over Cu/SiO2 catalyst
During the reaction period of 38-50 h, the ethanol selectivity
still showed a continuing decrease, although the Cu+/(Cu++Cu0)
value did not change significantly, the value of Cu+/(Cu++Cu0)
after 38 h were so small that the low Cu+ species density in
vicinity of metallic Cu could not hamper the migration of Cu
particles (Cu2O was more stable than Cu metal [46, 47]).
Moreover, the exothermic CO hydrogenation also boosted the
accumulation of Cu particles. Thus, the further decrease in
ethanol selectivity may be caused by the aggregation of Cu
particles when the reaction time increased from 38 to 50 h (Fig.
7). In addition, a further increase in selectivity of methanol was
also observed during the reaction period of 38-50 h (Fig. 1).
Arena et al. [50] used Cu/Zn/Zr catalysts with Cu particle size
from 2 to 25 nm for methanol synthesis and found that selectivity
to methanol was favored over catalysts possessing larger
copper nanoparticles. It suggested that larger Cu particles after
This article is protected by copyright. All rights reserved.

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
38 h promoted methanol formation in our present experiment.
In the entire process of CO hydrogenation, it should be
noted that despite a continuous decrease in ethanol selectivity,
the main factor which led to the result was different during the
different stages. Specifically, highly dispersed CuO species on
the calcined catalyst were reduced into Cu2O phase (Cu+
species) after reduction treatment, accompanied by a high
Cu+/(Cu++Cu0) value, which was a crucial factor in ethanol
formation from CO hydrogenation. After reduced, when the
catalyst was exposed under the H2/CO mixture gas, Cu2O (Cu+
species) gradually transformed into metallic Cu with increased
time. Apparently, both the Cu+/(Cu++Cu0) value and ethanol
selectivity declined, however, the Cu particle size occupied a
stable state. When Cu+/(Cu++Cu0) value decreased to a certain
value, the low amount of Cu+ species in vicinity of metallic Cu
could not hamper the metallic Cu migration and thus resulted in
the aggregation of Cu particles, which inevitably further led to
the loss of ethanol selectivity.
value and copper crystallite size were considered to be
contributed to ethanol formation from CO hydrogenation within
50 h. In the beginning 38 h, the decrease in Cu+/(Cu++Cu0) value
led to the loss of ethanol selectivity. During the reaction from 38
to 50 h, the rapid growth of Cu particle further decreased in
ethanol selectivity. Thus, designing
a catalyst possessing
relatively small Cu particles and high Cu+/(Cu++Cu0) value may
be a promising prospect to improving ethanol selectivity and
stability from CO hydrogenation over Cu-based catalysts.
Experimental Section
Materials
The analytical-grade chemical Cu(NO3)2·3H2O (99%) was
purchased from the Beijing Chemical Co., Ltd., and used directly
without further purification. The employed SiO2 was purchased
from Aladdin Industrial Co.
The preparation of Cu/SiO2 catalyst
Conclusions
The Cu/SiO2 catalyst was prepared using the ammonia
evaporation method, briefly described as follows. 5.33 g of
Cu(NO3)23H2O was dissolved in 100.00 mL of ethanol. Then,
100.00 mL of 28 % ammonia aqueous solution was added and
stirred for 20 min. Subsequently, 10.00 g SiO2 was added into
the copper ammonia complex solution, stirred in ultrasonic wave
for 1 h at room temperature. After that, the suspension was
transferred into a Teflon-lined stainless autoclave and heated at
180 °C for 6 h. The obtained precipitate was collected, filtered,
washed with ethanol and deionized water carefully. Finally, the
mixture was dried at 120 oC for 10 h and the dried precursor was
Studying the relation between Cu species evolution and
ethanol selectivity over Cu/SiO2 catalyst for ethanol synthesis
from CO hydrogenation would provide a theoretical basis for
designing an efficient catalyst in terms of improving ethanol
selectivity and stability. In the present work, a simplified catalytic
system Cu/SiO2 catalyst prepared by the ammonia evaporation
method was studied for ethanol synthesis from CO
hydrogenation and the correlation of catalyst structure evolution
and ethanol selectivity change was investigated. The Cu/SiO2
catalyst exhibited a high initial ethanol selectivity of 40.0 wt%.
