Mn(salen)-catalyzed sulfimidation: What are the real active species in sulfimidation?

Article abstract of DOI:10.1016/S0040-4039(01)00593-7

An unusual temperature effect on enantioselectivity was found in a study of Mn(salen)-catalyzed sulfimidation. The phenomenon was considered to be attributable to the following reasons: the reaction of Mn(salen) and a nitrenoid precursor (ArI=NTs) provides primarily an Mn-ArINTs adduct which undergoes sulfimidation and transformation to an Mn-nitrenoid species competitively; however, the rates of sulfimidation by the Mn-ArINTs adduct and the Mn-nitrenoid species and of the transformation are dependent on the structure of the catalyst, the nature of ArI=NTs and the presence or absence of a donor ligand.

Full text of DOI:10.1016/S0040-4039(01)00593-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 3885–3888
Mn(salen)-catalyzed sulfimidation: what are the real active
species in sulfimidation?
Chisa Ohta and Tsutomu Katsuki*
Department of Chemistry, Faculty of Science, Graduate School, Kyushu University 33, Hakozaki, Higashi-ku,
Fukuoka 812-8581, Japan
Received 26 February 2001; revised 2 April 2001; accepted 6 April 2001
Abstract—An unusual temperature effect on enantioselectivity was found in a study of Mn(salen)-catalyzed sulfimidation. The
phenomenon was considered to be attributable to the following reasons: the reaction of Mn(salen) and a nitrenoid precursor
(ArIꢀNTs) provides primarily an Mn–ArINTs adduct which undergoes sulfimidation and transformation to an Mn–nitrenoid
species competitively; however, the rates of sulfimidation by the Mn–ArINTs adduct and the Mn–nitrenoid species and of the
transformation are dependent on the structure of the catalyst, the nature of ArIꢀNTs and the presence or absence of a donor
ligand. © 2001 Elsevier Science Ltd. All rights reserved.
We recently disclosed that chiral (salen)manganese(III)
complexes [hereafter referred to as Mn(salen)] served as
catalysts for asymmetric sulfimidation.1 However, the
optimized reaction conditions for the sulfimidation
were found to vary with the substrates used. For exam-
ple, asymmetric sulfimidation of methyl phenyl sulfide
was well effected in chlorobenzene by using (R,R)-
Mn(salen) 1 as the catalyst in the presence of N-methyl-
morpholine N-oxide (NMO), while that of methyl
o-nitrophenyl sulfide in benzonitrile was well effected
by using (R,S)-Mn(salen) 2 in the absence of a donor
ligand (Scheme 1). On the other hand, sulfimidation of
TsN-
S+
1 (5 mol%), NMO (50 mol%)
S
Ph
Me
S
PhI=NTs, PhCl, rt
Ph
Me
84% ee, 55%
TsN-
S+
2 (5 mol%), MS 3A
o-(O2N)C6H4
Me
o-(O2N)C6H4
Me
PhI=NTs, C6H5CN, rt
90% ee, 77%
S
R
N
N
N
O
N
Mn+
Mn+
O
Ph Ph
O
O
3
3'
3'
3
Ph Ph
R
R
-
-
PF6
PF6
2
1
Scheme 1.
Keywords: (salen)manganese(III) complex; asymmetric catalysis; asymmetric sulfimidation; nonlinear effect.
* Corresponding author. Fax: +81 92 642 2607; e-mail: katsuscc@mbox.nc.kyushu-u.ac.jp
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)00593-7

3886
C. Ohta, T. Katsuki / Tetrahedron Letters 42 (2001) 3885–3888
Y
Y
Y= H or NO2
Me
S+ -NTs
Me
S
NTs
Mn
+
+
Mn
Me
S
Me
S+ -NTs
Me
S
+
N
N
N
NTs
Mn
O
O
NTs
O
O
O
O
+
Mn
Mn
Scheme 2.
methyl p-nitrophenyl sulfide with (R,S)-2 as the cata-
lyst showed moderate enantioselectivity (55% ee). To
understand the mechanism of these sulfimidation reac-
tions, we studied the enantioselectivity–temperature
relationship in the reactions using (R,R)-1 and (R,S)-2
as catalysts.
in the absence of a donor ligand: almost the same
enantiomer ratio was obtained at every temperature in
the reactions with (R,S)-2 as the catalyst and 3 and 4 as
the nitrenoid precursors and the ee-break was observed
at 40°C, while the enantiomer ratios observed in the
reactions with (R,R)-1 depended on the nitrenoid pre-
cursor used, that is, the ee-break was observed at ca.
