Theory-guided discovery of unique chemical transformations of cyclopropenes

Article abstract of DOI:10.1021/ol052752+

(Chemical Equation Presented) Chiral 2-cyclopropenyl-4-tolyl sulfones, available by the [2 + 1]-cycloaddition of tosyldiazomethane to acetylenes under catalysis by the Rh(II) complex 1, provide a number of unusual transformations and useful chiral product

Full text of DOI:10.1021/ol052752+

ORGANIC
LETTERS
2006
Vol. 8, No. 1
171-174
Theory-Guided Discovery of Unique
Chemical Transformations of
Cyclopropenes
Robin A. Weatherhead-Kloster and E. J. Corey*
Department of Chemistry and Chemical Biology, HarVard UniVersity,
12 Oxford Street, Cambridge, Massachusetts 02138
corey@chemistry.harVard.edu
Received November 14, 2005
ABSTRACT
Chiral 2-cyclopropenyl-4-tolyl sulfones, available by the [2
+ 1]-cycloaddition of tosyldiazomethane to acetylenes under catalysis by the Rh(II)
complex 1, provide a number of unusual transformations and useful chiral products. This chemistry is dominated by the unusual strain and
reactivity of the cyclopropene ring system.
We recently found that a mixed Rh2(II) acetate complex
containing three bridging N-triflyl (diphenylimidazolidinone)
Scheme 1. Catalyzed Enantioselective Addition of Ethyl
(DPTI) groups, Rh2(OAc)(DPTI)3 (1), is an extremely
effective catalyst for the enantioselective generation of chiral
cyclopropenes from ethyl diazoacetate and various terminal
acetylenes, as illustrated in Scheme 1.1,2 Syntheses of such
chiral cyclopropenes have also been reported by other groups
using different Rh2(II) complexes as catalysts.3 These chiral
cyclopropenes are of great interest since they serve as
versatile intermediates that can be elaborated to a wide
variety of more complex chiral cyclopropanes, especially
those having stereocenters at all three ring vertices.1 This
work was motivated not only by these considerations but
also by the major advances in recent years in our understand-
ing of strain effects in unsaturated cyclopropanes and of
variations in individual bond strength.4 The guidance from
these new insights and the ready availability of a wide range
of chiral cyclopropenes provide an excellent opportunity to
Diazoacetate to Terminal Acetylenes
discover new and unique aspects of the chemistry of
cyclopropenes that can then further enrich mainstream
synthetic methodology.5
Studies with Cyclopropenyl 4-Tolyl Sulfones. We took
advantage of the known 4-toluenesulfonyl (tosyl, Ts) diaz-
omethane6 and the highly selective catalyst 1 to synthesize
(1) Lou, Y.; Horikawa, M.; Kloster, R. A.; Hawryluk, N. A.; Corey, E.
J. J. Am. Chem. Soc. 2004, 126, 8916-8918.
(2) See also: Lou, Y.; Remarchuk, T. P.; Corey, E. J. J. Am. Chem.
Soc. 2005, 127, 14223-14230.
(3) (a) Doyle, M.; Protopopova, M.; Mu¨ller, P.; Ene, D.; Shapiro, E. J.
Am. Chem. Soc. 1994, 116, 8492-8498. (b) Protopopova, M.; Doyle, M.;
Mu¨ller, P.; Ene, D. J. Am. Chem. Soc. 1992, 114, 2755-2757. (c) Mu¨ller,
P.; Imoga¨ı, H. Tetrahedron: Asymmetry 1998, 9, 4419-4428.
(4) For an outstanding exposition of strain energies in unsaturated
cyclopropanes and their origins, see: Bach, R. D.; Dmitrenko, O. J. Am.
Chem. Soc. 2004, 126, 4444-4452.
