Nanoparticulate Ru on TiO2 exposed the {1?0?0} facets: Support facet effect on selective hydrogenation of benzene to cyclohexene

Article abstract of DOI:10.1016/j.jcat.2018.11.032

Support facet effect of supported metal catalysts for conversion of environmentally toxic feedstock to value-added product is challenging but of significant importance for the understanding of the reaction mechanism and accordingly rational design of newly high-efficient catalytic materials. We report here the effect of the exposed degree of TiO₂ {1 0 0} facets, which was facilely tuned by hydrothermal temperature, on benzene selective hydrogenation to cyclohexene over TiO₂-supported nanoparticulate Ru catalysts. It is found that the Ru/TiO₂ catalyst exposed the most {1 0 0} facets exhibited slightly lower intrinsic activity, but unexpectedly highest selectivity towards cyclohexene (initial selectivity of 93%). A positively linear correlation between the initial selectivity to cyclohexene and the amount of surface OH groups on the Ru/TiO₂ catalysts, which was determined by the exposed degree of TiO₂ {1 0 0} facets, was identified. This linearity rationalizes the selectivity enhancement of the Ru/TiO₂ catalyst exposed more {1 0 0} facets since the surface OH groups functioned in increasing the hydrophilicity of the catalysts, decreasing the type of adsorption sites of cyclohexene, and blocking the chemisorption sites of cyclohexene.

Full text of DOI:10.1016/j.jcat.2018.11.032

Journal of Catalysis 369 (2019) 352–362
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Nanoparticulate Ru on TiO2 exposed the {1 0 0} facets: Support facet
effect on selective hydrogenation of benzene to cyclohexene
⇑
Gongbing Zhou , Lan Jiang, Daiping He
Chongqing Key Laboratory of Green Synthesis and Applications, College of Chemistry, Chongqing Normal University, Chongqing 401331, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Support facet effect of supported metal catalysts for conversion of environmentally toxic feedstock to
value-added product is challenging but of significant importance for the understanding of the reaction
mechanism and accordingly rational design of newly high-efficient catalytic materials. We report here
the effect of the exposed degree of TiO2 {1 0 0} facets, which was facilely tuned by hydrothermal temper-
ature, on benzene selective hydrogenation to cyclohexene over TiO2-supported nanoparticulate Ru cata-
lysts. It is found that the Ru/TiO2 catalyst exposed the most {1 0 0} facets exhibited slightly lower intrinsic
activity, but unexpectedly highest selectivity towards cyclohexene (initial selectivity of 93%). A positively
linear correlation between the initial selectivity to cyclohexene and the amount of surface OH groups on
the Ru/TiO2 catalysts, which was determined by the exposed degree of TiO2 {1 0 0} facets, was identified.
This linearity rationalizes the selectivity enhancement of the Ru/TiO2 catalyst exposed more {1 0 0} facets
since the surface OH groups functioned in increasing the hydrophilicity of the catalysts, decreasing the
type of adsorption sites of cyclohexene, and blocking the chemisorption sites of cyclohexene.
Ó 2018 Elsevier Inc. All rights reserved.
Received 23 September 2018
Revised 19 November 2018
Accepted 26 November 2018
Keywords:
Ruthenium
TiO2 facet
Benzene
Cyclohexene
Hydrogenation
1. Introduction
the most studied {0 0 1}, {1 0 1}, and {1 0 0} facets, although the
stability decreased in the sequence of {0 0 1} > {1 0 1} > {1 0 0}
Design of TiO2 with specific crystal facets has been intensively
studied since the surface atomic coordination and configuration,
which essentially affect its catalytic performance, can be well
tuned by crystal facet control [1–3]. For instance, owing to a coop-
erative mechanism of surface atomic structure and surface elec-
tronic structure, the photoreactivity of {0 0 1}, {1 0 1}, and {0 1 0}
facets of anatase TiO2 crystals in photooxidation and photoreduc-
tion reactions was proved to be {0 1 0} > {1 0 1} > {0 0 1} [1].
Among the three natural phases of TiO2 (anatase, rutile, and broo-
kite), anatase is most intensively studied and applied in catalysis
and photocatalysis owing to its higher stability in strongly basic
solution or when the particle size is below 11 nm [4]. To investi-
gate the facet effect of anatase TiO2, synthesis of TiO2 crystals with
specific facets is highly required. Until now, the low-index facets
({0 0 1} [1,5], {1 0 1} [1,5], {1 0 0} [6], {0 1 0} [1], and {1 1 0} [7])
and high-index facets ({1 0 3}, {1 0 5}, {1 0 6}, {1 0 7}, {2 0 1},
{3 0 1}, and {4 0 1}) [4] of anatase crystals have been synthesized
by wet-chemistry route, epitaxial growth, gas oxidation route,
crystallization transformation, and topotactic transformation. For
since the relative surface energies increased on the contrary, the
{1 0 0} facets were more active and consequently showed higher
catalytic activity than the other two facets as predicted by DFT cal-
culations [8].
In contrast, studies regarding with the facet effect of TiO2 when
serving as a support for metal nanoparticles (NPs) are rare [9–11],
while it is remarkably important from both scientific and techno-
logical perspectives as it may provide new tuning factors or mech-
anisms to improve the catalytic performance of supported metal
catalysts other than the known physical effect of support and
‘‘strong metal-support interaction (SMSI)” [12–16]. Different facets
with diverse surface atomic configurations displayed unique inter-
actions with metal NPs, thereby exhibiting different catalytic per-
formances. Sui et al. demonstrated that the Pt atoms selectively
deposited on TiO2 with a proper percentage of exposed {0 0 1}
and {1 0 1} facets were more reactive than those on TiO2 with
dominated {1 0 1} facets or {0 0 1} facets in sunlight-driven hydro-
gen production from water splitting, owing to the enhanced ‘‘sur-
face heterojunction” effect between {0 0 1} and {1 0 1} facets on
the former [9]. Cao et al. reported that the Pd NPs enclosed by
{1 0 0} or {1 1 1} facets dosed on TiO2 (0 0 1) crystal plane exhib-
ited higher activity than which on TiO2 (1 0 1) plane in oxygen acti-
vation reaction [10]. Therefore, the selection of the support facets
⇑
Corresponding author.
E-mail address: gbzhou@cqnu.edu.cn (G. Zhou).
https://doi.org/10.1016/j.jcat.2018.11.032
0021-9517/Ó 2018 Elsevier Inc. All rights reserved.

G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
353
can offer a new tuning parameter to regulate the catalytic perfor-
mance of metal NPs.
obtained products were labeled TiO2-x, where x represents the
temperature of the second hydrothermal treatment step.
The above TiO2 samples were then used as supports to prepare
the Ru/TiO2 catalysts by the wetness impregnation–chemical
reduction strategy. 0.5 g of the as-prepared TiO2 sample was dis-
persed in 10 ml of deionized water. 1.5 ml of RuCl3Á3H2O aqueous
solution (0.40 M) was added. After stirring for 3.5 h, 1.5 ml of KBH4
aqueous solution (1.58 M) was added dropwise. The nominal Ru
loading was 10.7 wt% for the catalysts. The solids were washed
with deionized water until inexistence of ClÀ1 (AgNO3 test). The
obtained catalysts were denoted as Ru/TiO2-x according to the kind
of the respective TiO2 support. For comparison, the Ru/Al2O3, Ru/
SiO2, Ru/CoO, Ru/NiO, Ru/CuO, and Ru/ZnO catalysts were also pre-
pared via the same strategy.
