Robust Organic Photosensitizers Immobilized on a Vinylimidazolium Functionalized Support for Singlet Oxygen Generation under Continuous-Flow Conditions

Article abstract of DOI:10.1055/s-0039-1691582

Rose Bengal was immobilized on a vinylimidazolium functionalized support, and the heterogeneous organic photosensitizer thus prepared was applied for photooxidation reactions of organic molecules under continuous-flow conditions. Substituents of the cation part of the support were found to play a crucial role in determining the lifetime of the catalyst. More than 11 days continuous operation of a flow reaction was achieved.

Full text of DOI:10.1055/s-0039-1691582

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
©
Georg Thieme Verlag Stuttgart · New York
2020, 31, A–E
letter
A
en
Synlett
K. Masuda et al.
Letter
Robust Organic Photosensitizers Immobilized on a Vinylimidazoli-
um Functionalized Support for Singlet Oxygen Generation under
Continuous-Flow Conditions
Koichiro Masudaa,1
Yao Wanga,b
White LED light
O2 (Air)
Shun-ya Onozawa*b
Shigeru Shimadab
O
Immobilized catalyst
O
High yield
Nagatoshi Koumurab
Kazuhiko Satob
Continuous operation
>
300 h
Shū Kobayashi*a,b
0
0
0
0-
0
0
0
2-8
2
3
5-4
3
6
8
a
Department of Chemistry, School of Science, The University of Tokyo, 7-3-1 Hongo,
Bunkyo-ku, Tokyo, 113-0033, Japan
shu_kobayashi@chem.s.u-tokyo.ac.jp
b
Interdisciplinary Research Center for Catalytic Chemistry, National Institute of Ad-
vanced Industrial Science and Technology (AIST), Central 5, 1-1-1 Higashi, Tsukuba,
Ibaraki, 305-8565, Japan
s-onozawa@aist.go.jp
Received: 11.11.2019
Accepted after revision: 06.01.2020
Published online: 06.02.2019
practical advantages, a characteristic of heterogeneous sen-
sitizers is that the transparency of the liquid phase and
scattering on solid particles enhances the photo-harvesting
efficiency. Given their high photo-efficiency and productiv-
ity, flow-type reactors have also been studied for photosen-
DOI: 10.1055/s-0039-1691582; Art ID: st-2019-b0608-l
Abstract Rose Bengal was immobilized on a vinylimidazolium func-
tionalized support, and the heterogeneous organic photosensitizer thus
prepared was applied for photooxidation reactions of organic molecules
under continuous-flow conditions. Substituents of the cation part of
the support were found to play a crucial role in determining the lifetime
of the catalyst. More than 11 days continuous operation of a flow reac-
tion was achieved.
sitization or photocatalysis with homogeneous or heteroge-
neous sensitizers.1
actions,
1–14
Especially for the singlet oxygen re-
15,16
dye-immobilized
membrane-type reactors, and packed-bed systems
microfluidic
reactors,
17
18–21
have been described to date. However, previous approaches
to the development of immobilized/heterogeneous sensi-
tizers have suffered from the instability of the sensitizers.
The history of development is similar to that of more stable
and more specialized molecules that have been designed to
obtain higher durability. On the other hand, durability en-
hancement by modification of molecular structure often
limits the availability of structure and increases the cost of
synthesis. It would therefore be beneficial to enhance the
stability of the dyes by either applying immobilization
methods or by using supporting materials. In this article,
we report the application of new anion-exchange resins for
the immobilization of Rose Bengal (Figure 1) for use in sin-
glet oxygen reactions, and describe the implementation of
this system in a continuous-flow oxidation that was con-
ducted for more than 11 days.
Rose Bengal was chosen as the photosensitizer to be im-
mobilized, and several commercial ion-exchange resins
were examined as catalyst supports. Rose Bengal was im-
mobilized by applying a simple ion exchange in water. The
prepared heterogeneous sensitizers were examined by pho-
tooxidation reactions of -terpinene to ascaridole 1 under
batch conditions to confirm their catalytic activity and
properties.
Key words photooxidation, flow chemistry, organic dye, supported
catalyst, Rose Bengal, continuous-flow reaction
Singlet oxygen is a widely used oxidant in various fields
of science in both industry and academia, such as fine
chemical synthesis, wastewater treatment, blood steriliza-
2–4
tion, cancer therapy, insecticides, and herbicides.
