Titanium-Catalyzed, Asymmetric Sulfoxidation of Alkyl Aryl Sulfides with Optically Active Hydroperoxides

Article abstract of DOI:10.1021/jo980243y

The Ti-catalyzed, asymmetric oxidation of alkyl aryl sulfides by enantiomerically pure hydroperoxides (ee >99%) has been examined. Enantioselectivities with ee values up to ca. 80% were achieved for the oxygen transfer from (S)-(-)-1-phenylethyl hydroperoxide 2a to methyl phenyl and methyl p-tolyl sulfide 1a in CCl₄ as solvent, but with much overoxidation to the corresponding sulfone 4. Detailed mechanistic studies showed that the enantioselectivity of the sulfide 1a oxidation results from a combination of a rather low (ee values >20%) asymmetric induction in the sulfoxidation and an effective kinetic resolution (ee values ca. 80% at 85% sulfide conversion) of the sulfoxide 3a by enantioselective oxidation to the sulfone 4a. The overoxidation (loss of chemoselectivity) is due to sulfoxide coordination to the Ti metal to generate a template in which the oxygen atom is intramolecularly transferred from the bound and activated, optically active hydroperoxide to the ligated sulfoxide in a stereocontrolled manner.

Full text of DOI:10.1021/jo980243y

J . Org. Chem. 1998, 63, 3423-3428
3423
Tita n iu m -Ca ta lyzed , Asym m etr ic Su lfoxid a tion of Alk yl Ar yl
Su lfid es w ith Op tica lly Active Hyd r op er oxid es
Waldemar Adam, Marion N. Korb,* Konrad J . Roschmann,† and Chantu R. Saha-Mo¨ller
Institute of Organic Chemistry, University of Wu¨rzburg, Am Hubland, D-97218 Wu¨rzburg, Germany
Received February 10, 1998
The Ti-catalyzed, asymmetric oxidation of alkyl aryl sulfides by enantiomerically pure hydroper-
oxides (ee >99%) has been examined. Enantioselectivities with ee values up to ca. 80% were
achieved for the oxygen transfer from (S)-(-)-1-phenylethyl hydroperoxide 2a to methyl phenyl
and methyl p-tolyl sulfide 1a in CCl4 as solvent, but with much overoxidation to the corresponding
sulfone 4. Detailed mechanistic studies showed that the enantioselectivity of the sulfide 1a oxidation
results from a combination of a rather low (ee values <20%) asymmetric induction in the
sulfoxidation and an effective kinetic resolution (ee values ca. 80% at 85% sulfide conversion) of
the sulfoxide 3a by enantioselective oxidation to the sulfone 4a . The overoxidation (loss of
chemoselectivity) is due to sulfoxide coordination to the Ti metal to generate a template in which
the oxygen atom is intramolecularly transferred from the bound and activated, optically active
hydroperoxide to the ligated sulfoxide in a stereocontrolled manner.
In tr od u ction
titanium-containing complexes are known. For the lat-
ter, Kagan9 optimized their Sharpless-modified titanium-
catalyzed sulfoxidation [cumyl hydroperoxide/Ti(OiPr)4/
(+)-DET/iPrOH ) 5:1:4:4] to achieve enantiomeric excesses
up to 96% for alkyl aryl sulfides. Recently Scettri10
employed furyl hydroperoxides (ee values up to >95%)
instead of cumyl hydroperoxide, while Modena11 intro-
duced a nonracemic amino triol as chiral ligand (ee up
to 84%). Uemura12 utilized (R)-binaphthol as chiral
auxiliary, which not only oxidizes sulfides, but also
promotes the kinetic resolution of sulfoxides. The latter
has also been observed for (S,S)-1,2-diphenylethane-1,2-
diol13 and for amino triol11 as chiral auxiliaries.
Optically active sulfoxides have been increasingly used
as building blocks and chiral auxiliaries in the asym-
metric synthesis of pharmaceutical products;1 therefore,
their enantioselective preparation is still of importance
today. The main route to nonracemic sulfoxides consti-
tutes the Andersen method,2 namely the reaction of
diastereomerically pure sulfinates with organometallic
reagents. Alternatively, the asymmetric oxidation of
sulfides has received much attention during the past 15
years. The use of chiral oxaziridines3 as oxidants or the
Sharpless-modified procedures of Modena4 and Kagan5
led to high enantioselectivities (up to ee 90%), but the
sulfoxidations have been conducted stoichiometrically
with respect to Ti(OiPr)4 and the chiral auxiliary. Of
paramount interest is, therefore, the development of
efficient catalytic systems.
In such catalytic, enantioselective oxidations, the
asymmetric induction is exercised by a chiral diol ligand,
e.g., tartrate or binaphthol, in combination with an
achiral hydroperoxide as the oxygen donor. First at-
tempts to use optically active sugar-derived hydroperox-
ides for the asymmetric sulfoxidation have been made
for alkyl aryl sulfides, but only low ee values (up to 26%)
were obtained for the sulfoxide.14 In the following we
report the Ti-catalyzed asymmetric oxidation of methyl
p-tolyl sulfide 1a (Scheme 1) with optically active hydro-
In this context, besides the highly enantioselective
enzyme-catalyzed sulfoxidations,6 catalyzed sulfoxida-
tions with (salen)manganese,7 (salen)vanadium,8 and
† Undergraduate research participant, Spring 1997.
(1) (a) Mata, E. G. Phosphorus Sulfur Silicon 1996, 117, 231-286.
(b) Kagan, H. B.; Rebiere, F.; Samuel, O. Phosphorus Sulfur Silicon
1991, 58, 89-110. (c) Carren˜o, M. C. Chem. Rev. 1995, 95, 1717-1760.
(2) (a) Anderson, K. K. Tetrahedron Lett. 1962, 93-95. (b) Rebiere,
F.; Samuel, O.; Ricard, L.; Kagan, H. B. J . Org. Chem. 1991, 56, 5991-
5999. (c) Whitesell, J . K.; Wong, M.-S. J . Org. Chem. 1994, 59, 597-
601.
