Asymmetric photocycloadditions with an optically active allenylsilane: Trimethylsilyl as a removable stereocontrolling group for the enantioselective synthesis of exo-methylenecyclobutanes

Article abstract of DOI:10.1021/ja963436g

Allenylsilanes undergo enantioselective intramolecular photocycloaddition reactions with enones and enoates to give adducts in 76-99% enantiomeric excess and 67-90% yield. The silyl-substituted exo-methylenecyclobutane products undergo protodesilylation t

Full text of DOI:10.1021/ja963436g

VOLUME 119, NUMBER 11
MARCH 19, 1997
© Copyright 1997 by the
American Chemical Society
Asymmetric Photocycloadditions with an Optically Active
Allenylsilane: Trimethylsilyl as a Removable Stereocontrolling
Group for the Enantioselective Synthesis of
exo-Methylenecyclobutanes
Mary S. Shepard and Erick M. Carreira*
Contribution from the Arnold and Mabel Beckman Laboratory of Chemical Synthesis,
DiVision of Chemistry and Chemical Engineering, California Institute of Technology,
Pasadena, California 91125
ReceiVed October 1, 1996X
Abstract: Allenylsilanes undergo enantioselective intramolecular photocycloaddition reactions with enones and enoates
to give adducts in 76-99% enantiomeric excess and 67-90% yield. The silyl-substituted exo-methylenecyclobutane
products undergo protodesilylation to give the parent unsubstituted exo-methylenecyclobutane. Optically active
allenylsilanes may thus be used in enantioselective photocycloadditions in which the silyl moiety functions as a
removable stereochemical controlling group.
Introduction
cycloadditions afford key synthetic intermediates that are
otherwise not readily accessed.1,4,5
Two general strategies may be identified for control of relative
and absolute asymmetric induction in intramolecular photocy-
cloaddition reactions. These may be classified according to the
position of the stereochemical controlling group in relation to
the reacting partners and the cyclic adducts. The first type of
intramolecular diastereoselective photocycloaddition reactions
includes substrates in which the stereochemical controlling
Winkler, J. D.; Siegel, M. G.; Stelmach, J. E. Tetrahedron Lett. 1993, 34,
6509. (d) Kraus, G. A.; Zheng, D. Synlett 1993, 71. (e) Crimmins, M. T.;
Dudek, C. M.; Cheung, A. W.-H. Tetrahedron Lett. 1992, 33, 181. (f) Baker,
W. R.; Senter, P. D.; Coates, R. M. J. Chem. Soc., Chem. Commun. 1980,
1011.
(4) (a) Rawal, V. H.; Fabre, A.; Iwasa, S. Tetrahedron Lett. 1995, 36,
6851. (b) Crimmins, M. T.; Jung, D. K.; Gray, J. L. J. Am. Chem. Soc.
1993, 115, 3146. (c) Winkler, J. D.; Scott, R. D.; Williard, P. G. J. Am.
Chem. Soc. 1990, 112, 8971. (d) Pirrung, M. C.; Thomson, S. A. J. Org.
Chem. 1988, 53, 227. (e) Winkler, J. D.; Hershberger, P. M.; Springer, J.
P. Tetrahedron Lett. 1986, 27, 5177. (f) Crimmins, M. T.; DeLoach, J. A.
J. Am. Chem. Soc. 1986, 108, 800. (g) Smith, A. B.; Sulikowski, G. A.;
Sulikowski, M. M.; Fujimoto, K. J. Am. Chem. Soc. 1992, 114, 2567. (h)
Rawal, V. H.; Eschbach, A.; Dufour, C.; Iwasa, S. Pure Appl. Chem. 1996,
68, 675. (i) Pirrung, M. C. J. Am. Chem. Soc. 1979, 101, 7130.
(5) Meyers, A. I.; Fleming, S. A. J. Am. Chem. Soc. 1986, 108, 306.
The [2+2] photocycloaddition reaction of alkenes is an
important and powerful synthetic transformation for the con-
struction of complex molecules.1 The synthetic potential of this
reaction is formidable; in a single step a ring system is produced
along with two new C-C bonds and up to four stereogenic
centers. As an added benefit, the strained four-membered ring
products readily undergo ring-expansion reactions, providing
access to additional structures.2,3 Elegant applications of such
processes have been reported in which stereoselective [2+2]
X Abstract published in AdVance ACS Abstracts, March 1, 1997.
(1) (a) Crimmins, M. T. Chem. ReV. 1988, 88, 1453. (b) Crimmins, M.
T.; Reinhold, T. L. Organic Reactions; Wiley: New York, 1993; Vol. 44,
p 297. (c) Winkler, J. D.; Bowen, C. M.; Liotta, F. Chem. ReV. 1995, 95,
2003. (d) Schreiber, S. L. Science 1985, 227, 857. (e) Wiesner, K.
Tetrahedron 1975, 31, 1655.
(2) Bellus, D.; Ernst, B. Angew. Chem., Int. Ed. Engl. 1988, 27, 797.
(3) For selected, recent elegant examples of fragmentation and ring-
enlargement reactions of substituted cyclobutanes, see: (a) Crimmins, M.
T.; Huang, S.; Guise-Zawacki, L. E. Tetrahedron Lett. 1996, 37, 6519. (b)
Rawal, V. H.; Dufour, C.; Iwasa, S. Tetrahedron Lett. 1995, 36, 19. (c)
S0002-7863(96)03436-1 CCC: $14.00 © 1997 American Chemical Society

2598 J. Am. Chem. Soc., Vol. 119, No. 11, 1997
Shepard and Carreira
group, or stereogenic center, resides in the tether interconnecting
the reacting partners.4f Since the resident stereogenic center is
incorporated within one of the newly formed rings in the
product, these cycloaddition reactions are characterized by
intraannular chirality transfer.6 The second type of intra-
molecular diastereoselective photocycloaddition reactions is
exemplified by substrates in which the stereochemical control-
ling group resides outside of the newly formed rings in the
product; these reactions are characterized by extraannular
chirality transfer. Diastereoselective [2+2] photocycloadditions
have been reported using either of these approaches.4 However,
in general, the conformational and stereoelectronic constraints
accompanying ring formation resulting from intraannular chiral-
ity transfer afford greater control in reaction diastereoselectivity.
The strategies that have been employed in asymmetric
photocycloaddition reactions of monosubstituted allenes and
enones parallel those described in olefin-enone cycloaddition
reactions involving extraannular and intraannular chirality
transfer.4d However, the use of 1,3-disubstituted allenes affords
the opportunity to develop a novel class of [2+2] cycloaddition
reactions with the chiral allene itself serving as the stereocon-
trolling group. In contrast to photoadditions involving extraan-
nular and intraannular chirality transfer, the stereocontrolling
group in such allene-enone photocycloadditions is absent in
the product.7 In the course of the [2+2] cycloaddition reaction,
the stereogenic centers of the product are established at the
expense of the axial asymmetry of the allene. We have been
interested in developing this complementary cycloaddition
strategy into a useful method for the asymmetric synthesis of
complex ring systems.
Scheme 1
moiety functioning as a removable stereochemical controlling
group (eq 2).
We have reported the intramolecular cycloadditions of
optically active tert-butyl-substituted allenes with enones and
enoates to give photoadducts in excellent yields and up to 99%
asymmetric induction (eq 1).8 Although the products were
formed with high enantioselectivities, the bulky tert-butyl
substituent is a limitation to the synthetic elaboration of these
products. Thus, the synthetic utility of the enantioselective
allene-enone photocycloaddition reaction would benefit from
the development of a family of optically active allenes with a
controlling group analogous to the tert-butyl moiety which could
be removed following enantioselective photocycloaddition.
(2)
Results
The requisite optically active silylallene 6 was readily
prepared in multigram quantities from aldehyde 1, itself
synthesized from inexpensive, commercially available (S)-
(-)-malic acid (Scheme 1).9 Treatment of aldehyde 1 with the
Gilbert-Seyferth reagent (THF, -78 °C) afforded acetylene 2.10
Deprotonation of 2 (n-BuLi, THF) followed by addition of
Me3SiCl afforded the silylalkyne 3 (92%). Hydrolysis of the
benzylidene acetal (MeOH, TsOH) followed by selective
monosilylation of the isolated diol yielded alcohol 4 (80%, two
steps). The stereoselective conversion of this optically active
propargyl alcohol to the desired allenylsilane was effected using
reaction methodology recently reported by Myers.11 Thus,
treatment of 4 with [(o-nitrophenyl)sulfonyl]hydrazine under
Mitsunobu conditions (-15 °C, Ph3P, EtO2CNdNCO2Et)
afforded an intermediate alkylhydrazine which upon warming
(23 °C) gave allene 5 in 92% yield. Optically active allenyl-
silane-alcohol 6 was obtained quantitatively upon hydrolysis
of the O-silyl ether protecting group (HF/CH3CN, H2O).
