Hydrogenation of 3-hydroxypropanal to 1,3-propanediol over a Cu-V/Ni/SiO2 catalyst

Article abstract of DOI:10.1039/c7nj01746d

A Ni-based catalyst modified with copper and vanadium (5Cu-20V/30Ni/SiO₂) was synthesized and employed in the hydrogenation of 3-hydroxypropanal (3-HPA) to 1,3-propanediol (1,3-PDO). Catalysts were systematically characterized via XRD, TEM, HRTEM, SAED, H₂-TPD, N₂ physisorption, H₂-TPR, and XPS. It was indicated that the addition of Cu and V not only promoted the reduction of Ni²⁺ species and the generation of active hydrogen species, but also increased the dispersion of Ni species and the interaction between Ni particles and SiO₂. The catalytic performance evaluation showed that the (5Cu-20V/30Ni/SiO₂) could provide the largest yield of 1,3-PDO (above 76.8%) and highest TOF (4.32 × 10³ s^-1). The structure-activity relationship was clarified according to the characterization results. The optimized reaction conditions over the 5Cu-20V/30Ni/SiO₂ catalyst were obtained as the reaction temperature of 80 °C, a H₂ pressure of 2.0 MPa (H₂), and an LHSV of 0.4 h^-1. The employment of non-noble metals, the relatively low pressure of H₂, and the relatively high yield of 1,3-PDO make 5Cu-20V/30Ni/SiO₂ an efficient and economic catalyst that may have potential applications in industry.

Full text of DOI:10.1039/c7nj01746d

View Article Online
View Journal
NJC
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: L. Li, W. Yi, T. Liu,
C. Huang, A. Li, C. Luo and Z. Chao, New J. Chem., 2017, DOI: 10.1039/C7NJ01746D.
This is an Accepted Manuscript, which has been through the
Volume 40 Number
1 January 2016 Pages 1–846
Royal Society of Chemistry peer review process and has been
accepted for publication.
NJC
Accepted Manuscripts are published online shortly after
New Journal of Chemistry
www.rsc.org/njc
A journal for new directions in chemistry
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
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
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1144-0546
PAPER
Jason B. Benedict et al.
The role of atropisomers on the photo-reactivity and fatigue of
diarylethene-based metal–organic frameworks
rsc.li/njc

Page 1 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
Hydrogenation of 3-Hydroxypropanal to
1,3-Propanediol over Cu-V/Ni/SiO2 Catalyst
LiꢀJun Lia†, WenꢀJun Yia†, TianꢀWei Liua, Chen Huanga,
An Lia, CaiꢀWu Luoa, ZiꢀSheng Chaoa,b*
a State Key Laboratory of Chemo/Biosensing and
Chemometrics, College of Chemistry and Chemical
Engineering
China
， Hunan University, Changsha, 410082,
b College of Materials Science and Engineering,
Changsha University of Science & Technology,
Changsha, Hunan 410114, China
† These authors contributed equally to this work.
*Corresponding author: Prof. ZiꢀSheng Chao;
Tel.＆Fax: +86ꢀ731ꢀ88713257;
Eꢀmail: zschao@yahoo.com; chao_zs@aliyun.com; zschao@hnu.edu.cn.
1

New Journal of Chemistry
Page 2 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
Abstract
A Copper and Vanadium modified Niꢀbased catalyst (5Cuꢀ20V/30Ni/SiO2) was
synthesized and employed in the hydrogenation of 3ꢀhydroxypropanal (3ꢀHPA) to
1,3ꢀpropandiol (1,3ꢀPDO). The catalysts were systematically characterized by means
of XRD, TEM, HRTEM, SAED, H2ꢀTPD, N2ꢀphysisorption, H2ꢀTPR and XPS. It was
indicated that the addition of Cu and V not only promoted the reduction of the Ni2+
species and the generation of active hydrogen species, but also increase the dispersion
of Ni species and the interaction between the Ni particles and SiO2. The catalytic
performance evaluation showed that the (5Cuꢀ20V/30Ni/SiO2) could provide the
largest yield of 1,3ꢀPDO (above 76.8%) and highest TOF(4.32 × 103 Sꢀ1).
Structureꢀactivity relationship is clarified according to characterization results. An
optimized reaction conditions over the 5Cuꢀ20V/30Ni/SiO2 catalyst has been obtained
as the reaction temperature is 80 ℃, H2 pressure is 2.0 MPa (H2) and LHSV is 0.4 hꢀ1.
The application of nonꢀnoble metal, the relatively low pressure of H2 and the
relatively high yield of 1,3ꢀPDO make 5Cuꢀ20V/30Ni/SiO2 an efficient and economic
catalyst that may have potential applications in industry.
Keywords: CuꢀV/Ni/SiO2; 3ꢀhydroxypropanal; 1,3ꢀpropanediol; hydrogenation
2

Page 3 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
1. Introduction
1,3ꢀpropanediol (1,3ꢀPDO) is an important chemical intermediate, which is
widely applied in many fields, such as monomer of resins, adhesives, solvents,
cosmetics, detergents, freshnessꢀkeeping agents, biocides, feedstuff,1 especially for
the manufacture of polytrimethylene terephthalate (PTT).2,3 In the past decades,
biological and chemical routes have been successfully developed for the production of
1,3ꢀPDO. The biological route involves a process for the bioconversion the renewable
resources (such as, glucose or glycerol) to 1,3ꢀPDO.3,4 This route, however, was
accompanied with a few drawbacks, i.e., long reaction time, low 1,3ꢀPDO yield5,
complicated purification of 1,3ꢀPDO from fermentation broth and large discharge of
waste water.6 These factors have limited its development in the industrial field.
Nowadays, the chemical route to produce1,3ꢀPDO is the predominant method in
industry. The second routes involved either the hydrogenolysis of glycerol or the
hydrogenation of 3ꢀhydroxypropanal (3ꢀHPA) to 1,3ꢀPDO. The direct hydrogenolysis
of glycerol always involves low yields of 1,3ꢀPDO,7 harsh conditions8 and noble
metal catalyst.9 This method has not been applied in industry upon till now. The
hydrogenation of 3ꢀHPA (3ꢀhydroxypropanal) could be conducted under relatively
warm reaction condition and provides a very high yield of 1,3ꢀPDO, and is studied in
this work.
Much studies have been conducted to evaluate the hydrogenation of
3ꢀhydroxypropanal in recent years, among which the catalyst is most investigated.
Three groups of catalysts were mainly studied in these works. The first group
comprises noble metal catalysts. For example, a yield of 1,3ꢀPDO reached 89.0% over
5% Ru/SiO2 at 4.0 MPa,10 whereas a yield of 1,3ꢀPDO reached 98.7% over 2%
Pt/TiO2 at 15.0 MPa.11 Although Ptꢀbased catalysts give a higher yield relative to the
Ruꢀbased catalysts, harsh reaction conditions such as high pressure have to be
confronted. Nevertheless, for Ruꢀbased catalysts, the low yield of 1,3ꢀPDO and the
high loading (5 wt%) are still great challenges. The second group is Raney Ni and
nickel alloy catalysts. although the yields of 1,3ꢀPDO are higher than 99.5% over both
3