(7) (a) Palucki, M.; Hanson, P.; J acobsen, E. N. Tetrahedron Lett.
1992, 33, 7111-7114. (b) Noda, K.; Hosoya, N.; Yanai, K.; Irie, R.;
Katsuki, T. Tetrahedron Lett. 1994, 35, 1887-1890.
(8) (a) Nakajima, K.; Kojima, M.; Fujita, J . Chem. Lett. 1986, 1483-
1486. (b) Bolm, C.; Bienewald, F. Angew. Chem., Int. Ed. Engl. 1995,
34, 2883.
(3) (a) Davis, F. A.; McCauley, J . P., J r.; Chattopadhyay, S.; Harakal,
M. E.; Towson, J . C.; Watson, W. H.; Tavanaiepour, I. J . Am. Chem.
Soc. 1987, 109, 3370-3377. (b) Davis, F. A.; Thimma Reddy, R.;
Weismiller, M. C. J . Am. Chem. Soc. 1989, 111, 5964-5965. (c) Davis,
F. A.; Thimma Reddy, R.; Han, W.; Carroll, P. J . J . Am. Chem. Soc.
1992, 114, 1428-1437.
(4) Di Furia, F.; Modena, G.; Seraglia, R. Synthesis 1984, 325-327.
(5) (a) Pitchen, P.; Kagan, H. B. Tetrahedron Lett. 1984, 25, 1094-
1052. (b) Pitchen, P.; Dun˜ach, E.; Deshmukh, M. N.; Kagan, H. B. J .
Am. Chem. Soc. 1984, 106, 8188-8193. (c) Zhao, S. H.; Samuel, O.;
Kagan, H. B. Tetrahedron 1987, 43, 5135-5144. (d) Brunel, J .-M.;
Diter, P.; Duetsch, M.; Kagan, H. B. J . Org. Chem. 1995, 60, 8086-
8088.
(9) (a) Brunel, J .-M.; Kagan, H. B. Synlett 1996, 404-406. (b)
Brunel, J .-M.; Kagan, H. B. Bull. Soc. Chim. Fr. 1996, 133, 1109-
1115.
(10) (a) Scettri, A.; Bonadies, F.; Lattanzi, A. Tetrahedron: Asym-
metry 1996, 7, 629-632. (b) Lattanzi, A.; Bonadies, F.; Scettri, A. Tetra-
hedron: Asymmetry 1997, 8, 2141-2151. (c) Lattanzi, A.; Bonadies,
F.; Senatore, A.; Soriente, A.; Scettri, A. Tetrahedron: Asymmetry 1997,
8, 2473-2478.
(11) Di Furia, F.; Licini, G.; Modena, G.; Motterle, R.; Nugent, W.
A. J . Org. Chem. 1996, 61, 5175-5177.
(12) (a) Komatsu, N.; Hashizume, M.; Sugita, T. Uemura, S. J . Org.
Chem. 1993, 58, 4529-4533. (b) Komatsu, N.; Hashizume, M.; Sugita,
T. Uemura, S. J . Org. Chem. 1993, 58, 7624-7626.
(13) Superchi, S.; Rosini, C. Tetrahedron: Asymmetry 1997, 8, 349-
352.
(14) Hamann, H.-J .; Ho¨ft, E.; Mostowicz, D.; Mishnev, A.; Urbanczyk-
Lipkowska, Z.; Chmielewski, M. Tetrahedron 1997, 53, 185-192.
(6) (a) Fu, H.; Kondo, H.; Ichikawa, Y.; Look, G. C.; Wong, C.-H. J .
Org. Chem. 1992, 57, 7265-7270. (b) Ottolina, G.; Pasta, P.; Carrea,
G.; Colonna, S.; Dallavalle, S.; Holland, H. L. Tetrahedron: Asymmetry
1995, 6, 1375-1386.
S0022-3263(98)00243-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/22/1998

3424 J . Org. Chem., Vol. 63, No. 10, 1998
Adam et al.
Sch em e 1
Ta ble 1. Ch em o- a n d En a n tioselectivity of th e
Su lfoxid a tion of Meth yl p-Tolyl Su lfid e (1a ) w ith
(S)-1-P h en yleth yl Hyd r op er oxid e (2a ) Ca ta lyzed by
Ti(OiP r )4 u n d er Va r iou s Rea ction Con d ition s
F igu r e 1. Time profile of the enantioselectivity (% ee of
sulfoxide 3a ) and the chemoselectivity (% sulfone 4a ) in the
oxidation of sulfide 1a by R*OOH 2a /Ti(OiPr)4.
selectivitya
enantio
sulfone 4a ratio was observed as in CH2Cl2, but with a
higher ee value (47%). Still higher enantioselectivities
were obtained when the reaction was conducted in CCl4
(entry 7), but otherwise the same conditions as in entry
3 were used. Moreover, a variation of the reaction
temperature from +20 to -20 °C caused a significant
increase of the ee value from 40% at +20 °C (entry 4) to
75% at -20 °C (entry 6).
Most indicative is the variation of the amount of
hydroperoxide 2a that is used (entries 5-7) at -20 °C
in CCl4. An increase from 1.0 (entry 5) to 1.2 equiv (entry
6) raised the enantiomeric excess from 52 to 75%, which
parallels the increase in the amount of overoxidation to
the sulfone 4a from 64 to 84%. However, when 1.5 equiv
of hydroperoxide 2a (entry 7) is employed, no significant
change in the ee value or the sulfoxide 3a :sulfone 4a ratio
is observed compared to when 1.2 equiv is used (entry
6), except that ca. 20% more of the sulfide 1a is con-
sumed.
