Features and applications of [Rh(CO)2Cl]2-catalyzed alkylations of unsymmetrical allylic substrates

Article abstract of DOI:10.1021/jo701290b

(Chemical Equation Presented) A novel regio- and stereoselective [Rh(CO)₂Cl]₂-catalyzed allylic alkylation of unsymmetrical allylic carbonates was discovered. The regioselectivity of the reaction favors product ratios in which substitution occurs at the carbon bearing the leaving group. When an enantiomerically enriched carbonate (≥99% ee) was examined, the Rh(I)-catalyzed allylic alkylation proceeded stereoselectively to provide the alkylation product with retention of absolute stereochemistry (98% ee). To establish the scope of the [Rh(CO)₂Cl]₂-catalyzed allylic alkylation, a variety of carbon and heteroatom nucleophiles were examined and the results described. As an application of the Rh(I)-catalyzed allylic alkylation, a series of novel domino reactions have been developed that couple the unique regio- and stereoselective [Rh(CO)₂Cl]₂- catalyzed alkylation of allylic trifluoroacetates with an intramolecular Pauson-Khand annulation, a cycloisomerization, or a [5+2] cycloaddition. A unique aspect of the method described is the use of a single catalyst to effect sequential transformations in which the catalytic activity is moderated simply by controlling the reaction temperature. Implementation of such processes provides a rapid and efficient entry to a variety of bicyclic carbon skeletons from simple precursors.

Full text of DOI:10.1021/jo701290b

Features and Applications of [Rh(CO)2Cl]2-Catalyzed Alkylations of
Unsymmetrical Allylic Substrates
Brandon L. Ashfeld, Kenneth A. Miller, Anna J. Smith, Kristy Tran, and
Stephen F. Martin*
Department of Chemistry and Biochemistry, The UniVersity of Texas, Austin, Texas 78712
sfmartin@mail.utexas.edu
ReceiVed June 18, 2007
A novel regio- and stereoselective [Rh(CO)2Cl]2-catalyzed allylic alkylation of unsymmetrical allylic
carbonates was discovered. The regioselectivity of the reaction favors product ratios in which substitution
occurs at the carbon bearing the leaving group. When an enantiomerically enriched carbonate (g99%
ee) was examined, the Rh(I)-catalyzed allylic alkylation proceeded stereoselectively to provide the
alkylation product with retention of absolute stereochemistry (98% ee). To establish the scope of the
[Rh(CO)2Cl]2-catalyzed allylic alkylation, a variety of carbon and heteroatom nucleophiles were examined
and the results described. As an application of the Rh(I)-catalyzed allylic alkylation, a series of novel
domino reactions have been developed that couple the unique regio- and stereoselective [Rh(CO)2Cl]2-
catalyzed alkylation of allylic trifluoroacetates with an intramolecular Pauson-Khand annulation, a
cycloisomerization, or a [5+2] cycloaddition. A unique aspect of the method described is the use of a
single catalyst to effect sequential transformations in which the catalytic activity is moderated simply by
controlling the reaction temperature. Implementation of such processes provides a rapid and efficient
entry to a variety of bicyclic carbon skeletons from simple precursors.
Introduction
the sterically less hindered allylic terminus, irrespective of the
structure of the starting materials (e.g., 1 or 2), to yield
substitution products 3,1a,2 whereas Ru,3 Mo,4 Rh,5,6 Ir,7,8 and
W9 preferentially deliver products 4 arising from substitution
Transition metal-catalyzed allylic alkylations have had con-
siderable impact on organic synthesis, and such transformations
have been widely studied and applied since their discovery in
the mid-1970’s.1 Indeed, such reactions are arguably some of
the most versatile processes in organic chemistry as a wide range
of allylic and nucleophilic reaction partners may be efficiently
combined by using a variety of transition metal catalysts to form
products in a highly regiospecific and stereospecific manner.
The factors governing the regiochemical and stereochemical
outcomes in transformations involving symmetrical and unsym-
metrical allylic substrates have been explored extensively
revealing a number of general trends. For example, palladium-
catalyzed processes typically favor nucleophilic substitution at
(1) (a) Trost, B. M. Tetrahedron 1977, 33, 2615. (b) Trost, B. M. Acc.
Chem. Res. 1980, 13, 385. (c) Tsuji, J. Organic Synthesis with Palladium
Compounds; Springer-Verlag: Heidelberg, Germany, 1980. (d) Trost, B.
M.; Verhoeven, T. R. In ComprehensiVe Organometallic Chemistry;
Pergamon: Oxford, UK, 1982; Vol. 8, pp 799. (e) Tsuji, J.; Minami, I.
Acc. Chem. Res. 1987, 20, 140. (f) Godleski, S. A. In ComprehensiVe
Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford,
UK, 1991; Vol. 4, p 585. (g) Tsuji, J. Palladium Reagents and Catalysts:
InnoVations in Organic Synthesis; John Wiley & Sons: New York, 1995;
p 1. (h) Trost, B. M.; Van Vranken, D. L. Chem. ReV. 1996, 96, 395. (i)
Trost, B. M.; Heumann, A. In Transition Metals for Organic Synthesis;
Beller, M., Bolm, C., Eds.; Wiley-VCH: Weinheim, Germany, 1998; Vol.
1, p 3. (j) Trost, B. M.; Crawley, M. L. Chem. ReV. 2003, 103, 2921.
10.1021/jo701290b CCC: $37.00 © 2007 American Chemical Society
Published on Web 11/01/2007
9018
J. Org. Chem. 2007, 72, 9018-9031

