Effect of surface hydroxyl group of ultra-small silica on the chemical states of copper catalyst for dimethyl oxalate hydrogenation

Article abstract of DOI:10.1016/j.cattod.2019.06.031

Cu/SiO₂ catalyst prepared by ammonia evaporation method was reported to perform a great catalytic activity and selectivity in dimethyl oxalate (DMO) hydrogenation, which is one of the key steps in ethylene glycol (EG) synthesis from syngas. In recent years, significant advances have achieved on the nature of active sites and catalytic mechanism. However, the fabrication of Cu/SiO₂ catalyst with controllable copper species remains challenging. Here, we reported a facile and effective approach to manipulate the surface hydroxyls of silica, which was revealed to be important factor for the formation of active species. An excellent linear correlation between surface hydroxyl groups of silica and the amount of Cu⁺ species was demonstrated, indicating that the formation of copper phyllosilicate can be kinetically favoured with increasing surface hydroxyls during preparation. Furthermore, as the copper phyllosilicate formation is enhanced, the specific surface area is significantly increased and the polymerization of copper hydroxide may be slow down, resulting in a highly improved dispersion of metallic copper as well. The enlarged surface areas of Cu⁰ and Cu⁺ species greatly enhanced the catalytic performance of Cu/SiO₂ in DMO hydrogenation to EG. These understandings on the relationship between surface hydroxyl groups and chemical states of copper catalyst may lead to new possibilities in rational design of catalysts.

Full text of DOI:10.1016/j.cattod.2019.06.031

Accepted Manuscript
Title: Effect of surface hydroxyl group of ultra-small silica on
the chemical states of copper catalyst for dimethyl oxalate
hydrogenation
Authors: Yue Wang, Wenlong Yang, Dawei Yao, Shengping
Wang, Yan Xu, Yujun Zhao, Xinbin Ma
PII:
S0920-5861(19)30312-8
DOI:
Reference:
https://doi.org/10.1016/j.cattod.2019.06.031
CATTOD 12278
To appear in:
Catalysis Today
Received date:
Revised date:
Accepted date:
1 February 2019
31 May 2019
10 June 2019
Please cite this article as: Wang Y, Yang W, Yao D, Wang S, Xu Y, Zhao Y, Ma
X, Effect of surface hydroxyl group of ultra-small silica on the chemical states
of copper catalyst for dimethyl oxalate hydrogenation, Catalysis Today (2019),
https://doi.org/10.1016/j.cattod.2019.06.031
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Effect of surface hydroxyl group of ultra-small silica on the chemical
states of copper catalyst for dimethyl oxalate hydrogenation
Yue Wang, Wenlong Yang, Dawei Yao, Shengping Wang, Yan Xu, Yujun Zhao* and Xinbin Ma
Key Laboratory for Green Chemical Technology of Ministry of Education, Collaborative
Innovation Center of Chemical Science and Engineering, School of Chemical Engineering and
Technology, Tianjin University, Tianjin 300072, China
*Yujun Zhao: the corresponding author, to whom correspondence and proofs should be sent.
Email: yujunzhao@tju.edu.cn
Graphical abstract

Highlights



A highly efficient Cu/SiO2 catalyst for DMO hydrogenation was prepared.
Surface hydroxyls of silica sol favor the formation of copper phyllosilicate.
Ultra-small silica greatly improved the surface areas of both Cu0 and Cu+
species.
ABSTRACT: Cu/SiO2 catalyst prepared by ammonia evaporation method was
reported to perform a great catalytic activity and selectivity in dimethyl oxalate
(DMO) hydrogenation, which is one of the key steps in ethylene glycol (EG)
synthesis from syngas. In recent years, significant advances have achieved on the
nature of active sites and catalytic mechanism. However, the fabrication of Cu/SiO2
catalyst with controllable copper species remains challenging. Here, we reported a
facile and effective approach to manipulate the surface hydroxyls of silica, which was
revealed to be important factor for the formation of active species. An excellent linear
correlation between surface hydroxyl groups of silica and the amount of Cu+ species
was demonstrated, indicating that the formation of copper phyllosilicate can be
kinetically favored with increasing surface hydroxyls during preparation.
Furthermore, as the copper phyllosilicate formation is enhanced, the specific surface
area is significantly increased and the polymerization of copper hydroxide may be
slow down, resulting in a highly improved dispersion of metallic copper as well. The
enlarged surface areas of Cu0 and Cu+ species greatly enhanced the catalytic

performance of Cu/SiO2 in DMO hydrogenation to EG. These understandings on the
relationship between surface hydroxyl groups and chemical states of copper catalyst
may lead to new possibilities in rational design of catalysts.
KEY WORD: Cu/SiO2, surface hydroxyls, hydrogenation, ethylene glycol, syngas

1. Introduction
Ethylene glycol (EG), an important chemical material, is widely used in many
fields such as polyester manufacture and antifreeze production.[1] Traditionally, EG
is synthesized via petroleum-based ethylene oxide hydrolysis. With the decrease of
petroleum reserves and increasing demand of EG, alternative routes via non-
petroleum-based resources have drawn more and more attention. Synthesis of EG
from syngas, which can be derived from coal, biomass or natural gas, is one of the
most promising routes due to its good economy and environment friendly features.
This synthesis route includes two steps: coupling of CO with methyl nitrite ester to
produce dimethyl oxalate (DMO) and hydrogenation of DMO to EG.[2] Considerable
works have been devoted to the hydrogenation of DMO, but fabrication of an efficient
catalyst with both high activity and excellent stability is still one of the key problems
in promoting the industrialization of this process. Copper-based catalysts have been
reported with good hydrogenation activity of C=O/C-O bonds and low tendency for
C-C bond cleavage, which is extensively used in DMO hydrogenation.[3]
Numerous efforts have been exerted to understand the nature of active sites and
their catalytic behaviour of the copper-based catalysts in DMO hydrogenation.
Previous works suggested that the coexistence of Cu0 and Cu+ species is the crux for
displaying a remarkable catalytic performance.[4] A dual-site synergetic catalysis
mechanism was accordingly proposed, in which the Cu0 species activated H2 and the
Cu+ species adsorbed the carbonyl group.[5, 6] To obtained stable coexisted Cu0 and
Cu+ species in hydrogenation reaction (high temperature and abundant hydrogen

