A general synthetic route to differentially functionalized angularly and linearly fused [6-7-5] ring systems: A Rh(I)-catalyzed cyclocarbonylation reaction

Article abstract of DOI:10.1021/jo8007258

(Figure Presented) Investigations of a Rh(I)-catalyzed cyclocarbonylation reaction reveal its general synthetic utility for accessing highly functionalized tricyclic [6-7-5] linear and angular ring systems from allene-ynes. Three types of allene-ynes were

Full text of DOI:10.1021/jo8007258

A General Synthetic Route to Differentially Functionalized
Angularly and Linearly Fused [6-7-5] Ring Systems: A
Rh(I)-Catalyzed Cyclocarbonylation Reaction
Kay M. Brummond,* Daitao Chen, and Matthew M. Davis
Department of Chemistry, UniVersity of Pittsburgh, Pittsburgh, PennsylVania 15260
kbrummon@pitt.edu
ReceiVed April 1, 2008
Investigations of a Rh(I)-catalyzed cyclocarbonylation reaction reveal its general synthetic utility for
accessing highly functionalized tricyclic [6-7-5] linear and angular ring systems from allene-ynes.
Three types of allene-ynes were prepared and subjected to Rh(I)-catalyzed cyclocarbonylation conditions.
For three series of allene-ynes, the [6-7-5] ring systems were afforded in varying yields depending on
the substrate structure. One series of allene-ynes afforded the [6-6-5] ring system possessing an
R-alkylidene cyclopentenone as a result of a selective reaction with the proximal double bond of the
allene.
Introduction
accomplished, it has not been without difficulty.4 Accessing the
latter skeleton via this strategy has scarcely been accomplished,
but not for lack of trying.4c,d
Since the Stork-Eschenmoser hypothesis of polyolefinic
cyclizations,1 rapid and selective assembly of polycyclic com-
pounds from unsaturated precursors has intrigued the synthetic
community.2 Transition-metal catalysis has substantially in-
creased the substrate scope of polyolefinic cyclizations; however,
most are dependent upon selectivity factors imparted by both
the substrate and the reagent. Selectivity achieved through only
catalyst-based control elements has the highest potential for
synthetic utility. Recently, our group has demonstrated that in
nearly all intramolecular Rh(I)-catalyzed allenyl cyclocarbony-
lations and carbocyclizations examined to date the reactions
occur selectively with the distal double bond of an allene.3
Interestingly, this regioselective reaction affords bicyclo[4.3.0]
nonadienones and bicyclo[5.3.0]decadienones. While obtaining
the former skeleton via cyclocarbonylation chemistry has been
We5 and others6 have continued to explore the scope and
limitations of the Rh(I)-catalyzed cyclocarbonylation reaction
to form bicyclo[5.3.0]decadienones. Initially, we demonstrated
the synthetic potential of this reaction by rapidly assembling
the carbocyclic core of guanacastepene A.7 By varying the
position of the allene and alkyne on a cycloalkane scaffold (e.g.,
cyclohexane, A-D, Scheme 1), carbocycles are produced that
possess a unique positioning of the resulting dienone functional-
(3) (a) Bayden, A. S.; Brummond, K. M.; Jordan, K. D. Organometallics
2006, 25, 5204–5206. (b) Brummond, K. M.; Mitasev, B. Org. Lett. 2004, 6,
2245–2248. (c) Brummond, K. M.; Chen, H.; Fisher, K. D.; Kerekes, A. D.;
Rickards, B.; Sill, P. C.; Geib, S. J Org. Lett. 2002, 4, 1931–1934. Similar
observations were reported independently: (d) Mukai, C.; Inagaki, F.; Yoshida,
T.; Yoshitani, K.; Hara, Y.; Kitagaki, S. J. Org. Chem. 2005, 70, 7159–7171.
(e) Mukai, C.; Nomura, I.; Kitagaki, S. J. Org. Chem. 2003, 68, 1376–1385. (f)
Mukai, C.; Nomura, I.; Yamanishi, K.; Hanaoka, M. Org. Lett. 2002, 4, 1755–
1758.
(1) (a) Stadler, P. A.; Eschenmoser, A.; Shinz, H.; Stork, G. HelV. Chim.
Acta 1957, 40, 2191–2198. (b) Stork, G.; Burgstahler, A. W. J. Am. Chem. Soc.
1955, 77, 5068–5077. (c) Eschenmoser, A.; Ruzicka, L.; Jeger, O.; Arigoni, D.
