Origin of diastereoselectivity in the addition of allylsilane to chiral acyclic mixed acetals

Article abstract of DOI:10.1039/a904608i

Addition of allylsilane to chiral acyclic mixed acetals has been found to proceed via an S(N)1 mechanism and a model has been proposed to explain the observed diastereoselectivity.

Full text of DOI:10.1039/a904608i

Origin of diastereoselectivity in the addition of allylsilane to chiral acyclic
mixed acetals
Kavita Manju and Sanjay Trehan*
Department of Chemistry, Panjab University, Chandigarh-160 014, India. E-mail: strehan@panjabuniv.chd.nic.in
Received (in Cambridge, UK) 9th June 1999, Accepted 12th August 1999
Addition of allylsilane to chiral acyclic mixed acetals has
been found to proceed via an SN1 mechanism and a model
has been proposed to explain the observed diastereo-
selectivity.
equiv. of TMSOTf and 1.5 equiv. of allyltrimethylsilane in
toluene at 278 °C to give homoallylic ether 8a/8b in 79% yield.
1
The diastereomeric ratio was found to be 82+18 from its H
NMR, which was vastly different from its precursor acetals.
Identical diastereoselectivity was also obtained when hydro-
cinnamaldehyde was treated with the trimethylsilyl ether of (R)-
1-phenethyl alcohol and allyltrimethylsilane in the presence of
TMSOTf in toluene at 278 °C. These two experiments suggests
an SN1 mechanism whether the mixed acetal is prepared in situ
or is isolated before allylation.
Over the last couple of decades great strides have been made in
the Lewis acid catalysed reaction of chiral acetals with silyl
nucleophiles. Chiral cyclic acetals have been studied in detail
and the mechanism as well as the diastereofacial selectivity of
their reaction is now well-established.1 Despite the fact that
chiral mixed acyclic acetals prepared in situ give homoallylic
ethers in high diastereoselectivity,2 an understanding of the
origin of the diastereoselectivity is lacking. In order to
comprehend the origin of the diastereoselectivity one needs to
first establish the mechanism (SN1 or SN2) because there is
support for both type of mechanisms in the literature.†3 If the
reaction proceeds via an SN2 mechanism, then the origin of the
diastereoselectivity should lie in the diastereoselective forma-
tion of a silyl acetal intermediate which would undergo
allylation in a stereospecific manner (Scheme 1). In such a case
the diastereomeric ratio of the silyl acetal (3a+3b) and
homoallylic ether (4a+4b) should be the same. Alternatively, if
the reaction proceeds via an SN1 mechanism, a common
oxocarbenium ion 5 will be formed from both the silyl acetal
intermediates. The diastereoselectivity in this case will be
determined by the extent of 1,3-induction4 and will be
independent of the ratio of the silyl acetals (3a+3b) (Scheme 1).
Houk has proposed a theoretical model for the nucleophilic
addition on this type of oxocarbenium ion by invoking a weak
1,3-allylic interaction and has shown reaction occurring from
the most stable conformation 5A giving 4b as the major product.5
Herein, we report our results which establish the mechanism
and the diastereoselectivity of this reaction.
The configuration of the newly created chiral centre was
established to be S from the optical rotation of the correspond-
ing homoallylic alcohol 9a/9b {[a]D 212.6 (c 1, CHCl3); lit.7
+16.9 (c 1, CHCl3) for the R isomer with 80% ee} obtained by
treatment of 8a/8b with iodotrimethylsilane.8 Thus the major
diastereomer formed is (S,R)-8a, just as reported in earlier
work,2 and not (R,R)-8b as as identified mistakenly by Houk as
the major product obtained by Mukaiyama2b and Seebach2c
while applying his model.5 Since the Houk model is not
applicable to the system under investigation an alternate model
is desirable. The reaction of the oxocarbenium ion intermediate
is expected to be exothermic in nature, therefore in the absence
of any polar group in the chiral centre, the ground state
conformation of the oxocarbenium ion should be important in
determining the diastereoselectivity.9 In order to establish the
importance of steric or stereoelectronic effects on the conforma-
tional preferences of the oxocarbenium ion intermediate leading
to the product, various silyl ethers with R groups of increasing
steric bulk were synthesised and used in the allylation
reaction.10 The results are summarised in Table 1. The
configuration of the products was assigned as done earlier
except when silyl ethers with isopropyl groups were used,
because they could not be prepared in reasonable optical purity.
In these cases the configuration was established by correlation
of their 1H NMR with that of 8.
Chiral ester 6 prepared from (R)-1-phenethyl alcohol and
hydrocinnamic acid was treated with 1 equiv. of DIBAL-H and
the intermediate was treated with TMSOTf and pyridine6 to
give acetal 7a/7b as a mixture of diastereomers in the ratio of
54+46 (Scheme 2).‡ This mixture of acetals was treated with 0.1
Scheme 1
Scheme 2
Chem. Commun., 1999, 1929–1930
This journal is © The Royal Society of Chemistry 1999
1929

