Photocatalytic Reduction of CO2 to CH3OH Coupling with the Oxidation of Amine to Imine

Article abstract of DOI:10.1007/s10562-018-2412-6

Abstract: The photocatalytic reduction of CO₂ to organic molecules is one promising approache for both decreasing CO₂ concentration in the atmosphere and storing energies. However, most of the photocatalytic reduction of CO₂

Full text of DOI:10.1007/s10562-018-2412-6

Catalysis Letters
https://doi.org/10.1007/s10562-018-2412-6
Photocatalytic Reduction of CO to CH OH Coupling with the Oxidation
2
3
of Amine to Imine
Tian Yang1 · Qiming Yu1 · Hongming Wang1
Received: 12 February 2018 / Accepted: 14 May 2018
©
Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
The photocatalytic reduction of CO to organic molecules is one promising approache for both decreasing CO concentration
2
2
in the atmosphere and storing energies. However, most of the photocatalytic reduction of CO cannot avoid the utilization of
2
sacrificial agents, which are not atomic economy and restricts the practical application of the photocatalytic reduction of CO .
2
In this contribution, an atomic economy photocatalytic reduction of CO to CH OH coupling with the oxidation of amine
2
3
to imine by Cu/TiO was reported, which can avoid using the sacrificial agents in the reaction systems. CO was reduced to
2
2
CH OH by photo-induced electrons; meanwhile benzylamine was selected as the reductant to react with photo-generated
3
holes and converted into imine with high selectivity. The results showed that the maximum conversion of benzylamine to
−
1
imine is 88.7% with selectivity of 98%, and the highest yield of CH OH is 961.4 µmol g using ca.0.5 wt% Cu/TiO as the
3
2
catalyst. And also, the reaction mechanism was also investigated by DFT calculation, which would give a detailed explana-
tion for the reaction process.
Graphical Abstract
Schematic illustration is that maximum conversion of benzylamine to imine is 88.7% with selectivity of 98%, and the highest
−
1
yield of CH OH is 961.4 μmol g using ca.0.5 wt% Cu/TiO as the catalyst
3
2
Keywords Photocatalysis · Reduction of CO · DFT · Schiff base · Oxidation
2
1
Introduction
An intriguing research found that the H and O derived
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s10562-018-2412-6) contains
supplementary material, which is available to authorized users.
2
2
from water-slipping with a weak external voltage under the
light condition in 1972 [1]. Over the past few years, more
and more attentions were paid to the study of photocataly-
sis [2–4]. With the development of photocatalytic technol-
ogy, it can not only produce hydrogen by photocatalytic
water-splitting [5, 6], but also prepare a series of chemical
*
Hongming Wang
hongmingwang@ncu.edu.cn
1
College of Chemistry and Institute for Advanced Study,
Nanchang University, Nanchang 330031, Jiangxi, China
Vol.:(0
1
123456789)
3

T. Yang et al.
raw materials and small organic molecules by photocata-
2 Experiments
lytic reduction of CO [7–10]. To a certain extent, photo-
2
catalytic reduction of CO can solve both the energy crisis
2.1 The Preparation of Copper Loading Crystalline
2
and the greenhouse effects caused by the accumulation of
Titania (Cu/TiO ) Nanoparticle
2
CO in the atmosphere [11–14].
2
Motivated by the the above-mentioned advantages
The preparation of copper loading crystalline titania (Cu/
of photocatalytic reduction of CO , an increasing num-
TiO ) nanoparticle was according with a representative
2
2
ber of researchers launched out into the studies [15–23].
synthesis process. 10 mL tetrabutyl titanate was dissolved
Inoue et al. reported the photocatalytic reduction of CO
in 40 mL anhydrous ether at room temperature. A volume
2
to HCHO, CH OH, CH using TiO , ZnO, CdS, SiC
of 12 mL acetic acid (CH COOH) was added to the above-
3
4
2
3
and WO as the photocatalysts in 1979 [24]. Li et al.
mention solution by the intensively stirring. A volume of
5 mL acetic acid was added to 10 mL deionized water
(or a predetermined concentration of Cu(NO ) aqueous
3
found CO could be photocatalytically transformed into
2
−
HCOO under visible light over a photoactive Ti-contain-
3
2
ing MOF, NH -MIL-125(Ti) [25]. However, the before-
solution) as hydrolysis inhibitor solution in the meantime.
When the precursor solution was adequately stirred for
half an hour, the hydrolysis inhibitor solution was slowly
added to the precursor solution. When the hydrolysis
inhibitor solution was added, a large amount of white pre-
cipitate was forming in the mixed solution. After stirring
for 2 h under airtight condition, the suspension was aged
in the same temperature for 24 h. The supernatant was
removed, and the underlayer white deposition was dried
at 100 °C under vacuum for 24 h. In order to remove the
residual organic compounds, the solid was alternately
washed with deionized water and ethanol, and dried at
100 °C to remove moisture. And then, the sample was
calcined at various temperatures for 2 h in muffle furnace.
