Diastereoselection in the formation of contiguous quaternary carbon stereocenters by the intramolecular Heck reaction

Article abstract of DOI:10.1021/jo0523363

The second in a series of two papers, this study examines the origins of diastereoselection in the second ring closure of the highly diastereoselective double Heck cyclization of cyclohexenes 1 and 3 that form contiguous quaternary stereocenters. Seven mo

Full text of DOI:10.1021/jo0523363

Diastereoselection in the Formation of Contiguous Quaternary
Carbon Stereocenters by the Intramolecular Heck Reaction
Larry E. Overman* and Donald A. Watson
Department of Chemistry, 516 Rowland Hall, UniVersity of California, IrVine, California 92697-2025
leoVerma@uci.edu
ReceiVed NoVember 10, 2005
The second in a series of two papers, this study examines the origins of diastereoselection in the second
ring closure of the highly diastereoselective double Heck cyclization of cyclohexenes 1 and 3 that form
contiguous quaternary stereocenters. Seven model substrates were synthesized and cyclized to examine
the structural features responsible for imparting diastereoselection in the second intramolecular Heck
reaction. These studies demonstrate that stereoselection in the formation of the second spirooxindole
ring results from the avoidance of steric interactions in the insertion step with the spirooxindole formed
in the first Heck cyclization. An axial substituent (carbonyl or arene) is required at the allylic position
for high levels of diastereoselection to be realized.
SCHEME 1a
Introduction
In the context of recent total syntheses of meso- and (-)-
chimonanthine, we described the sequential double Heck
cyclization of C2-symmetric diiodides 1a and 3a.1 These
reactions, as well as those of the related ditriflate 1b or 3b,2
provided spirocyclic dioxindoles 2 and 4 with high levels of
diastereoselection (Scheme 1). Depending upon the nature of
the diol protecting group, the new spirocyclic oxindoles were
formed with either a trans or a cis stereorelationship. Double
Heck cyclizations of acetonides 1a and 1b were particularly
stereoselective, providing trans-hexacyclic dioxindole 2 as a
single detectable stereoisomer. The identical reaction of disiloxy
substrates 3 produced as the major product pentacyclic diox-
indole 4 having the alternative cis relationship of the newly
formed oxindole groups, although this transformation was less
clean than that of acetonide congeners 1. These reactions are
notable for their ability to stereoselectively form vicinal
quaternary carbon stereocenters, a challenging structural feature
that can be constructed by few extant synthetic methods.3
a Key: (a) 20 mol % Pd(PPh3)2Cl2, 10 equiv Et3N, DMA, 100 °C, 24 h.
As a result of the synthetic utility of the transformations
depicted in Scheme 1, we wished to further understand the
factors that control diastereoselection in these sequences. Studies
(1) Overman, L. E.; Paone, D. V.; Stearns, B. A. J. Am. Chem. Soc.
1999, 121, 7702-7003.
(2) Overman, L. E.; Watson, D. A. J. Org. Chem. 2006, 71, 2587-2599.
10.1021/jo0523363 CCC: $33.50 © 2006 American Chemical Society
Published on Web 02/28/2006
2600
J. Org. Chem. 2006, 71, 2600-2608

Diastereoselection in the Formation of Stereocenters
SCHEME 2a
pure diol 6 in 55% yield over the three steps.7 Silylation of 6
with TBDMSOTf and 2,6-lutidine gave disiloxy bromide 7 in
98% yield. Although palladium-catalyzed carbonylation of this
intermediate was unsuccessful,8 lithium/bromine exchange using
t-BuLi and quenching of the resultant vinyllithium intermediate
with CO2 provided the corresponding acid in good yield.
Coupling of this acid with N-benzyl-2-(triethylsiloxy)aniline
using N-methyl-2-chloropyridium iodide (Mukaiyama’s salt)9
provided disiloxy anilide 8. Removal of the triethyl silane
protecting group from this coupled product, followed by
triflation of the resulting phenol, provided disiloxy triflate 9 in
76% overall yield from diol alkenyl bromide 6.
The corresponding acetonide 11 precursor was accessed from
disiloxy triflate 9. The silyl protecting groups were removed
by the reaction of 9 with aqueous hydrofluoric acid to provide
diol 10. The use of other fluoride sources, such as Bu4NF,
resulted in slower deprotection at room temperature and
sporadically led to triflate cleavage, even when the reaction was
buffered with acetic acid. The conversion of diol 10 to acetonide
11 proved surprisingly challenging. Subjection of diol 10 to
various conditions commonly used to form acetonides failed to
generate acetonide 11.10 Similar difficulties have been described
in the conversion of trans-diol 6 to its acetonide derivative.7
Conditions were finally developed to carry out this transforma-
tion, albeit in moderate yield: the reaction of diol 10 with 1
equiv of 2,2-dimethoxypropane (2,2-DMP), catalytic pyridinium
p-toluenesulfonate (PPTS), and 4 Å molecular sieves (4 Å MS)
in benzene at 80 °C provided acetonide 11 in 42% yield.11
Presumably, the removal of methanol from the reaction media
by the molecular sieves facilitates formation of the otherwise
thermodynamically disfavored acetonide 11.12
a Key: (a) TMSOTf, Et3N; (b) m-CPBA, then HCl; (c) NaBH(OAc)3;
(d) TBDMSOTf, 2,6-lutidine; (e) t-BuLi, then CO2; (f) N-benzyl-2-
(triethylsiloxy)aniline, Mukaiyama’s salt, 2,4,6-collidine; (g) K2CO3, H2O;
(h) PhNTf2, Cs2CO3; (i) aq HF; (j) 1 equiv 2,2-DMP, 10 mol % PPTS, 4
Å MS, PhH, 80 °C.
reported in the preceding paper examined the origin of diaste-
reoselection in the first Heck cyclization step.2 This report
describes our studies into the origin of stereoselection in the
second ring-closing event of these unique transformations.
The preparations of cyclization substrates 17a, 17b, 21a, and
21b, which contain a spirooxindole substituent, are summarized
in Schemes 3 and 4. Methyl ester 122 was investigated initially
as a starting material for preparing triflatoanilide 17a (Scheme
3). However, attempts to cleave this ester under a variety of
conditions (LiOH in THF/H2O, KOTMS in Et2O,13 “anhydrous
hydroxide”,14 LiCl or LiI in pyridine,15 (Bu3Sn)2O,16 and several
Results
Our strategy was to examine this second Heck cyclization
with substrates that contained various structural components of
the initially formed spirooxindole intermediates 5 generated
upon the Heck reaction of ditriflates 1 and 3. The stereocontrol
imparted by the protected trans-diol and spirooxindole structural
fragments would be explored.
(6) For leading references to hydroxy-directed reductions of ketones,
see: Gribble, G. W. Sodium Triacetoxyborohydride. In Encyclopedia of
Reagents for Organic Synthesis; Paquette, L. A., Ed.; Wiley & Sons: New
York, 1995; Vol. 7, pp 4649-4653.
(7) For a previous synthesis of this intermediate, see: Banwell, M. G.;
Lambert, J. N.; Richards, S. L. Aust. J. Chem. 1991, 44, 939-950.
