Demonstration of a Common Concerted Mechanistic Pathway for the Acid-Catalyzed Cyclization of 5,6-Unsaturated Oxiranes in Chemical and Enzymatic Systems

Full text of DOI:10.1021/ja980096l

3526
J. Am. Chem. Soc. 1998, 120, 3526-3527
Demonstration of a Common Concerted Mechanistic
Pathway for the Acid-Catalyzed Cyclization of
shown by a study of the product ratio 7/(5 + 6) as a function of
acid concentration; no 7 can be detected when the cyclization is
conducted at 0.5 M ClCH
2
CO
2
H. Thus, the pseudo-first-order
5
,6-Unsaturated Oxiranes in Chemical and
Enzymatic Systems
E. J. Corey* and Donnette Daley Staas
Department of Chemistry and Chemical Biology
HarVard UniVersity
Cambridge, Massachusetts 02138
rate constant for the conversion of 4 to 8 is 0.22 × 10 s-1 at
-7
0
2 2
.1 M ClCH CO H and 25 °C. Under these conditions, the
substrate 1 reacts to form as the only products 2 and 9 in the
amounts and at the rates shown (dissected pseudo-first-order rate
constants). (The cyclized product 2 is a mixture of the three
possible olefinic isomers, as shown.) None of the pinacolic
rearrangement product 10 could be detected by GC analyses. Since
ReceiVed January 9, 1998
The acid catalyzed cyclization of chiral terminal epoxides of
polyprenoids is an extremely powerful synthetic construction
1
which is a key step in sterol and triterpene biosynthesis and also
an increasingly useful route for the enantioselective synthesis of
a wide variety of polycyclic terpenoids.2 Although it has recently
been demonstrated that in sterol biosynthesis the oxirane cleavage
and initial cyclization events are concerted,1 the relative timing
of the oxirane cleavage and cyclization steps in nonenzymatic
cyclization reactions of oxiranes has remained obscure. This
paper describes the study of acid-catalyzed monocyclization
processes such as 1 f 2 and provides compelling evidence that
such reactions proceed via a pathway in which oxirane C-O
cleavage and C-C bond formation are concerted rather than via
discrete carbocations such as 3.
a
the rate constant for cyclization of 1 (3.3 × 10 s-1) is 15 times
-7
the rate constant for formation of the acyclic carbocation 8 from
, it follows that the transformation 1 f 2 occurs by a concerted
4
pathway in which nucleophilic participation of the π-electrons
of the double bond accelerate oxirane C-O cleavage and not by
way of the discrete acyclic carbocation 3. The absence of the
pinacolic rearrangement product 10 in the acid-catalyzed cycliza-
tion of 1 also militates against the intermediacy of 3 in the
cyclization of 1 to 2.
The above conclusions were confirmed in a study of the
To evaluate the rate and nature of acid-catalyzed non-π-assisted
reactions of 1, the reaction of the saturated analogue 4 was
investigated by GC and H NMR analysis of kinetics and products.
cyclization of 11, the trimethylsilyl analogue of 1. In CDCl
3
(or
CHCl ) at 25 °C using 0.1 M ClCH CO H (10 equiv) as catalyst,
3
2
2
1
the unsaturated epoxide 11 was converted completely to mono-
3 3 2
In CDCl (or CHCl ) as solvent at 25 °C with ClCH COOH as
cyclic product 12 at a pseudo-first-order rate constant of 174 ×
3
catalyst (0.1 M, 10 equiv), the reaction of 4 followed pseudo
first-order kinetics and generated 5-7 as products in the indicated
amounts and with the indicated dissected pseudo first-order rate
constants for product formation. Kinetic measurements at several
-7 -1
10
s , indicating powerful nucleophilic assistance by the double
bond and a concerted oxirane cleavage-cyclization pathway. The
relative rates of cyclization and acyclic carbocation formation in
this case must be greater than 800. Examination of literature data
on the addition of fully formed carbocations to olefinic substrates
2
different concentrations of ClCH COOH showed very clearly that
the reaction is second order in this acid catalyst, implying that at
least 5 and 6 are probably formed by a process in which one
molecule activates the oxirane function by hydrogen bonding and
(
3) Chloroacetic acid was selected for the kinetic and product studies as a
1a,1b
mimic of D-456, the essential protic source in lanosterol synthase
and
also because it afforded a convenient rate of reaction at 25 °C. The rate of
reaction of substrate 1 using acetic acid as catalyst was in comparison very
another attacks to form 5 by S
elimination. The pinacolic rearrangement product 7 arises via
N
2 displacement or 6 by E2
slow at 25 °C, as is consistent with the known acidities (e.g., in H
pK values are 4.75 and 2.92 for CH CO H and ClCH CO
products of ClCH CO H-catalyzed reactions of oxiranes described in this paper
2
O at 25 °C
4
cation 8 in a reaction which is first order in ClCH
2
CO
2
H, as
a
3
2
2
2
H). Each of the
2
2
(1) (a) Corey, E. J.; Cheng, H.; Baker, C. H.; Matsuda, S. P. T.; Li, D.;
are primary products which remain unchanged under the reaction conditions.