27°C in the reaction with 3 and no ee-break in the
reaction with 4 (Fig. 2). Reactions of methyl o-nitro-
phenyl sulfide with (R,S)-2 as the catalyst and 3 and 4
We had reasoned that methyl o-nitrophenyl sulfide is a
better substrate for sulfimidation with (R,S)-2 as cata-
lyst, as the p-nitro group serves simply as an electron-
withdrawing group but the o-nitro group serves as a
coordinating group as well as an electron-withdrawing
group.1 Accordingly, we had also expected that methyl
phenyl sulfide and methyl p-nitrophenyl sulfide should
attack the Mn–nitrenoid species nucleophilically
(Scheme 2), and the enantiomer ratio of the resulting
sulfimide (p=major enantiomer/minor enantiomer)
would correlate linearly with the reaction temperature
through the Eyring equation (ln p=−DDH‡/RT+DDS‡/
R). On the other hand, methyl o-nitrophenyl sulfide
had been expected to make a complex with an Mn–
nitrenoid species which undergoes subsequent sulfimi-
dation. If this consideration is correct, there is some
chance for the latter reaction to show a non-linear
relationship between ln p and 1/T, because the forma-
tion of the Mn(nitrenoid)–sulfide complex is considered
to be reversible.2,3 Thus, we studied the relationship
between one enantiomer ratio and the reaction temper-
ature in these reactions using (R,R)-1 and (R,S)-2 as
the catalyst and PhIꢀNTs 3 and p-CH3OC6H4IꢀNTs 4
as the nitrenoid precursors.
Figure 1. Reactions of methyl pheny sulfide in chlorobenzene
in the presence of NMO.
Different from the above expectation, however, the
relationship between the enantiomer ratio and the reac-
tion temperature was found to depend not only on the
substrate, but also on the catalyst, the nitrenoid precur-
sor (ArIꢀNTs) and the presence or absence of a donor
ligand.
All sulfimidations of methyl phenyl sulfide in
chlorobenzene in the presence of NMO, irrespective of
the catalyst and the nitrenoid precursor, showed a
linear relationship between the enantiomer ratio and
the reaction temperature, though the reaction with
(R,R)-1 as the catalyst exhibited higher enantioselectiv-
ity than that with (R,S)-2 as the catalyst (Fig. 1). A
different type of relationship was observed in the sulfi-
midation of methyl p-nitrophenyl sulfide in benzonitrile
Figure 2. Reactions of methyl p-nitrophenyl sulifide in benzo-
nitrile.

C. Ohta, T. Katsuki / Tetrahedron Letters 42 (2001) 3885–3888
3887
ArI
Ts
as the nitrenoid precursors in benzonitrile also showed
N
almost the same enantiomer ratios at every temperature
and showed the ee-break at ca. 40°C, while the enan-
tiomer ratio–temperature relationship in the reactions
with (R,R)-1 depended on the nitrenoid precursor: the
reaction using 3 showed the ee-break at ca. 10°C, but
the reaction using 4 did not (Fig. 3). These results
indicated that the reactions of (R,R)-1 and the
nitrenoid precursors 3 and 4, in benzonitrile in the
absence of NMO, did not give the common nitrenoid
species but a different active species, respectively, below
some temperature (Figs. 2 and 3) (Scheme 3).4,5 Fur-
thermore, both the reactions of methyl o- and p-nitro-
phenyl sulfides with (R,S)-2 as the catalyst showed the
ee-break. This suggested that the ee-break was not
attributable to the reversible formation of the Mn(ni-
trenoid)–sulfide complexes but to some other factor(s),
though we could not completely rule out the possibility
of the complex formation.
Mn
L
b
Ar'
a
S
NTs
+
ArI=NTs
c
Mn
L
Ar'SMe
Mn
L
Ts
N
d
Mn
L
= salen ligand
Scheme 3.
showed the ee-break at ca. 20°C. Although the enan-
tiomer ratios observed in chlorobenzene and benzo-
nitrile in the absence of NMO are different, this is
considered to be partly attributable to the difference in
coordinating ability of these solvents.
