Tuning of active sites in M/TiO2 for photocatalytic cyanation of olefins with high regioselectivity

Article abstract of DOI:10.1016/j.apcata.2020.117787

The detailed structure of active sites plays an important role for the determination of catalytic performance. Herein, catalytic active sites of M/TiO₂ for photocatalytic cyanation of olefins are tuned by delicate manipulation of the metal kinds, metal loading amount and pretreatment processes. It was found that the 0.1% Pt/P25 catalyst reduced at 300 °C possessed high metal dispersion and suitable oxygen defects on TiO₂, and thus exhibited the best catalytic performance with high specific speed of time yield and high selectivity. A wide scope of olefin substrates could be converted to the corresponding nitriles with high atom efficiency and anti-Markovnikov regioselectivity under mild conditions for the optimized Pt/P25 catalyst. The reaction mechanism based on radical coupling of acetonitrile and olefins was also discussed. This approach offers an environmental-friendly platform for the selective activation of C-H bonds of acetonitrile and will bring potential applications for hydrofunctionalization of olefins.

Full text of DOI:10.1016/j.apcata.2020.117787

AppliedCatalysisA,General604(2020)117787
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Tuning of active sites in M/TiO2 for photocatalytic cyanation of olefins with
high regioselectivity
Shuyi Zhanga,b, Bo Wua,b, Min Huanga,c, Jingxian Baoa, Liangshu Zhonga,c,*, Yuhan Suna,c,
*
a CAS Key Laboratory of Low-Carbon Conversion Science and Engineering, Shanghai Advanced Research Institute, Chinese Academy of Sciences, Shanghai 201203, China
b University of the Chinese Academy of Sciences, Beijing 100049, China
c School of Physical Science and Technology, ShanghaiTech University, Shanghai201210, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
The detailed structure of active sites plays an important role for the determination of catalytic performance.
Herein, catalytic active sites of M/TiO2 for photocatalytic cyanation of olefins are tuned by delicate manip-
ulation of the metal kinds, metal loading amount and pretreatment processes. It was found that the 0.1% Pt/P25
catalyst reduced at 300 °C possessed high metal dispersion and suitable oxygen defects on TiO2, and thus ex-
hibited the best catalytic performance with high specific speed of time yield and high selectivity. A wide scope of
olefin substrates could be converted to the corresponding nitriles with high atom efficiency and anti-
Markovnikov regioselectivity under mild conditions for the optimized Pt/P25 catalyst. The reaction mechanism
based on radical coupling of acetonitrile and olefins was also discussed. This approach offers an environmental-
friendly platform for the selective activation of C-H bonds of acetonitrile and will bring potential applications for
hydrofunctionalization of olefins.
Photocatalysis
Nitriles
Cyanation
Olefins
Anti-Markovnikov
1. Introduction
produce nitriles by coupling olefins and HCN with nickel homogeneous
catalysts [1,13]. For addition reaction, it is challenging to avoid the use
Nitriles are widely used in the fields of pharmaceuticals, agro-
chemicals and natural products. The nitrile group is regarded as a va-
luable functional group for being a universal intermediate of diverse
organic chemicals, such as carboxylic acid, amide and aldehyde/ami-
nomethyl [1–3]. Nitriles can be synthesized by substitution reaction of
organic halide and cyanide, such as KCN, CuCN or NaCN (Rosenmuncd-
von Braun reaction) [4]. Sandmeyer reaction, the diazotization of
amide and substitution of cyano, is also employed (Scheme 1a) for ni-
triles production [5]. However, both of the above reactions require pre-
activated compounds (-X, -NH2), which leads to a complicated trans-
formation of C-H bond. To realize the direct activation of C-H bond,
different organic substrates and cyano sources have been well-docu-
mented via homogeneous catalysts. The alternative CN units can be
obtained directly from organics with CN group, such as trimethylsilyl
cyanide (TMSCN) [6,7], aryl(cyano) iodonium triflates [8] and PhN(Ts)
(CN) [9,10] or indirectly from additives containing N element, such as
CH3NO2 [11], N,N-dimethylform-amide(DMF) and NH3 [12]. Except
the substitution reaction, addition reaction as an alternative route to
obtain alkyl nitriles could be performed at milder reaction conditions
(Scheme 1b). Nowadays, DuPont Company still uses this method to
of toxic, explosive HCN and further improve the target product re-
gioselectivity. Among the various types of cyano alternatives, the usage
of low carbon-chain nitriles as CN source attracts much attention
[14,15]. And to obtain more valuable anti-Markovnikov products, it is
preferable to apply an α-cyano radical initiated reaction through the
activation of α-C(sp3)-H bond of nitriles via SET (single electron
transfer) mechanism [16–19]. In addition, the transfer of CN units from
sacrificial nitriles to target olefins driven by the gas evolution or strain
release was also reported with Ni-based catalysts [20]. Photocatalysis
has also been applied to this reaction due to its great advantages in
generating radicals [21] (Scheme 1c). Sonawane [22] generated cy-
naomethyl radicals by photoinduced decomposition of hydrogen per-
oxide, but alcohols as by-product were inevitable. Yamashita [23] de-
veloped an improved procedure to avoid the addition of peroxide and
the formation of by-products. However, benzophenone and its deriva-
tives were required as a hydrogen abstractor of nitriles upon light ir-
radiation.
