Silica/titania composite-supported NiCo catalysts with combined catalytic effects for phenol hydrogenation under fast and mild conditions

Article abstract of DOI:10.1016/j.apcata.2020.117409

The effect of titanium as a promoter for the nickel-cobalt/silica-titania (NiCo/Si-Ti) catalyst in the hydrogenation of phenol was comparatively studied. In the phenol hydrogenation under mild reaction conditions (1 MPa H₂ pressure and 100 °C), the Ti-promoter induces the formation of the Co^δ??O_V?Ti³⁺ active sites and enhances the hydrogenation activity of the NiCo/Si catalyst to convert phenol to cyclohexanol. The catalytic hydrogenation with the optimal composition of the NiCo/Si-1.0Ti (Si:Ti ratio of 8.5:1) catalyst was completed within 1 h (conversion 98.2 % and selectivity >99 %), which was more efficiently than that with other Ti-doped catalysts and NiCo/Si as baseline, indicating a strong positive synergistic effect between the metal (Ni-Co) and support (Ti-Si) components of the catalyst in this reaction. Additionally, the kinetics, pathway and catalytic mechanism of the hydrogenation of phenol are also examined. Furthermore, the NiCo/Si-1.0Ti catalysts also showed a good performance at reducing the unstable oxygenated compounds in the bio-oil.

Full text of DOI:10.1016/j.apcata.2020.117409

Journal Pre-proof
Silica/titania composite-supported NiCo catalysts with combined catalytic
effects for phenol hydrogenation under fast and mild conditions
Yucheng Li (Methodology) (Investigation), Jing Liu (Formal analysis)
(Writing - original draft), Jing He (Validation), Luying Wang (Writing -
review and editing), Jiandu Lei (Supervision)
PII:
S0926-860X(20)30002-8
DOI:
https://doi.org/10.1016/j.apcata.2020.117409
APCATA 117409
Reference:
To appear in:
Applied Catalysis A, General
Received Date:
Revised Date:
Accepted Date:
19 August 2019
31 December 2019
2 January 2020
Please cite this article as: Li Y, Liu J, He J, Wang L, Lei J, Silica/titania composite-supported
NiCo catalysts with combined catalytic effects for phenol hydrogenation under fast and mild
conditions, Applied Catalysis A, General (2020),
doi: https://doi.org/10.1016/j.apcata.2020.117409
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. 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.
©
2019 Published by Elsevier.

Silica/titania composite-supported NiCo catalysts with combined catalytic effects
for phenol hydrogenation under fast and mild conditions
*
*
Yucheng Li , Jing Liu liujing@bjfu.edu.cn, Jing He , Luying Wang , Jiandu Lei
ljd2012@bjfu.edu.cn
Beijing Key Laboratory of Lignocellulosic Chemistry, Beijing Forestry University,
Beijing 100083, P. R. China
*
Corresponding author. Tel: +86 10 6233 7251
Graphical abstract
Highlights
Hydrogenation phenol to cyclohexanol using NiCo/Si-Ti catalyst was completed

within 1 h under mild reaction conditions (1 MPa H
2
and 100°C).
1





NiCo/Si-1.0Ti (Si:Ti ratio of 8.5:1) catalyst was more efficiently (conversion
8.2% and selectivity >99%) than NiCo/Si and other Ti-doped catalysts.
9
δ−
3+
Ti-promoter induces the formation of the Co −O −Ti active sites and enhances
V
the hydrogenation activity.
The NiCo/Si-1.0Ti catalysts showed a good performance at reducing the unstable
oxygenated compounds in the real bio-oil.
The kinetics, pathway and catalytic mechanism of the hydrogenation of phenol
are examined.
Abstract
The effect of titanium as a promoter for the nickel-cobalt/silica-titania (NiCo/Si-Ti)
catalyst in the hydrogenation of phenol was comparatively studied. In the phenol
hydrogenation under mild reaction conditions (1 MPa H pressure and 100°C), the
2
δ−
3+
Ti-promoter induces the formation of the Co −O −Ti active sites and enhances
V
the hydrogenation activity of the NiCo/Si catalyst to convert phenol to cyclohexanol.
The catalytic hydrogenation with the optimal composition of the NiCo/Si-1.0Ti (Si:Ti
ratio of 8.5:1) catalyst was completed within 1 h (conversion 98.2% and
selectivity >99%), which was more efficiently than that with other Ti-doped catalysts
and NiCo/Si as baseline, indicating a strong positive synergistic effect between the
metal (Ni-Co) and support (Ti-Si) components of the catalyst in this reaction.
Additionally, the kinetics, pathway and catalytic mechanism of the hydrogenation of
phenol are also examined. Furthermore, the NiCo/Si-1.0Ti catalysts also showed a
good performance at reducing the unstable oxygenated compounds in the bio-oil.
2

Keywords: Silica/titania composite; NiCo catalyst; Phenol hydrogenation;
Cyclohexanol; Mild condition
1
. Introduction
Recognized as an environmentally-friendly alternative resource, biomass has been
increasingly drawing worldwide attention. As the most abundant component of
biomass [1], many efforts have been made to efficiently use lignocellulosic materials
by fast pyrolysis, gasification, combustion and liquefaction. Among these processes,
fast pyrolysis is considered as one of the most promising approaches for the efficient
conversion of biomass to bio-oil with high yields (up to 60 wt%) through thermal
decomposition and depolymerization of macromolecules [2]. Bio-oil can be produced
by pyrolysis, liquefaction, transesterification and other processes. It is mainly
composed of oxygenated aromatic compounds, such as phenols, furans, ketones,
aldehydes, alcohol, ether, acids and esters [3], with a content of phenolic aromatic
compounds over 55% [4]. Phenol, the most representative of the phenolic compounds,
as a common byproduct of the petrochemical industry, can cause considerable damage
to the environment [5].
The hydrogenation of phenol is an important process in the chemical synthesis of raw
materials (e.g. cyclohexanone and cyclohexanol) for the production of various
products, such as nylon, fuel, pharmaceuticals, plasticizers, surfactants, solvents, etc.,
which play a critical role in economic and societal development [6]. In fact, besides
its much lower toxicity than phenol and cyclohexanone [7], cyclohexanol is an
3