Although it is conceivable that the pinacolic product 7 arises by a process in
which oxirane cleavage and hydrogen shift are concerted, such a mechanism
seems very improbable for stereoelectronic reasons. In addition, were it to
occur, the rate of tertiary cation formation would have to be even slower and
all arguments made herein would still apply.
Song, X. J. Am. Chem. Soc. 1997, 119, 1277. (b) Corey, E. J.; Cheng, H.;
Baker, C. H.; Matsuda, S. P. T.; Li, D.; Song, X. J. Am. Chem. Soc. 1997,
1
19, 1289. (c) For a recent review, see: Abe, I.; Rohmer, M.; Prestwich, G.
D. Chem. ReV. 1993, 93, 2189.
(2) (a) Corey, E. J.; Roberts, B. E. Tetrahedron Lett. 1997, 38, 8921. (b)
Corey, E. J.; Liu, K. J. Am. Chem. Soc. 1997, 119, 9929. (c) Corey, E. J.;
Luo, G.; Lin, L. S. J. Am. Chem. Soc. 1997, 119, 9927. (d) Corey, E. J.;
Wood, H. B., Jr. J. Am. Chem. Soc. 1996, 118, 11982. (e) Corey, E. J.; Lin,
S. J. Am. Chem. Soc. 1996, 118, 8765. (f) Corey, E. J.; Lee, J.; Liu, D. R.
Tetrahedron Lett. 1994, 35, 9149. (g) Corey, E. J.; Lee, J. J. Am. Chem. Soc.
(4) The finding that the kinetics for the disappearance of 4 show a second-
order dependence on [ClCH
2 2
CO H] is not inconsistent with a first-order
dependence on [ClCH CO H] for the formation of the pinacolic product 7
2
2
from 4, since the pathway 4 f 7 represents only 6% of the total reaction;
i.e., our experimental data do not allow a distinction between order 2 and
1
993, 115, 8873.
2 2
1.94 in [ClCH CO H].
S0002-7863(98)00096-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 03/31/1998

Communications to the Editor
J. Am. Chem. Soc., Vol. 120, No. 14, 1998 3527
Scheme 1
synthesized from 20; and substrate 17 was synthesized from 19
via 22-25.
In conclusion, the results reported herein demonstrate that in
the acid-catalyzed cyclization of 5,6-unsaturated oxiranes to form
a six-membered ring, the oxirane cleavage and cyclization events
are concerted. The cyclization reaction is accelerated by the
olefinic linkage in proportion to its π-nucleophilicity. Taken
together the experimental data define transition state 26 for the
cyclization of 1 in which there is partial positive charge at C(1)
and C(5) (considerably less at the latter than the former) and
partial cleavage of the O-C(1) bond and partial formation of the
C(1)-C(6) bond.7 Additional driving force results when the
3
indicates roughly 10 -fold Me
3
Si rate enhancement, considerably
greater than the value of 53 which is calculated from the above
5
data for concerted cyclization of 11 relative to 1.
We have also carried out product and kinetic studies with the
substrates 13, 15, and 17 which are converted to the corresponding
cyclization products 14, 16, and 18 with the indicated pseudo-
unsaturated oxirane is conformationally constrained to favor the
geometry required for concerted cyclization. Although our studies
were carried out with protic acid catalysts, it is most likely that
the concerted mechanistic pathway also dominates with Lewis
first-order rate constants (0.1 M ClCH
2 2 3
CO H in CDCl at 25 °C)
(Scheme 1). Products 14 and 16 were a mixture of three position
isomeric olefins resulting from the three possible modes of proton
loss from the intermediate bicyclic carbocation. The rates of
reaction of 13, 15, and 17 relative to formation of carbocation 8
from 4 provide a measure of the driving force for cyclization:
acid catalysts such as MeAlCl
optimum reagents (in CH Cl ) for synthetic applications.