These leading works inspire us to explore the facet effect of TiO2
when serving as a support for Ru NPs on other industrially impor-
tant reactions, such as the conversion of environmentally toxic
feedstock to value-added product (benzene selective hydrogena-
tion to cyclohexene) addressed in the present work. The tetragonal
faceted-nanorods of anatase TiO2 single crystals with different
exposed degrees of highly reactive {1 0 0} facets were firstly syn-
thesized by hydrothermal route at different temperatures (433–
473 K), and then the Ru/TiO2 catalysts were prepared by a facile
wetness impregnation–chemical reduction method. We identified
that the hydrothermal temperature is essential to the exposed
degree of {1 0 0} facets, and the temperature to expose the most
{1 0 0} facets is 453 K. Meanwhile, O1s X-ray photoelectron spec-
troscopic (XPS) spectrum, Fourier transform infrared spectroscopy
(FTIR), and water contact angle demonstrated that the hydrophilic-
ity of the Ru/TiO2 catalysts was in line with the exposed degree of
TiO2 {1 0 0} facets. In the selective hydrogenation of benzene to
cyclohexene, the Ru/TiO2-453 catalyst exhibited higher selectivity
than the Ru/TiO2-433, Ru/TiO2-443, and Ru/TiO2-473 catalysts pre-
pared in the same manner. An excellent linear correlation between
the cyclohexene selectivity and the hydrophilicity of these cata-
lysts was confirmed, which rationalizes the significant impact of
TiO2 facets on selectivity enhancement in selective hydrogenation
of benzene over Ru/TiO2 catalysts.
2.2. Catalytic evaluation
Selective hydrogenation of benzene to cyclohexene was tested in
a 250 ml-capacity autoclave with a mechanical stirrer. After intro-
ducing 0.5 g of catalyst, 5.0 g of ZnSO4Á7H2O, 50 ml of H2O, and
25 ml of benzene, the autoclave was sealed and purged with H2.
The reaction conditions are temperature of 413 K, H2 pressure of
5.0 MPa, and stirring speed of 1000 rpm, which are typical for the
selective hydrogenation of benzene [12,17–21]. The product was
analyzed by a SC-3000B gas chromatograph with a thermal conduc-
tivity detector (TCD) and a PEG-20M packed column. The catalysts
were tested in duplicate, and the results agreed within 2%.
As the benzene conversion, the cyclohexene selectivity, and the
cyclohexene yield changed with reaction time (t), the turnover fre-
quency (TOF) of benzene and the initial selectivity towards cyclo-
hexene (S0), which were absolutely fair for all the catalysts, were
used to compare the intrinsic activity and selectivity, respectively.
For the calculation of the TOF, the initial weight-specific activity
(r0), that is, the reacted millimoles of benzene per gram of the cat-
alyst per minute at the beginning of the reaction, was acquired first
by the approach reported by Hu et al. [22]. Precisely, the polyno-
mial equation acquired by fitting the experimental curve of ben-
zene content and reaction time was differentiated, and the r0
was obtained by using zero to replace t. The TOF was calculated
according to the formula raised in our previous work [23]. For
the S0, the linear correlation of cyclohexene selectivity and t was
fitted, and the intercept is S0.
2. Experimental
2.1. Preparation
RuCl3Á3H2O was purchased from Adamas. P25 TiO2 was sup-
plied by Degussa. Benzene was purchased from Sigma-Aldrich.
Other unspecified chemicals were provide by Sinopharm Chemical
Reagent. All chemicals are analytical grade (A.R.). The gases were
purchased from Chongqing Ruike Gas Co.
The fabrication of the Ru/TiO2 catalysts is graphically illustrated
in Scheme 1. Firstly, TiO2 crystals exposed different degrees of
{1 0 0} facets were synthesized according to the strategy developed
by Li et al. with some modifications [6]. Specifically, 4 g P25 TiO2
and 80 ml of NaOH aqueous solution (10 M) were mixed and trans-
ferred to a 200 ml-capacity Teflon-lined stainless steel autoclave.
The autoclave was sealed and hydrothermally treated at 393 K
for 24 h. After cooling down to room temperature, the white pre-
cipitates were separated and washed with deionized water until
the pH of approximate 10.5, producing Na-titanate intermediates.
Then, 2 g of the wet intermediates were dispersed into 80 ml of
deionized water and transferred to a 100 ml-capacity Teflon-
lined stainless steel autoclave. The autoclave was sealed and
hydrothermally treated at 433, 443, 453, and 473 K for 24 h,
respectively. After cooling down to room temperature, the white
precipitates were separated by centrifugation and washed with
deionized water until neutrality, and dried at 333 K for 10 h. The
3. Results and discussion
3.1. Texture and morphology of TiO2 nanorods
Table 1 lists the physicochemical parameters of the TiO2 sam-
ples. The N2 adsorption–desorption isotherms of the samples
(Fig. S1) all belonged to type IV with H3 hysteresis loop. According
to Table 1, a sharp decrease of the multipoint Brunauer–Emmett–
Teller surface area (SBET) and the pore volume (Vpore) was observed
when the hydrothermal temperature was elevated from 433 to
Scheme 1. Schematic illustration of the synthesis of TiO2 single crystals with dominated {1 0 0} facets-supported Ru catalysts.

354
G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
Table 1
Physicochemical properties of TiO2 and the Ru/TiO2 catalysts.
Sample
Ru loadinga (wt%)
Dispersionb (%)
SBET (m2 gÀ1
)
dpore (nm)
Vpore (cm3 gÀ1
)
TiO2-433
TiO2-443
TiO2-453
TiO2-473
Ru/TiO2-433
Ru/TiO2-443
Ru/TiO2-453
Ru/TiO2-473
n.a.c
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
25.8
24.9
26.2
25.3
35
6
7
8.2
0.19
0.06
0.05
0.04
0.26
0.09
0.08
0.08
22.5
23.1
30.5
5.9
17.9
13.7
18.6
n.a.
3
10.5
10.4
10.5
10.6
90
14
16
12
a
b
c
Determined by ICP-AES.
Dispersion of Ru determined by CO chemisorption.
n.a.: not applicable.
443 K, while further elevation of the temperature induced slight
impact on SBET and Vpore. For the average pore diameter (dpore), it
increased gradually accompanying the elevation of the hydrother-
mal temperature. The powder X-ray diffraction (XRD) patterns of
the TiO2 samples as well as the Na-titanate intermediate are pre-
sented in Fig. 1. The peaks in Fig. 1a–d at 2h of 25.4°, 37.0°,
37.9°, 38.7°, 48.2°, 54.1°, 55.2°, 62.1°, 62.7°, and 68.9° are respec-
tively ascribed to the (1 0 1), (1 0 3), (0 0 4), (1 1 2), (2 0 0),
(1 0 5), (2 1 1), (2 1 3), (2 0 4), and (1 1 6) crystal planes of anatase
TiO2 (tetragonal, JCPDS 21-1272), showing that the Na-titanate
intermediate was transferred into an anatase phase after
hydrothermal treatment.