Table 1 Reaction of hydrocinnamaldehyde with various silyl ethers and
allylsilaneab
We conclude that C–C hyperconjugative stabilisation of the
oxocarbenium ion restricts the conformation of the chiral centre
to 11a and 11b leading to major and minor diastereomers. The
observed diastereoselectivity is due to the differential inter-
action of the nucleophile with the hydrogen in 11a and the aryl
groups in 11b. Accordingly higher diastereoselectivity is
observed for the o-tolyl series.
We thank CSIR for financial support. K. M. also thanks
U. G. C. for senior research fellowship.
Entry
Ar
R
Diastereomer ratioc
Yield (%)
Notes and references
† Mukaiyama has proposed an SN2 mechanism for the allylation of chiral
mixed acyclic acetals [ref. 2(b)].
‡ No attempt was made to assign the configuration of the newly generated
acetal chiral centre in the major and minor isomer.
§ Since C(sp3)–C(sp2) is a stronger bond therefore C(sp3)–C(sp2)
hyperconjugation may not be as important as C(sp3)–C(sp3) hyper-
conjugation.
¶ The silyl ether of 1-deutero-1-phenylethyl alcohol gave product in an
86+14 ratio (compare Table 1, entry 1). The increase in diastereoselectivity
indicates that the C–H or C–D bond stabilises the oxocabenium ion by
inductive effects and not by hyperconjugation (ref. 15).
1
2
3
4
5
6
a
Ph
Ph
Ph
o-Tolyl
o-Tolyl
o-Tolyl
Me
Et
82+18 (4.5+1)
85+15 (5.7+1)
86+14 (6.1+1)d
88+12 (7.3+1)
93+07 (13.3+1)
91+09 (10.1+1)d
89
82
85
82
88
80
Pri
Me
Et
Pri
All reactions were carried out in toluene at 278 °C under nitrogen
atmosphere. b Enantiomeric excess of silyl ethers used in entries 1, 2, 4 and
5 were 94, 89, 90 and 81% respectively (ref. 11) and (±)-silyl ethers were
used in entries 3 and 6. c Diastereomeric ratio was determined via 1H NMR.
d Configuration was established by correlation of 1H NMR with that of 8.
∑ Linderman has also invoked hyperconjugative stabilisation of the
oxocarbenium ion intermediate by R3Si in diastereoselective Mukaiyama-
type aldol reactions of silyl-substituted mixed acyclic acetals (ref. 16). We
thank one of the referees for bringing this reference to our notice.
If steric effects were important, then one would expect a
decrease in diastereoselectivity with a decrease in the steric
difference between the medium and large groups. However, the
opposite trend was observed and generally diastereoselectivity
increased as the size of the alkyl group increased in the two
series studied (Table 1, entries 1–3 and 4–6). This suggests that
stereoelectronic effects stabilise certain select conformations of
the oxocarbenium ion leading to the product. In the absence of
any electrostatic effects,12 the important stereoelectronic inter-
actions between the three groups of the chiral centre and the
CNO bond which can restrict the conformational mobility of the
chiral centre are s*–p* and p–p* (hyperconjugation).13,14
However, conformation 10 obtained using the s*–p* inter-
1 S. E. Denmark and N. G. Almstead, J. Am. Chem. Soc., 1991, 113, 8089;
R. Silverman, C. Edington, J. D. Elliott and W. S. Johnson, J. Org.
