Visible light-induced Minisci reaction through photoexcitation of surface Ti-peroxo species

Article abstract of DOI:10.1039/d1cy00248a

Photocatalytic Minisci-type functionalization of pyridine with tetrahydrofuran (THF) proceeded using hydrogen peroxide (H₂O₂) and a TiO₂photocatalyst under acidic conditions. Under UV light (λ= 360 nm), the reaction selectivity based on pyridine (S_py) was >99% while the selectivity based on THF (S_THF) was low such as 19%. In contrast, under visible light (λ= 400 or 420 nm)S_pywas similarly high (>99%) andS_THFwas two times higher than that under UV light. A surface peroxo complex formed upon contact of hydrogen peroxide with the TiO₂surface can be selectively photoexcited by visible light to inject the photoexcited electron to the conduction band of TiO₂. The electron can reduce H₂O₂to a reactive oxygen species (ROS) and promote selectively the Minisci-type cross-coupling reaction between pyridinium ions and THF. A reaction test with a hole scavenger (methanol) evidenced that the hole oxidation of H₂O₂under UV light is responsible for the lower selectivity, in other words, the higher selectivity under visible light would be due to suppression of the hole oxidation of H₂O₂. These results demonstrate a novel way to improve the selectivity of the photocatalytic cross-coupling reaction by using H₂O₂as an oxidant with the photoexcitation of surface Ti-peroxo species on TiO₂

Full text of DOI:10.1039/d1cy00248a

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Visible light-induced Minisci reaction through
photoexcitation of surface Ti-peroxo species†
Cite this: DOI: 10.1039/d1cy00248a
ab
ab
Shimpei Naniwa, a Akira Yamamoto
and Hisao Yoshida
*
Photocatalytic Minisci-type functionalization of pyridine with tetrahydrofuran (THF) proceeded using
hydrogen peroxide (H2O2) and a TiO2 photocatalyst under acidic conditions. Under UV light (λ = 360 nm),
the reaction selectivity based on pyridine (Spy) was >99% while the selectivity based on THF (STHF) was low
such as 19%. In contrast, under visible light (λ = 400 or 420 nm) Spy was similarly high (>99%) and STHF was
two times higher than that under UV light. A surface peroxo complex formed upon contact of hydrogen
peroxide with the TiO2 surface can be selectively photoexcited by visible light to inject the photoexcited
electron to the conduction band of TiO2. The electron can reduce H2O2 to a reactive oxygen species
(ROS) and promote selectively the Minisci-type cross-coupling reaction between pyridinium ions and THF.
A reaction test with a hole scavenger (methanol) evidenced that the hole oxidation of H2O2 under UV light
is responsible for the lower selectivity, in other words, the higher selectivity under visible light would be
due to suppression of the hole oxidation of H2O2. These results demonstrate a novel way to improve the
selectivity of the photocatalytic cross-coupling reaction by using H2O2 as an oxidant with the
photoexcitation of surface Ti-peroxo species on TiO2.
Received 9th February 2021,
Accepted 11th March 2021
DOI: 10.1039/d1cy00248a
rsc.li/catalysis
reactions, which are one of the most important reactions to
build a carbon–carbon bond in organic molecules. Only a few
1. Introduction
Titanium
dioxide
(TiO2),
the
first
semiconductor
studies, however, have been reported for oxidative cross-
coupling reactions so far7,8 because of low selectivity. To
apply TiO2 photocatalysis to cross-coupling at high ROS
concentration, the selectivity of the coupling must be
improved.
photocatalyst, has attracted much attention as one of the
most practical photocatalysts with high oxidation ability,
chemical stability, and low cost.1,2 In a typical oxidation
process with a TiO2 photocatalyst, reactive oxygen species
(ROS) such as hydrogen peroxide, hydroxyl radicals, and
superoxide anion radicals, have been considered as key
intermediates,3,4 which are strongly correlated with the
efficiency of the system.5 For instance, hydrogen peroxide can
be used to accelerate the formation of hydroxyl radicals and
improve the activity in photocatalytic degradation of organic
wastes.6
One of the ways to improve the selectivity of TiO2
photocatalysis is to use photoexcitation of a ligand-to-metal
charge transfer (LMCT) complex consisting of adsorbed
molecules and the surface of a photocatalyst.9 LMCT complex
systems have achieved several selective photooxidation
reactions of various compounds such as alcohols,10
amines,11,12 and sulphides.13 Recently, we found that the
LMCT system can also be applied to sp2C–sp3C cross-coupling
between an aromatic ring and an alkane through the LMCT of
aromatic surface complexes.14,15 In the present study, we
utilized the LMCT excitation of a surface peroxo complex on
TiO2 to achieve high selectivity of photocatalytic organic
transformation involving the generation of ROS. We chose the
There is no doubt that these ROS formed with a TiO2
photocatalyst are available for not only degradation but also
other synthetic reactions such as oxidative cross-coupling
a Graduate School of Human and Environmental Studies, Kyoto University, Yoshida
Nihonmatsu-cho, Sakyo-ku, Kyoto 606-8501, Japan.
