Flavin-cyclodextrin conjugates as catalysts of enantioselective sulfoxidations with hydrogen peroxide in aqueous media

Article abstract of DOI:10.1039/c0cc02562c

β-Cyclodextrin-flavin conjugates are highly efficient catalysts for the sulfoxidation of methyl phenyl sulfides with hydrogen peroxide in neat aqueous media operating at loadings down to 0.2 mol% and allowing for enantioselectivities up to 80% ee.

Full text of DOI:10.1039/c0cc02562c

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Flavin–cyclodextrin conjugates as catalysts of enantioselective
sulfoxidations with hydrogen peroxide in aqueous mediaw
a
b
Viktor Mojr,a Vladimı
´
r Herzig,ab Milos Budesı
´
nsky b
´
´
, Radek Cibulka* and Tomas Kraus*
Received 14th July 2010, Accepted 24th August 2010
DOI: 10.1039/c0cc02562c
b-Cyclodextrin–flavin conjugates are highly efficient catalysts
for the sulfoxidation of methyl phenyl sulfides with hydrogen
peroxide in neat aqueous media operating at loadings down to
0.2 mol% and allowing for enantioselectivities up to 80% ee.
p–p interaction between an oxidant (flavin hydroperoxide) and
an aromatic substrate.6b,7 It can be expected that the catalytic
properties of these systems could be further improved by an
approach inspired by enzyme-like catalysis which utilizes a
chiral substrate-binding site to control the stereoselectivity of
the transformation. In particular, cyclodextrins and their
derivatives have been extensively studied as enzyme mimics
capable of significant enhancement of the reaction rates of
various organic transformations in neat aqueous media.8–11 In
addition, cyclodextrin cavities provide a chiral environment,
thus allowing, in principle, enantioselective reactions.10
Aiming at the development of flavin-based organocatalysts
for enantioselective sulfoxidations,6b we have considered the
use of cyclodextrin macrocycle as chiral auxiliary covalently
attached to the flavinium moiety hoping that preorganization
of the substrate by complexation inside the cyclodextrin cavity
could enhance both the rate and enantioselectivity of the
sulfoxidation. Flavin–cyclodextrin conjugates have already
been investigated11 as catalysts (enzyme mimics) for redox
reactions and were found to be active catalysts of oxidation of
various substituted benzyl alcohols11c,d to corresponding
aldehydes and phenylmethanethiols11d to disulfides. None of
them, however, met all the requirements for a catalyst for
oxidation by hydrogen peroxide. To this end, we have
designed two new flavin–cyclodextrin conjugates 1 and 2
(Fig. 1) with N5-ethylated dihydroalloxazine and dihydro-
isoalloxazine moieties which are bonded through N1 and
N3 positions, respectively, to the primary C6 carbon of
b-cyclodextrin by amidomethylene linkage.
Enantiomerically pure sulfoxides are important chiral auxiliaries
employed in asymmetric syntheses as well as active substances
used in the pharmaceutical chemistry.1 Enantioselective
sulfoxidation of prochiral sulfides thus represents one of the
key transformations in organic synthesis. Various chemocatalytic
and biocatalytic systems for sulfoxidation have been developed,
the use of transition metal complexes with chiral ligands as
catalysts along with hydrogen peroxide as the terminal oxidant
being studied in particular.1–3 Metal-catalyzed sulfoxidations
with hydrogen peroxide are usually highly enantioselective
(ee 4 90%) and require low loadings of the catalyst (1–2%).
With most of these catalysts, however, high enantioselectivities
are achieved partly by overoxidation of one of the sulfoxide
enantiomers to sulfone (kinetic resolution) at the expense
of the yield.2,3 In contrast, sulfoxidations promoted by
organocatalysts are devoid of excessive oxidation.4 Among
organocatalytic oxidation methods, those utilizing 5-ethylflavin
hydroperoxides (5-EtFl-OOH) formed in situ from the corres-
ponding flavinium salts (5-EtFl+) and hydrogen peroxide
seem to be the most promising for sulfoxidations.5 Similarly,
reduced 5-ethylflavins (5-EtFlH) were used5 for the catalysis of
oxidations by H2O2, being activated by oxygen in the first
catalytic cycle (Scheme 1). When planarly chiral flavinium
catalysts6 were employed, substituted thioanisoles were oxidized
with enantiomeric excess up to 65%,6a provided that relatively
high loading of the catalyst (up to 12 mol%) was used.