As a representative, the scanning electron microscopy (SEM)
image of the TiO2-453 sample is showed in Fig. 2a. It indicated a
nanorod shape enclosed by four tetragonal facets, and the adjacent
facets of the nanorod were perpendicular (inset in Fig. 2a). The
transmission electron microscopy (TEM) images with correspond-
ing selected-area electron diffraction (SAED) patterns (Fig. 2b and
c) revealed more structural information of the TiO2 nanorod. The
SAED pattern in Fig. 2b revealed that the [0 0 1] and [1 0 0] direc-
tions are perpendicular, which is in line with the schematic model
of anatase TiO2 nanorod enclosed by {1 0 0} facets along the [0 1 0]
direction (inset in Fig. 2b). The high-resolution TEM (HRTEM)
image taken from the rectangle revealed two sets of lattice fringes
with spacings of 3.52 and 2.43 Å, attributable to (1 0 1) and (1 0 3)
single-crystalline structure of anatase exposed with {1 0 0} facets,
matching well with the schematic model enclosed by {1 0 0} facets
À
along the [0 2 1] direction (inset in Fig. 2c). Based on the above
analysis, the exposed lateral facets of the as-prepared TiO2 nanor-
ods are mainly the {1 0 0} facets, in agreement with the observa-
tion of Li et al. [6].
3.2. Physicochemical characteristics, morphology, and microstructure
of the Ru/TiO2 catalysts
With the as-synthesized TiO2 samples, we prepared the Ru/TiO2
catalysts. The physicochemical properties of the catalysts are listed
in Table 1. The Ru loadings on the catalysts are similar of about
10.5 wt%. The dispersion of Ru remained virtually constant at
approximate 25% irrespective of the kinds of supports. The N2
adsorption–desorption isotherms of the catalysts (Fig. S2) are anal-
ogous to the corresponding supports, while the SBET and Vpore of the
catalysts are higher and the dpore is lower than those of the corre-
sponding supports owing to the deposition of the Ru NPs. Fig. 3 dis-
played the XRD patterns of the Ru/TiO2 catalysts. After Ru loading,
the crystallographic form of the TiO2 samples was retained. Mean-
while, an extra weak and diffuse feature appeared at ca. 44°,
attributing to the (1 0 1) diffraction of the hexagonally close
packed (hcp) Ru [12]. This weak and diffuse feature was originated
from the high dispersion and small size of Ru on the Ru/TiO2 cata-
lysts, as indicated by CO chemisorption results (Table 1) and TEM
images shown below.
Fig. 4a exhibited the SEM image of the Ru/TiO2-453 catalyst. It is
obvious that the morphology of TiO2 remained perfectly after the
loading of Ru NPs, while the facets became rough. The distribution
of Ru on TiO2 was observed by the energy-dispersive X-ray (EDX)
elemental mapping analysis with corresponding high angle annu-
lar dark field-scanning transmission electron microscopy
(HAADF-STEM) image (Fig. 4b). It is clear that the Ru NPs were uni-
formly distributed in the entire region. Fig. 4c presents the TEM
image, SAED pattern, and particle size distribution (PSD) histogram
with Gaussian analysis fitting of the Ru/TiO2-453 catalyst. In this
figure, the Ru NPs verified by EDX shown in Fig. S3c were evenly
distributed on the (1 0 0) planes of TiO2, and the mean size is
3.4 nm with a PSD ranging in 2–5 nm. The SAED pattern inset in
Fig. 4c shows diffraction rings and random dots, corresponding to
the diffraction of the Ru NPs and TiO2, respectively, which agreed
with the short-range ordered and long-range disordered atomic
arrangement of the Ru NPs prepared by the chemical reduction
method [24]. The HRTEM image in Fig. 4d provides more structural
details of the Ru/TiO2-453 catalyst. The lattice fringes with the
interplanar spacings of approximate 2.05 and 3.52 Å ascribable to
the (1 0 1) planes of hcp Ru and TiO2, respectively, were visible,
which were further confirmed by the fast Fourier transform (FFT)
pattern inset in Fig. 4d.
planes of anatase, respectively. To confirm the exposed facets, the
À
nanorod along the [0 2 1] direction was showed in Fig. 2c. The
width of the nanorod along this direction is 1.4 times as that along
the [0 1 0] direction. Both the SEAD pattern and HRTEM image
inserted of Fig. 2c demonstrated that the TiO2 nanorod had a
Anatase JCPDS: 21-1272
d
c
b
a
Na-titanate intermediate
10
20
30
40
50
60
70
2 (degree)
Fig. 1. XRD patterns of Na-titanate intermediate and (a) TiO2-433, (b) TiO2-443, (c)
TiO2-453, and (d) TiO2-473.

G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
355
Fig. 2. (a) SEM image of TiO2-453. The inset is the top view of a single TiO2 nanorod; (b) TEM image of an individual nanorod of TiO2-453 viewed along the [0 1 0] direction.
The insets are the corresponding SEAD pattern, schematic model of an ideal nanorod projected along the [0 1 0] direction, and HRTEM image taken from the nanorod
À
indicated by the rectangle; (c) TEM image of an individual nanorod of TiO2-453 viewed along the [0 2 1] direction. The insets are the corresponding SEAD pattern, schematic
À
model of an ideal nanorod projected along the [0 2 1] direction, and HRTEM image taken from the nanorod indicated by the rectangle.
ducing the ideal TiO2 nanorods with dominated {1 0 0} facets.
Anatase JCPDS: 21-1272
Ru JCPDS: 06-0663
Excessively low temperature resulted in the partial formation of
the nanorods (Fig. 5a1), while too high temperature led to the
aggregation of the nanorods and consequently, the nanorods
became irregular (Fig. 5c1). The optimum temperature for obtain-
ing the perfect nanorods is 453 K. Fig. 5a2–c2 indicated that the Ru
NPs verified by EDX shown in Fig. S3 were uniformly distributed on
the supports, which was further proved by the EDX elemental
mapping analysis (Fig. S4). The average sizes of the Ru NPs are
identical of 3.3 nm and the PSD ranges are similar of 2–5 nm,
which is qualitatively in agreement with the results of CO
chemisorption. The SAED patterns of the Ru/TiO2-433, Ru/TiO2-
443, and Ru/TiO2-473 catalysts are also similar to that of the Ru/
TiO2-453 catalyst. The HRTEM images of these catalysts shown in
Fig. 5a3–c3 indicated that the Ru NPs are approximately spherical
and in close contact with the supports. The lattice fringes on the Ru
NPs had the same interplanar spacings of approximate 2.05 Å,
ascribing to the (1 0 1) planes of hcp Ru. The short-range ordered
and long-range disordered feature of these lattice fringes embod-
ied the amorphous alloy structure of the Ru NPs [24]. Moreover,
the lattice fringes with the interplanar spacings of approximate
2.43, 3.52, and 2.38 Å were visualized for the Ru/TiO2-433, Ru/
TiO2-443, and Ru/TiO2-473 catalysts, respectively, attributing to
the (1 0 3), (1 0 1), and (0 0 4) planes of TiO2.
d
c
b
a
10
20
30
40
50
60
70
2 (degree)
Fig. 3. XRD patterns of the (a) Ru/TiO2-433, (b) Ru/TiO2-443, (c) Ru/TiO2-453, and
(d) Ru/TiO2-473 catalysts.
For the Ru/TiO2-433, Ru/TiO2-443, and Ru/TiO2-473 catalysts,
the SEM images (Fig. 5a1–c1) showed that the morphologies of
TiO2 are nonuniform than that of the Ru/TiO2-453 catalyst, evi-
dencing the crucial role of the hydrothermal temperature for pro-

356
G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
Fig. 4. (a) SEM image with the top view of a single nanorod indicated by the rectangle, (b) mapping EDX and corresponding HAADF-STEM image, (c) TEM image with the
corresponding PSD histogram of the Ru NPs with Gaussian analysis fitting and SAED image, and (d) HRTEM image with FFT pattern indicated by the rectangle of the Ru/TiO2-
453 catalyst.