(8) Cacchi, S.; Ciattini, P. G.; Morera, E.; Ortar, G. Tetrahedron Lett.
1986, 37, 3931-3934.
(9) Bald, E.; Saigo, K.; Mukaiyama, T. Chem. Lett. 1975, 1163-1166.
(10) (a) Cai, J.; Davison, B. E.; Ganellin, C. R.; Thaisrivongs, S.
Tetrahedron Lett. 1995, 36, 6535-6536. (b) Larson, G. L.; Hernandez, A.
J. Org. Chem. 1973, 38, 3935-3936. (c) Rollin, P.; Pougny, J. R.
Tetrahedron 1986, 42, 3479-3490. (d) Tsunoda, T.; Suzuki, M.; Noyori,
R. Tetrahedron Lett. 1980, 21, 1357-1358.
Synthesis of Heck Cyclization Substrates. The syntheses
of cyclization substrates 9 and 11, that contain the protected
diol unit but lack the spirooxindole fragment, are summarized
in Scheme 2. These preparations start with 3-bromocyclohex-
enone,4 which was converted to the corresponding 2-siloxy-
1,3-diene by a reaction with TMSCl and Et3N. Rubottom
oxidation of this intermediate,5 followed by stereoselective
reduction of the resultant R-hydroxyketone with NaHB(OAc)3,6
provided trans-diol 6 with greater than 10:1 diastereoselectivity.
Recrystallization of this material provided diastereomerically
(11) Heating was required for this transformation; at room temperature,
only mono- and diprotected mixed methoxy ketals were isolated for this
class of diols.
(12) The acetonide of bromide 5 could also be prepared using this method
in 67% yield. For details, see: Watson, D. A. Investigation into the Origins
of Diastereoselection in Spirocyclic Oxindole Forming Intramolecular Heck
Cyclizations. Ph.D. Thesis, University of CaliforniasIrvine, Irvine, CA,
2004.
(13) Laganis, E. D.; Chenard, B. L. Tetrahedron Lett. 1984, 25, 5831-
5834.
(14) Gassman, P. G.; Schenck, W. N. J. Org. Chem. 1977, 42, 918-
920.
(3) Peterson, E. A.; Overman, L. E. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 11943-11948.
(15) (a) Liotta, C.; Markiewicz, W.; Santiesteban, H. Tetrahedron Lett.
1977, 18, 4365-4368. (b) Fisher, J. W.; Trinkle, K. L. Tetrahedron Lett.
1994, 35, 2505-2508.
(16) Salomon, C. J.; Mata, E. G.; Mascartti, O. A. Tetrahedron Lett.
1991, 32, 4239-4242.
(4) Piers, E.; Cheng, K. F.; Nagakura, I. Can. J. Chem. 1982, 60, 210-
223.
(5) Rubottom, G. M.; Vazquez, M. A.; Pelegrina, D. R. Tetrahedron
Lett. 1974, 15, 4319-4322.
J. Org. Chem, Vol. 71, No. 7, 2006 2601

Overman and Watson
SCHEME 3a
acyl chloride, coupled with N-benzyl-2-(tert-butyldimethylsi-
loxy)aniline (15), and the triflate functionality was installed by
the reaction of the unsaturated amide product with CsF, Cs2-
CO3, and PhNTf2 to provide triflatoanilide 17a in 31% overall
yield from ester 13a.
The analogous silyl-protected Heck cyclization substrate 17b
was prepared by a similar sequence (Scheme 3). Hydrogenolysis
of disiloxy benzyl ester 13b readily generated acid 14b.
However, amidation of this acid proved more difficult than in
the acetonide series. Attempts to access the corresponding acid
chloride by the treatment of this siloxy acid with oxalyl chloride,
resulted in a complex mixture of products. When this reaction
was carried out in the presence of various anilines, only trace
amounts of the desired unsaturated anilide were observed. The
use of O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium
hexafluorophosphate22 or PyBrOP23 also failed to provide the
coupled product. Likewise, the acyl fluoride derivative of acid
14b, prepared using FC(NMe2)2PF6,24 failed to react with aniline
nucleophiles. Formation of the acyl chloride derivative of 14b
in situ by a reaction with CCl4 and Ph3P (in solution or polymer
bound) and a reaction at 60 °C with aniline 15 delivered the
coupled anilide in ∼20-25% yield.25 However, unidentified
higher molecular weight side products were formed also,
requiring preparative HPLC purification to obtain pure samples
of anilide 16b. On the other hand, coupling of disiloxy acid
14b with aniline 15 in MeCN using bis(2-oxo-3-oxazolidinyl)-
phosphinic chloride (BOP-Cl)26 was cleaner, with only simple
column chromatography required to obtain pure samples of
anilide 16b. The conversion of this product to the triflate
derivative was straightforward, delivering disiloxy triflate 17b
in 38% yield over the three steps.
a Key: (a) H2, Pd/C, EtOAc; (b) (COCl)2, 2,6-lutidine, then 15; (c) 15,
BOP-Cl, i-Pr2NEt; (d) CsF, Cs2CO3, PhNTf2.
SCHEME 4a
The epimeric acetonide- and disiloxy-protected Heck cy-
clization precursors 21a and 21b were prepared from the
epimeric benzyl esters 18a and 18b20 using an analogous
sequence of reactions (Scheme 4).
Ditriflate 23, which lacks oxygen substituents at C4 and C5,
was synthesized from the Diels-Alder product 2227 (eq 1). The
regioselective hydrogenation of diene 22 was accomplished by
a brief exposure to hydrogen over Pd/C. A reaction of the
resultant diphenol with PhNTf2 in the presence of Cs2CO3
provided ditriflate 23 in 87% overall yield.
a Key: (a) H2, Pd/C, EtOAc; (b) (COCl)2, 2,6-lutidine, then 15; (c) 15,
BOP-Cl, i-Pr2NEt; (d) CsF, Cs2CO3, PhNTf2.
lipase/esterase conditions17) failed to provide the required acid
14a. These reactions either returned the starting ester or resulted
in the decomposition of the substrate. In particular, basic
conditions resulted in the rapid decomposition of ester 12,
presumably because of the acidity of the γ methine hydrogen
atom. Metal-mediated amidation conditions were also examined.
Both aluminum and bromomagenesium amides, generated by
the reaction of a variety of primary and secondary anilines with
Me3Al18 or i-PrMgBr,19 were investigated; however, only the
decomposition of the ester was observed. Similar results were
observed with the disiloxy congener of 12, as well as with
methyl esters having an epimeric spirooxindole fragment.
However, the benzyl ester analogue 13a cleanly provided the
desired acid 14a when subjected to mild hydrogenolysis
conditions.20,21 Acid 14a was transformed to the corresponding
Heck Cyclizations. This series of unsaturated triflates was
cyclized under identical reaction conditions: 20 mol % Pd-
(20) Benzyl esters 13a, 13b, 18a, and 18b were prepared in analogous
fashion to their methyl ester congeners. See ref 2 and Supporting Information
for details.
(21) Hydrogenation of the alkene was not observed in this transformation
after short reaction times.
