Direct chlorohydrin and acetoxy alcohol synthesis from olefins promoted by a Lewis acid, bis(trimethylsilyl) peroxide and (CH3)3SiX [16]

Full text of DOI:10.1021/ja993492s

J. Am. Chem. Soc. 2000, 122, 1245-1246
1245
Scheme 1. Direct Chlorohydrin and Acetoxy Alcohol
Direct Chlorohydrin and Acetoxy Alcohol Synthesis
from Olefins Promoted by a Lewis Acid,
Bis(trimethylsilyl) Peroxide and (CH3)3SiX
Synthesis from Olefins with BTSP, TMSX, and Lewis Acidsa
Isao Sakurada, Shingo Yamasaki, Richard Go¨ttlich,
Takehiko Iida, Motomu Kanai, and Masakatsu Shibasaki*
Graduate School of Pharmaceutical Sciences
The UniVersity of Tokyo, Hongo, Bunkyo-ku
Tokyo 113-0033, Japan
ReceiVed September 28, 1999
The use of bis(trimethylsilyl) peroxide (BTSP) offers a useful
method for selective Baeyer-Villiger oxidations in the presence
of C-C double bonds.1 The reaction proceeds efficiently upon
treatment with BTSP, catalytic amounts of SnCl4, and additives
such as (()-trans-1,2-(diamino)cyclohexane and (()-trans-N,N′-
bis(p-toluenesulfonyl)cyclohexane-1,2-diamine, giving the cor-
responding Baeyer-Villiger product in which the C-C double
bond remains unchanged. On the other hand, we have found that
olefins are directly functionalized to chlorohydrins and also
acetoxy alcohols in one pot by the combination of BTSP,
trimethylsilyl (TMS) X (X ) Cl, OAc), and a catalytic amount
of Lewis acid (Scheme 1). We describe these results, including a
possible reaction mechanism, as well as a preliminary application
to metal-catalyzed asymmetric direct chlorohydrin formation from
olefins, in this contribution.
The reaction using cyclohexene was examined in detail. It was
first found that the reaction of cyclohexene with SnCl4 (10 mol
%), BTSP (2 mol equiv), and TMSCl (2 mol equiv), in CH2Cl2
at -20 °C for 48 h, provided the chlorohydrin 1a in 69% yield
after acidic workup. On the other hand, in the absence of TMSCl,
almost quantitative starting material was recovered. The stereo-
chemistry of 1a was determined by comparison with an authentic
sample. The present new method for the direct synthesis of
chlorohydrins was successfully applied to a variety of olefins as
shown in Scheme 1.2,3 It is noteworthy that the chlorohydrins 6a
and 7a can be synthesized in a highly stereocontrolled manner.4
To get some insight into the mechanism of this novel
chlorohydrin synthesis, the reaction of SnCl4 with BTSP was first
examined. It was found that when 1 mol equiv of BTSP (13C: δ
) - 1.27, 29Si: δ ) 27.35) was reacted with SnCl4 in CH2Cl2 at
- 20 °C for 1 h, evolution of Cl2 gas was observed. This was
confirmed by the reaction of trans-5-decene to give the corre-
sponding erythro-dichloride in 5% yield together with the
chlorohydrin 5a (78%), although in other cases, no dichloride
could be isolated. Furthermore, the NMR spectrum showed the
quantitative formation of hexamethyldisiloxane (13C: δ ) 2.01,
29Si: δ ) 7.45). These results suggest that (Cl2SnO)n (8) is formed
as the active catalyst concomitantly with Cl2 and hexamethyl-
disiloxane by the reaction of SnCl4 and BTSP (Scheme 2).
Actually, (Cl2SnO‚1.5 pyridine complex)n was successfully
obtained by adding pyridine to the mixture, and the structure was
confirmed by comparison with an authentic sample (IR).5,6
a Reactions under conditions A were performed at -20 °C. Reactions
under conditions B and C were performed at ambient temperature. b4
mol equiv of BTSP was used.
Moreover, we were pleased to find that commercially available
(SnIIO)n could promote chlorohydrin formation (Scheme 1).3,7
Thus, the active species of the catalyst would be (Cl2SnO)n.
Second, the epoxide was found not to be the intermediate in this
reaction. During the reaction, none of the corresponding epoxide
was detected by TLC or NMR analysis. Furthermore, when the
mixture of cyclohexene oxide and cycloheptene was subjected
to the chlorohydrin-forming conditions (SnCl4 (10 mol %), BTSP
(1.2 mol equiv), and TMSCl (2 mol equiv)), trans-2-chlorocy-
cloheptanol was obtained in 84% yield with cyclohexene oxide
remained unchanged. Finally, from the 18O-labeling experiment,
it was found that the oxygen atom of the product chlorohydrin
was derived from the oxygen atom of (Cl2SnO)n species. We
(1) (a) Suzuki, M.; Takada, H.; Noyori, R. J. Org. Chem. 1982, 47, 902-
904. (b) Matsubara, S.; Takai, K.; Nozaki, H. Bull. Chem. Soc. Jpn. 1983,
56, 2029-2032. (c) Go¨ttlich, R.; Yamakoshi, K.; Sasai, H.; Shibasaki, M.