atmosphere), different supports and promoters were investigated.[7-11] Among them,
silica supported copper catalyst exhibited favourable selectivity of EG due to its weak
acidic and basic property, which are suggested as the active sites for the synthesis of
side products such as 1,2-propanediol and 1,2-butanediol.[12, 13]
Ammonia evaporation (AE) method is proved as an effectively way to prepare
Cu/SiO2 catalysts with both Cu0 and Cu+ sites after reduction. Chen and co-workers
systematically studied the effect of AE temperature on the chemical states of copper
species and proposed that the Cu0 and Cu+ are primarily originated from well
dispersed CuO nanoparticles and copper phyllosilicate.[14] HMS,[15-18] MCM-
41[19] and ordered hierarchical porous silica [20, 21] were applied as support
precursors to fabricate catalysts with high dispersion of copper species and a proper
ratio of Cu0/Cu+. Based on the understanding of active sites in the hydrogenation of
DMO to EG over silica supported copper catalyst, the increase of surface areas and
appropriate ratio are believed to be effective ways to improve catalytic activity. Our
previous work on studying the structure evolution of the catalyst during the
preparation process and the sensitivity of the preparation parameters have given some
meaningful experiential information on tuning the copper species distribution.[6, 21]
However, the fabrication of Cu/SiO2 catalyst with controllable copper species remains
challenging and deserves an intensive study to understand the formation process of
active sites.
As far as we know, silica sol is generally used as the silica resource for the
preparation of the Cu/SiO2 catalyst, nevertheless few attentions have been paid on the

role of silica sol on the formation of copper species. Zhang and co-workers reported
that different nano-sized silica sols would have a great impact on the structure and
properties of catalysts.[22] However, the origin of this effect, as well as the formation
process of active sites, is failed to be illuminated.
Here, a highly efficient Cu/SiO2 catalyst for DMO hydrogenation was prepared
by using ultra-small size silica sol. And, the influence of surface hydroxyl groups of
silica sol on the formation of copper phyllosilicate, as well as the surface copper
species distribution, was well demonstrated. The Cu/SiO2 catalyst prepared with ultra-
small size silica sol exhibited excellent activity and stability in DMO hydrogenation.
These insights on the formation of copper species would give a significant inspiration
on the design and fabrication of Cu/SiO2 catalysts.
2. Experimental
2.1. Catalyst preparation
Firstly, we papered a silica sol with ultra-small particles by injecting 0.125 g/mL
Na2SiO3·9H2O at a speed of 10 mL/min to a 0.9 mg/mL NaOH solution at 363 K and
then stirring for 4 h. In order to obtain silica sol with different particle sizes, the sol
was then transferred into autoclaves and kept at 473 K for 2 h, 4 h, and 12 h
respectively for their hydrothermal treatment. In this way, four silica sol samples with
different particle sizes were synthesized and further introduced into the following
catalysts preparation.
The Cu/SiO2 catalysts were prepared by the AE method described as below. 8.0 g
of Cu(NO3)2·3H2O (Tianjin Guangfu Technology Development Co. Ltd) was

dissolved in 50 mL of deionized water followed by the addition of 20 mL of 28%
ammonia aqueous solution (Tianjin Kermel Chemical Reagent Co. Ltd) and 20 mL of
deionized water and stirred for 30 min. Then the required amount of silica sol
(Qingdao EGF Silicon Products Co. Ltd) was added to the copper ammonia complex
solution and stirred for another 4 h with the initial pH of the suspension was 11-12.
All above operations were performed at room temperature. The suspension was
preheated at 353 K in a water bath to allow for the evaporation of ammonia and the
deposition of copper species on silica. The evaporation process was terminated when
the pH of the suspension decreased to 6-7. The mixture was filtered and washed with
deionized water followed by drying at 373 K for 4 h. The catalyst precursors were
calcined at 673 K for 4 h and the Cu/SiO2 catalyst was obtained. The samples are
labeled as Cu/SiO2-N, where N stands for the particle size of silica sol, which is
determined by the TEM images (Fig. 1).
2.2. Catalyst characterization
The copper content of the samples was determined by inductivity coupled
plasma optical emission spectrometry (ICP-OES) (Varian, Vista-MPX) operated at
high frequency emission power of 1.5 kW and plasma airflow of 15 L/min. Samples
were dissolved in the mixture of HF and HBO3 and then diluted with deionized water.
Laser Particle Size Analyzer (Malvern, Zetasizer Nano) was used to measure the
particle size distribution of silica sol. The silica sol sample was first diluted to 1 g/L
and filtered before injected to the sample cell. The scan speed was 1000 times per
second and the result was an average one of three times of measurements.