In this communication, we describe the syntheses of
conjugates 1 and 2 and their abilities to catalyze enantio-
selective oxidation of aryl methyl sulfides by hydrogen
peroxide. Their syntheses were achieved by PyBOP-promoted
condensation reactions of the corresponding N1- and
N3-carboxymethylflavin derivatives 3 and 4 with 6-amino-b-
cyclodextrin (Scheme 2) which gave conjugates 5 and 6 in 72%
Stereodiscrimination in oxidation reactions catalyzed by
chiral flavin derivatives studied so far is induced by weak
Scheme 1 Catalytic cycle of sulfoxidation by hydrogen peroxide
mediated by flavinium salts 5-EtFl+
.
a Department of Organic Chemistry, Institute of Chemical Technology,
´
Prague, Technicka 5, Prague 6, Czech Republic.
E-mail: cibulkar@vscht.cz; Fax: +420 220444288
b Institute of Organic Chemistry and Biochemistry AS CR, v.v.i.,
´
Flemingovo nam. 2, 166 10 Prague 6, Czech Republic.
E-mail: kraus@uochb.cas.cz; Fax: +420 220183560
w Electronic supplementary information (ESI) available: Experimental
details about synthesis and characterization of conjugates, and about
stereoselective sulfoxidations. See DOI: 10.1039/c0cc02562c
Fig. 1 Structures of flavin–cyclodextrin conjugates used in the study.
Chem. Commun., 2010, 46, 7599–7601 7599
c
This journal is The Royal Society of Chemistry 2010

View Article Online
and 40% yields, respectively. In order to be catalytically
active, the flavin moiety requires an alkyl (most often ethyl)
substituent at the N5 position. Thus, N5-ethylation of 5 and 6
was accomplished by modified12 reductive alkylation employing
Pd-catalyzed hydrogenation and a large excess of acetaldehyde
in a mixture of ethanol and water acidified with perchloric
acid. Since N5-ethylated flavins are sensitive to air oxidation,
making direct monitoring of the course of the reaction by
common techniques difficult, the progress of reductive alkylation
was followed by testing the activity of the catalyst in the
oxidation reaction of thioanisole. In this way, the optimal
reaction time for the reductive alkylation was determined
to be 24 hours. Parallel control experiments using native
b-cyclodextrin in place of the conjugates revealed that the
b-cyclodextrin macrocycle is stable under these reaction
conditions. Despite our great efforts, we were unable to record
clear NMR spectra of the N-alkylated conjugates in their
reduced forms due to peak broadening, probably as a result
of the presence of radical species. However, the expected
structures of the N5-alkylated conjugates were supported by
high resolution mass spectra; peaks of reduced species could
be observed in the spectrum of 2 recorded using a freshly
prepared sample, whereas the presence of flavinium salts was
evidenced in solutions of both 1 and 2 after exposing the
samples to air oxygen.13
series of six methyl phenyl sulfides (Table 1) differing in
substitution in the para position. The reactions were conducted
in buffered (sodium phosphate) aqueous solutions with 1 mol%
loadings15 of the catalysts along with an excess (2.3 equivalents)
of hydrogen peroxide. While conjugate 1 provided optimal
conversions at pH 7.5, the isoalloxazine analogue 2 was
employed in the phosphate buffer at pH 2.9.16
The starting thiols proved to be poorly soluble in the
aqueous environment and tended to form emulsions or fine
suspensions. It was found that vigorous shaking of the reaction
vials with a wrist shaker provided the best reproducibility. In
most cases the reaction mixture became homogeneous after
several minutes of reaction time. The poor solubility of the
starting materials in the reaction mixture did not allow
determination of kinetics of the reactions and quantification
of rate enhancements. Nevertheless, a comparison of efficiencies
of the catalysts was made by determining (i) the conversions
after a 10 minute period and (ii) the reaction time required to
achieve at least 90% conversion.
In general, alloxazine conjugate 1 produced higher conversions
for the oxidation reactions compared with isoalloxazine
conjugate 2. Under catalysis of 1 most substrates were
oxidized with a conversion of over 90% within 10 minutes.
The isoalloxazine conjugate showed lower acceleration [except
for 4-(methylsulfanyl)phenol (entry 6) where it provided
nearly quantitative oxidation within 10 minutes]. Nevertheless,
with the exception of 4-(methylsulfanyl)benzoic acid (entries
11 and 12) all oxidation reactions catalyzed with conjugates 1
or 2 proceeded to at least 90% conversion within 90 minutes.
4-(Methylsulfanyl)benzoic acid proved to be a difficult substrate
for oxidation for both catalysts, presumably because of the
electron withdrawing nature of the carboxylic function and its
very poor solubility in aqueous solutions. It is important to
note that no over oxidation of sulfides to sulfones was
observed in any of the oxidations catalyzed by 1 and 2.