Obviously, the use of CN unit from HCN, KCN and TMSCN en-
counters the problems of high toxicity or cost. Alkyl nitriles as an al-
ternative feedstock would be a better choice. To obtain high-carbon
⁎ Corresponding authors at: CAS Key Laboratory of Low-Carbon Conversion Science and Engineering, Shanghai Advanced Research Institute, Chinese Academy of
Sciences, Shanghai 201203, China.
E-mail addresses: zhongls@sari.ac.cn (L. Zhong), sunyh@sari.ac.cn (Y. Sun).
https://doi.org/10.1016/j.apcata.2020.117787
Received 19 June 2020; Received in revised form 30 July 2020; Accepted 13 August 2020
Availableonline15August2020
0926-860X/©2020ElsevierB.V.Allrightsreserved.

S. Zhang, et al.
AppliedCatalysisA,General604(2020)117787
Scheme 1. Reaction pathways for synthesis of nitriles. (a) Traditional reaction pathway to nitriles via substitution reaction; (b) Traditional reaction pathway to
nitriles via addition reaction; (c) Photocatalytic pathway to nitriles via addition reaction.
2

S. Zhang, et al.
AppliedCatalysisA,General604(2020)117787
nitriles with high anti-Markovnikov regioselectivity, α-cyano radical is
an important intermediate. However, high temperature and extra ad-
ditives, such as oxidants, co-oxidants, SET reagents and specific sol-
vents, are required in thermal catalysis. Photocatalysis has a natural
advantage in producing cynao radicals, but avoiding the formation of
by-product and addition of extra additives are still great challenges.
In our previous research, we found the TiO2 with noble metal
loading could realize the hydrofunctionalization of olefins via acti-
vating the C-H bond of methanol, acetonitrile, acetic acid, etc [24].
High atom efficiency and anti-Markovnikov regioselectivity were
achieved for a wide scope of substrates. However, the relationship of
structure and catalytic performance of M/TiO2 for hydro-
functionalization of olefins had not been fully explored. It is still of
great importance to carefully tune the catalytic active sites of M/TiO2
for the optimization of catalytic performance. Obviously, each step in
the catalyst preparation process was critical for the detailed structure of
active sites and the corresponding catalytic performance. Herein, the
effects of metal kinds, loading amount and pretreatment processes on
structure-performance of M/TiO2 were investigated. Acetonitrile was
chosen as solvent and raw material providing CN unit for the coupling
reaction under mild conditions without any extra additives or by-pro-
ducts.
Microscope (Thermo Scientific™ DXR™2xi) with 535 nm excitation laser
wavelength.
The X-ray photoelectron spectroscopy (XPS) was conducted on a
Thermo Fisher Scientific K-alpha spectrometer using Al Kα (hν =
1486.6 eV) beam as excitation source. The binding energies were ca-
librated using C 1s peak at 284.8 eV.
2.3. Catalyst evaluation
Photocatalytic coupling of olefins and acetonitrile was conducted in
a sealed 100 mL photocatalytic stainless steel reactor (CEL-HPR100 T,
Beijing Aulight) with a sapphire window (40 mm for diameter). For the
test of photocatalytic performance with time-on-stream, quartz kettle
was used to realize continuous sampling. 300 W Xe lamp (CEL-
HXUV300, Beijing Aulight) with 100 mW⋅ cm-1 light intensity and 200-
400 nm wavelength was chosen as light source. In a typical reaction,
50 mg photocatalyst was suspended in the mixed solution of 50 mL
acetonitrile, 1 mmol olefins and 0.44 mmol dodecane (internal stan-
dard). After stirring for 10 min, 1.5 mL solution was aspirated, filtered
and then stored in the refrigerator as pre-reaction sample. Afterwards,
the reactor was sealed and filled with Ar to ensure anaerobic condition.
Then, the reaction slurry was heated to 60 °C and the light was turned
on for 15 h with continuous stirring. After cooled down to room tem-
perature, 1.5 mL solution was collected in the same way and denoted as
post-reaction sample. All the samples were analyzed by GC-MS (Agilent
2890A).