Also remarkable is the concentration effect of the
sulfide 1a in this asymmetric sulfoxidation (entries 7 and
8). For the dilute (0.13 M, entry 7) solution, the ee value
is drastically higher than that at 0.33 M (entry 8), i.e.,
71 versus 21%. Again this trend to higher enantioselec-
tivity parallels the extent of overoxidation, as reflected
in the sulfoxide 3a :sulfone 4a ratio of 20:80 at 0.13 M
versus 47:53 at 0.33 M (entries 7 and 8). In other words,
the higher the overoxidation to sulfone 4a , the higher
the enantiomeric excess of the sulfoxide 3a .
In a control experiment, the oxidation of sulfide 1a by
the hydroperoxide (S)-(-)-2a was conducted without Ti-
(OiPr)4 catalyst at -20 °C (entry 9) to afford exclusively
the sulfoxide 3a in only 9% yield with no enantiomeric
excess. Thus, under these conditions, the direct sulfoxi-
dation by the optically active hydroperoxide 2a without
Ti(OiPr)4 catalyst is negligible and affords racemic sul-
foxide 3a .
sulfideb R*OOHc
sol-
T
t
convna chemo (S)-(-)-3a
entry 1a (M) 2a (equiv) vent (°C) (h) (%)
3a :4a
ee (%)
1
2
3
4
5
6
7
8
9
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.33
0.13
1.2
1.5
1.5
1.2
1.0
1.2
1.5
1.5
1.2
CH2Cl2 -20
CH2Cl2 -20
4
4
95
98
79:21
50:50
49:51
24:76
36:64
16:84
20:80d
47:53
100:0
20
31
47
40
52
75
71
21
0
iPrOH -20 48 >99
CCl4
CCl4
CCl4
CCl4
CCl4
CCl4
20
-20
-20
-20
-20
2
3
4
5
5
68
64
69
88
95
9
-20 16
a
Determined by HPLC analysis of the crude reaction mixture
on a chiral column (Daicel Chiralcel OD-H); error (2% of the
b
stated value, mass balance g90%. 1.0 equiv of sulfide 1a and
0.05 equiv of Ti(OiPr)4 except entry 3 (0.1 equiv) and entry 9,
without Ti(OiPr)4. c ee >99%. Yields of isolated products: sul-
d
foxide 3a (14%), sulfone 4a (65%).
peroxides (S)-(-)-2 as asymmetric inductors, readily
available for the preparative scale by horseradish per-
oxidase-catalyzed kinetic resolution.15 To assess the
scope of this catalytic, enantioselective sulfoxidation, a
variety of alkyl aryl sulfides were used in which the size
of the alkyl group and the electronic properties of the aryl
substituent have been probed.
Resu lts
To establish the optimal reaction conditions for the
asymmetric sulfoxidation with optically active hydro-
peroxides under Ti catalysis, the solvent, temperature,
amount of hydroperoxide, and sulfide concentration were
varied in the oxidation of methyl p-tolyl sulfide (1a ) by
the enantiomerically pure (ee >99%) (S)-(-)-1-phenyl-
ethyl hydroperoxide (2a ). The results of these optimiza-
tions are displayed in Table 1. When CH2Cl2 was used
as solvent, an enantiomeric excess (ee) of 20% (Table 1,
entry 1) was obtained for the sulfoxide (S)-(-)-3a . By
raising the amount of hydroperoxide 2a from 1.2 to 1.5
equiv (entries 1 and 2), the sulfoxide yield was reduced
from 79 to 50% due to increased overoxidation to the
sulfone 4a , while the ee values were raised from 20 to
31%. In both cases the conversion of the sulfide 1a was
almost complete. With 2-propanol as solvent (entry 3),
at complete conversion, the same (49:51) sulfoxide 3a :
A time profile of the oxidation of 1a with the hydro-
peroxide 2a (1.5 equiv) clearly shows the dependence of
the enantiomeric excess of the sulfoxide 3a on the amount
of sulfone 4a formation (Figure 1). Remarkable is the
very fast formation of the sulfone 4a , so that after 15
min the ratio 3a :4a is 43:57 at a conversion of 58%. The
enantiomeric excess of the sulfoxide 3a is only ca. 20%
at the very beginning (ca. 2 min) of the reaction, when
only little sulfone 4a is formed.
(15) (a) Adam, W.; Hoch, U.; Lazarus, M.; Saha-Mo¨ller, C. R.;
Schreier, P. J . Am. Chem. Soc. 1995, 117, 11898-11901. (b) Adam,
W.; Korb, M. N. Tetrahedron: Asymmetry 1997, 8, 1131-1142.

Titanium-Catalyzed, Asymmetric Sulfoxidation
J . Org. Chem., Vol. 63, No. 10, 1998 3425
Ta ble 3. Oxid a tion of Meth yl p-Tolyl Su lfid e (1a )a w ith
(S)-1-P h en yleth yl Hyd r op er oxid e (2a )b in th e P r esen ce of
Differ en t Ad d itives
Sch em e 2
t
convnc
(%)
ee (%)c
entry
1
additive
(equiv)d (h)
3a :4a c (S)-(-)-3a
(S)-3a
(0.05)
10
82
12:88
74
(ee 86%)
4a
Ph2SO
R*OHe
(ee >99%)
2
3
4
(0.05)
(0.05)
(1.50)
4
7
23
80
76
87
40:60
33:67
18:82
38
53
80
a
b
0.13 M sulfide 1a in CCl4 at -20 °C. 1.2 equiv of (S)-(-)-2a
(ee >99%), except run 1 and 4 (1.5 equiv). c Determined by HPLC
analysis of the crude reaction mixture on a chiral column (Daicel
d
Chiralcel OD-H), error (2% of the stated value. Relative to
sulfide 1a . e (S)-1-Phenylethanol (ee >99%).