Chem., 1987, 52, 180; K. Maruoka and H. Yamamoto, Angew. Chem.,
Int. Ed. Engl., 1985, 24, 668; P. A. Bartlett, W. S. Johnson and J. D.
Elliott, J. Am. Chem. Soc., 1983, 105, 2088; J. M. McNamara and Y.
Kishi, J. Am. Chem. Soc., 1982, 104, 7371.
2 (a) A. Mekhalfia and I. E. Marko´, Tetrahedron Lett., 1991, 32, 4779; (b)
T. Mukaiyama, M. Ohshima and N. Miyoshi, Chem. Lett., 1987, 1121;
(c) R. Imwinkelreid and D. Seebach, Angew. Chem., Int. Ed. Engl.,
1985, 24, 765.
3 D. Sames, Y. Liu, L. DeYoung and R. Polt, J. Org. Chem., 1995, 60,
2153; T. Sammakia and R. S. Smith, J. Am. Chem. Soc., 1994, 116,
7915; I. Mori, K. Ishihara, L. A. Flippin, K. Nozaki, H. Yamamoto, P. A.
Bartlett and C. H. Heathcock, J. Org. Chem. 1990, 55, 6107; S. E.
Denmark and T. M. Willson, J. Am. Chem. Soc., 1989, 111, 3475; I.
Mori, P. A. Bartlett and C. H. Heathcock, J. Am. Chem. Soc., 1987, 109,
7199.
4 R. W. Hoffmann, Chem. Rev., 1989, 89, 1841.
5 J. L. Broeker, R. W. Hoffmann and K. N. Houk, J. Am. Chem. Soc.,
1991, 113, 5006.
6 S. Kiyooka, M. Shirouchi and Y. Kaneko, Tetrahedron Lett., 1993, 34,
1491.
7 N. Minowa and T. Mukaiyama, Bull. Chem. Soc. Jpn., 1987, 60,
3697
8 M. E. Jung and M. A. Lyster, J. Org. Chem., 1977, 42, 3761.
9 G. Frenking, K. F. Ko¨hler and M. T. Reetz, Tetrahedron, 1991, 47,
9005.
10 E. P. Lodge and C. H. Heathcock, J. Am. Chem. Soc., 1987, 109,
3353.
11 E. J. Corey, R. K. Bakshi and S. Shibata, J. Am. Chem. Soc., 1987, 109,
5551; H. Takahashi, T. Kawakita, M. Ohno, M. Yoshioka and S.
Kobayashi, Tetrahedron, 1992, 48, 5691.
action10 is similar to Houk’s model5 and therefore does not
appear to be important. Therefore hyperconjugation between
the substituents of the chiral centre and the oxocarbenium ion
appears to be the predominant stereoelectronic interaction.
We suggest that the reaction proceeds through a conforma-
tion where the alkyl group occupies the anti position. In
conformations 11a and 11b besides normal a–b C–C hyper-
conjugative stabilisation of the oxocarbenium ion there is an
additional stabilisation due to bonding between the b–g
substituents, as depicted in 12.14c This stabilisation is greater for
a C–C bond than for a C–H bond.14c Therefore ethyl or
isopropyl groups should stabilise conformations 11a and 11b
more than a methyl group over other conformations. This
additional stabilisation is missing for bonding between the
chiral carbon and aryl groups or hydrogen.§¶∑
12 S. S. Wong and M. N Paddon-Row, J. Chem. Soc., Chem. Commun.,
1991, 327. Also see ref. 11.
13 N. T. Anh and O. Eisenstein, Nouv. J. Chim., 1977, 1, 62.
14 (a) S. Berger, B. W. K. Diehl and H. Ku¨nzer, Chem. Ber., 1987, 120,
1059; (b) W. J. Hehre and L. Salem, J. Chem. Soc., Chem. Commun.
1973, 754; (c) L. Radom, J. A. Pople and P. v. R. Schleyer, J. Am. Chem.
Soc., 1972, 94, 5935.
15 W. M. Schubert, R. B. Murphy and J. Robins, J. Org. Chem., 1970, 35,
951; V. J. Shiner Jr. and J. S. Humphrey Jr., J. Am. Chem. Soc., 1963,
85, 2416.
16 R. J. Linderman and T. V. Anklekar, J. Org. Chem., 1992, 57, 5078.
Communication 9/04608I
1930
Chem. Commun., 1999, 1929–1930