Subsequent attachment of 6 to the R,â-unsaturated carbonyl
compounds was accomplished following procedures described
previously.8 Although the conversion of optically active pro-
pargyl alcohols analogous to 4 to the corresponding allenes has
been shown to occur stereospecifically,11 we checked that allene
6 had been produced with complete induction by gas chromato-
(1)
In this paper we disclose the synthesis of optically active
allenylsilanes and their use in enantioselective intramolecular
photocycloadditions. The tethered, optically active allenyl-
silanes undergo intramolecular [2+2] photocycloadditions with
enones and enoates to give cyclobutane adducts in 76-99%
enantiomeric excess (ee) and 67-90% yield. Importantly, the
trimethylsilyl-substituted exo-methylenecyclobutanes undergo
protodesilylation to give the corresponding unsubstituted pho-
toadducts. Thus, the optically active allenylsilanes may be used
in enantioselective photocycloadditions with the trimethylsilyl
(6) This nomenclature parallels that which has been used in auxiliary-
based enolate alkylation reactions; see: Evans, D. A. In Asymmetric
Synthesis; Morrison, J. D., Ed.; Stereodifferentiating Addition Reactions
Part B; Wiley: New York, 1984; Vol. 3, Chapter 1.
(7) This is an example of chirality transfer that has been termed by
Mislow as self-immolative; see: Mislow, K. Introduction to Stereochemistry;
Benjamin: New York, 1965; p 131.
(8) Carreira, E. M.; Hastings, C. A.; Shepard, M. S.; Yerkey, L. A.;
Millward, D. B. J. Am. Chem. Soc. 1994, 116, 6622.
(9) For the preparation of aldehyde 1, see: Pawlak, J.; Nakanishi, K.;
Iwashita, T.; Borowski, E. J. Org. Chem. 1987, 52, 2896.
(10) (a) Gilbert, J. C.; Weerasooriya, U. J. Org. Chem. 1982, 47, 1837.
(b) Seyferth, D.; Marmor, R. S.; Hilbert, P. J. Org. Chem. 1971, 36, 1379.
(11) Myers, A. G.; Zheng, B. J. Am. Chem. Soc. 1996, 118, 4492.

Asymmetric Photocycloadditions with an Allenylsilane
J. Am. Chem. Soc., Vol. 119, No. 11, 1997 2599
Table 1. Photocycloaddition of Optically Active Silyl-Substituted
Table 2. Protodesilylation of Silyl-Substituted Photoadducts
Allenes
a Yields correspond to isolated material after purification by chro-
matography on silica gel. b Reaction was performed on the mixture of
alkene diastereomers (2-1.5:1) isolated from the photocycloaddition
reaction.
the optical purity of each diastereomer individually as shown
in Table 1.12
The stereochemical induction in these reactions ranged from
76 to 99% ee.13 The determination of asymmetric induction of
the products necessitated the development of several different
protocols, since in only one case (entry 3) was it possible to
assay directly the optical purity by GC on a Cyclodex-B chiral
column. The stereoisomeric purity of 13 (entry 1) was
established upon reduction of the ketone (NaBH4, MeOH) and
analysis of the alcohol product by gas chromatography. For
adduct 14 (entry 2) the enantioselectivity was assayed upon
reduction of the ketone (NaBH4, MeOH), conversion of the
isolated secondary alcohol to the corresponding (S)-MTPA ester,
a Isolated yields of both diastereomers after purification by chro-
matography on silica gel. b 2-1.5:1 ratio of olefin diastereomers
observed. c Isolated as a single alkene diastereomer. d In all cases, the
optical purity was established by comparison to authentic racemic
mixtures. The % ee’s were determined as follows: e Reduction of 13
to the corresponding alcohol and analysis by GC on a Cyclodex-B chiral
column. f Reduction of 14 to the corresponding alcohol, preparation
1
of the (S)-MTPA ester, and analysis by H NMR. g Direct analysis of
the adducts by GC on a Cyclodex-B chiral column. h Reduction of 16
and 17 to the corresponding phenol-alcohols, preparation of the bis[(S)-
MTPA esters], and analysis by 19F NMR.
1
and analysis by H NMR spectroscopy. For the coumarin-
derived products 16 and 17 reduction to the phenol-alcohol
and preparation of the bis[(S)-MTPA esters] permitted the extent
of asymmetric induction to be determined in the product by
19F NMR spectroscopy.
graphic analysis using a Cyclodex-B chiral column and com-
parison to authentic racemic material.
All photocycloaddition reactions were performed in cyclo-
hexane at ambient temperature with a Hanovia 450 W Hg
medium-pressure UV lamp in a water-cooled quartz immersion
apparatus through a Pyrex filter (eq 3). As shown in Table 1,
Having prepared the optically active photoadducts, we sought
reaction conditions that would excise the trimethylsilyl group.
In this regard, it was desirable to find protodesilylation
conditions to effect removal of the trimethylsilyl moiety without
compromising the exo-methylenecyclobutane ring. Under a
variety of different conditions, such as Bu4NF/THF, K2CO3/
MeOH, and KF/MeOH, we observed extensive decomposition
of the starting materials, giving highly polar products. When
HF‚py was employed, unreacted starting material was reisolated.
However, use of Bu4NF buffered with glacial acetic acid in THF
at 23 °C proved sufficiently mild, allowing for isolation of the
exo-methylenecyclobutanes 18-21 in good yields (eq 4 and
Table 2). The protodesilylation reactions were performed on
(3)
the photoadducts were isolated in 67-90% yield. In analogy
to observations made in our previously reported investigation
involving tert-butyl-substituted allenes, the coumarin-derived
adducts (entry 4, 16; entry 5, 17) were formed as single alkene
diastereomers as determined by 1H NMR spectroscopy. Enones
8-10 gave the substituted exo-methylenecyclobutane products
as a 2-1.5:1 mixture of alkene E and Z diastereomers which
could be readily separated by chromatography on silica gel.
Although for the purposes of this investigation the formation
of such alkene diastereomers is inconsequential, we determined
(12) The optical purity of 14Z could not be determined accurately as a
1
consequence of carbamate rotamers. The H and 19F NMR resonances for
the derived (S)-MTPA ester were poorly resolved to give an estimated 80-
85% ee in the product.
(13) The absolute stereochemistry of the adducts is presumed to parallel
that previously reported for the tert-butyl-substituted allenes, for which X-ray
crystallographic analysis of an appropriate derivative was possible (ref 8).
To date, we have been unable to obtain crystalline material of adducts 13-
17 and their derivatives.

2600 J. Am. Chem. Soc., Vol. 119, No. 11, 1997
Shepard and Carreira
(4)
the mixture of alkene diastereomers isolated from the photo-
cycloadditions. It should be noted that the E-substituted
diastereomers were routinely observed to undergo protodesily-
lation at a faster rate than the more sterically encumbered
Z-substituted adducts.
Discussion
The use of an allenylsilane in [2+2] cycloadditions consider-
ably expands the scope of products that may be accessed via
enantioselective allene-enone synthetic photochemistry. When
compared to the tert-butyl-substituted allenes we have previously
described and studied, the trimethylsilyl-substituted counterparts
provide adducts with slightly diminished levels of induction.
This observation is consistent with the effective size of the
stereochemical controlling groups (Me3C- versus Me3Si-) on
the allene terminus and its ability to block one of the allene
diastereofaces. Although the van der Waals radius of Me3Si-
is larger than that of Me3C-, the longer Si-C bond (1.81 Å)
versus C-C bond (1.54 Å) attenuates the effectiveness of the
former as a blocking group.14 In this regard, we have observed
higher levels of asymmetric induction when an optically active
triisopropylsilyl-substituted allene is employed in intramolecular
[2+2] photocycloaddition reactions (eq 5).15
Figure 1. Working model for the enantioselective photocycloadditions.
consequent loss of olefin geometry in reaction products such
as 13, 14, and 15 (Table 1).19,20 Interestingly, the coumarin
substrates gave photoproducts as single alkene diastereomers.
This result can be understood when it is considered that for
these cases the resulting 1,4-biradical intermediate is buttressed
by peri-interactions with the adjacent aromatic ring. We have
proposed that such steric interactions considerably constrain the
conformation of the 1,4-biradical and lead to rapid collapse to
products.8 The observation of asymmetric induction for the
cycloadditions reported herein is consistent with a mechanism
in which ring closure by the 1,4-biradical 23 is faster than
retroaddition, kc(23f24) . kr(23f22).21,22 In this regard, the
lifetime of 1,4-biradicals analogous to 23 can vary from 15 to
900 ns.23 The inversion barriers for vinyl and methylvinyl
radicals have been determined by ESR spectroscopy to be in
the range 2-3 kcal/mol, giving rise to bent conformations
possessing lifetimes with upper and lower limits of 10-8 and
10-10 s.18,24,25 Additionally, the barrier to rotation about the
CsC single bond in the methylvinyl radical (H2CdCMe) has
been observed to be 1300 cal/mol, giving lifetimes of 2 × 10-10
s for the conformational rotamers. Thus, the reversible forma-
tion of the 1,4-biradical 23 may be excluded for substrates giving
products with high enantioselectivities. Given the lifetimes for
bond rotation and vinyl radical inversion in relation to the
lifetime of 1,4-biradicals, reversible formation of 23 would likely
result in loss of allene enantiopurity. Supporting experimental
data for this hypothesis may be found in studies we have
(5)
The working model which accounts for the transfer of
asymmetry during the cycloaddition reaction parallels the Becker
model we have previously invoked (Figure 1). The regiochem-
istry of the initial bond formation uniformly follows Hammond’s
and Srinivasan’s “rule of five”.16 Formation of 23, including a
C(R)-radical, is consistent with Winkler’s observations in related
olefin-enone cycloadditions.17 The observation of high optical
activity in the photoadducts suggests a mechanism in which
the stereochemistry of the products is established kinetically
upon addition of the enone excited state to the least hindered
allene face (7 f 22 f 23, Figure 1). Importantly, within the
lifetime of the putative 1,4-biradical intermediate 23, C-C bond
rotation and inversion of the vinyl radical occurs18 with
(19) For mechanistic investigations of alkene-enone photocycloadditions,
see: (a) Andrew, D.; Weedon, A. C. J. Am. Chem. Soc. 1995, 117, 5647.
