Noble-metal-free deoxygenation of epoxides: Titanium dioxide as a photocatalytically regenerable electron-transfer catalyst

Article abstract of DOI:10.1021/cs500065x

Catalytic deoxygenation of epoxides into the corresponding alkenes is a very important reaction in organic synthesis. Early reported systems, however, require noble metals, high reaction temperatures (>373 K), or toxic reducing agents. Here, we report a noble-metal-free heterogeneous catalytic system driven with alcohol as a reducing agent at room temperature. Photoirradiation (λ <420 nm) of semiconductor titanium dioxide (TiO₂) with alcohol promotes efficient and selective deoxygenation of epoxides into alkenes. This noble-metal-free catalytic deoxygenation is facilitated by the combination of electron transfer from surface Ti³⁺ atoms on TiO₂ to epoxides, which promotes deoxygenation of epoxides, and photocatalytic action of TiO₂, which regenerates oxidized surface Ti atoms with alcohol as a reducing agent.

Full text of DOI:10.1021/cs500065x

Research Article
pubs.acs.org/acscatalysis
Noble-Metal-Free Deoxygenation of Epoxides: Titanium Dioxide as a
Photocatalytically Regenerable Electron-Transfer Catalyst
Yasuhiro Shiraishi,* Hiroaki Hirakawa, Yoshiki Togawa, and Takayuki Hirai
Research Center for Solar Energy Chemistry, and Division of Chemical Engineering, Graduate School of Engineering Science, Osaka
University, Toyonaka 560-8531, Japan
S
* Supporting Information
ABSTRACT: Catalytic deoxygenation of epoxides into the corre-
sponding alkenes is a very important reaction in organic synthesis.
Early reported systems, however, require noble metals, high reaction
temperatures (>373 K), or toxic reducing agents. Here, we report a
noble-metal-free heterogeneous catalytic system driven with alcohol
as a reducing agent at room temperature. Photoirradiation (λ <420
nm) of semiconductor titanium dioxide (TiO2) with alcohol
promotes efficient and selective deoxygenation of epoxides into alkenes. This noble-metal-free catalytic deoxygenation is
facilitated by the combination of electron transfer from surface Ti3+ atoms on TiO2 to epoxides, which promotes deoxygenation
of epoxides, and photocatalytic action of TiO2, which regenerates oxidized surface Ti atoms with alcohol as a reducing agent.
KEYWORDS: photocatalysis, titanium dioxide, surface defects, reduction, deoxygenation
V).15 This means that two-electron reduction of epoxides by
INTRODUCTION
■
the conduction band e− is thermodynamically unfavorable.
Deoxygenation of epoxides into the corresponding alkenes is an
Detailed study on the deoxygenation mechanism is therefore
essential reaction in biological systems for the reproduction of
necessary.
vitamin K in the vitamin K cycle.1,2 Deoxygenation is also a
It is well-known that trivalent titanium atom (Ti3+) behaves
very important reaction in organic synthesis for the
as a single-electron-transfer reagent, as often reported for
deprotection of the oxirane ring, a protecting group for C
C bonds.3−5 The reaction is usually carried out using a
stoichiometric or excess amount of toxic reducing agents, such
as phosphines, silanes, iodides, and heavy metals, with a
concomitant formation of copious amount of wastes.6 Catalytic
deoxygenation driven by easily recyclable heterogeneous
catalyst is therefore an ideal process. Several heterogeneous
systems have been proposed so far for catalytic deoxygenation
of epoxides.7−13 These systems employ gold or silver
nanoparticles loaded on metal oxide supports with molecular
hydrogen (H2), carbon monoxide (CO), or alcohol as a
reducing agent. They successfully promote selective deoxyge-
nation at the reaction temperatures ranging from room
temperature to 353 K, but require noble metals as the catalyst.
An alternative process, which promotes efficient and selective
deoxygenation with a noble-metal-free heterogeneous catalyst, is
necessary for green organic synthesis.
titanocene complexes.16,17 As shown in Scheme 1, the Ti3+
Scheme 1. Catalytic Cycle for Titanocene-Promoted
Deoxygenation of Epoxides
atom within a titanocene complex18 promotes a homolytic
cleavage of the oxirane ring via the single-electron-transfer to
the oxygen atom (a).19 The radical formed is reduced by the
reaction with the other Ti3+ atom (b).20 β-Elimination of the
intermediate produces the corresponding alkenes with a
formation of oxidized two Ti atoms (c). These reactions
proceed efficiently and selectively even at room temperature,
although the system suffers from the reduction step for oxidized
Ti atoms (d); regeneration of Ti3+ requires hazardous reducing
agents, such as silanes and heavy metals.21
Very recently, Li et al.14 reported that UV irradiation of
semiconductor titanium dioxide (TiO2) in alcohol with
epoxides successfully produces the corresponding alkenes.
The reaction proceeds without noble metals at room
temperature; therefore, it is a potential green catalytic system
for deoxygenation. They proposed that the deoxygenation
proceeds via the two-electron reduction of epoxide by the
conduction band electrons (e−) on the photoactivated TiO2.
