Mechanism of Chiral Lewis Acid Mediated Enantiotopic Group-Selective Ring Cleavage of Cyclic Acetals Derived from meso-1,2-Diols

Article abstract of DOI:10.1021/jo991071n

Diastereoselectivity and enantioselectivity of chiral oxazaborolidine-mediated ring-cleavage reactions of meso-2,4,5-trisubstituted 1,3-dioxolane acetals with a trimethylsilyl ketene acetal were investigated in detail and discussed in terms of a mechanism involving a contact ion pair as a product-forming intermediate. Both diastereomeric 2-phenyl derivatives syn- and anti-11a gave the same ring-cleavage product 13a. However, the reaction of 2-phenylethynyl derivatives syn- and anti-11b proceeded almost stereospecifically, giving rise to 13b and 14b, respectively. In all of the reactions, isomerization of diastereomeric acetals was not observed. On the basis of these results, it was deduced that the dissociation of a Lewis acid-acetal complex is the rate-determining step, and the resulting ion pair intermediate undergoes either isomerization to a diastereomeric acetal or attack by a nucleophile depending on the structure of the acetal. The possible enantioselection at the product-forming step was ruled out. It was proposed that the observed enantioselectivity is determined by enantiodifferentiating coordination of the acetal oxygen atom by the chiral Lewis acid.

Full text of DOI:10.1021/jo991071n

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
7594 J . Org. Chem. 1999, 64, 7594-7600
Mech a n ism of Ch ir a l Lew is Acid Med ia ted En a n tiotop ic
Gr ou p -Selective Rin g Clea va ge of Cyclic Aceta ls Der ived fr om
m eso-1,2-Diols
Toshiro Harada,* Tomohito Nakamura, Motoharu Kinugasa, and Akira Oku
Department of Chemistry, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku, Kyoto 606-8585, J apan
Received J uly 6, 1999
Diastereoselectivity and enantioselectivity of chiral oxazaborolidine-mediated ring-cleavage reactions
of meso-2,4,5-trisubstituted 1,3-dioxolane acetals with a trimethylsilyl ketene acetal were inves-
tigated in detail and discussed in terms of a mechanism involving a contact ion pair as a product-
forming intermediate. Both diastereomeric 2-phenyl derivatives syn- and anti-11a gave the same
ring-cleavage product 13a . However, the reaction of 2-phenylethynyl derivatives syn- and anti-
11b proceeded almost stereospecifically, giving rise to 13b and 14b , respectively. In all of the
reactions, isomerization of diastereomeric acetals was not observed. On the basis of these results,
it was deduced that the dissociation of a Lewis acid-acetal complex is the rate-determining step,
and the resulting ion pair intermediate undergoes either isomerization to a diastereomeric acetal
or attack by a nucleophile depending on the structure of the acetal. The possible enantioselection
at the product-forming step was ruled out. It was proposed that the observed enantioselectivity is
determined by enantiodifferentiating coordination of the acetal oxygen atom by the chiral Lewis
acid.
In tr od u ction
The Lewis acid mediated cleavage of cyclic acetals has
been intensively investigated in recent years and has
proven to be a versatile tool in asymmetric synthesis.
Titanium complex mediated reaction of chiral acetals 1,
derived from (2R,4R)-2,4-pentanediol, proceeds with a
highly stereoselective introduction of a variety of carbon
nucleophiles such as allyltrimethylsilane, affording a
powerful method for asymmetric carbon-carbon bond
formation (eq 1).1 The ring-cleavage reaction is also
Recent work in these laboratories revealed that the
enantioselective ring cleavage of cyclic acetals can be
achieved without the aid of chiral auxiliaries by using
an appropriate chiral Lewis acid.3-5 In the presence of a
catalytic amount (10 mol %) of chiral oxazaborolidine 4,
the reaction of dioxolane acetals 3 with enol silyl ethers
and silyl ketene acetals proceeds with high enantiose-
lectivity (78-93% ee) to give the ring-cleavage products,
which can be transformed to enantiomerically enriched
secondary alcohols (eq 3).3 A variety of meso-1,2-diols can
utilized in desymmetrization of prochiral diols. L-Men-
thone-derived spiroacetals 2 undergo exclusive equatorial
C-O bond cleavage, leading to the enantiomerically pure
(>95% ee) desymmetrized derivatives of the parent diols
(eq 2).