Texture properties of the samples were determined by a nitrogen adsorption
method using a Micromeritics Tristar II 3000 Analyzer at 77 K. Pore size distribution
was calculated by the Brunauer-Emmet-Teller (BET) method from the desorption
isotherms and the pore size distribution was determined by Barrett-Joyner-Halenda
(BJH) method from the adsorption branches.
X-ray powder diffraction (XRD) was performed on a Rigaku C/max-2500
diffractometer using Cu Kα radiation (λ=1.5406 Å) at room temperature. The
scanning speed was 12°/min from 2θ=10° to 2θ=90°. To avoid the phase
transformation from surface oxidation, the reduced catalysts were collected under H2
at room temperature and sealed in glass bottles before XRD analysis. The copper
crystallite size was calculated using the Scherrer equation.
Temperature programmed reduction (TPR) was performed on Micromeritics
Autochem II 2920. 50 mg of the samples were loaded into quartz tubes and dried in
Ar at 393 K for 1 h before reduction. Then the samples were heated to 1073 K in the
flow of 5%H2-Ar with a heating rate of 10 K/min. The amount of hydrogen
consumption was determined by a thermal conductivity detector (TCD).
The special surface areas of metallic copper were measured by N2O titration
carried out on Micromeritics Autochem II 2920. The samples were reduced under
10% H2-Ar flow at 623 K for 2 h and then cooled to 363K. The pure N2O was
induced for 1 h, to ensure surface Cu atoms completely oxidized to Cu2O. Then 10%
H2-Ar pulse reduction was conducted at 623 K and repeated until the pulse area no
longer changed, for reducing all surface Cu2O back to metallic Cu. The consumed

amount of H2 obtained from the TCD signal was used to calculate the surface area of
metallic copper.
Fourier transform infrared (FTIR) spectra of samples were recorded on a Nicolet
6700 spectrometry (Thermo Fisher Scientific) using the standard KBr disk method.
The effective region was from 4000 to 400 cm-1.
X-ray photoelectron spectra (XPS) and Auger electron spectroscopy (AES) were
conducted by using PHI 1600 ESCA (PE Company) equipped with an Al Kα X-ray
radiation source (hν = 1486.6 eV). The catalyst was pressed to thin disks and reduced
in a flow of H2 at 623 K for 4 h. After cooled to room temperature, the samples were
sealed with Ar, and then transferred to the holder in less than 5 minutes. The binding
energy was calibrated using the Si 2p peak at 103.2 eV as reference. The experimental
errors were within ±0.2 eV.
Transmission electron microscopy (TEM) images were obtained by a Philips
TECNAL G2 F20 system electron microscope at 100 kV. The sample powder was
dispersed in ethanol and applied onto a copper grid-supported transparent carbon foil
and dried in air.
2.3. Catalyst test
The catalytic test was carried out in a continuous-flow stainless steel fixed-bed
tubular reactor. Firstly, 0.4 g of catalyst (40-60 meshes) was packed in the tube. Then
the catalyst was reduced under H2 at 623 K for 4 h. After cooling to reaction
temperature of 473 K, the reactant (20 wt.% DMO in methanol solution) was injected
from the top of the reactor through a high-pressure pump (Elite P230 HPLC Pump)

with a system pressure of 2.5 MPa and a H2/DMO molar ratio of 80. The reaction
products were condensed and analyzed by Agilent GC 6820 with an HP-innowax
capillary column (Hewlett-Packard Company, 30 m×0.32 mm×0.50 μm) and a flame
ionization detector.
3. Results
3.1. Structural properties of silica sol
In order to discuss how the silica sol affects the active species formation process,
we prepared four silica sol samples with different particle size by hydrothermal
treatment and introduced them into the preparation of the Cu/SiO2 catalysts. First, the
structural properties of silica sol were characterized. The TEM images of silica sol
samples and their particle sizes were shown in Fig. 1. It is demonstrated with no
hydrothermal treatment, a silica sol with ultra-small particle size (about 5 nm) was
successfully obtained. As the hydrothermal treatment time increased to 2, 4, and 12 h,
the silica particles grew to 10, 17, and 25 nm respectively.
To further verify the particle size of silica sol, we measured the particle size
distribution by using a Malvern Laser Particle Size Analyzer (Fig. S1). It shows that
the particle sizes of all samples distribute in small ranges, and the average particle
size of silica sol was increased from 1.3 nm to 21 nm by a 12-hours hydrothermal
treatment. Notably, although there is a little difference between the results of particle
sizes obtained by LPSA and TEM images, their variation trends are similar. For the
convenience of the following discussion, the silica sol samples were labeled as SiO2-
N, where N is their average particle sizes obtained from TEM images.

Moreover, FTIR of these four samples were conducted to determine the surface
chemical structure of silica particles, and shown in Fig. 2A. The bands at 1110 and
800 cm-1 are respectively assigned to asymmetric and symmetric stretching of the Si-
O bonds in silica and the shoulder peak at 960 cm-1 is ascribed to surface OH bonds
vibration. In this case the relative amount of surface hydroxyl groups on silica
particles can be calculated by normalizing the integrated intensity of the band at 960
cm-1 to that of the band at 800 cm-1, which is regarded as I970/I800. Fig. 2B clearly
displays that the relative amount of surface hydroxyl groups nearly linearly dropped
with the increase of silica particle size, which might have a significant impact on the
following process of Cu/SiO2 preparation.
3.2. Textural properties of calcined Cu/SiO2 catalysts
The textural properties of calcined samples were shown in Table. 1. The copper
contents in all the catalysts were about 18 wt.%, which allows us to exclude the
influence of copper loading in following discussion. It was obvious that the particle
size of silica sol significantly influenced the specific surface areas and pore volumes
of the Cu/SiO2 catalysts. When the particle size of silica sol increases from 5 to 25
nm, the BET surface areas and pore volume decreases from 522 to 124 m2/g and 2.24
to 0.41 cm3/g, respectively. The particle size of silica sol slightly influenced the
average pore diameter as well. As shown in Fig. S2, both pores at 3.3 and 5.4 nm
demonstrate a decreasing trend with the increasing particle size of silica sol.
The compounds of phyllosilicate species can be determined by IR technique. [23,
24] Fig. 3A shows the FTIR spectra of calcined Cu/SiO2 samples in the frequency