[Rh(CO)2Cl]2-Catalyzed Allylic Alkylation
SCHEME 1
certain circumstances. It was thus with some excitement that
we recently discovered [Rh(CO)2Cl]2 catalyzes the allylic
alkylation of unsymmetrical allylic substrates with excellent
regio- and stereoselectivity to provide products arising from
substitution at the carbon atom bearing the leaving group,
irrespective of the structure of the starting carbonate. Herein
we report the details of these findings, together with their
application to the design of novel cascade reactions.12
Results and Discussion
Background to the Discovery. In the context of developing
concise strategies for the syntheses of hydroazulene sesquiter-
penes, we were intrigued by the observation that the cyclopro-
panations of divinyl diazoesters 8 to provide lactones 913 and
the [5+2] cycloadditions of enynes of the general structure 11
to yield bicyclo[5.3.0]decanes 1214 were both rhodium-catalyzed
processes (Scheme 3). It occurred to us that if the allylic
alkylation of 9 with malonate 10 could also be rhodium-
catalyzed, a novel cascade of reactions might be devised that
would rapidly transform structurally simple diallylic diazoesters
into complex hydroazulenes. Although Evans had previously
shown that allylic substitutions could in fact be promoted by a
modified Wilkinson’s catalyst, the major products of such
transformations resulted from nucleophilic attack at the more
substituted terminus of the allylic moiety.6b Literature precedent
thus argued persuasively that the regiochemical outcome of an
allylic alkylation involving 9 and 10 would be opposite of that
which we required. However, if a rhodium(I) catalyst were
capable of catalyzing the conversion of 9 to 11, the possibility
of inducing a subsequent intramolecular [5+2] cycloaddition
in situ was simply too attractive to ignore. Initial experiments
were thus conducted to assess the viability of this novel approach
to hydroazulenes.
at the more substituted allylic terminus regardless of whether 1
or 2 is the starting material (Scheme 1).1c,5,10
Transition metal-catalyzed nucleophilic substitutions of allylic
substrates are believed to proceed via metal-stabilized π-allyl
intermediates. The structure of such intermediates can vary along
the continuum between the η1-allyl complex 5, the enyl (σ+π)-
complex 6, and the η3-allyl complex 7 depending upon the
metal, its ligands, and the nature of the substituents R1-R4
(Scheme 2).1h,5a The regioselectivity of the ensuing nucleophilic
SCHEME 2
attack is therefore dictated by a combination of steric and
electronic factors that vary with the intermediate complex and
the nucleophile.11 However, it is generally possible to selectively
prepare either 3 or 4 through a judicious selection of catalyst
and allylic substrate 1 or 2. However, examination of the prior
art reveals that no single catalyst enables allylic substitution at
the carbon atom bearing the leaving group, such that the starting
material structure may map directly onto the structure of the
major product, a situation that might prove advantageous in
SCHEME 3
(2) (a) Pre´toˆt, R.; Pfaltz, A. Angew. Chem., Int. Ed. 1998, 37, 323. (b)
Helmchen, G.; Pfaltz, A. Acc. Chem. Res. 2000, 33, 336.
(3) (a) Kondo, T.; Ono, N.; Satake, N.; Mitsudo, T.-A.; Watanabe, Y.
Organometallics 1995, 14, 1945. (b) Morisaki, Y.; Kondo, T.; Mitsudo,
T.-A. Organometallics 1999, 18, 4742. (c) Trost, B. M.; Fraisse, P. L.;
Ball, Z. T. Angew. Chem., Int. Ed. 2002, 41, 1059. (d) Kawatsura, M.; Ata,
F.; Wada, S.; Hayase, H.; Uno, H.; Itoh, T. Chem. Commun. 2007, 298.
(4) (a) Trost, B. M.; Lautens, M. J. Am. Chem. Soc. 1982, 104, 5543.
(b) Trost, B. M.; Dogra, K.; Hachiya, I.; Emura, T.; Hughes, D. L.; Krska,
S. W.; Reamer, R. A.; Palucki, M.; Yasuda, N.; Reider, P. J. Angew. Chem.,
Int. Ed. 2002, 41, 1929.
(5) (a) Evans, P. A.; Nelson, J. D. J. Am. Chem. Soc. 1998, 120, 5581.
(b) Evans, P. A.; Leahy, D. K. J. Am. Chem. Soc. 2003, 125, 8974.
(6) (a) Tsuji, J.; Minami, I.; Shimizu, I. Tetrahedron Lett. 1984, 25, 5157.
(b) Evans, P. A.; Nelson, J. D. Tetrahedron Lett. 1998, 39, 1725. (c)
Takeuchi, R.; Kitamura, N. New J. Chem. 1998, 22, 659. (d) Hayashi, T.;
Okada, A.; Suzuka, T.; Kawatsura, M. Org. Lett. 2003, 5, 1713.
(7) Helmchen, G.; Dahnz, A.; Du¨bon, P.; Schelwies, M.; Weihofen, R.
Chem. Commun. 2007, 675.
Inasmuch as our primary interest was to gain access to
cyclopropyl enynes related to 11, we conducted preliminary
experiments in which the vinyl cyclopropyl lactone 13 was
treated with the substituted malonate anion 14 in the presence
of RhCl(PPh3)3/P(OMe)3 (eq 1).6b However, these initial efforts
failed to yield any alkylation product, even after extended
(8) (a) Takeuchi, R. Synlett 2002, 1954. (b) Takeuchi, R.; Kezuka, S.
Synthesis 2006, 3349.
(9) (a) Trost, B. M.; Hung, M.-H. J. Am. Chem. Soc. 1983, 105, 7757.
(b) Lehmann, J.; Lloyd-Jones, G. C. Tetrahedron 1995, 51, 8863.
(10) (a) Osborn, J. A.; Jardine, F. H.; Young, J. F.; Wilkinson, G. J.
Chem. Soc. A 1966, 1711. (b) Takeuchi, R.; Kashio, M. J. Am. Chem. Soc.
1998, 120, 8647. (c) Blacker, A. J.; Clarke, M. L.; Loft, M. S.; Mahon, M.
F.; Humphries, M. E.; Williams, J. M. J. Chem. Eur. J. 2000, 6, 353. (d)
Trost, B. M.; Andersen, N. G. J. Am. Chem. Soc. 2002, 124, 14320. (e)
Trost, B. M.; Jiang, C. Org. Lett. 2003, 5, 1563. (f) Bartels, B.; Garcia-
Yebra, C.; Helmchen, G. Eur. J. Org. Chem. 2003, 1097. (g) Murphy, K.
E.; Hoveyda, A. H. J. Am. Chem. Soc. 2003, 125, 4690.
(12) (a) Ashfeld, B. L.; Miller, K. A.; Martin, S. F. Org. Lett. 2004, 6,
1321. (b) Ashfeld, B. L.; Miller, K. A.; Smith, A. J.; Tran, K.; Martin, S.
F. Org. Lett. 2005, 7, 1661.
(13) (a) Martin, S. F.; Spaller, M. R.; Liras, S.; Hartmann, B. J. Am.
Chem. Soc. 1994, 116, 4493. (b) Doyle, M. P.; Austin, R. E.; Bailey, A. S.;
Dwyer, M. P.; Dyatkin, A. B.; Kalinin, A. V.; Kwan, M. M. Y.; Liras, S.;
Oalmann, C. J.; Pieters, R. J.; Protopopova, M. N.; Raab, C. E.; Roos, G.
H. P.; Zhou, Q.-L.; Martin, S. F. J. Am. Chem. Soc. 1995, 117, 5763.
(14) (a) Wender, P. A.; Dyckman, A. J. Org. Lett. 1999, 1, 2089. (b)
Wender, P. A.; Dyckman, A. J.; Husfield, C. O.; Kadereit, D.; Love, J. A.;
Rieck, H. J. Am. Chem. Soc. 1999, 121, 10442.
(11) Trost, B. M.; Hung, M.-H. J. Am. Chem. Soc. 1984, 106, 6837.
J. Org. Chem, Vol. 72, No. 24, 2007 9019

Ashfeld et al.
TABLE 1. [Rh(CO)2Cl]2-Catalyzed Allylic Alkylation of
Unsymmetrical Primary Carbonatesa
reaction times and elevated temperatures. Because the dimeric
rhodium(I) catalyst [Rh(CO)2Cl]2 was known to catalyze the
intramolecular [5+2] cycloadditions of cyclopropyl enynes,14a
we queried if perchance it might also induce allylic alkylations.
To address this question, a preliminary experiment was con-
ducted in which 13 and 14 were combined in the presence of
[Rh(CO)2Cl]2 to afford an E/Z-mixture (1:1) of 15 as the only
isolable product, albeit in 20% yield. This was an exciting and
important discovery that demonstrated the heretofore unknown
ability of [Rh(CO)2Cl]2 to catalyze allylic alkylations. Equally
intriguing was the observation that this reaction of an unsym-
metrical substrate proceeded with a regioselectivity opposite that
anticipated based upon considering examples found in the
literature.4 This fortuitous result prompted us to explore the
scope of the [Rh(CO)2Cl]2-catalyzed allylic alkylations with
more traditional substrates such as allylic acetates and carbon-
ates.
Reactions of Unsubstituted Malonates. In an initial series
of experiments, we examined the reactions of unsymmetrical
primary allylic carbonates 16a-h as well as secondary and
tertiary allylic carbonates 20a-l with sodiodimethyl malonate
(17) (Tables 1 and 2). Dimethyl malonate was chosen as the
nucleophilic partner in these preliminary experiments due to
its ready availability and ease of handling. However, a variety
of other nucleophiles proved to be viable partners in the
substitution reaction vida infra. We never observed bis-alkylation
products, a common and undesired byproduct in non-transition
metal-catalyzed SN2 reactions involving dimethyl malonate. The
corresponding allylic acetates were also found to be reactive
toward alkylation in the presence of [Rh(CO)2Cl]2. However,
compared to the corresponding allylic carbonates, the allylic
acetates required longer reaction times and provided the
substitution products in slightly diminished yields. In our initial
studies, we thus employed a methyl carbonate as the leaving
group to examine regio- and stereochemical trends in [Rh-
(CO)2Cl]2-catalyzed allylic substitutions.
a Conditions: 5 mol % of [Rh(CO)2Cl]2, 2.5 equiv of CH2(CO2Me)2,
2.0 equiv of NaH, 0.1 M in carbonate. b Isolated yields. c Ratios determined
by GLC. d THF, rt. e THF, 0 °C. f DMF, -20 °C. g DMF, rt. h Ratio of cis/
trans isomers.
Before embarking on an extensive examination of the scope
and limitations of [Rh(CO)2Cl]2-catalyzed allylic alkylations,
the various reaction parameters were optimized for the reaction
of 16a with the sodium salt of dimethyl malonate (17). This
reaction was highly regioselective, providing a mixture of the
alkylation products 18a and 19a in a 97:3 ratio. We first
conducted experiments to ascertain what solvent(s) would be
preferred and found that the allylic alkylation of 16a with 17
proceeded in good yields when DMF, THF, MeCN, or Et2O
was used as solvent. However, the reactions in DMF and THF
were much faster (1-2 h at room temperature) than those in
MeCN and Et2O (∼12 h at room temperature); reactions in
MeCN and Et2O were also less efficient, proceeding in 10-
15% lower yields. Reactions proceeded readily at concentrations
ranging from 0.01 to 0.1 M, and were generally performed at
0.1 M. Initial alkylations were conducted with 10 mol % of the
[Rh(CO)2Cl]2 dimer, but we found that lowering the catalyst
loading to 5 mol % did not adversely affect the efficiency of
the reaction, although the reaction was slightly slower, requiring
an additional 30-45 min to completely consume 16a. When
the catalyst loading was decreased to 1-2 mol %, yields of
alkylated product dropped to less than 20%. Although the
reactions proceeded with equal quantities of allylic substrate
and nucleophile, we found that the use of 2.5 equiv of dimethyl
malonate and 2.0 equiv of NaH as base was generally a good
starting point for further optimizations.
Having established the essential parameters for the reaction,
we conducted a series of studies to explore the scope and
regioselectivity of the [Rh(CO)2Cl]2-catalyzed allylic alkylations
of different allylic substrates, and these results are summarized
in Table 1. The reactions were typically performed by treating
the allylic substrate (1.0 equiv; 0.1 M) with dimethyl malonate
(2.5 equiv) and NaH (2. 0 equiv) in the presence of 5 mol % of
[Rh(CO)2Cl]2 in THF. However, if the reaction was sluggish
or proceeded with modest selectivity, DMF was used as the
solvent. The temperature was varied so that the reactions were
completed in a reasonable period of time.
9020 J. Org. Chem., Vol. 72, No. 24, 2007