Fig. 5. SEM images (left), TEM images with the corresponding PSD histograms of the Ru NPs with Gaussian analysis fittings and SAED images (middle), and HRTEM images
(right) of the (a1–a3) Ru/TiO2-433, (b1–b3) Ru/TiO2-443, and (c1–c3) Ru/TiO2-473 catalysts.

G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
357
Table 2
3.3. Chemical state and hydrophilicity of the Ru/TiO2 catalysts
XPS results of the Ru/TiO2 catalysts.
Fig. 6 displays the Ru3p and Ti2p spectra of the Ru/TiO2-433, Ru/
TiO2-443, Ru/TiO2-453, and Ru/TiO2-473 catalysts. In consideration
of the partly overlapping between the Ru3p3/2 and Ti2p peaks, the
deconvolution of the spectra is conducted to ascertain the relative
contributions. For all four catalysts, the Ti2p spectrum exhibited
two peaks at 458.8 eV (Ti 2p3/2) and 464.5 eV (Ti 2p1/2), assigning
to Ti4+ oxidation state [25]. The Ru3p1/2 BE and the 3p3/2–3p1/2 dou-
blet separation are 484.1 eV and 22.1 eV, respectively, evidencing
the metallic Ru [25]. This assignment can be further confirmed
by the Ru3d spectra shown in Fig. S5, since the Ru3d5/2 BE and
the 3d5/2–3d3/2 doublet separation are 280.1 eV and 4.2 eV, respec-
tively, for the four catalysts. The Ru/Ti surface molar ratios of the
Ru/TiO2 catalysts originated from Fig. 6 are listed in Table 2. All
four catalysts exhibited a higher Ru/Ti surface molar ratio than
the corresponding bulk one, signifying a surface segregation of Ru.
As pointed out by Li et al. [6], in the preparation of TiO2 nanor-
ods, the OHÀ released from Na-titanate intermediates would pref-
erentially adsorb onto {1 0 0} facets of anatase TiO2. We inferred
that this OHÀ adsorption would increase the hydrophilicity of
TiO2 nanorod with dominated {1 0 0} facets, and subsequently
increasing the hydrophilicity of the Ru/TiO2 catalyst. To verify this
assumption, FTIR and water contact angle analyses were carried
out. In Fig. S6, the FTIR spectra of the TiO2 samples after saturation
adsorption of H2O are exhibited. The weak band at 2353 cmÀ1 is
probably assigned to carbonaceous contaminants. The broad and
strong band at 3473 cmÀ1 is assigned to the stretching vibration
of hydrogen-bonded OH groups [26]. This band was increased in
intensity when the hydrothermal temperature increased from
433 K to 453 K, and then weakened at 473 K. The integral band
intensity obeyed the order of TiO2-433 < TiO2-473 < TiO2-
443 < TiO2-453. Meanwhile, the integral intensity of the band at
1649 cmÀ1 attributable to the bending vibration of intact water
Catalyst
Ru/Ti surface molar ratioa
OOH/Ti molar ratiob
Ru/TiO2-433
Ru/TiO2-443
Ru/TiO2-453
Ru/TiO2-473
1.18 (0.075)
1.29 (0.075)
1.29 (0.075)
1.88 (0.076)
0.48
0.62
0.68
0.60
a
Data in parentheses are the bulk molar ratios determined by inductively cou-
pled plasma-atomic emission spectroscopy (ICP–AES).
b
OOH, surface hydroxyl.
O 1s
3460
OL
34.1o
OOH
2353
OH2O
1634
d
d
20.1o
c
b
a
c
b
27.6o
39.2o
a
4000 3500 3000 2500 2000 1500
Wavenumber (cm-1)
536 534 532 530 528 526
Binding Energy (eV)
Fig. 7. (A) The FTIR spectra obtained at 298 K of the (a) Ru/TiO2-433, (b) Ru/TiO2-
443, (c) Ru/TiO2-453, and (d) Ru/TiO2-473 catalysts. Insets are the corresponding
water contact angles; (B) The O1s spectra of the (a) Ru/TiO2-433, (b) Ru/TiO2-443,
(c) Ru/TiO2-453, and (d) Ru/TiO2-473 catalysts.
exhibited the same sequence as that of the band at 3473 cmÀ1
.
summarized in Table 2. It is obvious that the Ru/TiO2-453 catalyst
with the highest exposed degree of {1 0 0} facets on TiO2 possessed
the highest OOH/Ti ratio, verifying that the OHÀ released from Na-
titanate intermediates would preferentially adsorb onto the {1 0 0}
facets of TiO2 and subsequently increase the hydrophilicity of the
Ru/TiO2 catalyst. The OOH/Ti surface molar ratios increased in the
sequence of Ru/TiO2-433 < Ru/TiO2-473 < Ru/TiO2-443 < Ru/TiO2-
453, conforming with the hydrophilicity sequence disclosed by
the FTIR and water contact angle characterizations.
After the loading of Ru NPs, the integral intensity order of the band
at 3460 cmÀ1 remained unchanged (Fig. 7A). The water contact
angle of the Ru/TiO2 catalysts inserted in Fig. 7A evolved in the
order of Ru/TiO2-433 > Ru/TiO2-473 > Ru/TiO2-443 > Ru/TiO2-453,
further demonstrating the contrary hydrophilicity sequence of
the Ru/TiO2 catalysts. The variation in the hydrophilicity of the
Ru/TiO2 catalysts can be interpreted by O1s XPS analysis. As shown
in Fig. 7B, the deconvolution of the O1s spectra afforded three
peaks at 530.1, 531.9, and 533.1 eV, attributing to lattice oxygen
in TiO2 (OL), oxygen in hydroxyl groups (OOH), and oxygen in H2O
(OH2O), respectively [27,28]. The OOH/Ti surface molar ratios are
3.4. Acidity of the Ru/TiO2 catalysts
Exposure of Lewis and Brønsted acid sites on oxide surface is an
essential demand for adsorption of reactive molecules and cat-
alytic application. FTIR spectroscopy of adsorbed small alkaline
molecules, such as pyridine, NH3, CO, CH3CN, or NO, is widely rec-
ognized as a powerful technique to characterize the nature and
concentration of acid sites. As an IR probe molecule, pyridine is
superior to others in terms of higher selectivity and stability than
NH3, stronger adsorption than CO and CH3CN, and relatively higher
sensibility to the strength of acid sites than NO [29]. Hence, FTIR of
adsorbed pyridine (Py-IR) characterization was conducted to
investigate the nature and concentration of acid sites on the Ru/
TiO2 catalysts. According to previous works [29–32], various vibra-
tion bands of pyridine originated from different adsorption sites
appeared at around 1446 (Lewis acid sites-coordinated pyridine,
L-Py), 1491 (ring vibration of adsorbed pyridine), 1581 (hydrogen
bonded pyridine, hb-Py), and 1596 cmÀ1 (strong L-Py) in Py-IR
spectrum of TiO2 at room temperature. If the concentration of acid
sites needs to be evaluated, the IR spectra should be recorded at
423 K to eliminate the physically adsorbed pyridine [30].
Ru 3p + Ti 2p
Ti 2p3/2
Ru 3p3/2
Ru 3p1/2
484.1
Ti 2p1/2
Ti satellite
d
c
b
a
490
480
470
460
Binding Energy (eV)
Fig. 6. The Ru3p and Ti2p spectra of the (a) Ru/TiO2-433, (b) Ru/TiO2-443, (c) Ru/
TiO2-453, and (d) Ru/TiO2-473 catalysts.