range of 400-2000 cm-1. The bands at 1110, 800 and 470 cm-1 are assigned to different
vibration modes of the Si-O bonds in the amorphous silica. The appearance of the
characteristic δOH band at 670 cm-1 and the shoulder peak at 1040 cm-1 suggested the
existence of copper phyllosilicate species. [14, 23] As the particle size of silica sol
increased, the adsorption band at 670 cm-1 decreased and almost vanished on
Cu/SiO2-25. The relative amount of copper phyllosilicate in these samples could be
evaluated by considering the integrated intensity of the band at 670 cm-1 normalized
to the integrated intensity of the Si-O symmetric stretching band of SiO2 at 800 cm-1,
which was termed as I670/I800.[23] Although the extinction coefficients of
corresponding IR bands are not known, this term could give a qualitative estimation
of the ratio of copper phyllosilicate to catalyst. As shown in Fig. 3B, the relative
amount of copper phyllosilicate decreased with the increasing particle size of silica
sol, which displays as the same variation trend as surface hydroxyl groups of silica
sol.
To further understand the structure of Cu/SiO2-N catalyst, TEM images were
taken and shown in Fig. 4. It could be easily observed that a large amount of copper
phyllosilicate was formed on the sample of Cu/SiO2-5 (Fig. 4A), which is identified
by its characteristic lamellar structure. Moreover, the amount of lamellar structure
decreased on Cu/SiO2-17 (Fig. 4B) and almost vanished when the particle size of
silica sol increased to 25 nm (Fig. 4C), which is agreed with the results from FTIR
spectra. It is notable that the spherical structure of silica particles was destroyed to
some extent. After introduced into the catalyst preparation, the silica particles were

found to totally disappear for SiO2-5. The above result indicates a dissolution of silica
in the base solution during the catalyst preparation procedure by the AE method. But
for the sample of Cu/SiO2-17 and Cu/SiO2-25, the shape of silica particles was
reserved in the catalyst due to the weak dissolution ability of silica with large particle
sizes.
3.3. Reduction behavior and chemical states of copper species of reduced
catalysts
Fig. 5A showed the XRD patterns of calcined samples. The strong and broad
peak at ca. 22° is attributed to amorphous silica. Moreover, the characteristic peaks of
CuO at 2θ of 35.4° and 38.6°is assigned to CuO crystallite (JCPDS05-0661). Their
absence in Cu/SiO2-5 and Cu/SiO2-10 samples indicates the copper species was
highly dispersed. Further increasing the particle size of silica sol induces a noticeable
growth of CuO crystallite.
When the catalysts were reduced, the characteristic peaks of CuO vanished in the
XRD patterns (Fig. 5B). It was well known that the reduction of copper phyllosilicate
ceased at Cu+ under present reduction condition (623 K) due to the strong interaction
between copper ions and the support while well dispersed CuO and bulk CuO were
reduced to metallic copper.[4, 15, 17] For the catalysts of Cu/SiO2-5 and Cu/SiO2-10,
only a small and broad peak at 2θ of 36.4°could be found, which was assigned to
Cu2O species (JCPDS05-0667), while the metallic copper crystallite was too small to
be detected. For the catalysts prepared by silica sol with relatively large particle sizes,
an obvious diffraction peak at 43.3° along with two weak ones at 51.1° and

74.1°emerged, which were identified as metallic copper species (JCPDS04-0836).
The reducibility of the Cu/SiO2-N was characterized by TPR, shown as Fig. 6. It
was reported that the reduction of well dispersed CuO species on ion-exchanged
Cu/SiO2 samples was 523 K, which is lower than that of bulk CuO. [25] For the
copper phyllosilicate, the reduction temperature occurred at 523 K which overlapped
with the reduction of well dispersed CuO to Cu. Further reduction of Cu+ to Cu
required a temperature higher than 873 K.[25, 26] Van der Grift et al. identified only
one reduction peak at 510 K for the catalyst prepared by deposition-precipitation
method, suggesting the identical reduction temperature for copper phyllosilicate to
Cu+ and for the well dispersed CuO to Cu. [27]
For the sample of Cu/SiO2-10, only one reduction peak centered at 468 K could
be observed. This peak was attributed to the collective contribution of the reduction of
the well dispersed CuO to metallic Cu and the copper phyllosilicate to Cu+. Besides
the main reduction peak, there was a shoulder peak at ca. 486 K occurred in the
pattern of Cu/SiO2-17 and became stronger on Cu/SiO2-25. This shoulder peak could
be ascribed to the reduction of large CuO particles to metallic Cu. It was worthwhile
to note that the Cu/SiO2-5 only displayed one reduction peak at high temperature.
According to the XRD result, the diffraction peak of metallic copper could not be
detected. So, the single reduction peak occurred at high temperature suggested the
presence of strong copper-support interaction on Cu/SiO2-5.
To further determine the chemical states in the reduced Cu/SiO2 catalysts, N2O
titration, XPS and AES spectra were conducted to qualitatively and quantitatively