The reduction potential of epoxide (e.g., styrene oxide) is,
however, −2.4 V (vs NHE, pH 0),12 which is significantly more
negative than the bottom of the TiO2 conduction band (∼0
In the present work, we clarified the mechanism for
deoxygenation of epoxides on the photoactivated TiO2 by
Received: January 17, 2014
Revised: April 15, 2014
Published: April 16, 2014
© 2014 American Chemical Society
1642
dx.doi.org/10.1021/cs500065x | ACS Catal. 2014, 4, 1642−1649

ACS Catalysis
Research Article
means of X-ray photoelectron spectroscopy (XPS) and diffuse-
reflectance infrared Fourier transform (DRIFT) analysis. The
active sites for deoxygenation are the Ti3+ atoms located at the
surface defects. As shown in Scheme 2, the TiO2 (110) surface
This photocatalytic cycle successfully regenerates the surface
Ti3+ atoms with alcohol as a reducing agent, thus promoting
efficient catalytic deoxygenation.
RESULTS AND DISCUSSION
■
Scheme 2. Catalytic Cycle for Deoxygenation of Epoxides on
the Surface of TiO2 (110) under Photoirradiation
Deoxygenation of styrene oxide was carried out with 2-PrOH in
the presence of TiO2 particles with different surface areas.
Table 1 summarizes the results obtained by photoirradiation (λ
> 300 nm, 6 h) of TiO2 (10 mg) in a 2-PrOH/MeCN mixture
(5/5 v/v, 5 mL) with styrene oxide (20 mM) under N2
atmosphere. All of the TiO2 particles (samples 1−9) produce
styrene with very high selectivity (84−95%), although the
styrene yields are different from each other. GC analysis of the
solutions recovered after the reaction detected minor amounts
of 2-phenylethanol as a byproduct. IR analysis of the solution
(Figure S2, Supporting Information) showed a band at ∼1110
cm−1 assigned to C−O stretching of the ether linkage,
suggesting that polymerized styrene oxide is also involved as
a byproduct, as observed in related catalytic systems.28 As
shown in Table 1 (sample 10), the dark reaction scarcely
promotes deoxygenation. In addition, photoirradiation without
alcohol (sample 11) also does not promote reaction. These
data indicate that photoirradiation with alcohol is necessary for
deoxygenation.
a
a
The light blue spheres that lie in the (001) azimuth are the Ob atoms,
The active sites for deoxygenation of epoxide are the surface
Ti3+ atoms on TiO2 (Scheme 2). XPS analysis confirms this. As
shown in Figure 1a, XPS chart for TiO2 (sample 9) shows Ti
2p1/2 and 2p3/2 signals, both of which consist of bulk Ti4+
(blue) and surface Ti3+ (red) atoms.29 In contrast, as shown in
Figure 1b, adsorption of styrene oxide onto TiO2 in the gas
phase leads to complete disappearance of the surface Ti3+ signal
(red), along with a formation of new Ti4+ signal at a higher
binding energy (green). This clearly indicates that the surface
Ti3+ atoms are oxidized to Ti4+30 via electron transfer to styrene
oxide.
and the parallel red spheres are the Ti atoms, respectively.
is characterized by alternate rows of 5-fold-coordinated Ti4+
atoms and bridging O atoms (Ob) that run in the (001)
direction.22 Surface defects are the Ob vacancies, where two
excess electrons associated with Ob are transferred to the empty
3d orbitals of neighboring Ti4+ atoms, producing two exposed
Ti3+ atoms.23 These Ti3+ atoms promote deoxygenation of
epoxides via the single electron transfers, producing alkenes and
oxidized Ti atoms. In contrast, photoexcitation of TiO2
produces electron (e−) and positive hole (h+) pairs.24−27 The
e−’s reduce the oxidized Ti atoms, and the h+’s oxidize alcohol.
The electron transfer from the surface Ti3+ atoms promotes
deoxygenation of epoxide. This is confirmed by DRIFT analysis
a
Table 1. Deoxygenation of Styrene Oxide on Various TiO2 Particles under Photoirradiation
crystalline
b
BET surface area
styrene oxide conv.
styrene yield
select.
(%)
NTi
vf
c
d
d
e
f
sample
catalyst
phase
(m2 g−1
)
(%)
(%)
(μmol g−1
)
(μmol h−1
)
g
1
2
3
4
5
6
7
8
9
ST-41
A
A
A
A
8
11
10
44
9
38
36
54
85
41
<1
69
95
<1
<1
84
86
84
87
88
84
2.9
34.2
12.8
59.4
1.4
6.1
g
ST-21
Aldrich anatase
67
43
5.7
g
ST-01
217
59
62
10.0
h
P25
A/R (83/17)
94
i
NS-51
R
R
R
R
7
49
15.9
1.9
7.7
<1
Wako rutile
15
<1
77
g
PT-101
25
90
95
81.6
168.0
11.4
25.7
h
JRC-TIO-6
100
>99
<1
<1
j
10
JRC-TIO-6
JRC-TIO-6
k
11
a
b
Photoirradiation was carried out using a 2 kW Xe lamp (light intensity at 300−450 nm is 27.3 W m−2). Determined by X-ray diffraction (XRD)
c
d
e
analysis (Figure S1, Supporting Information). Determined by N2 adsorption/desorption analysis (ref 37). Determined by GC analysis. The
number of surface Ti3+ atoms per gram TiO2, measured by DRIFT analysis of nitrobenzene adsorbed on TiO2 in the gas phase (Figure S3,
f
g
h
Supporting Information). Initial rate for styrene formation during 3 h photoreaction. Supplied from Ishihara Sangyo, Ltd. (Japan). Japan
Reference Catalyst supplied from the Catalyst Society of Japan. Supplied from Toho Titanium Co., Ltd. (Japan). Reaction was performed in the
dark. Reaction was performed in MeCN without 2-PrOH.