(1) (a) McNamara, J . M.; Kishi, Y. J . Am. Chem. Soc. 1982, 104,
7371. (b) Sekizaki, H.; J ung, M.; McNamara, J . M.; Kishi, Y. J . Am.
Chem. Soc. 1982, 104, 7372. (c) Bartlett, P. A.; J ohnson, W. S.; Elliott,
J . D. J . Am. Chem. Soc. 1983, 105, 2088. Review: (d) Alexakis, A.;
Mangency, P. Tetrahedron Asymmetry 1990, 1, 477. (e) Seebach, D.;
Imwinkelreid, R.; Weber, T. In Modern Synthetic Methods; Scheffold,
R., Ed.; Springer-Verlag: Berlin, 1986; Vol. 4, p 125. For leading
references, see also: (f) Denmark, S. E.; Almstead, N. G. J . Am. Chem.
Soc. 1991, 113, 8089. (g) Ishihara, K.; Hanaki, N.; Yamamoto, H. J .
Am. Chem. Soc. 1993, 115, 10695.
(2) (a) Harada, T.; Hayashiya, T.; Wada, I.; Iwa-ake, N.; Oku, A. J .
Am. Chem. Soc. 1987, 109, 527. (b) Harada, T.; Sakamoto, K.; Ikemura,
Y.; Oku, A. Tetrahedron Lett. 1988, 29, 3097. (c) Harada, T.; Wada, I.;
Oku, A. J . Org. Chem. 1989, 54, 2599. (d) Harada, T.; Ikemura, Y.;
Nakajima, H.; Oku, A. Chem. Lett. 1990, 1441. Harada, T.; Nakajima,
H.; Ohnishi, T.; Takeuchi, M.; Oku, A. J . Org. Chem. 1992, 57, 720.
(e) Harada, T.; Shintani, T.; Oku, A. J . Am. Chem. Soc. 1995, 117,
12346. Review: (f) Harada, T.; Oku, A. Synlett 1994, 95.
be successfully desymmetrized via ring-cleavage reaction
(3) Kinugasa, M.; Harada, T.; Oku, A. J . Org. Chem. 1996, 61, 6772.
(4) (a) Kinugasa, M.; Harada, T.; Oku, A. J . Am. Chem. Soc. 1997,
119, 9067. (b) Kinugasa, M.; Harada, T.; Oku, A. Tetrahedron Lett.
1998, 39, 4523.
(5) (a) Harada, T.; Egusa, T.; Kinugasa, M.; Oku, A. Tetrahedron
Lett. 1998, 39, 5531. (b) Harada, T.; Egusa, T.; Oku, A. Tetrahedron
Lett. 1998, 39, 5535. (c) Harada, T.; Nakamura, T.; Kinugasa, M.; Oku,
A. Tetrahedron Lett. 1999, 40, 503.
10.1021/jo991071n CCC: $18.00 © 1999 American Chemical Society
Published on Web 09/14/1999

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Mechanism of Ring Cleavage of Cyclic Acetals
J . Org. Chem., Vol. 64, No. 20, 1999 7595
of 2-phenylethynyl acetals syn-5 (eq 4).4 The reaction
Sch em e 1
proceeds with high enantioselectivity (85-96% ee) at -78
°C by using a stoichiometric amount of chiral oxazaboro-
lidine 6.
The ring cleavage of cyclic acetals possesses a unique
character in the mechanism and origin of stereoselectiv-
ity. Because acetals can be regarded as difunctional
compounds with two ethereal C-O groups, the cleavage
reactions are classified as group-selective reactions,6,2f
where Lewis acids play a role different from that in the
more common face-selective reactions. The diastereotopic
C-O bonds of spiroacetal 2 are differentiated through
the ring-cleavage reaction (eq 2), whereas the discrimina-
tion of enantiotopic C-O bonds in syn-5 leads to the
stereoselective product formation (eq 4). Group-selective
C-O bond cleavage as well as the stereochemical course
(namely, inversion or retention of the acetal carbon) of
the nucleophile introduction determines the stereoselec-
tivities in eqs 1 and 3.
