Preparation of flavin-containing mesoporous network polymers and their catalysis

Article abstract of DOI:10.1016/j.tetlet.2020.151710

Riboflavin tetramethacrylate (RFlTMA) was prepared as a flavin monomer and copolymerized with ethylene glycol dimethacrylate (EGDMA) under polymerization-induced phase separation conditions. The resulting flavin-containing mesoporous network polymer, poly(RFlTMA-co-EGDMA), was found to be a more effective catalyst than riboflavin tetraacetate (RFlTA), a soluble analogue, for aerobic hydrogenation of olefins despite its heterogeneity, which allowed for its multiple recovery and reuse through simple filtrations and washings without loss in catalytic activity. In addition, the polymeric flavin was demonstrated to be utilized also as an effective photocatalyst in the oxidation of benzyl alcohols.

Full text of DOI:10.1016/j.tetlet.2020.151710

Tetrahedron Letters 61 (2020) 151710
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Preparation of flavin-containing mesoporous network polymers
and their catalysis
Yukihiro Arakawa , Fumiaki Sato , Kenta Ariki , Keiji Minagawa a,b, Yasushi Imada a,
a
a
a
⇑
a
Department of Applied Chemistry, Tokushima University, Minamijosanjima, Tokushima 770-8506, Japan
Institute of Liberal Arts and Sciences, Tokushima University, Minamijosanjima, Tokushima 770-8502, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Riboflavin tetramethacrylate (RFlTMA) was prepared as a flavin monomer and copolymerized with ethy-
lene glycol dimethacrylate (EGDMA) under polymerization-induced phase separation conditions. The
resulting flavin-containing mesoporous network polymer, poly(RFlTMA-co-EGDMA), was found to be a
more effective catalyst than riboflavin tetraacetate (RFlTA), a soluble analogue, for aerobic hydrogenation
of olefins despite its heterogeneity, which allowed for its multiple recovery and reuse through simple fil-
trations and washings without loss in catalytic activity. In addition, the polymeric flavin was demon-
strated to be utilized also as an effective photocatalyst in the oxidation of benzyl alcohols.
Ó 2020 Elsevier Ltd. All rights reserved.
Received 26 December 2019
Revised 1 February 2020
Accepted 4 February 2020
Available online 4 February 2020
Keywords:
Flavin
Polymer-supported catalyst
Hydrogenation
Photooxidation
A variety of natural flavoenzymes catalyses, such as monooxyge-
nase catalysis and DNA photolyase catalysis, have inspired chemists
to develop metal-free thermal redox and photoredox reactions cat-
alyzed by simple flavin molecules [1]. For instance, we have so far
developed some flavin-catalyzed aerobic oxidation reactions
including sulfoxidation [2a,2e], amine oxidation [2e], Baeyer-
Villiger reaction [2a,2d,2f], and hydrogenation of olefins with
hydrazine [2b,2c] under environmentally friendly conditions.
Recently, we extended our attention to the immobilization of
2
Riboflavin (Vitamin B ) is readily available in commerce and its
tetraacylated derivatives, such as riboflavin tetraacetate (RFlTA),
can be readily synthesized in a single step [2b,5].
Thereupon we designed RFlTMA as a new flavin-containing
monomer, which was successfully prepared by the reaction of ribo-
flavin with methacrylic anhydride in pyridine under reflux condi-
tions (Fig. 1). According to the PIPS technique reported
previously [4], RFlTMA was then copolymerized with EGDMA
0
(RFlTMA:EGDMA = 0.06:0.94) using 2,2 -azobis(isobutyronitrile)
flavin catalysts onto
a
synthetic polymer to improve their
(AIBN) as an initiator in DMF/diglyme (3:1) under heating and inert
conditions (Fig. 2). The copolymerization proceeded smoothly and
the resulting gel was crushed and washed well with THF, metha-
nol, and diethyl ether and dried in vacuo to give a bright yellow
copolymer, poly(RFlTMA-co-EGDMA) (1a), in 80% yield. The flavin
loading of poly(RFlTMA-co-EGDMA) was estimated to be
practicability and reported highly reusable poly(styrene-co-
divinylbenzene)-supported flavin catalysts for the aerobic
reduction of olefins, which was the first example of immobilizing
a flavin catalyst onto insoluble supports via covalent bond [3].
In this article, we present a new type of flavin catalyst sup-
ported on a mesoporous network polymer. Mesoporous network
polymers are attractive supports to immobilize low-molecular-
weight catalysts because of their robustness, insolubility, and high
surface area, which can be prepared using well-established
polymerization-induced phase separation (PIPS) technique [4].
We have prepared a flavin-containing mesoporous network poly-
mer by copolymerization of ethylene glycol dimethacylate
ꢀ1
0.25 mmol g from its nitrogen content determined by elemental
analysis. The actual CHN content of poly(RFlTMA-co-EGDMA) was
in good agreement with the theoretical value, indicating that the
present copolymerization took place according to the feed molar
ratio. A strong fluorescence emission peak at around 540 nm was
observed in solid state of 1a (/ = 0.124), indicating that flavin units
could be dispersed to the polymer network rather homogeneously,
while RFlTMA showed little emission in its solid state (/ = 0.007)
due to concentration quenching (see Supporting information). For
comparison, the same copolymerization was carried out in DMF
in the absence of diglyme that is known to assist PIPS, which
(
EGDMA) and riboflavin tetramethacrylate (RFlTMA) under PIPS
conditions and explored its thermal redox and photoredox
catalytic activities.
ꢀ1
afforded another poly(RFlTMA-co-EGDMA) (1b, 0.27 mmol g
)
⇑
Corresponding author.
E-mail address: imada@tokushima-u.ac.jp (Y. Imada).
in 90% yield through the same work up (Fig. 2). Furthermore, we
https://doi.org/10.1016/j.tetlet.2020.151710
040-4039/Ó 2020 Elsevier Ltd. All rights reserved.
0