But it dramatically decreased to 9.6 wt% when the reaction time
was enhanced to 50 h. According to the characterization results
(at different stages), both copper valence and copper crystallite
size underwent changes during the CO hydrogenation process.
During the initial 38 h, the Cu+/(Cu++Cu0) value on the surface of
Cu/SiO2 catalyst rapidly dropped from 0.67 to 0.39, while the
copper crystallite size remained almost unchanged. During the
reaction period of 38-50 h, the Cu+/(Cu++Cu0) value did not
change significantly, but Cu+/(Cu++Cu0) value was so small that
could not hamper the Cu particles migration, thus resulting in an
inevitable occurrence of the Cu particles accumulation.
Combined with the experimental results, both the Cu+/(Cu++Cu0)
o
calcined under static air at 450 C for 5 h. The resulted Cu/SiO2
catalyst was tableted, crushed, sieved to 30-40 mesh, and
denoted as CS-calcined, where C and S indicated CuO, and
SiO2, respectively. The reduced catalyst was denoted as CS-
reduced and the used catalysts were simplified as CS-reaction-t
(t = 14.5, 28, 37 and 50 h), where “t” represented the reaction
time.
Catalyst characterization
The structure of CS-calcined, CS-reduced, CS-reaction-
14.5 h, CS-reaction-28 h, CS-reaction-38 h and CS-reaction-50
h catalysts was studied by XRD, XPS, TEM and N2O titration
techniques. Before characterization, in order to keep the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
2 × Y × Nav
X × MCu × 1.4 × 1019
1353 × Y (m2/g-Cu)
structure of CS-reduced, CS-reaction-14.5 h, CS-reaction-28 h,
CS-reaction-38 h and CS-reaction-50 h catalysts unchanged as
far as possible, after taken out from the fixed-bed reactor, those
catalysts were immediately filled with N2 and crushed in an inert
glove box.
SCu(0)
=
=
X
where Nav = Avogadro’s constant, MCu = relative atomic mass
(63.456 g/mol), 1.4 × 1019 came from that an equal abundance
of an average copper surface atom area of 0.0711 nm2 (1.4 ×
1019 copper atoms per square meter).
Here, SCu(0) represented the areas of surface Cu(0) at 300 oC
(both the reduction temperature and reaction temperature were
300 oC).
Powder X-ray diffraction patterns (XRD) were recorded over
o
the 2θ range from 5 to 90 using a Rigaku MiniFlex II X-ray
diffractometer, which was operated at 40 kV and 40 mA with Cu
Kα radiation (λ = 0.15418 nm).
In situ XRD patterns of the catalyst were recorded over the
Catalytic performance evaluation
o
2θ range from 30 to 80 equipped with an XRK900 instrument,
which was operated at 40 kV and 40 mA with Cu Kα radiation (λ
= 0.15418 nm). The CS-calcined catalyst was heated to 300 °C
at a rate of 10 °C/min under a flow of 10 % H2/N2 mixture gas,
kept at that temperature for 4 h. After reduction treatment,
CO/H2 (1/2.5) mixture gas was introduced into the flow system at
300 oC.
The catalyst evaluation experiments were carried out in a
stain-less fixed-bed reactor. The flow rate of feed gas was
controlled by a mass flow controller, and the exit gas was
measured by a wet flow meter. The reactions were conducted at
o
a temperature of 300 C and pressure of 3 MPa, using CO/H2 =
1/2.5 at a space velocity of 5000 h-1. Before reaction, the
catalyst was reduced according to the designed temperature
program, i.e. from room temperature to 300 oC, using a 10 %
H2/N2 mixture for 4 h at 300 oC. The products during the reaction
were analyzed using four GCs. The organic gas products,
consisting of hydrocarbons and methanol, were detected online
by flame ionization measurement using a GC4000A (GDX-403
column). The inorganic gas products were measured online by
thermal conductivity measurement using a GC4000A (carbon
molecular sieves column). The H2O and methanol products in
the liquid phase were detected by thermal conductivity
measurement using a GC4000A (GDX-401 column). The alcohol
products in the liquid phase were detected by flame ionization
measurement using a GC-7AG (Chromosorb 101).