New Journal of Chemistry
Page 4 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
Raney nickel12 and nickel alloy,13 the hard separation of catalyst fine powder and
reaction products and thus the difficult recycle use of catalyst and low purity of
product still remained as the great. The third group composed of supported transition
metal oxide catalysts, in particular, nickel oxide. For instance, a yield of ca. 81% yield
over an Ni/SiO2ꢀAl2O3 catalyst was obtained at 15.0 MPa.14 Compared to the
supported noble metal catalysts, supported nickel oxide catalysts usually provide
relatively low yields of 1,3ꢀPDO (above 80%). There is no doubt that the
development of a lowꢀcost and highly effective catalyst is currently highly desired.
As is well studied in literature, Cuꢀbased catalysts were also widely used for
17
aldehyde hydrogenation. CuꢀCo/SiO2,15 Cu/MCMꢀ4116 and Cu/SiO2 had been used
in the hydrogenation of cinnamaldehyde to cinnamon alcohol. Cu/CeO2 had also been
applied in the hydrogenation of crotonyl aldehyde to crotonyl alcohol.16 The Cuꢀbased
catalyst showed great catalytic performance in the hydrogenation of aldehyde
compounds. As is generally known, the reduction temperature of the Cu species
catalyst (220 oC) is lower than that of the Ni species (about 400 oC). NiꢀCu/TiO218 had
also been employed in the hydrogenation of cinnamaldehyde to cinnamon alcohol.
NiꢀCu/CNTs19 had been used in the hydrogenation of furfural to tetrahydrofurfuryl
alcohol. The combination of the Cu and Ni in the catalyst greatly enhance the yields
of the alcoholic products. The main drawback of Cu–Ni catalyst is that it could
deactivated easily because of coke deposition, and/or metal particle sintering during
the reaction. The Cu and Ni particles (about 50 nm) aggregated together in the
unmodified CuꢀNi/SiO2 catalyst, while the modified CuꢀNi/SiO2 showed more
uniform particle dispersion of the Cu and Ni particles.20 In a similar way, the amount
of deposited coke on Cu–Ni/SBAꢀ15 was significantly reduced throughout the
modification of SBAꢀ15 mesostructured silica support with Ca and Mg oxides.20 The
modification of the CuꢀNi catalyst is expected to obtain both good activation
performance and stability during the hydrogenation reaction.
VOx as a promoter has been studied in different catalysts and reactions recently.
For example, Mori et al.21 investigated the VOx promoter in Al2O3 supported Ru
4

Page 5 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
catalyst and proposed a promotion mechanism for dissociation of CO on Vꢀpromoted
Ru/Al2O3 catalyst during the hydrogenation of CO. Karaselçuk et al.22 studied the
VOx promoter in Al2O3 supported Ni catalysts. It revealed that V promotion increases
the CO hydrogenation activity and specific activity of Ni catalysts. Lu et al.23 also
studied the location of VOx promoter in bentonite supported Ni catalysts and revealed
that the electronic effect of VOx can promote dissociation of CO in the methanation
reaction. The addition of V could help the adsorption of H2, which might have
positive effect on the hydrogenation reaction. Also the addition of V could inhibit the
formation of coke in the 20Nꢀ3V/bentonite catalysts.23 V was found to greatly
enhance the performance of the lots of catalyst (Vꢀpromoted Ru/Al2O3,21 Vꢀpromoted
Ni/Al2O322 and Vꢀpromoted Ni/bentonite23) via promoting the dissociation of CO. It is
well conformed that V could greatly enhance the performance of the above catalyst
via promoting the dissociation of CO. While the C=O bond in the 3ꢀHPA may be
activated through the similar mechanism, the addition of V is expected to promote the
catalytic performance of the catalyst. By combining V and Cu with Ni/SiO2, the
trimetallic catalyst may display excellent catalytic performance in the synthesis of
1,3ꢀPDO.
In this paper, for the first time, the CuꢀV modified Niꢀbased catalyst was prepared
and employed in the hydrogenation of 3ꢀHPA toward 1,3ꢀPDO. The catalysts were
characterized by a series of techniques and their performance were evaluated. The
structureꢀactivity relationship was discussed for better insight into the catalyst. The
optimal conditions were also established for potential applications in industry.
2. Experimental
2.1. Chemicals
Nickel nitrate hexahydrate (Ni(NO3)2·6H2O), ammonium vanadate (NH4VO3),
copper dinitrate (Cu(NO3)2) and Colloidal silica solution (commercial grade,
SiO2·nH2O, 30.0 wt%, Na2O type, pH≈10) were purchased from Sinopharm Chemical
Reagent limited corporation. All chemicals were used without further purification.
5

New Journal of Chemistry
Page 6 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
2.2. Preparation of catalyst
The monometallic catalysts were prepared by the follow process: calculated
amounts of Ni(NO3)2·6H2O and colloidal silica solution were mixed at room
temperature with strong stirring for 8h. The resultant mixture was dried overnight at
o
o
120 C, and then, calcined in ambient air by increasing the temperature to 550 C at
o
10 C minꢀ1 and holding that temperature for 4 h, noted as yNi/SiO2 (y =15 or 30).
Herein, y refers to the theoretical weight percent of nickel in the catalyst.
The trimetallic catalysts were prepared as follow process: Calculated amounts of
NH4VO3, Cu(NO3)2, (Ni(NO3)2·6H2O and colloidal silica solution were mixed at
room temperature with strong stirring for 8h. Then the resultant mixture was dried
overnight at 120 oC, and calcined in ambient air by increasing the temperature to 550
o
oC at 10 C minꢀ1 and holding it for 4h. The obtained catalyst was noted as
5Cuꢀ20V/30Ni/SiO2. Herein, 5, 20 and 30 refers to the theoretical weight percent of
copper, vanadium and nickel in the catalyst, respectively.
For the purpose of comparison, a unsupported SiO2 sample was prepared as the
following procedure: A certain amount of colloidal silica solution was dried overnight
at 120 oC, and then, calcined in ambient air by increasing the temperature to 550 oC at
o
10 C minꢀ1 and holding it for 4 h. Pure CuO sample was prepared as the following
procedure: A certain amount of Cu(NO3)2 calcined in ambient air by increasing the
temperature to 550 oC at 10 oC minꢀ1 and holding it for 4 h. Pure NiO sample was also
prepared as the above procedure.
Before being employed in the hydrogenation of 3ꢀHPA to 1,3ꢀPDO, all the
catalysts were pressed into small discs, then crushed and sieved to 20ꢀ40 mesh.
Reduced samples were prepared as the following procedure: the calcined
samples were reduced in an ambient hydrogen atmosphere by increasing the
temperature to 550 oC at 10 oC minꢀ1 and holding it for 4 h. After that, the catalyst was
cooled to room temperature under the hydrogen atmosphere.
2.3. Catalyst characterization
6

Page 7 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
Xꢀray diffraction spectroscopy (XRD) was performed with a Bruker
D8ꢀAdvance XꢀRay diffractometer, under the following conditions: Cu target Kα ray
(λ = 1.54187 Å); scanning voltage 40 kV; scanning current 40 mA; scanning speed 6o
minꢀ1; scanning step 0.02o. The crystal sizes of Ni particles were calculated by the
Scherrer equation using the peak at 2θ = 44.4o.24
Transmission electron microscopy (TEM), high resolution transmission electron
microscopy (HRTEM) and selected area electron diffraction (SAED) were performed
with a JEMꢀ2100F transmission electron microscope operated at an acceleration
voltage of 200 kV. Before determination, the specimen was dispersed into ethanol by
ultrasonic treatment and then the resultant suspension was dropped onto a
carbonꢀcoated copper grid.
H2 temperature programmed desorption (H2ꢀTPD) was performed using a
Quantachrome Autosorbꢀ1 instrument equipped with a thermal conductivity detector
(TCD). The calcined catalyst (500 mg) was placed in a quartz reactor, increasing the
◦
◦
temperature from room temperature to 550 C at 10 C minꢀ1 and holding that
temperature for 4 h in a stream of 10% H2/Ar with a flow rate of 40 mL minꢀ1. Then,
the catalyst was cooled to 40 ◦C and exposed to a stream of 10% H2/Ar with flow rate
of 40 mL minꢀ1 till the saturated adsorption of H2. After that, the sample was purged
with N2 at room temperature for 30 min to remove the physically adsorbed H2. The
desorption of H2 was carried out in a stream of N2 with flow rate of 30 mL minꢀ1 from
◦
40 ◦C to 600 ◦C at a heating rate of 10 C minꢀ1. The amount of desorbed hydrogen
was determined by TCD. Calibration of the hydrogen uptake was done by injecting a
certain volume of pure hydrogen into the inlet stream of the reducing gas. Based on
the peak area of the H2ꢀTPD profiles and the standard calibration process, the
hydrogen uptake can be estimated, assuming that one H atom was adsorbed per metal
atom.25 This enabled the degree of dispersion to be calculated, according to equation
(1):26
D% = 1.17ꢀ/ꢁꢂ
(1)
Where X = H2 uptake in ꢁmole per g of catalyst, W = the weight percentage of
7