Synlett 1997, 971-973.
(2) For chlorohydrin synthesis from olefins, see: Larock, R. C. Compre-
hensiVe Organic Transformations; VCH: New York, 1989; 325-326.
(3) See Supporting Information for experimental details.
(4) The stereochemistry was unequivocally determined by converting to
(1R*,2S*,4R*,5S*)-1,2-bis(tert-butyldiphenylsilyloxymethyl)-4,5-epoxycyclo-
hexane, whose structure is known by X-ray analysis. Iida, T.; Yamamoto, N.;
Sasai, H.; Shibasaki, M. J. Am. Chem. Soc. 1997, 119, 4783-4784. For the
determination of the relative stereochemistry of other compounds, see
Supporting Information.
(6) For the discussion of oligomeric structure of oxotin complexes, see:
Davies, A. G.; Smith, P. J. In ComprehensiVe Organometallic Chemistry;
Wilkinson, G., Ed.; Pergamon Press: New York, 1982; Chapter 11, Vol. 2,
p 519.
(7) The use of (Bu2SnO)n and SnIVO2 gave less satisfactoy results in terms
of reaction rate.
(5) Dehnicke, V. K. Z. Anorg. Allg. Chem. 1961, 308, 72-78.
10.1021/ja993492s CCC: $19.00 © 2000 American Chemical Society
Published on Web 01/29/2000

1246 J. Am. Chem. Soc., Vol. 122, No. 6, 2000
Communications to the Editor
Scheme 2. Possible Mechanism of Chlorohydrin Formation
from Olefins
reaction, via an epoxide intermediate,13 which was confirmed by
1H NMR of the reaction mixture, as well as by the fact that the
epoxide reacted smoothly to give the acetoxy alcohol.3 Moreover,
in the case of trans-5-decene, the epoxide (5c) was the major
product, and only a trace amount of the acetoxy alcohol was
obtained. Judging from the stereochemistry of 6b and 7b,14 the
epoxide formation was found to take place from the opposite side
of the ester and ether groups. Importantly, of Zr(OiPr)4, BTSP,
and TMSOAc, each is essential for generating the active species.
Thus, in the absence of TMSOAc, even epoxidation did not take
place. Furthermore, the epoxide did not afford the acetoxy alcohol
in the absence of BTSP. Although the structure of the catalytic
species is not clear at this point, it appears that the actual catalytic
species may be an oxozirconium complex.15
These new reactions prompted us to investigate their ap-
plicability to catalytic asymmetric reactions.16,17 A preliminary
result using diisopropyl L-tartrate (L-DIPT) as a chiral ligand is
shown in Scheme 3.3,18-20
Scheme 3. Catalytic Asymmetric Synthesis of Chlorohydrins
from Olefins
prepared 18O-labeled (Cl2SnO)n by the reaction of SnCl4 and 18O-
labeled BTSP.3 The reaction using this pregenerated (Cl2Sn18O)n
(1 mol equiv), 16O-BTSP (2 mol equiv) and TMSCl (2 mol equiv)
afforded 18O-labeled product 6a in 44% yield. The incorporation
of 18O atom was detected by mass spectroscopy after conversion
to the corresponding epoxide. From this evidence, a possible
reaction mechanism for chlorohydrin formation is postulated in
Scheme 2. The first step may be an insertion of an olefin to a
Sn-O bond to give 9.8 Maybe this step is an equilibrium process,
in which starting 8 and the olefin are in a thermodynamically
more favored state.9 As a result, no reaction could proceed in the
absence of TMSCl. Reaction of 9 with TMSCl followed by the
attack of BTSP could give 11. Furthermore, the reaction of 11
with TMSCl possibly via the transition state 12 could give the
TMS-protected chlorohydrin and hexamethyldisiloxane with the
regeneration of the catalyst 8. This model also explains the
stereochemistry of 6a and 7a, in which the hydroxyl group exists
cis to the ester or ether group that can function as directing groups
for (Cl2SnO)n by coordination.10
Next, we extended this novel reaction to the formation of more
synthetically useful diol derivatives. Development of one-step
trans-diol derivative formation from olefin is quite challenging.11
We expected that if we used TMSOAc, instead of TMSCl, trans-
â-acetoxy alcohols could be obtained via the hypothetical
transition state 12 (X ) OAc). In fact, the SnCl4-catalyzed reaction
of TMSOAc afforded a 5 (61%):1 (12%) mixture of acetoxy
alcohol 1b and chlorohydrin 1a. Having obtained this exciting
result, we optimized the reaction conditions and found that Zr-
(OiPr)4 was the best Lewis acid for the acetoxy alcohol forma-
tion.12 Thus, in the presence of Zr(OiPr)4 (10 mol %), BTSP (2
mol equiv), and TMSOAc (2 mol equiv), trans-â-acetoxy alcohols
were obtained in good yields (Scheme 1).3 Interestingly, the Zr-
catalyzed reaction proceeds, contrasting with the Sn-catalyzed
In conclusion, we have developed a conceptually new method
for the direct synthesis of chlorohydrins and acetoxy alcohols from
olefins. These reactions are potentially extended to catalytic
asymmetric processes. Further studies for improvement of enan-
tioselectivities by designing new chiral ligands, are currently under
investigation.