10.1021/ol052752+ CCC: $33.50
© 2006 American Chemical Society
Published on Web 12/09/2005

tosylcyclopropenes 2a-2c from 1-heptyne, propargyl bro-
mide, and tert-butylacetylene, respectively.7 Most of our
Technologies, Inc.) with i-PrOH-hexane as elution solvent
at 23 °C and were found to be cleanly first order. From
measurements of the first-order rate constant over the
temperature range 30-70 °C and an Arrhenius plot of ln k1
vs 1/T (strictly linear) the reaction activation parameters were
found to be: Ea ) 15.9 kcal/mol; ∆Hq ) 15.3 kcal/mol;
and ∆Sq ) -32.5 eu.7 The Eyring plot of ln(k/T) vs 1/T was
also linear and yielded ∆Hq ) 14.9 kcal/mol. The first-order
rate constants were very close at 70 °C for the solvents
benzene (3.6 × 10-5 s-1), cyclohexane (2.6 × 10-5 s-1),
and acetonitrile (1.8 × 10-5 s-1), indicating that a polar
dissociation mechanism (e.g., to a cyclopropenium toluene-
sulfinate ion pair)8 is improbable, especially since the rates
were somewhat slower in CH3CN than in C6H6 or cyclo-
hexane. We believe that the most reasonable explanation for
the facile thermal racemization reaction is a reversible 2,3-
sulfone-sulfinate allylic rearrangement, a process that is very
well-known in the allylic sulfinate to sulfone direction.9,10
The pathway for racemization is outlined in Scheme 2. The
research was carried out with 2a and 2b because these
cyclopropenes were obtained in excellent enantiomeric purity
(91% ee for 2a and 94% ee for 2b, as compared to 78% ee
for 2c). The molecular structures of 2a, 2b, and 2c were
proven to be as shown by single-crystal X-ray diffraction
analysis (all were dextrorotatory).7 In addition, (+)-2a was
converted to the Diels-Alder adduct 3, the structure and
absolute configuration of which were demonstrated by X-ray
crystallography.7 The absolute configurations of 2a-2c are
the same as predicted by the mechanistic model previously
proposed for catalyst 1.1 To the best of our knowledge these
chiral 2-cyclopropenyl sulfones are the first such chiral
compounds to have been synthesized. One interesting feature
emerged from the X-ray crystal structures of 2a, 2b, and
2c. In each case one of the oxygens attached to sulfur (a) is
located above the cyclopropenyl ring and equidistant from
the two olefinic carbons as shown in 4. This geometry may
Scheme 2. Pathway for Racemization of 2a via
[2,3]-Sigmatropic Rearrangement
result from the simultaneous operation of two not unreason-
able conditions: (1) staggering of substituents about the
cyclopropenyl C-S bond and (2) location of oxygen above
the three-membered ring in preference to tolyl because the
larger bulk of the latter would lead to steric repulsion. The
same conformational preference is also evident in (saturated)
cyclopropyl tolyl sulfones.
A surprising observation was made initially with 2-n-amyl-
2-cyclopropenyl 4-tolyl sulfone 2a. When this sulfone of 91%
ee was chromatographed on silica gel or stirred with a
suspension of silica gel in benzene for 30 min at 30 °C,
racemic sulfone 2a was recovered in high yield. This
unexpected finding prompted us to examine the kinetics of
the racemization process in solution, where it was also found
to be facile. The kinetics of thermal racemization of 2a in
benzene solution (without silica gel) were measured by
HPLC analysis using a Chiralcel OD column (Chiral
rates of such [2,3]-sigmatropic rearrangements are known
not to be especially sensitive to solvent polarity. One possible
explanation for the unique facility of the racemization of 2a
as compared to nonstrained allylic systems may be derived
from the known weakening of allylic bonds in the cyclo-
propene system.4 Thus, the allylic C-H bonds of cyclopro-
pene are much weaker (100.4 kcal/mol) than the C-H bonds
of cyclopropane (110.3 kcal/mol) (both BDEs).4 If the bond
weakening in the sulfone 2a is greater than that in the
sulfinate esters 2a′ and 2a′′, an acceleration of racemization
would result.11 Alternatively, there may be stabilization of
the transition state by virtue of some three-center π-electron
delocalization around the three-membered ring. In this
possibility, there might be a slight degree of cyclopropenium
cation character in the transition state.8
The observed silica-surface catalysis of the racemization
could easily be explained by such slight charge separation
(5) For recent reviews of transformations on cyclopropenes, see: (a)
Baird, M. S. Cyclopropanes: Synthesis: From Cyclopropenes. In Houben-
Weyl; Thieme: Stuttgart, 1997; pp 114-255. (b) Nakamura, M.; Isobe,
H.; Nakamura, E. Chem. ReV. 2003, 103, 1295-1326.