i
j
k
1643
dx.doi.org/10.1021/cs500065x | ACS Catal. 2014, 4, 1642−1649

ACS Catalysis
Research Article
(νoxirane),33 observed for styrene oxide at ∼1250 cm−1 (Figure
2c), scarcely appears. The results clearly indicate that styrene
oxide adsorbed onto the TiO2 surface is immediately
transformed to styrene via the electron transfer from the
surface Ti3+ atoms under dark condition. The catalytic
mechanism for deoxygenation of epoxide by surface Ti3+
atoms can therefore be summarized as Scheme 3A. One Ti3+
Scheme 3. (A) Proposed Mechanism for Epoxide
Deoxygenation on TiO2 Surface with Alcohol under
Photoirradiation, and (B) Energy Diagram for Regeneration
of Surface Ti3+ by Photoexcitation of TiO2
Figure 1. XPS chart in the Ti 2p region of (a) TiO2 (sample 9), (b)
the sample after adsorption of styrene oxide in the gas phase, and (c)
the sample after photoirradiation (λ > 300 nm) in 2-PrOH for 6 h.
of styrene oxide adsorbed onto TiO2 (sample 9) in the gas
phase. Figure 2a shows the time-dependent change in the
spectrum after the injection of styrene oxide. A distinctive
absorption band appears at 1630 cm−1, which is assigned to the
CC stretching vibration (νCC) of the styrene formed,31,32 as
evidenced by the spectrum of styrene (Figure 2b). In this case,
a band assigned to the breathing mode of oxirane ring
atom promotes homolytic cleavage of the oxirane ring via the
Ti4+−O bond formation (a). The radical formed is reduced by
another Ti3+ and produces a bridged intermediate (b). β-
Elimination of the intermediate affords the corresponding
alkene, along with a formation of oxidized two Ti4+ atoms (c).
The surface Ti atoms oxidized by deoxygenation of epoxide
(Scheme 3A.c) are reduced by the conduction band e−
produced by photoexcitation of TiO2 with alcohol as the
reducing agent (d−f).34 This regenerates the surface Ti3+ atoms
and completes the catalytic cycle. XPS analysis confirms this. As
shown in Figure 1b, TiO2 after adsorption of styrene oxide
exhibits green signals assigned to the oxidized surface Ti atoms.
However, as shown in Figure 1c, the TiO2 sample, when
photoirradiated in 2-PrOH under N2 atmosphere, leads to
disappearance of the green signals, along with a regeneration of
Ti3+ signals (red), similar to those observed for pure TiO2
(Figure 1a). This indicates that, as shown in Scheme 3A.d−f,
the oxidized surface Ti atoms are reduced to Ti3+ by
photoirradiation with alcohol.
Figure 2. (a) Time-dependent change in the DRIFT spectra of styrene
oxide adsorbed onto TiO2 (sample 9) in the gas phase at 303 K.
Attenuated total reflection (ATR) spectra of (b) styrene and (c)
styrene oxide measured at 303 K. The absorption bands indicated by
Figure 3 shows the time-dependent change in the amounts of
substrate and products during the reaction of styrene oxide in a
2-PrOH solution with TiO2 (sample 9) under photoirradiation.
Reaction for 4 h leads to complete conversion of styrene oxide
●
○
,
, and Δ are assigned to phenyl ring stretching, CH2
deformation, and phenyl C−H stretching (refs 31 and 32),
respectively.
1644
dx.doi.org/10.1021/cs500065x | ACS Catal. 2014, 4, 1642−1649

ACS Catalysis
Research Article
atoms with alcohol, thereby promoting efficient deoxygenation
of epoxides.
The surface Ti3+ atoms on TiO2 behave as active sites for
deoxygenation; therefore, the number of surface Ti3+ atoms
(NTi) critically affects the deoxygenation activity. The NTi
values can be determined by DRIFT analysis of nitrobenzene
adsorbed onto the TiO2 surface in the gas phase;37,38 as shown
in Figure S3 (Supporting Information), the intensity of
symmetric vibrational band (1346 cm−1) assigned to the
nitrobenzene adsorbed onto the surface Ti3+ atoms allows
accurate determination of NTi. The NTi (μmol g−1) values for
the respective TiO2’s are summarized in Table 1, and Figure 5
Figure 3. Time-dependent change in the amounts of substrate and
products during photoreaction of styrene oxide with TiO2 (sample 9)
in a 2-PrOH solution. Conditions are identical to those in Table 1.
to styrene (100 μmol) with a formation of an almost equivalent
amount of acetone (100 μmol). GC analysis also detected the
formation of water. These data suggest that, as shown in
Scheme 3A.d−f, photoirradiation reduces the oxidized surface
Ti atoms with alcohol as the reducing agent, and the H atoms
of alcohol are removed as water.