measure the amounts of surface copper sites. The specific surface area of Cu0 was
0
characterized by N2O titration and listed in Table. 2. It can be seen that SCu decreased
from 38.3 to 9.7 m2/g as the particle size of silica sol increased from 5 to 25 nm. In Cu
2p XPS spectra (Fig. 7), there is no satellite peak at binding energy of 942-944 eV,
indicating no Cu2+ species remaining after the reduction. The two peaks centered at
binding energies of 932.0 and 951.8 eV could be assigned to Cu+ or Cu0 species. In
Cu LMM Auger electron spectra, the appearance of two obvious overlapping peaks at
569.9 eV and 573.0 eV have been applied to distinguish between Cu0 and Cu+ species.
+
0
+
+
The molar ratios of the surface Cu /(Cu + Cu ), which is denoted as XCu ,[6] are
apparently affected by the particle sizes of silica sol. In this case, the Cu+ surface area
0
+
+
can be estimated on the basis of Cu surface area and XCu , which were listed as SCu in
Table 2 as well. It is worthy noticing that both surface areas of Cu0 and Cu+ decreased
dramatically with the increasing particle size of silica sol.
3.4. Catalytic activity
Catalytic performance of the reduced Cu/SiO2 catalysts were examined in the
hydrogenation of DMO to EG. The hydrogenation of DMO proceeded via methyl
glycolate (MG) to EG (eq. 1-2). Other byproducts included ethanol (eq. 3), and little
1,2-propanediol, 1,2-butanediol and ether. The conversion of DMO, selectivity of EG
and MG as functions of liquid hourly space velocity (LHSV) were shown in Fig. 8.
The conversion of DMO was 100% with the LHSV lower than 1.5 h-1 and further
increasing LHSV induced a slight decrease in DMO conversion. The selectivity of EG
showed a volcano-type catalytic behavior in all samples while the selectivity of MG

showed an increasing trend with the LHSV increased from 0.5 to 5.5 h-1. The results
were in accordance with the character of these consecutive reactions, in which MG
was an intermediate product. When the LHSV increased, the residence time of DMO
and MG decreased, leading to a higher selectivity of MG. The catalytic activity of
Cu/SiO2-N increased steadily with the particle size of silica sol decreases and reached
the highest level on the Cu/SiO2-5. Moreover, a strong correlation between the
amount of Cu+ and space time yield of EG (STYEG) was obtained (Fig. 9B) while no
clear correlation between the amount of Cu0 and STYEG was observed (Fig. 9A).
Further discussion will be given in section 4.1.
(COOCH3)2+2H2→CH2OHCOOCH3+CH3OH
CH2OHCOOCH3+2H2→(CH2OH)2+CH3OH
(CH2OH)2+H2→CH3CH2OH+H2O
(1)
(2)
(3)
Stability was another crucial factor to evaluate the catalytic performance of
Cu/SiO2 catalyst. For the copper-based catalysts, agglomeration of metallic copper
strongly inhibited their long-time utility due to the low Hütting temperature of copper.
It was reported that the boron dopant could significantly improve the long-time run
stability of Cu/SiO2.[8, 28, 29] The authors proposed that the boron gathered on the
copper cluster and inhibited the agglomeration of metallic copper. Additionally,
MCM-41[19] and ordered mesoporous silica[20, 21] with unique pore structures were
proved as effective supports for finely dispersing the copper species and improving
stability. In this work, as was shown in Fig. 8D, Cu/SiO2-5 catalyst performed an
excellent long-time stability for hydrogenation of DMO at 473 K. It was known that

the most abundant copper phyllosilicate formed on Cu/SiO2-5 catalyst and the special
lamellar structure of copper phyllosilicate could strongly enlarge the surface area and
divide copper crystal into very small particles, which in terms dispersed the copper
species effectively. The agglomeration of copper species was inhibited due to the
abundant copper phyllosilicate and therefore excellent long-time stability was
obtained.
4. Discussion
4.1. Dominating active site of Cu/SiO2-N in DMO hydrogenation
Previous study has proved that the co-existence of Cu0 and Cu+ in the Cu/SiO2
catalyst was responsible for high efficiency in catalytic hydrogenation of DMO.[4]
The Cu0 species facilitates the H2 decomposition while Cu+ species can stabilize the
methoxy and acyl group, which are intermediates in DMO hydrogenation.[4, 5] Cu+
could also act as electrophilic or Lewis acidic sites to polarize the C=O bond via
electron lone pair in oxygen, thus activates the ester group in DMO.[15] It is believed
that a proper distribution of Cu0 and Cu+ species is the key in catalyst synthesis. And
in our previous work, there is a balancing effect between the Cu0 and Cu+ species on
their synergetic catalysis process.[6] When the amount of Cu0 species is insufficient,
the catalytic activity of hydrogenation is determined by the amount of Cu0 species,
whereas it is most affected by the Cu+ surface area.
As shown above, the space time yield of EG as functions of Cu+ and Cu0 specific
surface areas are examined and shown in Fig. 9. The STYEG showed a linearly
increase with raising amount of Cu0 surface area at the beginning as well. With