New Journal of Chemistry
Page 8 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
nickel, and f = the fraction of nickel reduced to the metal.
The average crystallite diameter, d, was calculated from the value of D%,
assuming that that the crystals were all spherical and possessed the same dimension,26
as shown by equation (2):
ꢃ
ꢄ
d nm = 97.1/ꢃꢅ%ꢄ
(2)
N2ꢀphysisorption was carried out with a Quantachrome Autosorbꢀ1C/TCD
automated physisorption and chemisorption analyzer at liquid N2 temperature. Before
the measurements, the reduced catalyst was degassed at 200 oC for 3 h under a 4×10−4
Pa. The BET method was employed to calculate the specific surface area, with a
correlation coefficient of above 0.9999. The total pore volume was recorded at
P/Po=0.99. The pore size distribution was determined by the standard BarrettꢀJoyner
ꢀHalenda (BJH) method.
H2 temperature programmed reduction (H2ꢀTPR) was conducted using a
Quantachrome Autosorbꢀ1 instrument equipped with a thermal conductivity detector
(TCD). The catalyst specimen (50 mg) was placed in a quartz reactor, pretreated in a
o
o
flow of N2 (40 mL min−1) at 400 C for 30 min and then cooled down to 50 C in a
flow of N2 (40 mL min−1). After that, the catalyst was exposed to a flow of 10% H2/Ar
with a flow rate of 40 mL min−1 while the temperature was raised from 100 oC to 700
oC at a rate of 10 oC min−1 and the signal was simultaneously recorded by TCD.
Xꢀray photoelectron spectroscopy (XPS) was performed using a KꢀAlpha XPS
system with a monochromatic Al Kα Xꢀray source and a charge neutralizer. The
binding energies (B.E.) of O 1s, Ni 2p3/2, V 2p3/2 and Cu 2p coreꢀlevels were recorded.
The B.E. was referenced to the C 1s peak at 284.8 eV for the surface adventitious
carbon.27ꢀ29 The peaks were fitted in order to determine the chemical states of specific
species present on the samples. Curve fitting was carried out using a nonlinear
leastꢀsquares fitting program with a mix of Gaussian and Lorentzian shapes and a
Shirley baseline.30
The contents of the Cu、V and Ni in the reduced and used 5Cuꢀ20V/30Ni/SiO2
catalysts were determined by inductively coupled plasma (ICP) optical emission
8

Page 9 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
spectroscopy (Varian Vista Axial).
2.4. Hydrogenation of 3-HPA to 1,3-PDO
The reaction was performed in a fixedꢀbed flowꢀtype stainless reactor, which was
mounted in a tubular electric heater. The catalyst (1.0 g) was loaded in the middle of
the reactor, and the upper space of the catalyst bed in the reactor was filled with inert
quartz granulates (10ꢀ20 mesh). The catalyst was first in situ activated by reducing in
hydrogen atmosphere at 550 oC for 4 h, and then it was cooled to a preset temperature
o
in the range of 50~100 C. Then, a flow of H2 at a certain pressure in the range of
1.0~5.0 MPa and a stream of reactant, i.e., 10 wt% aqueous solution of 3ꢀHPA, were
simultaneously introduced into the reactor. The flow rates of the reactant and H2 were
regulated so that the liquid hourly space velocity (LHSV) had a certain value in the
range of 0.2~1.0 h−1. The reaction was conducted under the above conditions. The
effluent from the bottom of the reactor was passed through a gas–liquid separator
cooled by ice water. The liquid products were collected for analysis.
Qualitative and quantitative analysis of liquid products mixture was carried out
with a Varian Saturn 2200/CP 3800 GC/MS instrument, equipped with a flame
ionization detector (FID) and two CPꢀWax 52CB fused silica capillary columns (15 m
× 0.32 mm).
Several possible reactions are supposed to happen during the hydrogenation of
3ꢀHPA as follows.31 Among which, reactions (4), (5), (6), and (8) are reversible, while
the (3), (7) and (9) are irreversible.
main reaction
3 − HPA + Hꢆ → 1,3 − PDO
(3)
(4)
side reactions
3 − HPA ↔ acrolein + HꢆO
3 − HPA + HꢆO ↔ 3 − HPA − hydrate
3 − HPA ↔ 3 − HPA − dimer
3 − HPA + acrolein → 4 − oxa − 1,7 − heptanedial
(5)
(6)
(7)
(8)
4 − oxa − 1,7 − heptanedial ↔ 4 − hydroxy − 3 − formylteteahydropyran
9

New Journal of Chemistry
Page 10 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
4 − oxa − 1,7 − heptanedial + Hꢆ → 4 − oxa − 1,7 − heptanediol
(9)
The conversion of 3ꢀHPA (X3ꢀHPA), selectivity for 1,3ꢀPDO (S1,3ꢀPDO), yield of
1,3ꢀPDO (Y1,3ꢀPDO) and turnꢀover frequency (TOF)32ꢀ34 were respectively calculated by
equations 10~13:
ꢒ
ꢓ
ꢐ
ꢉꢐ
ꢈꢑꢊꢋꢌ
ꢈꢑꢊꢋꢌ
ꢃ
ꢄ
ꢇꢈꢉꢊꢋꢌ ꢍꢎꢏ% =
× ꢓꢒꢒ
× ꢓꢒꢒ
(10)
(11)
ꢒ
ꢐ
ꢈꢑꢊꢋꢌ
ꢐ
ꢓ,ꢈꢑꢋꢕꢖ
ꢓ
ꢃ
ꢄ
ꢔꢓ,ꢈꢉꢋꢕꢖ ꢍꢎꢏ% =
ꢒ
ꢐ
ꢉꢐ
ꢈꢑꢊꢋꢌ
ꢈꢑꢊꢋꢌ
ꢃ
ꢄ
ꢗꢓ,ꢈꢉꢋꢕꢖ ꢍꢎꢏ% = ꢇꢈꢉꢊꢋꢌ × ꢔꢓ,ꢈꢉꢋꢕꢖ/ꢓꢒꢒ
(12)
(13)
denoted the molar content of 3ꢀHPA in the reactant
ꢘꢖꢙ ꢃꢚꢉꢓꢄ = ꢗ
× ꢙ
ꢓ,ꢈꢑꢋꢕꢖ
ꢈꢑꢊꢋꢌ
ꢛ × ꢕ
Where n0
and n1
3ꢀHPA
3ꢀHPA
and product, respectively; n1,3ꢀPDO referred to the molar content of 1,3ꢀPDO in the
product; W, F3ꢀHPA and D was the weight of catalyst, molar flow rate of 3ꢀHPA, and
the dispersion of active site, respectively.
3. Results and discussions
3.1 Characterizations
Fig. 1 shows the XRD patterns for the reduced catalysts and used catalysts. The
broad diffraction peak at 2θ = 22° present in Fig. 1a is attributed to the typical
amorphous state of the SiO2 vitreous matrix.35 The weak diffraction peak at 2θ =
24.3°, 33.0°, 36.2°, 53.9°, and 65.2° are attributed to V2O3 (JCPDS No. 34ꢀ0187).
This peak of metallic Cu was not to be observed over the reduced and used
5Cuꢀ20V/30Ni/SiO2 catalysts (Fig. 1d and g). It may be due to the fact that the copper
species, which are amorphous or very small, are highly dispersed within the matrix.36
The diffraction peaks at 2θ = 44.3o, 51.6o and 76.3o over the 15Ni/SiO2, 30Ni/SiO2
and 5Cuꢀ20V/30Ni/SiO2 (Fig. 1b~g) catalysts, and they are attributed to the metallic
Ni (JCPDS No. 04ꢀ0850).24,25 The intensity of these three peaks increases obviously
with increasing the content of Ni in the catalyst and decreased largely over the used
catalysts relative to the corresponding reduced ones (cf. Fig. 1b ~ g). The weak
diffraction peaks at 2θ = 37.3o, 43.3o and 62.9o are present over the used 15Ni/SiO2
10