(22) Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397-4398.
(23) Costa, J.; Frerot, E.; Jouin, P.; Castro, B. Tetrahedron Lett. 1991,
32, 1967-1970.
(17) For leading references on the enzymatic cleavage of methyl esters,
see: Green, T. W.; Wuts, P. G. M. ProtectiVe Groups in Organic Synthesis,
3rd ed.; Wiley and Sons: New York, 1999; p 385.
(18) Basha, A.; Lipton, M.; Weinreb, S. M. Tetrahedron Lett. 1977, 48,
4171-4174.
(19) Williams, J. M.; Jobson, R. B.; Yasuda, N.; Marchesini, G.; Dolling,
U. H.; Grabowski, E. J. J. Tetrahedron Lett. 1995, 36, 5461-5464.
2602 J. Org. Chem., Vol. 71, No. 7, 2006

Diastereoselection in the Formation of Stereocenters
SCHEME 5a
In contrast, triflate 17a, which contains a spirooxindole unit,
cyclized cleanly to provide dioxindole 2 as the only detectable
product (by 1H NMR analysis). This outcome was expected, as
spirooxindole 17a is the putative intermediate in the high-
yielding conversion of ditriflate 1b to dioxindole 2 (Scheme
1). The relative configuration of dioxindole 2 is known from
X-ray crystallographic analysis.1
Heck cyclization of epimeric oxindole triflate 21a occurred
also with a high degree of diastereoselection, with dioxindole
26 (isolated in 87% yield) being formed as the predominant
product together with traces (∼3-4%) of an unidentified side
product.29 The relative configuration of Heck product 26 was
assigned by chemical correlation: the reaction of this product
with dilute aqueous trifluoroacetic acid in refluxing THF cleaved
the isopropylidene fragment to provide the corresponding keto
alcohol, which had been characterized previously by X-ray
crystallographic analysis.27
The analogous silyl-protected substrates provided qualitatively
similar results upon Heck cyclization (Scheme 6). For example,
triflate 9 provided spirooxindoles 27 and 28 in a 57:43 ratio
upon Heck cyclization. These products were isolated in 91%
combined yield, and their configurations were assigned by NOE
analysis. The major product 27, which adopts a half-chair
conformation having the C4 siloxy substituent pseudoaxial
(JH-H),28 displayed a 2.9% NOE for the C12 hydrogen atom
upon irradiation of the pseudoaxial hydrogen atom at C6. The
minor product 28, which lacked the siloxy group at C3,
displayed a 1.5% NOE enhancement for the C12 hydrogen atom
and a 1.9% NOE enhancement for the pseudoaxial C4 methine
hydrogen atom upon irradiation of the pseudoaxial hydrogen
atom at C6. Spirooxindole 28 arises from the migratory insertion
to the alkene face of 9 cis to the proximal siloxy substituent.
As the resultant cyclohexyl palladium intermediate lacks a cis
â-hydrogen atom, â-siloxy elimination takes place.30
Spirooxindole 21b, which was the putative intermediate en
route to the major dioxindole product 4 in the double cyclization
of disiloxy ditriflate 3, underwent Heck cyclization to provide
products 4, 29, and 30 in a 75:9:16 ratio (Scheme 6).31,32 The
known cis dispirooxindole 42 was isolated in a 52% yield. Arene
30,2 which is the product of triflate reduction, was isolated in
a 14% yield. The 1H and 13C NMR spectra of the minor product
29 showed that it had C2 symmetry. Its relative configuration
was established by an independent synthesis by silyl protection
of a sample of the corresponding diol.33 Thus, facial selectivity
in the Heck cyclization of siloxy ditriflate 21b is 89:11. Product
29 arises from a reductive process analogous to that involved
a Key: (a) 20 mol % Pd(PPh3)2Cl2, 10 equiv Et3N, DMA, 100 °C, 24 h.
(PPh3)2Cl2, 10 equiv Et3N, DMA, 100 °C, 24 h; product ratios
1
were determined by the analysis of the H NMR spectra of
unpurified reaction mixtures. The results of the Heck cycliza-
tions of triflates 11, 17a, and 21a, containing acetonide units,
are shown in Scheme 5. In all three cyclizations, the relative
configuration of the newly formed spirooxindole (relative to
the acetonide) of the major product is the same as that of the
second spirooxindole produced in the double cyclization of
ditriflate 1b (Scheme 1).
Heck cyclization of triflate 11, which lacks an oxindole
fragment, provided spirooxindoles 24 and 25 in a 63:37 ratio.
These products, which differ in molecular weight by 2 mass
units and are epimeric at the quaternary stereocenter, were
isolated in 87% combined yield. The configuration of major
product 24 was assigned by the observation of a 3.4% NOE
enhancement for the C12 hydrogen atom upon irradiation of
the axial C5 hydrogen atom.28 The equatorial disposition of the
aryl moiety of the spirooxindole group of minor product 25 was
assigned by the observation of NOE enhancements for both the
axial C6 and the axial C2 hydrogen atoms upon irradiation of
the C12 hydrogen atom.28 The dihydro product 25 is undoubt-
edly formed because migratory insertion into the alkene face
cis to the proximal ether generates a cyclohexyl palladium
species lacking a â-hydrogen atom cis to the metal center. The
reduction of this intermediate, most likely by excess Et3N
present in solution, would lead to the cyclohexyl product 25.
(29) The low yield of this minor product, and the limited amounts of
starting material available, prevented successful isolation.
(30) For examples of palladium-mediated reactions terminated by
â-alkoxide eliminations, see: (a) Hacksell, U.; Daves, G. D., Jr. Organo-
metallics 1983, 2, 772-775. (b) Saito, S.; Hara, T.; Takahashi, N.; Hirai,
M.; Moriwake, T. Synlett 1992, 237-238. (c) Daun, J.; Cheng, C.
Tetrahedron Lett. 1993, 34, 4019-4022. (d) Hosokawa, T.; Sugafugi, T.;
Yamanaka, T.; Murahashi, S. J. Organomet. Chem. 1994, 470, 253-255.
(e) Moinet, C.; Fiaud, J. Tetrahedron Lett. 1995, 36, 2051-2052. (f)
Bejeguelal, K.; Joseph, L.; Bolitt, V.; Sinou, D. J. Carbohydr. Chem. 2000,
19, 221-232.
(31) The cyclizations of oxindoles 17b and 21b proceeded more slowly
than the previous substrates; after 24 h, significant quantities of starting
material remained (ca. 15-20%). Upon heating the mixture for 30 h, the
remaining starting material had been reduced to about 3-5%.
(32) As the 1H NMR spectra for these compounds were complex at room
temperature, the 1H NMR spectrum of the crude reaction mixture was
recorded at 100 °C in toluene-d8.
(24) Carpino, L. A.; El-Faham, A. J. Am. Chem. Soc. 1995, 117, 5401-
5402.
(25) Lee, J. B. J. Am. Chem. Soc. 1966, 88, 3440-3441.
(26) Diago-Meseguer, J.; Palomo-Coll, A. L.; Fernandez-Lizarbe, J. R.;
Zugaza-Bilbao, A. Synthesis 1980, 547-551.