HelV. Chim. Acta 1955, 38, 1890–1904.
(4) (a) Bon˜aga, L. V. R.; Krafft, M. Tetrahedron 2004, 60, 9795–9833, and
references therein. (b) Lovely, C. J.; Seshadri, H.; Wayland, B. R.; Cordes, A. W.
Org. Lett. 2001, 3, 2607–2610. (c) Ahmar, M.; Locatelli, C.; Colombier, D.;
Cazes, B. Tetrahedron Lett. 1997, 38, 5281–5284. (d) Shibata, T.; Koga, Y.;
Narasaka, K. Bull. Chem. Soc. Jpn. 1995, 68, 911–919.
(2) (a) For a review of metal-catalyzed cascade reactions, see: Malacria, M.
Chem. ReV. 1996, 96, 286–306. (b) For a recent review on transition metal-
catalyzed Alder-ene reactions, see: (c) Brummond, K. M.; Loyer-Drew, J. A.
C-C Bond Formation (Part 1) by Addition Reactions: Alder-ene Reaction. In
ComprehensiVe Organometallic Chemistry, 3rd ed.; Crabtree, R., Mungos, M.,
Eds.; Elsevier Science: New York, 2006; Vol. 10, pp 557-601.
(5) Brummond, K. M.; Chen, D. Org. Lett. 2008, 10, 705–708.
(6) Hirose, T.; Miyakoshi, N.; Mukai, C. J. Org. Chem. 2008, 73, 1061–
1066.
(7) Brummond, K. M.; Gao, D. Org. Lett. 2003, 5, 3491–3494.
5064 J. Org. Chem. 2008, 73, 5064–5068
10.1021/jo8007258 CCC: $40.75 2008 American Chemical Society
Published on Web 06/04/2008

Rh(I)-Catalyzed Cyclocarbonylation Reaction
SCHEME 1. Angularly and Linearly Fused [6-7-5] Ring Systems
SCHEME 2. Synthesis of Allene-Ynes 5a-k
ity relative to the six-membered ring (A′-D′, Scheme 1). Rapid
access to the bicyclo[5.3.0]decane ring system, in addition to
the differential positioning of the dienone, provides entry into
an abundance of biologically relevant compounds with function-
ally dense, angularly and linearly fused [6-7-5] ring systems
(e.g., rippertene, grayanotoxin III, resiniferatoxin, and guana-
castepene A, Scheme 1). The results of this scope and limitation
study of a Rh(I)-catalyzed cyclocarbonylation reaction of
allene-ynes of types A-C to form bicyclo[5.3.0]decadienones
are reported within.
Results and Discussion
Preparation of Allene-Ynes Types A-C. Preparation of
cyclocarbonylation precursors began with the synthesis of
allene-yne type A (Schemes 1). Two methods were imple-
mented for their formation. First, the Stork-Danheiser viny-
logous ester methodology8 was used; alkylation of the substi-
tuted 3-alkoxy-2-cyclohexenones 1a,b with allenes 2a-c9
afforded the corresponding enones 3a-d in 26-63% yield
(Scheme 2.) These low to moderate yields are attributed to a
competing E2 elimination reaction involving the homoallylic
iodide of the allene. Because allenes 2a-c are the most valuable
component in these reactions, we were reluctant to increase the
allene equivalents. Enone 3d was formed as a 1:1 mixture of
diastereomers which were not separated and carried on as a
mixture. Addition of acetylides 4a-d to enones 3a-d, followed
by acidic hydrolysis afforded allene-ynes 5a-k in 34-96%
yield.
An alternate route to allene-yne type A began by alkylating
ethyl 2-oxocyclohexanecarboxylate (6) with 5-iodopenta-1,2-
diene (2a) utilizing K2CO3 in refluxing acetone, affording 7 in
60% yield (Scheme 3). Subsequently, addition of ethynylmag-
nesium bromide (4a) or propynylmagnesium bromide (4b) to
the ketone gave allene-yne 8a and 8b in 89% and 97% yield,
respectively. Interestingly, 8a was obtained as an inseparable
1
mixture of diastereomers (2:1 based upon H NMR), and 8b
was obtained as a single diastereomer. Addition of the lithium
trimethylsilylacetylide (4d) to ketone 7 gave allene-yne 8c in
70% yield as a 2:1 mixture of diastereomers (based upon H
NMR). The diastereomers were separated by column chroma-
tography and taken on independently. Dehydration of allene-yne
8b with phosphorus oxychloride gave allene-yne 8d in 55%
yield.10
1
(8) Stork, G.; Danheiser, R. L. J. Org. Chem. 1973, 38, 1775–1776.