important intermediate for the synthesis of Nylon-6 and the component of
ketone-alcohol (KA) oil in petroleum industry [8, 9], Additionally, cyclohexanol is
also an effective oxygenate additive for the enhancement of the multi-component
diesel fuel used to improve emissions [10]. Thus, the selective hydrogenation of
phenol appears to be an effective alternative route to produce cyclohexanol, where a
highly-efficient catalyst is crucial for industrial applications [11].
According to the literature, commonly used catalysts for the hydrogenation of phenol
are transition metals (like Ni, Co, and Ce) [12-14], and noble metals (such as Pt, Pd,
Ru, etc.) [15-18], loaded into oxyphilic supports, such as Al
2
O
3
, SiO
2
, and ZrO
2
[
19-22]. Undoubtedly, noble metal catalysts are expensive and sulfurated catalysts [3]
cause environmental problems, which limit their application on an industrial scale.
Due to their low cost and abundance, the Ni-based bimetallic and trimetallic alloy
catalysts have been synthesized and used in the hydrogenation of phenol [23, 24]
which is usually chosen as a model bio-oils compound [25, 26]. Aiqin Li et al. [11]
prepared a bimetallic metal-organic framework (MOF)-derived Co-Ni@NC catalyst
for the selective hydrogenation of phenol and the sole product was cyclohexanol.
They found that the catalyst with the optimal composition (1Co-1Ni@NC-600) was
able to completely catalyze the hydrogenation of phenol to >99.9% cyclohexanol with
an activity up to 2.8 and 4.3-fold higher than that of the catalysts Co@NC and
Ni@NC, respectively. However, the complicated MOF-templated synthesis (more
than 72 h) and high calcination temperature (reached 600°C) of the catalyst may limit
its application in industry. Recent studies have reported that Ni- and Co-based
4

catalysts (Ni
x
Co
y
@C/ZrO ) supported by a Zr-based MOF have the best catalytic
2
performance at the desired temperature of 200°C for 240 min, with a 96.33%
conversion of phenol and a cyclohexanol selectivity of 91.14% [27]. Yuzhen Shi et al.
[
28] obtained a total mole yield of 80% and a selectivity of 80% using 15 wt% of the
Ni–Co/γ-Al catalyst. Therefore, high activity (>99%) non-noble metal catalysts at
2
O
3
faster reaction rate (<60 min) and under milder reaction condition (<100°C) are
coveted in phenol hydrogenation.
In this study, Ti-Si-supported Ni and Co metal catalysts were synthesized and used for
phenol hydrogenation to prepare cyclohexanol. Titanium oxide was used to enhance
the active sites on the catalyst. Besides the preparation of the catalysts, we focused on
its characterization and evaluation of its performance in the conversion, selectivity
and kinetics during phenol hydrogenation under mild conditions. In addition, the
pathway of phenol hydrogenation and metal−support interaction were also examined.
Moreover, the selected catalyst (NiCo/Si-1.0Ti) was experimentally used for the
hydrogenation of real bio-oil.
2
2
. Experimental and methods
.1 Material
Ammonia (14%), Cetyltrimethylammonium, Tetrabutyl titanate (99%; Macklin
Biochemical Co., Ltd., Shanghai, China), Ethyl orthosilicate (99%; Macklin
Biochemical Co., Ltd.), Ni(NO
3
)
2
, Co(NO
3
)
2
·6H O, Hexane (99%; Alfa-Aesar
2
Chemicals, Tewksbury, MA, USA), n-Heptane (99%; Aladdin Biochemical
5

Technology Co., Ltd.), n-Dodecane (99%; Macklin Biochemical Co., Ltd.) and
Phenol (AR, Xilong Scientific Co., Ltd.) are all commercially available and used
without further purification. Bio-oil as feedstock in this work was produced by
pyrolyzing walnut shell at 530 °C, which was provided by University of Science and
Technology of China. The pyrolysis-condensation device and experimental details
were described by Chu Wang et al [29].
2
.2 Synthesis of catalysts
Templated synthesis. A solution of ammonia water (3.25 mL),
cetyltrimethylammonium (CTACL, 16.58 g) and water (243 mL) was prepared by
mixing and stirring at room temperature for 12 h. Subsequently, ethyl orthosilicate
(TEOS) and butyl titanate (TBOT) were added dropwise to the above solution with
agitation. The amount of Si and Ti precursor was varied in order to prepare four
catalysts with different Si:Ti ratios of 8.5:1.0, 6.0:3.5, 3.5:6.0 and 1.0:8.5. The
resulting samples were denoted as Si-1.0Ti, Si-3.5Ti, Si-6.0Ti and Si-8.5Ti,
respectively. Then the obtained mixture was vacuum dried at 90°C for 3 h. After the
gelatin formed, it was ground into a powder to obtain the titanium-silicon carrier.
Metal impregnation. Ni(NO
3
)
2
(1 g), Co(NO
3
)
2
·6H O (1.5 g) and titanium-silicon
2
carrier (1.5 g) prepared above were dispersed in 20 mL normal hexane reflux for 12 h
at 70°C. After cooling, the resulting slurry was centrifuged in 3000 rpm for 5 min.
The precipitate was dried in air at room temperature for 12 h, then dried at 60°C for 3
h under vacuum. Subsequently, the solid products were calcined at 450°C for 4.5 h
-1
with heating reat of 10°C min . After cooling, the solid products were treated with a
6