[Rh(CO)2Cl]2-Catalyzed Allylic Alkylation
TABLE 2. [Rh(CO)2Cl]2-Catalyzed Allylic Alkylation of
Unsymmetrical Secondary and Tertiary Carbonatesa
Examination of the entries in Table 1 reveals a number of
important findings. Perhaps most striking is that primary allylic
carbonates having di- and trisubstituted carbon-carbon double
bonds provided linear alkylation products with generally >9:1
regioselectivity, even when the other terminus of the allylic
moiety is benzylic (entry 3). This mode of reactivity resembles
that normally observed with palladium catalysts,1j not rhodi-
um,5,6 iridium,7,8 molybdenum,4 or ruthenium3 catalysts, al-
though there are isolated reports of similar regioselectivities for
some rhodium5a and iridium8 catalysts. Indeed, that 16e (Entry
5) underwent any alkylation is noteworthy because 2,3,3-
trisubstituted allylic carbonates are inert to the modified
Wilkinson’s catalyst reported by Evans and typically require
forcing conditions when other transition metal catalysts are
used.5a,15 The observation that the Z-carbonate 16b (entry 2)
underwent alkylation with little carbon-carbon double bond
isomerization is noteworthy because extensive Z/E isomerization
of Z-allylic substrates normally occurs when other transition
metal catalysts are used, although there are some exceptions
involving iridium,8,10b,16 palladium,17 and tungsten18 catalysts.
The temperature and solvent played critical roles in maintaining
the Z-double bond geometry. For example, the Z/E-ratio fell to
86:13 when the reaction was conducted in THF at room
temperature instead of 0 °C. If the reaction was run in DMF at
temperatures even as low as -20 °C, the E-isomer was obtained
as the major product (>60:40). Finally, the presence of
conjugated enol ethers did not adversely affect the regiochem-
istry or efficiency of the reaction (entry 7), and propargylic
carbonates underwent substitution to give substituted alkynes
that were not contaminated with allenic products commonly
observed in palladium-catalyzed substitutions (entry 8).19
Secondary and tertiary allylic carbonates were found to be
excellent substrates in [Rh(CO)2Cl]2-catalyzed allylic alkyla-
tions, but substitution patterns influenced the regioselectivity
of the reaction to a greater extent than observed with primary
allylic carbonates (Table 2). For example, reactions of carbonates
and carboxylates derived from linear allylic carbinols tended
to give products resulting from substitution at the secondary
carbon atom bearing the leaving group, although the preference
was sometimes modest (entries 1, 2, 5, 6, 8, and 9). Conversely,
when the substituent attached to one terminus of the allylic
moiety was branched, carbon-carbon bond formation occurred
preferentially at the sterically less congested terminus (entries
3, 4, 7, 10, and 11). Comparing the results of entries 9 and 11
suggests that branching on the carbon atom attached to the allyl
moiety may be a more important controlling factor than
branching on a carbon atom contained within the allylic array.
Further support for this hypothesis is found in entry 12. Namely,
alkylation of the tertiary allylic carbonate 20l occurred at the
carbon atom bearing the leaving group even though a quaternary
center was generated in the process. In those cases examined,
changing solvent from THF to DMF and lowering the reaction
temperature sometimes led to modest increases in the observed
regioselectivity favoring alkylation of the carbon atom bearing
(15) (a) Takahashi, T.; Nemoto, H.; Tsuji, J. Tetrahedron Lett. 1983,
24, 2005. (b) Trost, B. M.; Lautens, M.; Chan, C.; Jebaratnam, J.; Mueller,
T. J. Am. Chem. Soc. 1991, 113, 636.
(16) Takeuchi, R. Polyhedron 2000, 19, 557.
(17) (a) Huntzinger, M. W.; Oehlschlager, A. C. J. Org. Chem. 1991,
56, 2918. (b) Sjogren, M. P. T.; Hansson, S.; Akermark, B. Organometallics
1994, 13, 1963. (c) Kazmaier, U.; Zumpe, F. L. Angew. Chem., Int. Ed.
2000, 39, 802.
(18) Frisell, H.; Akermark, B. Organometallics 1995, 14, 561.
(19) Keinan, E.; Bosch, E. J. Org. Chem. 1986, 51, 4006.
a Conditions: 5 mol % of [Rh(CO)2Cl]2, 2.5 equiv of CH2(CO2Me)2,
2.0 equiv of NaH, 0.1 M in carbonate. b Isolated yields. c Ratios determined
by GLC. d THF, rt. e DMF, -20 °C. f THF, 0 °C. g DMF, 0 °C. h The
corresponding carbonate was unstable.
J. Org. Chem, Vol. 72, No. 24, 2007 9021