(27) Stearns, B. A. A Formal Synthesis of Loracarbef and Total Synthesis
of Polypyrrolidinoindoline Alkaloids. Ph.D. Thesis, University of Californias
Irvine, Irvine, CA, 2000.
(33) Overman, L. E.; Larrow, J. F.; Stearns, B. A.; Vance, J. M. Angew.
Chem., Int. Ed. 2000, 39, 213-215.
(28) See Supporting Information for details.
J. Org. Chem, Vol. 71, No. 7, 2006 2603

Overman and Watson
SCHEME 6a
a Key: (a) 20 mol % Pd(PPh3)2Cl2, 10 equiv Et3N, DMA, 100 °C, 24 h; (b) 20 mol % Pd(PPh3)2Cl2, 10 equiv Et3N, DMA, 100 °C, 30 h.
SCHEME 7a
forms dispirooxindole epimers 34 and 35 (isolated in an 83%
combined yield and in a 3:1 ratio). The relative configurations
of these products were defined by catalytic hydrogenation of
the major isomer 34 to give rise to the C2-symmetric cyclohexyl
dioxindole 36, whose relative configuration was secured by
single-crystal X-ray analysis.
Discussion
The results of Heck cyclizations in the acetonide and disiloxy
series reported in Schemes 5 and 6 demonstrate the importance
of the initially formed spirooxindole in directing stereoselection
of the second Heck cyclization in the cascade double Heck
reactions depicted in Scheme 1. In both series, substrates
containing a spirooxindole moiety cyclized with significantly
higher levels of diastereoselection than substrates lacking this
functional group. For example, aryl triflates 17a and 21a, which
differ in relative configuration of the spirooxindole unit, cyclized
to give a single (dr > 20:1) dispirooxindole product. In contrast,
cyclization substrate 11 lacking the spirooxindole moiety
cyclized with little diastereoselection. In addition, the sense of
diastereoselection in forming the second spirooxindole fragment
was the same in both the acetonide and disiloxy series. Only
the magnitude of diastereoselection varied, being higher in the
acetonide series.
The interpretation of these results is straightforward in the
acetonide series, because the trans orientation of the acetonide
fixes the conformation of the cyclohexene ring. An analysis of
the ground-state conformations of these substrates (Figure 1)
provides insight into the role of the spirooxindole in regulating
stereoselection. In cyclization precursor 17a, the pseudoaxial
arene fragment of the spirooxindole creates significant steric
encumbrance about the lower face of the double bond, whereas
the oxindole carbonyl oxygen is positioned above the double
bond. As the Heck cyclization of spirooxindole 17a occurs
preferentially to the upper face of the double bond, stereose-
lection could derive from the steric influence of the pseudoaxial
a Key: (a) 20 mol % Pd(PPh3)2Cl2, 10 equiv Et3N, DMA, 100 °C, 24 h.
in the production of spirooxindole 25 (Scheme 5), with the
product of â-siloxy elimination not being observed. The reason
for the divergent behavior of the two putative â-siloxy palladium
alkyls en route to 28 and 29 is not understood at this time.
The epimeric oxindole precursor 17b produced pentacyclic
dispirooxindoles 31 and 32 in a 91:9 ratio.31 After chromato-
graphic separation, Heck products 31 and 32 were isolated,
respectively, in yields of 61 and 8%; a small amount (7%) of
the starting triflate was also recovered. The structure and relative
configuration of the major product 31 was determined by single-
crystal X-ray diffraction analysis. NMR and mass spectroscopy
showed that product 32 lacked a siloxy group at C4. As a result
of the small amount of this product isolated, the relative
configuration of 32 was not rigorously determined. However,
in analogy to the formation of the siloxy cyclohexene 28,
mechanistic reasoning strongly supports the assignment of this
minor Heck product as the cis dispirooxindole isomer 32.
Finally, ditriflate 23, which lacks the trans-diol functionality,
cyclized with low stereoselection (Scheme 7). This reaction
2604 J. Org. Chem., Vol. 71, No. 7, 2006

Diastereoselection in the Formation of Stereocenters
FIGURE 1. Depiction of the conformations of 11, 17a, and 21a. The
adjacent axial fragment of the spirooxindole dominates the steric
environment of the double bond.
aryl unit, coordination of palladium to the oxindole carbonyl
oxygen, or a combination thereof. The observation that spiroox-
indole epimer 21a also cyclizes preferentially from the upper
face rules out coordination as the major stereodetermining factor
in these Heck cyclizations. Thus, the steric influence of the
adjacent pseudoaxial substituent is largely regulating facial
stereoselection. Consistent with this interpretation is the low
facial stereoselectivity seen in the Heck cyclization of substrate
11, which lacks the spirooxindole substituent.34
FIGURE 2. Depiction of the two half-chair conformations of 9, 17b,
and 21b.
Stereoselection in the Heck cyclizations of the silyl-protected
substrates can be rationalized by similar steric considerations.
For these cyclization substrates, two half-chair ground-state
conformations must be considered (Figure 2). In the case of
Heck precursors 17b and 21b, which contain a spirooxindole
fragment, the analysis is straightforward as conformation A,
having the siloxy groups oriented pseudoequatorial, should be
much preferred; the alternate conformer B suffers from severe
1,3-diaxial interactions. As in the acetonide series, preferential
migratory insertion from the face opposite the pseudoaxial aryl
fragment of the spirooxindole rationalizes the observed stereo-
selection in the Heck cyclizations of precursors 17b and 21b.
The low stereoselectivity observed in the Heck cyclization of
substrate 9, which lacks the spirooxindole group, likely reflects
cyclization taking place from both half-chair conformations.35
FIGURE 3. Depiction of the two half-chair conformations of 33; Heck
insertion takes place with a 3:1 facial selectivity from the top face.
from the top face of the double bond reflects coordination of
the palladium atom to the carbonyl oxygen of the spirooxin-
dole.36
Finally, we consider the modest diastereoselection (3:1)
observed in the cyclization of ditriflate 23 (Scheme 7), which
lacks oxygen substitution at C4 and C5. Molecular mechanics
calculations (MM2) found two low-energy half-chair cyclo-
hexenes for intermediate 33 (Figure 3); these differ in energy
by 0.75 kcal/mol, corresponding to an ∼3:1 mixture of
conformers at 100 °C. Most likely the similar populations of
these conformers contributes to the low stereoselectivity ob-
served in this case. It is tempting to speculate that the small
preference observed for the Heck cyclization of 33 to take place
Conclusions
This study investigated the origin of stereoselection in the
second ring-closing event in the sequential Heck cyclizations
of C2-symmetric ditriflates 1 and 3, reactions that assemble
complex hexacyclic products containing vicinal quaternary
carbon stereocenters. The former substrate, in which the trans-
C4,C5 diol is masked as an acetonide, exclusively forms
(36) Coordination-controlled Heck reactions have been suggested for
various Lewis basic functional groups. For examples, see: (a) Earley, W.