Based on these results, we propose the following reac-
tion pathway for Mn(salen)-catalyzed sulfimidation, in
which both the Mn–ArINTs adduct and the nitrenoid
species take part (Scheme 3): the reaction of Mn(salen)
and ArIꢀNTs first provides the Mn–ArINTs adduct
(path a), which undergoes sulfimidation (path b) and
transformation to an Mn–nitrenoid species (path c)
We also examined the sulfimidation of methyl phenyl
sulfide in benzonitrile in the absence of NMO (Fig. 4).
The reactions with (R,S)-2 and nitrenoid precursors 3
and 4 exhibited almost the same enantiomer ratios at
every reaction temperature, though the enantioselectivi-
ties of the reactions were depressed considerably as
compared with those of the reactions in chlorobenzene
in the presence of NMO. On the other hand, the
enantioselectivities of the reactions with (R,R)-1
depended on the nitrenoid precursor used: the reaction
with 3 showed higher enantioselectivity than the reac-
tion with 4 at temperatures of −10 to 80°C.6 This result
suggested that the reactive intermediates generated by
treating (R,R)-1 with 3 and 4 in benzonitrile in the
absence of NMO were not the same ones. The reactions
in chlorobenzene in the absence of NMO also showed
similar results: the enantioselectivities of the reactions
with (R,R)-1 depended on the nitrenoid precursor used:
the reaction with 3 showed higher enantioselectivity
than the reaction with 4 at temperatures of −20 to
60°C, though the enantiomer ratio correlated linearly
with the temperature (Fig. 5). On the other hand, the
reactions with (R,S)-2 and nitrenoid precursors 3 and 4
Figure 4. Reactions of methyl phenyl sulfide in benzonitrile.
Figure 3. Reactions of methyl o-nitrophenyl sulfide in benzo-
nitrile.
Figure 5. Reaction of methyl phenyl sulfide in chlorobenzene
in the absence of NMO.

3888
C. Ohta, T. Katsuki / Tetrahedron Letters 42 (2001) 3885–3888
competitively. At lower temperature, sulfimidation by
the Mn–ArINTs adduct (path b) is the major pathway
and the enantioselectivity of the reactions with 2 is
dependent on the nitrenoid precursors used, while the
adduct transforms to the Mn–nitrenoid species (path c)
preferentially at higher temperature and the enantio-
selectivity becomes independent of the precursors (Figs.
2 and 3). As described above, the reactions of methyl
phenyl sulfide in the presence of NMO showed a linear
relationship, regardless of the catalysts used. This sug-
gests that coordination of NMO at an apical position
accelerates the conversion of the Mn–ArINTs adduct to
the nitrenoid species (path c).7
the product was determined by HPLC analysis using
Daicel Chiralpak AD, (hexane/2-propanol=1/1).
The reaction of methyl p-nitrophenyl sulfide was per-
formed under the same conditions as methyl o-nitro-
phenyl sulfide and the enantiomeric excess was
determined by HPLC analysis using Daicel Chiralcel
OJ, (hexane/2-propanol=1:1).
The same reaction was repeated three times at each
measurement point and the average ln p values are
given in Figs. 1–5. The scattering of the measured ln p
values was within 0.1.
In conclusion, we were able to demonstrate that two
active species, Mn–ArINTs adduct and Mn–nitrenoid,
take part in the Mn(salen)-catalyzed sulfimidation,
depending on the reaction conditions used.
Although ee-breaks were observed in both the reactions
with (R,R)-1 and (R,S)-2 as catalysts in the absence of
NMO, ln p difference between the reactions with 3 and
4 at lower temperature was observed mostly when
(R,R)-1 was used as catalyst. This suggests that the
aryliodo moiety of the Mn–ArINTs adduct derived
from (R,R)-1 is probably located close to the 3(3%)-sub-
stituents of the salen ligand, while the aryliodo moiety
of the adduct derived from 2 is distal from the 3(3%)-
substituents. However, more information on the struc-
tures of Mn–ArINTs adducts is required for further
discussion.
Acknowledgements
Financial support from a Grant-in-Aid for Scientific
Research (Priority Areas, No. 706: Dynamic Control of
Stereochemistry) from the Ministry of Education, Sci-
ence, Sports and Culture, Japan, is gratefully acknowl-
edged.
Typical experimental procedures were exemplified by
sulfimidation of methyl phenyl sulfide with (R,R)-1 and
that of methyl o-nitrophenyl sulfide with (R,S)-2.
References
1. Nishikori, H.; Ohta, C.; Oberlin, E.; Irie, R.; Katsuki, T.
Tetrahedron 1999, 55, 13937–13946.