2
Y. Arakawa et al. / Tetrahedron Letters 61 (2020) 151710
Fig. 1. Preparation of RFlTMA.
Fig. 3. Characterization of 1a, 1b, and 2: (a) nitrogen absorption isotherms at 77 K
and the resulting specific surface area determined by the BET method (SSABET), (b)
pore size distributions calculated by the DH method, (c) SEM images.
size distributions for 1a and 1b were plotted by the DH method
[
6], which suggested that both have a high mesopore ratio (pore
diameters between 2 and 50 nm) and also that the size of the
mesopores could be better controlled in 1a than 1b (Fig. 3b). Not
only the larger SSABET but also the better control of mesopore size
in 1a may be attributed to the positive effect of diglyme on PIPS in
their synthesis (Fig. 2). The surfaces of 1a, 1b, and 2 were then
visualized in scanning electron microscope (SEM), from which
numerous small cavities were clearly observed only in 1a and 1b
Fig. 2. Synthesis of poly(RFlTMA-co-EGDMA) and poly(RFlTMA-co-GLDMA).
(Fig. 3c). In addition, the degrees of swelling of 1a, 1b, and 2 in ace-
tonitrile, the solvent used for catalytic reactions mentioned later,
were found to be 290%, 250%, and 110%, respectively (see Support-
ing information), which suggested the specific surface area of 1a
would become highest also in acetonitrile.
were also interested in using glycerol dimethacrylate
GLDMA) instead of EGDMA because of the additional hydroxyl
functionality, so that the corresponding copolymerization was
performed according to the procedure that was used for preparing
(
We then evaluated the catalytic activities of 1a, 1b, and 2 in the
aerobic hydrogenation of olefins with hydrazine that was
previously developed by us as one of the unique flavin-catalyzed
thermal reactions [2b,2c]. 4-Phenyl-1-butene (3a) was chosen
as a test substrate and reactions were carried out with 3
equivalents of hydrazine monohydrate to the substrate in
acetonitrile under air at room temperature in the presence of
1
a, providing an orange copolymer, poly(RFlTMA-co-GLDMA)
ꢀ1
(2, 0.20 mmol g ), in 95% yield (Fig. 2).
The flavin-containing copolymers 1a, 1b, and 2 were further
characterized to evaluate their meso-porosity. Specific surface area
as well as pore size distribution were estimated by nitrogen
absorption and desorption isotherm measurement and analysis,
which revealed that 1a and 1b could have the BET specific surface
area (SSABET) of 570 m g and 490 m g , respectively, while the
surface area of 2 was not measurable due to significantly low N
absorption volumes in the isotherm measurement (Fig. 3a). Pore
2
ꢀ1
2
ꢀ1
2
mol% of either 1a, 1b, or 2 for 24 h. For comparison, the
reaction in the presence of RFlTA as a soluble catalyst and that in
the absence of any catalysts were also conducted under
2