Ta ble 2. Kin etic Resolu tion of th e Ra cem ic Meth yl
p-Tolyl Su lfoxid e (3a ) w ith (S)-1-P h en yleth yl
Hyd r op er oxid e (2a ) Ca ta lyzed by Ti(OiP r )4
3a f 4a , the high enantioselectivity (ee up to 80%)
achieved through the “overoxidation” 1a f 4a , as dis-
played in Figure 1, could not be obtained. (This low
efficiency of the kinetic resolution is also reflected in the
small stereoselectivity factor s16 of 1.8 and 1.9 in Table
2, entry 6 and 7.) For this reason, a number of additives
were tested to assess whether they influence the enan-
tioselectivity of the Ti-catalyzed sulfide 1a oxidation by
hydroperoxide (S)-2a (Table 3).
R*OOHa Ti(OiPr)4
entry 2a (equiv) (equiv) (°C)
T
t
convnb
eeb (%)
(h)
3a (%) (S)-(-)-3a sc
1
2
3
4
5
6
0.50
0.50
0.50
0.75
0.50
0.50
0.25e
0.50
0.25e
0.05
0.05
0.15d
0.25d
1.0
20
0
4
7
57
56
53
75
52
50
75
52
76
7
10
11
30
17
19
39
22
44
1.2
1.3
1.3
1.6
1.6
1.7
1.8
1.8
1.9
-20 148
-20
-20
-20
-20
-20
-20
96
4
0.75
0.75
0.5
0.5
5.0
7
5.0
Thus, equimolar (with respect to the Ti catalyst)
amounts of (S)-sulfoxide 3a (ee 86%) directly from the
start of the sulfide 1a oxidation (entry 1, Table 3)
exercised no appreciable effect in the ee value or 3a :4a
ratio (compare with entry 7, Table 1), except that the
reaction time was doubled for approximately the same
extent of sulfide 1a conversion. The addition of sulfone
4a (entry 2, Table 3) or the extraneous diphenyl sulfoxide
(entry 3, Table 3) actually lowered the enantioselectivity,
but the extent of “overoxidation” was reduced for ap-
proximately the same amount of sulfide 1a conversion
(compare with entry 6, Table 1). Finally, when (S)-1-
phenylethanol (entry 4, Table 3) was added at the start
of the reaction, again only the reaction time was pro-
longed (compare with entry 7, Table 1), while the chemo-
and enantioselectivities remained the same and the ee
value was raised by only 9%. In view of the large amount
of (S)-1-phenylethanol (1.5 equiv) that is present from
the very beginning of the reaction, this small increase in
the enantioselectivity is not sufficient to allow a conclu-
sion that this alcohol influences the enantioselectivity
significantly. Therefore, it may be stated that none of
these additives promote significantly the enantioselec-
tivity of this oxygen-transfer process.
a
ee >99%, 0.13 M sulfoxide 3a in CCl4 except entry 7 (0.03 M
b
3a ). Determined by HPLC analysis of the crude reaction mixture
on a chiral column (Daicel Chiralcel OD-H); error (2% of the
stated value; mass balance >90%. c Stereoselectivity factor, de-
termined according to ref 16. At the start 0.05 equiv of Ti(OiPr)4
d
was added, followed by additional aliquots (0.05 equiv) of Ti(OiPr)4
during the course of the reaction. e An additional 0.25 equiv of
hydroperoxide 2a was added.
To examine whether one of the enantiomers is prefer-
rably oxidized to the sulfone 4a , the separate optically
active sulfoxides (R)-3a (ee 80%) and (S)-3a (ee 86%) were
treated with 0.5 equiv of the hydroperoxide (S)-2a
(Scheme 2). This hydroperoxide (S)-2a was consumed
faster (ca. 10 times) in the reaction with the sulfoxide
(R)-3a than with the enantiomer (S)-3a , and the enan-
tiomeric excess of the sulfoxide (S)-3a increased from 86
to 95%.
All of these results evidently suggest that during the
course of the sulfide 1a oxidation the resulting sulfoxide
3a is kinetically resolved by “overoxidation” to the sulfone
4a . To confirm this, the racemic sulfoxide 3a was
submitted to the oxidation with hydroperoxide (S)-2a
(Table 2). Whereas at 20 and 0 °C the hydroperoxide
(S)-2a was completely consumed within 4-7 h (entries
1 and 2), the reaction at -20 °C (entry 3) took about 6
days with further addition of the Ti catalyst necessary.
In all cases (ca. 55% conversion) the ee value of the
remaining sulfoxide 3a was around 10%, but at a
conversion of 75% (entry 4) the ee value was raised to
30%. With stoichiometric or excess (5.00 equiv) amounts
of the Ti catalyst, the reaction times were reduced to 45
min for ca. 50% conversion, but the ee values were only
17-19% (compare entries 5 and 6 with 4). After further
addition of 0.25 equiv of hydroperoxide (S)-2a , the ee
value rose to 39% (entry 6). The last run (entry 7) was
conducted to test concentration effects (4 times more
dilute), but there were no significant changes in the
conversion and enantiomeric excess of the sulfoxide 3a .
While the ee values (up to ca. 40%) in Table 2 definitely
confirm that kinetic resolution is at play in the oxidation
The high extent of sulfone 4a formation (overoxidation)
implicates a pronounced nucleophilic character of the
R*OOH/Ti(OiPr)4 oxidant, especially since the oxidation
of the sulfide 1a by the hydroperoxide (S)-2a without the
Ti catalyst (Table 1, entry 8) gave exclusively the sul-
foxide 3a , which demonstrates the electrophilic character
of this hydroperoxide. To assess the electronic nature
of the R*OOH/ Ti(OiPr)4 oxidant, the established17 thi-
(16) Kagan, H. B.; Fiaud, J . C. Top. Stereochem. 1988, 18, 249-
330.
(17) (a) Adam, W.; Haas, W.; Sieker, G. J . Am. Chem. Soc. 1984,
106, 5020-5022. (b) Adam, W.; Du¨rr, H.; Haas, W.; Lohray, B. B.