The final sample was grinded into powder.
2
mentioned photocatalytic reduction of CO cannot avoid
2
the utilization of sacrificial agents, which is not atomic
economy and would restrict the practical application of
the photocatalytic reduction of CO .
2
In order to address this issue, an atomic economy pho-
tocatalytic (AEPC) reduction of CO is in urgent need.
2
In our previous work, photocatalytic reduction of CO
2
into CH OH coupling with aromatic alcohol high selec-
3
tive oxidation to aromatic aldehyde was carried out under
ambient conditions [10]. In order to improve the photo-
catalytic reduction efficiency of CO , in this paper, ben-
2
zylamine that has high ability of producing hydrogen, was
selected as reductant to react with photo-generated holes
and transfers to the imine. The product imine plays an
important role in the fields of medicine, catalysis, analyti-
cal chemistry, and photochromism [26–29]. This kind of
catalytic reactions can avoid using the sacrificial agents
and increase the atom utilization efficiency.
2.2 Photocatalysts Characterization
Photocatalyst titanium dioxide (TiO ) has always been
The crystalline phases of the photocatalyst were charac-
terized on a D8 diffractometer from Bruker instruments
inc by X-ray diffractometer (XRD, Cu Kɑ radiation,
λ=0.15406 nm, at 40 kV and 25 mA). The diffract grams
were obtained at 2θ range from 15° to 85° in steps of 0.02°/s.
The crystalline phases of the photocatalysts with different Cu
loading content were performed from diffraction peak inten-
sities of rutile and anatase corresponding to (101), (110)
and (120), respectively. According to the Debye–Scherrer
equation, the average crystalline size was obtained. Mor-
phology of the photocatalyst was investigated by a scan-
ning electron microscope (SEM, Quanta 200F FEI). The
microstructures of the photocatalyst was determined by
the transmission electron microscopy (TEM, JEM-2010F,
at 200 kV) equipped with EDX. The fluorescence lifetime
spectrum of the photocatalysts were examined by Fluoro-
Max-4 (λ = 365 nm) at room temperature. XPS spectrum
2
the first choice of photocatalyst, due to its well-known
advantages of abundance, nontoxicity, strong catalytic
activity and chemical and biological stability [30–33].
However, the photocatalytic activity of TiO nanoparticles
2
is higher under ultraviolet (UV) light irradiation and it is
widely used to photocatalytic reduction CO to CH OH
2
3
(
CH OH) [34–36]. Recently, Some new paths to solve the
3
above-mentioned problems are bubbling by alien ion dop-
ing to enhance photocatalytic activity, such as nonmetal
doping [37–40], self-doping [41–43], rare-earth metal
doping [44–47], transitional metal doping [48, 49] and
noble metal doping [50–52]. Copper loading crystalline
titania (Cu/TiO ) is actually exploited for the improve-
2
ment of the conversion efficiency of CO to CH OH from
2
3
under certain conditions [53]. In this paper, a sequence
of different concentrations for copper loading crystalline
titania (Cu/TiO ) were prepared by the emblematic sol–gel
of ca.0.5 wt% Cu/TiO was investigated on THERMO SCI-
2
2
method. These photocatalysts were devoted to photocata-
ENTIFIC K-Alpha surface analysis system, equipped with
lytic reduction of CO to CH OH coupling with the oxida-
650 µm Mono at Al Kɑ 1486.6 eV.
2
3
tion of amine to imine.
1
3

Photocatalytic Reduction of CO to CH OH Coupling with the Oxidation of Amine to Imine
2
3
2
.3 Photocatalytic Reduction of CO Coupling
3 Results and Discussions
2
with Benzylamine to the Corresponding Imine
Test
3.1 Catalyst Characterization
The reactions of photocatalytic CO2 reduction into
The microstructure of Cu/TiO nanoparticles was shown
2
CH OH coupling with benzylamine into N-benzyliden-
in Fig. 1 and Fig. S1 (see supporting information). From
Fig. 1a, this kind of photocatalyst is obviously nanopar-
ticles. The average particle size of the photocatalyst was
about 20 nm that was the same as the diameter calculated
by Debye–Scherrer equation from XRD results (Fig. S2),
which indicated that the photocatalyst was well crystal-
lized. From Fig. S1e and f, there are some shadows (cop-
per nanoparticles) on the surface of the photocatalyst
3
ebenzylamine were carried in 40 ml anhydrous acetoni-
trile with 0.06 g photocatalyst, 1 mmol benzylamine
(
purified by distilled before using). In order to adequately
remove diminutive quantities of oxygen, the reactor was
evacuated and highly purified by adding dry CO swiftly.