Y. Arakawa et al. / Tetrahedron Letters 61 (2020) 151710
3
otherwise identical conditions that were previously developed by
Table 1
a
Aerobic reduction catalyzed by 1a.
us for the reaction catalyzed by soluble flavin molecules [2b]. As
expected, the best catalytic activity and the highest yield (94%) of
butylbenzene (4a) were observed in the heterogeneous reaction
system with 1a, which was even better than that observed in the
homogeneous system with RFlTA (Fig. 4a). The catalytic activity
over the homogeneous system may be attributed to the
controlled mesoporosity as well as well-dispersed flavin
molecules on the support, successfully suppressing self-
quenching of redox-active flavin species. Reusability of the
catalyst 1a was tested through the simplified catalyst recycling
Entry
1
Substrate
Product
Yield (%)b
95
2
3
96
74
process using
a filter-equipped glass reactor as previously
reported [3,7], which demonstrated that 1a could be reused with-
out loss in the activity at least 5 times (Fig. 4b).
We furthermore verified the utility of 1a in the aerobic hydro-
genation of other alkene/alkyne substrates under the above condi-
tions (Table 1). The reduction of 1-decene (3b) to 1-decane (4b)
smoothly proceeded in 95% yield (entry 1). The alkene moiety
within 1,2-epoxy-9-decene (3c) was selectively hydrogenated to
give 1,2-epoxydecane (4c) in 96% yield without ring opening of
the epoxide moiety (entry 2). Although the catalytic activity was
found to somewhat diminish in the presence of coordinating func-
tional groups, the desired reactions still occurred with moderate to
high yields. For example, allyl phenyl sulfide (3d) was selectively
reduced to phenyl n-propyl sulfide (4d) in 74% yield without oxida-
tion of the sulfide group (entry 3), which is valuable because such a
4
53
76
5c
a
Reactions were performed using 250
hydrazine monohydrate in 2 mL of acetonitrile in the presence of 2 mol% of 1a
lmol of the olefin and 750 lmol of
under air at room temperature.
b
Determined by 1H NMR measurement with internal standard.
c
1.5 mmol of hydrazine monohydrate was used.
sulfur-containing substrate hardly undergoes hydrogenation under
conventional transition metal-based conditions. In addition,
2
-allylphenol (3e) was converted into 2-propylphenol (4e) with
a yield of 53% under identical conditions (entry 4), whereas
-octyn-3-ol (3f) underwent hydrogenation twice to 3-octanol
4f) in 76% yield when 6 equivalents of hydrazine monohydrate
1
(
were used (entry 5).
Finally, we preliminarily explored whether the flavin-contain-
ing mesoporous polymer 1a could also work as an effective photo-
catalyst (Table 2). On the basis of conditions previously reported
[
5m,8], a solution of 4-methoxybenzyl alcohol in a mixture of
acetonitrile and water (1:9) was shaken and irradiated with blue
LED light (465 nm) under air in the presence of 2 mol% of 1a,
which provided 4-methoxybenzaldehyde in 92% yield after 24 h
(
entry 1). Under identical conditions, the use of RFlTMA instead
of 1a resulted in a lower yield of the product (51%, entry 2). No
reaction was observed without any catalysts or without light
irradiation (entries 3 and 4). These results indicate the potential
Table 2
a
Photooxidation of 4-methoxybenzyl alcohol.
Entry
Catalyst
Yield (%)b
1
2
3
1a
RFlTMA
—
1a
92
51
0
c
4
0
a
Reactions were performed with 0.4 mmol of the substrate in a mixture of
acetonitrile and water (1:9) in the presence of 2 mol% of a catalyst under air and
blue LED light irradiation.
Fig. 4. Flavin-catalyzed aerobic reduction of 3a to 4a: (a) comparison of yields
between catalytic reactions and a blank reaction, (b) catalyst recovery and reuse
protocol (left) and reusability of 1a (right).
b
Determined by GC analysis.
c
Without light irradiation.