Angew. Chem., Int. Ed. Engl. 1986, 25, 101-103. (c) Adam, W.; Lohray,
B. B. Angew. Chem., Int. Ed. Engl. 1986, 25, 188-189. (d) Adam, W.;
Haas, W.; Lohray, B. B. J . Am. Chem. Soc. 1991, 113, 6202-6208. (e)
Adam, W.; Golsch, D.; Sundermeyer, J .; Wahl, G. Chem. Ber. 1996,
129, 1177-1182. (f) Adam, W.; Golsch, D. J . Org. Chem. 1997, 62, 115-
119.

3426 J . Org. Chem., Vol. 63, No. 10, 1998
Adam et al.
Ta ble 4. Oxid a tion of Th ia n th r en e 5-oxid e (SSO) by
(S)-(-)-1-P h en yleth yl Hyd r op er oxid e (2a ) Ca ta lyzed by
Ti(OiP r )4
Sch em e 3
product ratiob
R*OOHa
t
solvent 2a (equiv) (h) ∑SOSOc SSO2
SOSO2
XSO
d
e
f
CH2Cl2
CCl4
0.1
0.1
0.5
0.5
70.3
57.6
21.6
27.9
7.3
14.5
0.27
0.37
a
(S)-2a (ee >99%), equivalents relative to SSO (0.094 M).
b
Determined by HPLC analysis [reversed phase column (Knauer
Eurospher 100 C-18)] error (2% of the stated value; SSO
d
conversion ca. 10%. c Sum of cis- and trans-5,10-dioxides. 5,5-
Dioxide. e 5,5,10-Trioxides. f Calculated according to ref 17b.
Ta ble 5. Asym m etr ic Su lfoxid a tion of P r och ir a l Su lfid es
1 w ith (S)-1-P h en yleth yl Hyd r op er oxid e (2a ) Ca ta lyzed
by Ti(OiP r )4
asymmetric oxidation. Of interest is the ee value of 62%
for p-tolyl n-butyl sulfoxide, which is higher than reported
for the modified Sharpless procedure9 (ee 25%).
selectivitya
The change of the aryl group in the methyl sulfide 1
to probe electronic effects displayed no regular trends
(entries 5-7) in the enantio- and chemoselectivity. The
highest ee value of 79% was observed for the phenyl
group (entry 5), but much lower enantioselectivities were
obtained both for the electron-donating p-anisyl (entry
6, ee 31%) and the electron-accepting p-nitrophenyl group
(entry 7, ee 15%). These ee values are difficult to
compare because the extent of overoxidation (chemose-
lectivity) varies significantly (entries 5-7), since we have
seen previously for the model sulfide 1a that the enan-
tioselectivity increases with higher amount of sulfone 4a .
Also, the oxidation of the methyl 2-naphthyl sulfide with
1.2 equiv of the hydroperoxide 2a (entry 8) gave only a
moderate enantioselectivity (ee 49%). As expected for
dialkyl sulfides,9,12 low enantioselectivities were observed
in the oxidation of the benzyl (entry 9, ee 10%) and the
n-octyl (entry 10, ee 20%) methyl sulfides. This low
enantiomeric excess for the oxidation of methyl n-octyl
sulfide was raised only to 32% by double asymmetric
induction with (S)-(-)-1-phenylethyl hydroperoxide (2a )
and the optically active tartrate D-(-)-DET under Ka-
gan’s9 modified Sharpless conditions.
enantio
sulfideb
R1
p-tolyl
t
convnb chemo (S)-(-)-3
entry
R2
(h)
(%)
3:4
ee (%)
1
2
3
4
5
6
7
8
9
Me
Et
5
4
57
29
6
23
6
4
88
89
82
82
79
97
93
73
93
92
20:80
17:83
21:79
23:77
19:81
56:44
41:59
32:68
68:32
34:66
71
72
65
62
79
31
15
p-tolyl
p-tolyl
iPr
nBu p-tolyl
Me
Me
Me
Me
Me
Me
phenyl
p-anisyl
p-nitrophenyl
2-naphthyl
benzyl
24 (49)c
8
20
4
4
10
n-octyl
a
Determined by HPLC analysis of the crude reaction mixture
on a chiral column (Daicel Chiralcel OD-H or OB-H), error (2%
b
of the stated value, mass balance >90%. 0.13 M in CCl4 at -20
°C, 1.5 equiv of (S)-2a (ee >99%) and 0.05 equiv of Ti(OiPr)4,
preparative scale. c In parantheses for 1.2 equiv of hydroperoxide
(S)-2a .
anthrene 5-oxide (SSO) probe was employed and the XSO
values were determined in CH2Cl2 and CCl4 (Table 4).
Indeed, even at as little as 10% SSO conversion, not only
SSO2 (22-28%) but considerable amounts of SOSO2 (7-
15%) were formed; therefore, in CCl4 there was twice as
much overoxidation as in CH2Cl2. These results establish
unequivocally the high extent of overoxidation for the
R*OOH/ Ti(OiPr)4 oxidant.
Mech a n istic Discu ssion
The highest enantioselectivities (ee values up to 80%)
for the Ti(OiPr)4-catalyzed asymmetric oxidation of the
model sulfide 1a (0.13 M) by the hydroperoxide (S)-2a
(1.5 equiv) are achieved in CCl4 at -20 °C. More
important, such a high enantioselectivity (ee value) can
only be obtained at the expense of chemoselectivity
(sulfide 3a :sulfone 4a ratio), i.e., high degree (>80%) of
overoxidation of the sulfoxide 3a to its sulfone 4a . This
is most evidently expressed in Figure 1, in which at 40%
sulfide 1a consumption (after a 2-min reaction time,
during the earlier stage of the oxidation), the ee value of
sulfoxide 3a amounts to e20% (negligible overoxidation
to sulfone 4a ), but at high (g80%) sulfide 1a conversion
(during the later stage of the oxidation) the ee value
levels off to ca. 80% (extensive overoxidation to sulfone
4a ). Loss of enantioselectivity due to the unselective,
direct oxidation by the hydroperoxide (S)-2a [no Ti(OiPr)4
catalyst] is negligible, as established by a control experi-
ment (Table 1, entry 9).