(9) Homoallenic iodides were prepared by a modified four-step reaction
sequence on multigram scale and were freshly purified via flash chromatography
prior to the alkylation reaction. Dauben, W. G.; Shapiro, G. J. Org. Chem. 1984,
49, 4252–4258.
(10) Trost, B. M.; Sorum, M. T.; Chan, C.; Harms, A. E.; Ru¨hter, G. J. Am.
Chem. Soc. 1997, 119, 698–708.
J. Org. Chem. Vol. 73, No. 13, 2008 5065

Brummond et al.
SCHEME 5. Formation of Allene-Ynes 15a,b
SCHEME 3. Formation of Allene-Ynes 8a-d
SCHEME 6. Formation of Allene-Ynes 15c-e
SCHEME 4. Formation of Allene-Ynes 11a-e
removed from 11d with tetra-n-butylammonium fluoride to
provide the terminal alkynone 11e in 87% yield.
Allene-ynes type C were prepared as follows. Alkylation
of the N,N-dimethylhydrazone of 1,4-cyclohexanedione mono-
ethylene ketal (13) with (5-iodopent-1-ynyl)trimethylsilane13
gave 14 following a selective hydrolysis of the hydrazone in
the presence of the ketal with oxalic acid (Scheme 5).14
Hydrazone formation was necessary due to the low yields
obtained of the monoalkylation product when direct alkylations
of 1,4-cyclohexanedione monoethylene ketal were attempted
using a variety of conditions (LDA, THF; KHMDS, THF or
1:1 toluene/DMF; KH, THF). Addition of ethynylmagnesium
bromide (4a) or propynylmagnesium bromide (4b) followed by
in situ acetylation using acetyl chloride furnished the corre-
sponding propargyl acetates, which upon SN2′ addition of a
dialkyl cuprate gave allene-ynes 15a and 15b in 55% and 40%
yield, respectively, for three steps. Allene-yne 15a was obtained
as a 3:1 mixture of diastereomers.
Three protocols were examined for the formation of the 1,1-
disubstituted allene 15c (Scheme 6). Reaction of propargyl
acetate 16 (R1 ) Ac) with Stryker’s reagent ([(PPh3)CuH]6)15
gave allene-yne 15c in 70% yield. Treatment of propargyl
carbonate 16 (R1 ) CO2Me) to Tsuji’s Pd(0)-catalyzed hydro-
genolysis conditions16 gave allene-yne 15c in 82% yield.
Reaction of the propargyl alcohol 16 (R1 ) H) to AlH3
(generated in situ from LiAlH4 and AlCl3)17 gave 15c in 68%
yield. While the Tsuji protocol was highest yielding, the direct
Preparation of allene-yne precursors type B began by
alkylating the lithium enolate of 3-ethoxy-2-cyclohexenonone
(9) with propargyl bromide, followed by a Crabbe´ homologation
to convert the alkyne to terminal allene 10a (Scheme 4.)11
Alternatively, 10b was prepared by direct alkylation of 3-ethox-
ycyclohexenon-2-one (9) with 1-bromopenta-2,3-diene. With
10a and 10b in hand, the alkyne moiety was installed by addition
of the lithiate of 1-(trimethylsilyl)-1-propyne at -78 °C, giving
11a and 11b in 50% and 66% yield, respectively, after
hydrolysis with 1 N HCl.12 Addition of the lithium anion of
propargyl ether 12 to 10a followed by hydrolysis afforded 11c
in 32% yield. Oxidation of 11c with Jones’ reagent gave the
alkynone 11d in 72% yield. The trimethylsilyl group was
(13) Bra¨se, S.; Wertal, H.; Frank, D.; Vidovic´, D.; de Meijere, A. Eur. J.
Org. Chem. 2005, 19, 4167–4178.
(11) Searles, S.; Li, Y.; Nassim, B.; Lopes, M-T. R.; Tran, P. T.; Crabbe´, P.
J. Chem. Soc., Perkin Trans. 1 1984, 747–751.
(14) Enders, D.; Nu¨hring, A.; Runsink, J. Chirality 2000, 12, 374–377.
(15) Daeuble, J. F.; McGettigan, C.; Stryker, J. M. Tetrahedron Lett. 1990,
31, 2397–2400.