further increase of Cu0 species, the enhancement effect was slow down, which maybe
result from the sufficient of ability of H2 decomposition. Meanwhile, as to the
formation behaviour of copper species analysis above, the Cu0 surface area was
greatly enhanced by increasing copper phyllosilicate. There is a strong correlation
between the surface area of Cu+ and the STYEG, suggesting the insufficient ability of
the catalyst in adsorbing carbonyl groups. Therefore, the adsorption of carbonyl has
become a control step in the hydrogenation of DMO to EG. Thus, the insights for
enhancing surface area of Cu+ during catalyst formation is the key in the catalyst
design of DMO hydrogenation, which is exactly discussed in this work as follows.
4.2. Surface hydroxyl groups effect of silica sol on copper species of Cu/SiO2-N
Several studies and our previous work have indicated that the formation of
phyllosilicate involved the dissolution of silica.[23, 30-32] Iler et al. proposed that the
dissolution of silica in base solution underwent a mechanism showed in Fig. S3.[33]
The dissolution rate of silica was strongly influenced by the amounts of surface
hydroxyl groups and the interface area between silica and solution. Therefore, more
silica could be dissolved when the silica sol with smaller particle size was applied in
the catalyst preparation. During the ammonia evaporation (AE) procedure, the
departure of ammonia pushed the balance moving rightward (eq. 4) and the
Cu(OH)2(H2O)4 gradually formed. The Cu(OH)2(H2O)4 was further involved in two
competing reactions: formation of copper phyllosilicate and polymerization of copper
hydroxide. With the addition of silica sol with smaller particle size, the formation of
copper phyllosilicate was more kinetically favoured due to the strong solubility of

silica in the solution. Therefore, more copper phyllosilicate should be formed when
silica sol with small particle size was used during the preparation. Whereas
polymerization between Cu(OH)2(H2O)4 to copper hydroxide was favoured when the
particle size of silica sol was larger, since the dissolution rate of silica is too slow to
satisfy the requirement by the formation of copper phyllosilicate.
2+
+
Cu(NH3)4 +6H2O→Cu(OH)2(H2O)4+2NH4 +2NH3
(4)
It is reported the copper phyllosilicate unique structure of Cu-O-Si bonds is the
main contribution for remaining such a lot of surface Cu+ species after reduction.[4,
14] As shown in Fig. 10, the correlation establishes an excellent linear fit between
surface hydroxyl groups of silica sol and the amount of Cu+ surface area
correspondingly synthesized. The Cu+ surface area showed a linearly increase with
raising amount of copper phyllosilicate as well.
Furthermore, the increasing amount of copper hydroxide and the declined
surface area of catalysts, which resulted from the decreasing amount of copper
phyllosilicate were responsible for the increasing size of metallic copper particles.[27]
Therefore, when the particle size of silica was increased and surface hydroxyl groups
was decreased, the enlarged Cu particles induced a dramatically decline in Cu0
surface area as well as the decreasing amount of copper phyllosilicate, resulting in the
decrease in surface area of Cu+. These results strongly supported the above analysis
on the behaviour of surface hydroxyl groups of silica sol during the formation of the
copper-based catalysts, which further demonstrates the effect of surface hydroxyl
groups of silica sol on enhancing the formation of well dispersed Cu+ species and Cu0

species.
Beside the particle size of silica sol resources, the surface hydroxyls could be
adjusted by the preparation process and parameters as well, which probably is the key
origin of their influence on the amount of copper phyllosilicates and distribution of
copper species based on the above analysis. Chen and co-workers[14] reported high
ammonia-evaporation temperature was favorable to increasing the contents of copper
phyllosilicate, which maybe resulted from the easier generation of surface hydroxyls
on silica. Yin et al. [34] reported the similar results for the initial precipitation
temperature. In our previous works, we also found the aging process and pretreatment
of supports would have significant influences on these properties of Cu/SiO2
catalysts.[6, 21, 35] For example, the formation of copper phyllosilicate was much
easier after pretreatment of silica sol with ammonia solution at low temperature. In a
low pretreatment temperature, there may be a lot of hydroxyl groups formed on the
silica particles. But when the pretreatment temperature was as high as 353 K, the
silica aggregated together, leading to a sharp decrease of the copper phyllosilicate.[6]
Thus, these understanding of the role of silica surface hydroxyls on copper
phyllosilicate formation and copper species distribution could provide practical
guidance for further design and fabrication of metal-silicate catalysts.
5. Conclusion
A simple and effective method to controllable manipulate the copper states in
Cu/SiO2 catalyst is established, which performed high catalytic activity and stability
for the hydrogenation of DMO to EG. It is demonstrated that surface hydroxyl groups

of silica sol dominate a crucial role in the preparation of Cu/SiO2 catalyst and the
formation of copper species. For the Cu/SiO2 catalyst prepared by silica sol with
increasing surface hydroxyl groups, the formation of copper phyllosilicate was more
kinetically favoured. And an enlarged amount of copper phyllosilicate displays a large
specific surface area of catalyst, resulting in the highly dispersed CuO nanoparticles
as well. The correlation establishes an excellent linear fit between surface hydroxyl
groups of silica sol and the amount of Cu+ surface area, which strongly supports the
insights of catalyst formation process. Moreover, the enhanced surface area of Cu+
and Cu0 species significantly increased the catalytic activity for the hydrogenation of
DMO to EG and displayed an excellent long-time stability. The understanding of
effect of surface hydroxyl groups on catalyst formation process may lead to new
possibilities in rational design of Cu/SiO2 catalysts.
Acknowledgement
We are grateful to the financial support from the National Key R&D Program of
China (2018YFB0605803) and the National Nature Science Foundation of China
(21878227, U1510203, 21706184).
References
[1] G. Xu, Y. Li, Z. Li, H. Wang, Kinetics of the hydrogenation of diethyl oxalate to ethylene
glycol, Ind. Eng. Chem. Res. 34 (1995) 2371-2378.
[2] H. Yue, Y. Zhao, X. Ma, J. Gong, Ethylene glycol: Properties, synthesis, and applications,
Chem. Soc. Rev. 41 (2012) 4218-4244.
[3] D. Brands, E. Poels, A. Bliek, Ester hydrogenolysis over promoted Cu/SiO2 catalysts, Appl.
Catal. A 184 (1999) 279-289.