E-mail: yoshida.hisao.2a@kyoto-u.ac.jp
Minisci-type
functionalization
of
pyridine
with
b Elements Strategy Initiative for Catalysts & Batteries (ESICB), Kyoto University,
1-30 Goryo-Ohara, Nishikyo-ku, Kyoto 615-8245, Japan
tetrahydrofuran (THF) by using hydrogen peroxide under
acidic conditions as the first trial (eqn (1)).8 The effect of
irradiation wavelength on the reaction was investigated and
† Electronic supplementary information (ESI) available: Experimental details,
information, and some results. See DOI: 10.1039/d1cy00248a
This journal is © The Royal Society of Chemistry 2021
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
several scavenging tests were also carried out to clarify the
reaction mechanism. We found that that selective excitation
of the peroxo surface complex by visible light can improve the
selectivity of cross-coupling by suppressing the undesired
oxidation of H2O2.
filter to remove the photocatalyst, and then analyzed by GC-MS
(Shimadzu, GCMS-QP2020). The analysis was carried out
without neutralization.
The desired product, tetrahydrofuranylpyridine (TP), was
quantified by using
a
GC-MS calibration curve of
2-cyclohexylpyridine. Three peaks for the isomers of TP were
obtained and their peak areas were always in a ratio of 3 : 1 :
1, the sum of which was treated as the desired products (TPs)
in the present study and used to calculate the TP yield in the
reaction tests. By-products observed were some oxidized THF
(1)
2. Experimental
2.1 Catalysts
(OTs:
2-hydroxytetrahydrofuran,
γ-butyrolactone,
and
α-hydroxy-γ-butyrolactone), which were quantified using the
calibration curve of authentic samples purchased from
companies. The remaining H2O2 in the reaction mixture was
quantified by iodometry (ESI†).
Three TiO2 powder samples donated by the Catalysis Society
of Japan were employed: JRC-TIO-6 (rutile phase, 100 m2 g−1),
JRC-TIO-14 (anatase phase, 338 m2 g−1), and JRC-TIO-4 (a
mixture of rutile and anatase phases, 50 15 m2 g−1).
In some experiments, a Pd-loaded TiO2 catalyst prepared by a
photodeposition method (ESI†) was used since our previous
studies showed the positive effect of a Pd cocatalyst on some
organic reactions.16–18 The catalysts were referred to as Pd(x)/
TiO2, where x indicates the loading amount of the metal in wt%.
2.3 Diffuse reflectance UV-vis spectroscopy
UV-vis spectra of the powder samples were recorded in a
diffuse reflectance (DR) mode. The samples were diluted 100
times with BaSO4. For the samples with an adsorbate, 50 μL
of the reaction mixture (a mixture of pyridine, THF, TFA, and
H2O2 aq.) or 30% H2O2 aqueous solution was added to the
diluted samples, followed by 10 min of mixing. Then, the
desired amount of the sample was taken into the cell so as to
fix the amount of TiO2 in the samples and the spectrum was
measured using a UV-vis spectrophotometer (JASCO V-570)
equipped with an integrating sphere, where BaSO4 was used
as a reference.
2.2 Reaction test
2.2.1 Materials. All the chemicals employed for the
photocatalytic reaction tests were of analytical grade and were
used without further purification: pyridine (Nacalai Tesque,
99.7%), tetrahydrofuran (THF, Nacalai Tesque, >99.5%),
hydrogen peroxide (Nacalai Tesque, 30%), trifluoroacetic acid
(TFA, Santa Cruz Biotechnology, >99%), difluoroacetic acid
(Aldrich, 98%), acetic acid (Nacalai Tesque, 99%),
trifluoroethanol (Tokyo Chemical Industry, >99%), and
tribromoacetic acid (Aldrich, 99%).
3. Results and discussion
Several experiments showed that a Pd(0.1)/TiO2(rutile) sample
was the best photocatalyst among the Pd-loaded samples
(Tables S4 and S5, ESI†). The Pd cocatalyst with an optimal
loading amount (0.1 wt%) increased the photocatalytic
production of the aimed coupling products (1.8 times) with a
slightly lower selectivity (Table S4†). However, to simplify the
discussion for spectroscopic results and the reaction
mechanism, the results of bare TiO2(rutile) were mainly
shown rather than the results with Pd/TiO2 samples (Tables
S1–S6, and Fig. S3†) in the present study.
After performing optimization experiments for additives,
H2O2 and acid (Tables S1–S3, ESI†), 0.2 mmol of H2O2 and
0.2 mmol of TFA were employed for further experiments.