2
The CO was purified to remove small quantities of
2
H O and O through asbestos, allochroic silica gel and
2
2
activated carbon introducing the reactor. Subsequently,
the suspension was stirred at a high rate of speed for an
hour under the complete darkness. Then, the high pres-
sure mercury lamp (λ = 365 nm) was turned on (10 cm
from the reactor). After the reactor was irradiated for
which show that CuO nanoparticles and TiO nanoparti-
2
cles were clearly distinguished from each other. And Cu,
Ti, O elements are almost uniformly distributed, forming
a series of CuO/TiO nanocomposites. In fact, the uni-
2
form distribution of copper nanoparticles on the surface
1
0 h, gas chromatography (GC) with flame ionization
of TiO nanoparticles would be beneficial to improve pho-
2
detector (FID) was applied to quantitatively analyze the
product. And the liquid products were also determined
by a Bruker Avance 400 MHz spectrometer (NMR).
The GC-MS [equipped with a HP-5 capillary column
tocatalytic activity. A bathochromic shift absorption was
observed with the increasing of copper content, and it’s the
well-absorbed in the visible region, which may be caused
by the Cu plasmonic effect (Fig. S3a). XPS analyses of
Cu were showed in Fig. S3b, Based on the appearance of
(
30 m × 0.25 mm × 0.25 mm)], was applied to identify
2
p
the component of products.
the satellite feature, existence of CuO could be confirmed
[
57, 58].
Furthermore, a polycrystal diffraction ring was observed
2.4 Computation Method
for ca. 0.5 wt% Cu/TiO nanoparticles, which can attribute
2
to the (101) crystalline phase of TiO as shown in Fig. S1a.
2
First-principle calculations for the reaction on the anatase
And also, HR-TEM was used to investigate the interfacial
TiO (101) surface were carried out by using the Vienna
structures between CuO and TiO (Fig. 1c, d). In Fig. 1c,
2
2
ab initio simulation package (VASP) based on the all-elec-
tron projected augmented wave (PAW) method [54, 55].
The k-points sampling was generated following the Monk-
horst–Pack procedure with a 2 × 3 × 1 mesh for anatase
d, the lattice fringes of CuO and TiO showed a surface-
2
junction. The (110) lattice fringe of the CuO nanoparticles
was 0.278 nm, and the (101) lattice fringe of anatase was
0.356 nm. To demonstrate the success of the Cu load-
TiO (101) surface. The generalized gradient approxima-
ing, EDX spectrum of ca.0.5 wt% Cu/TiO nanoparticles
2
2
tion (GGA) with the spin-polarized Perdew–Burke–Ernz-
erhof (PBE) functional was used for all of the calculations.
During structural optimization, the bottom layers of the
slab were fixed at the bulk truncated position, while the
top two layers and the adsorbates were fully relaxed. To
prevent the artificial interaction between repeated slab
along z-direction, 15 Ǻ vacuum was introduced with
correction of the dipole moment. All calculation for the
formation of imine were performed using Gaussian 09
package [56]. Theoretical calculation for the geometrical
optimizations was performed by density functional theory
was determined, and the results were shown in Fig. S1f.
Ti, O and Cu elements were detected on the surface of
ca.0.5 wt% Cu/TiO nanoparticles, which indicates that
2
the photocatalyst was consist of TiO and CuO.
2
3.2 Catalytic Performance of Photocatalytic
Oxidation of Amine to Imine Coupling
with the Reduction of CO to CH OH
2
3
The photocatalytic oxidation of amine to imine coupling
with the reduction of CO to CH OH was carried out at
(
DFT) using B3LYP method. The 6-311G (d, p) basis sets
2
3
were employed for C, H, O and N atoms. The vibrational
frequency was calculated at the same level to characterize
the nature of the stationary points as true minima (with
no imaginary frequency) or transition states (with unique
imaginary frequency).
0.1 Mpa CO pressure under ultraviolet light (λ= 365 nm)
2
irradiation. The liquid products were determined by GC
(Fig. S4) and with flame ionization detector. At the same
1
time, liquid products have also been confirmed by H
1
3

T. Yang et al.
Fig. 1 a Representative SEM and b TEM image of ca.0.5 wt% Cu/TiO nanoparticles. c, d HR-TEM image of ca.0.5 wt% Cu/TiO nanoparticles
2
2
NMR spectroscopy (Fig. S5) and GC-MS (Fig. S6). All
these indicate that the CO was reduced to CH OH, and
increasing copper loading content in standard reaction con-
dition. The photocatalytic conversion of benzylamine into
2
3
benzylamine was converted to the corresponding imine
with high selectivity.