[4] J. Gong, H. Yue, Y. Zhao, S. Zhao, L. Zhao, J. Lv, S. Wang, X. Ma, Synthesis of Ethanol via
syngas on Cu/SiO2 catalysts with balanced Cu0–Cu+ Sites, J. Am. Chem. Soc. 134 (2012) 13922-
13925.
[5] E. Poels, D. Brands, Modification of Cu/ZnO/SiO2 catalysts by high temperature reduction,
Appl. Catal. A 191 (2000) 83-96.
[6] Y. Wang, Y. Shen, Y. Zhao, J. Lv, S. Wang, X. Ma, Insight into the balancing effect of active
Cu species for hydrogenation of carbon–oxygen bonds, ACS Catal. 5 (2015) 6200-6208.
[7] Y. Zhao, S. Zhao, Y. Geng, Y. Shen, H. Yue, J. Lv, S. Wang, X. Ma, Ni-containing Cu/SiO2
catalyst for the chemoselective synthesis of ethanol via hydrogenation of dimethyl oxalate, Catal.
Today 276 (2016) 28-35.
[8] S. Zhao, H. Yue, Y. Zhao, B. Wang, Y. Geng, J. Lv, S. Wang, J. Gong, X. Ma, Chemoselective
synthesis of ethanol via hydrogenation of dimethyl oxalate on Cu/SiO2: Enhanced stability with
boron dopant, J. Catal. 297 (2013) 142-150.
[9] X. Zheng, H. Lin, J. Zheng, X. Duan, Y. Yuan, Lanthanum oxide-modified Cu/SiO2 as a high-
performance catalyst for chemoselective hydrogenation of dimethyl oxalate to ethylene glycol,
ACS Catal. 3 (2013) 2738-2749.
[10] C. Wen, A. Yin, Y. Cui, X. Yang, W.-L. Dai, K. Fan, Enhanced catalytic performance for
SiO2-TiO2 binary oxide supported Cu-based catalyst in the hydrogenation of dimethyloxalate,
Appl. Catal. A 458 (2013) 82-89.
[11] Y. Wang, Y. Zhao, J. Lv, X. Ma, Facile synthesis of Cu@CeO2 and its catalytic behavior for
the hydrogenation of methyl acetate to ethanol, ChemCatChem, 9 (2017) 2085-2090.
[12] C. Carlini, M. Di Girolamo, A. Macinai, M. Marchionna, M. Noviello, A. Galletti, G. Sbrana,
Selective synthesis of isobutanol by means of the Guerbet reaction: Part 2. Reaction of
methanol/ethanol and methanol/ethanol/n-propanol mixtures over copper based/MeONa catalytic
systems, J. Mol. Catal. A 200 (2003) 137-146.
[13] Y.B. Song, J. Zhang, J. Lv, Y.J. Zhao, X.B. Ma, Hydrogenation of dimethyl oxalate over
copper-based catalysts: Acid-base properties and reaction paths, Ind. Eng. Chem. Res. 54 (2015)
9699-9707.
[14] L. Chen, P. Guo, M. Qiao, S. Yan, H. Li, W. Shen, H. Xu, K. Fan, Cu/SiO2 catalysts prepared
by the ammonia-evaporation method: Texture, structure, and catalytic performance in
hydrogenation of dimethyl oxalate to ethylene glycol, J. Catal. 257 (2008) 172-180.
[15] A. Yin, X. Guo, W. Dai, K. Fan, The nature of active copper species in Cu-HMS catalyst for
hydrogenation of dimethyl oxalate to ethylene glycol: New insights on the synergetic effect
between Cu0 and Cu+, J. Phys. Chem. C 113 (2009) 11003-11013.
[16] A. Yin, X. Guo, W. Dai, H. Li, K. Fan, Highly active and selective copper-containing HMS
catalyst in the hydrogenation of dimethyl oxalate to ethylene glycol, Appl. Catal., A 349 (2008)
91-99.
[17] A. Yin, X. Guo, K. Fan, W. Dai, Influence of copper precursors on the structure evolution and
catalytic performance of Cu/HMS catalysts in the hydrogenation of dimethyl oxalate to ethylene
glycol, Appl. Catal. A 377 (2010) 128-133.
[18] A. Yin, X. Guo, K. Fan, W. Dai, Ion-exchange temperature effect on Cu/HMS catalysts for
the hydrogenation of dimethyl oxalate to ethylene glycol, ChemCatChem, 2 (2010) 206-213.
[19] X. Ma, H. Chi, H. Yue, Y. Zhao, Y. Xu, J. Lv, S. Wang, J. Gong, Hydrogenation of dimethyl
oxalate to ethylene glycol over mesoporous Cu/MCM-41 catalysts, AIChE J. 59 (2013) 2530-