The oxidized surface Ti atoms are reduced by photo-
excitation of TiO2. Action spectrum analysis confirms this.
Figure 4 shows the apparent quantum yield for styrene
Figure 5. Relationship between the number of surface Ti3+ atoms
(NTi) and the average initial rate for styrene formation (vf) during
photoreaction (3 h) of styrene oxide on the respective TiO2. The
sample numbers correspond to those listed in Table 1.
shows the relationship between NTi and the average initial rate
for styrene formation (vf/μmol h−1) on the respective TiO2
determined by 3 h photoreaction. The proportional relation-
ship between NTi and vf suggests that surface Ti3+ atoms indeed
behave as the deoxygenation sites, and the TiO2 particles with a
larger number of surface Ti3+ is effective for deoxygenation.
The reaction scope of this TiO2 system is summarized in
Table 2. Photoirradiation of TiO2 (sample 9) in a 2-PrOH
solution promotes deoxygenation of various kinds of epoxides
into alkenes. For example, styrene oxides with several types of
substituents are successfully converted to styrenes (entries 1−
9). In particular, the reducible substituents such as halogen and
cyano groups are retained unchanged during deoxygenation
(entries 5−8). In addition, as shown in entries 2 and 3, the
TiO2 catalyst recovered after photoreaction of styrene oxide
(entry 1), when reused for further reaction, exhibits almost the
same yields of styrene. This indicates that the catalyst is
reusable without any loss of activity and selectivity.
In contrast, as shown by entries 10 and 12, aliphatic epoxides
such as 1,2-epoxyoctane and glycidyl phenyl ether show low
yields of alkenes (31 and 41%), although the selectivities are
high (82 and 88%). This is due to the low efficiency for
electron transfer from Ti3+ to the oxygen atom of their oxirane
ring, suppressing their ring-opening (Scheme 3a). As shown in
Figure S4 (Supporting Information), ab initio calculation based
on the density functional theory (DFT) revealed that the
electron density of the oxygen atom for LUMO of the aliphatic
epoxides is significantly smaller than that of aromatic
epoxides.39 This suggests that it is difficult for the electron to
transfer from Ti3+ to the oxygen atoms of aliphatic epoxide.
Figure 6a shows the DRIFT spectra of 1,2-epoxyoctane
adsorbed onto the surface of TiO2 (sample 9) in the gas
Figure 4. (line) Absorption spectra of TiO2 (sample 9), styrene oxide,
and styrene. (black circle) Action spectrum for styrene formation
obtained by the photoreaction of styrene oxide on the catalyst. The
apparent quantum yield for styrene formation was determined using
the equation ΦAQY (%) = (styrene formed ×2)/(photon number
entering into the reaction vessel) × 100 (see the Experimental
Section).
formation (ΦAQY), obtained by photoreaction of styrene
oxide with TiO2 (sample 9) using the monochromatic lights.
The quantum yields are consistent with the absorption
spectrum of TiO2, suggesting that the regeneration of surface
Ti3+ is promoted by band gap excitation of TiO2. As shown in
Scheme 3B, the edges of the conduction band (ECB) and
valence band (EVB) of TiO2 (rutile) are located at 0 and 3.0 V
(vs NHE, pH 0),15 respectively. In contrast, the donor levels of
surface Ti3+ atoms are located at 0.12−0.3 eV below ECB.35 As a
result of this, band gap excitation of TiO2 and subsequent
trapping of the conduction band e− by the oxidized Ti atoms36
regenerate surface Ti3+ atoms, while the positive hole (h+)
formed on TiO2 oxidizes alcohol. This photocatalytic cycle
(Scheme 3B) allows successful regeneration of surface Ti3+
1645
dx.doi.org/10.1021/cs500065x | ACS Catal. 2014, 4, 1642−1649

ACS Catalysis
Research Article
a
Table 2. Deoxygenation of Various Epoxides on TiO2 under Photoirradiation
a
b
Reaction conditions: TiO2 (sample 9, 10 mg), substrate (100 μmol), 2-PrOH/MeCN mixture (5/5 v/v, 5 mL), N2 (1 atm), Xe lamp. Determined
c
d
e
f
g
by GC. First reuse. Second reuse. Substrate (20 μmol). Catalyst (100 mg). 1,2-Diphenylethanol is formed as a byproduct. The yields are 42%
(entry 16), 11% (entry 17), 20% (entry 18), and 13% (entry 19), respectively.