N-benzylidenebenzylamine and the yields of CH OH were
3
enhanced with the increasing of copper loading content. It
The detailed results for photocatalytic reductions of CO
was found that the ca.0.5 wt% Cu/TiO was optimal to attain
2
2
to CH OH coupling with of amine to imine catalyzed by
the maximum conversion of benzylamine (67.9%) and the
3
−
1
TiO -base catalysts were showed in Table 1. From Table 1,
highest yield of CH OH (837.8 µmol g ) after irradiating
2
3
it is found that the CO was reduced to CH OH, and ben-
for 10 h. It could be interpreted that the CuO nanoparticles
can inhibit the recombination of photogenerated charge and
carrier to enhance the photocatalytic activity under ultravio-
let light irradiation [57, 58].
2
3
zylamine was converted to the corresponding imine with
the high selectivity for all catalysts. The yield of CH OH is
3
−
1
2
01.8 µmol g , and the conversion of benzylamine is 16.5%
catalyzed by pure TiO . However, the yield of CH OH and
However, the photocatalytic activity decreases when
2
3
conversion of benzylamine showed a tremendous growth
copper loading exceeds ca.0.5 wt% Cu/TiO . With the
2
catalyzed by Cu/TiO nanoparticles, which turned out that
increasing copper loading content from 0.5 to 1.0%,
2
the copper loading can effectively improve the photocata-
lytic activity. Moreover, two blank reactions were studied
as well, which showed that the conversion of benzylamine
the conversion of benzylamine and the yield of CH OH
3
were declined from 67.9 to 34.8% and from 837.8 to
−
1
410.8 µmol g , respectively. It might be interpreted that
superfluous copper nanoparticles cover the surface of
with the low selectivity is 23.7 and 9.5% oxidized by O in
2
the absence and presence of catalyst.
TiO activity sites that could absorb UV photon. There-
2
The effects of copper loading content on photocatalytic
fore, absorption photon capacity of the photocatalyst was
reduced. Meanwhile, when the concentration of loading
metal were higher, the exceeding copper nanoparticles
activity of Cu/TiO were evaluated. The catalytic activity of
2
ca.0.1–0.5 wt% Cu/TiO nanoparticles was improved with
2
1
3

Photocatalytic Reduction of CO to CH OH Coupling with the Oxidation of Amine to Imine
2
3
Table 1 The yields of the
ME/SBe
Entry
Catalyst
Gas
CH OH
Amine
3
reduction of CO coupling with
2
the conversion of benzylamine
to N-benzylidenebenzylamine
with high selectivity under UV
light irradiationa after 10 h
(µmol/g)
Conv. (%)
Sel. (%)
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
TiO2
TiO2
P25
CO2
N2
201.8
–
16.5
5.3
99
99
99
99
99
99
99
99
99
99
99
99
99
99
–
1/6.6
0
CO2
CO2
CO2
CO2
CO2
CO2
CO2
CO2
CO2
CO2
CO2
N2
285.5
317.6
342.7
396.2
505.6
837.8
764.9
563.7
458.4
429.6
410.8
–
24.7
28.3
30.5
34.4
43.8
67.9
57.4
46.7
40.9
36.1
34.8
7.8
–
11.3
–
23.7
9.5
1/7.3
1/7.2
1/7.4
1/7.2
1/7.2
1/6.7
1/6.2
1/6.9
1/7.4
1/7
1/7.1
0
0
1/6.4
0
–
0.1%Cu/TiO2
0.2%Cu/TiO2
0.3%Cu/TiO2
0.4%Cu/TiO2
0.5%Cu/TiO2
0.6%Cu/TiO2
0.7%Cu/TiO2
0.8%Cu/TiO2
0.9%Cu/TiO2
1.0%Cu/TiO2
0.5%Cu/TiO2
–
0
1
2
3
4
5
6
7
8
9
a
b
c
CO2
CO2
CO2
O2
–
0.5%Cu/TiO2
0.5%Cu/TiO2
0.5%Cu/TiO2
–
145.1
–
–
99
–
63
57
d
O2
–
–
1
mmol benzylamine; 40 mL CH CN, 60 mg catalyst
3
a
Without catalyst
b
c
d
e
4
4
0 mL benzylamine without CH CN
3
0 mL CH CN without benzylamine
3
Without catalyst
The molar ratio of the formation of CH OH and imine
3
could also act as recombination centers of electron and
hole, which also gave rise to the declining of the catalytic
the progression of the reaction, the CH OH also reacts
3
with photo-generated holes to produce H and CO .