2539.
[20] Y. Zhao, Z. Guo, H. Zhang, B. Peng, Y. Xu, Y. Wang, J. Zhang, Y. Xu, S. Wang, X. Ma,
Hydrogenation of diesters on copper catalyst anchored on ordered hierarchical porous silica: Pore
size effect, J. Catal. 357 (2018) 223-237.
[21] Y. Zhao, Y. Zhang, Y. Wang, J. Zhang, Y. Xu, S. Wang, X. Ma, Structure evolution of
mesoporous silica supported copper catalyst for dimethyl oxalate hydrogenation, Appl. Catal. A
539 (2017) 59-69.
[22] C. Zhang, D. Wang, M. Zhu, F. Yu, B. Dai, Effect of different nano-sized silica sols as
supports on the structure and properties of Cu/SiO2 for hydrogenation of dimethyl oxalate,
Catalysts 7 (2017) 75.
[23] T. Toupance, M. Kermarec, J.-F. Lambert, C. Louis, Conditions of formation of copper
phyllosilicates in silica-supported copper catalysts prepared by selective adsorption, J. Phys.
Chem. B 106 (2002) 2277-2286.
[24] Z. Huang, F. Cui, J. Xue, J. Zuo, J. Chen, C. Xia, Synthesis and structural characterization of
silica dispersed copper nanomaterials with unusual thermal stability prepared by precipitation-gel
method, J. Phys. Chem. C 114 (2010) 16104-16113.
[25] A. Marchi, J. Fierro, J. Santamaria, A. Monzon, Dehydrogenation of isopropylic alcohol on a
Cu/SiO2 catalyst: a study of the activity evolution and reactivation of the catalyst, Appl. Catal. A
142 (1996) 375-386.
[26] C. Van Der Grift, A. Wielers, A. Mulder, J. Geus, The reduction behaviour of silica-supported
copper catalysts prepared by deposition-precipitation, Thermochim. Acta 171 (1990) 95-113.
[27] C. Van Der Grift, P. Elberse, A. Mulder, J. Geus, Preparation of silica-supported copper
catalysts by means of deposition-precipitation, Appl. Catal. 59 (1990) 275-289.
[28] S. Zhu, X. Gao, Y. Zhu, Y. Zhu, H. Zheng, Y. Li, Promoting effect of boron oxide on Cu/SiO2
catalyst for glycerol hydrogenolysis to 1,2-propanediol, J. Catal. 303 (2013) 70-79.
[29] P. Ai, M. Tan, N. Yamane, G. Liu, R. Fan, G. Yang, Y. Yoneyama, R. Yang, N. Tsubaki,
Synergistic effect of a boron-doped carbon-nanotube-supported Cu catalyst for selective
hydrogenation of dimethyl oxalate to ethanol, Chem.-Eur. J. 23 (2017) 8252-8261.
[30] P. Burattin, M. Che, C. Louis, Molecular approach to the mechanism of deposition-
precipitation of the Ni (II) phase on silica, J. Phys. Chem. B 102 (1998) 2722-2732.
[31] D. Yao, Y. Wang, Y. Li, Y. Zhao, J. Lv, X. Ma, A High-performance nanoreactor for carbon–
oxygen bond hydrogenation reactions achieved by the morphology of nanotube-assembled hollow
spheres, ACS Catal. 8 (2018) 1218-1226.
[32] D. Yao, Y. Wang, K. Hassan-Legault, A. Li, Y. Zhao, J. Lv, S. Huang, X. Ma, Balancing effect
between adsorption and diffusion on catalytic performance inside hollow nanostructured catalyst,
ACS Catal. 9 (2019) 2969-2976.
[33] Iler, The Chemistry of Silica, Wiley & Sons, New York, 1979.
[34] A. Yin, X. Guo, W.-L. Dai, K. Fan, Effect of initial precipitation temperature on the structural
evolution and catalytic behavior of Cu/SiO2 catalyst in the hydrogenation of dimethyloxalate, Catal.
Commun. 12 (2011) 412-416.
[35] W. Yang, Y. Zhao, S. Wang, X. Ma, Formation of copper phyllosilicate in silica supported copper
catalyst, Chem. Ind. Eng. (China) 33 (2016) 1-5.

40%
30%
20%
10%
0%
(A)
(B)
3.0 3.5 4.0 4.5 5.0 5.5 6.0
Particle Size (nm)
40%
30%
20%
10%
0%
6
7 8 9 10 11 12 13 14 15
Particle Size (nm)
(C)
(D)
30%
20%
10%
0%
30%
20%
10%
0%
18 20 22 24 26 28 30 32 34
Particle Size (nm)
12 13 14 15 16 17 18 19 20 21 22
Particle Size (nm)
Fig. 1 TEM images and counted particle sizes (inset) of silica sol samples with
hydrothermal treatment of 0 h (A), 2 h (B), 4 h (C) and 12 h (D).

Fig. 2 FTIR spectra (A) of SiO2-N samples and their I960/I800 ratios (B).

1110
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
(B)
(A)
25
20
15
10
5
1040
470
800
670
Cu/SiO2-25
Cu/SiO2-17
Cu/SiO2-10
Cu/SiO2-5
Cu/SiO2-5 Cu/SiO2-10Cu/SiO2-17Cu/SIO2-25
Catalysts
2000
1600
1200
800
400
Wavenumber (cm-1)
Fig. 3 FTIR spectra (A) of calcined Cu/SiO2-N catalysts and their I670/I800 ratios (B).

Fig. 4 TEM images of calcined Cu/SiO2-5, Cu/SiO2-17 and Cu/SiO2-25 (C).

■ SiO2
▼ Cu2O
◆ Cu
■ SiO2
▼ CuO
(A)
(B)
◆
■
▼
▼
■
◆
Cu/SiO2-25
◆
Cu/SiO2-25
Cu/SiO2-17
▼
Cu/SiO2-17
Cu/SiO2-10
Cu/SiO2-5
Cu/SiO2-10
Cu/SiO2-5
20
40
60
80
20
40
60
80
2Ɵ (degree)
2Ɵ (degree)
Fig. 5 XRD profiles of Cu/SiO2-N catalyst upon calcination (A) and reduction (B).

486K
468K
Cu/SiO2-25
Cu/SiO2-17
Cu/SiO2-10
Cu/SiO2-5
373
473
573
673
773
Temperature (K)
Fig. 6 TPR profiles of the Cu/SiO2-N catalysts.