sodium hydroxide solution (15% wt%) for 5 h at room temperature, to obtain the
NiCo/Si-1.0Ti, NiCo/Si-3.5Ti, NiCo/Si-6.0Ti and NiCo/Si-8.5Ti catalysts. NiCo/Si
catalyst was also prepared by using above method with no Ti. Ultimately, the catalysts
were reduced under a flow of 50% v/v N
2
/H (50mL/min) at 300°C for 3 h. The
2
typical synthetic method for the preparation of the NiCo/Si-Ti catalyst is
schematically illustrated in Fig. 1.
Fig. 1 Schematic of synthesis of NiCo/Si-Ti catalysts and their hydrogenation
performance of phenol.
2
2
.3 Catalyst evaluation
.3.1 Phenol hydrogenation
About 0.05 g of phenol and 0.1 mL n-tetradecane (as the internal standard substance)
were dissolved in 10 mL of n-heptane and placed in a 50-mL stainless steel cylinder
reactor, followed immediately by the addition of 0.1 g of fresh catalyst into the reactor.
The reactor was charged with N
air from the chamber, and then filled up three times with H
displace the N . After repeated charging and venting, the H
2
gas (99.999%) three times at 0.2 MPa to displace the
2
gas at 0.2 MPa to
2
2
pressure was raised to 1
MPa and the reactor was placed on a heater with matched temperature controller to
7

maintain the reactor at 100°C with stirring at a speed of 300 rpm. After heating for 2 h,
the reactor was removed from the heater and allowed to cool to room temperature
with a fan. The products of the phenol hydrogenation reaction were analyzed by gas
chromatography (GC) using the Agilent 7890B GC system (Agilent Technologies,
Santa Clara, CA, USA), equipped with a flame ionization detector (FID), using a 330
mm × 0.32 mm × 0.25 μm Agilent HP-5 capillary column. The GC system was
operated at 40°C for 4 minutes and increased to 260°C at a rate of 15°C/min. The
results were calculated using the follow equations [10]:
mol of Phenol after reaction
Phenol conversion = (1-
)100%
mol of Phenol intial
mol of Cyclohexanol
Cyclohexanol selectivity =
100%
mol of the amounts of products
2
.3.2 Hydrogenation of real biofuel.
Hydrogenation of bio-oil was carried out in a 50-mL stainless steel cylinder reactor.
The prepared catalyst (0.5 g) and bio-oil (10 mL) were added into the reactor. The
reactor was then subjected to the same in-reactor gas treatment described above.
Afterwards, the reactor was heated to the reaction temperature of 250°C with stirring
at 300 rpm. Eventually, the reactor was quickly cooled to room temperature with a fan.
The liquid reaction products were separated from the catalysts by centrifugation and
analyzed by Gas Chromatography-Mass Spectrometry (GC-MS) using the Agilent
7
890A/5975C GC-MS system (Agilent Technologies), equipped with a BPX-70
capillary column. The GC-MS system was operated as described for the GC system
above. Functional groups of raw bio-oil and hydrogenated bio-oil were characterized
by nuclear magnetic resonance spectroscopy (NMR) using a Bruker AVIII 400 MHz
8

NMR (Bruker, Germany) as described in the literature [30].
.4 Catalysts characterization
2
The phase composition and crystal structure of the catalysts were investigated by
X-ray diffraction (XRD) analysis, performed on a Rigaku D/MAX-IIIA X-ray
diffractometer (Rigaku Corp., Tokyo, Japan) operated at 3 kW and using Cu K
radiation. The scanning speed was at 2°/min in the scanning region of 10-90°. N
adsorption-desorption isotherms of the catalysts were recorded at 77K on a
2
Micromeritics TriStar 3000 apparatus (Micromeritics Instrument Corp., Norcross, GA,
USA). H temperature-programmed desorption (H -TPD) analyses of the final catalyst
were performed using a TP-5080 Multi-functional Automatic Adsorption Instrument
Tianjin Xianquan Company, Tianjin, China). First, 100 mg of the catalyst samples
were pretreated for 20 min at 400°C, under a N atmosphere. Then, the adsorption of
was measured under a mixed flow of H (30 mL/min) and N (30 mL/min) at
00°C for 1 h and then slowly decreased from 400 to 20°C. Next, the H weakly
absorbed on the sample was removed by flushing with a N stream (30 mL/min, 20°C,
5 min) and the strongly absorbed H was detected by H -TPD method. The H
desorption curves were obtained under an N stream of 30 mL/min at a temperature
increasing linearly from 20, to 1,000°C at 10°C/min. The desorbed H was monitored
2
2
(
2
H
2
2
2
4
2
2
1
2
2
2
2
2
on-line using a GC system equipped with a thermal conductivity detector (TCD). The
specific surface areas were determined by Brunauer-Emmett-Teller (BET) method and
the pore size distribution and pore volume were calculated by the
Barrett-Joyner-Halenda (BJH) method. The microscopic features such as the surface
9