To explore whether other optically active hydroperox-
ides are more effective in the asymmetric sulfoxidation
of sulfide 1a , the derivatives (S)-(-)-2b and (S)-(-)-2c
were employed (Scheme 1) as alternatives. The oxidation
of sulfide 1a by the indanyl hydroperoxide (S)-(-)-2b (1.5
equiv) afforded after 98% conversion an enantiomeric
excess of 22% for sulfoxide 3a . In contrast to the
hydroperoxide 2a , the main sulfoxide enantiomer pos-
sessed the (R) configuration. With the furyl hydroper-
oxide (S)-(-)-2c (1.5 equiv), the sulfoxide 3a was obtained
with an ee value of only 7% after 96% conversion. In
view of these low enantioselectivities, further work with
these optically active hydroperoxides was abandoned.
For general scope of the asymmetric sulfoxidation by
the (S)-(-)-1-phenylethyl hydroperoxide (2a ), the set of
structurally varied alkyl aryl sulfides (Table 5) were
oxidized under the optimal conditions worked out in
Table 1 [1.5 equiv of (S)-2a , CCl4, -20 °C, g80% conver-
sion of the sulfide]. Replacement of the methyl substitu-
ent in the p-tolyl sulfide 1 by sterically more demanding
alkyl groups (Table 5, entries 1-4) resulted in no
significant change in the enantioselectivity (ee values
range between 62 and 72%) and the chemoselectivity
[sulfoxide 3:sulfone 4 ratio (ca. 20:80)] for this catalytic
These results are mechanistically readily rationalized
in terms of Scheme 3. The relatively ineffective (ee
e20%) asymmetric sulfoxidation of sulfide 1a is followed
by a more efficient (ee value ca. 80%) kinetic resolution
of the sulfoxide 3a by its enantioselective oxidation to

Titanium-Catalyzed, Asymmetric Sulfoxidation
J . Org. Chem., Vol. 63, No. 10, 1998 3427
reaction (3a f 4a ), the sulfoxide concentration is much
higher than in the oxidation of the corresponding sulfide
1a , in which sulfoxide 3a is generated in situ. The excess
sulfoxide presumably blocks the small amount of the Ti
catalyst, and thus diminution of the reaction rate is
expected in the former process. Such inhibition of the
Ti catalyst activity may be due to multiple sulfoxide
coordination to the metal center.5b,18 This may also be
the reason why in the oxidation of the sulfide 1a a higher
amount of the hydroperoxide (S)-2a (1.5 versus 1.2 equiv,
see Table 1, entry 6 and 7) does not result in a higher
amount of overoxidation to the sulfone 4a and, therefore,
does not change the ee value of the sulfoxide 3a . As a
consequence, the generation of the active oxidizing spe-
cies in this reaction is hindered. Consequently, to avoid
such an inhibitory effect, a sufficient amount of Ti
catalyst (g1.0 equiv) must be present in the kinetic
resolution of the racemic sulfoxide 3a (Table 2, entries
5-7); however, in the presence of excess catalyst (> 1.0
equiv), lower ee values (ca. 40%) were obtained for the
sulfoxide 3a at a conversion of 75% (Table 2, entry 6 and
7) in comparison to the catalytic sulfide 1a oxidation (ee
71-75%, Table 1, entries 6 and 7). This difference may
be explained by the fact that in the case of sulfide
oxidation the enantiomerically enriched (ee ca. 20%)
sulfoxide 3a resulting from the asymmetric sulfoxidation
(1a f 3a ) is kinetically resolved by subsequent enantio-
selective oxidation (3a f 4a ). Therefore, at the same
sulfoxide conversion, the ee values in the sulfide oxidation
(1a f 3a ) should be at least 20% higher than in the case
of the kinetic resolution of the racemic sulfoxide (3a f
4a ).
F igu r e 2. Postulated templates in the oxidation of (S)-
sulfoxide 3a (A) and (R)-3a (B) with (S)-(-)-1-phenylethyl
hydroperoxide (2a ) under Ti(OiPr)4 catalysis.
the sulfone 4a (overoxidation). Precedents for such a
sequence of asymmetric sulfoxidation followed by sulfox-
ide kinetic resolution are documented in the literature.10-13
Evidence for the kinetic resolution of the racemic
sulfoxide 3a in this oxidation is provided by the data in
Table 2, although maximum ee values of only ca. 40%
were obtained (entry 7). More significant is the fact that
the (R) enantiomer of sulfoxide 3a is faster oxidized by
the optically active hydroperoxide (S)-2a than the (S)
enantiomer (Scheme 2). Thus, the combination (R)-3a /
(S)-2a is more effective in the Ti(OiPr)4-catalyzed oxygen
transfer than the (S)-3a /(S)-2a one. This is difficult to
rationalize without invoking a template effect, i.e.,
simultaneous coordination to the titanium metal center
of both the oxygen acceptor (the sulfoxide 3a substrate)
and the oxygen donor (the optically active hydroperoxide
2a ). Moreover, the hydroperoxide 2a is not only ligated
to the titanium by means of an oxygen-metal bond but
activated electrophilically for oxygen transfer by coordi-
nation of the remote oxygen atom of the peroxide bond
(Figure 2). Such sulfoxide coordination to the titanium
metal has been reported18 and accounts adequately for
the appreciable overoxidation even in the earlier stages
of the oxidation.19 In the proposed monomeric template,
inspection of molecular models suggest that the methyl
and phenyl substituents of the hydroperoxide should
point away from the metal center for steric reasons.
Thus, for the sulfoxide (S)-3a /hydroperoxide (S)-2a dia-
stereomeric combination (Figure 2, complex A), signifi-
cant steric interaction is expected between the p-tolyl
substituents of the sulfoxide (S)-3a and the hydroperox-
ide (S)-2a during the oxygen transfer. This steric
hindrance is reduced for the diastereomeric combination
(R)-3a /(S)-2a (Figure 2, complex B), since now a methyl
group of (R)-3a is facing the hydroperoxide phenyl
substituent. In this way, the faster oxidation of the (R)
enantiomer of the sulfoxide 3a becomes evident and the
kinetic resolution of the sulfoxide 3a may be rationalized.