(8) (a) Norden, T. D.; Staley, S. W.; Taylor, W. H.; Harmony, M. D. J.
Am. Chem. Soc. 1986, 108, 7912-7918. (b) Breslow, R.; Groves, J. T.;
Ryan, G. J. Am. Chem. Soc. 1967, 89, 5048-5048. (c) Breslow, R. J. Am.
Chem. Soc. 1957, 79, 5318-5318.
(6) (a) Van Leusen, A. M.; Strating, J. Recl. TraV. Chim. Pays-Bas 1965,
84, 151-164. (b) Van Leusen, A. M.; Strating, J. Recl. TraV. Chim. Pays-
Bas 1965, 84, 140-151. (c) Van Leusen, A. M.; Strating, J. Org. Synth.
1977, VI, 981-988.
(9) For a review, see: Hoffmann, R. W. Angw. Chem., Int. Ed. Engl.
1979, 18, 563-640.
(10) (a) Braverman, S.; Mechoulam, H. Tetrahedron 1974, 30, 3883-
3890. (b) Cope, A. C.; Morrison, D. E.; Field, L. J. Am. Chem. Soc. 1950,
72, 59-67.
(7) For full details, see Supporting Information.
172
Org. Lett., Vol. 8, No. 1, 2006

in the transition state. Since silica gel behaves effectively
like a weak protic acid, hydrogen bonding to one oxygen of
the migrating TolSO2 group in the transition state would be
an obvious source of acceleration of racemization by the
pathway outlined in Scheme 2. Given the sizable magnitude
of the acceleration of racemization by silica gel, it is even
possible that this process occurs at the ion-pair extreme of
the mechanistic spectrum. Deprotonation of 2a R to the
TolSO2 group is clearly not a reasonable mechanistic
pathway for the silica gel promoted racemization because
silica gel is obviously not basic enough to form a high energy
antiaromatic (4π-type) carbanion intermediate.
not in deprotonation R to the TolSO2 group, but at the other
ring position as shown by subsequent reaction with Me3-
SiCl to form the TMS derivative 11 (91% ee, 77% yield).
Similarly, deprotonation of 2a (91% ee) by t-BuLi in THF
at -78 °C followed by treatment with iodine gave the chiral
iodo sulfone 12 (91% ee, 64% yield). The enhanced acidity
of the olefinic C-H of 2a is in part a consequence of the
considerably increased s-character (approaching an acetylenic
C-H) of the olefinic carbon.
Strong confirmatory evidence for the pathway of (solution-
phase) racemization of 2a via a [2,3]-sigmatropic rearrange-
ment to the achiral cyclopropenyl sulfinate 2a′ and achiral
sulfone 2a′′′ (Scheme 2) was obtained by a simple trapping
experiment. When 2a was heated at 55 °C in dry CD3OD
solution under an inert atmosphere the products 5, 6, 7, and
Hydrogenation of sulfone 2a gave stereoselectively the
corresponding cis-cyclopropyl sulfone 13, which was readily
deprotonated by n-BuLi-TMEDA in THF at -78 °C to give
after treatment with NH4Cl, methyl iodide, allyl bromide,
methyl chloroformate, or n-decyl iodide the sulfones 14a-
14e, again stereoselectively. The results indicate that the
conjugate base of 13 undergoes an inversion to the diaster-
eomer in which the n-C5H11 and TolSO2 groups are trans to
one another and that this more stable anion undergoes
alkylation (or electrophilic attack) with stereochemical reten-
tion.