Ashfeld et al.
the leaving group (entries 2, 4, 8, and 9). Interestingly,
alkylations of substrates bearing additional double bonds are
highly regioselective (entries 5 and 6), even when the alternate
mode of reaction would generate a conjugated product. Such
selectivity is not without precedent as it has been observed that
pendant olefins can direct the regiochemistry of transition metal-
catalyzed allylic alkylation reactions, presumably via a weak
coordination of the carbon-carbon double bond to the
metal.20
Stereochemistry of [Rh(CO)2Cl]2-Catalyzed Allylic Sub-
stitutions. To determine the stereochemical outcome of [Rh-
(CO)2Cl]2-catalyzed allylic substitutions with malonate nucleo-
philes, enantioenriched allylic carbonate (+)-23 was synthesized
in g99% ee in three steps from cinnamaldehyde via Sharpless
kinetic resolution of the intermediate allylic alcohol.21 When
allylic carbonate (+)-23 was treated with the sodium salt of
dimethyl malonate in the presence of [Rh(CO)2Cl]2, malonate
(+)-24 was obtained in 93% yield and 98% ee (regioselectivity
) 93:7) (eq 2). On the basis of this experiment, [Rh(CO)2Cl]2
appears to catalyze substitutions of secondary allylic carbonates
with net retention of configuration as has been observed with
Pd,1g,h,j Ru,3c Mo,22 Rh,5a and Ir10b catalysts.
centers were generated (entries 8 and 10). It is again noteworthy
that the substitution reaction of the Z-carbonate 16b (entry 2)
with 25a proceeded regioselectively to furnish enyne 26b with
minimal olefin isomerization. Inasmuch as rhodium(I) catalysts
are known to interact in a nonproductive fashion with terminal
acetylenes,26 it is significant that reactions involving 25b are
efficient.
Other Carbon Nucleophiles. The anions derived from
â-ketoesters and R-sulfonyl esters such as 28 and 29 were also
found to be competent nucleophiles in the [Rh(CO)2Cl]2-
catalyzed allylic alkylation (Table 4). The regiochemical trends
in these reactions parallel those with malonate nucleophiles in
which the predominant product arises from substitution at the
carbon bearing the leaving group, even when contiguous
quaternary centers are formed (entry 3).
We were also intrigued by the possibility that [Rh(CO)2Cl]2
might catalyze allylic substitutions of unstabilized carbon
nucleophiles. This inquiry was inspired in part by Evans, who
reported that allylic hexafluoroisopropyl carbonates underwent
regio- and stereoselective alkylation upon treatment with aryl
zinc reagents in the presence of TpRh(C2H4)2, LiBr, and
dibenzylidene acetone.27 We were thus pleased to find that when
enantioenriched methyl carbonate (+)-23 (g99% ee) was simply
treated with a premixed solution of PhLi and ZnBr2 in the
presence of [Rh(CO)2Cl]2 (5 mol %), the arylated product
(-)-32 was obtained in 99% yield (eq 3). Notably, the reaction
Reactions of Substituted Malonates. In accord with our plan
of developing approaches toward hydroazulene sesquiterpenes
and other cyclic ring systems, we initiated a brief survey of
[Rh(CO)2Cl]2-catalyzed allylic alkylations in which substituted
malonates were employed as nucleophiles. These studies, which
are summarized in Table 3, were focused on the homopropargyl
malonates 25a (R5 ) 2-butyn-1-yl) and 25b (R5 ) propynyl)
because the 1,6-enynes that would be formed as products were
known to be substrates for a variety of transition metal-catalyzed
reactions, including Pauson-Khand annulations,23 cycloisomer-
izations,24 and ring-closing metatheses.25 Additionally, malonate
25c (R5 ) n-Bu) was employed to examine what effect an alkyl
chain lacking a coordinating alkyne moiety would have on the
alkylation. The yields in these reactions were generally excellent,
again providing the substitution product arising from nucleo-
philic substitution at the carbon atom bearing the carbonate
group with excellent regioselectively. This regiochemical trend
persisted even in cases when two adjacent quaternary carbon
proceeded with >95:5 regioselectivity without significant loss
of enantiopurity (92% ee). Comparison of the optical rotation
of (-)-32 with the literature data showed that the reaction
proceeded with an overall net inVersion of absolute configura-
tion, an outcome consistent with nucleophilic attack of the aryl
zinc reagent on the allyl metal center followed by reductive
elimination.28
Allylic Alkylations of Oxygen and Nitrogen Nucleophiles.
Transition metal-catalyzed allylic etherifications and aminations
are well-known although the former are less commonly en-
countered in target directed synthesis.7,29 The regioselectivities
of these reactions vary with the nature of the transition metal
complex, but the general trends are similar to those observed
with carbon nucleophiles.30 For example, palladium-catalyzed
reactions of alcohol and amine nucleophiles tend to give
products arising from substitution at the less sterically hindered
allylic terminus,30b although this selectivity can be reversed when
C2-symmetric ligands are used to yield products in enantiose-
lective processes.30d Rhodium and iridium catalysts typically
(20) Krafft, M. E.; Sugiura, M.; Abboud, K. A. J. Am. Chem. Soc. 2001,
123, 9174.
(21) (a) Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.; Masamune,
H.; Sharpless, K. B. J. Am. Chem. Soc. 1987, 109, 5765. (b) Stary, I.;
Kocovsky, P. J. Am. Chem. Soc. 1989, 111, 4981.
(22) (a) Trost, B. M.; Lautens, M. J. Am. Chem. Soc. 1987, 109, 1469.
(b) Dvorak, D.; Stary, I.; Kocovsky, P. J. Am. Chem. Soc. 1995, 117, 6130.
(c) Lloyd-Jones, G. C.; Krska, S. W.; Hughes, D. L.; Gouriou, L.; Bonnet,
V. D.; Jack, K.; Sun, Y.; Reamer, R. A. J. Am. Chem. Soc. 2004, 126, 702.
(23) (a) Jeong, N.; Lee, S.; Sung, B. K. Organometallics 1998, 17, 3642.
(b) Brummond, K. M.; Kent, J. L. Tetrahedron 2000, 56, 3263. (c) Jeong,
N.; Sung, B. K.; Choi, Y. K. J. Am. Chem. Soc. 2000, 122, 6771. (d) Park,
K. H.; Son, S. U.; Chung, Y. K. Tetrahedron Lett. 2003, 44, 2827. (e)
Bon˜aga, L. V. R.; Krafft, M. E. Tetrahedron 2004, 60, 9795.
(24) (a) Cao, P.; Wang, B.; Zhang, X. J. Am. Chem. Soc. 2000, 122,
6490. (b) Kobayashi, T.; Koga, Y.; Narasaka, K. J. Organomet. Chem. 2001,
624, 73. (c) Tong, X.; Li, D.; Zhang, Z.; Zhang, X. J. Am. Chem. Soc.
2004, 126, 7601.
(26) Wender, P. A.; Sperandio, D. J. Org. Chem. 1998, 63, 4164.
(27) Evans, P. A.; Uraguchi, D. J. Am. Chem. Soc. 2003, 125, 7158.
(28) (a) Matsushita, H.; Negishi, E.-i. J. Chem. Soc., Chem. Commun.
1982, 160. (b) Fiaud, J. C.; Legros, J. Y. J. Org. Chem. 1987, 52, 1907.
(29) (a) Elliott, M. C.; Williams, E. J. Chem. Soc., Perkin Trans. 1 2001,
2303. (b) Negishi, E.-i. Handbook of Organopalladium Chemistry for
Organic Synthesis; John Wiley: New York, 2002; Vol. 2, p 1845.
(30) (a) Stanton, S. A.; Felman, S. W.; Parkhurst, C. S.; Godleski, S. A.
J. Am. Chem. Soc. 1983, 105, 1964. (b) Keinan, E.; Sahai, M.; Roth, Z.;
Nudelman, A.; Herzig, J. J. Org. Chem. 1985, 50, 3558. (c) Keinan, E.;
Seth, K. K.; Lamed, R. J. Am. Chem. Soc. 1986, 108, 3474. (d) Trost, B.
M.; McEachern, E. J.; Toste, F. D. J. Am. Chem. Soc. 1998, 120, 12702.
(e) Trost, B. M.; Tenaglia, A. Tetrahedron Lett. 1998, 29, 2931.
(25) Diver, S. T.; Giessert, A. J. Chem. ReV. 2004, 104, 1317.
9022 J. Org. Chem., Vol. 72, No. 24, 2007

[Rh(CO)2Cl]2-Catalyzed Allylic Alkylation
TABLE 3. Regioselectivity in the [Rh(CO)2Cl]2-Catalyzed Allylic Alkylation of Unsymmetrical Allylic Carbonates Utilizing r-Substituted
Malonate Nucleophilesa
a Conditions: 10 mol % of [Rh(CO)2Cl]2, 1.5 equiv of malonate 24, 1.4 equiv of NaH, 0.1 M in carbonate. b Isolated yields. c Ratios determined by GLC
1
or H NMR (400 MHz). d THF, rt. e THF, -20 °C. f DMF, -20 °C. g Ratio of cis/trans isomers.
lead to the formation of the more highly substituted alkylation
products.31 There is no transition metal catalyst that reliably
promotes allylic substitutions with oxygen or nitrogen nucleo-
philes so that the structure of the product maps directly onto
the starting material. It was thus a logical extension of our
studies of carbon nucleophiles to determine the regioselectivity
of [Rh(CO)2Cl]2-catalyzed alkylations of selected oxygen and
nitrogen nucleophiles (Table 5).
J. Org. Chem, Vol. 72, No. 24, 2007 9023