phase at 303 K. An absorption band assigned to the CC
stretching vibration (νCC) of 1-octene observed at ∼1630
cm−1 (Figure 6c) does not appear, although a distinctive νCC
band is observed in the case of styrene oxide (Figure 2a). This
suggests that aliphatic epoxides are, indeed, less active than
aromatic epoxides. In this case, a vibrational band assigned to a
Ti−O−C bond, as often observed at ∼1260 cm−1,40,41 scarcely
appears in the spectra. This suggests that ring-opened
intermediates (Scheme 3A.b,c) are scarcely produced during
the DRIFT analysis and the electron transfer from Ti3+ to the
oxygen atom of epoxide (oxirane ring-opening, Scheme 3A.a) is
the rate-determining step for deoxygenation. The above ab
initio calculation and DRIFT analysis indicate that the low
reactivity of aliphatic epoxides is ascribed to the low efficiency
for the electron transfer from Ti3+ to the oxygen atoms, which
suppresses ring-opening of the oxyrane ring (Scheme 3A.a).
For these unreactive aliphatic epoxides, the reaction at
elevated temperature is effective. As shown by entries 11 and
13, the reaction at 343 K enhances deoxygenation and affords
the corresponding alkenes with moderate yields (55 and 71%).
As shown in Figure 6b, the DRIFT analysis of 1,2-epoxyoctene
performed at 343 K shows the νCC band at ∼1630 cm−1,
although at a weak level. This suggests that the electron transfer
from Ti3+ to the oxygen atom of the aliphatic epoxide is,
indeed, enhanced at elevated temperature. In contrast, as
shown by entries 14 and 15, the internal epoxide exhibits very
low reactivity, even at elevated temperature, as also observed in
other catalytic system.10 This is probably due to the steric
hindrance by adjacent alkyl chains, which suppresses the
electron transfer from Ti3+ to the oxygen atom of epoxide.
In the case of (Z)- and (E)-stilbene oxides (entries 16 and
18), the alkene selectivities are very low (48 and 71%) because
large amounts of 1,2-diphenylethanol are formed as a
byproduct. Several experimental results42 suggest that the
byproduct is produced by photoexcitation of a ring-opened
intermediate (Scheme 3A.b) via the absorption of short-
wavelength UV light (∼<350 nm) followed by abstraction of
hydrogen atom from 2-PrOH, as is also observed for related
diphenylethane derivatives.43,44 In these cases, photoexcitation
of TiO2 with relatively longer-wavelength light is effective for
1646
dx.doi.org/10.1021/cs500065x | ACS Catal. 2014, 4, 1642−1649

ACS Catalysis
Research Article
immersed in a temperature-controlled water bath (error: 0.5
K),47 and photoirradiated with magnetic stirring using a 2 kW
Xe lamp (USHIO Inc.; λ > 300 nm).48 The intensity of light at
300−450 nm was determined to be 27.3 W m−2. In some cases,
a grass filter was used to give light wavelength of λ > 370 nm,
where the intensity of light at 370−450 nm was 12.8 W m−2.
After the photoreaction, the catalyst was separated by
centrifugation, and the resulting solution was analyzed by
GC−FID. The substrate and product concentrations were
calibrated with authentic samples. Analysis was performed at
least three times, and the errors were 0.2%.
Action Spectrum Analysis. The reaction was carried out
using a 2-PrOH/MeCN (5/5 v/v) mixture (2 mL) containing
styrene oxide (0.04 mmol) and JRC-TIO-6 TiO2 (sample 9, 4
mg) within a Pyrex glass tube (ϕ 12 mm; capacity, 20 mL).
After ultrasonication and N2 bubbling, the solution was
photoirradiated using a 2 kW Xe lamp for 1 h, where the
light was monochromated by band-pass glass filters (Asahi
Techno Glass Co.).49 The full-width at half-maximum (fwhm)
of the light was 11−16 nm. Apparent intensity of light entered
into the reaction vessel was determined with a spectroradi-
ometer USR-40 (USHIO Inc.).50 The apparent quantum yields
(ΦAQY) for deoxygenation of styrene oxide (two-electron-
reduction reduction process) was calculated using the following
equation.51
Figure 6. Time-dependent change in the DRIFT spectra of 1,2-
epoxyoctane adsorbed onto TiO2 (sample 9) in the gas phase at (a)
303 and (b) 343 K. ATR spectra of (c) 1-octene and (d) 1,2-
●
epoxyoctane measured at 303 K. The absorption bands indicated by
,
○
, and Δ are assigned to CH2 scissoring, CH3 symmetric bending, and
CH2 wagging, respectively (refs 53 and 54).
[styrene formed (μmol)] × 2 × N × hc
Φ
=
AQY
(1)
H × A × λ × t
where N is the Avogadro′s number (mol−1), H is the apparent
intensity of light entered into the tube (W m−2), A is the
absorption cross section of the solution (m2), h is the Planck′s
constant (J s), c is the light speed (m s−1), λ is the light
wavelength (m), and t is the time for photoirradiation (s),
respectively.
selective alkene formation. As shown by entries 17 and 19, the
reaction of (Z)- and (E)-stilbene oxides, when performed by
the irradiation of >370 nm light, exhibits enhanced alkene
selectivity (both 85%).45,46 These data suggest that reaction
temperature and irradiation wavelengthes are the most
important factors for efficient and selective deoxygenation of
unreactive epoxides.