G.; Oh, T.; Overman, L. E. Tetrahedron Lett. 1988, 29, 3785-3788. (b)
Madin, A.; Overman, L. E. Tetrahedron Lett. 1992, 33, 4859-4862. (c)
Nilsson, P.; Larhed, M.; Hallberg, A. J. Am. Chem. Soc. 2001, 123, 8217-
8225. (d) Itami, K.; Nokami, T.; Ishimura, Y.; Mitsudo, K.; Kamel, T.;
Yoshida, J. J. Am. Chem. Soc. 2001, 123, 11577-11585. (e) Buezo, N. D.;
de la Rosa, J. C.; Priego, J.; Alonso, I.; Carretero, J. C. Chem.sEur. J.
2001, 7, 3890-3900. (f) Nilsson, P.; Larhed, M.; Hallberg, A. J. Am. Chem.
Soc. 2003, 125, 3430-3431. (g) Oestreich, M.; Sempere-Culler, F.;
Machotta, A. B. Angew. Chem., Int. Ed. 2005, 44, 149-152. (h) Osetreich,
M. Eur. J. Org. Chem. 2005, 783-792.
(34) The small level of diastereoselection observed in the cyclization of
substrate 11 is consistent with the avoidance of developing eclipsing
interactions with the C1 hydrogen atoms, in analogy to the model developed
in the previous article.2
(35) Because of steric interactions, trans-vicinal siloxy groups destabilize
the diequatorial conformation, making it energetically similar to the diaxial
conformation. See: Marzabadi, C. H.; Anderson, J. E.; Gonzalez-Outeirino,
J.; Gaffney, P. R. J.; White, C. G. H.; Tocher, D. A.; Todaro, L. J. J. Am.
Chem. Soc. 2003, 125, 15163-15173.
J. Org. Chem, Vol. 71, No. 7, 2006 2605

Overman and Watson
SCHEME 8. Intermediates 17a and 21b, Formed Preferentially in the First Heck Cyclization Step, Undergo the Second Heck
Cyclization from the Less-Sterically-Hindered Top Face to Preferentially Form Products 2 and 4
Column chromatography (79:20:1, hexanes/EtOAc/Et3N) afforded
45.7 mg (87%) of 24 and 25 as a mixture of congeners as a colorless
oil. Preparative HPLC (40 mL/min; 89:10:1, hexanes/EtOAc/Et3N;
300 × 50 mm, 5 µm silica gel column) allowed for the separation
of these compounds for characterization.
hexacyclic product 2 having trans-related spirooxindole frag-
ments. In contrast, disiloxy congener 3 preferentially forms
hexacyclic product 4 in which the newly formed spirooxindoles
are cis-related, although stereoselection is lower than in the
acetonide series. The presence of the spirooxindole formed in
the first step of these sequential cyclizations plays the decisive
role in dictating facial stereoselection in the second Heck
insertion step. Intermediates 17a and 21b are produced prefer-
entially in the first Heck cyclization step.2 They undergo the
second migratory insertion step with high facial selectivity from
the less-hindered alkene face (opposite the bulky pseudoaxial
fragment of the spirooxindole) in the low-energy conformations
depicted in Scheme 8. Substrates analogous to 17a and 21b
that lack the spirooxindole substituent cyclize with low stereo-
selection.
Data for 24: IR (film) 2989, 2941, 1706, 1609, 1219, 1127 cm-1
;
1H NMR (500 MHz, C6D6) δ 7.09-7.16 (m, 3H), 7.03 (t, J ) 7.1
Hz, 2H), 6.98 (t, J ) 7.2 Hz, 1H), 6.87 (td, J ) 7.7, 0.8 Hz, 1H),
6.75 (t, J ) 7.6 Hz, 1H), 6.43 (d, J ) 7.8 Hz, 1H), 4.74 (d, J )
15.7 Hz, 1H), 4.52-4.56 (m, 2H), 4.46 (qd, J ) 5.5, 1.6 Hz, 1H),
2.20 (td, J ) 14.0, 3.3 Hz, 1H), 1.94-1.99 (m, 2H), 1.82-1.90
(m, 1H), 1.49 (dt, J ) 13.6, 3.1 Hz, 1H), 1.43 (s, 3H), 1.24 (s,
3H); 13C NMR (125 MHz, C6D6) δ 180.3, 156.9, 142.3, 137.1,
136.5, 129.3, 128.9, 128.0, 127.8, 124.4, 122.9, 111.9, 109.4, 92.2,
73.5, 50.5, 44.2, 30.9, 27.4, 27.1, 24.9; LRMS (ESI, m/z) 384.11
(M + Na)+; HRMS (EI+, m/z) calcd for C23H23NO3, 361.1678 (M)+;
found, 361.1680.
Experimental Section37
Data for 25: IR (film) 2988, 2934, 1702, 1363, 1096 cm-1; 1H
NMR (500 MHz, C6D6) δ 7.07 (d, J ) 7.3 Hz, 2H), 6.93-7.01
(m, 3H), 6.84 (td, J ) 7.7, 1.3, 1H), 6.77 (td, J ) 7.5, 1.1 Hz, 1H),
6.67 (dd, J ) 7.4, 0.8 Hz, 1H), 6.37 (d, J ) 7.7 Hz, 1H), 4.96
(ddd, J ) 16.2, 8.2, 4.2 Hz, 1H), 4.60 (d, J ) 15.7 Hz, 1H), 4.54
(d, J ) 15.7 Hz, 1H), 3.47 (ddd, J ) 12.2, 8.8, 3.8 Hz, 1H), 2.66
(ddd, J ) 24.6, 12.0, 4.1 Hz, 1H), 2.18 (ddd, J ) 12.5, 4.1, 1.5
Hz, 1H), 1.85-1.94 (m, 2H), 1.56 (s, 3H), 1.52 (s, 3H), 1.46 (td,
J ) 12.8, 4.2 Hz, 1H); 13C NMR (125 MHz, C6D6) δ 179.8, 142.6,
137.0, 134.3, 139.4, 128.7, 128.4, 128.0, 127.8, 109.9, 109.4, 81.2,
75.5, 48.0, 43.6, 37.1, 33.7, 28.0, 27.7, 25.6; LRMS (ESI, m/z)
386.15 (M + Na)+; HRMS (CI+, NH3, m/z) calcd for C23H25NO3,
363.1834 (M)+; found, 363.1835.
Heck Cyclization of Triflate 17a. According to the general
procedure, triflate 17a (43.1 mg, 60.0 µmol), Pd(PPh3)2Cl2 (8.4 mg,
12.0 µmol), Et3N (60.7 mg, 77 µL, 600 µmol), and DMA (0.6 mL)
were heated to give rise to dioxindole 2 as a single isomer. Column
chromatography (90:10, hexanes/EtOAc) afforded 31.1 mg (91%)
2 as a colorless oil. The analytical data for this compound matched
that previously reported.1
General Procedure for Heck Reactions. In a glovebox under
a nitrogen atmosphere, the triflate substrate, Pd(PPh3)2Cl2, Et3N,
and DMA were combined in a base-washed glass vial containing
a magnetic stir bar. The vial was sealed with a Teflon-lined cap
and placed in a 100 °C aluminum heating-block (bored to the
diameter of the vial) atop a magnetic stir plate. The reaction was
maintained at 100 °C with stirring for 24 h, during which time the
initial yellow suspension became a deep red, homogeneous solu-
tion.38 At the end of the indicated time, the reaction was cooled to
room temperature, removed from the glovebox, opened to the
atmosphere, and diluted with Et2O. The resultant solution was
washed with water twice and once with brine, dried over MgSO4,
and concentrated in vacuo. A 1H NMR analysis of the crude reaction
mixture was then used to determine the diastereomeric ratio of the
products formed.