4
Y. Arakawa et al. / Tetrahedron Letters 61 (2020) 151710
effectiveness of immobilized catalysts with such a high meso-
porosity in the photo-induced flavin catalyses [5].
c) G. de Gonzalo, M.W. Fraaije, ChemCatChem 5 (2013) 403–415;
d) Y. Imada, T. Naota, Chem. Rec. 7 (2007) 354–361.
2] a) Y. Arakawa, K. Yamanomoto, H. Kita, K. Minagawa, M. Tanaka, N. Haraguchi,
S. Itsuno, Y. Imada, Chem. Sci. 8 (2017) 5468–5475;
b) Y. Imada, T. Kitagawa, T. Ohno, H. Iida, T. Naota, Org. Lett. 12 (2010) 32–35;
c) Y. Imada, H. Iida, T. Naota, J. Am. Chem. Soc. 127 (2005) 14544–14545;
d) Y. Imada, H. Iida, S.-I. Murahashi, T. Naota, Angew. Chem. Int. Ed. 44 (2005)
[
In summary we introduced a flavin-containing mesoporous net-
work polymer, poly(RFlTMA-co-EGDMA), that was prepared by
copolymerization of RFlTMA, a novel flavin-containing multi-vinyl
monomer, with EGDMA under PIPS conditions. Characterizations
of poly(RFlTMA-co-EGDMA) by elemental analysis, absorption
and desorption isotherm analysis, and SEM analysis showed its
high meso-porosity. Poly(RFlTMA-co-EGDMA) exhibited high cat-
alytic activities comparable with low-molecular-weight flavin cat-
alysts in aerobic hydrogenation of olefins as well as photooxidation
of benzyl alcohols, which was readily recovered by filtration and
reused without a loss in activity.
1
704–1706;
e) Y. Imada, H. Iida, S. Ono, S.-I. Murahashi, J. Am. Chem. Soc. 125 (2003) 2868–
869;
f) S.-I. Murahashi, S. Ono, Y. Imada, Angew. Chem. Int. Ed. 41 (2002) 2366–
368.
2
2
[3] Y. Arakawa, R. Kawachi, Y. Tezuka, K. Minagawa, Y. Imada, J. Polym. Sci. Part A:
Polym. Chem. 55 (2017) 1706–1713.
4] S. Ogasawara, S. Kato, J. Am. Chem. Soc. 132 (2010) 4608–4613.
[
[
5] For selected studies that utilize riboflavin or RFlTA as a catalyst, see: a) N.P.
Ramirez, B. König, J.C. Gonzalez-Gomez Org. Lett. 21 (2019) 1368–1373;
b) G. Tang, Z. Gong, W. Han, X. Sun, Tetrahedron Lett. 59 (2018) 658–662;
c) T. Morack, J.B. Metternich, R. Gilmour, Org. Lett. 20 (2018) 1316–1319;
d) M. März, M. Kohout, T. Nevesel y´ , J. Chudoba, D. Prukała, S. Nizi n´ ski, M.
Sikorski, G. Burdzi n´ ski, R. Cibulka, Org. Biomol. Chem. 16 (2018) 6809–6817;
e) J.B. Metternich, D.G. Artiukhin, M.C. Holland, M. von Bremen-Kühne, J.
Neugebauer, R. Gilmour, J. Org. Chem. 82 (2017) 9955–9977;
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
f) R. Martinez-Haya, M.A. Miranda, M.L. Marin, Eur. J. Org. Chem. 2017 (2017)
2164–2169;
g) B. Mühldorf, R. Wolf, ChemCatChem 9 (2017) 920–923;
h) J.B. Metternich, R. Gilmour, J. Am. Chem. Soc. 138 (2016) 1040–1045;
i) T. Nevesel y´ , E. Svobodová, J. Chudoba, M. Sikorski, R. Cibulka, Adv. Synth.
Catal. 358 (2016) 1654–1663;
Acknowledgments
This work was supported by Grant-in-Aid for Scientific
Research on Innovative Areas ‘Advanced Molecular Transforma-
tions by Organocatalysts’ from MEXT.
j) T. Hering, B. Mühldorf, R. Wolf, B. König, Angew. Chem. Int. Ed. 55 (2016)
5342–5345;
k) B. Mühldorf, R. Wolf, Angew. Chem. Int. Ed. 55 (2016) 427–430;
l) J.B. Metternich, R. Gilmour, J. Am. Chem. Soc. 137 (2015) 11254–11257;
m) C. Feldmeier, H. Bartling, K. Magerl, R.M. Gschwind, Angew. Chem. Int. Ed.
5
4 (2015) 1347–1351;
n) J. Dad’ová, E. Svobodová, M. Sikorski, B. König, R. Cibulka, ChemCatChem 4
2012) 620–623.
[6] D. Dollimore, G.R. Heal, J. Appl. Chem. 109 (1964) 109–114.
Appendix A. Supplementary data
(
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tetlet.2020.151710.
[
7] Y. Arakawa, M. Wiesner, H. Wennemers, Adv. Synth. Catal. 353 (2011) 1201–
206.
1
[
8] a) J. Svoboda, H. Schmaderer, B. König, Chem. Eur. J. 14 (2008) 1854–1865;
b) H. Schmaderer, P. Hilgers, R. Lechner, B. König, Adv. Synth. Catal. 351 (2009)
References
163–174.
[
1] a) H. Iida, Y. Imada, S.-I. Murahashi, Org. Biomol. Chem. 13 (2015) 7599–7613;
b) R. Cibulka, Eur. J. Org. Chem. 2015 (2015) 915–932;