Espe-
cially in organic synthesis, its characteristic electrophilic
nature promotes [4+2]-type cycloaddition of dienes, ene re-
5–7
actions, heteroatom oxidation reactions, and so on. High-
ly reactive singlet oxygen can be generated by chemical or
photophysical methods. Among the approaches, photosen-
sitization of molecular oxygen is the most common method
from the standpoint of green and sustainable chemistry.
Molecular sensitizers (organic dyes or organometallic com-
plexes) or solid semiconductors can activate ground-state
triplet oxygen to a reactive molecule and promote oxida-
tion reactions of organic molecules. Further demands, such
as the repeated use of catalysts and the avoidance of the
need for separation, encouraged a recent study on the im-
mobilization of homogeneous sensitizers led to the devel-
8–10
opment of heterogeneous sensitizers.
In addition to
©
2020. Thieme. All rights reserved. Synlett 2020, 31, A–E
Georg Thieme Verlag KG, Rüdigerstraße 14, 70469 Stuttgart, Germany

B
Synlett
K. Masuda et al.
Letter
Cl
idazolium support was then selected, and, to increase the
level of conversion, the catalyst was filled into a PTFE tube
that had twice the width (3/2 mm OD/ID). The system was
illuminated with 33 bulbs of LED tape light, and a substrate
solution and air were flowed at 0.04 and 1.0 mL/min, re-
spectively (entry 2). Under these conditions, the reaction
showed a very high conversion, with high yield of the de-
sired product 1, together with 9% of the dehydrogenated
product 2. Reduced flow rates of the solution and gas led to
full conversion and 80% yield of the product, but the
amount of 2 also increased. It was suspected that intense
light caused undesired dehydrogenation reactions. We then
conducted an experiment with three bulbs of LED tape
light. At a flow rate of 0.04 mL/min of the substrate solution
and 1.0 mL/min of air, a moderate conversion was observed
with 61% yield of the product (entry 4). With a reduced
flow rate (0.02 mL/min of substrate solution and 0.5
mL/min air), almost full conversion was achieved, and the
product was obtained in 80% yield (entry 5). The amount of
undesired product 2 was also reduced under these condi-
tions. In addition, under these optimal conditions no leach-
ing or bleaching was observed by UV/Vis spectroscopy (see
Figure S6 in the Supporting Information).
Cl
Cl
Cl
O– Na+
O
I
I
O– Na+
O
O
I
I
Figure 1 Structure of Rose Bengal (sodium salt)
All commercial resins examined provided good catalytic
activities, and 73–82% yields of the desired product were
achieved (Table 1, entries 1–5). However, both the IRA900
(
entry 1) and IRA910 (entry 3) solutions had a pink color,
indicating partial leaching of Rose Bengal during the reac-
tions. Other supports (Amberlyst A26, DS-5 and FPA 60) did
not show any visible leaching after 3 h (entries 2, 4, 5). The
electrochemical stability of quaternary ammonium cations
22,23
depends on their substituents,
and benzyl substituted
derivatives are known to easily fragment through radical
processes. Singlet oxygen processes or radical processes
promoted by activated dyes may lead to decomposition re-
actions of either the dye itself or the supporting materials.
The imidazolium cation is also known to be a chemically
stable species, and is often employed in ion-exchange res-
Table 1 Evaluation of Immobilized Rose Bengal Catalysts in Batcha
24–26
ins, ionic liquids, cationic membranes, etc.
A report that
described high activities of a Rose Bengal-imidazolium sys-
White LED light
Immobilized catalyst (30 mg)
27
O
tem also encouraged us to prepare new heterogeneous
supports for anionic dyes. Thus, alkyl or vinyl substituted
imidazoles were immobilized on Merrifield resin to prepare
imidazolium resins as new catalysts (PS-alkylimidazolium).
As a control, a synthetic ammonium resin was also pre-
pared by using the same protocol (PS-triethylammonium).
When the new catalysts were subjected to the batch reac-
tions, high conversions with moderate to high yields of the
products were observed (entries 6–11). None showed any
leaching of Rose Bengal, indicating that the catalyst sup-
ports are less prone to decomposition, although lower
yields were observed in the reactions with supports bear-
ing a hydrophobic side chain (entries 9, 10).