1
8 shown in Scheme 3 were detected by H NMR spectro-
Scheme 3
Yet another facet of the unusual chemistry of sulfone 2a
was revealed by its behavior when treated with tetra-n-
butylammonium cyanide in THF at 23 °C. Under these
conditions, it was readily converted to a mixture of E and Z
forms of the exocyclic olefinic isomer 15. Although this
facile isomerization appears surprising at first glance, since
cyanide ion is generally not considered to be sufficiently
basic to effect allylic deprotonation of an olefin, it seems
not unreasonable when it is recalled that the difference in
strain energy between 1-methylcyclopropene and methylene
cyclopropane is ca. 15 kcal/mol.4
Relief of ring strain as a driving force for reactions of
cyclopropenes via carbanion intermediates adds to the
richness and uniqueness of cyclopropene chemistry. With
this in mind we have investigated the chemistry of 2-bro-
momethyl-2-cyclopropenyl 4-tolyl sulfone (2b). When this
compound was treated with 1 equiv of Na2CO3 in CD3OD
at 23 °C in an NMR tube and analyzed by 1H NMR
spectroscopy, a very rapid exchange of the vinylic proton at
scopy.7 Methanolysis of 2a′′ or ent-2a′′ produces 2H3 methyl
toluenesulfinate (5), which was also isolated, and cyclopro-
penol 9, which was too unstable to isolate but which gave
rise to the â-deuterated R,â-enal 6 and the corresponding
2H3-methyl acetal 7. The instability of 2-cyclopropenols,
which has previously been noted,12 is a consequence of the
high strain energy (ca. 55 kcal/mol) and the availability of
a carbonyl-forming elimination process which relieves that
strain. Ring strain also accounts for the ring cleavage process
that converts sulfone 2a′′′ via intermediate 10 to the
deuterated methoxy sulfone 8.
The action of the strong base t-BuLi on the cyclopropenyl
sulfone 2a (91% ee) in THF at -78 °C results, as expected,
(11) It is highly relevant that (S)-(1,2,3-triphenylcyclopropenyl)-O-
ethyldithiocarbonate undergoes [2,3]-sigmatropic rearrangement (degenerate)
with ∆Hq ) 14.8 kcal/mol and ∆Sq ) -10 eu, k298 ) 0.43 s-1. (a)
Mikhailov, I. E.; Dushenko, G. A.; Dorogan, I. V.; Minyaev, R. M.;
Negrebetskii, V. V.; Zschunke, A.; Minkin, V. I. MendeleeV Commun. 1994,
9-11. (b) Minkin, V. I.; Mikhailov, I. E.; Dushenko, G. A.; Kompan, O.
E.; Zschunke, A. Russ. Chem. Bull. 1998, 47, 884-894.
(12) For C-C cleavage as a consequence of the high strain energy of
2-cyclopropenols, see: Wolff, S.; Agosta, W. C. J. Am. Chem. Soc. 1984,
106, 2363-2367.
Org. Lett., Vol. 8, No. 1, 2006
173

C-3 to form 16 was observed (within 3 min). After 24 h, 16
was completely converted to the trideuterated exocyclic
olefin 17 (reaction incomplete after 12 h). When chiral 2b
Scheme 4
of 91% ee was treated with 1 equiv of carbonate in CH3OH
under the same conditions, the nondeuterated isomer 17
of 89% ee was obtained. This result indicates that the kine-
tically controlled H/D exchange 16 f 17 occurs with reten-
tion of configuration at the carbon R to the tosyl group.
Again, the driving force of endocyclic f exocyclic olefinic
isomerization dominates cyclopropene chemistry. This effect
also shows up in nucleophilic displacements of 2b, which is
converted to 18a-18d upon treatment with the nucleophiles
Et3N/THF, NaI/acetone, TolSO2-Na+/15-C-5/CH3CN or Bu4-
N+CN-/THF. Nucleophilic displacement at the secondary
carbon in preference to the primary carbon is unusual but
readily understandable because of the larger strain energy
of a cyclopropene relative to a methylene cyclopropane.
ylsilane with ethyl diazoacetate in the presence of 0.5% of
the chiral Rh catalyst 1 afforded the cyclopropene carboxylate
20 in 98% ee.7 Desilylation of 20 with tris(dimethylamino)-
sulfonium difluorotrimethyl silicate (TASF) in dry CH2Cl2
under a CO2 atmosphere at -78 °C yielded the monoethyl
ester of Feist’s acid (S,S-enantiomer) (21),16 which upon
esterification with phenyldiazomethane gave the (-)-(S,S)-
enantiomer of Feist’s acid ethyl, benzyl ester (22 of 98% ee
(65% overall from 20)). Reduction of 20 by LiAlH4 in ether
produced the trans-(S,S)-(+)-cyclopropyl carbinol 23.