Analyses of the enantiomeric excesses of the isolated
sulfoxides by HPLC on chiral phases (see Table 1 and ESIw
for details) revealed that both catalysts 1 and 2 were able to
exert various degrees of enantioselectivity across the series of
thioanisole oxidations. Remarkably, conjugate 1 consistently
provided enantioselectivities in the range 64% up to 80% ee,
the highest value being observed in the case of the oxidation
of methyl 4-methylphenyl sulfide (entry 13). The observed
differences in enantioselectivities exhibited by 1 and 2 could
be caused by different contribution of non-stereoselective
intermolecular oxidation (reaction of flavin hydroperoxide
with a sulfide outside the cavity) or by different orientation
of the flavin moiety with respect to the bound substrate. Blank
experiments (entries 3 and 4) carried out under the same
conditions using a mixture of b-cyclodextrin and 5-ethyl-1,3-
dimethylalloxazinium salt 7 or 5-ethyl-3,10-dimethylalloxazinium
salt 8 (Fig. 2) in place of conjugates 1 and 2, respectively,
showed lower efficiency in terms of conversions and provided
no detectable enantioselectivities, therefore proving that covalent
linking of both flavin and b-cyclodextrin components is
essential.
Next, the abilities of the N5-alkylated alloxazine and
isoalloxazine conjugates 1 and 2 to catalyze hydrogen
peroxide-promoted oxidation of sulfides were tested using a
Scheme 2 Syntheses of flavin–cyclodextrin conjugates 1 and 2;
(a) compound 3, PyBOP, DIPEA, DMF; (b) CH3CHO, Pd/C, H2,
EtOH, H2O, HClO4; (c) compound 4, PyBOP, DIPEA, DMF;
(d) CH3CHO, Pd/C, H2, EtOH, H2O, HClO4.
The high conversion rates achieved with conjugate 1
prompted us to further decrease the catalyst loadings. Thus,
when 0.5 mol% of conjugate 1 was employed for the oxidation
c
7600 Chem. Commun., 2010, 46, 7599–7601
This journal is The Royal Society of Chemistry 2010

View Article Online
Table 1 Enantioselective H2O2-sulfoxidations of aryl methyl sulfides
catalyzed by conjugates 1 and 2a
down to 0.2 mol% with a turnover number up to 395.
Structural modifications of both the flavin and cyclo-
dextrin components, which may bring about further improve-
ment in efficiencies (namely stability of the catalyst) and
enantioselectivities of the sulfoxidations as well as broad-
ening the substrate scope, are under investigation in our
laboratories.
Optimized conversion
Conv.
Catalystc 10 min (%)
Timed/min Conv. (%)
e
ee (%)
This work was financially supported by Czech Science
Foundation (Grant No. 203/07/1246) and IOCB (Z40550506).
Entry R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
H
H
H
H
1
2
83
20
20
90
120
120
—
—
—
36
—
30
60
120
—
120
60
93
99
80
82
—
—
—
90
—
96
34
23
—
95
99
79
64
21
0
0
73
9
72
28
76
21
73
35
80
41
77
77
7 + b-CDf 17
Notes and references
8 + b-CDf 17
OH
OH
OCH3
OCH3
ORb
ORb
COOH 1
COOH 2
CH3
CH3
CH3
CH3
1
2
1
2
1
2
91
95
97
78
98
46
30
4
1 (a) I. Ferna
´
(b) M. C. Carreno, G. H. Herna
´
ndez and N. Khiar, Chem. Rev., 2003, 103, 3651;
ndez-Torres, M. Ribagorda and
A. Urbano, Chem. Commun., 2009, 6129; (c) J. Legras, J. R. Dehli
and C. Bolm, Adv. Synth. Catal., 2005, 347, 19.
2 (a) K. P. Bryliakov and E. P. Talsi, Curr. Org. Chem., 2008, 12, 386;
(b) H. B. Kagan, in Organosulfur Chemistry in Asymmetric Synthesis,
ed. T. Toru and C. Bolm, Wiley-VCH, Weinheim, 2008, pp. 1–29.
3 Recent examples of metal catalyzed enantioselective sulfoxidations:
(a) Y. Wu, J. Liu, X. Li and A. S. C. Chan, Eur. J. Org. Chem., 2009,
2607; (b) K. Matsumoto, T. Yamaguchi and T. Katsuki, Chem.
Commun., 2008, 1704; (c) T. Yamaguchi, K. Matsumoto, B. Saito
and T. Katsuki, Angew. Chem., Int. Ed., 2007, 46, 4729;
(d) E. Kiromichi and T. Katsuki, Synlett, 2008, 1543.