2. Experimental
2.1. Catalyst preparation
Moles of target product was calculated by the following equation：
Commercial P25 (TiO2) catalyst was purchased from ACROS ORD-
ANICS.
Moles of target product = (peak area of target product) ⋅ (relative
correction factor) / ((peak area of internal standard) ⋅ (moles of internal
standard))
Noble metal supported catalysts (M/P25) were prepared by im-
pregnation method. Typically, 5 g P25 was dispersed in 25 mL ultrapure
water (containing a certain amount of noble metal). After stirring at
room temperature for 1 h, the slurry was dried at 60 °C with vacuum
overnight. The obtained powder was ground homogeneously, and de-
noted as X% M/P25-raw (X was the weigh loading amount of noble
metal).
Speed of time yield (STY) was calculated by the following equa-
tion：
STY = (moles of target product) / ((mass of catalyst) ⋅ (reaction time))
In a typical reduction procedure, the raw material was reduced in
pure H2 (99.9%) flow of 100 mL·g-1 min-1 at different temperatures
(200, 300, 400 and 500 °C) for 3 h with a heating ramp of 1 °C min-1.
After cooled to room temperature, the catalyst was passivated in 1%
O2/Ar (99.9%) flow of 10 mL·g-1 min-1 for 1 h, and the obtained cata-
lysts were denoted as M/P25-200H2, M/P25-300H2, M/P25-400H2 and
M/P25-500H2, respectively. In a typical calcination procedure at
300 °C, all the steps were the same except for replacing H2 with air, and
the obtained catalyst was denoted as M/P25-300 cal. The M/P25-
300 cal-300H2 sample was obtained via the calcination and reduction
process.
Specific speed of time yield (STY) was calculated by the following
equation：
Specific STY = (moles of target product) / ((mass of noble metal) ⋅
(reaction time))
Turnover number (TON) was calculated by the following equation：
TON = (moles of target product) / (moles of noble metal)
Selectivity (%) was calculated by the following equation：
Selectivity (%) = (moles of target product) / (moles of all products)
2.2. Catalyst characterization
Anti-Markovnikov regioselectivity (%) was calculated by the fol-
lowing equation：
The transmission electron microscopy (TEM) and high-resolution
transmission electron microscopy (HRTEM) were employed on a JEM-
2100 F microscope.
Anti-Markovnikov regioselectivity (%) = (moles of product with anti-
Markovnikov) / (moles of products with anti-Markovnikov and not)
The H2 temperature-programmed reaction (H2-TPR) measurements
were performed on the AutoChem II 2920 equipped with an in-situ
mass spectrometry (MS) (CIRRUS 2). Samples were pretreated at 200 °C
in the He flow for 1 h. Then a gas flow of 10% H2/Ar (40 ml⋅ min-1) was
charged and the temperature was programmed up to 800 °C with a
constant ramp rate of 10 °C min-1.
3. Result and Discussion
3.1. Photocatalytic performances of M/P25 photocatalysts with different
metal kinds
The dispersion of Pt was determined on AutoChem II 2920 appa-
ratus. 100 mg sample was pretreated with He (40 ml⋅ min-1) at 200 °C
for 30 min, then cooled to 50 °C. After the baseline was stabilized, CO
pulses were injected until the peaks saturated.
Acetonitrile was chosen as the source of CN group for the hydro-
functionalization of olefins via photocatalysis using P25 as the model
catalyst (Table 1). Controlled experiments indicated that both light ir-
radiation and photocatalyst were essential for the coupling of acetoni-
trile and olefins as no product was detected in the absence of any of
The content of Pt in the TiO2 was measured by inductively coupled
plasma-optical emission spectroscopy (ICP-OES) on Perkin Elmer.
The Raman measurements were collected on a Raman imaging
3

S. Zhang, et al.
AppliedCatalysisA,General604(2020)117787
Table 1
Table 2
Catalytic performances of hydrofunctionalization of olefins to nitriles via pho-
tocatalytic coupling.
Catalytic performances of Pt/TiO2 with different Pt loading amount.
Entrya Metal amount STY (μmol·gcata-1·h-1
(wt %)
)
Pt dispersion
Specific STY
(%)
(mmol·gPt-1·h-1
)
Entrya Catalysts Metal
Substrates
Corresponding
nitriles yield
(μmol)
STY for nitriles
loading
(wt %)
(μmol·gcata-1·h-1
)
1
2
3
4
0.11
0.22
0.57
1.26
4.22
2.79
1.73
2.54
91.6
24.3
7.9
4.01
1.25
0.30
0.20
1b
2c
3
P25
–
1-decene
1-decene
1-decene
1-decene
1-decene
1-decene
1-decene
1-decene
1-nonene
1-octene
–
–
4.6
–
–
–
–
P25
–
0.08
0.20
–
0.11
0.25
–
Reaction condition: 1 mmol of 1-decene, 50 mg of photocatalyst in 50 mL of
acetonitrile. All Pt/TiO2 catalysts with different Pt amount were reduced at
300 °C. The reaction was conducted under argon atmosphere at 60 °C and Xe
lamp irradiation (200-400 nm) with light intensity of 100 mW·cm-2. The reac-
tion time was 15 h.