Page 11 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
and 30Ni/SiO2 catalysts (Fig. 1e and f), with their intensity increasing as the increase
of Ni content in the catalyst, but they are absent over the corresponding reduced ones
(Fig. 1b and c). These three peaks are attributed to cubic NiO (JCPDS No 47ꢀ1049).
Besides, these peaks for cubic NiO are absent over the used 5Cuꢀ20V/30Ni/SiO2
catalyst. This result indicates that, during the reaction, a proportion of metallic Ni
species have been transformed into NiO species over the catalysts containing no VꢀCu,
but the above transformation of Ni species can be inhibited due to the loading of
VꢀCu onto the catalyst. The reason for the above transformation of Ni species is most
probably due to the fact that there is the interaction between Ni species and SiO2
network, forming the SiꢀOꢀNi species. During the reaction, the SiꢀOꢀNi species can be
hydrolyzed into SiꢀOH and Ni(OH)x by the water in 3ꢀHPA solution, and the
dehydrogenation of Ni(OH)x leads to the formation of NiO species. However, the
presence of Cu can initiate easily the generation of active H species at a relatively low
temperature, and this kind of active H species are transferred onto the NiO species via
hydrogen spillꢀover process, and thus, enables the NiO species to be reduced to
metallic Ni species, during the hydrogenation of 3ꢀHPA to 1,3ꢀPDO. In this sense, the
presence of Cu stabilizes the metallic Ni species, avoiding the transformation of
metallic Ni into NiO.
• SiO2
• SiO2 ♦ Ni ∇ NiO
♦ Ni ◊ V2O3
♦
♦
♦
g
♦
◊
◊
◊
◊
•
d
♦
◊
♦
•
♦
♦
♦
∇
♦
♦
•
∇
∇
∇
∇
♦
♦
c
f
•
♦
•
•
•
e
b
a
∇
♦
10
20
30
40
50
60
70
80 10
20
30
40
50
60
70
80
2-Theta (degree)
2-Theta (degree)
Fig. 1 XRD patterns of reduced (left) and used (right) catalysts. (a) SiO2, (b)
11

New Journal of Chemistry
Page 12 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
15Ni/SiO2, (c) 30Ni/SiO2, (d) 5Cuꢀ20V/30Ni/SiO2, (e) used 15Ni/SiO2, (f) used
30Ni/SiO2, (g) used 5Cuꢀ20V/30Ni/SiO2. The catalysts were used for 4 h for (e) and
(f). The catalyst was used for 75 h for (g).
The TEM micrographs and particle size distributions for various reduced
catalysts are shown in Fig. 2. One can see from the TEM images that the particles of
30Ni/SiO2 catalyst are aggregated to a larger extent and possess a broader size
distribution as well as larger average size, relative to those of 15Ni/SiO2 catalyst (cf.
Fig. 2a and b and Fig. 2c and d). Compared to the 30Ni/SiO2 catalyst, both the extent
of aggregation and average size of particles become smaller and the particle size
distribution narrower for the 5Cuꢀ20V/30Ni/SiO2 catalyst (cf. Fig. 2c and d and Fig.
2e and f). The above result can be due to the fact that the increase in the Ni content
leads to the generation of larger Ni particles; however, the sintering of Ni particles is
inhibited, to some extent, by the presence of Cu. The EDS mapping of
5Cuꢀ20V/30Ni/SiO2 catalyst was shown in Fig. S1. According to the EDS mapping of
the 5Cuꢀ20V/30Ni/SiO2 catalyst, the Ni, V and Cu dispersed on the surface of the
catalyst uniformly. It confirmed that the Cu, VOx and Ni do not aggregated to form
large particles.
12

Page 13 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
30
(a)
(b)
25
20
15
10
5
0
10
10
10
15
15
15
20
20
20
25
25
25
30
30
30
35
35
35
40
20
15
10
5
(c)
(d)
0
40
25
20
15
10
5
(f)
(e)
0
40
Particle size (nm)
Fig. 2 TEM micrographs (left) and particle size distributions (right) for various
reduced catalysts. (a, b) 15Ni/SiO2; (c, d) 30Ni/SiO2; (e, f) 5Cuꢀ20V/30Ni/SiO2.
The HRTEM micrograph and SAED pattern for the reduced 5Cuꢀ20V/30Ni/SiO2
catalyst are shown in Fig. 3. One can see that some particles exhibit lattice fringes of
Ni, Cu and V2O3 on their surface. The HRTEM image of 5Cuꢀ20V/30Ni/SiO2 (Fig. 3a)
displays overlapping lattice fringes of 0.20 nm, 0.25 nm, 0.27 nm, and 0.37 nm which
are attributed to the Ni (111) crystalline plane (JCPDS No. 04ꢀ0850), Cu (111)
crystalline plane (JCPDS No. 78ꢀ2076), V2O3 (104) crystalline plane and V2O3 (012)
crystalline plane (JCPDS No. 34ꢀ0187), respectively. This indicates that Cu particles
and V2O3 particles are in close contact with Ni particles. The above results can be
further supported by H2ꢀTPR. The SAED pattern of reduced 5Cuꢀ20V/30Ni/SiO2 (Fig.
3b) exactly matches the reflections corresponded to metallic Ni, metallic Cu and V2O3.
The SAED studies confirm the nanocrystalline nature of the Ni particles, Cu particles
and V2O3 particles.
13

New Journal of Chemistry
Page 14 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
(b)
(a)
0.25 nm Cu (111)
0.27 nm V O (104)
0.20 nm Ni (111)
2
3
0.20 nm Ni (111)
0.37 nm V O (012)
2 3
0.2nm Ni(111)
0.27 nm V O
(104)
3
2
0.37 nm V O (012)
2
3
0.25 nm Cu (111)
Fig. 3 HRTEM micrographs (a) and SAED pattern (b) for the reduced
5Cuꢀ20V/30Ni/SiO2 catalyst.
Table 1 lists the degree of dispersion (D) and average particle size (d) of Ni over
various reduced catalysts. For the yNi/SiO2 (y = 15 or 30) catalysts, the degree of
dispersion of Ni decreases and the average particle size of Ni become larger with
increasing Ni content. In comparison with the 30Ni/SiO2 catalyst, the degree of
dispersion of Ni increased and average particle size become smaller over the
5Cuꢀ20V/30Ni/SiO2 catalyst. This indicates that addition of CuꢀV could increase the
Ni dispersion and hinder the formation of larger Ni particles. This result is in
agreement with that of TEM.
Table 1 Degree of dispersion and average particle size of Ni over various
reduced catalysts
d (nm)
Catalyst
D a (%)
H2ꢀTPD b
7.7
XRD c
25.4
15Ni/SiO2
30Ni/SiO2
12.6
7.9
12.3
27.3
5Cuꢀ20V/30Ni/SiO2
9.6
10.1
18.8
a Degree of dispersion determined by H2ꢀTPD;
b Average particle sizes determined by H2ꢀTPD;
c Crystal sizes calculated by the Scherrer equation using the peak at 2θ = 44.4o.
14