Recently, the mechanism of the ring-cleavage reaction
has been studied extensively to clarify the origin of the
high stereoselectivity observed for chiral acetals 1.1f,7
These studies revealed a wide mechanistic spectrum of
the reaction. Heathcock et al.7b and Denmark et al.1f
reported that the reaction of diastereomeric acetals syn-
and anti-7 was not stereospecific. The stereoselectivity
was demonstrated to be influenced by the structure of
the parent aldehyde residue (R), the nature of the Lewis
acid, the solvent, and the nucleophile. The observations
were incompatible with a mechanism involving direct
displacement of a Lewis acid-acetal complex. Denmark
et al. proposed a unified mechanistic scheme involving
three types of ion pairs with different degrees of dissocia-
tion, each with a different stereochemical profile (Scheme
1).1f The three species are (1) a contact ion pair (8), (2)
an external ion pair (9), and (3) a free oxocarbenium ion
(10). The contact ion pair undergoes highly selective
reactions in an invertive manner. The external ion pair
reacts with modest selectivity due to greater access at
the acetal center. The separated ion pair reacts with no
selectivity due to the minimal influence of the remote
stereogenic center. The degree of dissociation is enhanced
by stronger Lewis acids, by sterically demanding or
cation-stabilizing substituents, by polar solvents, and by
less reactive nucleophiles, and the enhanced dissociation
results in low selectivities.
To extend the scope of chiral Lewis acid mediated
enantioselective ring-cleavage reactions, it is essential to
understand the mechanism and origin of the selectivity.
For this purpose, we investigated the stereochemical
aspects of the chiral oxazaborolidine-mediated ring cleav-
age of dioxolane acetals syn- and anti-11a ,b derived from
meso-2,3-butanediols. The study demonstrated that dis-
sociation of a Lewis acid-acetal complex is the rate-
determining step, and the resulting contact ion pair
intermediate undergoes either rapid isomerization to a
diastereotopic ion pair or attack by a nucleophile. The
origin of enantioselectivity is discussed in terms of the
above mechanistic scheme.
Resu lts
In the course of preliminary study on the effect of an
aldehyde-derived moiety in the ring cleavage of dioxolane
acetals, we examined the reaction of phenyl derivatives
syn- and anti-11a .4a Treatment of syn-11a with silyl
ketene acetal 12 in the presence of oxazaborolidine 6 (0.3
equiv) at -20 °C for 15 h gave ring-cleavage product 13a
in 72% yield with 22% ee (eq 5). Although the enantio-
(6) For nonenzymatic enantiotopic group-selective reactions, see: (a)
Ward, R. S. Chem. Soc. Rev. 1990, 19, 1. (b) Gais, H.-J . Methods of
Organic Chemistry (Houben-Weyl); Helmchen, G., Hoffmann, R. W.,
Mulzer, J ., Schumann, E., Eds.; Georg Thieme Verlag: Stuttgart, 1995;
Vol. E21a, p 589. (c) Vedejs, E.; Daugulis, O.; Diver, S. T. J . Org. Chem.
1996, 61, 430 and references therein.
(7) (a) Yamamoto, Y.; Nishi, S.; Yamada, J . J . Am. Chem. Soc. 1986,
108, 7116. (b) Mori, I.; Ishihara, K.; Flippin, L. A.; Nozaki, K.;
Yamamoto, H.; Bartlett, P. A.; Heathcock, C. H. J . Org. Chem. 1990,
55, 6107. (c) Denmark, S. E.; Almstead, N. G. J . Org. Chem. 1991, 56,
6458. (d) Sammakia, T.; Smith, R. S. J . Org. Chem. 1992, 57, 2997. (e)
Sammakia, T.; Smith, R. S. J . Am. Chem. Soc. 1992, 114, 10998. (f)
Sammakia, T.; Smith, R. S. J . Am. Chem. Soc. 1994, 116, 7915.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
7596 J . Org. Chem., Vol. 64, No. 20, 1999
Harada et al.