(b) Haddad, N.; Abramovich, Z. J. Org. Chem. 1995, 60, 6883. For
computational studies, see: (c) Broeker, J. L.; Eksterowicz, J. E.; Belk, A.
J.; Houk, K. N. J. Am. Chem. Soc. 1995, 117, 1847.
(20) For a [2+2] photocycloaddition reaction of an alkene with an enone
which intercepts the 1,4-biradical, see: Becker, D.; Morlender, N.; Haddad,
N. Tetrahedron Lett. 1995, 36, 1921.
(21) For an example in which retroaddition competes with ring closure,
see: (a) McCullough, J. J.; Ramachandran, B. R.; Snyder, F. F.; Taylor, G.
N. J. Am. Chem. Soc. 1975, 97, 6767. (b) Hastings, D. J.; Weedon, A. C.
J. Am. Chem. Soc. 1991, 113, 8525. (c) Maradyn, D. J.; Sydnes, L. K.;
Weedon, A. C. Tetrahedron Lett. 1993, 34, 2413. (d) Rudolph, A.; Weedon,
A. C. Can. J. Chem. 1990, 68, 1590.
(22) For recent theoretical discussion of allene-enone photocycloaddi-
tions, see: Froese, R. D. J.; Lange, G. L.; Goddard, J. D. J. Am. Chem.
Soc. 1996, 61, 952.
(23) Schuster, D. I.; Lem, G.; Kaprinidis, N. A. Chem. ReV. 1993, 93, 3.
(24) (a) Fossey, J.; Lefort, D.; Sorba, J. Free Radicals in Organic
Chemistry; Wiley: Chichester, 1995; Chapter 3. (b) Lowry, T. H.;
Richardson, K. S. Mechanism and Theory in Organic Chemistry; Harper
Collins: New York, 1987; Chapter 9.
(25) In elegant mechanistic work, Weedon has reported trapping the
triplet 1,4-biradical intermediates in allene-enone photocycloadditions; see
ref 21c. The conditions of the trapping experiment (H2Se) are analogous
to those previously employed by these researchers in trapping the corre-
sponding 1,4-biradical intermediate produced in alkene-enone photocy-
cloadition reactions.19a
(14) This same effect is manifest in the observed A values for -CMe3
(-∆G° ) 4.7 kcal/mol) when compared to -SiMe3 (-∆G° ) 2.5 kcal/
mol). See: Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic
Compounds; Wiley: New York, 1994; Chapter 11, p 695.
(15) Although the triisopropyl-substituted allenes typically gave products
with higher levels of induction than the corresponding trimethylsilyl-
substituted allenes, the desilylation of these photoadducts proceeded in
diminished yields. We are currently investigating this novel class of allenes,
and their synthetic chemistry which will be the subject of future work in
our laboratories.
(16) (a) Liu, R. S. H.; Hammond, G. S. J. Am. Chem. Soc. 1967, 89,
4936. (b) Srinivasan, R.; Carlough, K. H. J. Am. Chem. Soc. 1967, 89,
4932. (c) For recent investigations which provide verification to the rule of
five, see: Maradyn, D. J.; Weedon, A. C. J. Am. Chem. Soc. 1995, 117,
5359.
(17) Winkler, J. D.; Shao, B. Tetrahedron Lett. 1993, 34, 3355.
(18) (a) Fessenden, R. W.; Schuler, R. H. J. Chem. Phys. 1963, 39, 2147.
(b) Kochi, J. K. AdV. Free Radical Chem. 1975, 5, 189.

Asymmetric Photocycloadditions with an Allenylsilane
J. Am. Chem. Soc., Vol. 119, No. 11, 1997 2601
Figure 2. Transition-state models for the addition of allene and enone.
Scheme 2
previously reported involving the intramolecular [2+2] photo-
cycloaddition reaction of a tethered optically active allene with
cyclopentenone, a class of substrates routinely giving the lowest
levels of asymmetric induction. We have observed that inter-
ruption of the reaction at various stages of conversion and
analysis of the recovered starting material affords allene with
diminished levels of optical purity.8
Although allene regioselectivity and face selectivity may be
understood on the basis of well-known kinetic and steric
preferences in related photochemical processes, the origins of
the enone face selectivity (Si versus Re) are not clear. Shown
in Figure 2 are the possible arrangements between allene and
enone CdC bonds (two synclinal and one anticlinal) to be
considered for the attack of the allene on each of the diaste-
reofaces of the enone. Although a more detailed analysis awaits
additional data on the nature of transition states for such [2+2]
photocycloaddition reactions, we have chosen to consider
idealized staggered orientations which minimize nonbonded
steric interactions between the substituents on the two sp2-
hybridized carbons.26 Examination of molecular models permits
conformers 22c and 25a to be excluded since the ethylene tether
interconnecting the enone and allene must incur significant
angular deformation in such arrangements. In structures 22b
and 25b the tethered allene is disposed synclinal to the enone
CdC and endo to the preexisting ring. In both of these
arrangements, the distal allenyl CsH points directly toward the
cyclic enone, resulting in steric interactions with the ring.
Importantly, these interactions are absent in 22a and 25c.
Differing only in the relative orientation of the enone and allene
CdC (synclinal 22a versus anticlinal 25c), these two structures
possess otherwise similar nonbonded interactions and are thus
difficult to differentiate. The stereochemistry of the products
that we have routinely obtained in allene-enone cycloadditions
is only consistent with the reaction proceeding through structure
22a. Structure 25c would yield diastereomeric product 27,
possessing a trans-fused 4-5 ring system (Scheme 2) which
we have not observed in the photocycloaddition reactions
performed to date.27,28
arrangement 22a over the anticlinal arrangement 25c may result
from stabilizing secondary orbital interactions between the
nascent singly occupied orbitals on the central allene carbon
and the enone R-position.29 It is also possible that a bias for
synclinal orientation may be established in the formation of an
exciplex prior to carbon-carbon bond formation. In the more
commonly employed stereoselective, synthetic reactions involv-
ing ground-state polar intermediates, numerous steric, inductive,
and stereoelectronic effects have been documented to bias the
stereochemical outcome in the products in substantial ways. By
contrast, a detailed understanding of similar effects in stereo-
selective synthetic photochemistry is still lacking, and thus
additional mechanistic investigations of this fundamental process
are warranted.
The successful development of the protodesilylation reaction
conditions required careful optimization and examination of
numerous fluoride sources. Importantly, the optimal conditions
necessitated the use of fluoride buffered with acetic acid. Under
basic conditions (Bu4NF/THF, K2CO3/MeOH, and KF/MeOH)
we observed extensive decomposition of the starting materials
with the formation of highly polar products. We speculate that
under these basic conditions Grob fragmentation takes place to
give an enolate that subsequently undergoes â-elimination
(Scheme 3); the 2-en-5-ynone so produced would be expected
The energetic difference between 22a and 25c likely results
from a combination of subtle effects which dictate the orientation
of the allene and enone during the formation of 1,4-biradical
intermediate 23. An inherent kinetic preference for the synclinal
(28) It is important to note that formation of the first spirocyclic ring
from 22a and 25c, for example, is an exothermic process and likely proceeds
through an early transition state that should not resemble the cis- or trans-
fused, 4-5 ring systems present in the ultimate products 24 and 27,
respectively.
(26) Houk, K. N.; Paddon-Row, M. N; Rondan, N. G.; Wu, Y.-D.;
Brown, F. K.; Spellmeyer, D. C.; Metz, J. T.; Li, Y.; Loncharich, R. J.
Science 1986, 231, 1108.
(27) For a discussion of the relative orientation between alkene and enone,
and the resulting consequences in the stereochemistry of the photoadducts,
see ref 19.
(29) Similar issues in the addition of allylmetal reagents to aldehydes
have been discussed. For a leading reference, see: Keck, G. E.; Dougherty,
S. M.; Savin, K. A. J. Am. Chem. Soc. 1995, 117, 6210.

2602 J. Am. Chem. Soc., Vol. 119, No. 11, 1997
Shepard and Carreira
Scheme 3
ditions of achiral monosubstituted allenes with enones, a
synthetic transformation which lacks precedence (eq 6).
(6)
Conclusion
We have demonstrated that optically active allenylsilanes will
undergo intramolecular [2+2] photocycloadditions with enones
and enoates to give cyclobutane products in high enantioselec-
tivity. Importantly, this represents the first use of allenylsilanes
as substrates in [2+2] photocycloadditions. Additionally,
conditions have been developed which effect protodesilylation
of these adducts, thereby affording the unsubstituted exo-
methylenecyclobutanes. The salient features of this process
include (1) an optically active allenylsilane which is easily
prepared, (2) compatibility of the allenylsilane with the photo-
reaction conditions, and (3) the ease with which the trimethyl-
silyl controlling group is excised from the reaction products.