(12) Cramer, N.; Buchweitz, M.; Laschat, S.; Frey, W.; Baro, A.; Mathieu,
D.; Richter, C.; Schwalbe, H. Chem. Eur. J. 2006, 12, 2488–2503.
(16) Tsuji, J.; Sugiura, T.; Minami, I. Synthesis 1987, 603–606.
5066 J. Org. Chem. Vol. 73, No. 13, 2008

Rh(I)-Catalyzed Cyclocarbonylation Reaction
TABLE 1. Cyclocarbonylation Reactions of Allene-Ynes 5a-k
Affording 17a-k
TABLE 2. Rh(I)-Catalyzed Cyclocarbonylation Reaction of 8a-d
Affording 18a-c and 19
yield
entry
reaction
time
dr
(%)
entry
reaction
time
yield (%)
a
b
c
d
e
f
g
h
i
5a: R1 ) R2 ) R3 ) R4 ) H
100 min
130 min
40 min
35 min
30 min
70 min
20 h
85
67
71
64
67
56
58
62
62
78
57
a
b
c
d
8a f 18a: R ) H
8b f 18b: R ) Me
8c f 18c: R ) TMS
8d f 19: R ) Me
190 min
40 min
23 h
32
37
13
29
5b: R1 ) R2 ) R4 ) H, R3 ) Me
5c: R1 ) R2 ) R4 ) H, R3 ) Ph
5d: R1 ) R3 ) R4 ) H, R2 ) Me
5e: R1 ) R4 ) H, R2 ) R3 ) Me
5f: R1 ) R2 ) R3 ) H, R4 ) Me
5g: R1 ) R2 ) H, R3 ) R4 ) Me
5h: R1 ) Me, R2 ) R4 ) R3 ) H
5i: R1 ) R3 ) Me, R2 ) R4 ) H
160 min
TABLE 3. Rh(I)-Catalyzed Cyclocarbonylation Reaction of
Allene-Ynes 11a-eAffording 20a-d
30 min
100 min 2:1
1.5:1
j
k
5j: R1 ) Me, R2 ) R4 ) H, R3 ) Ph 120 min 1:1
5k: R1 ) R2 ) R4 ) H, R3 ) TMS
9.5 h
conversion of the propargylic alcohol using AlH3 was more
convenient for accessing 15c, due the ease of substrate prepara-
tion and the relative simplicity of the reaction and subsequent
purification. Removal of the trimethylsilyl group from the alkyne
of 15c (R2 ) TMS) was accomplished using K2CO3 in methanol
to give allene-yne 15d (R2 ) H) in 86% yield. The alkyne
terminus was methylated using LDA and MeI to give allene-yne
15e (R2 ) Me) in 64% yield.
entry
reaction
time (min) yield (%)
a
b
c
d
e
11a: R1 ) R2 ) R3 ) H, R4 ) TMS
11b: R1 ) Me, R2 ) R3 ) H, R3 ) TMS
11c: R1 ) R2 ) H, R3 ) OH, R4 ) TMS
11d: R1) H, R2 ) R3 ) O, R4 ) TMS
11e: R1 ) H, R2 ) R3 ) O, R4 ) H
120
180
50
48
68
84
10
76
a
a Decomposition was observed.
diastereomer of 8c afforded the cycloadduct 18c in 13% yield,
and the minor diastereomer gave only trace quantities of 18c
Cyclocarbonylation of Allene-Ynes Types A-C. The
allene-ynes types A-C were next subjected to the Rh(I)-
cyclocarbonylation reaction conditions. For each series of
allene-ynes, the reaction conditions were not varied, in an effort
to examine the effects of substrate functionality and substitution.
Allene-ynes 5a-k were reacted with 5 mol % of rhodium
biscarbonyl chloride dimer [Rh(CO)2Cl]2 in toluene at 90 °C
under a balloon of carbon monoxide. Cycloadducts 17a-j were
obtained in yields ranging from 56-85% (Table 1.) Especially
interesting is the reaction of 5g, where R3 and R4 are both methyl
groups which afford the cyclocarbonylation product 17g in 58%
yield, albeit in 20 h. This long reaction time is likely due to the
developing A (1,3) strain in the transition state between the
methyl groups. Terminal alkynes were tolerated as exemplified
by the formation of products 17a, 17d, 17f, and 17h. Allene-ynes
5h-j were subjected to the cyclocarbonylation reaction condi-
tions as 1:1 mixtures of diastereomers (entries h-j). The
resulting cycloadducts 17h and 17i were obtained in enriched
diastereomeric ratios (1.5:1 and 2:1, respectively). While the
reason for this diastereoselectivity is not known, it is most likely
due to the steric bulk of the methyl group at R1 slightly biasing
the cyclocarbonylation reaction of one diastereomer of 5h and
5i over the other.