Page 15 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
Fig. 4 shows the N2 adsorptionꢀdesorption isotherms and pore size distribution
profiles (PSDs) for the reduced and used catalysts. For the reduced catalysts, all the
isotherms show the type IV isotherms in the IUPAC classification with type H2
hysteresis loops,37 which demonstrates the presence of mesopores.38,39 The PSDs
determined from the isotherms indicate that there are a number of mesopores of sizes
of around 3~22 nm in the samples, which are attributed to particle piled pore.40 The
pore size (Dp) increased with an increase in the content of Ni, while decreased with
the addition of Cu and V. A possible reason for the increase in the Dp values is that
small amount of Ni could be incorporated in the silica network and thus hindered the
sintering of catalyst particles.41 A possible reason for the decrease in the Dp values is
that the pore is blocked by V2O3 particles or Cu particles.28,42 For the used catalysts,
all the isotherms also correspond to type IV isotherms with type H2 hysteresis loops,
which demonstrates that the mesoporous structure of the catalysts remain. The pore
size of used catalysts decreased significantly. It may be caused by that the coke
formed in the side reactions blocked the pores.
Table 2 lists the textural properties of reduced and used catalysts. For the reduced
catalysts, one can see that the specific surface areas (SBET) are display in the order:
15Ni/SiO2 > 30Ni/SiO2 > 5Cuꢀ20V/30Ni/SiO2. This can be due to the aggregation or
agglomeration of the Ni particles43 and/or the pore was blocked by V2O3 particles or
Cu particles.28,42 Nevertheless, the pore volumes and average pore size in these
catalysts are both increased with increasing Ni content. The main reasons are as
follows: (i) the smaller sized pores are blocked by Ni particles;44 (ii) the
agglomeration of Ni particles may generate some new pores, thereby, increasing the
pore volume.41 After addition of CuꢀV, the pore volume and the average pore size are
all decreased because the pore are blocked by V2O3 particles or Cu particles.28,42 For
the used 15Ni/SiO2, 30Ni/SiO2 and 5Cuꢀ20V/30Ni/SiO2 catalysts, the increase of the
SBET may greatly due to the increase of the Smic. The increase of the Smic may be
caused by the coke formed from side reactions. After use for 75 h, the Smes of
5Cuꢀ20V/30Ni/SiO2 decrease from 72.7 m2 gꢀ1 to 46.9 m2 gꢀ1, while the Smic increase
15

New Journal of Chemistry
Page 16 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
from 2.7 m2 gꢀ1 to 38.8 m2 gꢀ1.
(A)
(B)
g
f
c
a
d
g
b
e
d
c
e
f
b
a
0.0
0.2
0.4
0.6
0.8
1.0
5
10
15
20
25
Relative Pressure (P/Po)
Diameter (nm)
Fig. 4 N2 adsorption–desorption isotherms (A) and pore size distribution (B) for
the reduced and used catalysts. (a) SiO2, (b) 15Ni/SiO2, (c) 30Ni/SiO2, (d)
5Cuꢀ20V/30Ni/SiO2, (e) used 15Ni/SiO2, (f) used 30Ni/SiO2, (g) used
5Cuꢀ20V/30Ni/SiO2. (e) and (f) were both used for 4 h and (g) for 75 h, respectively.
Table 2 Textural properties for reduced and used catalysts
SB
Sm
Sm
Vtot
D
ET
ic
es
al
p
Catalyst
(
(
(
(c
(
m2 gꢀ1) m2 gꢀ1) m2 gꢀ1) m3 gꢀ1) nm)
13
6.2
0.
5.
2.
0.
14
13
6.2
0.2
0.3
0.3
0.2
0.3
3
1
1
1
1
SiO2
0
3
7
0
4
8
9
9
.5
98
93
72
11
93
90
72
98
15Ni/SiO2
30Ni/SiO2
.6
.2
.7
.3
.5
.7
5.3
6.8
5.8
5Cuꢀ20V/30Ni/Si
O2
DE
15Ni/SiO2
16

Page 17 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
3.1
98
.3
4
.8
.5
.9
2
2
3
1.0
3.1
0.5
7.
91
46
0.3
0.2
1
1
DE
30Ni/SiO2
.9
5Cuꢀ20V/30Ni/Si
85
38
DE*
O2
.7
.8
Note: SBET, Smic and Smes refers to specific surface area, micropore surface area and mesopore
surface area, respectively, and SBET = Smic + Smes. Vtotal refers to total pore volume; Dp refers to
average pore size. DE: the catalyst was used for 4 h. DE*: The catalyst was used for 75 h.
Fig. 5 shows the H2ꢀTPR profiles for various unreduced catalysts. The broad
peaks in the profiles have been deconvolved and fitted to the component peaks, and
the temperatures at the maxima and the area percentages of peaks, which
corresponding to the reduction temperatures and percentage contents of Ni species,
are summarized in Table S1 (ESI*). One can see from Fig. 5a that single peak at 300
o
oC (pure CuO) and 375 C (pure NiO) are found, respectively. Over the 15Ni/SiO2
o
(Fig. 5b) catalyst, the reduction peaks appear 396.1 C, 461.1 oC and 590 oC, which
can be respectively ascribed to the small NiO particles (SꢀNiO particles),45 NiO
particles weakly interacted with SiO2 (wꢀNiOꢀSiO2 particles),46 and NiO particles
strongly interacted with SiO2 (SꢀNiOꢀSiO2 particles).47 As shown in Table S1
(15Ni/SiO2), the area percentages of these three kinds of NiO particles were 65.7%,
32.0% and 2.3%, respectively. Over the 30Ni/SiO2 (Fig. 5c) catalyst, the reduction
o
peaks appear 300.0 C, 381.6 oC and 456.7 oC, which can be respectively ascribed to
the large NiO particles48 (LꢀNiO particles, 2.3%), SꢀNiO particles (74.6%) and
wꢀNiOꢀSiO2 particles (23.1%), respectively. There is no evident reduction peak
attributed to the reduction of SꢀNiOꢀSiO2 particles. Over the 5Cuꢀ20V/30Ni/SiO2 (Fig.
o
5d) catalyst, the reduction peaks appear 241.8 oC, 273.2 oC, 365.1 oC, 407.7 C, 486.2
o
oC and 520.4 C, which can be respectively ascribed to the well dispersed CuO
(42.4%), bulk CuO (57.6%),36 SꢀNiO particles (18.2%), wꢀNiOꢀSiO2 particles
(81.8%), the reduction peak of V5+ to V4+ (32.9%) and V4+ to V3+ (67.1%).49
According to the H2ꢀTPR result, the reduction temperature of the SꢀNiO particles
17