2.2.2 Procedure for the reaction tests. The typical
procedure is as follows. Before a photocatalytic reaction test, a
photocatalyst sample (TiO2 or Pd(0.1)/TiO2, 50 or 100 mg) in a
Pyrex glass tube (20 or 50 mL) was subjected to pre-treatment
for 30 min under the light from a ceramic xenon lamp
(PE300BUV, 300 W) to clean its surface. Then, a reaction
mixture (2.05 mL in total) of pyridine, THF, H2O2 aqueous
solution (30%), and TFA was added into the test tube. The
resultant mixture was a yellowish suspension (Fig. S1, ESI†).
After that, the test tube was sealed with a rubber septum and
the mixture was bubbled by argon for 10 min under magnetic
stirring to remove air. The reaction mixture was irradiated by
using another 300 W xenon lamp (Asahi Spectra Co., Ltd. MAX-
302) with stirring for a desired reaction time at room
temperature. The irradiation wavelength was limited by a band-
pass filter (MX0360 (λ = 360 10 nm), MX0400 (λ = 400 10
nm), MX0420 (λ = 420 10 nm)). After irradiation, 0.5 mL of the
gaseous phase was collected using an air-tight syringe and
analyzed by GC-TCD (Shimadzu, GC-8A). The reaction mixture
in the liquid phase was sampled using a syringe with a PTFE
3.1 Diffuse reflectance UV-vis spectroscopy
The visible-light responsive species in this system was first
confirmed by diffuse reflectance UV-vis spectroscopy (Fig. 1).
The TiO2 (rutile) sample in the presence of the reaction
mixture (pyridine, THF, and TFA) without H2O2 was white
and its spectrum (Fig. 1b) was almost similar to that of the
sample without the reaction mixture (Fig. 1a), where the
increase in the baseline was probably due to the liquid
phase. This suggests that a reported surface complex formed
with an acid–base interaction between a pyridine molecule
and a Ti cation on the TiO2 surface15 is not the case under
the present acidic conditions. This would be due to
protonation of pyridine molecules by the acid (TFA). In
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Catalysis Science & Technology
Paper
(λ
=
360
10 nm) and afforded 1.8 μmol of
tetrahydrofuranylpyridine (TP) as the cross-coupling product,
which corresponds to 0.85% yield in 1 h (Table 1, entry 1).
By-products from pyridine such as 2,2′-bipyridyl, 2,3′-
bipyridyl, and 2,4′-bipyridyl were not observed. This means
that the selectivity to TPs based on pyridine (Spy) was >99%.
On the other hand, 6.8 μmol of some oxidized by-products of
THF (OTs) was obtained, in which the selectivity to TPs based
on THF (STHF) was low such as 19% even in the initial stage
of the reaction.
When the reaction test was carried out under visible light (λ
= 400 10 nm, Table 1, entry 2), the amount of the products was
higher than that under UV light due to higher light intensity
from the lamp; 8.6 μmol of TPs and 15 μmol of OTs were
produced. The yield and STHF were five and two times higher
than that under UV-light such as 4.3% and 37%, respectively.
The TP selectivity based on pyridine was as high as the case
under UV light (Spy > 99%). The reaction proceeded even under
longer-wavelength light (λ = 420 10 nm) and the selectivity
obtained was the highest (such as 40%) among them (Table 1,
entry 3). The higher STHF under visible light than that under UV
light means that the formation of OTs was suppressed under
visible light irradiation. The amount of H2O2 in the solution
was decreased after reaction in all cases. The yield of TPs was
drastically decreased in the absence of H2O2 under both UV
and visible light (Table 1, entries 4 and 5), which indicates that
H2O2 as an oxidant is essential for the production of TPs. The
reaction also hardly proceeded in an air atmosphere without
H2O2 under UV light (Table 1, entry 6). TPs were not produced
in the absence of TFA under both light conditions (Table 1,
entries 7 and 8), suggesting that TPs are formed through a
Minisci-type mechanism, which involves protonation of
Fig. 1 DR UV-vis spectra of (a) the rutile TiO2 sample diluted with
BaSO4, (b) the diluted sample with the reaction mixture (pyridine, THF,
TFA) except for H2O2, and (c) that with the reaction mixture including
H2O2.
contrast, the H2O2 containing sample was yellowish and its
spectrum showed an additional visible-light absorption band
in the wavelength range of 400–500 nm compared with the
other two samples (Fig. 1c). This is attributed to the presence
of surface Ti-peroxo species.19
It is reported that H2O2 is adsorbed on TiO2 to form some
surface Ti-peroxo species and changes the catalyst's color into
yellow,19 which was confirmed also on the three TiO2
samples employed here (Fig. S2†). Three types of surface
peroxo species have been proposed: end-on, bridged,20 and
side-on (Scheme 1).21,22 The former two species exist mainly
on anatase TiO2, while the latter does on rutile TiO2.21 The
color change by H2O2 adsorption was also observed in zeolitic
titanium-silicate systems23 and the side-on peroxo is regarded
as the visible-light responsive species because of its similar
spectroscopic features with the model compound.24 Thus, it
is suggested that the visible-light responsive species in the
present system with the rutile TiO2 sample is the side-on
species. The end-on species is another candidate since the
side-on species would be protonated under acidic conditions
in the presence of TFA (eqn (2)). Since TiO2 is excited by UV
light only, the reaction under visible light would start from
the photoexcitation of the surface peroxo species.