The methodology described herein should prove useful in the
preparation of fused ring systems that are not otherwise readily
synthesized in optically active form.
Figure 3. Computationally minimized structure of 13 with graphical
representation of the calculated LUMO.
to decompose extensively under the reaction conditions.30 In
this regard, the use of added acetic acid as a buffer in the
protodesilylation reaction ensures that the putative vinylsili-
conate 28 rapidly protonates to afford the observed product 29
at the expense of fragmentation leading to decomposition.31,32
Experimental Section
General Procedures. All reagents were commercially obtained
except where noted. Where appropriate, reagents were purified prior
to use. All nonaqueous reactions were performed using oven-dried
glassware under an atmosphere of dry nitrogen. Air- and moisture-
sensitive liquids and solutions were transferred Via syringe or stainless
steel cannula. Organic solutions were concentrated by rotary evapora-
tion below 35 °C at ∼25 mmHg (water aspirator). Photoreactions were
performed in Pyrex flasks (λ > 293 nm) using a Hanovia 450 W Hg
medium-pressure UV lamp in a water-cooled quartz immersion
apparatus as the light source. Diethyl ether and tetrahydrofuran (THF)
were distilled from sodium-benzophenone ketyl prior to use. Dichlo-
romethane, pyridine, and triethylamine were distilled from calcium
hydride prior to use. Dimethyl sulfoxide (DMSO) and N,N-dimeth-
ylformamide (DMF) were distilled under reduced pressure from calcium
hydride and stored over 4 Å molecular sieves. Methanol was distilled
from magnesium methoxide prior to use. Spectroscopy grade cyclo-
hexane was used in photoreactions. Chromatographic purification of
products was accomplished using forced flow chromatography on EM
Science Geduran silica gel 60 according to the method of Still.36 Thin-
layer chromatography was performed on EM Reagents 0.25 mm silica
gel 60F plates (230-400 mesh). Visualization of the developed
chromatogram was performed by fluorescence quenching, ethanolic
p-anisaldehyde stain, or aqueous ceric ammonium molybdate (CAM)
stain.
It is remarkable that protodesilylation of the silyl-substituted
exo-methylenecyclobutanes can compete effectively with ring
fragmentation. Analysis of the computationally-minimized
molecular model of photoadduct 13 reveals an ideal stereoelec-
tronic alignment among the exocyclic C-Si bond, the cyclobutyl
C-C bond, and the carbonyl antibonding orbital (Figure 3).33
Moreover, calculation and inspection of the LUMO orbital in
13 using semiempirical methods (AM1) reveal that this anti-
bonding orbital has considerable contributions from the adjacent
cyclobutyl C-C and C-Si bonds.
To the best of our knowledge, the results described herein
represent the first example of asymmetric photocycloadditions
involving chiral, optically active allenylsilanes. In combination
with the photocycloaddition of Si-tethered alkenes and enones
described by Crimmins34 and the photoremovable silyl protect-
ing group developed by Pirrung,35 the methodology we delineate
further expands the use of silanes in synthetic photochemistry.
The sequential intramolecular photocyclization of allenylsilanes
with enones and protodesilylation of the trimethylsilyl-
substituted products provide access to optically active exo-
methylenecyclobutanes incorporating complex fused ring sys-
tems. The products obtained from such a two-step procedure
are formally the photoadducts of enantioselective photocycload-
NMR spectra were recorded on a Bruker AM-500 operating at 500
1
and 470 MHz for H and 19F, respectively, or a General Electric QE
Plus operating at 300 and 75 MHz for 1H and 13C, respectively. Data
1
for H are reported as follows: chemical shift (δ, ppm), multiplicity
(s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet), integration,
and coupling constant (J, Hz). IR spectra were recorded on a Perkin-
Elmer Paragon 1000 FTIR spectrometer using NaCl salt plates and are
reported in terms of frequency of absorption (ν, cm-1). Optical rotations
were determined on a JASCO DIP-1000 digital polarimeter operating
at the sodium D line and are reported as follows: [R]Na, concentration
(g/100 mL), and solvent. Gas chromatographic analyses were per-
formed on a Hewlett-Packard 6890 gas chromatograph with a flame
ionization detector and a 30 m J&W Cyclodex-B capillary column.
High-resolution mass spectra were obtained from the Mass Spectrometry
Laboratory, Division of Chemistry and Chemical Engineering, Cali-
fornia Institute of Technology.
(30) Wovkulich, P. M. In ComprehensiVe Organic Synthesis; Trost, B.
M., Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 1.
(31) For an excellent leading reference concerning siliconates, see: Chuit,
C.; Corriu, R. J. P.; Reye, C.; Young, J. C. Chem. ReV. 1993, 93, 1371.
(32) For the facile Lewis acid mediated Grob fragmentation of fused
4-6 ring systems with an exocyclic Me3SiCH2 moiety, see: Ochiai, M.;
Arimoto, M.; Fujita, E. J. Chem. Soc., Chem. Commun. 1981, 460.
(33) The calculation of the LUMO of 13 was performed using AM1
semiempirical methods supported by the Spartan molecular modeling
package available from Wavefunction, Inc., Irvine, CA.
(34) Crimmins, M. T.; Guise, L. E. Tetrahedron Lett. 1994, 35, 1657.
(35) Pirrung, M. C.; Lee, Y. R. J. Org. Chem. 1993, 58, 6961.
(36) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923.

Asymmetric Photocycloadditions with an Allenylsilane
J. Am. Chem. Soc., Vol. 119, No. 11, 1997 2603
Terminal Alkyne 2. A 1.0 M solution of KOtBu (48 mL, 48 mmol,
1.0 equiv) in THF was diluted with 100 mL of THF and cooled to
-78 °C. A solution of dimethyl (diazomethyl)phosphonate (7.2 g, 48
mmol, 1.0 equiv) in 50 mL of THF was added dropwise over 10 min
Via cannula. After stirring at -78 °C for 25 min, a solution of the
crude aldehyde 1 (48 mmol, 1.0 equiv) in 90 mL of THF was transferred
Via cannula whereupon effervescence occurred. The reaction mixture
was stirred at -78 °C for 10 h and then at 23 °C for an additional 2 h.
The mixture was poured into 500 mL of H2O and extracted with 3 ×
250 mL of CH2Cl2. The combined organic extracts were washed with
500 mL of saturated aqueous NaCl, dried over anhydrous Na2SO4, and
concentrated in Vacuo. Purification of the residue by chromatography
on silica gel (gradient elution, 5% Et2O in pentane to 10% Et2O in
pentane) afforded 4.1 g (45%) of alkyne 2: TLC Rf ) 0.22 (5% Et2O
9H), 0.09 (s, 3H), 0.08 (s, 3H); 13C NMR (CDCl3, 75 MHz) δ 106, 89,
62, 61, 39, 26, 18, -0.1, -5.5; IR (thin film) ν 3418 (br), 2958, 2930,
2859, 2172; HRMS (FAB+) calcd for C14H31O2Si2+ (MH+) 287.1863,
found 287.1851.
Allene 5. To triphenylphosphine (0.93 g, 3.5 mmol, 1.3 equiv) in
10 mL of THF at -15 °C was added diethyl azodicarboxylate (0.56
mL, 3.5 mmol, 1.3 equiv) dropwise. After 5 min a solution of propargyl
alcohol 4 (0.78 g, 2.7 mmol, 1.0 equiv) in 10 mL of THF was added
dropwise Via cannula over 10 min. The mixture was stirred at -15
°C for an additional 10 min before a solution of [(o-nitrophenyl)-
sulfonyl]hydrazine (0.77 g, 3.5 mmol, 1.3 equiv) in 10 mL of THF
was added dropwise Via cannula over 10 min. The resultant orange
solution was stirred at -15 °C for 30 min and then was allowed to
warm slowly to 23 °C over 23 h. The reaction mixture was
concentrated in Vacuo, and the residue was purified by chromatography
on silica gel (pentane) to afford 0.67 g (92%) of allene 5: TLC Rf )
1
in pentane); [R]Na -31.2° (c ) 0.50, CHCl3); H NMR (CDCl3, 300
MHz) δ 7.52-7.48 (m, 2H), 7.40-7.33 (m, 3H), 5.51 (s, 1H), 4.67
(dt, 1H, J ) 11.6, 2.3 Hz), 4.28 (ddd, 1H, J ) 11.7, 5.1, 1.1 Hz), 3.97
(td, 1H, J ) 12, 2.5 Hz), 2.54 (d, 1H, J ) 2.1 Hz), 2.27 (m, 1H), 1.79
(dm, 1H, J ) 12.9 Hz); 13C NMR (CDCl3, 75 MHz) δ 138, 129, 128,
126, 102, 82, 74, 67.2, 66.6, 32; IR (thin film) ν 3287, 3066, 3036,
1
0.91 (4:1 hexanes/EtOAc); [R]Na +76.7° (c ) 1.0, CHCl3); H NMR
(CDCl3, 300 MHz) δ 4.90-4.87 (m, 1H), 4.77 (q, 1H, J ) 7.0 Hz),
3.63 (t, 2H, J ) 7.3 Hz), 2.23-2.14 (m, 2H), 0.90 (s, 9H), 0.09 (s,
9H), 0.06 (s, 6H); 13C NMR δ 210, 82, 80, 63, 32, 26, 18, -0.9, -5.2;
IR (thin film) ν 2957, 2930, 2859, 1939 (s).