XPS Analysis. The analysis was performed using a JEOL
JPS-9000MX spectrometer with Mg Kα radiation as the energy
source.52 Measurements were carried out as follows: TiO2
(sample 9, 50 mg) was set in a vacuum line and evacuated
(0.9 Pa) at 423 K for 3 h. The sample was placed on a sample
stage immediately and evacuated at room temperature for 6 h,
and measurement was started (Figure 1a). After pretreatment
of TiO2 (0.9 Pa, 423 K, 3 h), styrene oxide (20 μmol) was
introduced in the gas phase at 303 K, and the cell was left for 30
min. The sample was placed on a sample stage immediately and
evacuated at room temperature for 6 h, and measurement was
started (Figure 1b). After the measurement, the sample (50
mg) was added to 2-PrOH (10 mL) and photoirradiated for 6 h
under N2 atmosphere in a manner similar to that for
photoreaction of epoxide. The resulting powders were
recovered by centrifugation and placed on a sample stage
immediately. The sample was evacuated at room temperature
for 6 h, and measurement was started (Figure 1c).
DRIFT Analysis. The spectra were measured on a FT/IR
610 system equipped with a DR-600B in situ cell (Jasco
Corp.).38 TiO2 (sample 9; 50 mg) was placed in a DR cell and
evacuated (0.9 Pa) at 423 K for 3 h. Epoxide (20 μmol) was
introduced to the cell at 303 K, and measurement was started.
Absorption bands in Figure 6 were assigned according to the
literature.53,54
Other Analysis. DR UV−vis spectra were measured on an
UV−vis spectrophotometer (Jasco Corp.; V-550 with Inte-
grated Sphere Apparatus ISV-469) with BaSO4 as a reference.
Absorption spectra were measured on an UV−vis photodiode
CONCLUSION
■
We clarified the mechanism for deoxygenation of epoxides on
the photoactivated TiO2 by means of XPS and DRIFT analysis.
This catalytic deoxygenation is facilitated by the combination of
electron transfer from surface Ti3+ atoms on TiO2 to the
oxygen atom of epoxides, promoting deoxygenation of
epoxides, and the photocatalytic action of TiO2, promoting
regeneration of surface Ti3+ atoms with alcohol as a reducing
agent. This photoprocess has significant advantages over the
early reported processes: (i) noble-metal-free and inexpensive
heterogeneous catalyst (TiO2), (ii) safe and inexpensive
reducing agent (alcohol), (iii) milder reaction conditions
(relatively low temperature (<343 K) and atmospheric
pressure). This process, therefore, has a potential to enable
green deoxygenation of epoxides. The role of surface Ti3+
atoms on the photoactivated TiO2 clarified here may help open
a new strategy toward the design of more efficient photo-
catalysts and the creation of new methods for photocatalysis-
based organic synthesis.
EXPERIMENTAL SECTION
■
Photoreaction. Each respective epoxide was dissolved in a
2-PrOH/MeCN mixture (5/5 v/v, 5 mL). The solution and
catalyst were added to a Pyrex glass tube (ϕ 12 mm; capacity,
20 mL), and the tube was sealed with a rubber septum cap. The
catalyst was dispersed well by ultrasonication for 5 min, and N2
gas was bubbled through the solution for 5 min. The tube was
1647
dx.doi.org/10.1021/cs500065x | ACS Catal. 2014, 4, 1642−1649

ACS Catalysis
Research Article
(20) Gansauer, A.; Lauterbach, T.; Narayan, S. Angew. Chem., Int. Ed.
̈
array spectrophotometer (Shimadzu; Multispec-1500). XRD
patterns were obtained by Philips X’Pert-MPD spectrometer.
Computational Details. Ab initio calculations were
performed at the DFT level within the Gaussian 03 program,55
using the B3LYP/6-31G(d) basis set. The excited states of
respective epoxides were calculated with the time-dependent
TD-DFT at the same level of optimization. Cartesian
coordinates for the respective epoxides are summarized at the
end of the Supporting Information.
2003, 42, 5556−5573.
́
(21) Justicia, J.; Jimenez, T.; Morcillo, S. P.; Cuerva, J. M.; Oltra, J. E.
Tetrahedron 2009, 65, 10837−10841.
(22) Papageorgiou, A. C.; Beglitis, N. S.; Pang, C. L.; Teobaldi, G.;
Cabailh, G.; Chen, Q.; Fisher, A. J.; Hofer, W. A.; Thornton, G. Proc.
Natl. Acad. Sci. U.S.A. 2010, 107, 2391−2396.
(23) Rodriguez, J. A.; Jirsak, T.; Liu, G.; Hrbek, J.; Dvorak, J.; Maiti,
A. J. Am. Chem. Soc. 2001, 123, 9597−9605.
(24) Maldotti, A.; Molinari, A.; Amadelli, R. Chem. Rev. 2002, 102,
3811−3836.
(25) Palmisano, G.; Augugliaro, V.; Pagliano, M.; Palmisano, L.
Chem. Commun. 2007, 3425−3437.