Ta ble 1. Rin g Clea va ge of Aceta ls syn - a n d a n ti-11a ,b
Ta ble 2. Isom er iza tion of Aceta ls syn - a n d a n ti-11a ,b
w ith Oxa za bor olid in e 6 a n d Silyl Keten e Aceta l 12a
w ith Oxa za bor olid in e 6a
ring-cleavage product
recovery of acetal
yield
entry
acetal
6 (equiv)
time (h)
syn:antib
time yield
1
2
3
4
5
6c
7
8
syn-11a
0.3
1.2
0.3
1.2
1.2
1.2
0.3
1.2
1
8
1
8
8
72
8
8
3.0:1
3.0:1
3.2:1
3.2:1
1:1.4
1:1.6
1:2.2
1:1.5
entry
acetal
(h)
(%)
13:14b ee (%)c
(%)
syn:antib
1
2
3
4
5
6
7
8
9
syn-11a
8
0.25
8
1
0.25
4
94
82
96
76
53
82
34
59
43
46
9
>50:1
>50:1
>50:1
>50:1
>50:1
>50:1
>50:1
33
27
52
53
49
70
68
anti-11a
syn-11b
anti-11b
8
5
21
30
>50:1
1:>50
1:>50
1:>50
anti-11a
syn-11b
0.25
43
41
51
49
75
>50:1
1:>50
1:>50
1:>50
1:>50
a
anti-11b 20
1:16 96 (71)d
1:16 96 (73)d
1:10 94 (65)d
1:8.4 90 (60)d
All reactions were carried out in CH2Cl2 (0.4 M) at -20 °C.
Unless otherwise noted, the starting acetal was recovered in >95%
4
20
20
10e
11e,f
b
yield. Determined by 500 MHz 1H NMR analysis. c The acetal
was recovered in 60% yield.
a
Unless otherwise noted, reactions were carried out by using
11a or from anti-11a , the equilibration was reached
within 1 h even with 0.3 equiv of 6.
6 (1.2 equiv) and 12 (3 equiv) in CH2Cl2 (0.4 M) at -20 °C.
b
Determined by 500 MHz 1H NMR analysis. c Determined by 500
d
MHz 1H NMR analysis of the MTPA ester derivative. Ee (%) of
A similar set of experiments was carried out for
phenylethynyl derivatives syn- and anti-11b at -20 °C
(eqs 6 and 7). The reaction of syn-11b also proceeded with
high diastereoselectivity to give diastereomer 13b as the
sole product (Table 1, entries 6 and 7). In comparison
with the phenyl derivative, enantioselectivity was higher
(ca. 70% ee)8 while the reaction was slower (entry 2 vs
7). Again, no isomerization was observed for the recov-
ered acetal. In sharp contrast to the phenyl derivative,
anti-11b afforded another diastereomer 14b together
with a minor formation of 13b (14b:13b ) 16:1) (entries
8 and 9). Interestingly, the enantioselectivity of major
isomer 14b was high (96% ee) while that of minor isomer
13b was moderate (71% ee). When the reaction was
carried out with a reduced amount (1.2 equiv) of nucleo-
phile 12, the diastereoselectivity was lowered to 10:1
(entry 10). Further lowering (8.4:1) was observed under
dilute conditions using twice as much solvent (entry 11).
Even under the standard conditions, the ring-cleavage
reaction of anti-11b was slow and did not attain comple-
tion after 20 h (entry 8).9 Isomerization of the recovered
acetal was not observed at all. Control experiments
showed that, in the absence of the nucleophile, syn- and
anti-11b underwent isomerization to give an equilibrat-
ing mixture (syn-11b:anti-11b ) ca. 2:3) in 8 h (Table 2,
entries 5-8). J udging from the incomplete equilibration
in the experiment using 0.3 equiv of 6 (entry 7), the rate
of isomerization for syn- and anti-11b is slower than that
for the phenyl derivatives.
minor diastereomer 13b. e A 1.2 equiv portion of 12 was used. f The
reaction was carried out at 0.2 M.