+
2967, 2931, 2858, 2126; HRMS (FAB+) calcd for C12H13O2 (MH+)
189.0916, found 189.0917.
Allenyl Alcohol 6. To 0.74 g of 5 (2.7 mmol, 1.0 equiv) was added
a solution of HF (95:5:1.5 CH3CN/48% HF/H2O, 2.7 mL). After
stirring at 23 °C for 30 min, the reaction mixture was loaded directly
onto a column of silica gel and subjected to chromatography (4:1
pentane/Et2O) to afford 0.43 g (100%) of allenyl alcohol 6: TLC Rf )
Silylalkyne 3. To a solution of 2 (0.81 g, 4.3 mmol, 1.0 equiv) in
43 mL of THF at 0 °C was added a 1.6 M solution of nBuLi (4.0 mL,
6.5 mmol, 1.5 equiv) in hexane. The mixture was stirred at 0 °C for
15 min before chlorotrimethylsilane (1.1 mL, 8.6 mmol, 2.0 equiv)
was added. After 15 min the reaction was poured into 75 mL of H2O
and 50 mL of Et2O. The organic layer was collected, and the aqueous
layer was extracted with 2 × 50 mL of Et2O. The combined organic
extracts were washed with 50 mL of saturated aqueous NaCl, dried
over anhydrous Na2SO4, and concentrated in Vacuo. Purification of
the residue by chromatography on silica gel (4% EtOAc in hexanes)
yielded 1.0 g (92%) of a white crystalline solid: mp 59-62 °C
(hexanes/EtOAc); TLC Rf ) 0.71 (4:1 hexanes/EtOAc); [R]Na -40.1°
(c ) 1.0, CHCl3); 1H NMR (CDCl3, 300 MHz) δ 7.53-7.50 (m, 2H),
7.40-7.33 (m, 3H), 5.49 (s, 1H), 4.66 (dd, 1H, J ) 11.5, 2.5 Hz),
4.26 (dm, 1H, J ) 12 Hz), 3.95 (td, 1H, J ) 12.1, 2.4 Hz), 2.27-2.19
(m, 1H), 1.77 (dq, 1H, J ) 13.6, 2.0 Hz), 0.18 (s, 9H); 13C NMR
(CDCl3, 75 MHz) δ 138, 129, 128, 126, 103, 102, 90, 68, 67, 32, -0.2;
IR (thin film) ν 3068, 3037, 2963, 2853, 2187; HRMS (FAB+) calcd
for C15H21O2Si+ (MH+) 261.1311, found 261.1316.
1
0.48 (3:1 hexanes/EtOAc); [R]Na +87.8° (c ) 1.0, CHCl3); H NMR
(CDCl3, 300 MHz) δ 4.98-4.93 (m, 1H), 4.76 (q, 1H, J ) 6.9 Hz),
3.66 (td, 2H, J ) 6.3, 1.6 Hz), 2.26-2.18 (m, 2H), 1.8 (br s, 1H), 0.09
(s, 9H); 13C NMR (CDCl3, 75 MHz) δ 210, 83, 79, 62, 31, -1.0; IR
(thin film) ν 3316 (br), 2956, 2898, 1939 (s).
Allenyl Mesylate. To allenyl alcohol 6 (160 mg, 1.0 mmol, 1.0
equiv) in 4 mL of CH2Cl2 at 0 °C was added 210 µL (1.5 mmol, 1.5
equiv) of triethylamine followed by 94 µL (1.2 mmol, 1.2 equiv) of
methanesulfonyl chloride. The reaction mixture was stirred at 0 °C
for 30 min and then was poured into 20 mL of pH 7 phosphate buffer.
The aqueous layer was extracted with 3 × 10 mL of Et2O, and the
combined organic layers were washed with 15 mL of saturated aqueous
NaCl, dried over anhydrous Na2SO4, and concentrated in Vacuo. The
resultant oil was purified by chromatography on silica gel (1:1 CH2-
Cl2/hexanes) to afford 0.23 g (97%) of the mesylate: TLC Rf ) 0.71
1
(5:1 CH2Cl2/hexanes); H NMR (CDCl3, 300 MHz) δ 5.00 (dt, 1H, J
Alkynediol. To a solution of 3 (1.0 g, 3.8 mmol, 1.0 equiv) in 16
mL of MeOH was added 7 mg (0.04 mmol, 1 mol%) of p-
toluenesulfonic acid monohydrate. The reaction was stirred at 23 °C
for 70 min and then was poured into 25 mL of saturated aqueous
NaHCO3 and 15 mL of Et2O. The organic layer was collected, and
the aqueous layer was extracted exhaustively with CH2Cl2. The
combined organic layers were washed with saturated aqueous NaCl,
dried over anhydrous Na2SO4, and concentrated in Vacuo. Purification
of the residue by chromatography on silica gel (2:1 hexanes/EtOAc
and then 1:1 hexanes/EtOAc) provided 0.66 g (100%) of the diol: TLC
) 6.7, 3.6 Hz), 4.76 (q, 1H, J ) 6.9 Hz), 4.24 (t, 2H, J ) 6.8 Hz),
3.02 (s, 3H), 2.46-2.38 (m, 2H), 0.11 (s, 9H).
Photosubstrate 8. To a solution of 6 (130 mg, 0.80 mmol, 1.0
equiv) in 8 mL of THF was added 120 mg of 1,3-cyclohexanedione
(1.0 mmol, 1.3 equiv), 270 mg of triphenylphosphine (1.0 mmol, 1.3
equiv), and 170 µL of diethyl azodicarboxylate (1.0 mmol, 1.3 equiv).
The mixture was stirred at 23 °C for 20 min, concentrated in Vacuo,
and subjected to chromatography on silica gel (4:1 hexanes/EtOAc) to
afford 170 mg (84%) of 8: TLC Rf ) 0.15 (4:1 hexanes/EtOAc); [R]Na
+69.2° (c ) 1.0, CHCl3); 1H NMR (CDCl3, 300 MHz) δ 5.30 (s, 1H),
4.9 (m, 1H), 4.74 (q, 1H, J ) 6.8 Hz), 3.82 (t, 2H, J ) 6.6 Hz), 2.27-
2.38 (m, 6H), 1.9 (m, 2H), 0.46 (s, 9H); 13C NMR (CDCl3, 75 MHz)
δ 210, 199, 178, 103, 84, 79, 68, 37, 29, 27, 21, -1.1; IR (thin film)
ν 3072 (w), 2953, 2896, 1940 (s), 1667 (s), 1606 (s); HRMS (FAB+)
calcd for C14H22O2Si 250.1389, found 250.1391.
1
Rf ) 0.21 (2:1 hexanes/EtOAc); [R]Na -30.8° (c ) 1.0, CHCl3); H
NMR (CDCl3, 300 MHz) δ 4.62 (dd, 1H, J ) 6.7, 5.0 Hz), 4.01-3.93
(m, 1H), 3.87-3.80 (m, 1H), 2.82 (br s, 2H), 2.04-1.92 (m, 2H), 0.16
(s, 9H); 13C NMR (CDCl3, 75 MHz) δ 106, 90, 62, 60, 39, -0.2; IR
(thin film) ν 3317 (br), 2959, 2173; HRMS (EI+) calcd for C8H15O2Si
(M - 1) 171.0841, found 171.0834.
Propargyl Alcohol 4. To 560 mg of the diol (3.3 mmol, 1.0 equiv)
in 30 mL of CH2Cl2 at 0 °C were added triethylamine (680 µL, 4.9
mmol, 1.5 equiv), BuMe2SiCl (540 mg, 3.6 mmol, 1.1 equiv), and
Photoadduct 13. A solution of 8 (99% ee, 150 mg, 0.59 mmol) in
80 mL of cyclohexane was deoxygenated by nitrogen sparge for 10
min. The flask was sealed, and the solution was irradiated for 90 min.