New Journal of Chemistry
Page 18 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
is in the order of 15Ni/SiO2 > 30Ni/SiO2 > 5Cuꢀ20V/30Ni/SiO2. The reduction
temperature of wꢀNiOꢀSiO2 shows the similar trend. While the percentage of SꢀNiO
particles is in the order of 30Ni/SiO2 > 15Ni/SiO2 > 5Cuꢀ20V/30Ni/SiO2. The
percentage of wꢀ NiOꢀSiO2 particles shows the opposite order.
300
375.0
(a)
250
200
300.0
150
CuO
100
50
0
NiO
100
200
200
200
300
400
396.1
500
500
500
600
700
20
15
10
5
(b)
461.1
15Ni/SiO2
590.2
600
0
100
300
400
381.6
700
(c)
50
40
30
20
10
0
30Ni/SiO2
456.7
300.0
300
100
400
600
700
(d)
407.7
30
25
20
15
10
5
NiVCu/SiO2
365.1
486.2
241.8
520.4
273.2
0
-5
100
200
300
400
500
600
700
Temperature (oC)
Fig. 5 H2ꢀTPR profiles for various unreduced catalysts. (a) CuO+NiO, (b)
15Ni/SiO2, (c) 30Ni/SiO2, (d) 5Cuꢀ20V/30Ni/SiO2.
Comparing with 15Ni/SiO2 catalyst, the reduction temperature of the Ni species
in 30Ni/SiO2 catalyst shifted to lower temperature. On the one hand, the Ni particle
size of 30Ni/SiO2 (27.3 nm) was large than that of 15Ni/SiO2 (25.4 nm), which caused
the reduction of Ni species to shift to the lower temperature. As is evaluated by
Pawelec et al.50 larger NiO particles size have a smaller interface with the SiO2
surface, which means that the interaction between SiO2 and the NiO particles is
weaker, and thus makes the nickel particles more easy to be reduced. On the other
hand, the area percentage of the SꢀNiO particles in 30Ni/SiO2 (74.6%) was higher
than that in 15Ni/SiO2 (65.7%), which could also cause the reduction of NiO species
to shift to lower temperature. As is known to all, the reduction temperature of SꢀNiO
18

Page 19 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
particles is lower than that of the wꢀNiOꢀSiO2 particles because of the interactions
with SiO2 are weaker.
Comparing with the pattern of 30Ni/SiO2 catalyst, the reduction temperature of
the Ni species in 5Cuꢀ20V/30Ni/SiO2 catalyst also shifted to lower temperature. It
may be due to the presence of Cu species. which causes the interꢀparticle hydrogen
spillover from Cu particles to Ni particles.51,52 As to the hydrogen spillover, Cu
species are first reduced to Cu0 by H2, while H2 molecules are dissociated on the
surface of Cu0 to form the active hydrogen species; then, the active hydrogen species
can be transferred to the neighboring Ni species in close proximity, due to the contact
between the Cu0 and NiO particles and/or the contact of SiO2 with both the Cu0 and
NiO particles, and this enables the Ni species to be easily reduced to Ni0 at an
obviously lower temperature, relative to the normal reduction of NiO by H2 without
hydrogen spillover.51,52 The presence of Cu0 and Ni0 and their contact each other have
been evidenced by the above HRTEM characterization.
Fig. 6 shows the XPS spectra of various elements for various unreduced catalysts
and the corresponding binding energies (B.E.) are summarized in Table S2. One can
see that all the catalysts exhibit a broad main peak with B.E. of ca. 855.0 eV and a
shoulder peak with B.E. of 860.4 ~ 860.9 eV. The main peak is deconvolved and
fitted into two components situating at B.E. = 854.1 ± 0.2 eV and 855.9 ± 0.2 eV,
which can be attributed to Ni2+ species in bulk NiO35 and that interacting with SiO2
(Ni2+ꢀSiO2),35 respectively. while the shoulder peak corresponds to the satellite of Ni
2p3/2 peak.30 As shown in Table S2, the percentages of Ni2+ꢀSiO2 species is in the
order of 30Ni/SiO2 (33.8%) < 15Ni/SiO2 (35.3%) < 5Cuꢀ20V/30Ni/SiO2 (40.8%). It
indicates that increasing the content of Ni results in a decrease in the percentage of
Ni2+ꢀSiO2 species. Whereas, the addition of V and Cu species is in favor of enhancing
the interaction between Ni2+ and SiO2. As the XPS result explained the change of Ni
species on the catalyst surface, these results show the same order of the percentage of
the isolated Ni and the SꢀNiO particles (H2ꢀTPR). Also the Ni dispersion (calculated
from H2ꢀTPD) would decrease if the percentage of the isolated Ni (reduced from
19

New Journal of Chemistry
Page 20 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
SꢀNiO particles) increase.
15Ni/SiO2
Ni 2p3/2
O 1s
15Ni/SiO2
854.0
855.8
532.0
860.4
529.8
850 852 854 856 858 860 862 864 866
Binding Energy (eV)
526 528 530 532 534 536 538 540 542 544
Binding Energy (eV)
Ni 2p
3/2
30Ni/SiO2
O 1s
30Ni/SiO2
854.0
532.4
855.8
860.9
529.7
850 852 854 856 858 860 862 864 866526 528 530 532 534 536 538 540 542 544
Binding Energy (eV)
5Cu-20V/30Ni/SiO2
Binding Energy (eV)
5Cu-20V/30Ni/SiO2
Ni 2p3/2
O 1s
532.4
854.5
856.1
861.1
529.5
850 852 854 856 858 860 862 864 866526 528 530 532 534 536 538 540 542 544
Binding Energy (eV) Binding Energy (eV)
Cu 2p3/2
V 2p3/2
5Cu-20V/30Ni/SiO2
5Cu-20V/30Ni/SiO2
933.8
516.2
517.3
941.6
523.9
514 516 518 520 522 524 526
Binging Energy (eV)
927 930 933 936 939 942 945
Binging Energy (eV)
Fig. 6 Ni 2p3/2, O 1s, V 2p3/2 and Cu 2p XPS spectras for various unreduced catalysts.
In the O 1s region (526~544 eV), for all the samples, there are two peaks at B.E.
20

Page 21 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
= 529.7 ± 0.2 eV and 532.4 ± 0.4 eV, respectively. The former is ascribed to the lattice
oxygen of NiO,53 while the latter is ascribed to the O 1s line of SiO2.27 In the V 2p3/2
region (513ꢀ527 eV), there are two peaks at B.E. = 516.9 eV and 523.9 eV,
respectively. The first main peak at 516.9 eV is fitted to two overlapping curves with
peak positions at 516.2 eV (corresponding to V4+ ion,28 86.5%) and 517.3 eV
(corresponding to V5+ ion,29 13.5%). The second main peak at 523.9 eV is
corresponds to the satellite peak of V5+ ion.29 It shows that the V species is mainly in
form of V4+ ion. In the Cu 2p3/2 region (925ꢀ947 eV), there are two peaks at B.E. =
933.8 eV and 941.6 eV, which are respectively attributed to the Cu2+ species54), and a
satellite peak of Cu2+ species.54
3.2 Structure-activity relationship in the hydrogenation of 3-HPA over various
catalysts
As the mechanism of the hydrogenation of 3ꢀHPA has always been thought to
follow the LangmuirꢀHinshelwood mechanism,55 in which the reactants (3ꢀHPA and
H2) are absorbed and then activated to form the target product (1,3ꢀPDO). The ability
to adsorb and active the reactants is the most significant property of the catalyst. In
particular, the activation of H2 is an important process in the hydrogenation of
3ꢀHPA.56 The conversion of 3ꢀHPA (X3ꢀHPA), selectivity for 1,3ꢀPDO (S1,3ꢀPDO), yield
of 1,3ꢀPDO (Y1,3ꢀPDO) and TOF over various catalysts used in the hydrogenation of
3ꢀHPA are listed in Table 3.
Table 3 Results for the hydrogenation of 3ꢀHPAa
X3ꢀHPA
(mol%)
73.26
S1,3ꢀPDO
(mol%)
62.60
Y1,3ꢀPDO
(mol%)
45.86
TOF
(×103 Sꢀ1)
3.47
Catalysts
15Ni/SiO2
30Ni/SiO2
62.83
93.50
58.74
2.76
5Cuꢀ20V/30Ni/SiO2
94.87
80.97
76.81
4.32
a X3ꢀHPA: conversion of 3ꢀHPA; S1,3ꢀPDO: selectivity for 1,3ꢀPDO; Y1,3ꢀPDO: yield
of 1,3ꢀPDO; Reaction conditions: T = 80 oC, P = 2.0 MPa, LHSV = 0.4 hꢀ1 TOS = 2 h
,
For the 30Ni/SiO2 catalyst, the X3ꢀHPA was 62.83% and the S1,3ꢀPDO was 93.50%.
For the 15Ni/SiO2, the X3ꢀHPA was 10.43% higher than that of the 30Ni/SiO2, while
21