The reaction of allene-ynes 8a-c to the rhodium(I)-catalyzed
cyclocarbonylation reaction conditions afforded 18a-c and none
of the expected [5-7-6] ring systems 19 (Table 2). We
speculate that the hydroxyl group coordinates to the rhodium
metal, directing the reaction to the proximal double bond of
the allene, giving R-alkylidene cyclopentenones 18. Indeed, this
hypothesis is supported by the cyclocarbonylation reaction of
8d, which affords 19 as a result of the selective reaction of the
distal double bond in the absence of the hydroxyl group, albeit
in low yield. These low-yielding cyclocarbonylation reactions
were a concern, so the diastereomers of 8c were separated and
subjected independently to the reaction conditions. The major
1
(by H NMR) along with significant decomposition. It is not
clear at this time why the yield obtained for 18c was notably
lower than 18a and 18b; however, the significantly longer
reaction time necessary for consumption of 8c compared to 8a
and 8b likely contributes to the low yields. In addition, the crude
1H NMR spectra of 8b and 8c show substantial quantities of a
substance(s) possessing broad peaks (δ 4.3-4.0 and 2.0-1.0).
There have been no previously reported Rh(I)-catalyzed allenic
cyclocarbonylation reactions showing exclusive reactivity for
the proximal double bond of the allene.18
Allene-ynes 11a-e were reacted with 5 mol % of
[Rh(CO)2Cl]2 to give cyclocarbonylation adducts 20a-d (Table
3). For substrates 11c and 11d, bearing an R-hydroxyl or
carbonyl group adjacent to the alkyne (entries c and d),
cyclocarbonylation resulted in high yields of the [5-7-6]
cycloadducts. A trimethylsilyl group on the terminus of the
alkynone was necessary for this series of allene-ynes, with 11e
possessing a terminal alkyne resulting in rapid decomposition.
This result is not unexpected given the increased acidity of
proton on the terminus of the alkyne and its higher propensity
to react with the rhodium catalyst to form rhodium vinylidenes.19
Attempts to prepare allene-ynes possessing either a methyl or
phenyl group on the terminus of the alkyne were unsuccessful
due to substantial quantities of byproduct formed during the
addition of the corresponding propargyl anions to the vinylogous
ester 10a.
(17) Hung, S.-C.; Wen, Y.-F.; Chang, J.-W.; Liao, C.-C.; Uang, B.-J. J. Org.
Chem. 2002, 67, 1308–1313.
(18) Mukai and co-workers have found that electronically tuning the allene
gives mixtures of cyclocarbonylation products with both the distal and proximal
olefin reacting; the reaction with the distal olefin predominates in all cases. Inagaki,
F.; Kawamura, T.; Mukai, C. Tetrahedron 2007, 63, 5154–5160.
(19) (a) Brummond, K. M.; Chen, D.; Painter, T. O.; Mao, S.; Seifried, D. D.
Synlett 2008, 5, 759–764. (b) Trost, B. M.; Phan, L. T. Tetrahedron Lett. 1993,
34, 4735–4738.
J. Org. Chem. Vol. 73, No. 13, 2008 5067

Brummond et al.
TABLE 4. Rh(I)-Catalyzed Cyclocarbonylation Reaction of 15a-e
Affording 21a,c-e
of diastereomers (Scheme 9.) The diastereomers were readily
separated by chromatography and the major diastereomer was
assigned as 23 from the X-ray crystal structure.21
Conclusions
Four series of allene-ynes (types A-C) were prepared and
subjected to a catalytic Rh(I)-catalyzed cyclocarbonylation
reaction. Linearly and angularly fused tricyclic [6-7-5] ring
systems were afforded for three of these series. Low yields of
tricyclic [6-6-5] ring systems were obtained for one series of
type A allene-ynes (8a-c) possessing a tertiary propargylic
hydroxyl group, which is postulated to have a directing effect
on the constitutional group selectivity of this reaction. Moreover,
each series of allene-ynes behaved uniquely in the cyclocar-
bonylation reaction. For example, all allene-ynes 5a-j (type
A) underwent cycloaddition reactions with ease affording the
cyclocarbonylation products 17a-k in yields ranging from
56-85%. The substitution on the terminus of the alkyne or the
allene did not substantially affect this reaction. Alternatively,
allene-ynes 11 (type B) and 15 (type C) required that the alkyne
terminus be substituted with a TMS group. Preliminary inves-
tigations into the possible functional group manipulations of
the resulting dienones show that both epoxidation and dihy-
droxylation reactions occur with complete regioselectively
affording epoxide 22 and diol 23, respectively. Finally, the
Rh(I)-catalyzed cyclocarbonylation reaction shows excellent
functional group compatibility as evidenced by the tolerance
of skipped enynes, ketals, enones, hydroxyl groups, and esters.