topography and element distribution of the catalyst were examined by transmission
electron microscope (TEM) on a JEM-1010 transmission electron microscope (JEOL
Ltd., Tokyo, Japan), scanning electron microscope (SEM) on a SU8010 scanning
electron microscope (JEOL Ltd.) equipped with an energy-dispersive X-ray
spectrometer (EDS) accessory, and high-resolution transmission electron microscopy
(HRTEM) on a JEM-2100F transmission electron microscopy (JEOL Ltd.) equipped
with an EDS. The content and chemical states of the elemental constituents were
analyzed by X-ray photoelectron spectroscopy (XPS) on an EscaLab 250Xi X-ray
photoelectron spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). FT-IR
-1
spectra in the region 500-4000 cm were recorded using a Bruker VERTEX 70 FT-IR
spectrophotometer (Bruker Optik GmbH, Ettlingen, Germany) with KBr pellets.
3
3
. Results and discussion
.1 Characterizations of the catalysts.
The SEM images showing the surface morphology of the Si-1.0Ti, NiCo/Si-1.0Ti and
NiCo/Si-8.5Ti catalyst are displayed in Fig. 2(a)-(c), respectively, while Fig. 2(d)-(f)
display the TEM images of the NiCo/Si, NiCo/Si-1.0Ti and NiCo/Si-8.5Ti catalysts.
HRTEM images of NiCo/Si-1.0Ti and NiCo/Si-8.5Ti catalysts were showed in Fig. S1,
and some micropore below 2 nm appeared in NiCo/Si-1.0Ti catalyst. These SEM and
TEM images revealed the characteristic porous surface of all the catalysts examined.
They show that the porosity decreased when Si:Ti ratio from 8.5:1 change to 1:8.5
(proved by BET results below). The highly uniform porous structure is great
10

beneficial to improve the catalytic performance. Additionally, the HRTEM-EDS (Fig.
g) images show that all elements are evenly distributed with no formation of
2
aggregates. The estimated Ti, Ni and Co content was 4.12, 13.68 and 22.12 wt%,
respectively.
Fig. 2 SEM images of Si-1.0Ti (a), NiCo/Si-1.0Ti (b) and NiCo/Si-8.5Ti(c); TEM
images of NiCo/Si (d), NiCo/Si-1.0Ti (e) and NiCo/Si-8.5Ti (f); HRTEM-EDS
elemental mappings for NiCo/Si-1.0Ti (g).
11

The N
catalysts are displayed in Fig. 3, and the details of their physical properties are shown
in Table 1. All the samples displayed type IV N adsorption-desorption isotherms, as
2
adsorption-desorption isotherms and pore size distribution curves of the
2
can be seen in Fig. 3a, and showed an H4 hysteresis loop (according to the IUPAC
classification), indicating its dominating mesoporous structure [31]. Also, the
hysteresis loop becomes narrower with the addition of Ti, which indicates the
decrease of the mesoporosity [32], leading to decrease of BJH pore volume (see Table
1
). Additionally, as shown in Fig. 3b, the porosity decreased with the increase in the
Ti content for Ti-doped catalysts. The data presented in Table 1 revealed that
2
NiCo/Si-1.0Ti catalyst had the highest specific surface area (56.64 m /g), BJH pore
3
3
volume (0.12 cm /g) and SF micropore volume (0.02 cm /g) among all catalysts.
However, the specific surface area and pore volume of the Ti-doped catalyst slightly
decrease with increasing Ti content, which can be attributed to the blockage of the
small pores of NiCo/Si-1.0Ti. Compared to the baseline NiCo/Si, there is an increase
in surface area for all of the Ti-doped catalysts and an increase in pore diameter for
higher Ti-doped catalysts (3.5Ti, 6.0Ti and 8.5Ti). Such an increase is probablly
because Ti-precursor decomposed leading to pore formation.
12

Fig. 3 The N adsorption-desorption isotherms (a) and pore size distribution (b) of the
2
NiCo/Si, NiCo/Si-1.0Ti, NiCo/Si-3.5Ti, NiCo/Si-6.0Ti and NiCo/Si-8.5Ti catalysts.
Table 1 Surface area, pore size and pore volume of catalysts.
BET surface
BJH pore
BJH pore
SF micropore
Catalyst
NiCo/Si
2
3
3
area (m /g)
diameter (nm) volume (cm /g) volume (cm /g)
29.83
11.65
11.00
11.98
12.81
14.47
0.12
0.12
0.10
0.09
0.08
0.01
0.02
0.02
0.01
0.01
NiCo/Si-1.0Ti 56.64
NiCo/Si-3.5Ti 44.15
NiCo/Si-6.0Ti 33.33
NiCo/Si-8.5Ti 32.06
The XRD patterns of the NiCo/Si, NiCo/Si-1.0Ti, NiCo/Si-3.5Ti, NiCo/Si-6.0Ti and
NiCo/Si-8.5Ti catalysts are shown in Fig. 4(a). The XRD pattern of all catalysts
clearly reveal the oxide phases of cobalt (Co
3
O
4
2θ at 19.08, 32.03, and 36.80°) [9,
3
5
1
3], and nickel (NiO 2θ at 37.70, 43.13 and 62.88°) [34, 35]. The weak peak at 2θ =
1.44° assigned to metallic Ni (JCPDS No. 04-0850) [36] or metallic Co (JCPDS No.
5-0806) [37], confirming the presence of slight metallic Ni and/or Co in the all
catalysts. In fact, the crystalline phases of metallic Ni and Co have similar diffraction
patterns, making their identification difficult. The SiO exhibits amorphous XRD
patterns with some small peaks at 2θ = 23-25° (ICSD No. 47219) [38]. No crystalline
peaks corresponding to TiO are observed in the spectra of the all catalysts. We
considered the lower concentration of TiO was so uniformly dispersed on the surface
of catalysts that it is not detectable by XRD analysis. However, for NiCo/Si-1.0Ti, the
2
2
2
13