4
Pd/P25
Ru/P25
Au/P25
Ag/P25
Pt/P25
Pt/P25
Pt/P25
Pt/P25
Pt/P25
0.43
0.38
0.43
0.46
0.57
0.57
0.57
0.57
0.57
5
6
0.46
0.83
1.36
1.54
0.48
0.59
1.10
1.73
2.00
0.64
1.29
1.21
7
8
9
10
11
12
1- heptene 1.00
TEM images and size distributions of Pt/P25-300H2 samples with
various Pt content (0.1, 0.2, 0.5 and 1.0 wt%) were shown in Fig.1. For
the samples with low Pt loading (0.1, 0.2 wt %), small Pt nanoparticles
were observed with uniform particle size distribution (1-1.7 nm).
However, larger Pt nanoparticles were observed with wide size dis-
tribution when the loading was increased over 0.5 wt%, which were
consistent with the changing trend of Pt dispersion in Table 2. Based on
the results of specific STY and structure characterization, it was sug-
gested that Pt nanoparticles on TiO2 were the active sites and low Pt
content loading led to high metal dispersion and intrinsic activity.
styrene
0.91
aReaction condition: 1 mmol of olefin substrate, 50 mg of photocatalyst in
50 mL of acetonitrile. All the catalysts were reduced at 300 °C. The reaction was
conducted under argon atmosphere at 60 °C and Xe lamp irradiation (200-
400 nm) with light intensity of 100 mW·cm-2. b Entry 1: no light irradiation with
photocatalyst. c Entry 2: no photocatalyst with irradiation.
them (Entries 1,2). As the Fermi levels of noble metals and TiO2 are
distinct, electrons tend to migrate from high level to low level until they
are the same when noble metals are contacted with TiO2. This electron
migration results in the formation of Schottky barrier, which is bene-
ficial for the separation of photogenerated electrons and holes to en-
hance the photocatalytic activity [25,26]. In order to select the most
suitable metal for this reaction, different kinds of noble metals were
impregnated in the P25 with about 0.5% mass content and the corre-
sponding catalytic performance of coupling 1-decene and acetonitrile
were displayed in Entries 4-8. The catalytic performances of P25 were
significantly improved with the addition of noble metal (except Ru).
Among them, Pt exhibited the best activity. The yield and STY for the
corresponding nitriles reached 1.36 μmol and 1.73 μmol gcata-1 h-1, re-
spectively, which were almost 16 times of that of pure P25. In addition,
all of the studied catalysts showed 100% selectivity and anti-Markov-
nikov regioselectivity for the corresponding high-carbon nitriles. The
scope of substrate olefins was also investigated to demonstrate the
universality for this protocol. As shown in Entries 8-12, all of the stu-
died olefins could effectively couple with acetonitrile to the corre-
sponding high-carbon nitriles with 100% atom efficiency.
3.3. Effect of pretreatment atmosphere on structure of active sites and
catalytic performance
To figure out the effect of pretreatment atmosphere on the structure
of active sites and catalytic performance, 0.1% Pt/P25 was heated at
300 °C in H2 or air atmosphere to test for the coupling of acetonitrile
and 1-decene under identical reaction conditions. The reduced and
calcined catalysts exhibited no difference in the product selectivity.
However, the activity of H2 reduction was about 6 times of that with air
calcination. This indicated that the pretreatment of H2 reduction pro-
moted the coupling reaction (Fig.2a). In addition, better performance
was achieved when the calcined sample was treated with reduction
process, further confirming H2 reduction process was conducive to
higher activity.
Characterizations were conducted to study the structure changes
caused by atmosphere. Pt 4f XPS data of 0.1% Pt/P25 samples with
different treatment processes was shown in Fig.2b. The binding energy
at 74.5 and 77.7 eV of Pt/P25-raw and Pt/P25-300 cal were corre-
sponded to Pt (4f7/2) and Pt (4f5/2) of Pt4+, respectively [28]. Com-
pared with the strong peaks of Pt/P25-raw and 300 cal, the peak in-
tensity of the reduced samples (Pt/P25-300H2 and 300 cal-300H2) was
sharply decreased. This phenomenon was mainly due to the strong
metal-support interactions (SMSI). Under inert or reductive atmo-
sphere, mobilized TixOy could be formed and then encapsulated Pt
nanoparticles, leading to low Pt peak intensity in XPS characterization
[29,30]. Although the peak intensity of reduced samples was low, the
tendency of shifting to lower binding energy was obvious, indicating Pt
nanoparticles were reduced and partially covered by TiO2 layers in H2
process.