358
G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
On basis of the above results, we collected the Py-IR spectra of
the Ru/TiO2 catalysts at 423 K, as shown in Fig. 8. Interestingly, dif-
fering from the previous observation that only the Lewis acid sites
existed on TiO2 [30,33], an additional band at 1545 cmÀ1 attribut-
ing to Brønsted acid sites-coordinated pyridine appeared for all the
catalysts. According to Arrouvel et al. [34], a rather high amount of
OH groups, which represented the Brønsted acid sites, will gener-
ate on the (1 0 0) surface of anatase TiO2 when it was hydrotreated.
Therefore, the origin of the Brønsted acid sites on the Ru/TiO2 cat-
alysts can be assigned to the OH groups on the {1 0 0} facets of TiO2
as corroborated by O1s spectra. By integrating the intensity of the
bands at 1446 and 1545 cmÀ1, the amounts of Lewis acid sites
(nL-Py) and Brønsted acid sites (nPyH+), as well as the total amount
of acid sites determined by Py-IR (nPy) were calculated, as shown
in Table 3. It is obvious that the nPyH+ and nPy of the Ru/TiO2 catalysts
decreased with the elevation of the hydrothermal temperature.
The effect of the exposed degree of the {1 0 0} facets of TiO2 on
the amount and strength of the acid sites on the Ru/TiO2 catalysts
was further investigated by temperature-programmed desorption
(TPD) of NH3 (NH3-TPD), as displayed in Fig. 9. A similar three
peak-profile at 452, 666, and 791 K attributing to NH3 adsorbed
on weak acid sites, acid sites with medium strength, and strong
acid sites [35], respectively, was exhibited for all four catalysts,
suggesting that the exposed degree of the {1 0 0} facets of TiO2
had no impact on the strength of the acid sites. However, the peak
intensity of the desorbed NH3 decreased with the elevation of the
hydrothermal temperature. By integrating the peak areas, the over-
all amounts of acid sites (nNH3) were calculated and displayed in
Table 3. It is evident that the NH3-TPD and Py-IR matched well
on the variation of the acid amounts, and the observation that
nNH3 is always higher than nPy is in line with the smaller kinetic
diameter and higher basicity of NH3 than pyridine.
666
452
791
d
c
b
a
400 500 600 700 800 900 1000
Temperature (K)
Fig. 9. NH3-TPD profiles of the (a) Ru/TiO2-433, (b) Ru/TiO2-443, (c) Ru/TiO2-453,
and (d) Ru/TiO2-473 catalysts.
rGmθ
+3H2
=
rGmθ
=
+2H2
+H2
Second step
First step
Scheme 2. The pathway and thermodynamics of benzene hydrogenation.
3.5. Selective hydrogenation of benzene to cyclohexene
L-Py
Strong L-Py + hb-Py
1446
Owing to the reactive C@C bond, cyclohexene molecules are
facilely converted to highly value-added chemicals such as cyclo-
hexanol, adipic acid, and caprolactam by conventional olefin reac-
tions. However, the achievement of a high cyclohexene yield from
the environmentally toxic feedstock-eliminated, atom-economic,
and operation-simplified route of benzene hydrogenation is
severely restricted to thermodynamics, that is, the standard free
energy change for the formation of complete hydrogenation pro-
PyH+
1596
1581
1545
d
c
b
duct (cyclohexane) from benzene hydrogenation (À98 kJ molÀ1
)
is more negative than that of the formation of intermediate cyclo-
hexene (À23 kJ molÀ1) [36,37], as shown in Scheme 2. Besides,
first-principles calculations substantiated that the energy barrier
of the first hydrogenation step is higher than that of the second
step even on the most selective Ru metal [38]. These limitations
resulted in the relatively low cyclohexene yield on a majority of
reported Ru-based catalysts.
a
1650
1600
1550
1500
1450
1400
Wavenumbers (cm-1)
Fig. 8. Py-IR spectra of the (a) Ru/TiO2-433, (b) Ru/TiO2-443, (c) Ru/TiO2-453, and
(d) Ru/TiO2-473 catalysts at 423 K. The spectra had been normalized by sample
mass.
Table 3
The acidic properties of the Ru/TiO2 catalysts.
AL-Py (cmÀ1 gÀ1
)
APyH+ (cmÀ1 gÀ1
)
nL-Py (mmol gÀ1
)
nPyH+ (mmol gÀ1
)
nPy (mmol gÀ1
e
)
nNH3 (mmol gÀ1
)
a
b
c
d
f
Catalyst
Ru/TiO2-433
Ru/TiO2-443
Ru/TiO2-453
Ru/TiO2-473
163
179
150
201
111
75
47
7
231
253
212
284
209
141
88
440
394
300
297
2965
1412
1386
1319
13
a
b
c
d
e
f
The integral absorption band intensity of Lewis acid sites-bonded pyridine (L-Py) at 1446 cmÀ1 in the Py-IR spectra collected at 423 K.
The integral absorption band intensity of Brønsted acid sites-bonded pyridine (PyH+) at 1545 cmÀ1 in the Py-IR spectra collected at 423 K.
The amount of pyridine adsorbed on Lewis acid sites determined by Py-IR.
The amount of pyridine adsorbed on Brønsted acid sites determined by Py-IR.
The total amount of pyridine adsorbed on Lewis and Brønsted acid sites determined by Py-IR.
The amount of desorbed NH3 determined by NH3-TPD.

G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
359
Remarkably, Fig. 10 shows that the regulation of the facets of
TiO2 support is an efficient approach for Ru to break through these
limitations. As illustrated in Fig. 10, the desired intermediate cyclo-
hexene and the by-product cyclohexane are the only detected
products. The benzene content decreased and the cyclohexane
content increased monotonically as the reaction proceeded. For
the cyclohexene content, a maximum appeared at a certain reac-
tion time, which conformed with the known feature of the consec-
utive reaction. Among the four catalysts, the Ru/TiO2-433 catalyst
exhibited the lowest cyclohexene yield of 21%, while the reaction
proceeded the fastest that only 45 min was required to convert
90% of benzene. A dramatical increment for the cyclohexene yield
appeared on the Ru/TiO2-443 catalyst (52%). On the Ru/TiO2-453
catalyst, cyclohexene amounted to the highest yield of 54%, and
then declined to 40% on the Ru/TiO2-473 catalyst. Additionally,
the curve of the cyclohexene selectivity against the benzene con-
version of the optimal Ru/TiO2-453 catalyst at the third run per-
fectly mimics the first run (Fig. S7), manifesting the good
stability of the catalyst.
Table 4 lists the catalytic results of the selective hydrogenation
of benzene on the Ru/TiO2 catalysts. The r0 decreased gradually
with the elevation of the hydrothermal temperature of TiO2.
According to the r0 and the dispersion data in Table 1, the TOFs
of benzene were calculated. It is observed that the TOFs showed
the same decreased tendency with the r0. The highest TOF of the
Ru/TiO2-433 catalyst was close to that of the Ru/anatase catalyst
we reported previously [39].
As the particle size and chemical state of the Ru NPs on the
Ru/TiO2 catalysts were similar, it is reasonable to exclude the
possibility that the Ru NPs caused the difference in the TOFs.