2
2
activity of Cu/TiO nanocomposites. The above-men-
In order to compare with above results, a series of blank
experiments were carried out under the above-mentioned
conditions. When N was used to replaced CO , the conver-
2
tioned conclusions can be proved by the UV–Vis diffuse
reflectance spectroscopy and photoluminescence spectra
of the photocatalysts showed in Fig. S3a and S7. From
this Fig. S7, when the copper loading content was lower
than 0.5 wt%, the fluorescence intensity of the photocata-
lyst were decreases with the increase of copper loading
content. However, when the copper loading content was
higher than 0.5 wt%, the fluorescence intensity of the pho-
tocatalyst was gradually increasing with the increasing of
copper loading content. It was well known that the photo-
catalytic activity of the photocatalyst is negative correla-
tion with the fluorescence intensity of the photocatalyst,
which were consistent with experimental conclusions. In
order to know the stoichiometry of this reaction, the ratio
2
2
sion of benzylamine are only 5.3 and 7.8%, and there are no
small organic molecule to be observed, catalyzed by TiO
2
and ca.0.5 wt% Cu/TiO , respectively. These indicated that
2
the CO was able to be conversed to CH OH in standard
2
3
reaction condition. Moreover, without benzylamine in reac-
tion system, no product is observed catalyzed by ca.0.5 wt%
Cu/TiO in CH CN. These results certificated that CO can
2
3
2
be reduced to CH OH coupling with the conversion of ben-
3
zylamine to the corresponding imines.
Furthermore, the photocatalytic reductions of CO cou-
2
pled with selective oxidation of various amines to imines
have been performed over ca.0.5 wt% Cu/TiO under the
2
of the formation of CH OH and imine has been calculated
same reaction conditions. The results were shown in Table 2.
The photocatalytic oxidation of the benzylamine to the cor-
responding imines and the conversion of CO into CH OH
3
(
the last column in Table 1). The result shows that the ratio
will be keep in a range from 1:6.5 to 1:7.5 for all catalysts,
2
3
which is smaller than the reaction stoichiometry (1:3). The
were observed. It indicated that the groups on the aromatic
ring have an effect on the photocatalytic performance. There
are 10.9–67.9% conversions for different kind of amines in
first reason is that side-products are formed, such as H ,
2
CH , HCOOH, HCHO etc. The second reason is that with
4
1
3

T. Yang et al.
Table 2 The yields of the reduction of CO coupling with the conver-
to know the electronic effect of substituted group, the elec-
tron-donating groups (CH O- and CH -) and the electron-
2
sion of amine to corresponding imine with ca.0.5 wt% Cu/TiO cata-
2
3
3
lyst under UV light irradiation after 10 h
drawing groups (F- and Cl-) benzylamine were investigated
in detailed. The methoxy benzylamine could be converted
into the corresponding imines with conversions of 37.6%
a
b
Entry Reactant Product
Yield
Conv. [%]
Sel.
b
[
%]
−
1
and the yield of CH OH of 234.2 µmol g after reacting for
3
1
0 h, respectively. However, the conversion of 4-fluoroben-
NH
2
1
2
3
N
837.8
157.2
213.7
67.9
21.1
30.2
99
99
98
zylamin is only 18.9% and the yield of CH OH is lower
3
−
1
(
135.9 µmol g ) under the same reaction conditions. The
NH
2
results showed the electron donating groups on the aromatic
N
N
ring favor CH OH production and benzylamine oxidation.
3
However, the electron withdrawing groups on the aromatic
NH
2
ring resulted in a low conversion of benzylamine and the
yield of CH OH. Compared with 4-chlorobenzylamine,
3
2
,4-dichlorobenzylamine were easy to be oxidated into the
NH
2
corresponding imine (entry 8). As can be seen from Table 2,
the reaction of heteroatom-containing substrates amine
derivatives were also converted to corresponding imine with
high selectivity under photocatalysis (entries 9–10). How-
ever, there are no products to be observed when the aliphatic
amine used in this reaction under same reacting conditions
4
5
6
N
271.4
135.9
221.5
54.3
18.9
34.5
99
98
98
NH
2
N
F
F
F
NH
2
N
(
entries 11–12).
Cl
Cl
Cl
The effect of irradiation time on photocatalytic reduction
of CO into CH OH and the conversion of benzylamine were
2
3
NH
2
7
8
N
234.2
247.1
37.6
41.2
99
99
conducted on ca.0.5 wt% Cu/TiO nanoparticles under the
O
O
O
2
same reaction condition (Fig. 2). The conversion of ben-
zylamine was clearly enhanced with the increasing of irra-
diation time. However, the selectivity was slightly declined
due to the growth of by-products in the reaction system. At
NH
2
N
Cl
Cl
Cl
Cl Cl
Cl
S
S
S
9
1
1
1
NH2
N
N
193.7
64.8
29.8
10.9
98
98
the same time, the yield of CH OH was gradually increased
3
with the elongation of irradiation time. When the irradia-
tion time reached 15 h, the yield of CO into CH OH was
O
O
O
2
3
0
1
2
NH
2
−1
9
61.4 µmol g (Fig. 2b), and the conversion of benzylamine
to N-benzylidenebenzylamine was 88.7 with 98% selectivity
(Fig. 2a).