diffraction peak (2θ=36.80°) of Co
3
4
O shifts toward high angles (2θ=37.21°),
indicating the probable formation of Co-Ti alloy [11, 39]. It is strange that
NiCo/Si-3.5Ti, NiCo/Si-6.0Ti and NiCo/Si-8.5Ti catalysts were hardly observed the
slight variation in this peak position. We consider that Co-Ti alloy appeared in
NiCo/Si-1.0Ti is the most significant. This indicates the occurrence of structural and
physicochemical changes in the hydrogenation catalyst, which may result in
differences in the catalytic performance.
As shown in Fig. 4(b), the H -TPD spectra obtained for NiCo/Si, NiCo/Si-1.0Ti,
2
NiCo/Si-3.5Ti, NiCo/Si-6.0Ti and NiCo/Si-8.5Ti catalysts displayed a broad profile at
low temperature (around 130°C). These spectra show that Ti-doped catalysts
markedly enhance H
intensity of the Ti-doped catalysts are all higher than that of the NiCo/Si catalyst. It
can be also observed that H desorption peak at 136°C for NiCo/Si-1.0Ti catalysts
2
uptake compared with the NiCo/Si catalyst, and the peak
2
slightly shifts to lower temperature (126°C for NiCo/Si-8.5Ti catalysts) with
increasing Ti content, suggesting that more Ti content leads weaker interaction
between active metals and adsorbed H
high-temperature desorption peak (around 530°C) only in the NiCo/Si-1.0Ti,
NiCo/Si-3.5Ti, and NiCo/Si-6.0Ti catalysts, indicating that H strongly chemisorbed
2
. It is important to note that there is a
2
on the surface of these catalysts. But NiCo/Si-8.5Ti catalyst showed no
high-temperature desorption peak below 600°C. All in all, NiCo/Si-1.0Ti catalyst
displayed the most intense H
2
desorption peak in the H -TPD profiles, suggesting that
2
it had the most abundant dissociated active H, and the strongest interaction between
14

activating centers and adsorbed H [31].
2
Fig. 4 XRD patterns (a) and H
2
-TPD profiles (b) of the NiCo/Si, NiCo/Si-1.0Ti,
NiCo/Si-3.5Ti, NiCo/Si-6.0Ti and NiCo/Si-8.5Ti catalysts.
−
1
In the FT-IR spectra of all samples (Fig. 5), the broad peaks at around 3434 cm were
assigned to the stretching vibrations of Si-OH present in the support and the O-H
associated to free water and structural water contained in the sample [40]. A sharp
-1
peak observed at 1633 cm was assigned to the O-H bending vibration of the
−
1
absorbed water molecules [41]. The IR absorption at 984 cm was ascribed to the
vibration of Si-OH [42-44]. As shown in Fig. 5, there is no obvious difference
between the spectra of the fresh NiCo/Si-1.0Ti and used NiCo/Si-1.0Ti, with the
−
1
−1
expected new characteristic peaks at around 2920 cm and 1148-1123 cm ,
attributed to the stretching vibrations of C-H and the C-O-C groups [45], resulting
from the carbonaceous accumulation around the used catalyst. It can also be seen that
-1
the IR spectra of NiCo/Si showed a strong vibration absorption band at 1015 cm ,
which was assigned to the vibration of Si−O−Si bonds [46-48]. However, for
-1
NiCo/Si-1.0Ti, the vibration band was shifted to 1031 cm , indicating the effective
15

coordination interaction between Si and Ni, which was further confirmed by the XPS
analysis, which will be discussed later. It is strange that no Si−O−Si vibration peak
was present in NiCo/Si-8.5Ti, we speculate that it is not detected by FT-IR
−
1
spectroscopy due to its lower Si content. All samples had peaks below 600 cm ,
which were ascribed to the symmetric stretching vibration of metal-oxygen groups
−
1
[
49], and the band at about 562 cm was assigned to Ni-O [50]. The characteristic
−
1
peak at around 462ꢀcm was ascribed to the stretching vibration absorption of Si-O,
Ti-O bonds [51-53] or Ni-O and Co-O bonds [54]. Importantly, although the Ti
content in NiCo/Si-1.0Ti was lower than that in NiCo/Si-8.5Ti, the intensity of the
metal-oxygen groups was higher. This may lead to stronger interaction between the
activating centers and adsorbed H
2
(proved by H
peaks around 500 cm were attributed to the stretching vibration of Ti-O-Ti [55, 56],
indicating that TiO was still remain in the spent catalysts. According to the relevant
2
-TPD). In addition, the absorption
−
1
2
-1
literature [57], the absorption bands at around 950 cm corresponding to the Ti−O−Si
bond was not detected in the spectrum, revealing that there was not obvious
interaction between Ti and Si.
16

Fig. 5 FT-IR spectra of NiCo/Si, NiCo/Si-1.0Ti, NiCo/Si-8.5Ti and spent
NiCo/Si-1.0Ti catalysts.
The wide scanning spectra (Fig. 6a) indicates that the NiCo/Si, NiCo/Si-1.0Ti and
NiCo/Si-8.5Ti catalysts are mainly composed of Ni, Co, O, Si and/or Ti elements. The
high-resolution XPS spectra of Ni 2p, Co 2p, O 1s, Si 2p and Ti 2p are shown in Fig.
6
(b-f). The analysis of the Ni 2p spectra (Fig. 6b) indicates out that Ni is mainly in the
oxidized state [58]. Four characteristic peaks, i.e. Ni 2p3/2 main peak and its satellite
peak at ca. 855 and 861 eV, respectively, as well as Ni 2p1/2 main peak and its satellite
2
+
peak at ca. 873 and 879 eV, respectively, are in good agreement with Ni [59, 60].
The peak at the binding energy of 101.46 eV is assigned to Si 2p (Fig. 6c). Compared
with the NiCo/Si catalyst, the characteristic peaks attributed to the Ni 2p3/2 and Ni
2p1/2 species in the NiCo/Si-1.0Ti sample were shifted to lower binding energy (-0.54
eV and -0.67 eV, respectively). As anticipated, a positive shift (+0.27 eV) of the peak
ascribed to Si 2p in the NiCo/Si-1.0Ti sample was observed compared with that in
17