100
100
b
a
80
60
40
20
0
80
60
40
20
0
0
10
20
30
40
50
0
10
20
30
40
50
Reaction Time (min)
Reaction Time (min)
100
80
60
40
20
0
100
80
60
40
20
0
d
c
0
10
20
30
40
50
60
0
20
40
60
80
100
Reaction Time (min)
Reaction Time (min)
Fig. 10. The time courses of benzene hydrogenation over the (a) Ru/TiO2-433, (b) Ru/TiO2-443, (c) Ru/TiO2-453, and (d) Ru/TiO2-473 catalysts. Reaction conditions: 0.5 g of
catalyst, 25 ml of benzene, 50 ml of H2O, 5.0 g of ZnSO4Á7H2O, temperature of 413 K, H2 pressure of 5.0 MPa, and stirring rate of 1200 rpm. (j) benzene, (d) cyclohexene, and
(▲) cyclohexane.
Table 4
Results of the selective hydrogenation of benzene over the Ru/TiO2 catalysts.a
tb (min)
Conv.b (%)
SCHE (%)
YCHE (%)
S0 (%)
r0
TOF (minÀ1
)
b
b
c
d
Catalyst
Ru/TiO2-433
Ru/TiO2-443
Ru/TiO2-453
Ru/TiO2-473
20
35
40
40
69
84
84
64
31
62
64
63
21
52
54
40
61
87
93
77
26.5
19.7
15.8
11.7
98.9
76.9
58.0
44.1
a
b
c
Reaction conditions: 0.5 g of catalyst, 25 ml of benzene, 50 ml of H2O, 5.0 g of ZnSO4Á7H2O, temperature of 413 K, H2 pressure of 5.0 MPa, and stirring rate of 1200 rpm.
Values recorded at the maximum yield of cyclohexene.
Initial selectivity towards cyclohexene.
Initial weight specific activity, unit in mmolC6H6
d
À1
g
minÀ1
.
cat

360
G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
We previously identified that the additional hydrogenation
between adsorbed benzene molecules and spillover hydrogen
(Hso) from metal sites was existed on the acid sites of supports of
the Ru/B-ZrO2 and Ru/AlOOH–SiO2 catalysts, which improved the
TOF of benzene [12,23]. For the anatase TiO2 supported-Ru10 clus-
ters, Chen et al. found that the adsorbed H2 molecules on the metal
cluster became thermodynamically favorable to dissociate and
spill onto the support surface at high hydrogen coverage [40].
Analogously, in the light of the close contact between Ru NPs
and TiO2 surfaces on the Ru/TiO2 catalysts revealed by HRTEM
and the high H2 pressure (5.0 MPa) in reaction, it is rational to infer
that in benzene hydrogenation over the Ru/TiO2 catalysts, the H2
molecules adsorbed on the Ru NPs would dissociate and spill onto
the TiO2 surfaces, generating the Hso species.
On the other hand, we investigated the adsorption capacity of
benzene on the Ru/TiO2 catalysts by TPD, as shown in Fig. 11. A
similar three peak-profile was exhibited for all four catalysts. The
peak at 364 K was attributed to the residual physisorbed benzene
as the desorption temperature approached to the boiling point of
benzene. The peaks at 515 and 819 K are assigned to chemisorbed
benzene. The integral peak intensity of the chemisorbed benzene
decreased with the elevation of the hydrothermal temperature,
which is consistent with the decrement in the amount of acid sites
(Table 3). This variation should be originated from the adsorption
of benzene on the acid sites of TiO2 supports since Lin et al. [41]
and Simon et al. [42] demonstrated that aside from the adsorption
of benzene on the Ru NPs, benzene can also adsorb on the Lewis
and Brønsted acid sites of supports. In addition, when plotting
the natural logarithm of benzene concentration with the t derived
from Fig. 10, a good linear relationship was emerged (Fig. S8),
revealing that the reaction is first order relative to benzene. There-
fore, the higher the hydrothermal temperature of TiO2 is, the less
the benzene adsorbed on the acid sites reacting with Hso that is
also adsorbed on TiO2, and as a result, a lower TOF. By plotting
the TOFs with the nPy in Table 3, an excellent linearity appeared
in Fig. 12A, further verifying the acid sites-induced activity
enhancement for the Ru/TiO2 catalysts.
515
364
819
d
c
b
3.6. Implications of the TiO2 facets on the cyclohexene selectivity
As the hydrothermal temperature of TiO2 elevated, an firstly
increased and then declined tendency emerged for the S0 (Table 4).
Meanwhile, it is delighted that supporting the Ru NPs on anatase
TiO2 with the highest degree of exposed {1 0 0} facets significantly
promoted the cyclohexene selectivity: the S0 on the Ru/TiO2-453
catalyst attained to 93%, which is higher than that of our previously
reported Ru/anatase catalyst (82%) [39], as well as the Ru/Al2O3,
Ru/SiO2, Ru/CoO, Ru/NiO, Ru/CuO, and Ru/ZnO catalysts
a
300 400 500 600 700 800 900
Temperature (K)
Fig. 11. The TPD profiles of benzene on the (a) Ru/TiO2-433, (b) Ru/TiO2-443, (c) Ru/
TiO2-453, and (d) Ru/TiO2-473 catalysts.
k1/k2 (10-2 l mol-1)
3.0
3.5
4.0
4.5
95
90
85
80
75
70
65
60
100
Ru/TiO2-453
Ru/TiO2-433
90
Ru/TiO2-443
80
Ru/TiO2-443
Ru/TiO2-473
70
60
Ru/TiO2-453
50
Ru/TiO2-473
40
Ru/TiO2-433
0.5 0.6
OOH/Ti Molar Ratio
300
350
400
450
0.4
0.7
0.8
nPy ( mol g-1)
Fig. 12. (A) Correlation of TOF and nPy, and (B) correlation of S0 and the OOH/Ti molar ratio and k1/k2 ratio on the Ru/TiO2 catalysts.

G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
361
(Table S1). According to our previous work [23], the acid sites on
the support of the Ru/B-ZrO2 catalyst affected the cyclohexene
selectivity by altering the rates of the two hydrogenation steps
shown in Scheme 2. To examine the validity of this viewpoint on
the Ru/TiO2 catalysts, the rate constants of these two steps (k1
and k2) were calculated from the slopes of the linearities between
the natural logarithm of benzene concentration (ln[C6H6]) against t
as well as the cyclohexane concentration ([C6H12]) against t based
on the reaction data shown in Fig. 10 and the rate equations raised
tivity towards cyclohexene is higher. A well linear relationship was
established when correlating the S0 and OOH/Ti surface ratio origi-
nated from the O1s spectra (Fig. 12B), further manifesting the cru-
cial role of the hydrophilicity of the Ru/TiO2 catalysts in benzene
selective hydrogenation. In agreement with this explanation, Peng
et al. found that Ru supported on ZrO2 heterophase structure
nanocrystal consisting of both monoclinic and tetragonal phases
was more efficient that which on single-phase ZrO2, attributing
to the higher concentration of the surface OH groups on the former
catalyst [46]. Spod et al. detected the formation of Zn(OH)À3 during
benzene selective hydrogenation over the Ru/La2O3-ZnO catalyst,
which improved the hydrophilicity of the catalyst and conse-
quently, promoted the cyclohexene yield [47]. Yu et al. observed
a significant increment in cyclohexene selectivity of the Ru@TiO2
catalyst when comparing with that of bare Ru, owing to the sup-
pressed diffusion of cyclohexene onto Ru particles by the hydro-
philic coating film [48]. Wang et al. discovered that more
structural H2O and surface OH groups were existed on the RuB/Al2-
by Liu et al. [43], that is, ln[C6H6] = ln[C6H6]0 À k1t and [C6H12
]
= k2t, in which [C6H6]0 is initial benzene concentration. As summa-
rized in Table 5, both k1 and k2 decreased with the elevation of the
hydrothermal temperature, while the degrees of the decrements
were different. The evolutionary process of the k1/k2 ratio matched
with that of the S0, evidencing the validity of the above acid sites-
selectivity relationship. By plotting the S0 against the k1/k2 values,
an excellent linearity appeared in Fig. 12B, which finely proves the
significance of the acid sites on the enhancement of the cyclohex-
ene selectivity by modifying the kinetics of benzene
hydrogenation.