pyridine by TFA, an attack of
a THF radical to a
pyridinium ion (discussed below), and subsequent
elimination of a hydrogen atom (Scheme 2). The active
species produced from H2O2 could promote the production
of THF radicals.
In these reaction tests (Table 1, entries 1–3), the amount
of consumed H2O2 was much larger than the product yield.
Thus, it must be noted that H2O2 molecules and the surface
peroxo species would largely decompose during the reaction.
We carried out a long-term reaction test under visible light
(Fig. 2), where the reaction scale was two times larger than
usual. The reaction stopped and the yellowish color of the
catalyst disappeared after 8 h when the yield of TPs and
conversion of H2O2 reached 14% and 81%, respectively. As
suggested from the result of DR UV-vis spectroscopy, the
reaction under visible light would start not from the
photoexcitation of the TiO2 photocatalyst but the
photoexcitation of the surface Ti-peroxo species. The
termination of the reaction with decolorization also indicates
that the surface peroxo species is necessary to promote the
reaction under visible-light irradiation, i.e., the reaction
cannot proceed without the surface peroxo species. Since the
peroxo species would be present on the surface in
equilibrium with the H2O2 molecules in the liquid phase,23 it
O2–Ti (side‐on) + H+ ⇄ HOO–Ti (end‐on)
(2)
3.2 Reaction test
Next, reaction tests were carried out using the pristine rutile
TiO2 sample (JRC-TIO-6) under the light of variously limited
wavelength (Table 1). The reaction took place under UV light
Scheme 1 Possible structures of surface Ti-peroxo species: (a) end-
on, (b) bridged, and (c) side-on.
This journal is © The Royal Society of Chemistry 2021
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Catalysis Science & Technology
Paper
Scheme 3 Photoexcitation of side-on peroxo species.
Scheme 2 Minisci reaction between pyridine and THF in the presence
of TFA. 2,2′-Tetrahydrofuranylpyridine was chosen as a representative
example for TPs.
the reactions under visible light. The hole (O˙) would also
activate the substrates or decompose the surface Ti-peroxo
species. The higher selectivity under visible light than that
under UV light suggests the presence of the decomposition
pathway before the oxidation of the substrates to OTs.
should almost disappear when the H2O2 concentration in the
liquid phase decreased. As another possibility, the surface
peroxo species might decompose during the reaction. An
additional long-term reaction test was carried out with a Pd
loaded sample (Fig. S3, ESI†), followed by the addition of
H2O2 after 12 h. It was found that the catalyst's color and
activity were restored with the addition of 0.4 mmol of H2O2
after the reaction termination. This indicates that the surface
Ti-peroxo species can be regenerated when H2O2 was
supplied. However, further addition of H2O2 after 21 h did
not improve the yield but promoted a successive reaction to
give several bifunctionalized pyridines with two THF
molecules (Fig. S3, ESI†). This means that the products
should be separated before the successive reaction takes
place to obtain a higher yield.
Under UV light, TiO2 is excited to generate photogenerated
electrons and holes in TiO2, which can activate the
substrates. The photogenerated electrons and holes would
react with H2O2 and THF to produce radical species such as a
hydroxyl radical (˙OH), a hydroperoxyl radical (HO2˙), and a
THF radical (˙C4H7O) (eqn (3)–(5)). Thus, under UV light
conditions, these radical species would form some kinds of
oxidation products (OTs) as the major by-products.
Meanwhile, hole oxidation of pyridine would be difficult
since pyridine molecules are in a cationic form. The hydroxyl
radical would also produce a THF radical (eqn (6)) while the
hydroxyl radical and the hydroperoxyl radical accelerate the
production of 2-hydroxytetrahydrofuran (eqn (7) and (8)), of
which subsequent oxidation gives other OT species.
Under visible light, ligand-to-metal-charge-transfer (LMCT)
takes place in the Ti-peroxo species to transfer an electron
from its peroxo oxygen (O−) to the Ti(IV) cation, producing O˙
(hole) and the Ti(III) cation (electron) in the surface complex
(Scheme 3).23 The photoexcited electron would be further
transferred to the conduction band of TiO2 to react with the
substrates. It is reported that visible-light excitation of
surface Ti-peroxo species produces a hydroxyl radical in the
H2O2–TiO2 system through the reduction of a H2O2 molecule
in the liquid phase by the photoexcited electron (eqn (3)).25
Therefore, it is also possible that the photogenerated electron
formed from the Ti-peroxo species through LMCT excitation
reduces another H2O2 molecule to produce a hydroxyl radical.