X-ray photoelectron spectroscopy (XPS) and X-ray auger
electron spectroscopy (XAES) were used to analyze the change
in the surface composition using an AXIS ULTRA DLD
instrument equipped with Al Kα (hν = 1486.6 eV). Charging
effects were corrected using the C 1s peak due to adventitious
carbon with EB fixed at 284.5 eV.
The information about the catalyst morphology was
recorded using
a
field emission transmission electron
microscope (HRTEM, JEM-2100 F), which was operated at 200
kV.
Copper surface area of the catalyst in different stages
o
during reaction was determined by N2O titration at 50 C using
the procedure described by H. Xi et al. [32] and Z. Yuan et al.
[33]. In order to obtain the copper surface area (SCu(0)) under
reaction temperature (300 oC), the used catalyst was pretreated
o
o
Acknowledgements
by 5 % oxygen at 300 C for 30 min and cooled to 30 C, which
totally oxidized the metallic Cu. After that, the catalyst was
reduced by 10 % H2/Ar (15 mL/min) with the temperature
increasing from 30 to 300 oC and keeping for 1 h. The amount of
consumed H2 was measured and marked as X. And then, the
catalyst was cooled down to 50 °C under Ar. The surface copper
atoms were oxidized to Cu2O in 5 % N2O/Ar (15 mL/min) at 50
oC for 20 min. Finally, the catalyst was reduced by 10 % H2/Ar
(15 mL/min) with the temperature increasing to 300 oC at a rate
of 10 oC/min. The amount of consumed H2 in the second TPR
was denoted as Y. The exposed Cu surface area at 300 oC was
calculated by the following equation:
The authors are grateful for the financial support of National
Natural Science Foundation of China (21573269) and the Key
Research and Development Program of Shanxi Province.
Keywords: ethanol, CO hydrogenation, Cu-based catalyst, Cu+-
Cu0 species, copper crystallite size.
References
[1] L. X. Wang, L.Wang, J. Zhang, X. L. Liu, H. Wang, W. Zhang, Q.Yang, J. Y.
Ma, X. Dong, S. J. Yoo, J. G. Kim, X. J. Meng, F. S. Xiao, Angew. Chem. Int.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
Ed. 2018, 130, 6212-6216.
L. Gong, X. B. Ma, J. Catal. 2013, 297, 142-150.
[2] Y. Wang, Y. J. Zhao, J. Lv, X. B. Ma, ChemCatChem 2017, 9, 2085-2090.
[3] H. R. Yue, X. B. Ma, J. L. Gong, Acc. Chem. Res. 2014, 47, 1483-1492.
[4] H. T. Luk, C. Mondelli, D. C. Ferre, J. A. Stewart, J. Pe´rez-Ramı´rez,
Chem. Soc. Rev. 2017, 46, 1358-1426.
[31] J. W. Zheng, J. F. Zhou, H. Q. Lin, X. P. Duan, C. T. Williams, Y. Z. Yuan,
J. Phys. Chem. C. 2015, 119, 13758-13766.
[32] S. Soled, Science, 2015, 350(6265), 1171-1172.
[33] J. J. Huang, T. Ding, K. Ma, J. M. Cai, Z. R. Sun, Y. Tian, Z. Jiang, J.
Zhang, L. R. Zheng, X. G. Li, ChemCatChem, 2018, 10, 3862-3871.
[34] D. X. Hu, H. L. Hu, H. Zhou, G. Z. Li, C. L. Chen, J. Zhang, Y. Yang, Y. P.
Hu, Y. Y. Zhang, L. Wang, Catal. Sci. Technol. 2018, 8, 6091-6099.
[35] I. Obregón, I. Gandarias, A. Ocio, I. García-García, N. A. de Eulate, P. L.
Arias, Appl. Catal. B. 2017, 210, 328-341.
[5] M. Devarapalli, H. K. Atiyeh, Biofuel Res. J. 2015, 7, 268-280.
[6] K. Gꢀransson, U. Sꢀderlind, J. He, W. Zhang, Renewable Sustainable
Energy Rev. 2011, 15, 482-492.