This sulfoxide coordination also explains the appre-
ciable formation of SSO2 and SOSO2 at only ca. 10%
consumption of SSO in its Ti-catalyzed oxidation by
hydroperoxide (S)-2a (Table 4). Thus, the genuine elec-
trophilic character (XSO ca. 0.4 in CCl4) of the (S)-2a /Ti-
(OiPr)4 oxidant is masked by coordination effects to the
titanium metal.17c,d
i
The pronounced retardation of the oxidation in PrOH
compared to that in CCl4 (Table 1, entries 3 and 7)
becomes also interpretable in terms of sulfoxide coordina-
tion to the titanium center. On one hand, the sulfoxide
i
3a is more strongly solvated in the protic PrOH and
should, therefore, diminish sulfoxide coordination; on the
other hand, iPrOH competes with sulfoxide coordination
by facile ligand exchange.
Con clu sion
The asymmetric sulfoxidation of aryl alkyl sulfides by
(S)-(-)-1-phenylethyl hydroperoxide (2a ) at -20 °C in
CCl4 affords good to high enantiomeric excesses for
methyl phenyl and p-tolyl alkyl sulfides (ee up to ca. 80%)
at the expense of substantial overoxidation to the sulfone.
Neither electronic (variation of the aryl group) nor steric
(variation of the alkyl group) effects helped to enhance
the enantiomeric excess. A time profile of the oxidation
of the model substrate p-tolyl methyl sulfide (1a ) with
(S)-(-)-1-phenylethyl hydroperoxide (2a ) showed that the
asymmetric induction of the sulfoxidation is rather low
(ee <20%), but the enantioselectivity is significantly
enhanced through concomitant kinetic resolution of the
sulfoxide 3a ; however, this comes at the expense of
chemoselectivity through overoxidation. Sulfoxide coor-
dination to the titanium center is responsible for the
kinetic resolution through a template effect. Thus, the
oxygen atom is transferred intramolecularly between the
hydroperoxide and sulfoxide, which are simultaneously
Also the long reaction times (g4 days) at -20 °C may
be accounted for in the kinetic resolution of the racemic
sulfoxide 3a when only a catalytic amount of Ti(OiPr)4
is present (Table 2, entries 3 and 4). In this catalytic
(18) (a) Bonchio, M.; Calloni, S.; Di Furia, F.; Licini, G.; Modena,
G.; Moro, S.; Nugent, W. A. J . Am. Chem. Soc. 1997, 119, 6935-6936.
(b) Bonchio, M.; Licini, G.; Modena, G.; Moro, S.; Bortolini, O.; Traldi,
P.; Nugent, W. A. Chem. Commun. 1997, 869-870.
(19) For convenience and simplicity, we assume a mononuclear
template with one Ti atom, but little if any structural data are available
for such metal complexes. The recently suggested theoretical model
(ref 18) is adopted for the Ti-catalyzed oxygen-transfer process.

3428 J . Org. Chem., Vol. 63, No. 10, 1998
Adam et al.
The enantiomeric excess of the sulfoxide 3a , the conversion of
the sulfide 1a , and the sulfoxide 3a :sulfone 4a ratio were
determined on the crude reaction mixture as described above.
Gen er a l P r oced u r e of th e Oxid a tion of th e Su lfid es 1
in CCl4 w ith th e Op tica lly Active Hyd r op er oxid es (S)-
(-)-2a on th e P r ep a r a tive Sca le. The reaction was carried
out as described above by starting with 14.8 µL (50.0 µmol) of
Ti(OiPr)4 in 3.5 mL of CCl4 and the addition of a solution of
sulfide 1 (1.00 mmol) in 2 mL of CCl4 and a solution of 82.9
mg (1.5 mmol) of hydroperoxide (S)-2a in 2 mL of CCl4. After
workup and the determination of the enantiomeric excess of
the sulfoxide 3, the conversion of the sulfide 1, and the
sulfoxide 3:sulfone 4 ratio as described above, the products
were isolated by flash column chromatography [silica gel,
petroleum ether (30-50 °C)/diethyl ether (1:1) f EtOAc]. The
spectral data of the sulfoxides 3 matched those reported.5d
P r oced u r e for th e Kin etic Stu d y of th e Ti-Ca ta lyzed
Oxid a tion of Meth yl p-Tolyl Su lfid e (1a ) w ith th e Hy-
d r op er oxid e (S)-2a . The reaction solution was prepared as
described above and stirred at -20 °C. After the reaction
times stated in Figure 1, 0.1 mL of the mixture was taken out
and immediately hydrolyzed with 8-10 drops aqueous, satu-
rated NH4F solution at room temperature. After 5 min of
stirring, this sample was diluted with 1 mL of CH2Cl2 and
dried over molecular sieves (4 Å), and the suspended material
was filtered. After the solvent was evaporated (20 °C/20 Torr),
the enantiomeric excess of the sulfoxide 3a , the conversion of
the sulfide 1a , and the sulfoxide 3a :sulfone 4a ratio were
determined as described above.
Gen er a l P r oced u r e of th e Kin etic Resolu tion of Ra -
cem ic Meth yl p-Tolyl Su lfoxid e (3a ) by (S)-(-)-1-P h en yl-
eth yl Hyd r op er oxid e (2a ). To a solution of Ti(OiPr)4 (0.05-
5.0 equiv) and (S)-(-)-1-phenylethyl hydroperoxide (2a ) (0.5-
0.75 equiv) in 2 mL of CCl4 was added at -20 °C a solution of
61.7 mg (400 µmol) of methyl p-tolyl sulfoxide (3a ) in 1 mL of
CCl4. The mixture was stirred at -20 °C until the hydroper-
oxide (S)-2a was consumed as demonstrated by a negative
peroxide test (KI). The solution was worked up, and the
enantiomeric excess of the sulfoxide 3a and the sulfoxide 3a :
sulfone 4a ratio were determined on the crude reaction
mixture as described above.
ligated to the titanium metal. For high enantioselectivi-
ties, the in situ generation of the sulfoxide by asymmetric
sulfide oxidation with R*OOH/Ti(OiPr)4 is more effective
than employing racemic sulfoxide.