NiCo/Si. This is attributed to certain charge transfer from Si to the Ni atom. These
results indicated that there must be stronger interaction between Ni and Si in
NiCo/Si-1.0Ti than that in NiCo/Si. Thus, Ti doping in the NiCo/Si catalysts leads to
change in the electron density and charge distribution of the atoms on the surface of
the catalyst, which is beneficial for improving the activity of the catalyst.
The spectrum of O 1s (Fig. 6d) shows one peak at 530.85 eV for NiCo/Si. The
asymmetrical O 1s region of NiCo/Si-1.0Ti and NiCo/Si-8.5Ti can be divided into
two peaks at ca. 529 and 531 eV. The peaks at 529.34 and 529.06 eV are related to the
TiO lattice oxygen (Ti-O bond) for NiCo/Si-1.0Ti and NiCo/Si-8.5Ti, respectively,
2
and those at 530.85 and 530.67 eV are attributed to adsorbed water or hydroxyl
groups on the surface of catalysts [61], which was confirmed by the FT-IR
spectroscopy results.
According to the data in Fig. 6e, the Ti 2p3/2 and Ti 2p1/2 binding energies were
observed at ca. 457 eV and 463 eV for the two samples that were identified as TiO
2
[
62, 63], and the binding energy of Ti 2p decreased with increasing Ti content.
Moreover, in Fig. 6f, the Co 2p3/2 peaks in the spectra of NiCo/Si-1.0Ti and
NiCo/Si-8.5Ti are located at 780.07 and 779.79 eV, respectively, which exhibited
slight downshifts compared to that of NiCo/Si (780.19 eV). These shifts to lower
values can be indicative of the occurrence of metallic synergic effects [11], and the
electron density gradually increases at the interface of the Co atoms and forms Coδ−
[
64].
The strong metal−support interaction (SMSI) could explain this phenomenon [34].
18

The overflowing H
2
hydrogen from the surface of metallic Co diffuses onto the
reducible TiO surface to form partially reduced titania (TiO2−x). The transfer of
2
electron from TiO2−x to the adjacent metal Co atoms causes electron enrichment at the
interface of Co atoms. In the meantime, the strong Co−TiO2−x electronic interaction
can promote the formation of the oxygen vacancies of titania species (i.e., O
V
-Ti3+
δ−
−Ti3+
sites) near the perimeter of the metal−support interface and form the Co −O
V
interface sites, which play an important role as active sites for the catalytic
hydrogenation of phenol [65-67]. Furthermore, the shapes of the Si 2p and Ni 2p
spectra of NiCo/Si-1.0Ti and NiCo/Si-8.5Ti remained similar for different Ti loadings,
and almost no electrons from the Si 2p and Ni 2p were transferred between
NiCo/Si-1.0Ti and NiCo/Si-8.5Ti, indicating that Ti content does not affect the state
of the Si and Ni species.
19

Fig. 6 XPS spectra of NiCo/Si, NiCo/Si-1.0Ti and NiCo/Si-8.5Ti.
.2 Catalyst evaluation
3
In this work, the effect of different metal combinations and carrier types on the
hydrogenation of phenol was studied in the presence of 0.1 g catalyst and 0.05 g
phenol with 10 mL n-heptane at 100°C for 0.5 h. As shown in Table 2, for the Ni/Si
catalyst, the Ti loadings have negative effect on the conversion and increasing the
20

amount of Ti can intensify the negative effect and result in slight fluctuations on the
selectivity by contrasting group 1-3. On the other hand, the comparison of group 4-6
revealed that Ti loading can enhance the conversion of Co-containing catalyst and
along with the increasing amount of added Ti, the conversion was greatly improved
from 19.65 to 48.29%. For the NiCo-containing catalyst, adding Ti enormously
increases (2.6 times) the phenol conversion to reach to a rate of 70.16% and has no
negative effect on cyclohexanol selectivity (>99%), but further addition of the Ti
decreases the phenol conversion rate to 58.85%. It is strange that the effect of the Ti
content on catalytic activity over Ni/Si, Co/Si, and NiCo/Si were very different. We
found that Co and Ti had strong metal−support interaction (proven by XPS results),
δ−
−Ti3+
leading to the formation of the active sites for phenol hydrogenation (Co −O
V
interface sites). In other words, when the content of Ti was low (Si-Ti ratio of 8.5:1),
adding Ni had a positive effect on the metal−support interaction, and remarkably
increased the catalytic activity of NiCo/Si. In contrast, when the content of Ti was
high (Si-Ti ratio of 1:8.5), adding Ni decreased the interaction between Co and Ti, and
led to the decline of the catalytic activity. These findings indicated that the addition of
a certain amount Ti has a very positive effect on the catalytic activity of bimetal
NiCo/Si catalyst. Ultimately, NiCo/Si-1.0Ti (group 8) was selected for further
research.
Table 2 Catalyst screening of the hydrogenation on phenol.
Phenol conv.
%)
Cyclohexanol
sel.(%)
Group
Catalysts
(
21