[7] P. Wang, S. Y. Chen, Y. X. Bai, X. F. Gao, X. L. Li, K. Sun, H. J. Xie, G. H.
Yang, Y. Z. Han, Y. S. Tan, Fuel. 2017, 195, 69-81.
[8] L. Tan, G. H. Yang, Y. Yoneyama, Y. L. Kou, Y. S. Tan, T. Vitidsant, N.
Tsubaki, Appl. Catal. A. 2015, 505, 141-149.
[36] L. P. Han, L. Zhang, G. F. Zhao, Y. F. Chen, Q. F. Zhang, R. J. Chai, Y. Liu,
Y. Lu, ChemCatChem, 2016, 8, 1065-1073.
[9] Y. Z. Xiang, R. Barbosa, X. N. Li, N. Kruse, ACS Catal. 2015, 5, 2929-2934.
[10] S. F. Zaman, N. Pasupulety, A. A. Al-Zahrani, M. A. Daous, S. S. Al-
Shahrani, H. Driss, L .A. Petrov, Appl. Catal. A. 2017, 532, 133-145.
[11] M. R. Morrill, N. T. Thao, H. Shou, R. J. Davis, D. G. Barton, D. Ferrari, P.
K. Agrawal, C. W. Jones, ACS Catal. 2013, 3, 1665-1675.
[12] M. Konarova, F. Q. Tang, J. L. Chen, G. Wang, V. Rudolph, J. Beltramini,
ChemCatChem 2014, 6, 2394-2402.
[37] X. L. Li, Q. D. Zhang, H. J. Xie, X. F. Gao, Y. Q. Wu, G. H. Yang, P. Wang,
S. P. Tian, Y. S. Tan, ChemistrySelect, 2017, 2, 10365-10370.
[38] P. Wang, S. Y. Chen, Y. X. Bai, X. F. Gao, X. L. Li, K. Sun, H. J. Xie, G. H.
Yang, Y. Z. Han, Y. S. Tan, Fuel, 2017, 195, 69-81.
[39] H. R. Yue, X. B. Ma, J. L. Gong, Acc. Chem. Res. 2014, 47, 1483-1492.
[40] H. J. Xi, X. N. Hou, Y. J. Liu, S. J. Qing, Z. X. Gao, Angew Chem. Int. Ed.
2014, 53, 11886-11889.
[13] Y. Choi, P. Liu, J. Am. Chem. Soc. 2009, 131, 13054-13061.
[14] M. Gupta, M. L. Smith, J. J. Spivey, ACS Catal. 2011, 1, 641-656.
[15] G. L. Liu, T. Niu, A. Cao, Y. X. Geng, Y. Zhang, Y. Liu, Fuel. 2016, 176, 1-
10.
[41] Z. L. Yuan, L. Wang, J. H. Wang, S. X. Xia, P. Chen, Z. Y. Hou, X. M.
Zheng, Appl. Catal. B. 2011, 101, 431-440.
[42] Z. Q. Wang, Z. N. Xu, S. Y. Peng, M. J. Zhang, G. Lu, Q. S. Chen, Y.
Chen, G. C. Guo, ACS Catal. 2015, 5, 4255-4259.
[16] J. J. Su, Z. P. Zhang, D. L. Fu, D. Liu, X. C. Xu, B. F. Shi, X. Wang, R. Si,
Z. Jiang, J. Xu, Y. F. Han, J. Catal. 2016, 336, 94-106.
[17] G. Prieto, S. Beijer, M. L. Smith, M. He, Y. Au, Z. Wang, D. A. Bruce, K. P.
de Jong, J. J. Spivey, P. E. de Jongh, Angew. Chem. Int. Ed. 2014, 53, 6397 –
6401.
[43] H. R. Yue, Y. J. Zhao, L. Zhao, J. Lv, S. P. Wang, J. L. Gong, X. B. Ma,
AIChE J. 2012, 58, 2798-2809.
[44] J. L. Gong, H. R. Yue, Y. J. Zhao, S. Zhao, L. Zhao, J. Lv, S. P. Wang, X.
B. Ma, J. Am. Chem. Soc. 2012, 134, 13922-13925.
[45] Q. Wang, L. Qiu, D. Ding, Y. Z. Chen, C. W. Shi, P. Cui, Y. Wang, Q. H.