This hydroxyl radical would be the major active species for
H2O2 + e− → ˙OH + OH−
H2O2 + h+ → HO2˙ + H+
(3)
(4)
(5)
(6)
(7)
(8)
C4H8O (THF) + h+ → ˙C4H7O + H+
C4H8O (THF) + ˙OH → ˙C4H7O + H2O
˙C4H7O + ˙OH → C4H8O2(OTs)
˙C4H7O + HO2˙ → C4H8O2(OTs) + 1/2O2
When the reaction test was carried out in the dark, OTs
were initially produced with the consumption of H2O2
(Table 1, entry 9) but the amount of OTs did not increase
with time (Table 1, entry 10). On the other hand, the
amounts of OTs and consumed H2O2 were larger in the
absence of TFA in the dark (Table 1, entry 11) and they
increased with time (Table 1, entry 12). This means that
OTs can be produced upon decomposition of H2O2, and
TFA suppresses the decomposition. OTs would be produced
from a reaction between THF and ˙OH adsorbed on the
TiO2 surface.26 TFA would suppress the decomposition of
H2O2 by poisoning the active sites on the TiO2 surface. The
suppression effect was also observed in the reaction under
UV light (Table 1, entries 1 and 7) or visible light (Table 1,
entries 3 and 8). Since TFA also promotes the production of
TPs as mentioned above, the roles of TFA in the present
system are as follows: protonation of pyridine for the
Minisci reaction and suppression of thermal decomposition
of H2O2.
Fig. 2 Time course of the Minisci reaction between pyridine and THF
using the pristine rutile TiO2 (JRC-TIO-6, 100 mg). Pyridine (0.4 mmol),
THF (48 mmol), H2O2 (0.4 mmol), and TFA (0.4 mmol) were used. The
irradiation wavelength was λ = 420 10 nm. The reaction tests were
carried in a Pyrex glass tube (50 mL).
This journal is © The Royal Society of Chemistry 2021
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
not observed. Methanol can consume the hole efficiently and
suppress the oxidation of the substrates sacrificially. The
increase in the selectivity with the suppression of H2O2
consumption indicates that the oxidation of H2O2 (eqn (4)) is
responsible for the formation of by-products. In addition, the
yield of TPs was not changed so much by the addition of
methanol. It indicates that methanol consumed only holes
on TiO2 selectively and the photoexcited electrons were used
for TP formation (eqn (3)).
Fig. 3 Difference of the photogenerated hole between the band-to-
band transition in TiO2 (A) and the LMCT transition in the surface Ti-
peroxo species (B).
Instead of methanol, the addition of 2-propanol was
examined under UV light irradiation. 2-Propanol is often used
as a hydroxyl radical scavenger while in some cases it is
regarded as a hole scavenger. As a result, the amount of
obtained OTs increased from 6.9 to 11 (Table 2, entry 4), which
is opposite to the case of methanol addition. It is clarified that
2-propanol cannot act as a hole scavenger under the present
conditions although the reason for the increase was uncovered.
On the other hand, the obtained TPs decreased from 1.8 to 1.2
μmol (Table 2, entry 4), meaning that 2-propanol reacted with a
hydroxyl radical in this system in competition with THF. This
also confirms that the hydroxyl radical participates in the
formation of TPs; the hydroxyl radical would produce a THF
radical which attacks a protonated pyridine. Therefore, it is
suggested that the THF radical is generated mainly from the
reaction of THF with the hydroxyl radical (eqn (6)), not with the
photogenerated hole (eqn (5)). This would be due to the high
reactivity of THF with the hydroxyl radical. The rate constant for
the reaction with the hydroxyl radical in aqueous solution is
higher for THF (4.0 × 109) than that for 2-propanol (1.9 × 109),29
which is a well-known hydroxyl radical scavenger and used in
the present study. Therefore, THF would mainly react with the
hydroxyl radical before reacting with the photogenerated hole.
Scavenging tests were also carried out under visible light
to obtain further insight into the reaction mechanism
(Table 3). The yield of TPs decreased with increasing the
amount of 2-propanol (Table 3, entries 1–3) as in the case
under UV-light, confirming that the hydroxyl radical activates
THF to form TPs. The yield of TPs also decreased with
increasing the amount of an electron scavenger, AgNO3
aqueous solution (Table 3, entries 1 and 4–6), while the
3.3 Scavenging test
UV light irradiation induces band-to-band transition in TiO2
(Fig. 3A) while visible light irradiation promotes LMCT
transition in the surface Ti-peroxo species (Fig. 3B). In both
cases, the photoexcited electron is generated in the
conduction band of TiO2 while the hole is generated in the
valence band of TiO2 in the former case and on the surface
peroxo species in the latter case. The presence of the
photoexcited electrons in the conduction band of TiO2 is
necessary and enough to promote the aimed coupling
reaction since they can activate hydrogen peroxide to start
the reaction through the formation of a hydroxyl radical and
then a THF radical, as mentioned above. The properties of
the excited electron are common in both cases. Thus, the
different selectivities under different wavelength-light would
be due to the different properties of the photogenerated hole.