Heck Cyclization of Triflate 11. According to the general
procedure, triflate 11 (74.1 mg, 145 µmol), Pd(PPh3)2Cl2 (20.3 mg,
28.6 µmol), Et3N (146 mg, 0.20 mL, 1.45 mmol), and DMA (1.4
mL) were heated to give a 63:37 mixture of congeners 24 and 25.
Heck Cyclization of Triflate 21a. According to the general
procedure, triflate 21a (50.7 mg, 70.5 µmol), Pd(PPh3)2Cl2 (9.9 mg,
14.1 µmol), Et3N (71.3 mg, 90 µL, 705 µmol), and DMA (0.7 mL)
were heated to give rise to dioxindole 26 as the major product
(>95%). Traces (ca. 3-4%) of an additional, unidentified product
were observed in the 1H NMR spectrum of the crude reaction
mixture. Repeated attempts failed to isolate this material. Column
(37) (a) General experimental details have been described: MacMillan,
D. W. C.; Overman, L. E.; Pennington, L. D. J. Am. Chem. Soc. 2001,
123, 9033-9044. (b) CCDC 288976-288978 contains the supplementary
crystallographic data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
(38) On occasion, Pd black was deposited toward the end of the reaction.
2606 J. Org. Chem., Vol. 71, No. 7, 2006

Diastereoselection in the Formation of Stereocenters
chromatography (90:10, hexanes/EtOAc) afforded 35.1 mg (87%)
of 26 as a colorless oil: IR (film) 3058, 2991, 2937, 1713, 1609,
1466, 1216 cm-1; 1H NMR (500 MHz, CDCl3) δ 7.48 (d, J ) 7.4
Hz, 1H), 7.35 (d, J ) 7.5 Hz, 2H), 7.27 (t, J ) 7.3 Hz, 2H), 7.17-
7.23 (m, 2H), 7.04-7.12 (m, 3H), 7.00 (t, J ) 7.5 Hz, 2H), 6.56
(d, J ) 7.9 Hz, 1H), 6.51 (t, J ) 8.1 Hz, 2H), 6.45 (d, J ) 7.6 Hz,
2H), 5.78 (d, J ) 7.4 Hz, 1H), 5.53 (d, J ) 9.5 Hz, 1H), 5.06 (d,
J ) 15.6 Hz, 1H), 4.97 (d, J ) 15.9 Hz, 1H), 4.92 (d, J ) 15.5
Hz, 1H), 4.78 (d, J ) 1.9 Hz, 1H), 4.26 (d, J ) 15.9 Hz, 1H), 3.55
(dd, J ) 14.3, 9.6 Hz, 1H), 2.18 (d, J ) 14.4 Hz, 1H), 1.71 (s,
3H), 1.64 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 176.8, 176.5,
157.6, 143.4, 142.8, 135.7, 135.1, 132.8, 128.5, 128.8, 128.62,
128.56, 128.5, 127.30, 127.25, 127.0, 126.4, 125.9, 124.2, 122.2,
122.1, 112.1, 109.14, 109.07, 92.4, 70.2, 54.1, 53.3, 44.3, 43.8,
33.9, 27.5, 24.2; LRMS (ESI, m/z) 569.22 (M + H)+, 591.20 (M
+ Na)+; HRMS (ESI, m/z) calcd for C37H32N2O4, 591.2260 (M +
Na)+; found, 591.2239.
Heck Cyclization of Triflate 9. According to the general
procedure, triflate 9 (211 mg, 301 µmol), Pd(PPh3)2Cl2 (42.3 mg,
60.2 µmol), Et3N (305 mg, 0.42 mL, 3.01 mmol), and DMA (3.0
mL) were heated to give a 57:43 mixture of congeners 27 and 28.
Column chromatography (92.5:7.0:0.5, hexanes/Et2O/Et3N) afforded
75.3 mg (45%) of silyl enol ether 27 and 57.8 mg (46%) of alkene
28, both as colorless oils.
Hz, 1H), 1.02 (s, 9H), 0.89 (s, 9H), 0.15-0.25 (m, 9H), -0.14 (br
s, 3H); 13C NMR (125 MHz, C6D6, rt) δ 177.1, 176.4, 156.5, 144.6,
143.7, 137.2, 136.3, 131.3, 130.6, 130.1, 129.3, 129.2, 129.1, 129.0,
128.7, 127.7, 127.4, 126.8, 125.9, 122.2, 121.8, 109.7, 109.4, 106.4,
67.1, 55.1, 52.0, 45.0, 43.7, 37.2, 26.6, 26.4, 19.1, 18.5, -3.3, -3.4,
-3.9, -4.2; LRMS (ESI, m/z) 757.73 (M + H)+, 779.7 (M + Na)+;
HRMS (ESI, m/z) calcd for C46H56N2O4Si2, 779.3676 (M + Na)+;
found, 779.3684. The 1H NMR spectrum was also recorded in
toluene-d8 at 100 °C and CDCl3 at room temperature.28
Data for 29: IR (film) 2929, 2856, 1698, 1611, 1362, 1096 cm-1
;
1H NMR (500 MHz, C6D6) δ 7.43 (dd, J ) 7.4, 0.9 Hz, 2H), 6.98-
7.16 (m, 10H), 6.62 (td, J ) 7.7, 1.3 Hz, 2H), 6.57 (td, J ) 7.6,
1.1 Hz, 2H), 6.11 (d, J ) 7.1 Hz, 2H), 5.28-5.34 (m, 2H), 4.86
(d, J ) 15.6 Hz, 2H), 4.30 (d, J ) 15.6 Hz, 2H), 3.53-3.62 (m,
2H), 2.05 (dd, J ) 13.9, 4.0 Hz, 2H), 1.02 (s, 18H), 0.28 (s, 6H),
0.21 (s, 6H); 13C NMR (125 MHz, C6D6) δ 178.1, 142.9, 136.7,
130.1, 129.3, 128.8, 128.1, 125.0, 123.1, 109.3, 71.4, 53.0, 43.8,
38.9, 26.8, 18.7, -3.1, -4.1; LRMS (ESI, m/z) 759.73 (M + H)+,
781.71 (M + Na)+; HRMS (ESI, m/z) calcd for C46H58N2O4Si2,
781.3833 (M + Na)+; found, 781.3836. The H NMR spectrum
1
was also recorded in toluene-d8 at 100 °C.28
Heck Cyclization of Triflate 17b. In a modification of the
general procedure, triflate 17b (111 mg, 122 µmol), Pd(PPh3)2Cl2
(17.1 mg, 24.4 µmol), Et3N (123 mg, 0.17 mL, 1.22 mmol), and
DMA (1.2 mL) were heated for 30 h at 100 °C to give a 91:9
mixture of compounds 31 and 32, along with traces of remaining
starting material (ca. 5%). As a result of the complexity at room
Data for 27: IR (film) 2954, 2929, 2858, 1713, 1657, 1611, 1250,
1
1081 cm-1; H NMR (500 MHz, C6D6) δ 7.27 (d, J ) 7.3 Hz,
1H), 7.14-7.17 (m, 3H), 7.06 (t, J ) 7.3 Hz, 2H), 6.99 (t, J ) 7.4
Hz, 1H), 6.88 (t, J ) 7.7 Hz, 1H), 6.81 (t, J ) 7.4 Hz, 1H), 6.44
(d, J ) 7.4 Hz, 1H), 4.74-4.79 (m, 2H), 4.54 (d, J ) 15.6 Hz,
1H), 4.20 (br s, 1H), 2.90 (br t, J ) 13.3 Hz, 1H), 2.31 (t, J ) 12.1
Hz, 1H), 1.70-1.82 (m, 2H), 1.05 (s, 9H), 0.97 (s, 9H), 0.25 (s,
3H), 0.10-0.13 (m, 9H); 13C NMR (125 MHz, C6D6) δ 179.1,
156.5, 142.6, 137.0, 135.5, 129.0, 128.4, 127.7, 127.5, 124.0, 122.9,
109.0, 106.1, 68.2, 49.5, 43.5, 29.3, 28.2, 26.22, 26.17, 18.5, 18.4,
-3.98, -4.01, -4.04, -4.5; LRMS (ESI, m/z) 550.4 (M + H)+,
572.4 (M + Na)+; HRMS (ESI, m/z) calcd for C25H47NO3Si2,
572.2992 (M + Na)+; found, 572.3004.