New Journal of Chemistry
Page 22 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
S1,3ꢀPDO was 30.90% lower than that of the 30Ni/SiO2. Several possibilities, either
acting together or separately, may account for the difference in the catalytic
performance of the two catalysts. Firstly, the conversion of 3ꢀHPA over the catalysis
of 15Ni/SiO2 was higher than that of 30Ni/SiO2. It may be because that the dispersion
of Ni particles in the 15Ni/SiO2 (12.6%) was higher than that of the 30Ni/SiO2 (7.9%);
the particle size of 15Ni/SiO2 (7.7 nm) was smaller than that of the 30Ni/SiO2 (12.3
nm); the special surface area of 15Ni/SiO2 (98.6 m2 gꢀ1) was higher than that of the
30Ni/SiO2 (93.2 m2 gꢀ1). In this case, 3ꢀHPA could be easier to be absorbed and then
activated on the Ni site in the 15Ni/SiO2 during hydrogenation. In addition, this result
is in agreement with the XPS and H2ꢀTPR results, in which the percentage of isolated
Ni (reduced from the SꢀNiO particles) was higher in 30Ni/SiO2 than in 15Ni/SiO2.
These changes in the Ni species in these two catalysts are related to the Ni dispersion,
whereby a higher percentage of SꢀNiO particles means a lower Ni dispersion.57
Secondly, the selectivity for 1,3ꢀPDO for the 30Ni/SiO2 catalyst was much higher
than that for the 15Ni/SiO2 catalyst. This may be because that the 3ꢀHPA would be
easier to be absorbed and activated in the 15Ni/SiO2 catalyst, while H2 would be
activated more hardly due to the higher reduction temperature of the Ni species in it.
Therefore, there would be the surplus of the activated 3ꢀHPA but the deficiency of the
activated H2 over the 15Ni/SiO2 catalyst, and this increases the extent for the side
reactions and thus decreases the selectivity to 1,3ꢀPDO, relative to the 30Ni/SiO2
catalyst.55 Another important reason for the improvement in the selectivity for
1,3ꢀPDO under the catalysis of 30Ni/SiO2 might be the increase in the average pore
size (16.8 nm). As the removal of the 1,3ꢀPDO would be easier in larger pore.
For the 5Cuꢀ20V/30Ni/SiO2 catalyst, the X3ꢀHPA was the highest among the three
catalyst. The S1,3ꢀPDO was lower than that of 30Ni/SiO2, but it is still much higher than
that of the 15Ni/SiO2. On the one hand, the conversion of 3ꢀHPA under the catalysis
of 5Cuꢀ20V/30Ni/SiO2 was greatly increased due to the lower reduction temperature
and greater number of active sites induced by Cu and V. The reduction peak of Ni
species shifted to lower temperature because of hydrogen spillover from Cu particles
22

Page 23 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
to Ni particles (as shown in H2ꢀTPR). In addition, V species show the ability to active
C=O bond,21ꢀ23 thus there would be more 3ꢀHPA be activated in the
5Cuꢀ20V/30Ni/SiO2 catalyst. On the other hand, the selectivity for 1,3ꢀPDO under the
catalysis of 5Cuꢀ20V/30Ni/SiO2 was higher than that for the 15Ni/SiO2. It might be
explained by that the activation of H2 were more easily because of the lower reduction
temperature of Ni species in the 5Cuꢀ20V/30Ni/SiO2, which means that the activated
3ꢀHPA could react with the activated H2 rather than with acrolein. While the
selectivity to 1,3ꢀPDO under the catalysis of 5Cuꢀ20V/30Ni/SiO2 is lower than that of
the 30Ni/SiO2. As the activation of the H2 is harder than the activation of 3ꢀHPA, the
activated 3ꢀHPA may suffer from side reactions due to relatively lower concentration
of activated H2. As the activation of the 3ꢀHPA under the catalysis of
5Cuꢀ20V/30Ni/SiO2 is much higher than that of the 30Ni/SiO2. The surplus of the
activated 3ꢀHPA would lead to more sideꢀreactions.
Scheme 1 Possible mechanism for the hydrogenation of 3ꢀHPA to 1,3ꢀPDO over
the CuꢀV/Ni/SiO2 catalyst.
According to the results of the characterization and reaction, we proposed a
possible mechanism for the hydrogenation of 3ꢀHPA as shown in Scheme 1. Activated
3ꢀHPA would form 1,3ꢀPDO by reacting with the activated H2. Whereas, activated
3ꢀHPA would be attacked by acrolein to form 4ꢀoxaꢀ1,7ꢀheptanedial in the absence of
23

New Journal of Chemistry
Page 24 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
activated H2, as acrolein could also be activated and usually be more active. The
ability to activate H2 greatly effects the performance of the catalyst, which could be
greatly enhanced by adding Cu and V. Meanwhile, the textural properties, particle size
and Ni distribution would have a great effect on the hydrogenation via limiting the
mass transfer of 3ꢀHPA and 1,3ꢀPDO.
3.3 Optimization of the reaction conditions
Optimization of the reaction conditions was conducted to further improve the
yields of 1,3ꢀPDO for potential applications in industry. Fig. 7 illustrates the effect of
the reaction conditions on the hydrogenation of 3ꢀHPA over the 5Cuꢀ20V/30Ni/SiO2
catalyst, including the reaction temperature, reaction pressure, liquid hourly space
velocity (LHSV) and time on stream (TOS).
Fig. 7a shows the effect of the reaction temperature on the hydrogenation of
3ꢀHPA over the 5Cuꢀ20V/30Ni/SiO2 catalyst. With an increase in the reaction
temperature, the conversion of 3ꢀHPA increased approximately linearly. This is
because higher temperature would increase the mass transfer rate of the reactant and
favor the activation of the reactants.56 The selectivity for 1,3ꢀPDO increased initially
with an increase in the reaction temperature. When the temperature was higher than
70 °C, a further increase in the reaction temperature lead to a decrease in selectivity
for 1,3ꢀPDO. This result can be explained by the fact that the hydrogenation reaction
rate and hydrogen transfer rate could both be accelerated by increasing the reaction
temperature. However, the hydrogenation reaction of 3ꢀHPA is suppressed due to its
exothermic nature at high reaction temperature. At the same time, high temperature
could promote side reaction, which would lead to a decrease of selectivity for
1,3ꢀPDO.56,58 Considering the high conversion of 3ꢀHPA and high selectivity for
1,3ꢀPDO over the 5Cuꢀ20V/30Ni/SiO2 catalyst, the optimal reaction temperature is
80 °C.
Another important influencing factor that affects the reaction is H2 pressure. Fig.
7b shows the effect of the reaction pressure on the hydrogenation of 3ꢀHPA over the
5Cuꢀ20V/30Ni/SiO2 catalyst. One can see that the conversion of 3ꢀHPA, the
24