Zhang, R. Liu, H. Shen, Catal. Commun. 2018, 108, 68-72.
[46] H. Oguchi, H. Kanai, K. Utani, Y. Matsumura, S. Imamura, Appl. Catal. A.
2005, 293, 64-70.
[18] Z. He, H. Q. Lin, P. He, Y. Z. Yuan, J. Catal. 2011, 227, 54-63.
[19] X. L. Zheng, H. Q. Lin, J. W. Zheng, X. P. Duan, Y. Z. Yuan, ACS Catal.
2013, 3, 2738-2749.
[20] Y. Wang, J. Y. Liao, J. Zhang, S. P. Wang, Y. J. Zhao, X. B. Ma, AIChE J.
2017, 63 (7), 2839-2849.
[47] H. Oguchi, T. Nishiguchi, T. Matsumoto, H. Kanai, K. Utani, Y. Matsumura,
S. Imamura, Appl. Catal. A. 2005, 281, 69-73.
[21] P. P. Ai, M. H. Tan, N. Yamane, G. G.Liu, R. G. Fan, G. H. Yang, Y.
Yoneyama, R. Q. Yang, N. Tsubaki, Chem. Eur. J. 2017, 23, 8252-8261.
[22] L. P. Han, L. Zhang, G. F. Zhao, Y. F. Chen, Q. F. Zhang, R. J. Chai, Y. Liu,
Y. Lu, ChemCatChem 2016, 8, 1065-1073.
[48] X. F. Zhai, J. Shamoto, H. J. Xie, Y. S. Tan, Y. Z. Han, N. Tsubaki, Fuel.
2018, 87, 430-434.
[49] Y. H. Zhao, K. J. Sun, X. F. Ma, J. X. Liu, D. P. Sun, H. Y. Su, W. X. Li,
Angew Chem. Int. Ed. 2011, 50, 5335-5338.
[23] Y. Wang, Y. Shen, Y. Zhao, J. Lv, S. Wang, X. Ma, ACS Catal. 2015, 5,
6200-6208.
[50] F. Arena, K. Barbera, G. Italiano, G. Bonura, L. Spadaro, F. Frusteri, J.
Catal. 2007, 249, 185-194.
[24] B. Wang, Y. Y. Cui, C. Wen, X. Chen, Y. Dong, W. L. Dai, Appl. Catal. A.
2016, 509, 66-74.
[25] T. Witoon, T. Numpilai, T. Phongamwong, W. Donphai, C. Boonyuen, C.
Warakulwit, M. Chareonpanich, J. Limtrakul, Chem. Eng. J. 2018, 334, 1781-
1791.
[26] R. P. Ye, L. Lin, J. X. Yang, M. L. Sun, F. Li, B. Li, Y. G. Yao, J. Catal.
2017, 350, 122-132.
[27] Y. Z. Yang, X. Z. Qi, X. X. Wang, D. Lv, F. Yu, L. S. Zhong, H. Wang, Y. H.
Sun, Catal. Today. 2016, 270, 101-107.
[28] J. T. Sun, I. S. Metcalfe, M. Sahibzada, Ind. Eng. Chem .Res. 1999, 38,
3868-3872.
[29] X. M. Huang, M. Ma, S. Miao, Y. P. Zheng, M. S. Chen, W. J. Shen, Appl.
Catal. A. 2017, 531, 79-88.
[30] S. Zhao, H. R. Yue, Y. J. Zhao, B. Wang, Y. C. Geng, J. Lv, S. P. Wang, J.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201801888
ChemCatChem
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Both the Cu+/(Cu++Cu0) values and
copper crystallite sizes had great
effects on the ethanol selectivity of
Cu/SiO2 catalyst
Xiao-Li Li, Guo-Hui Yang, Meng Zhang,
Xiao-Feng Gao, Hong-Juan Xie,
Yun-Xing Bai, Ying-Quan Wu, Jun-Xuan
Pan, and Yi-Sheng Tan*
Page No. – Page No.
.
Insight into the correlation between
Cu species evolution and ethanol
selectivity in the direct ethanol
synthesis from CO hydrogenation
This article is protected by copyright. All rights reserved.