3.2. Effect of Pt loading amount on structure of active sites and catalytic
performance
Compared with other noble metals, Pt/P25 exhibited very pro-
mising catalytic performance for hydrofunctionalization of olefins to
nitriles via photocatalytic coupling, probably due to the large work
function of Pt [27]. Pt size and dispersion dramatically affected the
surface properties of active support for their distinct interaction
strength. Herein, the effect of Pt loading amount on catalytic perfor-
mance was investigated. As displayed in Table 2, P25 with 0.1% Pt
loading exhibited the highest STY of 4.22 μmol gcata-1 h-1. A steady de-
crease for STY was observed as the Pt content increased to 0.5% and
followed with a slight increase for the 1.0% sample. As the STY did not
increase linearly with the increase of Pt loading, the dispersion of Pt
may affect the catalytic performance. Hence, CO pulse was carried out
to determine the Pt dispersion as displayed in Table 2. It was found that
Pt dispersion and the specific STY sharply decreased as the loading
amount raised. The 0.1 % Pt/P25 with the highest metal dispersion
(91.6%) exhibited the highest specific STY of 4.01 mmol·gPt-1 h-1, while
the 1.0% Pt/P25 with the lowest dispersion (4.6%) showed the worst
specific STY of 0.20 mmol·gPt-1 h-1.
Raman spectra were shown in Fig. 2c to study the structure changes
of TiO2. Compared with the samples of Pt/P25-raw and Pt/P25-300 cal,
the Pt/P25-300H2 sample exhibited blue shift after reduction. The peak
around 142 cm-1 was regard to Ti-O stretching vibration mode [31,32].
According to Hooke’s law, the blue shift suggests a decrease of Ti-O
bond force constant, which may be caused by finite crystal size
(< 10 nm) or the presence of defects [33,34]. As the crystal size of all
samples was around 21.7 nm based on the XRD results calculated by
Scherrer's equation, it was thus suggested that the blue shift was due to
the generation of oxygen defects in TiO2 by H2 reduction.
4

S. Zhang, et al.
AppliedCatalysisA,General604(2020)117787
Fig. 1. TEM images and size distributions of Pt nanoparticles for the Pt/P25-300H2 samples with different Pt loading amount. (a) 0.1% Pt; (b) 0.2% Pt; (c) 0.5% Pt;
(d) 1.0% Pt; (e) Size distributions of Pt nanoparticles.
Given from the XPS and Raman results, H2 reduction process gen-
erated the reduced Pt and oxygen defects. The reduced Pt could better
capture photogenerated electrons, while oxygen defects could stabilize
the photogenerated holes. In this way, photogenic charge carriers could
be better separated, which was essential to enhance catalytic perfor-
mance. Therefore, the sample of 0.1% Pt/P25-300H2 achieved the
highest activity.
and the Pt/P25-300H2 sample with the highest metal dispersion
showed the best activity.
H2-TPR curves of 0.1% Pt/P25 samples were shown in Fig.3b. It had
been reported that platinum oxide on TiO2 surface could be reduced at
around 100 °C under reductive atmosphere [35]. In addition, due to
strong support metal interaction, the metallic nanoparticles promoted
the reduction of support due to the hydrogen spillover effect [36]. Thus,
we deduced that the strong sharp peak and the broad peak at about
100 °C and 270 °C could be assigned to the reduction of Pt and TiO2,
which could also be proved by the in-situ MS spectrum.
3.4. Effect of reduction temperature on structure of active sites and catalytic
performance
The reduced Pt nanoparticles are generally considered as the sites to
capture photogenerated electrons during photocatalysis process. In
order to study the reductive states of Pt, XPS of various samples under
different procedures was conducted as displayed in Fig.3c. It was im-
portant to note that all of the reduced samples went through passivation
(purged with 1%O2/Ar at room temperature for 1 h) to avoid sponta-
neous combustion of metallic Pt in contact with air, so the Pt was
partially oxidized. The Pt peak intensity of various reduced samples
decreased drastically with the increase of reduction temperature as the
surface Pt nanoparticles were gradually covered by TixOy layers due to
To further reveal the effect of reduction process on catalytic per-
formance, the influence of reduction temperature was discussed using
0.1% Pt/P25 as a model catalyst. The specific STY showed a volcanic
curve with the increase of reduction temperature, in which the sample
reduced at 300 °C rendered the best performance (Fig.3a). The metal
dispersion was also depicted to visualize the effect of reduction tem-
perature on metal dispersion, which exhibited a volcanic curve with the
increase of reduction temperature as well. It was found that the higher
metal dispersion always corresponded to the higher catalytic activity
Fig. 2. (a) Catalytic performances; (b) Pt 4f XPS data; (c) Raman spectra of 0.1% Pt/P25 with different treatment processes.