Exp er im en ta l Section
Gen er a l P r oced u r e. For the sulfoxidation reactions, all
glassware was dried under vacuum (ca. 150 °C/0.1 Torr) and
all reactions were run under an argon gas atmosphere. CH2Cl2
and CCl4 were distilled under an argon gas atmosphere from
i
CaH2 and PrOH from magnesium; Ti(OiPr)4 was distilled prior
to use. The optically active hydroperoxides (S)-(-)-2a -c were
prepared by kinetic resolution of the racemic hydroperoxide
with horseradish peroxidase as previously described.15b Horse-
radish peroxidase was obtained from Boehringer, Mannheim.
The sulfides 1 were prepared according to literature proce-
dures.20
Enantiomeric excesses of the sulfoxides 3, the conversions
of the sulfides 1, and the sulfoxide 3:sulfone 4 ratios were
determined directly on the crude production mixture by HPLC
analysis on chiral columns (Daicel Chiralcel OD-H column, 35
× 0.46 cm), with UV detection at 220 nm, n-hexane/2-propanol
(9:1) as eluent and a flow rate of 0.5 mL/min (reactions with
p-Tol-S-Me, p-Tol-S-Et); n-hexane/2-propanol (99:1), a flow rate
of 0.5 mL/min (reactions with n-octyl-S-Me) or Daicel Chiralcel
OB-H, n-hexane/2-propanol (8:2), a flow rate of 0.6 mL/min
(reactions with p-Tol-S-i-Pr); n-hexane/ethanol (8:2), a flow rate
of 0.4 mL/min (reactions with n-Bn-S-Me) or of 0.5 mL/min
(reactions with Ph-S-Me, 2-naphthyl-S-Me) or of 0.6 mL/min
(reactions with p-nitrophenyl-S-Me); n-hexane/ethanol (95:5)
f n-hexane/2-propanol (8:2); a flow rate of 0.5 mL/min
(reactions with p-Tol-S-n-Bu); n-hexane/2-propanol (8:2) f (7:
3), a flow rate of 0.5 mL/min (reactions with p-anisyl-S-Me).
On the OD-H column the (R) enantiomer of sulfoxide 3
always eluted before the (S) one; on the OB-H column the (S)-
enantiomer eluted first.
For the conditions to determine the ee values of the isolated
sulfoxides 3, see ref 9b.
Absolute configurations of the sulfoxides 3 were assigned
by direct comparison of the specific rotation determined on a
polarimetric Chiralyser with literature values.5b,9b
Gen er a l P r oced u r e for th e Oxid a tion of Th ia n th r en e
5-Oxid e (SSO) by (S)-(-)-1-P h en yleth yl Hyd r op er oxid e
(2a )/Ti(OiP r )4. To a solution of 186 mg (800 µmol) of thian-
threne 5-oxide (SSO) in 8 mL of CCl4 or CH2Cl2 was added
11.8 µL (40.0 µmol) of Ti(OiPr)4. After the mixture was stirred
for 10 min at room temperature, a solution of 11.1 mg (80.0
µmol) of (S)-(-)-1-phenylethyl hydroperoxide (2a ) in 0.5 mL
of solvent was added. Stirring was continued at room tem-
perature until the entire hydroperoxide (S)-2a was consumed
(30 min), as confirmed by the negative peroxide test. After
addition of 40 µL of aqueous, saturated NH4F solution, the
mixture was worked up as described above. The product ratios
(Table 4) were determined by HPLC analysis of the crude
product mixture on a reversed phase column (Knauer Euro-
spher 100 C-18) with UV detection at 254 nm, methanol/water/
acetonitrile (64:34:2) as eluent, and a flow rate of 0.4 mL/min.
Gen er a l P r oced u r e of th e Oxid a tion of th e Su lfid e 1a
in CCl4 w ith th e Op tica lly Active Hyd r op er oxid es (S)-
(-)-2a -c on th e An a lytica l Sca le. To a solution of 5.9 µL
(20.0 µmol) of Ti(OiPr)4 in 1 mL of CCl4 was added a solution
of sulfide 1a (400 µmol) in 1 mL of CCl4. The mixture was
stirred for 10 min at room temperature (ca. 20 °C) and then
cooled to -20 °C. After 5 min, a solution of the hydroperoxide
[480 µmol (1.2 equiv) or 600 µmol (1.5 equiv)] in 1 mL of CCl4
was added, and the reaction mixture was stirred at -20 °C
for the time stated in Table 1. The reaction progress was
1
monitored by TLC or H NMR analysis directly on the reaction
mixture. After complete consumption of the hydroperoxide,
the catalyst was destroyed by the addition of 20 µL of aqueous,
saturated NH4F solution, and the mixture was stirred for
1-1.5 h at room temperature. After filtration of the suspended
material over Celite and thorough washing of the residue with
CH2Cl2 (5 × 2 mL), the solvent was evaporated (20 °C/20 Torr).
Ack n ow led gm en t . We thank the Deutsche For-
schungsgemeinschaft (SFB-347 “Selektive Reaktionen
Metallaktivierter Moleku¨le” and the Fond der Chemi-
schen Industrie for generous financial support.
(20) (a) Wardell, J . L.; McM. Wigzell, J . J . Chem. Soc., Dalton Trans.
1982, 2321-2326. (b) Zweig, A.; Maurer, A. H.; Roberts, B. G. J . Org.
Chem. 1967, 32, 1322-1329.
J O980243Y