Among the catalysts, PS-ethylimidazolium was selected
and applied for flow reactions with a 2/1 mm (OD/ID) PFA
tube. The reaction was conducted in a hand-made photo-
flow device: the catalyst was packed into the tube, which
was then capped with filter devices on both ends, and a
substrate solution and air were flowed under ambient con-
ditions. The device was illuminated with an LED tape light
attached in a beaker in a cooling bath to control the reac-
tion temperature (see the Supporting Information). The re-
action proceeded in the flow system to provide 7.8% yield of
the desired product; however, bleaching of Rose Bengal was
observed during the operation (Table 2, entry 1). A vinylim-
+
O2
O
DCE (3 mL)
RT, 3 h
(
air)
0.3 mmol
1
Entry
Catalyst support
Yield (%)b
Leachingc
1
2
3
4
5
6
7
8
9
IRA900 (ORGANO)
82
74
74
73
77
75
79
80
68
60
70
77
yes
Amberlyst A26 (Aldrich)
IRA910CT (ORGANO)
DS-5 (ORGANO)
no
yes
no
FPA 60 (ORGANO)
no
PS-triethylammonium
PS-methylimidazolium
PS-ethylimidazolium
PS-n-butylimidazolium
PS-tert-butylimidazolium
PS-vinylimidazolium
PS-allylimidazolium
(bleaching)d
(bleaching)d
no
no
1
0
1
no
1
no
12
(bleaching)d
a
Reaction conditions: -terpinene (0.3 mmol), immobilized Rose Bengal
catalyst (30 mg), dichloroethane (0.1 M), air balloon.
Determined by NMR spectroscopic analysis using 1,3,5-trimethoxyben-
b
zene as internal standard.
c
Visible leaching into a reaction solution.
The catalyst lost its color, and the reaction solution was not colored.
d
©
2020. Thieme. All rights reserved. Synlett 2020, 31, A–E

C
Synlett
K. Masuda et al.
Letter
Table 2 Flow Reactions with Synthetic Supportsa
tube filter
White LED
O
+
O2
Immobilized RB catalyst
ø2 x 100 mm, PTFE tube
+
O
(
dry air)
0.1 mol/L (DCE)
2
1
White LED
Entry
LED bulbs
Catalyst support
Flow rate (soln., mL/min) Flow rate (air, mL/min) Conv. (%)b
Yield (%)b
1
2
1c
2
33
33
33
3
PS-ethylimidazolium
PS-vinylimidazolium
PS-vinylimidazolium
PS-vinylimidazolium
PS-vinylimidazolium
0.02
0.04
0.02
0.04
0.02
0.5
1.0
0.5
1.0
0.5
11 (18 h)
96
7.8
77
4.7
9
3
full
80
10
6
4
73
61
5
3
>99
80
8
a
Reaction conditions: -terpinene (0.1 mol/L), pentadecane (0.018 mol/L) in DCE, kept at 0 °C in an ice bath.
Determined by GC analysis using pentadecane as internal standard.
ID 1 × 100 (mm) PFA tube was used as a column. Severe catalyst bleaching was observed and the catalyst became colorless within two days.
b
c
Among the catalysts, only the vinylimidazolium support
ference in the stability of the Rose Bengal molecule. To clar-
ify the reasons underlying these observations, newly pre-
pared vinylimidazolium resin, a fresh catalyst after immo-
bilization, and the used catalyst were compared by solid-
state 13C NMR spectroscopic analysis. The synthesized vi-
nylimidazolium resin showed a clear peak at around 110–
115 ppm, indicating the presence of a vinyl group (see Fig-
ure S7 in the Supporting Information). This peak remained
when Rose Bengal was immobilized but disappeared after
flow reactions. It is known that under aqueous conditions,
the triplet state of Rose Bengal molecule decomposes via an
attack by oxygen, or dehalogenation from reduced anion
enabled excellent reactivity and stability of Rose Bengal, es-
pecially under flow conditions. Other supports with similar
imidazolium cations that were examined are summarized
in Table 3. An analogous ethylimidazolium catalyst showed
good reactivity and was suitable for use under batch condi-
tions, but bleaching of Rose Bengal was observed under
flow conditions, and the catalyst became completely color-
less after 2 days. Another analogous support incorporating
allylimidazolium showed rapid bleaching even after 3 h of a
batch operation. It is very interesting that such a simple
structural change (vinyl, ethyl, or allyl) led to such a big dif-
Table 3 Comparison of Imidazolium Functionalities and Stability of Catalysts
Catalyst support/structure
PS-vinylimidazolium
Batch (3 h)
stable
Flow
stable over 11 daysa
N
N
PS
PS-ethylimidazolium
stable
complete bleaching within 2 days
N
N
PS
PS-allylimidazolium
bleachingb
–
N
N
PS
a
See Figure S4 in the Supporting Information.