5

S. Zhang, et al.
AppliedCatalysisA,General604(2020)117787
Fig. 3. (a) Catalytic performances; (b) H2-TPR curves; (c) Pt 4f XPS data of 0.1% Pt/P25 with different reduction temperatures.
Table 3
Relative content of different O species based on XPS spectra of O 1s.
Catalysts
Percentage (%)
Olattice
Odefect
Ohydroxyls
Pt/P25-300 cal
Pt/P25-200H2
Pt/P25-300H2
Pt/P25-400H2
Pt/P25-500H2
90.7
89.1
83.6
77.9
75.2
0.0
9.3
3.0
6.0
9.0
10.2
7.9
10.4
13.0
14.6
the effect of strong metal support interaction. As the reduced Pt na-
noparticles were the active sites for reduction half reaction, the cov-
erage of Pt by TixOy layers would weaken the catalytic activity. The
effect of TiO2 defects in photocatalysis is still controversial. Some re-
searchers believe that defects are the recombination and trapping
centers for photogenetic charge carriers, while the others suggest de-
fects promote the charge separation [37]. In recent years, a new theory
sheds light on this debate, in which the defects are divided into surface
oxygen defects and bulk oxygen defects [38]. According to this theory,
the introduction of surface oxygen defects facilitates the separation of
charge carriers by stabilizing h+ while the formation of bulk defects
suppresses separation as recombination sites. In our study, O 1s XPS
characterization were conducted to prove the existence of surface
oxygen vacancies. The spectra were de-convoluted into three peaks
corresponding to lattice oxygen, oxygen defects and surface hydroxyls,
respectively. The relative content of different O species was displayed in
Table 3. Compared with the calcined sample, the reduced samples had
much more defects on the surface, which enhanced the separation of
photogenerated carriers. The percentage of surface oxygen defects in-
creased as the reduction temperature raised. When the reduction tem-
perature was lower than 300 °C, catalytic activity was consistent with
the percentage of surface oxygen defects. However, the activity sig-
nificantly decreased at higher reduction temperature, probably due to
the formation of bulk defects.
Fig. 4. Photocatalytic performance with time-on-stream over 0.1%Pt/P25-
300H2. Reaction condition: 1 mmol of 1-decene, 50 mg photocatalyst in 50 mL
of acetonitrile. The reaction was conducted under argon atmosphere at 60 °C
and Xe lamp irradiation (200-400 nm) with light intensity of 100 mW·cm-2.
catalyst. In the first few hours, no products were detected. With the
time-on-stream, olefins were gradually converted and the catalyst still
remained high activity after 42 h (Fig. 4). Obviously, this reaction had
an activation stage. According to XPS analysis, the Pt species of fresh
samples were partially oxidized. During the first few hours of irradia-
tion, the photogenerated electrons could be captured by Pt species [39].
Once the surface oxidized Pt species were reduced into Pt0 by photo-
generated electrons, high activity for photocatalytic cyanation of ole-
fins was achieved as metallic Pt was regarded as the active sites.
Radical intermediates always play an important role in photo-
catalysis. In our study, the existence of radical intermediates was con-
firmed as no product was detected after adding hydroquinone as radical
scavenger. In order to confirm the formation of cyanomethyl radical,
2,2,6,6-Tetramethylpiperidinooxy (TEMPO) was used as radical trap.
After irradiation for 15 h, TEMPO-CH2CN appeared and the target
product of dodecanenitrile was not observed, indicating the deproto-
nated MeCN radical was the intermediate. It was suggested that re-
duced Pt nanoparticles were the active sites for the reduction half re-
action, while the surface oxygen defects were the active sites for the
oxidation half reaction. A radical-mediated mechanism for the coupling
of acetonitrile and olefins over M/P25 was proposed as shown in
Scheme 2. In the photocatalytic cycle, electrons in semiconductor could
be excited from the valence band to the conduction band, leaving holes
in the valence band (step Ⅰ). Afterwards, the electron was further
transferred to the noble metal by the effect of Schottky barrier (step Ⅱ).