If that is the case, the selectivity under UV light might be
improved by suppressing the hole oxidation of the substrates.
Therefore, we carried out some reaction tests under UV light
in the presence of methanol as a hole scavenger (Table 2,
entries 2 and 3). As expected, the production of OTs was
suppressed with increasing the amount of methanol, leading
to higher selectivity such as 46% with 16 mmol of methanol
(Table 2, entry 3). H2O2 consumption was also suppressed by
the addition of methanol. Although alcohols can also
promote Minisci-coupling,27,28 such kinds of products were
Table 2 Effect of a hole scavenger (MeOH) and a hydroxyl radical scavenger (2-PrOH) on the product yield and the amount of consumed H2O2 after
the reaction under UV lighta
Productsb/μmol
Selectivityd (%)
Consumed
H2O2 /μmol
c
Entry
Scavenger/mmol
TPs
OTs
Spy
STHF
1
2
3
4
—
—
8
16
8
1.8
1.7
1.9
1.2
6.9
4.7
2.2
11
63
57
7
>99
>99
>99
>99
19
27
46
10
MeOH
MeOH
2-PrOH
67
a
Reaction conditions: pyridine (0.2 mmol), THF (24 mmol), 30% H2O2 aq. (22.5 μL, 0.2 mmol of H2O2), TFA (0.2 mmol), and a scavenger
10 nm, and reaction time was 1 h. The
with the TiO2 photocatalyst (JRC-TIO-6, 50 mg) were used, irradiation wavelength was 360
reaction tests were carried out in a Pyrex test tube (20 mL). The light intensity was measured at 355 45 nm in wavelength. TPs: total
amount of tetrahydrofuranylpyridine. OTs: total amount of 2-hydroxytetrahydrofuran, γ-butyrolactone, and α-hydroxy-γ-butyrolactone. The
b
c
d
amount of consumed H2O2 after the reaction calculated as (amount of added H2O2 (μmol)) − (amount of H2O2 in the solution after the
reaction (μmol)).
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Catalysis Science & Technology
Paper
Table 3 Effect of a hydroxyl radical scavenger (2-PrOH), an electron scavenger (Ag+), and a hole scavenger (MeOH) on the product yield and the
amount of consumed H2O2 after the reaction under visible lighta
Productsb/μmol
Selectivityd (%)
Consumed
H2O2 /μmol
c
Entry
Scavenger/mmol
TPs
OTs
Spy
STHF
1
2
3
—
—
4
8
0.03
0.1
0.2
—
8
6.9
4.7
3.2
4.8
1.6
n.d.
8.0
2.8
10
11
82
75
74
84
86
78
92
67
>99
>99
>99
>99
>99
0
40
29
26
54
26
0
2-PrOH
2-PrOH
AgNO3
AgNO3
AgNO3
—
9.1
4.1
4.6
5.8
7.4
3.1
4e
5e
6e
7f
8
>99
>99
52
48
MeOH
a
b
Reaction conditions: Irradiation wavelength was 420 10 nm. Other conditions were the same as those described in Table 2. The light
c
intensity was measured at 415
55 nm in wavelength. TPs: total amount of tetrahydrofuranylpyridine. OTs: total amount of
d
2-hydroxytetrahydrofuran, γ-butyrolactone, and α-hydroxy-γ-butyrolactone. The amount of consumed H2O2 after the reaction calculated as
e
f
(amount of added H2O2 (μmol)) − (amount of H2O2 in the solution after the reaction (μmol)). 0.15 M AgNO3 aq. was used. Same amount of
water as that of entry 4 (0.2 mL) was added.
addition of the same amount of water did not decrease the
yield of TPs (Table 3, entry 7). This means that the decrease
in the yield of TPs by the addition of AgNO3 aqueous solution
is due to electron scavenging by the Ag+ cation. Thus, it is
also confirmed that the hydroxyl radical is produced through
the reduction of H2O2 by the photogenerated electron and
activates THF molecules under visible light.
The addition of 0.03 mmol of AgNO3 also decreased OT
production from 10 to 4.1 μmol (Table 3, entries 1 and 4),
suggesting that the photoexcited electron also produces OTs.