See Figure S5 in the Supporting Information.
b
©
2020. Thieme. All rights reserved. Synlett 2020, 31, A–E

D
Synlett
K. Masuda et al.
Letter
radical.2
moieties
8–31
2,33
Some studies of Rose Bengal with histidine
and recent reports of Rose Bengal interaction
Table 4 Scope of the Reactionsa
Entry SM
Conv. (%)b Product
3
with collagen to modulate photophysical/photochemical
Yield (%)b
34,35
properties
prompted us to consider some scenarios to
explain the high durability of our catalyst; the double bond
adjacent to an imidazolium ring may function as a sacrifi-
cial functionality to prevent an oxygen attack, or imidazoli-
um–Rose Bengal electron pair modulates triplet lifetime to
avoid formation of an anion radical, leading to dehalogena-
tion. This support-dependent dye protection is potentially
applicable to other anionic dyes, and such long-lifetime
heterogeneous dyes could be applicable for a wide range of
photosensitization and photocatalysis, especially under
continuous-flow conditions. The optimal conditions were
employed, and the flow reaction was conducted for an ex-
tended period to confirm the stability of the system. In-
deed, the flow reaction was conducted continuously for
more than 11 days without loss of reactivity or any visible
O
O
1
>99
80
2c
Ph
3
P
>99
>99
Ph
3
P=O
97
82
OH
OH
d
OH
S
3
S
O
S
16
O
O
a
Reaction conditions: substrate (0.1 mol/L), pentadecane (0.018 mol/L) in
DCE, 0.02 mL/min; air, 0.5 mL/min; catalyst was packed in ID 2 × 100
(mm) tube and illuminated with three white LED bulbs.
Determined by GC analysis using pentadecane as internal standard.
b
c
Flow rates: 0.005 mL/min for solution, 0.25 mL/min for air.
36
d
deactivation of the catalyst (Figure 2).
3
CHCl was used as solvent.
Acknowledgment
This work was partially supported by New Energy and Industrial
Technology Development Organization (NEDO).
Supporting Information
Supporting information for this article is available online at
https://doi.org/10.1055/s-0039-1691582.
S
u
p
p
orit
n
g Inform ati
o
n
S
u
p
p
orit
n
g Inform ati
o
n
References and Notes
Figure 2 Extended operation with PS-vinylimidazolium catalyst (for
conditions, see Table 2, entry 5); 1 (blue), 2 (red).
(1) New address: K. Masuda, Interdisciplinary Research Center for
Catalytic Chemistry, National Institute of Advanced Industrial
Science and Technology (AIST). Central 5, 1-1-1 Higashi,
Tsukuba, Ibaraki, 305-8565, Japan.
The developed catalyst system was then employed for
the oxidation of other molecules under continuous-flow
conditions (Table 4). Triphenylphosphine was successfully
oxidized, and its oxide was obtained in nearly quantitative
yield (entry 2). Oxidation of a thioether functionality also
successfully led to efficient conversion of ethylthioethanol
into its sulfoxide (82%) and sulfone (16%) (entry 3).
In summary, we have developed a robust organic photo-
sensitizer immobilized on a vinylimidazolium functional-
ized support. The immobilized Rose Bengal enabled contin-
uous-flow photooxidation reactions to be conducted for
over 11 days without leaching or bleaching. A significant
dye stabilization effect was only observed with vinylimid-
azolium functionalized support, and solid-state NMR analy-
sis indicated that the vinyl group played a key role. Further
detailed study of the dye protection mechanism and appli-
cations of the catalyst support to the other reactions are
under investigation in our group.