Ashfeld et al.
TABLE 4. Regioselectivity in the [Rh(CO)2Cl]2-Catalyzed Allylic Alkylation of Unsymmetrical Allylic Carbonates Utilizing Other Enolatesa
a Conditions: 10 mol % of [Rh(CO)2Cl]2, 1.5 equiv of nucleophile, 1.4 equiv of NaH, 0.1 M in carbonate, THF, rt. b Isolated yields. c Ratios determined
by GLC.
Phenols have been found to be excellent coupling partners
in transition metal-catalyzed etherifications.32 Consequently, we
first examined the [Rh(CO)2Cl]2-catalyzed reaction of carbonate
16a with the phenoxide ion generated from 33 by deprotonation
with LiHMDS or Et2Zn,33 but little of the desired ether was
formed. Following the lead of Evans31b,c and Hartwig,31d we
discovered that copper(I) phenoxides were excellent nucleo-
philes, and O-allylated phenols were obtained in good to
excellent yields and regioselectivities when primary and second-
ary disubstituted E-allylic carbonates were employed as sub-
strates (Table 5, entries 1-3). On the other hand, the more
highly substituted carbonates such as 16e and 20l were less
reactive with no appreciable product formation observed after
24 h at room temperature. We also found that [Rh(CO)2Cl]2-
catalyzed substitutions of primary allenic substrates as illustrated
by the efficient and regioselective coupling of carbonate 16i
with phenoxide 33 (entry 4). Although only exploratory experi-
ments were conducted, we were unable to induce allylations of
lithium or copper alkoxides derived from secondary alcohols.
Inasmuch as allylic amines are important synthetic intermedi-
ates, we then queried whether [Rh(CO)2Cl]2 would catalyze
alkylations of sulfonamide salts or secondary amines. Gratify-
ingly, we discovered that 16a underwent facile reaction with
the lithium salts 33c and 33d with good regioselectivity (entries
5 and 6). In preliminary experiments the corresponding reactions
of the secondary carbonate 20h with the salts of secondary
sulfonamides were found to be low yielding. Initial attempts to
induce the [Rh(CO)2Cl]2-catalyzed allylation of secondary
amines were unsuccessful, and the starting allylic carbonates
were recovered in near-quantitative yield. In thinking about what
modifications to the procedure we might explore, we were
inspired to examine the use of iodide ion as an additive based
upon reports of Lautens,34,3535 who has proposed that exchanging
the bridging chloride ion of [Rh(CO)2Cl]2 with an iodide ion
generates a more stable Rh(I)-I complex that is less prone to
react with an amine nucleophile to give an unreactive complex.
In this event, we discovered that allylic aminations with
pyrrolidine and methylbenzylamine as nucleophiles proceeded
readily and efficiently at room temperature in the presence of
tetrabutylammonium iodide (TBAI) (20 mol %) (entries 7-10).
The regioselectivity of the process was invariably excellent, but
dependent upon the nature of the starting allylic substrate.
Namely, the preferred product generally arose from nucleophilic
substitution at the carbon bearing the carbonate moiety (entries
6-8); however, amination of the dimethallyl substrate 16e
proceeded with the opposite regioselectivity (entry 10). The
allenic carbonate 16i also proved to be an excellent coupling
partner for allylic aminations, giving a nearly quantitative yield
of 34j with excellent regioselectivity (entry 11). The application
(31) (a) Takeuchi, R.; Ue, N.; Tanabe, K.; Yamashita, K.; Shiga, N. J.
Am. Chem. Soc. 2001, 123, 9525. (b) Evans, P. A.; Leahy, D. K. J. Am.
Chem. Soc. 2002, 124, 7882. (c) Evans, P. A.; Leahy, D. K.; Andrews, W.
J.; Uraguchi, D. Angew. Chem., Int. Ed. 2004, 43, 4788. (d) Shu, C.;
Hartwig, J. F. Angew. Chem., Int. Ed. 2004, 43, 4794.
(34) Lautens, M.; Fagnou, K.; Yang, D. J. Am. Chem. Soc. 2003, 125,
(32) Takacs, J. M. In ComprehensiVe Organometallic Chemistry II; Abel,
E. W., Stone, F. G. A., Wilkinson, G., Eds.; Elsevier Science: New York,
1995; Vol. 12, p 814.
14884.
(35) (a) Widenhoefer, R. A.; Buchwald, S. L. Organometallics 1996,
15, 2755. (b) Fagnou, K.; Lautens, M. Angew. Chem., Int. Ed. 2002, 41,
26.
(33) Kim, H.; Lee, C. Org. Lett. 2002, 4, 4369.
9024 J. Org. Chem., Vol. 72, No. 24, 2007

[Rh(CO)2Cl]2-Catalyzed Allylic Alkylation
TABLE 5. Regioselectivity in the [Rh(CO)2Cl]2-Catalyzed Allylic Alkylation of Oxygen and Nitrogen Nucleophiles with Unsymmetrical Allylic
Carbonatesa
a Conditions: 10 mol % of [Rh(CO)2Cl]2, 2.0 equiv of nucleophile, 0.1 M in carbonate. b Isolated yields. c Ratios determined by GLC. d LiHMDS (1.0
M in THF), CuI, THF, rt. e LiHMDS (1.0 M in THF), THF, rt. f TBAI (20 mol %), DCE, rt.
J. Org. Chem, Vol. 72, No. 24, 2007 9025

Ashfeld et al.
SCHEME 4
ates 38 and 39 is slow relative to nucleophilic attack, then the
structure of the preferred product will correlate more closely
with the substitution pattern in the starting material. The relative
stabilities of enyl intermediates 38 and 39, the rates of their
equilibration (k1 and k-1), and the relative rates of nucleophilic
attack on these complexes (k2 and k3) will be dictated by a
combination of steric and electronic effects. Although predicting
how the interplay of these factors influences the regioselectivity
of the allylic alkylation is a complex matter, a few observations
may be made. The modified Wilkinson’s catalysts used by Evans
have phosphite ligands that are more electron donating than the
CO ligands in the [Rh(CO)2Cl]2 complex, so the metal center
in the latter complex is more electron deficient.37 Since we find
that [Rh(CO)2Cl]2-catalyzed reactions preferentially furnish
products in which the nucleophile becomes bonded to the allylic
carbon atom bearing the leaving group, increasing the Lewis
acidity of rhodium appears to result in slower equilibration of
the enyl intermediates 38 and 39 and/or an increased rate of
nucleophilic attack on these complexes. Supporting this hy-
pothesis, Evans has observed that allylic alkylations using
Wilkinson’s catalyst modified by triphenylphosphite, which is
less electron-donating than trimethylphosphite, also tend to give
products in which substitution occurs at the carbon atom bearing
the leaving group.5a
Although we find that modest steric factors do not play a
significant role in influencing the regiochemistry of nucleophilic
attack on rhodium enyl (σ+π) intermediates such as 38 and
39, steric effects are not inconsequential. For example, substitu-
tion reactions of the allylic substrates 20c, 20d, and 20k did
not proceed preferentially to provide the expected regioisomer
as the major product (Table 2, entries 3, 4, and 11). In these
cases, it appears that branching on the carbon atoms attached
to the π-allyl fragment influences the site of nucleophilic attack.
The presence of unbranched alkyl groups on the allylic moiety
does not appear to have as large an effect on regioselectivity
(Table 2, entries 8 and 9). On the other hand, the degree of
substitution on the three carbon atoms comprising the π-allyl
moiety does not seem to impart any steric influence on the
regioselectivity of the alkylation reaction (Table 2, entry 12).
The differential effects of substituents on the allylic starting
material is consistent with nucleophilic attack onto the π-allyl
metal complex at an oblique angle, but this proposal must be
considered speculative in the absence of additional information.
Synthetic Applications. We have thus demonstrated that [Rh-
(CO)2Cl]2 has an uncommon propensity to catalyze allylic
substitutions at the carbon atom bearing the leaving group on a
variety of structurally different substrates. The question that
naturally arises is whether this regioselectivity can be exploited
in synthesis. We are currently pursuing this query on a number
of fronts, and some preliminary results will be presented herein.
The observation that allylic substitutions of Z-alkenes pro-
ceeded with retention of double bond geometry suggested that
[Rh(CO)2Cl]2 might be used to catalyze cyclizations to give
medium and large rings containing Z-olefins. That [Rh(CO)2Cl]2
might be especially well-suited to such constructions was based
upon the fact that most metal-stabilized π-allyl species exist
preferentially as anti complexes that lead to E-olefins. Inasmuch
as the synthesis of eight-membered rings via transition metal-
catalyzed allylic alkylations is particularly demanding,38 we first
examined whether eight-membered lactones might be formed.
In the event, stirring the salt of the â-ketoester 42 with
of this Rh(I)-catalyzed allylic amination reaction could find
utility in the synthesis of complex alkaloid natural products,
and we are currently exploring several such applications, the
results of which will be reported in due course.
Mechanistic Considerations. These investigations show that
[Rh(CO)2Cl]2-catalyzed alkylations of carbon-, oxygen-, and
nitrogen-based nucleophiles with unsymmetrical allylic sub-
strates are unique. No other transition metal catalyst has been
reported to catalyze these reactions regioselectively to give
products arising predominantly from new bond formation at the
allylic carbon atom bearing the leaving group, irrespective of
the structure of the allylic starting material. The structure of
the starting material thus maps directly onto that of the product
in the majority of [Rh(CO)2Cl]2-catalyzed alkylations we have
studied. This phenomenon may be regarded as a type of
“memory effect”, but it is distinctly different from the classical
examples involving palladium-catalyzed allylic alkylations of
enantioenriched and racemic secondary allylic substrates having
the same number of substituents on the allyl moiety.36 Of course,
the observed regiochemical trend in these [Rh(CO)2Cl]2-
catalyzed allylations is not completely independent of allylic
substitution as steric effects do seem to play a role in certain
situations, especially in the case of secondary substrates. In this
context, it is noteworthy that Evans has observed a similar
memory effect in several symmetrically substituted systems
involving carbon-based nucleophiles using a RhCl(PPh3)3/
P(OPh)3 complex.5a
Mechanistic studies must be conducted to understand the
origin of the regioselectivity in [Rh(CO)2Cl]2-catalyzed allylic
alkylations and why it differs from analogous reactions promoted
by other transition metal catalysts. However, one can set forth
some general hypotheses if one accepts the mechanism proposed
by Evans for allylic alkylations catalyzed by modified Wilkin-
son’s catalysts (Scheme 4) as being applicable to our results.5a
Oxidative additions of a rhodium(I) complex to simple allylic
substrates of the general types 36 and 37 are thus envisioned to
generate the corresponding enyl (σ+π) intermediates 38 and
39 that then undergo nucleophilic attack to generate the observed
products 40 and 41, respectively. Inasmuch as Evans found that
the branched isomers 40 were generally found to be the favored
products irrespective of the substitution of the allylic substrate,
he proposed that the equilibrium between 38 and 39 favored
the former.
Accepting the general mechanistic features outlined in
Scheme 4, it follows that if equilibration of the enyl intermedi-
(36) (a) Fiaud, J. C.; Malleron, J. L. Tetrahedron Lett. 1981, 22, 1399.
(b) Hayashi, T.; Kawatsura, M.; Uozumi, Y. J. Am. Chem. Soc. 1998, 120,
1681. (c) Poli, G.; Scolastico, C. Chemtracts: Org. Chem. 1999, 12, 837.
(d) Malkov, A. V.; Gouriou, L.; Lloyd-Jones, G. C.; Stary´, I.; Langer, V.;
Spoor, P.; Vinader, V.; Kocovsky´, P. Chem. Eur. J. 2006, 12, 6910.
(37) Tolman, C. A. J. Am. Chem. Soc. 1970, 92, 2953.
9026 J. Org. Chem., Vol. 72, No. 24, 2007