However, it started to increase slightly by a further increase of
AgNO3 (Table 3, entries 5 and 6) and 5.8 μmol of OTs was still
observed with the addition of 0.2 mmol of AgNO3 (Table 3, entry
6), which completely inhibited the production of TPs. Since OTs
can be produced in the dark (Table 1, entry 8), the OTs observed
under these conditions would be produced by dark reactions
between THF and H2O2 thermally catalyzed by TiO2 or Ag metal
species photodeposited on TiO2.
The reaction test under visible light was also carried out
using a hole scavenger (Table 3, entry 8). If hole oxidation of
H2O2 takes place under visible light, the addition of the hole
scavenger would improve the selectivity as with the reaction
under UV light. However, the selectivity was hardly improved;
the amounts of both TPs and OTs were decreased, which is
different from the UV-light case. This different effect of the
hole scavenger under visible light with that under UV light
supports the conclusion in the reaction tests under the light
of variously limited wavelength: hole oxidation of H2O2 takes
place to a great extent under UV light while it seems less
major under visible light. The decrease in the amounts of
both products would be due to inhibition of the adsorption
of H2O2 on TiO2 by methanol.
Finally, reaction tests were carried out using quinoline
instead of pyridine to investigate the generality of the present
system (Table S6, ESI†). Derivatives of quinoline are known
as important molecules in the field of pharmacy. The
reaction successfully proceeded to afford mainly two isomers
of tetrahydrofuranylquinoline (TQ) in a ratio of 3 : 2. The
selectivities to TQs under UV light were high such as >99%
based on quinoline (Squino) and 84% based on THF (STHF) (λ
= 360 10 nm, Table S6, entry 1, ESI†), suggesting that THF
radicals are selectively used for the functionalization of
quinoline even under UV light irradiation. Thus, the use of
visible light to improve the selectivity was not so effective;
Squino and STHF under visible light (λ = 400 10 nm, Table S6,
entry 2, ESI†) were >99% and 87%, respectively. The yield in
2 h was 14% under UV light and 20% under visible light,
which is much higher than that of the reaction with pyridine
(Table 1, entries 1 and 2). Thus, it is noticed that the present
reaction system using visible light has a certain limitation for
available substrates.
3.4 Reaction mechanism
Based on all of the results discussed above, a tentative
reaction mechanism is proposed as follows.
Fig. 4 Schematic illustration of the reaction mechanism under UV
light (A), visible light (B), and in the dark (C).
This journal is © The Royal Society of Chemistry 2021
Catal. Sci. Technol.

View Article Online
Paper
Under UV light (Fig. 4A), TiO2 is excited to generate the
Catalysis Science & Technology
Promotion of Science (JSPS), and the Program for Element
photoexcited electron and hole in the conduction and valence
bands, respectively. The electron and hole migrate to the
surface to react with the substrates. The former reduces
hydrogen peroxide to produce a hydroxyl radical (eqn (3))
while the latter oxidizes hydrogen peroxide into radical
species such as a hydroperoxo radical (eqn (4)). The hydroxyl
radical subtracts a hydrogen atom from the THF molecule to
produce a THF radical (eqn (5)) and this radical attacks the
pyridinium ion protonated by TFA. Removal of a hydrogen
atom from the transition state gives TPs. The hydroperoxo
radical produces OTs by reacting with THF (eqn (6)).
Under visible light (Fig. 4B), the surface Ti-peroxo species
absorbs light via LMCT excitation from the peroxy oxygen
atom to the Ti atom. The electron on the Ti atom is further
transferred to the conduction band of TiO2 and reduces
hydrogen peroxide into a hydroxyl radical, which produces a
THF radical and finally TPs. This reaction in TP formation is
the same as in the UV-light case. The hole, on the other
hand, is generated on the peroxy oxygen. This positively-
charged peroxy oxygen would decompose although it also
oxidizes the substrates. The presence of the decomposition
pathway suppresses the oxidation of H2O2 and thus the
formation of OTs, leading to higher TP selectivity (STHF).
In both cases, TFA promotes the production of TPs and
suppresses the thermal decomposition of H2O2.
Strategy Initiative for Catalysts
JPMXP0112101003) commissioned by the MEXT of Japan.
&
Batteries (ESICB,
Notes and references
1 K. Nakata and A. Fujishima, J. Photochem. Photobiol., C,
2012, 13, 169.
2 J. Schneider, M. Matsuoka, M. Takeuchi, J. Zhang, Y.
Horiuchi, M. Anpo and D. W. Bahnemann, Chem. Rev.,
2014, 114, 9919.
3 T. Hirakawa and Y. Nosaka, Langmuir, 2002, 18, 3247.
4 T. Hirakawa, K. Yawata and Y. Nosaka, Appl. Catal., A,
2007, 325, 105.
5 M. Cho, H. Chung, W. Choi and J. Yoon, Water Res.,
2004, 38, 1069.
6 A. P. Rivera, K. Tanaka and T. Hisanaga, Appl. Catal., B,
1993, 3, 37.