The ready availability of chiral 2-cyclopropenyl tosyl sul-
fones has allowed the exploration of the chemical behavior
of this class of compounds and the discovery of several unu-
sual transformations. These include (1) facile racemization
at room temperature in benzene solution in the presence of
silica gel or at temperatures in the range 50-70 °C in its
absence; (2) identification of a reversible [2,3]-sigmatropic
sulfone-sufinate ester rearrangement pathway as the likely
reason for this racemization that is facilitated by the high
strain energy of cyclopropenes and weakness of the allylic
bonds; (3) stereoselective base-promoted alkylation of cis-
2-substituted cyclopropyl sulfones with inversion of config-
uration due to the pyramidal character of the intermediate
anion and a preference for trans geometry; and (4) an unusual
preference for nucleophilic allylic displacement at the secon-
dary versus primary carbon, another consequence of the
thermodynamic driving force associated with greater strain
for endocyclic cyclopropenes as compared to isomeric
exocyclic structures.
An Application: A Simple, Highly Enantioselective
Route to the Feist Ester Series. The intriguing C2-sym-
metric acid 19 and its esters, first prepared in racemic form
by Feist in 1893, has been the subject of much research.13
The (R,R)-(+)-enantiomer has been obtained either by
resolution of the quinine salt14a or by HPLC separation of
the dimethyl ester using a chiral column.14b We have applied
the fundamental knowledge described above to the develop-
ment of a convenient enantioselective synthesis of differenti-
ated diesters of 19 by the pathway outlined in Scheme 4.15
We believe that this is the first enantioselective synthesis in
the Feist ester/acid series.15 Reaction of propargyltrimeth-
Supporting Information Available: Experimental pro-
cedures and characterization data for all new products and
X-ray diffraction data (CIF) for 2a, 2b, 2c, 3, 17, 18c, and
18d. This material is available free of charge via the Internet
at http://pubs.acs.org.
(13) (a) Feist, F. Ber. Dtsch. Chem. Ges. 1893, 26, 747-764. (b) Feist,
J. Liebigs Ann. Chem. 1924, 436, 125-153. (c) Ettlinger, M. G. J. Am.
Chem. Soc. 1952, 74, 5805-5806. (d) Lloyd, D.; Downie, T. C.; Speakman,
J. C. Chem. Ind. (London) 1954, 222-223. (e) Lloyd, D.; Downie, T. C.;
Speakman, J. C. Chem. Ind. (London) 1954, 492-492. (f) Ettlinger, M.
G.; Kennedy, F. Chem. Ind. (London) 1956, 166-167. (g) Petersen, D. R.
Chem. Ind. (London) 1956, 904-905. (h) Kende, A. S. Chem. Ind. (London)
1956, 437-437. (i) Bottini, A. T.; Roberts, J. D. J. Org. Chem. 1956, 1169-
1170. (j) Lloyd, D. In Topics in Carbocyclic Chemistry; Lloyd, D., Ed.;
Plenum: New York, 1969; Vol. 1, pp 249-268. (k) Ramasubbu, N.;
Venkatesan, K. Acta Crystallogr. 1982, B38, 976-978.
OL052752+
(15) In general, methodology for the highly enantioselective synthesis
of methylenecyclopropanes is poorly developed. See: (a) Lautens, M.;
Delanghe, P. H. M. J. Org. Chem. 1993, 58, 5037-5039. (b) Baldwin, J.;
Adlington, R. M.; Bebbington, D.; Russell, A. T. J. Chem. Soc., Chem.
Commun. 1992, 1249-1251. (c) Achmatowicz, B.; Kabat, M.; Krajewski,
J.; Wicha, J. Tetrahedron 1992, 48, 10201-10210. (d) Nemoto, T.; Ojika,
M.; Sakagami, Y. Tetrahedron Lett. 1997, 38, 5667-5670.
(16) Approximately 9% of the cis-isomer of Feist’s mono ester 21 was
produced in the carboxylation of 20. This cis impurity was removed by
silica gel chromatography after conversion to the diester 22.
(14) (a) Doering, W. v. E.; Roth, H. D. Tetrahedron 1970, 26, 2825-
2835. (b) Baldwin, J. E.; Sakkab, D. H. J. Org. Chem. 1995, 60, 2635-
2637.
174
Org. Lett., Vol. 8, No. 1, 2006