1
temperature, the H NMR spectrum of the crude reaction mixture
was recorded at 100 °C in toluene-d8. Column chromatography (95:
5-85:15, hexanes/EtOAc) afforded 59.8 mg (65%) of 31 as a
colorless solid, 6.4 mg (8%) of 32 as a colorless solid, and 7.7 mg
(7%) of recovered 17b. Slow diffusion of n-hexane into an EtOAc
solution of 31 provided crystals suitable for X-ray crystallographic
analysis.
Data for 31: IR (film) 2934, 2860, 1710, 1660, 1640, 1220, 1100
1
cm-1; H NMR (500 MHz, C6D6, rt) δ 7.70-7.75 (m, 1H), 7.47
(d, J ) 6.8 Hz, 1H), 7.00-7.10 (m, 11H), 6.66-6.80 (m, 3H),
6.61 (td, J ) 6.6, 1.0 Hz, 1H), 6.13-6.18 (m, 2H), 4.98 (s, 1H),
4.82 (d, J ) 15.8 Hz, 1H), 4.75 (d, J ) 6.3 Hz, 1H), 4.69 (d, J )
15.7 Hz, 1H), 4.51 (dd, J ) 15.9, 2.1 Hz, 1H), 3.93 (dd, J ) 14.8,
6.5 Hz, 1H), 2.06 (d, J ) 14.8 Hz, 1H), 1.07 (s, 9H), 1.05 (s, 9H),
0.34 (s, 3H), 0.30 (s, 3H), 0.25 (s, 3H), 0.02 (s, 3H); 13C NMR
(125 MHz, C6D6, rt) δ 177.4, 175.3, 156.3, 143.6, 143.5, 136.9,
136.6, 131.5, 130.4, 129.2, 129.2, 129.1, 128.9, 128.7, 128.6, 128.1,
128.03, 127.96, 127.8, 127.0, 125.0, 123.2, 122.5, 109.3, 109.1,
104.5, 66.7, 54.8, 49.9, 44.3, 43.8, 39.3, 26.8, 26.7, 19.2, 18.7,
-3.07, -3.11, -3.5, -4.1; LRMS (ESI, m/z) 757.43 (M + H)+,
779.41 (M + Na)+; HRMS (ESI, m/z) calcd for C46H56N2O4Si2,
Data for 28: IR (film) 2954, 2929, 2856, 1711, 1611, 1250, 1081
cm-1; 1H NMR (500 MHz, C6D6) δ 7.14-7.17 (m, 2H), 7.04 (t, J
) 7.3 Hz, 2H), 6.98 (t, J ) 7.4 Hz, 1H), 6.88-6.94 (m, 2H), 6.82
(t, J ) 7.5 Hz, 1H), 6.46 (d, J ) 7.7 Hz, 1H), 6.04 (d, J ) 9.9 Hz,
1H), 5.23 (d, J ) 9.9 Hz, 1H), 4.68 (d, J ) 15.6 Hz, 1H), 4.56 (d,
J ) 15.6 Hz, 1H), 4.34 (br s, 1H), 2.59-2.70 (m, 1H), 2.10-2.18
(m, 1H), 1.84-1.91 (m, 1H), 1.66 (t, J ) 11.4 Hz, 1H), 1.01 (s,
9H), 0.14 (s, 3H), 0.13 (s, 3H); 13C NMR (125 MHz, C6D6) δ 179.1,
156.5, 142.6, 137.0, 135.5, 129.0, 128.4, 127.7, 127.5, 124.0, 122.9,
109.0, 106.1, 68.2, 49.5, 43.5, 29.3, 28.2, 26.22, 26.17, 18.5, 18.4,
-3.98, -4.01, -4.04, -4.5; LRMS (ESI, m/z) 420.07 (M + H)+,
437.09 (M + Na)+; HRMS (ESI, m/z) calcd for C26H33NO2Si,
420.2359 (M + H)+; found, 420.2372.
1
779.3676 (M + Na)+; found, 779.3669. The H NMR spectrum
was also recorded in toluene-d8 at 100 °C.28
Heck Cyclization of Triflate 21b. In a modification of the
general procedure, triflate 21b (74.0 mg, 81.6 µmol), Pd(PPh3)2-
Cl2 (11.4 mg, 16.3 µmol), Et3N (82.5 mg, 110 µL, 815 µmol), and
DMA (0.8 mL) were heated for 30 h to give a 75:9:16 mixture of
compounds 4, 29, and 30 along with traces of starting material (ca.