7 T. Caronna, C. Gambarotti, L. Palmisano, C. Punta, M.
Pierini and F. Recupero, J. Photochem. Photobiol., A,
2007, 189, 322.
8 T. Caronna, C. Gambarotti, L. Palmisano, C. Punta and F.
Recupero, J. Photochem. Photobiol., A, 2005, 171, 237.
9 G. Zhang, G. Kim and W. Choi, Energy Environ. Sci., 2014, 7,
954.
10 S. Higashimoto, N. Kitao, N. Yoshida, T. Sakura, M. Azuma,
H. Ohue and Y. Sakata, J. Catal., 2009, 266, 279.
11 S. Higashimoto, Y. Hatada, R. Ishikawa, M. Azuma, Y. Sakata
and H. Kobayashi, Curr. Org. Chem., 2013, 17, 2374.
12 X. Lang, W. Ma, Y. Zhao, C. Chen, H. Ji and J. Zhao, Chem. –
Eur. J., 2012, 18, 2624.
4. Conclusion
LMCT excitation of surface Ti-peroxo species can be used for
the selective Minisci-type functionalization of pyridine. Both
H2O2 and TFA are necessary to promote this reaction. The Ti-
peroxo species decompose after photoexcitation, and this
vulnerability of the photogenerated hole suppresses the
oxidation of H2O2, leading to higher selectivity.
13 X. Lang, W. Hao, W. R. Leow, S. Li, J. Zhao and X. Chen,
Chem. Sci., 2015, 6, 5000.
14 A. Yamamoto, T. Ohara and H. Yoshida, Catal. Sci. Technol.,
2018, 8, 2046.
The yield in the present system is not satisfactory and
thus a further study is required to improve the yield by some
methods such as continuous addition of H2O2 with
separation of the products from the system. Nevertheless,
this study provided a unique way to improve the selectivity of
photocatalysis involving ROS generation. We anticipate the
present study to trigger studies to expand the application of
aerobic photocatalysis as an efficient and selective organic
reaction system.
15 S. Naniwa, A. Tyagi, A. Yamamoto and H. Yoshida, Phys.
Chem. Chem. Phys., 2018, 20, 28375.
16 A. Tyagi, T. Matsumoto, T. Kato and H. Yoshida, Catal. Sci.
Technol., 2016, 6, 4577.
17 H. Yoshida, Y. Fujimura, H. Yuzawa, J. Kumagai and T.
Yoshida, Chem. Commun., 2013, 49, 3793.
18 E. Wada, T. Takeuchi, Y. Fujimura, A. Tyagi, T. Kato and H.
Yoshida, Catal. Sci. Technol., 2017, 7, 2457.
19 A. H. Boonstra and C. A. H. A. Mutsaers, J. Phys. Chem.,
1975, 79, 1940.
20 R. Nakamura and Y. Nakato, J. Am. Chem. Soc., 2004, 126,
1290.
21 T. Ohno, Y. Masaki, S. Hirayama and M. Matsumura,
J. Catal., 2001, 204, 163.
22 Y. Shiraishi, S. Kanazawa, D. Tsukamoto, A. Shiro, Y. Sugano
and T. Hirai, ACS Catal., 2013, 3, 2222.
Conflicts of interest
The authors declare no competing financial interests.
Acknowledgements
This work was financially supported by ISHIZUE 2020 of the
Kyoto University Research Development Program, the joint
research program of the Artificial Photosynthesis, Osaka City
23 F. Bonino, A. Damin, G. Ricchiardi, M. Ricci, G. Spanò, R.
D'Aloisio, A. Zecchina, C. Lamberti, C. Prestipino and S.
Bordiga, J. Phys. Chem. B, 2004, 108, 3573.
24 S. Bordiga, A. Damin, F. Bonino, G. Ricchiardi, C. Lamberti
and A. Zecchina, Angew. Chem., Int. Ed., 2002, 41, 4734.
University, and
(19K15359),
a
Grant-in-Aid for Young Scientists
a
Grant-in-Aid for Challenging Research
(Exploratory, 20K21108) from the Japan Society for the
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Catalysis Science & Technology
Paper
25 X. Li, C. Chen and J. Zhao, J. Am. Chem. Soc., 2001, 17, 4118.
26 C. M. Lousada, A. J. Johansson, T. Brinck and M. Jonsson,
J. Phys. Chem. C, 2012, 116, 9533.
28 A. Vijeta and E. Reisner, Chem. Commun., 2019, 55,
14007.
29 G. V. Buxton, C. L. Greenstock, W. P. Helman and A. B. Ross,
J. Phys. Chem. Ref. Data, 1988, 17, 513.
27 J. Jin and D. W. C. MacMillan, Nature, 2015, 525, 87.
This journal is © The Royal Society of Chemistry 2021
Catal. Sci. Technol.