The hole stabilized by the oxygen defects could directly oxidize the
absorbed acetonitrile to produce cyanomethyl radical and hydrogen
As the degree of Pt and TiO2 reduction highly depended on the re-
duction conditions, H2 reduction process should be optimized. For the
sample reduced at 200 °C, the low activity may be due to the lack of
adequate oxygen defects on TiO2 (Table 3). For the sample reduced at 400
or 500 °C, one possible reason for their low activity was the decrease of Pt
active sites due to the coverage of TiO2 layers as shown in XPS results
(Fig. 3c). Another possible reason might be the formation of bulk defects
caused by harsh conditions. Therefore, reduction at 300 °C generated
reduced Pt and suitable surface defects, rendering the best efficiency.
3.5. Study of photocatalytic performance with time-on-stream and reaction
mechanism
The photocatalytic performance with time-on-stream was in-
vestigated using 1-decene as the substrate over 0.1%Pt/P25-300H2
6

S. Zhang, et al.
AppliedCatalysisA,General604(2020)117787
Scheme 2. Proposed photocatalytic mechanism for the photocatalytic coupling of acetonitrile and olefins.
proton (step Ⅲ). Subsequently, the desorbed free radical selectively
attacked the terminal position of olefins due to the steric effect and
radical stability. Therefore, the linear alkyl nitrile radical was generated
through C-C coupling (step Ⅳ). Finally, the alkyl nitrile radical was
reduced by an electron and combined with a proton on the surface of Pt
to produce high-carbon nitrile with 100% Anti-Markovnikov regios-
electivity (step Ⅴ). C-H bond is widely spread in various compounds,
and the ability of activating C-H bond has a great potential for the
activation and transformation of organics.
Declaration of Competing Interest
The authors report no declarations of interest.
ACKNOWLEDGMENTS
This work was supported by the Natural Science Foundation of
China (91945301), National Key R&D Program of China
(2017YFB0602202), Key Research Program of Frontier Sciences, CAS
(Grant No. QYZDB-SSW-SLH035), the “Transformational Technologies
for Clean Energy and Demonstration” and Strategic Priority Research
Program of the Chinese Academy of Sciences (Grant No.
XDA21020600) and the Youth Innovation Promotion Association of
CAS.
4. Conclusion
In summary, the catalytic performance of M/TiO2 for photocatalytic
cyanation of olefins was optimized by careful manipulation of active
sites. High atom efficiency and high regioselectivity were obtained, and
Pt was the most suitable metal for the M/TiO2 photocatalysts. The re-
duced Pt species and surface oxygen defects on TiO2 generated by re-
duction pretreatment process were regarded as the active sites for re-
duction and oxidation half reaction, respectively. Lowering the loading
amount increased Pt dispersion to enhance the catalytic activity and
maximize the usage of noble metal. Reduction treatment was essential
to generate active sites - reduced Pt and surface defects. In addition,
tuning the reduction conditions significantly changed the interaction
strength of Pt and support. With elaborate operation, a high specific
STY of 4.01 mmol·gPt-1 h-1 were achieved in 0.1%Pt/P25-300H2 due to
its high Pt dispersion and appropriate reduced states, which facilitated
the separation of photogenic charge carriers. Catalytic mechanism was
proposed based on the experimental results, which involved the radical-
initiated α-C-H cleavage and C-C bond formation caused by the pho-
togenerated electrons and holes. Our green strategy provides a pro-
mising route to synthesize nitriles from olefins, and could be expected
to have a promising potential for green synthesis of high-valued che-
micals derived from C-H bond activation.
References
[1] L. Bini, C. Müller, D. Vogt, ChemCatChem 2 (2010) 590–608.
[2] D.J. Segobia, A.F. Trasarti, C.R. Apesteguía, Chinese J. Catal 40 (2019) 1693–1703.
[3] Y.-C. Wong, K. Parthasarathy, C.-H. Cheng, Org. Lett. 12 (2010) 1736–1739.
[4] G.P. Ellis, T.M.R. Alexander, Chem. Rev. 87 (1987) 779–794.
[5] P. Anbarasan, T. Schareina, M. Beller, Chem. Soc. Rev. 40 (2011) 5049–5067.
[6] X. Wang, A. Studer, Angew. Chem. Int. Ed. 57 (2018) 11792–11796.
[7] J.B. McManus, D.A. Nicewicz, J. Am. Chem. Soc 139 (2017) 2880–2883.
[8] Z. Shu, W. Ji, X. Wang, Y. Zhou, Y. Zhang, J. Wang, Angew. Chem. Int. Ed. 53
(2014) 2186–2189.
[9] J. Li, L. Ackermann, Angew. Chem. Int. Ed. 54 (2015) 3635–3638.
[10] M. Chaitanya, P. Anbarasan, Org. Lett. 17 (2015) 3766–3769.
[11] X. Chen, X. Hao, C.E. Goodhue, J. Yu, J. Am. Chem. Soc. 128 (2006) 6790–6791.
[12] S. Ding, N. Jiao, J. Am. Chem. Soc 133 (2011) 12374–12377.