(2) DeRosa, M. C.; Crutchley, R. J. Coord. Chem. Rev. 2002, 233–234,
351.
(
(
(
(
(
(
3) Clennan, E. L.; Pace, A. Tetrahedron 2005, 61, 6665.
4) Ogilby, P. R. Chem. Soc. Rev. 2010, 39, 3181.
5) Ohloff, G. Pure Appl. Chem. 1975, 43, 481.
6) Wasserman, H. H.; Ives, J. L. Tetrahedron 1980, 37, 1825.
7) Ghogare, A. A.; Greer, A. Chem. Rev. 2016, 116, 9994.
8) Blossey, E. C.; Neckers, D. C.; Thayer, A. L.; Schaap, A. P. J. Am.
Chem. Soc. 1973, 95, 5820.
(9) Schaap, A. P.; Thayer, A. L.; Blossey, E. C.; Neckers, D. C. J. Am.
Chem. Soc. 1975, 97, 3741.
10) Herrmann, J. M. Top. Catal. 2005, 34, 49.
(
(
11) Su, Y.; Straathof, N. J. W.; Hessel, V.; Noël, T. Chem. Eur. J. 2014,
20, 10562.
(
12) Romero, N. A.; Nicewicz, D. A. Chem. Rev. 2016, 116, 10075.
(13) Shaw, M. H.; Twilton, J.; MacMillan, D. W. C. J. Org. Chem. 2016,
81, 6898.
14) Plutschack, M. B.; Pieber, B.; Gilmore, K.; Seeberger, P. H. Chem.
Rev. 2017, 117,18 11796.
15) Carofiglio, T.; Donnola, P.; Maggini, M.; Rossetto, M.; Rossi, E.
Adv. Synth. Catal. 2008, 350, 2815.
(
(
©
2020. Thieme. All rights reserved. Synlett 2020, 31, A–E

E
Synlett
K. Masuda et al.
Letter
(
(
(
(
16) Lumley, E. K.; Dyer, C. E.; Pamme, N.; Boyle, R. W. Org. Lett.
mixture was stirred (overhead stirring) under reflux condition
until the consumption of the imidazole became constant (mon-
itored by GC). After filtration, the residue was washed with ace-
tonitrile, ethyl acetate, acetone, water, acetone, ethyl acetate
and dichloromethane. Drying at 40 °C in a vacuum oven
afforded the PS-vinylimidazolium support (3.0359 g, loading of
imidazole: 0.674 mmol/g, determined by the consumption of
imidazole). Rose Bengal (3.1246 g, 3.07 mmol, 1.5 equiv) was
dissolved in pure water (250 mL), and PS-vinylimidazolium
support (3.0359 g, 2.05 mmol imidazolium group, 1.0 equiv)
was added to the resulting solution. This mixture was stirred at
room temperature for 2 h. After filtration, the residue was
washed with water, methanol, dichloromethane and finally
washed with dichloromethane in a Soxhlet extractor under
reflux conditions until the solvent in the extractor became col-
orless. Filtration and desiccation afforded the RB@PS-vinylimid-
azolium catalyst (3.9775 g, loading of RB: 0.12 mmol/g (deter-
mined by EDX analysis)).
2012, 14, 5724.
17) Dambruoso, P.; Ballestri, M.; Ferroni, C.; Guerrini, A.; Sotgiu, G.;
Varchi, G.; Massi, A. Green Chem. 2015, 17, 1907.
18) Han, X.; Bourne, R. A.; Poliakoff, M.; George, M. W. Chem. Sci.
2011, 2, 1059.
19) Amara, Z.; Bellamy, J. F. B.; Horvath, R.; Miller, S. J.; Beeby, A.;
Burgard, A.; Rossen, K.; Poliakoff, M.; George, M. W. Nat. Chem.
2015, 7, 489.
(
20) Peng, F.; Zhi, P.; Ji, H.; Zhao, H.; Kong, F.-Y.; Liang, X.-Z.; Shen, Y.-M.
RSC Adv. 2017, 7, 19948.
(
21) Tobin, J. M.; McCabe, T. J. D.; Prentice, A. W.; Holzer, S.; Lloyd, G.