Acknowledgment. This study was financially supported by CREST,
The Japan Science and Technology Corporation (JST). We thank Dr.
Higuchi in The University of Tokyo and Dr. Fujisawa in Tsukuba
University for showing us a procedure for the preparation of 18O-labeled
H2O2.
Supporting Information Available: Experimental procedures and
characterization of the products; details for 18O-labeling experiments.
NMR studies for Zr catalytic species (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
JA993492S
(13) Metal-catalyzed epoxidation of olefins by BTSP: Irie, R.; Hosoya,
N.; Katsuki, T. Synlett 1994, 255-256. Yudin, A. K.; Sharpless, K. B. J. Am.
Chem. Soc. 1997, 119, 11536-11537.
(14) 6b and 7b were found to be kinetic products: see Supporting
Information.
(15) Reaction of cyclohexene with (AcO)2ZrdO, BTSP, and TMSOAc gave
the epoxide (26%) and 1b (3%) for 48 h, and 1b (73%) and the epoxide (20%)
for 498 h.
(16) For a catalytic asymmetric synthesis of chlorohydrins from olefins,
see: (a) El-Qisairi, A.; Hamed, O.; Henry, P. M. J. Org. Chem. 1998, 63,
2790-2791. (b) Hamed, O.; Henry, P. M. Organometallics 1998, 17, 5184-
5189.
(17) For a catalytic asymmetric synthesis of halohydrins from epoxides,
see: (a) Denmark, S. E.; Barsanti, P. A.; Wong, K.-T.; Stavenger, R. A. J.
Org. Chem. 1998, 63, 2428-2429. (b) Nugent, W. A. J. Am. Chem. Soc.
1998, 120, 7139-7140.
(18) The absolute configuration of 1S,2S-1a was determined by the
comparison of the optical rotation; Fukazawa, T.; Shimoji, Y.; Hashimoto, T.
Tetrahedron Asymmetry 1996, 7, 1649-1658. The absolute configuration of
1S,2S-2a was determined by the Mosher method; Dale, J. A.; Mosher, H. S.
J. Am. Chem. Soc. 1973, 95, 512-519.
(8) For the reactivity of a Sn-O bond, see ref 6.
(9) 9 was not detected by NMR.
(10) One of the reviewers suggested an interesting alternative mechanism
as follows. Reaction of the unstable tin species containing a chlorine with
olefin gives the chloronium ion with an oxo(chloro)tin cluster as counterion.
Collapse of this ion pair proceeds with carbon-oxygen bond formation.
However, the above mechanism can be denied by the following experimental
results. Treatment of cyclohexene with SnCl4, BTSP, and TMSOAc gave 1b
(61%) and 1a (12%). Moreover, reaction of cyclohexene with SnCl4, BTSP,
and TMSN3 gave the trans-azidohydrin (48%) and 1a (17%).
(11) Wilson, C. V. In Organic Reactions; Adams, R., Ed.; John Willey &
Sons Inc. New York, 1957; Chapter 6, Vol. 9, p 350.
(19) The following results for 1a were obtained using other tartaric acid
derivatives. dimethyl L-tartrate (64% yield, 45% ee); diethyl L-tartrate (58%
yield, 51% ee); di-tert-butyl L-tartrate (24% yield, 18% ee); N,N,N′,N′-
tetramethyl-L-tartaramide (12% yield, 2% ee).
(20) Addition of a catalytic amount (30 mol %) of H2O, thus generating
HCl in situ, afforded a positive effect on enantioselectivity. The ee values of
1a and 2a increased to 66% from 56% and 63% from 58%, respectively, in
the presence of H2O (both in 70% yields).
(12) The use of other Lewis acids such as Zr(OtBu)4, Ti(OiPr)4, and (SnIIO)n
gave less satisfactory results.