selectivity was moderate, the reaction exhibited high
diastereoselectivity. The formation of diastereomeric
product 14a was not detected. A mechanistically infor-
mative result was obtained in the reaction of diastereo-
meric acetal anti-11a . The acetal was considerably less
reactive under similar conditions, affording exclusively
the same ring-cleavage product 13a (25% ee) in 10%
yield. No isomerization was observed for the recovered
acetal. The observed absence of stereospecificity sug-
gested the intervention of ion pairs in the oxazaboroli-
dine-mediated ring-cleavage reaction. To gain further
insight into the mechanism, we examined the reactions
of 2-phenyl and 2-phenylethynyl derivatives syn- and
anti-11a ,b with nucleophile 12 (3 equiv) by using a
stoichiometric amount (1.2 equiv) of 6 at -20 °C (eqs 5-7,
Table 1).
The absolute structures of ring-cleavage products
13a ,b were determined by correlation to (S)-MTPA esters
17a ,b (Scheme 2). TiCl4-mediated ring cleavage of chiral
acetal (4R,5R)-15a ,b stereoselectively gave ring-cleavage
products 16a ,b.1a-c Esterification of 16a ,b with (S)-MTPA
under the Mitsunobu conditions afforded esters 17a ,b
with stereochemical inversion.10 Ester 17a thus prepared
was identical in all respects with the major diastereomer
formed by the reaction of syn- and anti-11a with (R)-
MTPACl. Ester 17b was identical with the major isomer
obtained by hydrogenation of the (S)-MTPA ester of 13b.
Ring cleavage of 2-phenylethynyl derivative (4R,5R)-15c
mediated by TiCl4 proceeded nonselectively to give a 1.3:1
mixture of diastereomeric products 16c and 18 in 76%
yield. (Scheme 3). Esterification of the mixture with (S)-
MTPA under the Mitsunobu conditions afforded a 1.3:1
mixture of 17c and 19. Esters 17c and 19 thus obtained
The reaction of syn-11a for 8 h afforded 13a of 33% ee
exclusively in 94% yield (entry 1). When the reaction was
stopped after 0.25 h, the starting acetal was recovered
in 8% yield without isomerization to anti-11a (entry 2).
Under these stoichiometric conditions, the reaction of
anti-11a also afforded 13a exclusively (entries 3-5). The
anti acetal was less reactive but exhibited uniformly
higher ee (ca. 50%) than the syn isomer. Isomerization
of the recovered acetals was not observed even in the
higher conversion. In the absence of nucleophile 12, both
syn-11a and anti-11a underwent rapid isomerization
leading to an equilibrating mixture (syn-11a :anti-11a )
ca. 3:1) (Table 2, entries 1-4). Starting either from syn-
(8) The reaction at -78 °C gave higher ee of 96%.4a
(9) No ring cleavage was observed for anti-13b at -78 °C.
(10) Mitsunobu, O. Synthesis 1981, 1.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Mechanism of Ring Cleavage of Cyclic Acetals
Sch em e 2
J . Org. Chem., Vol. 64, No. 20, 1999 7597
(k2 . k5[12]). In these reactions, syn-11a was not detected
in the recovered acetal. This strongly suggests that the
ion pair syn-21a thus formed is exclusively captured by
the nucleophile to give 13a before it undergoes ring
closure to complex syn-20a (k4[12] . k-1). Ring cleavage
of syn-11a proceeds also through ion pair syn-21a .