[Rh(CO)2Cl]2-Catalyzed Allylic Alkylation
SCHEME 5
[Rh(CO)2Cl]2 in DMF at 0 °C provided lactone 43 in 68% yield
(eq 4) with none of the corresponding six-membered lactone
being observed. Although a palladium-catalyzed cyclization of
a â-sulfonyl ester to give an eight-membered lactone is
known,38a the formation of 43 represents to our knowledge the
first example of a transition metal-catalyzed allylic alkylation
of a â-ketoester to give an eight-membered lactone. When this
same reaction was conducted with Pd(PPh3)4 as catalyst, a
mixture (60:40) of eight- and six-membered lactones was
obtained. In a related experiment directed toward constructing
a carbocyclic seven-membered ring, we discovered that the [Rh-
(CO)2Cl]2-catalyzed cyclization of 44 in DMF at room tem-
perature gave exclusively the O-alkylation product 45 in 71%
yield (eq 5); none of the desired carbocycle 46 was obtained.
SCHEME 6
One of the initial driving forces that led us to explore [Rh-
(CO)2Cl]2-catalyzed allylic alkylations was the desire to develop
new and concise strategies for the synthesis of hydroazulenes
as outlined in Scheme 3.43 Toward this objective, initial studies
were focused upon an allylic alkylation/[5+2] cycloaddition
cascade involving carbonate 16d and malonate 25a to yield 26c
as the first step en route to cycloadduct 57 (Scheme 6). Although
the first step occurred smoothly, we were unable to induce the
[5+2] cycloaddition in the same pot. A variety of solvents and
temperatures were examined to no avail.
We then queried whether the methyl carbonate anion formed
as a product in the first step of the reaction might somehow
inhibit the [5+2] cycloaddition of 26c. Consistent with this
hypothesis, we discovered that when the acetate derivative of
16d was employed as the alkylating agent we were able to
isolate 57, albeit in poor yield. Because we had previously found
that allylic acetates reacted more sluggishly than their carbonate
we switched to trifluoroacetate 58, which proved to be an
excellent substrate for the cascade reaction. Accordingly,
treatment of 58 with 1.5 equiv of malonate 25a in the presence
of [Rh(CO)2Cl]2 (5 mol %) at room temperature gave interme-
diate cyclopropyl 1,6-enyne 26c that underwent facile [5+2]
cycloaddition to give 57 by simply elevating the bath temper-
ature to 80 °C (eq 6). Similarly, subjecting allylic trifluoroac-
etates 59 and 61 to the same conditions provided cycloadducts
60 and 62, respectively, in excellent overall yields (eqs 7 and
8). The diastereoselectivity and regiochemical trends observed
in these [5+2] cycloadditions are in accord with precedent
established by Wender.14a
Novel [Rh(CO)2Cl]2-Catalyzed Cascade Reactions. The
development of methods that transform relatively simple starting
materials into structurally complex intermediates with high
efficiency and atom economy is one of the major goals of
contemporary synthetic organic chemistry. In the context of
transition metal-catalyzed reactions, one tactic to achieve this
objective is to design one-pot, multistep sequences wherein the
product of the first serves as the starting material for the second,
and where each transformation is catalyzed by the same
transition metal.39 Hence, the discovery and development of
multifunctional catalysts that can be utilized to promote
mechanistically discrete cascade or domino reactions has
emerged as a worthy endeavor.12b It was well-known that [Rh-
(CO)2Cl]2 and other rhodium(I) complexes catalyze cyclizations
of 1,6-enynes via [5+2] cycloadditions,40 Pauson-Khand
annulations (PKR),23b,d,24b and cycloisomerizations.41,24a,b It
occurred to us that we might couple the [Rh(CO)2Cl]2-catalyzed
allylic alkylation of an R-alkynyl substituted malonate with such
cyclizations to enable the rapid formation of complex molecular
architectures from simple starting materials according to the
basic plan adumbrated in Scheme 5. Indeed, other sequential
reaction processes in which an allylic substitution serves as the
initial construction have been reported.42
(38) (a) Trost, B. M.; Verhoeven, T. R. J. Am. Chem. Soc. 1980, 102,
4743. (b) Trost, B. M. Angew. Chem., Int. Ed. Engl. 1989, 28, 1173. (c)
Trost, B. M.; Vos, B. A.; Brzezowski, C. M.; Martina, D. P. Tetrahedron
Lett. 1992, 33, 717.
(39) (a) Tietze, L. F. Chem. ReV. 1996, 96, 115. (b) Nicolaou, K. C.;
Edmonds, D. J.; Bulger, P. G. Angew. Chem., Int. Ed. 2006, 45, 7134.
(40) (a) Wender, P. A.; Husfeld, C. O.; Langkopf, E.; Love, J. A.; Pleuss,
N. Tetrahedron 1998, 54, 7203. (b) Wender, P. A.; Dyckman, A. J.; Husfeld,
C. O.; Scanio, M. J. C. Org. Lett. 2000, 2, 1609. (c) Trost, B. M.; Shen, H.
C. Org. Lett. 2000, 2, 2523.
Seeking to expand the breadth of [Rh(CO)2Cl]2-catalyzed
cascade reactions, we explored the feasibility of a process
commencing with an allylic alkylation and terminating with a
Pauson-Khand reaction. If successful, this reaction would
provide a facile entry to bicyclic cyclopentenones. We examined
(41) (a) Trost, B. M.; Krische, M. J. Synlett 1998, 1. (b) Lloyd-Jones,
G. C. Org. Biomol. Chem. 2003, 1, 215.
(42) (a) Evans, P. A.; Robinson, J. E. J. Am. Chem. Soc. 2001, 123,
4609. (b) Yamamoto, Y.; Nakagai, Y.-i.; Itoh, K. Chem. Eur. J. 2004, 10,
231.
(43) (a) Ashfeld, B. L.; Martin, S. F. Org. Lett. 2005, 7, 4535. (b) Ashfeld,
B. L.; Martin, S. F. Tetrahedron 2006, 62, 10497.
J. Org. Chem, Vol. 72, No. 24, 2007 9027