O3ÁxH2O catalyst than on the RuB/
c-Al2O3 catalyst, which
improved the hydrophilicity and accordingly the cyclohexene yield
On the other hand, as discussed above, aside from the implica-
tion of the TiO2 facets on the amount of acid sites, the amount of
hydroxyl groups was also altered by the exposed degree of TiO2
{1 0 0} facets, and as a result, the hydrophilicity of the Ru/TiO2 cat-
alysts was changed. For the benzene selective hydrogenation on
Ru-based catalysts, Struijk et al. corroborated that the H2O layer
formed on the catalyst surface was vital for the improvement of
the cyclohexene selectivity by (i) occupying the active sites where
cyclohexane is preferentially generated, and (ii) promoting the
desorption of cyclohexene from the catalyst surface and inhibiting
its readsorption [44,45]. Therefore, as illustrated in Scheme 3, the
catalyst with more surface OH groups rendered the higher amount
of adsorbed H2O in benzene hydrogenation, and thereby the selec-
[49].
In addition to the role of the OH groups on the TiO2 {1 0 0}
facets in stabilizing the stagnant H2O layer on the Ru/TiO2 cata-
lysts, we investigated the direct interaction of cyclohexene with
the OH groups by cyclohexene TPD, employing the Ru/TiO2-433
and Ru/TiO2-453 catalysts as representatives. As shown in
Fig. 13, four desorption peaks at 366, 511, 539, and 825 K appeared
on the Ru/TiO2-433 catalyst, while only three peaks at 366, 529,
and 825 K were existed on the Ru/TiO2-453 catalyst. The peak at
366 K can be assigned to the residual physisorbed cyclohexene
since the desorption temperature is close to the boiling point of
cyclohexene, and other peaks were assigned to chemisorbed cyclo-
hexene. One less desorption peak on the Ru/TiO2-453 catalyst
demonstrated that the surface OH groups decreased the type of
adsorption sites. Simultaneously, although the intensity of the
peak at 825 K slightly increased for the Ru/TiO2-453 catalyst, the
peak at 529 K was dramatically attenuated, revealing that the OH
groups can also block the adsorption sites for cyclohexene. As men-
tioned above, first-principles calculations testified that the hydro-
genation energy barrier of benzene to cyclohexene is higher than
Table 5
The rate constants for the hydrogenation of benzene to cyclohexene (k1) and
cyclohexene to cyclohexane (k2) over the Ru/TiO2 catalysts fitted from reaction data
in Fig. 10 according to the kinetic equations raised by Ref. [43].
Catalyst
k1
k2
k1/k2
(10À2 minÀ1
)
(10À2 mol lÀ1 minÀ1
)
(10À2 l molÀ1
)
Ru/TiO2-433
Ru/TiO2-443
Ru/TiO2-453
Ru/TiO2-473
5.7
4.7
4.3
2.6
193
107
95
3.0
4.4
4.5
3.9
67
529
825
366
b
a
511
539
300 400 500 600 700 800 900
Temperature (K)
Scheme 3. Schematic illustration of the mechanism of benzene selective hydro-
Fig. 13. The TPD profiles of cyclohexene on the (a) Ru/TiO2-433 and (b) Ru/TiO2-
genation over the Ru/TiO2 catalysts.
453 catalysts.

362
G. Zhou et al. / Journal of Catalysis 369 (2019) 352–362
cyclohexene to cyclohexane on the Ru(0 0 0 1) surface [38]. This
signifies that cyclohexene will be facilely saturated once it was
formed on the catalyst. Therefore, the decreased type of adsorption
sites and the suppressed adsorption of cyclohexene on the Ru/
TiO2-453 catalyst by more surface OH groups derived from higher
exposed degree of TiO2 {1 0 0} facets may be additionally critical
factors for the remarkable selectivity enhancement in benzene
selective hydrogenation.
References
[1] J. Pan, G. Liu, G.Q. Lu, H.M. Cheng, Angew. Chem. Int. Ed. 50 (2011) 2133–2137.
[2] G. Longoni, R.L.P. Cabrera, S. Polizzi, M. D’Arienzo, C.M. Mari, Y. Cui, R. Ruffo,
Nano Lett. 17 (2017) 992–1000.
[3] M. Chen, J. Ma, B. Zhang, F. Wang, Y. Li, C. Zhang, H. He, Appl. Catal. B: Environ.
223 (2018) 209–215.
[4] G. Liu, H.G. Yang, J. Pan, Y.Q. Yang, G.Q. Lu, H.M. Cheng, Chem. Rev. 114 (2014)
9559–9612.
[5] L. Liu, Y. Jiang, H. Zhao, J. Chen, J. Cheng, K. Yang, Y. Li, ACS Catal. 6 (2016)
1097–1108.
[6] J. Li, D. Xu, Chem. Commun. 46 (2010) 2301–2303.
[7] B. Zhang, F. Wei, Q. Wu, L. Piao, M. Liu, Z. Jin, J. Phys. Chem. C 119 (2015) 6094–
6100.
4. Conclusions
[8] A. Beltrán, J.R. Sambrano, M. Calatayud, F.R. Sensato, J. Andrés, Surf. Sci. 490
(2001) 116–124.
[9] Y. Sui, S. Liu, T. Li, Q. Liu, T. Jiang, Y. Guo, J.L. Luo, J. Catal. 353 (2017) 250–255.
[10] M. Cao, Z. Tang, Q. Liu, Y. Xu, M. Chen, H. Lin, Y. Li, E. Gross, Q. Zhang, Nano Lett.
16 (2016) 5298–5302.
[11] C. Liu, X.G. Han, S.F. Xie, Q. Kuang, X. Wang, M.S. Jin, Z.X. Xie, L.S. Zheng, Chem.
Asian J. 8 (2013) 282–289.
[12] G. Zhou, Y. Dong, L. Jiang, D. He, Y. Yang, X. Zhou, Catal. Sci. Technol. 8 (2018)
1435–1446.
[13] S.J. Tauster, S.C. Fung, R.L. Garten, J. Am. Chem. Soc. 100 (1978) 170–175.
[14] C. Wen, A.Y. Yin, Y.Y. Cui, X.L. Yang, W.L. Dai, K.N. Fan, Appl. Catal. A: Gen. 458
(2013) 82–89.
[15] L.Y. Zou, Y.Y. Cui, W.L. Dai, Chin. J. Chem. 32 (2014) 257–262.
[16] B. Wang, C. Wen, Y.Y. Cui, X. Chen, Y. Dong, W.L. Dai, RSC Adv. 5 (2015) 29040–
29047.
[17] X. Xue, J. Liu, D. Rao, S. Xu, W. Bing, B. Wang, S. He, M. Wei, Catal. Sci. Technol.