1
2
3
4
5
6
7
8
9
Ni/Si
39.36
37.03
25.68
19.65
22.61
48.29
26.71
70.16
58.85
96.87
>99
Ni/Si-1.0Ti
Ni/Si-8.5Ti
Co/Si
95.46
>99
Co/Si-1.0Ti
Co/Si-8.5Ti
NiCo/Si
90.64
>99
>99
NiCo/Si-1.0Ti
NiCo/Si-8.5Ti
>99
>99
Reaction conditions : 0.05 g phenol, 10 mL n-heptane as solvent, 0.1 g catalyst,
00 °C, 1.0 MPa H and reaction time = 0.5 h.
The effect of H pressure on the phenol hydrogenation performance was
investigated by varying the H pressure in the range of 0.5–3 MPa. As can be seen in
Fig. 7a, the phenol conversion showed a trend of sharply rising and then achieved the
full conversion with the increase of the H pressure. It is indicated that optimum H
1
2
2
2
2
2
pressure is 1 MPa. The tests of reaction temperature on the conversion of phenol was
carried out and the results were illustrated in Fig. 7b. It is obvious that the
hydrogenation of phenol was considerably accelerated with an increase of reaction
temperature from 60°C to 120°C. It was also worth noting that cyclohexanol was the
only detected product in all tested reaction temperatures, revealing the excellent
selectivity of NiCo/Si-1.0Ti can maintain over a wide range of temperatures.
22

Fig. 7 Influence of H
NiCo/Si-1.0Ti catalyst in the hydrogenation reaction of phenol. Reaction conditions:
.05 g phenol, 10 mL n-heptane as solvent, 0.1 g catalyst, 100 °C (a), 1.0 MPa H (b),
and reaction time = 1 h.
2
pressure (a) and temperature (b) on the activity of the
0
2
The hydrogenation reaction of phenol was investigated at 100°C under 1.0 MPa using
the NiCo/Si, NiCo/Si-1.0Ti, NiCo/Si-3.5Ti, NiCo/Si-6.0Ti and NiCo/Si-8.5Ti
catalysts. As shown in Fig. 8a, with the increase of the reaction time, the
NiCo/Si-1.0Ti catalyst showed the best performance, and achieved almost full phenol
conversion (98.2%) with excellent cyclohexanol selectivity (>99%) in 1 h of reaction
time. In comparison, NiCo/Si-3.5Ti, NiCo/Si-6.0Ti, NiCo/Si-8.5Ti and NiCo/Si only
achieved 87.60%, 83.35%, 76.1% and 34.1%, respectively. Compared to NiCo/Si,
catalytic performance of Ti-doped catalysts were improved, but that of high Ti content
catalysts decreased their catalytic activity, maybe due to the decrease in porosity, as
shown by the SEM and BET results. The differences in the –ln(C/C ) versus time (t)
0
for the NiCo/Si, NiCo/Si-1.0Ti, NiCo/Si-3.5Ti, NiCo/Si-6.0Ti and NiCo/Si-8.5Ti are
shown in Fig. 8b. The reaction equation (1) is expressed as follows [61]:
C
-ln
 kt
C0
(1)
23

where C denotes the reaction concentration of phenol, C represents the initial
0
concentration of phenol, k is the apparent rate constant and t is the reaction time. The
apparent rate constant (k) of phenol hydrogenation can be obtained from the slope of
the linear fitted plot. The specific value of k for the five kinds of catalysts are listed in
Table 3, and it can be seen that the k value followed the trend NiCo/Si-1.0Ti >
NiCo/Si-3.5Ti > NiCo/Si-6.0Ti > NiCo/Si-8.5Ti > NiCo/Si. It reveals that doped Ti
can significantly increase the hydrogenation of phenol to produce cyclohexanol, and
the k value of NiCo/Si-1.0Ti was 3.6-fold higher than that of NiCo/Si, respectively.
These differences suggest that NiCo/Si-1.0Ti has a considerably higher hydrogenation
ability than the other four catalysts. Material balance is calculated using Eq. (2) [68]:
n(Phenol) + n(Cyclohexanol)
Material balance 
n0 (Phenol)
(2)
where n(phenol) and n(cyclohexanol) are the amounts of phenol and cyclohexanol
in hydrogenation reaction products, respectively, and n (phenol) is the amount of
0
phenol before the hydrogenation reaction. As shown in Table 3, the material balance
was 100%±1%. Turnover frequency (TOF) is calculated as bellow [34]:
nphenol Cphenol
TOF 
nmetal T
(3)
where nphenol is mol of substrate; C is phenol conversion (%); nmetal is mol of metal
atom; T is reaction time (h). In Table 3, phenol could be nearly completely converted
to the cyclohexanol over the NiCo/Si-1.0Ti (conversion 98% and selectivity >99%)
-1
with a TOF value of 4.76 h , which is approximately 2.06 times higher than that of
-1
δ−
3+
the NiCo/Si (2.31 h ). This indicates that the interface site (e.g., Co −O −Ti ) in
V
24

NiCo/Si-1.0Ti, but absent in NiCo/Si, is a key factor governing the catalytic
efficiency [34]. Moreover, high Ti-loaded catalyst showed low activity for this
reaction because TOF value of NiCo/Si-3.5Ti, NiCo/Si-6.0Ti and NiCo/Si-8.5Ti were
slightly lower than that of NiCo/Si-1.0Ti, indicating the high amount of Ti was not
very effective for creating highly active alloy-based sites.
Fig. 8 Comparison of the catalytic activity of the NiCo/Si-1.0Ti, NiCo/Si-3.5Ti,
NiCo/Si-6.0Ti, NiCo/Si-8.5Ti and NiCo/Si catalysts: conversion vs. reaction time of
phenol (a). Plots of –ln(C/C
NiCo/Si-6.0Ti, NiCo/Si-8.5Ti and NiCo/Si (b). Reaction conditions: 0.05 g phenol, 10
mL n-heptane as solvent, 0.1 g catalyst, 100°C, and 1.0 MPa H
0
) versus time (t) for the NiCo/Si-1.0Ti, NiCo/Si-3.5Ti,
2
.
Table 3 Catalytic performance of different catalysts in the hydrogenation reaction of
phenol to cyclohexanol.a
Conv. Sel.(%)
Rate
Material
TOF b
Entry Catalysts
-1
(%)
(Cyclohexanol) constant k balance (%) (h )
1
2
NiCo/Si-1.0Ti 98.19 ＞99
NiCo/Si-3.5Ti 87.60 ＞99
2.126
1.714
99.50
4.76
3.48
100.45
25