–
–
–
–
–
–
–
–
NH
2
In order to test the reusability of the Cu/TiO nanopar-
2
ticle, the used catalyst was recovered by centrifugation
and thoroughly washed with anhydrous acetonitrile. The
photocatalyst was recycled up to three times. As shown in
Fig. 3, the photocatalytic activity of ca.0.5 wt% was only
slightly deteriorated after three consecutive photocatalytic
experiments. It could be interpreted that a small portion of
copper nanoparticles were leached and aggregated after the
photocatalytic experiment. So that photocatalytic activity
was slightly decreased. In summary, these kinds of catalyst
NH
2
1
mmol amine, 40 mL CH CN, 60 mg catalyst, 0.1 Mpa
3
b
−1
Unit: [µmol g
Conversion of amine
]
c
d
Selectivity of the corresponding imine
standard reaction condition. There are obvious differences
has high stability for photocatalytic reduction of CO into
2
both the conversion of amines and the yield of CH OH
CH OH coupling with the conversion of benzylamine into
3
3
among three kinds of methyl substituted benzylamine (at
ortho-, meta-, para-positions of the benzene ring, Table 2,
entries 2–4). The orders for the conversion of amines and
N-benzylidenebenzylamine.
3.3 Catalytic Mechanism
the yield of CH OH are both ortho- > meta- > para-, which
3
indicated that space steric hindrance of substituted ben-
To understand the photocatalysis reaction process, the
DFT calculations were carried out to study the reaction
zylamine had large effect on the catalytic activity. In order
1
3

Photocatalytic Reduction of CO to CH OH Coupling with the Oxidation of Amine to Imine
2
3
mechanism. The mechanism for photocatalytic reduction of
CO to CH OH coulping with the conversion of amine to
2
3
corresponding imine was showed in Fig. 4 and as follows:
∗
Catalyst + hv → catalyst + e
−
+
+ p
(1)
(2)
(3)
(4)
(5)
cond
val
+
+
PhCH NH (aq) + 2p
2
→ PhCH = NH + 2H
→ HCOOH
2
val
cond
+
−
CO + 2H
2
+ 2e
cond
cond
+
−
+ 2e
HCOOH + 2H
→ HCHO + H O
cond 2
cond
+
−
+ 2e
HCHO + 2H
→ CH OH
cond 3
cond
PhCH NH + PhCH = NH → PhCH = NCH Ph + NH
3
2
2
2
(
6)
PhCH = NH + H O → PhCHO + NH
2
2
3
(7)
PhCHO + PhCH NH → PhCH = NCH Ph + H s
2
2
2
2
(8)
Photogenerated holes in the VB of pholocatalyst oxi-
dize benzylamine to corresponding imine and generating
+
H (Eq. 2), and CO was reduced by photogenerated elec-
2
trons in CB following the sequence of reactions to produce
HCOOH, HCHO and CH OH (Eqs. 3, 4, 5). The formed
3
imine in step 2 was conversed to Schiff base via two possible
reaction channels, the concerted (channel 1) and stepwise
(
channel 2) mechanisms. For the channel 1, the imine was
reacted directly with benzylamine to form imine (step 6).
The channel 2 was divided into two step, the first step was
Fig. 2 a The influence of irradiation time on conversion of ben-
zylamine and selectivity of corresponding imine with ca.0.5 wt% Cu/
−
1
TiO catalyst. b The yields of CH OH production (unit: µmol g
)
the hydrolysis of imine to give NH and aldehyde that was
2
3
3
with the prolonging of irradiation time. (Reaction condition: 1 mmol
then reacted with amine to produce the imine and H O (step
2
benzylamine, 60 mg photocatalyst, 40 mL CH CN)
3
7
and 8).
Figure 5 presents the calculated potential energy surface
of the dissociation of benzylamine on anatase TiO (101)
2
(
Fig. S9). The benzylamine can be adsorbed at Ti atom on
TiO (101) surface to form a monodentate adsorbate without
2
an intrinsic transition state, where the N–Ti bond length was
2
.31 Å and the adsorption energy was 22.50 kcal/mol. Then
a hydrogen transfer was carried from N atom of benzylamine
to the O atom on TiO (101) surface, forming an O–H bond
2
via TS1 with a reaction barrier of 27.91 kcal/mol. The
Fig. 3 The conversion of benzylamine to N-benzylidenebenzylamine
Fig. 4 The proposed mechanism of photocatalytic reduction of CO2
with ca.0.5 wt% Cu/TiO catalyst in different recycling runs at
to CH OH coupling with the conversion of amine to corresponding
2
3
0
.10 MPa CO under UV light irradiation
imine
2
1
3

T. Yang et al.
Fig. 5 The calculated potential energy of the dissociation of ben-
zylamine on anatase TiO (101)
2
second hydrogen transfer step from C atom of benzylamine
to O atom on TiO (101) surface takes place by overcoming
2
reaction barrier of 18.33 kcal/mol at TS2, producing ben-
zylimine with an endothermicity of 15.64 kcal/mol.