7 (2017) 650–657.
[18] G. Zhou, H. Wang, J. Tian, Y. Pei, K. Fan, M. Qiao, B. Sun, B. Zong, ChemCatChem
10 (2018) 1184–1191.
[19] Z. Peng, X. Liu, H. Lin, Z. Wang, Z. Li, B. Li, Z. Liu, S. Liu, J. Mater. Chem. A 4
(2016) 17694–17703.
[20] D. Rao, X. Xue, G. Cui, S. He, M. Xu, W. Bing, S. Shi, M. Wei, Catal. Sci. Technol. 8
(2018) 236–243.
[21] P. Zhang, T. Wu, T. Jiang, W. Wang, H. Liu, H. Fan, Z. Zhang, B. Han, Green Chem.
15 (2013) 152–159.
[22] S.C. Hu, Y.W. Chen, Ind. Eng. Chem. Res. 36 (1997) 5153–5159.
[23] G. Zhou, Y. Pei, Z. Jiang, K. Fan, M. Qiao, B. Sun, B. Zong, J. Catal. 311 (2014)
393–403.
This work presents a comprehensive study on the support
facet effect of the Ru/TiO2 catalysts for benzene selective hydro-
genation to cyclohexene. Anatase TiO2 nanorods exposed differ-
ent degrees of {1 0 0} facets were successfully synthesized by
hydrothermal route at different temperatures, in which 453 K
is identified to be the optimal temperature for exposing the most
{1 0 0} facets. It is found that the increment of the exposed
degree of TiO2 {1 0 0} facets imposed no impact on the size,
microstructure, and chemical state of the Ru NPs, but introduced
more surface OH groups on the Ru/TiO2 catalyst, resulting in
improved hydrophilicity. Besides, the amount of acid sites on
the Ru/TiO2 catalysts decreased with the elevation of the
hydrothermal temperature of supports. In benzene selective
hydrogenation, the TOFs of benzene decreased gradually with
the elevation of the hydrothermal temperature, while the cyclo-
hexene selectivity displayed a volcanic-type evolution, passing
through the highest initial selectivity of 93% and the maximum
yield of 54% on the Ru/TiO2-453 catalyst. The decrement in TOFs
was attributed to the decreased amounts of acid sites on TiO2
owing to that benzene can be additionally hydrogenated on the
acid sites by Hso. For the S0 of cyclohexene, a positive linearity
was established between which and the amount of OH groups,
supporting the crucial character of the H2O layer on the catalysts
in determining the cyclohexene selectivity. In addition, we found
that the surface OH groups can essentially suppress the cyclo-
hexene adsorption by decreasing the type of adsorption sites
and blocking the chemisorption sites, which is identified to be
another key factor for the selectivity enhancement by support
facet regulation. The intriguing support facet-performance rela-
tionship of the Ru/TiO2 catalysts demonstrated in this work pro-
vides a new way for the design of more efficient Ru catalysts for
benzene selective hydrogenation by regulating the facets of sup-
port other than TiO2.
[24] Y. Pei, G. Zhou, N. Luan, B. Zong, M. Qiao, F. Tao, Chem. Soc. Rev. 41 (2012)
8140–8162.
[25] J.F. Moulder, W.F. Stickle, P.E. Sobol, K.D. Bomben, Handbook of X-ray
Photoelectron Spectroscopy, Perkin-Elmer, Minnesota, 1992.
[26] H. Ishikawa, J.N. Kondo, K. Domen, J. Phys. Chem. B 103 (1999) 3229–3234.
[27] Y. Zhu, L. Tian, Z. Jiang, Y. Pei, S. Xie, M. Qiao, K. Fan, J. Catal. 281 (2011) 106–
118.
[28] K. Lan, Y. Liu, W. Zhang, Y. Liu, A. Elzatahry, R. Wang, Y. Xia, D. Al-Dhayan, N.
Zheng, D. Zhao, J. Am. Chem. Soc. 140 (2018) 4135–4143.
[29] M.I. Zaki, M.A. Hasan, F.A. Al-Sagheer, L. Pasupulety, Colloids Surf. A 190 (2001)
261–274.
[30] G. Zhou, Y. Dong, D. He, Appl. Surf. Sci. 456 (2018) 1004–1013.
[31] P.A. Jacobs, C.F. Heylen, J. Catal. 34 (1974) 267–274.
[32] I.S. Pieta, M. Ishaq, R.P.K. Wells, J.A. Anderson, Appl. Catal. A: Gen. 390 (2010)
127–134.
[33] F. Liu, T. Wang, Y. Zheng, J. Wang, J. Catal. 355 (2017) 17–25.
[34] C. Arrouvel, M. Digne, M. Breysse, H. Toulhoat, P. Raybaud, J. Catal. 222 (2004)
152–166.
Acknowledgements
[35] W.H. Chen, H.H. Ko, A. Sakthivel, S.J. Huang, S.H. Liu, A.Y. Lo, T.C. Tsai, S.B. Liu,
Catal. Today 116 (2006) 111–120.
[36] L. Foppa, J. Dupont, Chem. Soc. Rev. 44 (2015) 1886–1897.
[37] H. Sun, Z. Chen, C. Li, L. Chen, Z. Peng, Z. Liu, S. Liu, Catalysts 8 (2018) 104–116.
[38] C. Fan, Y.A. Zhu, X.G. Zhou, Z.P. Liu, Catal. Today 160 (2011) 234–241.
[39] G. Zhou, R. Dou, H. Bi, S. Xie, Y. Pei, K. Fan, M. Qiao, B. Sun, B. Zong, J. Catal. 332
(2015) 119–126.
[40] H.Y.T. Chen, S. Tosoni, G. Pacchioni, ACS Catal. 5 (2015) 5486–5495.
[41] S.D. Lin, M.A. Vannice, J. Catal. 143 (1993) 539–553.
[42] L.J. Simon, J.G. van Ommen, A. Jentys, J.A. Lercher, Catal. Today 73 (2002) 105–
112.
[43] S.C. Liu, Y.Q. Guo, X.L. Yang, Y.L. Ji, G. Luo, Chin. J. Catal. 24 (2003) 42–46.
[44] J. Struijk, J.J.F. Scholten, Appl. Catal. A: Gen. 82 (1992) 277–287.
[45] J. Struijk, M. Angremond, W.J.M. Regt, J.J.F. Scholten, Appl. Catal. A: Gen. 83
(1992) 263–295.
This work was supported by National Natural Science Founda-
tion of China (No. 21703024), Natural Science Foundation of
Chongqing, China (No. cstc2018jcyjAX0622, cstc2016jcyjA0392),
and the Scientific and Technological Research Program of Chongq-
ing Municipal Education Commission (No. KJ1500305).
Competing interests statement
There are no competing interests to declare.
[46] Z. Peng, X. Liu, S. Li, Z. Li, B. Li, Z. Liu, S. Liu, Sci. Rep. 7 (2017) 39847–39857.
[47] H. Spod, M. Lucas, P. Claus, ChemCatChem 8 (2016) 2659–2666.
[48] X.L. Yu, Y. Li, S.M. Xin, P.Q. Yuan, W.K. Yuan, Ind. Eng. Chem. Res. 57 (2018)
1961–1967.
[49] J.Q. Wang, P.J. Guo, S.R. Yan, M.H. Qiao, H.X. Li, K.N. Fan, J. Mol. Catal. A: Chem.
222 (2004) 229–234.
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jcat.2018.11.032.