[13] C.A. Tolman, J. Chem. Educ. 63 (1986) 199–201.
[14] D. Wang, F. Wang, P. Chen, Z. Lin, G. Liu, Angew. Chem. Int. Ed. 56 (2017)
2054–2058.
[15] F. Wang, D. Wang, X. Wan, L. Wu, P. Chen, G. Liu, J. Am. Chem. Soc 138 (2016)
15547–15550.
[16] X.-Q. Chu, D. Ge, Z.-L. Shen, T.-P. Loh, ACS Catal. 8 (2017) 258–271.
[17] A. Bunescu, T.M. Ha, Q. Wang, J. Zhu, Angew. Chem. Int. Ed. 56 (2017)
10555–10558.
[18] T. Wu, X. Mu, G. Liu, Angew. Chem. Int. Ed. 50 (2011) 12578–12581.
[19] Z. Li, Y. Xiao, Z.-Q. Liu, Chem. Commun. 51 (2015) 9969–9971.
[20] P.Y. Xianjie Fang, Bill Morandi, Science 351 (2016) 832–836.
[21] S.Y. Zhang, J.X. Bao, B. Wu, L.S. Zhong, Y.H. Sun, Acta Phys.-Chim. Sin. 35 (2019)
923–939.
Justification for publication
Our work described herein is accordant with the scopes of Applied
Catalysis A: General and has not been published nor has been accepted
elsewhere. All of the authors listed have approved the manuscript that
is enclosed. We believe the paper may be of particular interest to the
readers of this journal.
[22] H.R. Sonawane, N.S. Bellur, V.G. Shah, J. Chem. Soc. Chem. Commun. 22 (1990)
1603–1605.
[23] T. Yamashita, J. Org. Chem. 61 (1996) 6438–6441.
[24] Y. Fan, S. Li, J. Bao, L. Shi, Y. Yang, F. Yu, P. Gao, H. Wang, L. Zhong, Y. Sun, Green
Chem 20 (2018) 3450–3456.
7

S. Zhang, et al.
AppliedCatalysisA,General604(2020)117787
[25] W.-N. Wang, W.-J. An, B. Ramalingam, S. Mukherjee, D.M. Niedzwiedzki,
S. Gangopadhyay, P. Biswas, J. Am. Chem. Soc. 134 (2012) 11276–11281.
[26] S. Xie, Y. Wang, Q. Zhang, W. Fan, W. Deng, Y. Wang, Chem. Commun. 49 (2013)
2451–2453.
Appl. Phys 16 (2016) 826–829.
[33] K.-R. Zhu, M.-S. Zhang, Q. Chen, Z. Yin, Phys. Lett. A 340 (2005) 220–227.
[34] X. Xin, T. Xu, J. Yin, L. Wang, C. Wang, Appl. Catal., B 176-177 (2015) 354–362.
[35] H. Liang, Y. Zhang, Y. Liu, J. Nat, Gas. Chem 17 (2008) 403–408.
[36] S.S. Kim, H.H. Lee, S.C. Hong, Appl. Catal., B 119-120 (2012) 100–108.
[37] K. Zhang, W. Zhou, L. Chi, X. Zhang, W. Hu, B. Jiang, K. Pan, G. Tian, Z. Jiang,
ChemSusChem 9 (2016) 2841–2848.
[27] A. Li, Q. Cao, G. Zhou, B.V.K.J. Schmidt, W. Zhu, X. Yuan, H. Huo, J. Gong,
M. Antonietti, Angew. Chem. Int. Ed. 58 (2019) 14549–14555.
[28] K. Taira, H. Einaga, Catal. Lett. 149 (2019) 965–973.
[29] S.J. Tauster, Acc. Chem. Res. 20 (1987) 389–394.
[38] H. Zhang, J.M. Cai, Y.T. Wang, M.Q. Wu, M. Meng, Y. Tian, X.G. Li, J. Zhang,
L.R. Zheng, Z. Jiang, J.L. Gong, Appl. Catal., B 220 (2018) 126–136.
[39] S. Shen, X. Wang, Q. Ding, S. Jin, Z. Feng, C. Li, Chinese J. Catal 35 (2014)
1900–1906.
[30] Q. Fu, F. Yang, X. Bao, Acc. Chem. Res. 46 (2013) 1692–1701.
[31] A. Naldoni, M. Allieta, S. Santangelo, M. Marelli, F. Fabbri, S. Cappelli, C.L. Bianchi,
R. Psaro, V. Dal Santo, J. Am. Chem. Soc. 134 (2012) 7600–7603.
[32] T. Potlog, M. Dobromir, D. Luca, P. Onufrijevs, A. Medvids, A. Shamardin, Curr.
8