The resultant yellow solution was concentrated in Vacuo and purified
by chromatography on silica gel (gradient elution, 8:1 hexanes/EtOAc
to 4:1 hexanes/EtOAc) to give 71 mg of the (Z)-olefin and 55 mg of
the (E)-olefin (86%). 13Z: TLC Rf ) 0.42 (4:1 hexanes/EtOAc); [R]Na
t
4-(N,N-dimethylamino)pyridine (20 mg, 0.16 mmol, 0.05 equiv)
successively. The reaction mixture was allowed to warm slowly to 23
°C over 4.5 h and then was poured into 50 mL of H2O. The organic
layer was washed with 25 mL of 1.0 M aqueous KH2PO4, and then
with 25 mL of saturated aqueous NaCl. The combined aqueous washes
were extracted with 35 mL of Et2O. The combined organic layers were
dried over anhydrous Na2SO4 and concentrated in Vacuo. Purification
of the residue by chromatography on silica gel (6% EtOAc in hexanes)
afforded 740 mg (80%) of alcohol 4: TLC Rf ) 0.78 (2:1 hexanes/
EtOAc); [R]Na -41.5° (c ) 1.0, CHCl3); 1H NMR (CDCl3, 300 MHz)
δ 4.61 (m, 1H), 4.06-4.04 (m, 1H), 3.86-3.81 (m, 1H), 3.4 (br s,
1H), 1.99-1.97 (m, 1H), 1.88-1.87 (m, 1H), 0.90 (s, 9H), 0.17 (s,
1
) +135° (c ) 0.5, C6D6); H NMR (C6D6, 300 MHz) δ 5.53 (t, 1H,
J ) 2.4 Hz), 3.9 (m, 1H), 3.7 (m, 1H), 3.52 (t, 1H, J ) 3.2 Hz), 2.9
(m, 1H), 2.2 (dt, 1H, J ) 16.9, 4.0 Hz), 1.8-1.2 (m, 7H), 0.32 (s, 9H);
13C NMR (C6D6 , 75 MHz) δ 206, 153, 127, 86, 67, 62, 54, 40, 33, 32,
20, -0.4; IR (thin film) ν 2948, 2864, 1709, 1643. 13E: TLC Rf )
0.32 (4:1 hexanes/EtOAc); [R]Na ) +13.4° (c ) 0.1, cyclohexane); 1H
NMR (C6D6, 300 MHz) δ 5.87 (t, 1H, J ) 2 Hz), 3.8 (m, 1H), 3.6 (m,
1H), 3.53 (t, 1H, J ) 3 Hz), 2.98 (dm, 1H, J ) 9 Hz), 2.23 (dt, 1H, J

2604 J. Am. Chem. Soc., Vol. 119, No. 11, 1997
Shepard and Carreira
) 18.1, 3.6 Hz), 1.79-1.33 (m, 7H), 0.30 (s, 9H); 13C NMR (C6D6, 75
MHz) δ 205, 151, 124, 84, 67, 62, 53, 39, 33, 32, 20, -0.4; IR (thin
film) ν 2952, 2865, 1701 (s), 1637.
(c ) 0.5, cyclohexane); 1H NMR (C6D6, 300 MHz) (peaks are
broadened due to rotamers) δ 5.7 (m, 1H), 3.9 (m, 1H), 3.6 (m, 1H),
3.1 (m, 1H), 2.9 (m, 1H), 2.35 (d, 1H, J ) 18 Hz), 2.0 (m, 1H), 1.6
(m, 2H), 1.4 (m, 13H), 0.00 (s, 9H); 13C NMR (C6D6, 75 MHz) (some
peaks are broadened due to rotamers) δ 205, 154, 124, 102, 80, 77,
63, 54, 49, 39, 31, 28, 20, -0.5; IR (thin film) ν 2952, 2875, 1697 (s),
1638; HRMS (FAB+) calcd for C19H31NO3Si 349.2073, found 349.2073.
Analysis of Asymmetric Induction in Photoadduct 14. A sample
of 14E (10 mg, 30 µmol, 1 equiv) in 1 mL of MeOH at 23 °C was
treated with 2 mg (50 µmol, 2 equiv) of NaBH4. After stirring for 2
h, the mixture was quenched with 3 drops of 1 M aqueous KH2PO4
and concentrated in Vacuo. The resultant residue was purified by
chromatography on silica gel (12% EtOAc in hexanes) to afford 1 mg
of the minor alcohol epimer (Rf ) 0.49, 3:1 hexanes/EtOAc). The
alcohol (3 µmol) was dried by azeotropic removal of water with toluene
and taken up in 0.2 mL of CH2Cl2. Triethylamine (20 µL, 140 µmol,
50 equiv) was added, followed by 4-(N,N-dimethylamino)pyridine (2.0
mg, 16 µmol, 5 equiv) and a solution of (R)-MTPA-Cl (0.1 M in CH2-
Cl2, 400 µL, 40 µmol, 13 equiv).38 The reaction was stirred at 23 °C
for 40 min and then was purified by chromatography on silica gel (10%
Analysis of Asymmetric Induction in Photoadduct 13. A sample
of 13Z (1 mg, 4 µmol, 1 equiv) in 0.5 mL of MeOH at 23 °C was
treated with 2 mg (50 µmol, 50 equiv) of NaBH4. After stirring for 12
h, the mixture was quenched with 2 drops of pH 7 phosphate buffer
and concentrated in Vacuo. The resultant residue was purified by
chromatography on silica gel (4:1 hexanes/EtOAc). The epimer with
an Rf of 0.39 (4:1 hexanes/EtOAc) was isolated and subjected to GC
analysis (170 °C; retention time 18.7 min (minor), 19.4 min (major)),
indicating 92% enantiomeric excess.
A sample of 13E (1 mg, 4 µmol, 1 equiv) in 0.5 mL of MeOH at 23
°C was treated with 2 mg (50 µmol, 50 equiv) of NaBH4. After stirring
for 12 h, the mixture was quenched with 2 drops of pH 7 phosphate
buffer and concentrated in Vacuo. The resultant residue was purified
by chromatography on silica gel (4:1 hexanes/EtOAc). The epimer
with an Rf of 0.17 (4:1 hexanes/EtOAc) was isolated and subjected to
GC analysis (170 °C; retention time 25.8 min (minor), 27.2 min
(major)), indicating 86% enantiomeric excess.
1
EtOAc in hexanes). Integration of the H NMR resonances (CDCl3,
500 MHz) at δ 0.23 ppm (major) and δ 0.08 ppm (minor) indicated
85% enantiomeric excess.
exo-Methylenecyclobutane 18. To a mixture of olefin isomers 13
(8.0 mg, 32 µmol, 1.0 equiv) in 0.5 mL of THF was added 0.35 mL of
TBAF/AcOH (0.1 M/0.05 M in THF, 35 µmol/17 µmol, 1.1/0.55 equiv).
The reaction mixture was stirred at 23 °C for 60 min and then was
poured into pH 7 phosphate buffer. The aqueous layer was extracted
3× with Et2O, and the combined organic layers were washed with
saturated aqueous NaCl, dried over anhydrous Na2SO4, and concentrated
in Vacuo. The residue was purified by chromatography on silica gel
(2% MeOH in 9:1 hexanes/EtOAc) to afford 4.0 mg (70%) of terminal
olefin 18: TLC Rf ) 0.47 (2:1 hexanes/EtOAc); [R]Na ) +7.8° (c )
exo-Methylenecyclobutane 19. To vinylsilane 14 (7 mg, 20 µmol,
1.0 equiv) in 0.2 mL of THF was added 240 µL of TBAF/AcOH (0.1
M/5 µM in THF, 24 µmol/1.2 µmol, 1.2/0.06 equiv). The reaction
mixture was stirred at 23 °C for 19 h and then was purified by
chromatography on silica gel (15% EtOAc in hexanes) to afford 4 mg
(67%) of terminal olefin 19: TLC Rf ) 0.42 (3:1 hexanes/EtOAc);
[R]Na +104° (c ) 0.5, cyclohexane); 1H NMR (C6D6, 300 MHz) (peaks
are broadened due to rotamers) δ 5.3-5.2 (m, 1H), 4.8 (br s, 1H), 3.7
(m, 1H), 3.6-3.5 (m, 1H), 3.3 (m, 1H), 2.7 (m, 1H), 2.3 (m, 1H), 2.0
(m, 1H), 1.5-1.2 (m, 15H); 13C NMR (C6D6, 75 MHz) (some peaks
are broadened due to rotamers) δ 110, 59, 54, 51, 48, 40, 30, 29, 20,
14; IR (thin film) ν 3082 (w), 2929, 2875, 1695 (s); HRMS (FAB+)
1
0.05, cyclohexane); H NMR (C6D6, 300 MHz) δ 5.28 (m, 1H), 4.79
(m, 1H), 3.8 (m, 1H), 3.6 (m, 1H), 3.40 (t, 1H, J ) 3.0 Hz), 2.8 (m,
1H), 2.22 (dm, 1H, J ) 17.9 Hz), 1.75-1.27 (m, 7H); IR (thin film)
ν 3075 (w), 2943, 2862, 1702 (s), 1669.
Photosubstrate 9. A 60% dispersion of NaH in mineral oil (10
mg, 0.26 mmol, 1.2 equiv) was weighed into a flask, washed with
hexanes, and dried under vacuum. An atmosphere of nitrogen was
restored, 1.5 mL of DMF was added, and the mixture was cooled to 0
°C. The Boc-protected 3-amino-2-cyclohexenone (prepared from the
known enaminone37) (54 mg, 0.26 mmol, 1.2 equiv) was added, and
stirring was continued at 0 °C for 15 min and then at 23 °C for 90
min. A solution of the allenyl mesylate (50 mg, 0.21 mmol, 1.0 equiv)
in 0.5 mL of DMF was added Via cannula. The reaction was heated
to 100 °C for 5 h, cooled to 23 °C, and poured into H2O, saturated
aqueous NaCl, and Et2O. The aqueous layer was extracted with Et2O,
and the combined organic layers were washed with saturated aqueous
NaCl, dried over anhydrous Na2SO4, and concentrated in Vacuo. The
residue was purified by chromatography on silica gel (14% EtOAc in
hexanes) to afford 14 mg (18%) of the desired product 9: TLC Rf )
+
calcd for C16H24NO3 287.1761, found 278.1756.