The formation of Schiff base from benzylimine was able
to proceed via two channels (channel 1 and channel 2 in
Fig. 4). The benzylamine was able to react with formed ben-
zylimine via an electrophilic addition of N to C=N bond to
form the intermediate (IN1a) by overcoming energy bar-
rier of 11.29 kcal/mol (channel 1 in Figs. 4, 6a). Then, a
hydrogen transfer step between two N atoms takes place via
TS2b with energy barrier of 11.53 kcal/mol to attain the last
product Schiff base.
If the existence of water, the reaction would take place via
another reaction channel (channel 2 in Fig. 4). The calcu-
lated energy profile for this channel was showed in Fig. 6b,
c. In the first step, the H O was reacted with benzylimine
2
via transition state TS1a in that the hydrogen atom of H O
2
Fig. 6 The calculated potential energy for the producing imine from
benzylimine via channel 1 (a) and channel 2 (b, c)
was transferred to N atom of benzylimine, producing IN1b
with an energy barrier of 10.48 kcal/mol. Then the second
hydrogen of H O was continually transferred to N atom of
2
benzylimine via TS2a with energy barrier of 7.71 kcal/mol,
channel 1 with channel 2, it is found that the energy barrier
in former is higher than that of the latter, which means that
then the NH and benzaldehyde was attained. Our experi-
3
ments also indicate that the NH was the main gas product
the introduction of H O can weakly accelerate the reaction
3
2
base on GC analysis (Fig. S6), which is good agreement with
process.
the calculating results.
It is well know that the benzaldehyde is able to react eas-
ily with benzylamine to form Schiff base. Figure 6c pre-
sents the calculated potential energy of the reaction between
benzaldehyde and benzylamine via two elementary reac-
tion steps. For the first step, the N atom of benzylamine
nucleophilicly attack the carbanyl group of benzaldehyde
to produce IN1c in that the hydrogen atom on N atom was
transferred to O atom of carbanyl group with a low energy
barrier of 9.96 kcal/mol. Then the second H on N atom was
transfer to O atom of carbanyl group via TS2b with energy
4 Conclusions
In this paper, an reduction of CO into CH OH coupling
2
3
with the conversion of aromatic amine into corresponding
imine by a series of photocatalysts have been conducted
under UV irradiation. This kind of catalytic reactions can
avoid using the sacrificial agents and increase largely the
atom utilization efficiency. The optimal photocatalyst was
ca.0.5 wt% Cu/TiO nanoparticles. The highest yield of the
2
−1
barrier 10.67 kcal/mol to form the H O and imine. Compare
conversion of CO into CH OH is 961.4 µmol g after 14 h.
2
2
3
1
3

Photocatalytic Reduction of CO to CH OH Coupling with the Oxidation of Amine to Imine
2
3
Meanwhile, the maximum conversion of benzylamine to
corresponding imine is 88.7% with high selectivity of 98%.
Therefore, it provides a new route for photocatalytic reduc-
tion of CO to CH OH, and also to produce another kind of
21. Liu L, Zhao C, Pitts D et al (2014) Catal Sci Technol 4:1539–1546
2
2. Li H, Zhang X, MacFarlane DR (2015) Adv Energy Mater
5
:1401077
2
3. Liang Y, Vijayan BK, Gray KA et al (2011) Nano Lett
11:2865–2870
2
3
significant chemical products.
24. Inoue T, Fujishima A, Konishi S et al (1979) Nature 277:637–638
2
5. Fu Y, Sun D, Chen Y et al (2012) Angew Chem Int Ed
1
24:3420–3423
2
6. Simon T, Bouchonville N, Berr MJ et al (2014) Nat Mater
13:1013–1018
5
Supporting Information
2
2
7. Huang D, Yin L, Niu J (2016) Environ Sci Technol 50:5857–5863
8. Sun Y, Cheng H, Gao S et al (2012) Angew Chem Int Ed
The Supporting Information is available free of charge on
The SEM, TEM, emission spectra, GC, GC-MS results and
additional data.
5
1:8727–8731
2
3
9. Guan M, Xiao C, Zhang J et al (2013) J Am Chem Soc
135:10411–10417
0. Spadavecchia F, Ceotto M, Presti L et al (2014) Chin J Chem
32:1195–1213
Acknowledgements We are grateful to the support of the National
Natural Science Foundation of China (21363014 and 21103082).
This work was supported by the Natural Science Foundation of
the Jiangxi Province of China (Grant Nos. 20142BCB23003 and
31. Yamada N, Suzumura M, Koiwa F et al (2013) Water Res
47:2770–2776
32. De Angelis F, Di Valentin C, Fantacci S et al (2014) Chem Rev
114:9708–9753
2
0143ACB21021).