ASSOCIATED CONTENT
■
S
* Supporting Information
(26) Fagnoni, M.; Dondi, D.; Ravelli, D.; Albini, A. Chem. Rev. 2007,
107, 2725−2756.
XRD patterns of respective TiO2 (Figure S1), IR spectrum of
polymerized byproduct (Figure S2), DRIFT spectra of
nitrobenzene adsorbed onto the respective TiO2 (Figure S3),
interfacial plots of epoxides (Figure S4), absorption spectra of
stilbenes (Figure S5), and Cartesian coordinates of epoxides.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(27) Shiraishi, Y.; Hirai, T. J. Photochem. Photobiol. C 2008, 9, 157−
170.
(28) Seung, S. L. N.; Young, R. N. J. Polym. Sci. Polym. Lett. Ed. 1980,
18, 89−96.
(29) Reiβ, S.; Krumm, H.; Niklewski, A.; Staemmler, V.; Woll, C. J.
̈
Chem. Phys. 2002, 116, 7704−7713.
(30) Cherian, S.; Wamser, C. C. J. Phys. Chem. B 2000, 104, 3624−
3629.
AUTHOR INFORMATION
■
(31) Klust, A.; Madix, R. J. Surf. Sci. 2006, 600, 5025−5040.
(32) Wei, W.; Huang, W. X.; White, J. M. Surf. Sci. 2004, 572, 401−
408.
Corresponding Author
*E-mail: shiraish@cheng.es.osaka-u.ac.jp.
(33) Nyquist, R. A.; Fiedler, S. Vib. Spectrosc. 1994, 7, 149−162.
(34) Brinkley, D.; Engel, T. J. Phys. Chem. B 1998, 102, 7596−7605.
(35) Panayotov, D. A.; Burrows, S. P.; Morris, J. R. J. Phys. Chem. C
2012, 116, 4535−4544.
Notes
The authors declare no competing financial interest.
(36) Wang, X.; Feng, Z.; Shi, J.; Jia, G.; Shen, S.; Zhou, J.; Li, C. Phys.
Chem. Chem. Phys. 2010, 12, 7083−7090.
ACKNOWLEDGMENTS
■
This work was supported by a Grant-in-Aid for Scientific
Research (No. 23360357) from the Ministry of Education,
Culture, Sports, Science and Technology, Japan (MEXT).
(37) Shiraishi, Y.; Togawa, Y.; Tsukamoto, D.; Tanaka, S.; Hirai, T.
ACS Catal. 2012, 2, 2475−2481.
(38) Shiraishi, Y.; Hirakawa, H.; Togawa, Y.; Sugano, Y.; Ichikawa, S.;
Hirai, T. ACS Catal. 2013, 3, 2318−2326.
REFERENCES
(39) Bessac, F.; Frenking, G. Inorg. Chem. 2003, 42, 7990−7994.
(40) Lei, P.; Wang, F.; Zhang, S.; Ding, Y.; Zhao, J.; Yang, M. ACS
Appl. Mater. Interfaces 2014, 6, 2370−2376.
■
(1) Silverman, R. B. J. Am. Chem. Soc. 1981, 103, 3910−3915.
(2) Preusch, P. C.; Suttie, J. W. J. Org. Chem. 1983, 48, 3301−3305.
(3) Corey, E. J.; Su, W. G. J. Am. Chem. Soc. 1987, 109, 7534−7536.
(4) Kraus, G. A.; Thomas, P. J. J. Org. Chem. 1988, 53, 1395−1397.
(5) Johnson, W. S.; Plummer, M. S.; Reddy, S. P.; Bartlett, W. R. J.
Am. Chem. Soc. 1993, 115, 515−521.
(41) Lei, P.; Wang, F.; Gao, X.; Ding, Y.; Zhang, S.; Zhao, J.; Liu, S.;
Yang, M. J. Hazard. Mater. 2012, 227−228, 185−194.
(42) The following experimental results were obtained: (1)
Photoirradiation (>300 nm) of (Z)- or (E)-stilbene with TiO2 in
the presence of water under N2 does not promote any reactions; (2)
photoirradiation (>300 nm) of (Z)- or (E)-stilbene oxide with TiO2
under N2 in the presence of zeolite as a dehydration agent still
produces 1,2-diphenylethanol with yields similar to those obtained
without zeolite; (3) photoirradiation (>300 nm) of (Z)- or (E)-
stilbene oxide with TiO2 under N2 produces 1,2-diphenylethanol, even
at the initial stage of reaction, and the selectivity scarcely changes with
photoirradiation time. These data suggest that the water and stilbenes
produced by deoxygenation of stilbene oxides are not involved in the
formation of 1,2-diphenylethanol. As reported (refs 43 and 44),
photoexcitation of diphenylethane derivatives leads to abstraction of
hydrogen atoms from donor molecules, such as alcohols and amines.
In the present case, the ring-opened intermediate (Scheme 3A.b) is
probably photoexcited by short-wavelength UV light (∼<350 nm). Its
excited state may lead to abstraction of a hydrogen atom from 2-
PrOH, producing 1,2-diphenylethanol.