Photosubstrate 10. To a solution of 6 (16 mg, 0.10 mmol, 1.0
equiv) in 1 mL of THF was added 13 mg of 1,3-cyclopentanedione
(0.13 mmol, 1.3 equiv), 34 mg of triphenylphosphine (0.13 mmol, 1.3
equiv), and 20 µL of diethyl azodicarboxylate (0.13 mmol, 1.3 equiv).
The mixture was stirred at 23 °C for 30 min and then was concentrated
in Vacuo and purified by chromatography on silica gel (3:2 Et2O/pentane
and then 2:1 Et2O/pentane) to afford 21 mg (88%) of the desired product
10: TLC Rf ) 0.08 (4:1 hexanes/EtOAc); [R]Na +70° (c ) 1.0, CHCl3);
1H NMR (CDCl3, 300 MHz) δ 5.29 (s, 1H), 4.98 (dt, 1H, J ) 6.6, 3.7
Hz), 4.79 (q, 1H, J ) 6.8 Hz), 3.99 (t, 2H, J ) 6.7 Hz), 2.6 (m, 2H),
2.4 (m, 4H), 0.08 (s, 9H); 13C NMR (CDCl3, 75 MHz) δ 210, 206,
190, 105, 84, 79, 71, 34, 29, 27, -1.0; IR (thin film) ν 2954, 1939,
1705, 1682, 1593 (s); HRMS (FAB+) calcd for C13H21O2Si+ 237.1311,
found 237.1310.
1
0.22 (3:1 hexanes/EtOAc); [R]Na +102° (c ) 0.5, CHCl3); H NMR
Photoadduct 15. A solution of 10 (99% ee, 15 mg, 60 µmol) in 15
mL of cyclohexane was deoxygenated by nitrogen sparge for 10 min.
The flask was sealed, and the mixture was irradiated for 30 min. The
resultant yellow solution was concentrated in Vacuo and purified by
chromatography on silica gel (3:1 hexanes/EtOAc) to give 14 mg of
an 11:7 mixture of (Z)- and (E)-olefins, respectively (67%). 15Z: TLC
Rf ) 0.68 (2:1 hexanes/EtOAc); [R]Na +124° (c ) 0.5, cyclohexane);
1H NMR (C6D6, 300 MHz) δ 5.45 (t, 1H, J ) 2 Hz), 3.86 (t, 1H, J )
8 Hz), 3.6 (m, 1H), 3.25 (br s, 1H), 2.91 (br d, 1H, J ) 8 Hz), 2.1 (m,
2H), 1.96 (m, 1H), 1.7 (m, 1H), 1.4-1.6 (m, 2H), 0.27 (s, 9H); 13C
NMR (C6D6, 75 MHz) δ 211, 151, 126, 88, 69, 60, 55, 38, 33, 31,
-0.6; IR (thin film) ν 2951, 2864, 1738 (s), 1642; HRMS (FAB+)
calcd for C13H21O2Si+ (MH+) 237.1311, found 237.1308.
(CDCl3, 300 MHz) δ 5.72 (s, 1H), 4.94 (m, 1H), 4.71 (q, 1H, J ) 6.9
Hz), 3.60 (m, 2H), 2.72 (t, 2H, J ) 6.0 Hz), 2.38 (t, 2H, J ) 6.6 Hz),
2.23 (m, 2H), 1.98 (m, 2H), 1.50 (s, 9H), 0.09 (s, 9H); 13C NMR
(CDCl3, 75 MHz) δ 210, 200, 163, 153, 115, 83, 82, 80, 49, 37, 31,
28, 27, 23, -1.0; IR (thin film) ν 2956, 1938, 1715, 1665; HRMS
(FAB+) calcd for C19H32NO3Si+ (MH+) 350.2151, found 350.2158.
Photoadduct 14. A solution of 9 (99% ee, 26 mg, 74 µmol) in 15
mL of cyclohexane was deoxygenated by nitrogen sparge for 5 min.
The flask was sealed, and the mixture was irradiated for 5 min. The
resultant yellow solution was concentrated in Vacuo and purified by
chromatography on silica gel (12% EtOAc in hexanes) to give 7 mg
of the (Z)-olefin and 13 mg of the (E)-olefin (77%). 14Z: TLC Rf )
1
0.62 (3:1 hexanes/EtOAc); [R]Na +250° (c ) 0.29, cyclohexane); H
Analysis of Asymmetric Induction in Photoadduct 15. The
mixture of 15Z and 15E was subjected to GC analysis (150 °C; retention
time 24.8 min (minor Z), 25.3 min (major Z), 34.3 min (minor E),
34.8 min (major E)), which indicated 76% enantiomeric excess for the
(Z)-olefin and 78% enantiomeric excess for the (E)-olefin.
NMR (C6D6, 300 MHz) (peaks are broadened due to rotamers) δ 5.5
(br s, 1H), 3.9-3.4 (m, 3H), 2.9 (m, 1H), 2.6 (m, 1H), 2.3-2.0 (m,
3H), 1.4 (m, 13H), 0.32 (s, 9H); 13C NMR (C6D6, 75 MHz) (some
peaks are broadened due to rotamers) δ 206, 176, 125, 79, 61, 60, 54,
53, 48.3, 47.7, 41, 30, 28.9, 28.5, 27, 22, 21, -0.4; IR (thin film) ν
2951, 1694 (s); HRMS (FAB+) calcd for C19H31NO3Si 349.2073, found
349.2072. 14E: TLC Rf ) 0.55 (3:1 hexanes/EtOAc); [R]Na +150°
Photosubstrate 11. To a solution of alcohol 6 (31 mg, 0.20 mmol,
1.0 equiv) in 2 mL of THF was added 43 mg of 4-hydroxycoumarin
(37) Baraldi, P. G.; Simoni, D.; Manfredini, S. Synthesis 1983, 902.
(38) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512.

Asymmetric Photocycloadditions with an Allenylsilane
J. Am. Chem. Soc., Vol. 119, No. 11, 1997 2605
(0.27 mmol, 1.3 equiv), 70 mg of triphenylphosphine (0.27 mmol, 1.3
equiv), and 45 µL of diethyl azodicarboxylate (0.27 mmol, 1.3 equiv).
The mixture was stirred at 23 °C for 30 min, concentrated in Vacuo,
and purified by chromatography on silica gel (6:1 hexanes/EtOAc) to
afford 43 mg (72%) of white crystalline product 11: mp 54-58 °C
(hexanes/EtOAc); TLC Rf ) 0.31 (3:1 hexanes/EtOAc); [R]Na +87° (c
by flash chromatography on silica gel (7% EtOAc in hexanes) to yield
35 mg (69%) of 12: TLC Rf ) 0.59 (3:1 pentane/Et2O); [R]Na +94° (c
1
) 1.0, CHCl3); H NMR (CDCl3, 300 MHz) δ 7.80-7.71 (m, 1H),
7.55-7.50 (m, 1H), 7.34-7.23 (m, 2H), 6.14 (s, 1H), 5.06 (m, 1H),
4.87 (q, 1H, J ) 6.6 Hz), 3.07 (t, 2H, J ) 7.3 Hz), 2.49-2.41 (m,
2H), 0.11(s, 9H); 13C NMR (CDCl3, 75 MHz) δ 209, 159, 156, 152,
132, 124.0, 123.8, 118, 117, 107, 85, 81, 31, 26, -0.9; IR (thin film)
ν 3072 (w), 2956, 1938, 1718 (s); HRMS (FAB+) calcd for C17H21O2-
SSi+ (MH+) 317.1032, found 317.1028
1
) 1.0, CHCl3); H NMR (CDCl3, 300 MHz) δ 7.85-7.82 (m, 1H),
7.58-7.52 (m, 1H), 7.34-7.24 (m, 2H), 5.67 (s, 1H), 5.0 (m, 1H),
4.88 (q, 1H, J ) 6.8 Hz), 4.17 (t, 2H, J ) 6.6 Hz), 2.61-2.53 (m,
2H), 0.09 (s, 9H); 13C NMR (CDCl3, 75 MHz) δ 210, 166, 163, 153,
132, 124, 123, 117, 116, 90, 84, 79, 69, 27, -1.0; IR (thin film) ν
3041, 2956, 1940, 1743 (s), 1610; HRMS (FAB+) calcd for C17H21O3-
Si+ 301.1247, found 301.1260.
Photoadduct 16. A solution of 11 (99% ee, 20 mg, 67 µmol) in 20
mL of cyclohexane was deoxygenated by nitrogen sparge for 10 min.