Ashfeld et al.
be orchestrated. We were thus gratified to discover that reaction
of the allylic trifluoroacetate 65 with malonate 25a in the
presence of [Rh(CO)2Cl]2 (10 mol %) gave the expected 1,6-
enyne (Z/E > 95:5), and when the reaction mixture was simply
heated to 110 °C (bath temperature) in a sealed tube, the diene
66 was isolated in 72% yield (eq 11). The corresponding allylic
methyl carbonate 15b was not a suitable substrate for this one-
pot allylic alkylation/cycloisomerization tandem sequence. It was
necessary to use an excess of 65 (2.5 equiv), rather than an
excess of the malonate as required in other domino sequences,
to ensure reasonable yields in the overall reaction.
this sequence by reacting allyl trifluoroacetate 63 with 1.3 equiv
of malonates 25a, 25b, and 25d in the presence of [Rh(CO)2Cl]2
(10 mol %) and a CO atmosphere. Gratifyingly, the 1,6-enyne
intermediates formed after the allylic alkylation underwent
cyclization upon heating to furnish bicyclic enones 64a-c in
good overall yields (eq 9). Additionally, we found that the
subsequent [2+2+1] cycloaddition could be effected through
microwave heating. Treatment of allylic trifluoroacetate 63b
with the malonate 25d in the presence of [Rh(CO)2Cl]2 under
an atmosphere of CO followed by microwave heating yielded
cyclopentenone 64d in moderate yield (eq 10). These transfor-
mations represent an attractive complement to a related tandem
Rh(I)-catalyzed allylic alkylation/Pauson-Khand annulation
protocol reported by Evans that led to bicyclo[3.3.0]octenones
having a different substitution pattern.42a
Although we found that the cyclic allylic trifluoroacetate 67
could be converted to 68 in a similar cascade sequence (eq 12),
this reaction proved to be rather capricious, despite extensive
efforts to optimize the process. The initial allylic alkylation failed
with allylic carbonates and acetates, and was sluggish even with
the highly reactive trifluoroacetate as a leaving group. Not only
is the allylic alkylation step problematic, but the high temper-
ature and reaction times required to effect cycloisomerization
facilitate double bond isomerization, leading to inseparable
mixtures of olefinic isomers. Isomerization could be minimized
by the use of microwave heating, often yielding 68 as the major
isomer, but the conjugated isomer 69 was always present.
Although yields of 68 as high as 65% were sometimes achieved,
the most reproducible yields were in the range of 40-
45%.
Inspired by the work of Ito and Brummond,44 we then queried
whether the scope of [Rh(CO)2Cl]2-catalyzed cascade allylic
alkylations and carbocyclizations might be extended to allenic
substrates. In preliminary experiments, we found that allenic
trifluoroacetates were difficult to handle, and allenic acetates
were unreactive toward the initial alkylation. Although allenic
carbonates underwent the initial [Rh(CO)2Cl]2-catalyzed sub-
stitution, the resultant products could not be induced to undergo
cycloisomerization in the same operation. Considering our prior
experiences with the use of carbonate leaving groups in domino
sequences, this was not a surprising result. Ultimately, allenic
trichloroacetates were found to be suitable reaction partners for
the alkylation step. Toward implementing the cascade sequence,
allenic trichloroacetate 70 was treated with 1.5 equiv of malonate
71 in the presence of [Rh(CO)2Cl]2 (10 mol %), but the
subsequent cycloisomerization of the intermediate allenyl olefin
to give 73 was not successful (eq 13).We reasoned that the
cycloisomerization might be facilitated by pre-generating an
open coordination site on the metal center,24a and after some
experimentation we found that addition of PPh3 and AgSbF6
prior to the alkylation step led to the formation of 73 in a
moderate 55% yield. Ito and co-workers have reported that the
While conducting exploratory experiments toward developing
this cascade of allylic alkylation and Pauson-Khand reactions,
we discovered that [Rh(CO)2Cl]2 catalyzed the isomerization
of 1,6-enynes to yield vinyl alkylidene cyclopentanes. This was
an interesting result as neutral Rh(I) species have not been
utilized effectively as enyne cycloisomerization catalysts. On
the other hand, cationic Rh(I) complexes are known to catalyze
the cycloisomerization of a variety of 1,6-enynes,24a,c but most
such examples involve Z-enynes. Indeed, we are aware of only
one example of the Rh(I)-catalyzed cycloisomerization of an
E-enyne, and in this case two products were obtained in
approximately equal amounts.24c
Having discovered that the [Rh(CO)2Cl]2-catalyzed allylic
substitutions of Z-substrates proceeds with little Z/E-isomer-
ization, it occurred to us that a novel cascade process involving
an allylic alkylation followed by a cycloisomerization might
(44) (a) Brummond, K. M.; Chen, H.; Mitasev, B.; Casarez, A. D. Org.
Lett. 2004, 6, 2161. (b) Makino, T.; Itoh, K. J. Org. Chem. 2004, 69, 395.
9028 J. Org. Chem., Vol. 72, No. 24, 2007