3
3
3
3
3. Zheng Z, Zhao J, Yuan Y et al (2013) Chem Eur J 19:5731–5741
4. Jiao W, Wang L, Liu G et al (2012) ACS Catal 2:1854–1859
5. Zeng G, Qiu J, Li Z et al (2014) ACS Catal 4:3512–3516
6. Ramesha GK, Brennecke JF, Kamat PV (2014) ACS Catal
4:3249–3254
Compliance with Ethical Standards
Conflict of interest The authors declare no conflict of interest.
3
7. Liu S, Liu J, Li X et al (2015) Plasma Processes Polym
1
2:422–430
3
3
4
8. Marschall R, Wang L (2014) Catal Today 225:111–135
9. Li Y, Xu D, Oh J II et al (2012) ACS Catal 2:391–398
0. Li M, Zhang S, Peng Y et al (2015) RSC Adv 5:7363–7369
References
1
2
3
.
.
.
Fujishima A, Honda K (1972) Nature 238:37–38
41. Liu L, Chen X (2014) Chem Rev 114:9890–9918
42. Zuo F, Bozhilov K, Dillon RJ et al (2012) Angew Chem Int Ed
124:6327–6330
Pal B, Sharon M (2002) Mater Chem Phys 76:82–87
Park H-A, Choi JH, Choi KM et al (2012) J Mater Chem
2
2:5304–5307
43. Xing M, Li X, Zhang J (2014) Sci Rep 4:5493
44. De Angelis F, Fantacci S, Selloni A et al (2007) Nano Lett
7:3189–3195
4
.
Bessekhouad Y, Mohammedi M, Trari M (2002) Sol Energy
Mater Sol Cells 73:339–350
5
6
7
8
.
.
.
.
Moomaw WR, Unruh GC (1997) Environ Dev Econ 2:451–463
Takeda H, Cometto C, Ishitani O et al (2016) ACS Catal 7:70–88
Li Y, Lu G, Li S (2001) Appl Catal A 214:179–185
Nguyen T-V, Kim KJ, Yang OB (2005) J Photochem Photobiol A
45. Zanella R, Rodríguez-González V, Arzola Y ACS Catal 2:1–11
46. Borlaf M, Colomer MT, de Andrés A et al (2014) Eur J Inorg
Chem 30:5152–5159
47. Liyanage AD, Perera SD, Tan K et al (2014) ACS Catal 4:577–584
48. Song K, Han X, Shao G (2013) Alloys Compd 551:118–124
49. Yu X, Hou T, Sun X et al (2012) Chem Phys Chem 13:1514–1521
50. Priebe JB, Radnik J, Lennox AJJ et al (2015) ACS Catal
5:2137–2148
1
73:56–63
9
.
Tamaki Y, Ishitani O (2017) ACS Catal 7:3394–3409
1
0. Wang L, Zhang X, Yang L et al (2015) Catal Sci Technol
5
:4800–4805
1
1. Fujishima A, Nagahara LA, Yoshiki H et al (1994) Electrochim
Acta 39:1229–1236
51. Chen P, Khetan A, Yang F et al (2017) ACS Catal 7:1197–1206
52. Du Q, Wu J, Yang H (2013) ACS Catal 4:144–151
53. Ola O, Maroto-Valer MM (2015) J Photochem Photobiol C
24:16–42
1
1
1
2. Lin L, Ou H, Zhang Y et al (2016) ACS Catal 6:3921–3931
3. Wang D, Liu Z, Yang W (2017) ACS Catal 7:2744–2752
4. Schneider TW, Ertem MZ, Muckerman JT et al (2016) ACS Catal
54. Blöchl PE (1994) Phys Rev B 50:17953
6
:5473–5481
55. Kresse G, Joubert D (1999) Phys Rev B 59:1758–1775
56. Frisch MG, Trucks W, Schlegel HB et al (2009) Gaussian, Inc.,
Wallingford CT
1
1
5. Habisreutinger SN, Schmidt-Mende L, Stolarczyk JK (2013)
Angew Chem Int Ed 52:7372–7408
6. Zhou Y, Tian Z, Zhao Z et al (2011) ACS Appl Mater Interfaces
57. Poulston S, Parlett PM, Stone P et al (1996) Surf Interface Anal
24:811–820
3
:3594–3601
1
1
1
2
7. Xing M, Shen F, Qiu B et al (2014) Sci Rep 4:6341
58. Ghodselahi T, Vesaghi MA, Shafiekhani A et al (2008) Appl Surf
Sci 255:2730–2734
8. Yang Y, Zhang T, Le L et al (2014) Sci Rep 4:6341
9. Pan J, Wu X, Wang L et al (2011) Chem Commun 47:8361–8363
0. Tan L, Chai S, Mohamed AR (2012) ChemSusChem 5:1868–1882
1
3