3
4
NiCo/Si-6.0Ti 83.35 ＞99
NiCo/Si-8.5Ti 76.11 ＞99
1.357
1.301
0.587
99.33
2.93
2.24
2.31
100.34
100.12
5
NiCo/Si
34.08 ＞99
a
Reaction condition: 0.05 g phenol, 10 mL n-heptane as solvent, 0.1 g catalyst, 100°C,
.0 MPa H and reaction time = 1 h.
Mole number of phenol converted per surface metal atom per second.
1
2
b
A possible pathway for the selective hydrogenation of phenol to cyclohexanol was
proposed in Fig. 9a. In pathway (i), the hydrogenation of phenol is a two-step reaction,
i.e. phenol is firstly hydrogenated to cyclohexanone, and cyclohexanone then be
further hydrogenated to cyclohexanol [69-72]. In pathway (ii), the hydrogenation
reaction proceeds through a one-step process, i.e. cyclohexanol is formed directly by
the one-step hydrogenation of phenol [11]. k
1
, k and k are the reaction rate constants
2
for each step in phenol hydrogenation. In this work, cyclohexanone was not detected
and cyclohexanol was the only product throughout the entire reaction process even at
a short reaction time (less than 10 min) and a very low phenol conversion rate (lower
than 15%), and thus, k
1
≪ k
2
. After calculated, clearly, rate constant of cyclohexanone
=2.581, see Fig. S2) was almost 10-fold higher than
hydrogenation to cyclohexanol (k
2
that of phenol hydrogenation to cyclohexanol (k=0.2611) over NiCo/Si-1.0Ti. This
indicated that cyclohexanone is too rapidly hydrogenated into cyclohexanol to be
detected under this two-step reaction over NiCo/Si-1.0Ti catalyst.
A possible catalytic mechanism for the hydrogenation of phenol using the NiCo/Si-Ti
catalyst is proposed in Fig. 9b. the phenol is adsorbed on the surface of NiCo/Si-1.0Ti
26

through the interaction between the OH and Ni-Co. The XPS analysis result of the
2
+
2+
NiCo/Si-1.0Ti catalyst indicated that the Ni and Co sites can potentially accept
lone pairs of electrons from donating groups of Si and Ti oxide species, respectively,
leading to the increase of the electron density around the Ni and Co atoms. The
interaction of Ni and Co species with the π-electrons of the phenol molecules led to
the formation of strong π-complex bonds between the Ni-Co and the oxygen atom of
the phenol molecule, which accelerated the adsorption of phenol. Thus, the electronic
interaction between Ni and Si, Co and Ti greatly influenced the adsorption capacity of
phenol. Then, Ni-Co attached to the adsorbed phenol to form an electron rich
intermediate. Additionally, the H
2
molecules were also dissociated into highly active
H atoms under the synergic effect of the metal−support, so as to form the
δ−
3+
Co −O −Ti interface sites. Afterwards, the H atoms attack the π electron of the
V
intermediate, which leads to the hydrogenation of this intermediate. With the
saturation of the benzene ring, the strength of the OH is considerably reduced, thus
the interaction between the OH and Ni-Co is remarkably weakened, causing the
dissociation of the final cyclohexanol product.
27

Fig. 9 Reaction pathways (a) and possible catalytic mechanism (b) of phenol
hydrogenation over the NiCo/Si-1.0Ti catalyst.
We compared our newly developed catalysts for the hydrogenation of phenol with
other reported non-sulfided inexpensive metal catalysts (Table 4). The catalytic
performance of NiCo/Si-1.0Ti was comparable or even superior to that of some of the
previously reported inexpensive metal catalysts. However, although the reaction
conditions and apparatus required for the reaction among the catalysts are different,
the shorter time, lower temperature and lower pressure required with our catalysts are
quite obvious under the respective optimal condition. In short, our catalysts
(NiCo/Si-1.0Ti) showed high phenol conversion and cyclohexanol selectivity under
28

mild conditions.
Table 4 Comparison of our newly developed catalysts with other non-sulfided
inexpensive metal catalysts for the hydrogenation of phenol.
Time
min)
Temp. H
2
Phenol
Cyclohexanol
Catalysts
(
(°C)
100
100
200
250
200
275
280
67
(MPa)
con. (%) sel. (%)
NiCo/Si-1.0Tia
NiCo@CN [11]
60
1
>99%
>99%
96.33
82.36
100
>99%
>99%
91.14
85
720
240
240
240
300
240
480
300
120
120
960
0.8
2
NiCo@C/ZrO
2
[27]
NiCo/γ-Al
2
O
3
[28]
2
Ni/SiO
Ni/SiO
2
2
[73]
[74]
1
100
10
4
100
≈90
--
Ni/Al
Ni/NiO [75]
Ni/CeO [13]
Raney Ni [12]
Ni P-CePO [76]
@porous carbon [9]
2
O
3
[19]
83
2
92
97
2
275
220
150
150
10
2
100
≈90
51.9
≈100
76.5
86.75
≈90
3
4
4
CoO
x
3
98
a
Reaction conditions: catalyst 0.1ꢀg, phenol 50ꢀmg, n-heptane 10ꢀmL.
The data presented in Fig. 10a reveal that the reaction mixture became clarified very
quickly when magnets were placed on the right of the bottle after 20 seconds, and this
simple method was used in the recovery of the catalyst. The recovered catalysts were
reused in the next run after reduction by H at 300°C for 3 h. The results in Fig. 10b
2
29