Page 25 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
selectivity for 1,3ꢀPDO and the yield of 1,3ꢀPDO all increased monotonously with an
increase in the reaction pressure from 1.0 MPa to 5.0 MPa. This indicates that the H2
pressure exhibits a positive effect on the conversion of 3ꢀHPA and the selectivity for
1,3ꢀPDO.59 This is due to the fact that, as is well known, the gasꢀliquidꢀsolid mass
transfer resistance of hydrogen has constituted a very important limiting factor for
many hydrogenation reaction, and therefore, the increase in the H2 pressure promotes
the accessibility of H2 at the active sites of catalyst, favoring the hydrogenation
conversion of 3ꢀHPA to 1,3ꢀPDO.60 It should be particularly addressed that a
relatively high conversion of 3ꢀHPA (ca. 85.0%) at high selectivity to 1,3ꢀPDO (ca.
72.0%) has been achieved at a appreciably low H2 pressures (2.0 MPa) over the
5Cuꢀ20V/30Ni/SiO2 catalyst. The relatively low H2 pressure employed in this work is
of significant importance for the industrial application of the hydrogenation of 3ꢀHPA
to 1,3ꢀPDO, since a lower reaction pressure implies both the lower requirements for
facilities and energy consumption.
The effect of LHSV on the hydrogenation of 3ꢀHPA over the
5Cuꢀ20V/30Ni/SiO2 catalyst is shown in Fig. 7c. One can see that the conversion of
3ꢀHPA decreased slightly with an increase in the LHSV from 0.10 hꢀ1 to 0.95 hꢀ1.
Similar results were shown for the selectivity for 1,3ꢀPDO. This indicates that a lower
LHSV is favorable for the conversion of 3ꢀHPA and the selectivity for 1,3ꢀPDO. At a
high LHSV, which corresponding to a short residence time, the reactants are not
sufficiently in contact with the catalyst. This result in a decrease in the conversion of
3ꢀHPA and the selectivity for 1,3ꢀPDO.59 Considering the production capacity,
requirements to facilities and energy consumption during the production of 1,3ꢀPDO,
the optimized condition for the hydrogenation of 3ꢀHPA are a reaction temperature is
80 oC, a reaction pressure of H2 of 2.0 MPa and LHSV of 0.4 hꢀ1.
Fig. 7d shows the effect of the time on stream (TOS) on the yield of 1,3ꢀPDO
over the 5Cuꢀ20V/30Ni/SiO2 catalyst. One can see that the yield of 1,3ꢀPDO was
above 70% during 50 h, and the yield of 1,3ꢀPDO maintained about 60% after 75 h. In
addition, the composition of the reduced and used 5Cuꢀ20V/30Ni/SiO2 catalysts was
25

New Journal of Chemistry
Page 26 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
determined by ICP as shown in Table 4. The ICP results show that the content of Cu,
V and Ni decreased by 4.6, 10.3 and 31.8 mg gꢀ1 after 75 h of reaction in the used
5Cuꢀ20V/30Ni/SiO2 catalyst. The loss of these metal elements may be because that
isolated Ni particles (reduced from SꢀNiO particles) would dissolve in the liquid
phase. According to the characterization of the XRD, BET, and ICP results, the
reasons for the deactivation of 5Cuꢀ20V/30Ni/SiO2 catalyst seem to be: 1) impurities
exist in the raw materials or organic oligomer are adsorped on the catalyst surface; 2)
the mesopores are blocked by coke (organic oligomers) or damaged during the
hydrogenation reaction;59 and 3) the loss of the active sites (Cu and Ni). Whereas, the
addition of Cu and V could protect the Ni site from oxidation and increase the
interaction between Ni and SiO2. It make the 5Cuꢀ20V/30Ni/SiO2 a promising
catalyst for the industrial application.
95
90
85
80
75
70
65
60
55
100
90
80
70
60
50
(a)
(b)
Conversion
Selectivity
Yield
Conversion
Selectivity
Yield
50
60
70
80
90
100
1
2
3
4
5
Temperature (oC)
Pressure (MPa)
100
90
80
70
60
50
40
30
100
80
60
40
20
0
(d)
(c)
Conversion
Selectivity
Yield
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
LHSV (h-1)
0
10 20 30 40 50 60 70 80
Time on stream (h)
Fig. 7 Effect of reaction conditions on the hydrogenation of 3ꢀHPA over the
5Cuꢀ20V/30Ni/SiO2 catalyst.
(a) Reaction conditions: P = 2.0 MPa, LHSV = 0.51 hꢀ1；
(b) reaction conditions: T = 80 oC, LHSV = 0.51 hꢀ1;
(c) reaction conditions: T = 80 oC, P = 2.0 MPa;
(d) reaction condition: T = 80 oC, P = 2.0 MPa, LHSV = 0.4 hꢀ1.
26

Page 27 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
Table 4 Composition of the reduced and used 5Cuꢀ20V/30Ni/SiO2
Component
Cu (mg gꢀ1)
23.7
V (mg gꢀ1)
94.8
Ni (mg gꢀ1)
Reduced 5Cuꢀ20V/30Ni/SiO2
Used 5Cuꢀ20V/30Ni/SiO2
142.2
110.4
19.1
84.5
Reaction conditions: T = 80 oC, P = 2.0 MPa, LHSV = 0.4 hꢀ1 TOS = 75 h.
,
Table 5 Comparison between the results in this work and in the literature
Temperature Pressure Conversion Selectivity Yield
Catalyst
Reference
(oC)
(MPa)
15.0
15.0
15.0
6.5
(%)
100
100
100
100
(%)
78.0
81.0
74.5
85.0
(%)
78.0
81.0
74.5
85.0
Ni/Al2O3/SiO2
Ni/Al2O3/SiO2
55
[61]
[14]
[11]
[62]
This
work
55
Ni/Al2O3/SiO2
50
50%Ni/Al2O3/SiO2
70
5Cuꢀ20V/30Ni/SiO2
80
2.0
94.9
81.0
76.9
Table 5 gives a comparison between the yield of 1,3ꢀPDO in this work and those
in the literature11,61,62 under different conditions. One can see that the yield of
1,3ꢀPDO in this work is relatively high among the various synthetic routes. Given the
extremely low pressure (2.0 MPa) and the high yield of 1,3ꢀPDO, the catalyst
prepared in this work could have great potential for industrial application.
4. Conclusion
Cu and V modified Ni/SiO2 (5Cuꢀ20V/30Ni/SiO2) catalyst is a promising
catalyst for the hydrogenation of 3ꢀhydroxypropanal to 1,3ꢀpropanediol. A high
3ꢀHPA conversion (94.9%) and relatively high selectivity to 1,3ꢀPDO (81.0%) can be
achieved even at low reaction pressure (2.0 MPa) on this catalyst, which is
advantageous in industrial application. The structureꢀactivity relationship is clarified
according to kinds of characterization results. (i) The dispersion of Ni active sites has
great effect on the absorption and activation of the reactants (3ꢀHPA and H2); (ii) the
addition of the Cu and V could lower reduction temperature of the Ni species in
catalyst, which would increase the ability to activate the reactants; (iii) the percentage
of the wꢀNiOꢀSiO2 particles in catalyst has great influence on the stability of catalyst.
27

New Journal of Chemistry
Page 28 of 33
View Article Online
DOI: 10.1039/C7NJ01746D
Therefore, the low reaction pressure, the stable structure and the high catalytic
performance of the 5Cuꢀ20V/30Ni/SiO2 catalyst enable it to have potential application
in industry.
28

Page 29 of 33
New Journal of Chemistry
View Article Online
DOI: 10.1039/C7NJ01746D
Acknowledgments
This work was supported by the National Natural Science Foundation of China
(Grant. No. 21376068), Program for New Century Excellent Talents in University, the
Ministry of Education of P. R. China, and the Program for FuꢀRong Scholar in Hunan
Province, P. R. China.
29