1
2
97
18
1 (0.5%) 67
1 (0.2%) 35
60
a
Conditions: substrate (0.1 mmol), H2O2 (2.3 equiv.), phosphate
b
4 Organocatalytic systems for sulfoxidations are reviewed in: (a)
J.-E. Backvall, in Modern Oxidation Methods, ed. J.-E. Backvall,
buffer pH 7.5 (for 1) or 2.9 (for 2), RT. RQO(CH2)2O(CH2)2OCH3.
Catalyst loading 1 mol% (related to the substrate) if not stated
¨
¨
c
Wiley-VCH Verlag, Weinheim, 2004, pp. 193–222; (b) A. Armstrong,
in Enantioselective Organocatalysis, ed. P. I. Dalko, Wiley-VCH
Verlag, Weinheim, 2007, pp. 403–424.
5 (a) F. G. Gelalcha, Chem. Rev., 2007, 107, 3338; (b) I. Yasushi and
N. Takeshi, Chem. Rec., 2007, 7, 354.
6 (a) S. Shinkai, T. Yamaguchi, O. Manabe and F. Toda, Chem.
d
otherwise. Time necessary to achieve at least 90% conversion,
e
maximally 120 min. Enantiomer ratios were determined by HPLC
f
on a chiral stationary phase (see ESIw). For structures of compounds
7 and 8, see Fig. 2.
Commun., 1988, 1399; (b) R. Jurok, R. Cibulka, H. Dvora
F. Hampl and J. Hodacova, Eur. J. Org. Chem, 2010, DOI:
10.1002/ejoc.201000592.
kova,
´ ´
´
7 S.-I. Murahashi, S. Ono and Y. Imada, Angew. Chem., Int. Ed.,
2002, 41, 2366.
8 R. Breslow and S. D. Dong, Chem. Rev., 1998, 98, 1997.
9 (a) L. G. Marinescu, M. Molbach, C. Rousseau and M. Bols,
J. Am. Chem. Soc., 2005, 127, 17578; (b) L. G. Marinescu and
M. Bols, Angew. Chem., Int. Ed., 2006, 45, 4590.
10 K. Takahashi, Chem. Rev., 1998, 98, 2013.
Fig. 2 Structures of flavin catalysts 7 and 8.
11 (a) I. Tabushi and M. Kodera, J. Am. Chem. Soc., 1987, 109, 4734;
(b) D. Rong, H. Ye, T. R. Boehlow and V. T. D’Souza, J. Org.
Chem., 1992, 57, 163; (c) H. P. Ye, W. Tong and V. T. D’Souza,
J. Am. Chem. Soc., 1992, 114, 5470; (d) H. Ye, W. Tong and
V. T. D’Souza, J. Chem. Soc., Perkin Trans. 2, 1994, 2431;
(e) V. T. D’Souza, Supramol. Chem., 2003, 15, 221.
of methyl 4-methylphenyl sulfide (entry 15), quantitative
conversion was accomplished within 60 minutes of reaction
time. With 0.2 mol% loading of 1 79% conversion was
achieved within the same period (entry 16). Further increase
of reaction time produced no effect, indicating that the catalyst
was inactive after 1 hour of reaction time.
12 C. Smit, M. W. Fraaije and A. J. Minnaard, J. Org. Chem., 2008,
73, 9482.
13 The presence of reduced form of 1 could not be evidenced by MS
despite great efforts devoted to avoiding oxidation of the sample by
air oxygen. Higher sensitivity of dihydroalloxazine moiety towards
oxidation compared to dihydroisoalloxazine can be explained by
lower oxidation potential of the former; see ref. 14.
14 Y. Imada, H. Iida, S. Ono, Y. Masui and S.-I. Murahashi,
Chem.–Asian J., 2006, 1, 136.
15 The concentration of the catalysts 1 and 2 corresponds to
the concentration of the conjugates 5 and 6 prior to reductive
amination assuming quantitative conversion.
In conclusion, we have synthesized two b-cyclodextrin–
flavin conjugates, which proved to be highly efficient catalysts
for the oxidation of electron-rich methyl phenyl sulfides to
sulfoxides by hydrogen peroxide. In particular, the catalytic
system based on b-cyclodextrin–alloxazine conjugate 1 is
distinguished from other organocatalysts by fast near-
quantitative conversions and high enantioselectivities, reaching
up to 80% ee. It is also remarkable that the reactions proceed
in neat aqueous media with very low loadings of the catalyst
16 R. Cibulka, L. Baxova, H. Dvorakova, F. Hampl, P. Menova,
´ ´ ´ ´ ´
V. Mojr, B. Plancq and S. Sayin, Collect. Czech. Chem. Commun.,
2009, 74, 973.
c
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 7599–7601 7601