The flask was sealed, and the mixture was irradiated for 120 min. The
resultant yellow solution was concentrated in Vacuo and purified by
chromatography on silica gel (9:1 hexanes/EtOAc) to give 18 mg (90%)
of the (E)-olefin 16 only: TLC Rf ) 0.58 (3:1 hexanes/EtOAc); [R]Na
+157° (c ) 0.5, CHCl3); 1H NMR (C6D6, 300 MHz) δ 7.34-7.31 (m,
1H), 6.87-6.78 (m, 3H), 6.16 (t, 1H, J ) 2.5 Hz), 3.92 (td, 1H, J )
7.8, 1.4 Hz), 3.85 (t, 1H, J ) 3.2 Hz), 3.64-3.55 (m, 1H), 3.27 (dm,
1H, J ) 8.9 Hz), 1.78-1.64 (m, 1H), 1.52 (ddd, 1H, J ) 12.5, 5.8, 1.5
Hz), -0.10 (s, 9H); 13C NMR (C6D6 , 75 MHz) δ 163, 151, 148, 130,
127.5, 127.1, 125, 122, 118, 80, 68, 60, 53, 33, -0.6; IR (thin film) ν
3072 (w), 3041 (w), 2954, 2864, 1757 (s), 1646, 1618; HRMS (FAB+)
calcd for C17H20O3Si 300.1182, found 300.1196.
Analysis of Asymmetric Induction in Photoadduct 16. A sample
of 16 (7 mg, 20 µmol, 1 equiv) in 0.3 mL of THF at 23 °C was treated
with 30 µL (60 µmol, 3 equiv) of 2.0 M LiBH4 in THF. After stirring
at 23 °C for 12 h, the reaction was quenched cautiously with 1 M
aqueous KH2PO4 and concentrated in Vacuo. The residue was purified
by chromatography on silica gel (4:1 hexanes/EtOAc) to afford 6 mg
of the diol (Rf ) 0.33, 3:1 hexanes/EtOAc). The alcohol (10 µmol)
was dried by azeotropic removal of water with toluene and dissolved
in 0.2 mL of CH2Cl2. Triethylamine (10 µL, 70 µmol, 7 equiv) was
added, followed by 4-(N,N-dimethylamino)pyridine (2 mg, 20 µmol, 2
equiv) and a solution of (R)-MTPA-Cl (0.1 M in CH2Cl2, 200 µL, 20
µmol, 2 equiv). The reaction mixture was stirred at 23 °C for 25 min
and was purified by chromatography on silica gel (15% EtOAc in
hexanes). Integration of the 19F NMR resonances (CDCl3, 470 MHz)
at δ -71.03 and -71.20 ppm (major) indicated 99% enantiomeric
excess (minor diastereomer peaks at δ -70.96 and -71.22 ppm were
not detected).
Photoadduct 17. A solution of 12 (99% ee, 4.5 mg, 14 µmol, 1
equiv) in 5 mL of cyclohexane was deoxygenated by nitrogen sparge
for 5 min. The reaction vessel was sealed, and the solution was
irradiated for 25 min. After concentration in Vacuo the residue was
purified by chromatography on silica gel (3% EtOAc in hexanes) to
afford 3.6 mg (80%) (E)-olefin 17: TLC Rf ) 0.64 (3:1 hexanes/
1
EtOAc); [R]Na +121° (c ) 1.0, cyclohexane); H NMR (C6D6, 300
MHz) δ 7.56 (dd, 1H, J ) 5.6, 2.5 Hz), 6.83-6.75 (m, 3H), 6.05 (t,
1H, J ) 2.5 Hz), 4.10 (t, 1H, J ) 3.4 Hz), 3.53-3.48 (m, 1H), 2.83-
2.74 (m, 1H), 2.67-2.61 (m, 1H), 1.89-1.82 (m, 2H), -0.15 (s, 9H);
13C NMR (C6D6, 75 MHz) δ 163, 151, 149, 130, 129, 127, 125, 123,
118, 67, 57, 56, 39, 35, -0.6; IR (thin film) ν 3061 (w), 2952, 1758
(s), 1643; HRMS (FAB+) calcd for C17H21O2SSi+ (MH+) 317.1032,
found 317.1022.
Analysis of Asymmetric Induction in Photoadduct 17. A sample
of 17 (3.6 mg, 11 µmol, 1 equiv) in 0.5 mL of THF at 23 °C was
treated with 30 µL (60 µmol, 5 equiv) of 2.0 M LiBH4 in THF. After
stirring at 23 °C for 1 h, the reaction was quenched cautiously with 3
drops of 1 M aqueous KH2PO4. The mixture was purified by
chromatography on silica gel (3:1 hexanes/EtOAc) to afford 2 mg of
the diol (Rf ) 0.22, 3:1 hexanes/EtOAc). The alcohol (6 µmol) was
dried by azeotropic removal of water with toluene and dissolved in
0.2 mL of CH2Cl2. Triethylamine (20 µL, 140 µmol, 23 equiv) was
added, followed by 4-(N,N-dimethylamino)pyridine (2.0 mg, 16 µmol,
3 equiv) and a solution of (R)-MTPA-Cl (0.1 M in CH2Cl2, 300 µL,
30 µmol, 5 equiv). The mixture was stirred at 23 °C for 30 min and
then was purified by chromatography on silica gel (9:1 hexanes/EtOAc).
Integration of the 19F NMR resonances (CDCl3, 470 MHz) at δ -70.32
and -71.05 ppm (minor) and δ -70.45 and -70.95 ppm (major)
indicated 89% enantiomeric excess.
exo-Methylenecyclobutane 21. To 17 (7 mg, 20 µmol, 1.0 equiv)
in 0.4 mL of THF was added 240 µL of TBAF/AcOH (0.1 M/0.15 M
in THF, 24 µmol/36 µmol, 1.1/1.6 equiv). The mixture was stirred at
23 °C for 20 min and then was purified by chromatography on silica
gel (5% EtOAc in hexanes) to afford 3 mg (60%) of terminal olefin
21: TLC Rf ) 0.60 (3:1 hexanes/EtOAc); [R]Na +45° (c ) 0.25,
exo-Methylenecyclobutane 20. To vinylsilane 16 (7 mg, 20 µmol,
1.0 equiv) in 0.6 mL of THF was added an excess of TBAF/AcOH
(0.1 M/0.15 M in THF). The reaction mixture was stirred at 23 °C for
15 min and then was subjected to chromatography on silica gel (15%
EtOAc in hexanes) to afford 3 mg (60%) of the desired terminal
olefin: TLC Rf ) 0.43 (25% EtOAc in hexanes); [R]Na +158° (c )
1
cyclohexane); H NMR (C6D6, 300 MHz) δ 7.54-7.50 (m, 1H), 6.8
(m, 3H), 5.30 (m, 1H), 4.57 (m, 1H), 3.93 (q, 1H, J ) 3.2 Hz), 3.06
(m, 1H), 2.82-2.72 (m, 1H), 2.58-2.51 (m, 1H), 1.61-1.54 (m, 2H);
13C NMR (C6D6, 75 MHz) δ 163, 151, 143, 130, 129, 125, 123, 118,
112, 64, 57, 55, 37, 35; IR (thin film) ν 3071 (w), 2925, 2854, 1756
(s), 1675; HRMS (FAB+) calcd for C14H13O2S+ (MH+) 245.0636, found
245.0635.
1
0.5, cyclohexane); H NMR (C6D6, 300 MHz) δ 7.26-7.23 (m, 1H),
6.88-6.79 (m, 3H), 5.39 (m, 1H), 4.73 (t, 1H, J ) 2 Hz), 3.88 (t, 1H,
J ) 8.4 Hz), 3.69 (t, 1H, J ) 3.1 Hz), 3.60-3.51 (m, 1H), 2.89 (dm,
1H, J ) 8 Hz), 1.60-1.46 (m, 1H), 1.32 (dd, 1H, J ) 12.5, 5.5 Hz);
13C NMR (C6D6 , 75 MHz) δ 164, 151, 142, 130, 127, 125, 122, 117,
112, 80, 68, 58, 50, 32; IR (thin film) ν 3072, 2947, 2863, 1755 (s),
1675, 1612; HRMS (EI+) calcd for C14H12O3 228.0786, found 228.0779.
Photosubstrate 12. A solution of 6 (25 mg, 0.16 mmol, 1.0 equiv)
in 2.0 mL of THF was transferred Via cannula to 4-mercaptocoumarin39
(160 mg, 0.87 mmol, 5 equiv). Triphenylphosphine (55 mg, 0.21 mmol,
1.3 equiv) was added, followed by 33 µL (0.21 mmol, 1.3 equiv) of
diethyl azodicarboxylate. The resultant mixture was stirred at 23 °C
for 3 h and then was concentrated in Vacuo. The residue was purified
Acknowledgment. We are grateful to Dr. Wheeseong Lee
and Mr. Vincent Tai for carrying out some preliminary
investigations on the protodesilylation of 16. M.S.S. is grateful
to the Office of Naval Research for a graduate fellowship. We
are grateful to Mr. Travis Williams (Caltech Summer Under-
graduate Research FellowsSURF) for assistance in the scale-
up of the allene synthesis. This research has been generously
supported by an award from the Packard Foundation, Camille
and Henry Dreyfus Foundation, and Sloan Foundation, funds
provided by the NIH and NSF, and gifts from Eli Lilly, Merck,
Pfizer, Upjohn, and Zeneca.
(39) (a) Vishnyakova, G. M.; Smirnov, T. V. Zh. Vses. Khim. OVa. im.
D. I. MendeleeVa 1978, 23, 235. (b) Eiden, F.; Zimmerman, E. Arch. Pharm.
(Weinheim, Ger.) 1976, 309, 619. (c) Peinhardt, G.; Reppel, L. Pharmazie
1970, 25, 68.
JA963436G