O.; Paterson, M. J.; Arrighi, V.; Cormack, P. A. G.; Vilela, F. ACS
Catal. 2017, 7, 4602.
(
(
(
(
(
22) Ross, S. D.; Finkelstein, M.; Petersen, R. C. J. Am. Chem. Soc. 1960,
82, 1582.
23) Mousavi, M. P. S.; Kashefolgheta, S.; Stein, A.; Bühlmann, P.
J. Electrochem. Soc. 2016, 163, H74.
24) Gu, F.; Dong, H.; Li, Y.; Si, Z.; Yan, F. Macromolecules 2014, 47,
Continuous-flow reactions: RB@PS-vinylimidazolium catalyst
(220 mg) were packed into a PTFE tube (ID: 2 mm; OD: 3 mm;
L: 10 cm), with in-line filter devices for both ends. This tube
was used as the column for the flow reaction. -Terpinene was
prepared as the reaction solution in 1,2-dichloroethane (0.1
mol/L, kept at 0 °C), with 0.018 mol/L pentadecane as an inter-
nal standard. This mixture (flow rate: 0.02 mL/min) was com-
bined with air (0.5 mL/min) and introduced to the column.
Three white LED bulbs were used as the light source. The reac-
tion mixture was collected for 1 h and analyzed by GC to deter-
mine the yield.
208.
25) Kazemiabnavi, S.; Zhang, Z.; Thornton, K.; Banerjee, S. J. Phys.
Chem. B 2016, 120, 5691.
26) Xue, Z.; Qin, L.; Jiang, J.; Mu, T.; Gao, G. Phys. Chem. Chem. Phys.
2018, 20, 8382.
(
(
(
(
(
(
(
(
27) Fall, A. Open Org. Chem. J. 2012, 6, 21.
28) Wilson, W. W.; Heitz, J. R. J. Agric. Food Chem. 1984, 32, 615.
29) Lambert, C. R.; Kochevar, I. E. Photochem. Photobiol. 1997, 66, 15.
30) Zakrzewski, A.; Neckers, D. C. Tetrahedron 1987, 43, 4507.
31) Oster, G.; Oster, G. K.; Karg, G. J. Phys. Chem. 1962, 66, 2514.
32) Westhead, E. W. Biochemistry 1965, 4, 2139.
33) Rippa, M.; Picco, C.; Pontremoli, S. J. Biol. Chem. 1970, 245, 4977.
34) Simpson, M. J.; Poblete, H.; Griffith, M.; Alarcon, E. I.; Scaiano, J.
C. Photochem. Photobiol. 2013, 89, 1433.
1
Ascaridole (1): Product was confirmed by H NMR analysis of
1
crude material and yields were determined by GC analysis. H NMR
(CDCl ):  = 6.50–6.40 (dd, J = 8.6, 33.5 Hz, 2 H, -CH=CH-), 2.04–
3
i
2.00 (m, 2 H, -CH - near - Pr), 1.96–1.89 (m, 1 H, -CH(CH ) ),
2
3 2
(35) Alarcon, E. I.; Poblete, H.; Roh, H. G.; Couture, J. F.; Comer, J.;
Kochevar, I. E. ACS Omega 2017, 2, 6646.
1.53–1.51 (d, J = 10.0 Hz, 2 H, -CH - near -Me), 1.38 (s, 3 H,
2
-CH ), 1.01–0.99 (d, J = 7.0 Hz, 6 H, -CH(CH ) ). GC: R = 38.5 min
3
3
2
t
(
36) Catalyst preparation: To a solution of 1-vinylimidazole (0.48
mL, 5.25 mmol, 1.05 equiv) and mesitylene (ca. 0.5 mL, internal
standard) in acetonitrile (50 mL), Merrifield resin (1.9 mmol/g
(50 °C: 30 min, then heating, rate: 10 °C/min to 250 °C: 10 min).
p-Cymene (2): Detected by GC analysis and identified by com-
parison to commercial product. GC: R = 24.9 min (50 °C: 30
t
-Cl, 2.6316 g, 5 mmol, 1.0 equiv) and NaI (37.5 mg, 0.25 mmol,
min, then heating, rate: 10 °C/min to 250 °C: 10 min).
0.05 equiv) were added at room temperature. The reaction
©
2020. Thieme. All rights reserved. Synlett 2020, 31, A–E