Therefore, it is deduced that the dissociation of complexes
syn-20a to form contact ion pairs syn-21a is the rate-
determining step of the reaction of syn-11a .11
The results obtained for phenylethynyl derivatives syn-
and anti-11b can also be discussed according to the
proposed pathway (Scheme 4). In contrast to the phenyl
derivatives, they underwent ring cleavage almost ste-
reospecifically: the syn isomer gave ring-cleavage product
13b exclusively, whereas the anti isomer afforded iso-
meric product 14b with high stereoselectivity (16:1). The
observation can be rationalized if we assume that the
isomerization of ion pair anti-21b to ion pair syn-21b is
slower than its reaction with the nucleophile (k2 < k5-
[12]). Isomerization between syn- and anti-11b was
observed in the absence of nucleophile 12, but it was
slower than that of the phenyl derivatives. The rate of
isomerization might be dependent on the difference in
relative energy between anti-21 and syn-21. In compari-
son with the phenyl group, a sterically less demanding
phenylethynyl group may reduce the energy difference
leading to the slower isomerization of anti-21b.12 Accord-
ing to the pathway, the difference between k2 and k5[12]
will be decreased by carrying out the reaction at lower
concentration of 12, resulting in the increase of the
formation of minor product 13b. Indeed, a 10:1 mixture
of 14b and 13b was obtained in the reaction using 1.2
equiv of 12 (Table 1, entry 10). The diastereoselectivity
was further lowered (8.4:1) when the above reaction was
carried out by using twice as much solvent (entry 11).
No isomerization of recovered anti-11b was observed
in its reaction even under the conditions where the
isomerization of the initially produced ion pair anti-21b
competes with its capture by the nucleophile. J udging
from the observation, the ring-closure of ion pair syn-
21b to give syn-11b is again slow in comparison with its
capture by the nucleophile to give a minor diastereomer
13b. Therefore, dissociation of Lewis acid-acetal complex
syn-20 to form contact ion pair syn-21 is deduced to be
the common rate-determining step for syn-11a and anti-
11b. For the reactions of anti-11a ,b, we can assume
similar rate-determining dissociation of complex anti-20
because it is less probable that anti-20 undergoes ring-
closure much faster than isomeric syn-20. The absence
of significant rate depression in the reaction of anti-11b
with a reduced amount of the nucleophile (entry 10)
provides support for the assumption.
Sch em e 3
were identical to the major (S)-MTPA esters derived from
13b and 14b, respectively.
Discu ssion
Facile isomerization between syn-11a ,b and anti-11a ,b
observed in the absence of nucleophile 12 implies the
involvement of intermediates with the dissociated acetal
carbon-oxygen bond in the oxazaborolidine-mediated
enantioselective ring cleavage of dioxolane acetals. J udg-
ing from the high diastereoselectivity observed, external
ion pairs and free oxocarbenium ions are less likely for
the intermediates. We propose a mechanism involving
contact ion pairs as the product-forming intermediates
(Scheme 4).
Phenyl derivative syn-11a gave ring-cleavage product
13a (and ent-13a ) with complete diastereoselectivity.
According to Scheme 4, major enantiomer 13a is pro-
duced through a pathway involving initial formation of
the complex syn-20a in which the pro-R oxygen atom is
coordinated to oxazaborolidine 6 (designated as BL*3),
followed by dissociation to contact ion pair syn-21a and
subsequent attack of the nucleophile in an invertive
manner. Similarly, minor enantiomer ent-13a is formed
via complex syn-20′a , in which the pro-S oxygen atom is
coordinated, and via ion pair syn-21′a . Diastereomeric
dioxolane anti-11a also gave 13a (and ent-13a ) exclu-
sively. The formation of 13a indicates the isomerization
of the initially formed ion pair anti-21a to syn-21a .
Isomerization between the ion pair intermediates has
been discussed previously for six-membered cyclic
acetals.1f,7b-d The absence of diastereomeric ring-cleavage
product 14a indicates that the isomerization of anti-21a
proceeds much faster than its capture by the nucleophile
Ion pairs syn- and anti-21 are also involved in the
isomerization of syn- and anti-11a ,b in the absence of
nucleophile 12. Rate-determining dissociation to the ion
pairs requires equal rates for ring-cleavage and isomer-
ization reactions. Isomerization of anti- and anti-11a ,b,
however, proceeded faster than the corresponding ring-
cleavage reaction. The apparent discrepancy might be
derived from the catalytic nature of isomerization that
ensures constant concentration of oxazaborolidine 6
(11) In the above discussion, the reaction pathway leading to 13a
(i.e., the right half of Scheme 4) was considered. A similar discussion
can be made for the left half of the scheme leading to ent-13a .
(12) Alternatively, relative slow isomerization might be due to the
lower degree of dissociation for these ion pairs in comparison to that
for the phenyl derivatives.