Efforts are continuing in our laboratory to expand the scope of
this reaction and to apply this methodology to the synthesis of
more complex molecules.
entry
reaction
time
22 h
72 h
2 h
40 min
3 h
yield (%)
a
b
c
d
e
15a: R1 ) Me, R2 ) H, R3 ) TMS
15b: R1 ) R2 ) Me, R3 ) TMS
15c: R1 ) R2 ) H, R3 ) TMS
15d: R1 ) R2 ) R3 ) H
87
b
91
27%a
29
15e: R1 ) R2 ) H, R3 ) Me
a Reaction run at 60 °C. b 40-60% starting material recovered.
SCHEME 7. Desilylation of 21c Affording 21d
SCHEME 8. m-CPBA Epoxidation of 21c Affording
Epoxide 22
SCHEME 9. Upjohn Dihydroxylation of 21c Affording Diol
23
Experimental Section
General Procedure for [Rh(CO)2Cl]2-Catalyzed Cyclocarbo-
nylation Reaction. A flame-dried test tube equipped with a Teflon-
coated stir bar was charged with allene-yne and toluene (10 mL/
mmol). The tube was evacuated for 3-5 s (via a needle through
the septa) and refilled with CO(g) (from a balloon) (3×). To the
allene-yne solution was added [Rh(CO)2Cl]2 (0.05 equiv for 5, 8,
and 11 and 0.10 equiv for 15) in one portion and the test tube was
evacuated and refilled with CO(g) (3×). The test tube was placed
in a preheated 90 °C oil bath and stirred under CO(g). After the
reaction was complete by TLC, the mixture was cooled to rt and
filtered directly through a short pad of silica gel. Further purification
was accomplished as necessary via flash chromatography.
Allene-ynes 15a and 15c readily cyclized under Rh(I)-
cyclocarbonylation conditions to give 21a and 21c (Table 4).
For this series of allene-yne substrates, the yield of the
cyclocarbonylation reactions was dependent upon allene and
alkyne substitution. For example, low yields of 21d and 21e
were observed when the alkyne terminus possessed a methyl
group or hydrogen atom. The reaction appears to tolerate some
steric bulk at the terminus of the allene with 15a and 15c giving
nearly identical yields of 21a and 21c, respectively. However,
for 15b, the only example investigated using a tetrasubstituted
allene, no reaction was observed under the standard Rh(I)
conditions (entry b), even when allowed to react several days.
Ultimately, 21d could be obtained by removal of the TMS group
from 21c in 75% yield over two steps (Scheme 7).20
Acknowledgment. We thank the National Institutes of Health
(GM54161) for financial support of this project.
While formation of a number of angular and linear [6-7-5]
ring systems is useful, the ability to selectively introduce further
functionality extends the synthetic value of this methodology.
Having successfully obtained the [6-7-5] core ring systems,
further functionalization was briefly examined. Substrate 21c
was subjected to Na2HPO4-buffered m-CPBA conditions to give
epoxide 22 in 72% yield as a 2.8:1 mixture of diastereomers
(Scheme 8.) The major diastereomer was tentatively assigned
as 22 based upon a crystal structure of 23.
Supporting Information Available: Full experimental pro-
tocols and characterization of new compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
JO8007258
(20) Trost, B. M.; Yang, H.; Probst, G. D. J. Am. Chem. Soc. 2004, 126,
48–49.
(21) For crystallographic data, see the Supporting Information. This data
(CCDC 629739) may also be obtained free of charge from the Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
In addition, dihydroxylation of 21c using OsO4/NMO condi-
tions gave diol 23, exclusively, in 81% yield as a 2:1 mixture
5068 J. Org. Chem. Vol. 73, No. 13, 2008