3%). Shorter reaction times resulted in a lower conversion of the
starting material. As a result of the uninterpretable complexity at
Data for 32: IR (film) 2929, 2856, 1723, 1609, 1488, 1466, 1100
1
cm-1; H NMR (500 MHz, C6D6, rt) δ 7.61 (d, J ) 6.6 Hz, 1H),
7.29 (d, J ) 7.4 Hz, 2H), 7.01 (t, J ) 7.2 Hz, 2H), 6.94 (dd, J )
10.7, 7.4 Hz, 2H), 6.87 (t, J ) 7.7 Hz, 2H), 6.83 (td, J ) 6.8, 1.5
Hz, 2H), 6.75 (td, J ) 7.8, 1.2 Hz, 1H), 6.40 (d, J ) 7.7 Hz, 1H),
6.25-6.35 (m, 4H), 6.19 (dd, J ) 7.2, 1.4 Hz, 2H), 5.40 (dd, J )
9.2, 1.9 Hz, 1H), 5.01-5.06 (m, 1H), 4.97 (d, J ) 16.2 Hz, 1H),
4.88 (d, J ) 15.8 Hz, 1H), 4.71 (d, J ) 15.5 Hz, 1H), 4.25 (dd, J
) 12.6, 9.6 Hz, 1H), 3.85 (d, J ) 16.2 Hz, 1H), 2.00 (q, J ) 6.3
Hz, 1H), 1.03 (s, 9H), 0.14 (s, 3H), 0.10 (s, 3H); 13C NMR (125
MHz, C6D6, rt) δ 175.8, 174.7, 144.9, 143.7, 137.1, 137.0, 136.0,
131.4, 129.9, 129.4, 129.3, 129.2, 129.1, 128.7, 127.9, 127.8, 127.4,
126.8, 126.4, 126.0, 122.2, 121.7, 110.1, 109.6, 67.5, 56.5, 53.6,
45.2, 43.5, 34.9, 26.5, 18.8, -3.6, -3.7; LRMS (ESI, m/z) 627.62
(M + H)+, 649.59 (M + Na)+; HRMS (ESI, m/z) calcd for
1
room temperature, the H NMR spectrum of the crude reaction
mixture was recorded at 100 °C in toluene-d8. Column chroma-
tography (88:12-80:20, hexanes/Et2O) afforded 32.0 mg (52%) of
4 as a colorless foam, 3.7 mg (6%) of 29 as a colorless solid, and
8.5 mg (14%) of 30 as a colorless oil. No starting material was
recovered. The analytic data for 4 and 30 matched that previously
reported.2
Data for 4: IR (film) 2954, 2929, 2858, 1719, 1652, 1611, 1360,
1250, 1171 cm-1; 1H NMR (500 MHz, C6D6, rt) δ 8.43 (br s, 1H),
7.29 (br s, 2H), 7.05 (br s, 2H), 6.82-6.70 (m, 7H), 6.76 (br s,
1H), 6.15-6.43 (m, 5H), 4.93-5.10 (m, 3H), 4.64-4.80 (m, 2H),
4.26 (br s, 1H), 3.80 (br d, J ) 15.0 Hz, 1H), 2.01 (br d, J ) 13.0
1
C40H42N2O3Si, 649.2863 (M + Na)+; found, 649.2852. The H
NMR spectrum was also recorded in toluene-d8 at 100 °C.28
Heck Cyclization of Triflate 23. According to the general
procedure, triflate 23 (242 mg, 304 µmol), Pd(PPh3)2Cl2 (42.6 mg,
J. Org. Chem, Vol. 71, No. 7, 2006 2607

Overman and Watson
60.8 µmol), Et3N (307 mg, 0.42 mL, 3.04 mmol), and DMA (3.0
mL) were heated to give a 75:25 mixture of epimers 34 and 35.
Column chromatography (90:10-85:15, hexanes/EtOAc) afforded
96.6 mg (64%) of 34 as a colorless solid and 28.7 mg (19%) of 35
as a colorless foam.
15.8 Hz, 1H), 4.74 (d, J ) 15.8 Hz, 1H), 3.85-3.96 (m, 2H), 3.39-
2.49 (m, 1H), 2.20-2.32 (m, 1H), 1.37 (dd, J ) 13.3, 6.9 Hz, 1H);
13C NMR (125 MHz, toluene-d8, rt) δ 176.9, 175.3, 144.6, 143.5,
137.5, 137.3, 136.9, 136.1, 132.5, 131.6, 130.2, 129.2, 128.6, 128.3,
127.3, 127.1, 126.6, 126.2, 125.5, 125.3, 122.0, 121.4, 109.33,
109.26, 53.1, 52.6, 44.8, 43.2, 25.1, 22.7; LRMS (ESI, m/z) 497.26
(M + H)+, 519.24 (M + Na)+; HRMS (ESI, m/z) calcd for
C34H28N2O2, 497.2229 (M + Na)+; found, 497.2245.
Data for 34: IR (film) 3029, 2916, 1704, 1609, 1488, 1466, 1360,
1173 cm-1; 1H NMR (500 MHz, toluene-d8, rt) δ 7.46 (d, J ) 7.4
Hz, 2H), 6.93-7.03 (m, 10H), 6.67-6.72 (m, 2H), 6.58-6.64 (m,
2H), 6.22 (ddd, J ) 9.9, 4.8, 2.5 Hz, 1H), 6.14 (d, J ) 7.7 Hz,
1H), 5.34 (dt, J ) 9.9, 1.9 Hz, 1H), 4.60 (d, J ) 15.7 Hz, 2H),
4.50-4.55 (m, 2H), 3.00-3.11 (m, 1H), 2.24 (dt, J ) 18.4, 4.9
Hz, 1H), 1.53 (dd, J ) 13.9, 6.9 Hz, 1H); 13C NMR (125 MHz,
toluene-d8, rt) δ 176.4, 176.0, 143.4, 143.0, 136.5, 136.3, 132.2,
131.0, 130.7, 129.2, 128.8, 128.7, 128.5, 128.23, 128.16, 127.7,
127.6, 127.5, 127.4, 126.2, 125.1, 123.9, 108.8, 108.6, 53.5, 50.4,
44.0, 43.4, 26.8, 22.3; LRMS (ESI, m/z) 497.25 (M + H)+, 519.23
(M + Na)+; HRMS (ESI, m/z) calcd for C34H28N2O2, 497.2229
(M + Na)+; found, 497.2231.
Data for 35: IR (film) 3029, 2912, 1715, 1609, 1488, 1466, 1349,
1177 cm-1; 1H NMR (500 MHz, toluene-d8, rt) δ 7.53 (d, J ) 6.9
Hz, 1H), 7.30 (d, J ) 7.4 Hz, 2H), 7.00-7.04 (m, 2H), 6.91-6.96
(m, 2H), 6.82-6.90 (m, 4H), 6.77 (t, J ) 7.7 Hz, 1H), 6.22-6.40
(m, 5H), 6.11 (d, J ) 7.4 Hz, 1H), 6.02 (dt, J ) 9.7, 3.4 Hz, 1H),
5.45 (d, J ) 9.9 Hz, 1H), 4.93 (d, J ) 16.0 Hz, 1H), 4.84 (d, J )
Acknowledgment. This research was supported by the NIH
National Institutes of General Medical Sciences (GM-30859)
and by predoctoral fellowships from Pharmacia Co. and the UC
Regents for D.A.W. NMR and mass spectra were determined
at UC Irvine using instruments acquired with the assistance of
NSF and NIH shared instrumentation grants.
Supporting Information Available: Complete experimental
details for the preparation of the substrates and the chemical
1
correlation of the products, copies of H and 13C NMR spectra of
all new compounds, and NOE data used in the assignment of the
configurations of the products. This material is available free of
charge via the Internet at http://pubs.acs.org.
JO0523363
2608 J. Org. Chem., Vol. 71, No. 7, 2006