2. (a) Corey, E. J.; Noe, M. C. J. Am. Chem. Soc. 1996, 118,
11038–11053; (b) Corey, E. J.; Noe, M. C. J. Am. Chem.
Soc. 1996, 118, 319–329.
Sulfimidation of methyl phenyl sulfide: Complex (R,R)-1
(2.3 mg, 2.5 mmol) and N-methylmorpholine N-oxide (3
mg, 25 mmol) were suspended in dry toluene (0.5 ml),
azeotropically concentrated in vacuo, and resuspended
in chlorobenzene (0.25 ml). To this suspension was
added methyl phenyl sulfide (11.7 ml, 0.1 ml) and it was
allowed to cool or warm to the appropriate tempera-
ture. [N-(p-Toluenesulfonyl)imino]phenyliodinane (18.6
mg, 0.05 mmol) was added to the mixture and stirred
for 6 h at the temperature (60, 40 and 27°C, respec-
tively) [for 12 h at 0 and −20°C]. The reaction mixture
was directly subjected to silica-gel column chromato-
graphy (hexane:ethyl acetate=7:3–3:7) to give the cor-
responding sulfimide. The enantiomeric excess of the
product was determined by HPLC analysis using Daicel
Chiralcel OD-H, (hexane/2-propanol=1/1).
3. Bushmann, H.; Scharf, H.-D.; Hoffmann, N.; Esser, P.
Angew. Chem., Int. Ed. Engl. 1991, 30, 477–515.
4. A copper–nitrenoid species has been considered to be the
active intermediate in Cu(diimine)-catalyzed asymmetric
aziridination, based on the experimental data that enan-
tioselectivities of aziridination of various olefins are not
affected by [N-(p-toluenesulfonyl)imino]aryliodinanes
used. However, all the experiments were performed at 25°C:
Li, Z.; Quan, R. W.; Jacobsen, E. N. J. Am. Chem. Soc.
1995, 117, 5889–5890.
5. In Mn(porphyrin)-catalyzed aziridination, an MnIV
–
PhINTs adduct has been proposed as the active intermedi-
ate based on UV–Vis and EPR studies: Lai, T.-S.; Kwong,
H.-L.; Che, C.-M.; Peng, S.-M. Chem. Commun. 1997,
2373–2374.
Sulfimidation of methyl o-nitrophenyl sulfide: Complex
(R,S)-2 (2.3 mg, 2.5 mmol) was dissolved in dry toluene
(0.5 ml), azeotropically concentrated in vacuo, and
re-dissolved in benzonitrile (0.25 ml). To this solution
were added methyl o-nitrophenyl sulfide (16.9 mg, 0.1
mmol) and MS-3A (25 mg) and stirred for half an hour
at room temperature. After the reaction temperature of
the mixture was adjusted to the appropriate one, [N-(p-
toluenesulfonyl)imino]phenyliodinane (18.6 mg, 0.05
mmol) was added under nitrogen and stirred for 3 h at
the temperature (80, 60, 40 and 27°C, respectively) [for
6 h at 10 and 0°C; for 12 h at −10°C]. The reaction
mixture was directly subjected to silica-gel column
chromatography (hexane:ethyl acetate=7:3–3:7) to give
6. As the freezing point of benzonitrile is −13°C, the enan-
tiomer ratios of the reactions in benzonitrile were measured
at temperatures of −10 to 80°C.
7. For similar trans-donor ligand effects, see: (a) Renaud, J.-
P.; Battioni, P.; Bartoli, J. F.; Mansuy, D. J. Chem. Soc.,
Chem. Commun. 1985, 888–889; (b) Anelli, P. L.; Banfi, S.;
Montanari, F.; Quici, S. J. Chem. Soc., Chem. Commun.
1989, 779–780; (c) Yamaguchi, K.; Watanabe, Y.; Mor-
ishima, I. Inorg. Chem. 1992, 31, 156–157; (d) Schwenkreis,
T.; Berkessel, A. Tetrahedron Lett. 1993, 34, 4785–4788; (e)
Irie, R.; Hosoya, N.; Katsuki, T. Synlett 1994, 255–256; (f)
Yamada, T.; Imagawa, K.; Nagata, T.; Mukaiyama Bull.
Chem. Soc. Jpn. 1994, 67, 2248–2256.
the corresponding sulfimide. The enantiomeric excess of
.