2 2
and Me AlCl, which are the
2
,8
2
2
Indeed, one reason for the success of these acid-induced cycliza-
tions of unsaturated epoxides in synthesis is that the acid-
coordinated oxirane has sufficient lifetime to allow proper
positioning of the nucleophilic double bond and sufficient
38 for 13, 540 for 15, and 3700 for 17. The greater driving force
for 15 f 16 (540) than for 1 f 2 (15) is readily understood as
a consequence of a favorable conformational restriction of 15,
2
q
electrophilic reactivity to induce C-C bond formation. The
relative to 1, i.e., a less negative ∆S for 15 than for 1. The
concertedness of the reaction obviates the intermediacy of an
initiating carbocation and associated side reactions such as
pinacolic rearrangement and proton loss. In addition, the present
results make it understandable that, for enzymatic cyclizations
such as the conversion of 2,3-oxidosqualene to lanosterol, optimal
folding of the substrate by the enzyme should accelerate cycliza-
tion (as well as control stereochemistry) and permit a modestly
acidic group such as D456 of lanosterol synthase to serve as an
effective activator.1 Finally, it is now abundantly clear that
further advances in the development of cationic cyclization
processes for synthesis will depend on the control of the level of
activation in the initiation step and also of substrate conformation.
smaller driving force for 13 f 14 relative to 15 f 16 clearly
indicates that acid-catalyzed cleavage of the oxirane ring is more
difficult in the former case since that involves generation of a
partial positive charge at a primary rather than tertiary carbon.
In addition, although the trimethylsilyl group in 17 provides
additional driving force for cyclization (3700) relative to 15 (540),
the difference is only a factor of 7, far from what might have
been predicted.5 Although surprising at first glance, this small
a
acceleration may be due to substantial steric repulsion involving
the Me
early transition state with regard to C-C bond formation.
Parallel studies of substrates 1, 4, 13, and 15 using Cl CCO
as acid catalyst instead of ClCH CO H under standard conditions
CDCl , 25 °C) afforded comparable results, although the
individual rate constants were greater, as expected from the greater
acidity of Cl CCO H. The Cl CCO H induced cyclizations of
3
Si group during the cyclization of 17 as well as to an
6
3
2
H
2
2
Acknowledgment. This research was supported by the National
Institutes of Health and by a National Science Foundation Graduate
Fellowship and a Harvard Graduate Prize Fellowship to D.D.S. This paper
is dedicated to the memory of two great pioneers in this field of research,
Professors Saul Winstein and William S. Johnson.
(
3
3
2
3
2
4
, 13, and 15 were each concerted with oxirane C-O cleavage,
since rate accelerations due to nucleophilic participation of the
olefinic π-electrons were observed in each case.
Supporting Information Available: Experimental procedures, spec-
troscopic data for synthetic intermediates and kinetic data (40 pages, print/
PDF). See any current masthead page for ordering information and Web
access instructions.
As described in the Supporting Information, substrate 15 was
synthesized from (2S,5R)-(-)-isopulegone (19) via aldehyde 20
and diene 21 (by selective epoxidation of 21); substrate 13 was
JA980096L
(
5) Mayr, H.; Patz, M. Angew. Chem., Int. Ed. Engl. 1994, 33, 938.
6) Examination of HGS stereochemical models for the cyclization of 17
(
(7) For a recent theoretical study of transition states for cation-olefin
cyclization, see: Jenson, C.; Jorgensen, W. L. J. Am. Chem. Soc. 1997, 119,
10846.
(8) It is also relevant that these conditions also disfavor S
cleavage, since both the catalyst and medium are nonnucleophilic.
in the conformation allowing maximum delocalization of the C-SiMe
3
σ-electrons toward the developing carbocation in the transition state reveals
a major steric repulsion between the Me
preexisting ring.
3
Si group and a nearby CH
2
of the
N
2-type oxirane