(6) Larock, R. C. Comprehensive Organic Transformations; Wiley: New
York, 1999; p 272.
(7) Mitsudome, T.; Mikami, Y.; Matoba, M.; Mizugaki, T.; Jitsukawa,
K.; Kaneda, K. Angew. Chem., Int. Ed. 2012, 51, 136−139.
(8) Mitsudome, T.; Noujima, A.; Mikami, Y.; Mizugaki, T.; Jitsukawa,
K.; Kaneda, K. Angew. Chem., Int. Ed. 2010, 49, 5545−5548.
(9) Mitsudome, T.; Noujima, A.; Mikami, Y.; Mizugaki, T.; Jitsukawa,
K.; Kaneda, K. Chem.Eur. J. 2010, 16, 11818−11821.
(10) Noujima, A.; Mitsudome, T.; Mizugaki, T.; Jitsukawa, K.;
Kaneda, K. Angew. Chem., Int. Ed. 2011, 50, 2986−2989.
(11) Ni, J.; He, L.; Liu, Y. M.; Cao, Y.; He, H. Y.; Fan, K. N. Chem.
Commun. 2011, 47, 812−814.
(12) Ke, X.; Sarina, S.; Zhao, J.; Zhang, X.; Chang, J.; Zhu, H. Chem.
Commun. 2012, 48, 3509−3511.
(13) Ke, X.; Zhang, X.; Zhao, J.; Sarina, S.; Barry, J.; Zhu, H. Green
Chem. 2013, 15, 236−244.
(43) Rajesh, C. S.; Givens, R. S.; Wirz, J. J. Am. Chem. Soc. 2000, 122,
(14) Li, Y.; Ji, H.; Chen, C.; Ma, W.; Zhao, J. Angew. Chem., Int. Ed.
2013, 52, 12636−12640.
611−618.
(44) Kandanarachchi, P. H.; Autrey, T.; Franz, J. A. J. Org. Chem.
2002, 67, 7937−7945.
(15) Maruska, H. P.; Ghosh, A. K. Sol. Energy 1978, 20, 443−458.
(16) RajanBabu, T. V.; Nugent, W. A.; Beattie, M. S. J. Am. Chem.
Soc. 1990, 112, 6408−6409.
(45) As shown in Table 2 (entries 16−19), the selectivity for (Z)-
and (E)-stilbenes depends on the irradiation wavelengths. This is due
to the photoisomerization of product stilbenes (ref 44). Under >300
nm irradiation conditions, (E)-stilbene exists mainly as a result of the
trans-to-cis photoisomerization (entries 16 and 18). In contrast, under
>370 nm irradiation conditions (entries 17 and 19), (Z)-stilbene exists
(17) RajanBabu, T. V.; Nugent, W. A. J. Am. Chem. Soc. 1994, 116,
986−997.
(18) Gansauer, A.; Bluhm, H. Chem. Rev. 2000, 100, 2771−2788.
̈
(19) Gansauer, A.; Pierobon, M.; Bluhm, H. Angew. Chem., Int. Ed.
̈
1998, 37, 101−103.
1648
dx.doi.org/10.1021/cs500065x | ACS Catal. 2014, 4, 1642−1649

ACS Catalysis
Research Article
mainly because stilbenes do not absorb light (Figure S5, Supporting
Information) and the Z form is thermodynamically more stable than
the E form.
(46) Arai, T.; Tokumaru, K. Chem. Rev. 1993, 93, 23−39.
(47) Tsukamoto, D.; Shiraishi, Y.; Sugano, Y.; Ichikawa, S.; Tanaka,
S.; Hirai, T. J. Am. Chem. Soc. 2012, 134, 6309−6315.
(48) Shiraishi, Y.; Sugano, Y.; Tanaka, S.; Hirai, T. Angew. Chem., Int.
Ed. 2010, 49, 1656−1660.
(49) Shiraishi, Y.; Sakamoto, H.; Sugano, Y.; Ichikawa, S.; Hirai, T.
ACS Nano 2013, 7, 9287−9297.
(50) Sugano, Y.; Shiraishi, Y.; Tsukamoto, D.; Ichikawa, S.; Tanaka,
S.; Hirai, T. Angew. Chem., Int. Ed. 2013, 52, 5295−5299.
(51) An, X.; Li, K.; Tang, J. ChemSusChem 2014, DOI: 10.1002/
cssc.201301194.
(52) Shiraishi, Y.; Tanaka, K.; Shirakawa, E.; Sugano, Y.; Ichikawa, S.;
Tanaka, S.; Hirai, T. Angew. Chem., Int. Ed. 2013, 52, 8304−8308.
(53) Limberg, C.; Koppe, R. Inorg. Chem. 1999, 38, 2106−2116.
̈
(54) Severcan, F.; Toyran, N.; Kaptan, N.; Turan, B. Talanta 2000,
53, 55−59.
(55) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R., Jr.; Montgomery, J. A.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J.
J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision B.05;
Gaussian, Inc.: Wallingford, CT, 2004.
1649
dx.doi.org/10.1021/cs500065x | ACS Catal. 2014, 4, 1642−1649