[Rh(CO)2Cl]2-Catalyzed Allylic Alkylation
stirred for 10 min at the indicated temperature (see text). In a
separate flask the pronucleophile (17, 25, 28, or 29) (1.5 mmol)
was added to a slurry of NaH (60% w/w in mineral oil, 1.4 mmol)
in THF (5 mL) at room temperature, and the mixture was stirred
for 20 min. The resulting sodium enolate was added via syringe to
the solution of allylic substrate and [Rh(CO)2Cl]2 at the indicated
temperature (see text). The mixture was sealed in a screw cap vial
under an atmosphere of argon, and stirring was continued until the
starting carbonate was consumed (as indicated by TLC analysis).
The reaction was then filtered through a short plug of silica gel
eluting with Et2O (50 mL), and the filtrate was concentrated under
reduced pressure. The crude residue was purified by flash chro-
matography, eluting with pentane/Et2O (5:1 or 10:1) to furnish a
mixture of 18/19, 21/22, (+)-23, or 26/27. The mixture was then
analyzed by either GLC or 1H NMR (400 MHz) (see text) to
determine the designated ratio of regioisomers.
efficiencies of allen-ene cycloisomerizations were improved
when the reaction was performed under an atmosphere of carbon
monoxide.44b To our delight, when the argon atmosphere was
exchanged with carbon monoxide following completion of the
allenic alkylation step, the yield of the overall process improved
to 72%. Extension of these optimized conditions to the reaction
of 70 with the anion of malonate 72 led to the production of 74
in 75% yield.
trans-2-Pent-2-enylmalonic Acid Dimethyl Ester (18a). Ma-
lonate 18a was obtained in 84% yield (0.34 mmol scale) after 2 h
in THF at room temperature as a clear, colorless oil after
chromatography (pentane/Et2O ) 5:1) in a 98:2 regioisomeric
ratio: 1H NMR (400 MHz) δ 5.56 (dddt, J ) 15.2, 6.8, 5.2, 1.6
Hz, 1 H), 5.34 (dddt, J ) 15.2, 7.2, 5.2, 1.6 Hz, 1 H), 3.73 (s, 3
H), 3.41 (t, J ) 7.6 Hz, 1 H), 2.58 (app dt, J ) 8.0, 1.2 Hz, 2 H),
2.02-1.95 (m, 2 H), 0.94 (t, J ) 7.2 Hz, 3 H); 13C NMR (100
MHz) δ 169.2, 135.5, 123.9, 52.4, 52.0, 31.9, 25.6, 13.8; IR (CDCl3)
2955, 1731, 1436, 1272, 1232, 1158 cm-1; mass spectrum (CI)
m/z 201.1117 [C10H17O4 (M + 1) requires 201.1127] 201 (base),
169.
cis-2-Pent-2-enylmalonic Acid Dimethyl Ester (18b). Malonate
18b was obtained in 86% yield (0.34 mmol scale) after 12 h in
THF at 0 °C as a clear, colorless oil after chromatography (pentane/
Et2O ) 5:1) in a 99:1 regioisomeric ratio and 97:3 cis/trans ratio:
1H NMR (400 MHz) δ 5.58 (dddt, J ) 10.4, 8.8, 7.2, 1.2 Hz, 1 H),
5.25 (dddt, J ) 10.8, 9.2, 7.6, 2.0 Hz, 1 H), 3.73 (s, 6 H), 3.39 (t,
J ) 7.2 Hz, 1 H), 2.67-2.63 (m, 2 H), 2.11-2.03 (m, 2 H), 0.96
(t, J ) 7.6 Hz, 3 H); 13C NMR (100 MHz) δ 169.2, 134.8, 123.6,
52.5, 51.8, 26.7, 20.6, 14.2; IR (CDCl3) 2956, 2258, 1732, 1437,
1240, 1158 cm-1; mass spectrum (CI) m/z 201.1124 [C10H17O4 (M
+ 1) requires 201.1127] 201 (base), 169.
General Procedure for the [Rh(CO)2Cl]2-Catalyzed Allylic
Alkylation with Phenolic Nucleophiles. A 1.0 M solution of
LiHMDS (0.66 mL, 0.66 mmol) was added to a slurry of phenol
33 (0.69 mmol) and CuI (132 mg, 0.69 mmol) in THF (1.5 mL) at
room temperature. The mixture was stirred at room temperature
for 30 min. In a separate flask, [Rh(CO)2Cl]2 (13 mg, 0.034 mmol)
was dissolved in THF (2 mL), stirred for 5 min at room temperature,
then transferred via syringe to the flask containing phenoxide.
Allylic carbonate 16 or 20 (0.34 mmol) was then added to the
mixture, and the reaction was stirred at room temperature for 24 h.
The mixture was filtered through a short plug of SiO2 eluting with
Et2O (50 mL). The eluent was concentrated under reduced pressure,
and the crude residue was purified by flash chromatography eluting
with pentane/Et2O (5:1) to provide aryl ether 34.
2-Phenyl-1-[(E)-pent-2-enyloxy]benzene (34a). Ether 34a was
obtained in 84% yield (0.34 mmol scale) in THF after 24 h at room
temperature as a clear, brown oil after chromatography (pentane/
Et2O ) 5:1) in a g95:5 regioisomeric ratio: 1H NMR (500 MHz)
δ 7.57-7.54 (comp, 2 H), 7.41-7.37 (comp, 2 H), 7.34-7.28
(comp, 3 H), 7.06-6.96 (comp, 2 H), 5.78 (dtt, J ) 15.4, 7.8, 1.4
Hz, 1 H), 5.60 (dtt, J ) 15.4, 5.6, 1.6 Hz, 1 H), 4.48 (ddt, J ) 5.6,
2.5, 1.6 Hz, 2 H), 2.08-2.02 (m, 2 H), 0.98 (t, J ) 7.5 Hz, 3 H);
13C NMR (125 MHz) δ 155.7, 138.6, 137.8, 136.1, 130.9, 129.6,
127.9, 126.8, 124.0, 121.0, 113.4, 69.4, 25.3, 13.3; IR (CHCl3) 2964,
1596, 1480, 1434, 1255, 1216, 912, 846 cm-1; mass spectrum (CI)
m/z 239.1426 [C17H19O1 (M + 1) requires 239.1436] 243, 239, 199,
171 (base).
Conclusions
In summary, we have discovered that the commercially
available rhodium(I) species [Rh(CO)2Cl]2 may be used to
catalyze the facile and highly regioselective allylic alkylation
of unsymmetrical substrates. Among transition metal-catalyzed
allylic substitution reactions, these processes are unique in the
sense that absent significant steric effects the major products
arise from substitution at the allylic carbon atom bearing the
leaving group, irrespective of the structure of the allylic starting
material. These reactions also proceed with a high level of
stereoselectivity in which enantioenriched starting materials
provide the substitution products with minimal loss of enan-
tiomeric excess. Additionally, the carbon-carbon double bond
geometry of primary Z-allylic substrates is maintained, a rare
mode of selectivity in the realm of transition metal-catalyzed
allylic alkylations. In addition to the unusual regioselectivity
in [Rh(CO)2Cl]2-catalyzed alkylations, there are several other
noteworthy features attending its use. The reactions may be
easily conducted without exercising stringent precautions to
exclude oxygen, and the mild conditions under which the
reaction is typically performed make it useful for thermally
sensitive substrates. The procedure is operationally simple
because additives or other ligands are not required for allylic
alkylations.
Because [Rh(CO)2Cl]2 is known to catalyze other transforma-
tions, a number of synthetic applications may be envisaged in
which allylic alkylations are combined with other transforma-
tions, resulting in an atom economical cascade process to rapidly
assemble complex structures from relatively simple starting
materials. This concept was illustrated by development of the
first cascade sequences involving allylic alkylation/[5+2] cy-
cloaddition and allylic alkylation/cycloisomerization. The method
was also applicable to a cascade allylic alkylation/Pauson-
Khand annulation that nicely complements existing procedures.
The ability to exploit multifunctional catalysts to promote two
or more sequential reactions in a single operation has significant
potential for the rapid preparation of structurally complex targets.
Studies to explore these and other [Rh(CO)2Cl]2-catalyzed
cascade reactions are in progress and will be reported in due
course.
Experimental Section
General Procedure for the [Rh(CO)2Cl]2-Catalyzed Allylic
Alkylation with Stabilized Carbon Nucleophiles. Allylic substrate
(16, 20, or 23) (1.0 mmol) was added to a solution of [Rh(CO)2Cl]2
(5 mol %) in THF or DMF (see text) (5 mL), and the solution was
1-Pent-2-enyloxy-2-vinylbenzene (34b). Ether 34b was obtained
in 77% yield (0.25 mmol scale) in THF after 24 h at room
temperature as a clear, colorless oil after chromatography (hexane)
J. Org. Chem, Vol. 72, No. 24, 2007 9029

Ashfeld et al.
in a g95:5 regioisomeric ratio: 1H NMR (400 MHz) δ 7.48 (dd,
J ) 7.2, 1.6 Hz, 1 H), 7.20 (dt, J ) 8.4, 1.6 Hz, 1 H), 7.09 (dd, J
) 17.6, 11.2 Hz, 1 H), 6.92 (t, J ) 7.6 Hz, 1 H), 6.86 (d, J ) 8.4
Hz, 1 H), 5.89 (dt, J ) 15.2, 6.4, Hz, 1 H), 5.74 (dd, J ) 17.6, 1.6
Hz, 1 H), 5.71 (m, 1 H), 5.24 (dd, J ) 11.6, 2.0 Hz, 1 H), 4.49
(dd, J ) 6.0, 1.2 Hz, 2 H), 2.11 (app p, J ) 6.4 Hz, 2 H), 1.03 (t,
J ) 7.6 Hz, 3 H); 13C NMR (100 MHz) δ 155.9, 136.6, 131.7,
128.7, 127.0, 126.4, 123.9, 120.6, 114.2, 112.4, 69.2, 25.3, 13.2;
IR (CHCl3) 3033, 2967, 2934, 2874, 1625, 1597, 1485, 1452, 1239,
1107, 1003, 969 cm-1; mass spectrum (CI) m/z 189.1278 [C17H19O1
(M + 1) requires 189.1279] 189 (base), 122, 107.
General Procedure for the [Rh(CO)2Cl]2-Catalyzed Allylic
Amination of Unsymmetrical Allylic Carbonates with Sulfona-
mides. [Rh(CO)2Cl]2 (5 mol %) was dissolved in dry, degassed
THF (5 mL), the allylic substrate 16 (1.0 mmol) was added, and
the solution was stirred for 30 min at room temperature. In a
separate flask, a 1.0 M solution of LiHMDS in THF (2.0 mmol)
was added to a solution of sulfonamide 33 (2.5 mmol) in THF (5
mL) and the mixture was stirred for 20 min at room temperature.
The resulting amide was added via syringe to the solution of allylic
substrate and [Rh(CO)2Cl]2 at room temperature. The mixture was
then sealed in a screw cap vial under argon and stirred for 2 h at
room temperature. The resulting dark brown solution was filtered
through a short plug of silica gel eluting with Et2O (50 mL), and
the combined filtrate and washings were concentrated under reduced
pressure. The crude residue was purified by flash chromatography
eluting with pentane/Et2O to provide the amination products in the
specified ratios.
1310, 1167, 965, 748, 692; MS (CI) m/z 202.1586 [C14H20N1 (M
+ 1) requires 202.1596] 202 (base), 186, 159, 131.
Benzyl-1,1-dimethylallylmethylamine (34i). Amine 34i was
obtained in 89% yield (0.25 mmol scale) as a yellow oil after
chromatography (hexanes/EtOAc ) 9:1) in g95:5 regioisomeric
ratio: 1H NMR (300 MHz) δ 7.60-7.20 (comp, 5 H), 6.03 (dd, J
) 17.7, 10.8 Hz, 1 H), 5.13 (dd, J ) 17.7, 1.5 Hz, 1 H), 5.09 (dd,
J ) 10.5, 1.5 Hz, 1 H), 3.52 (s, 2 H), 2.14 (s, 3 H), 1.25 (s, 6H);
13C NMR (75 MHz) δ 147.0, 141.3, 128.5, 128.1, 126.5, 112.0,
58.6, 55.7, 34.5, 22.8; IR (neat) 2973, 2842, 2794, 1494, 1453,
1411, 1355, 1181, 1001, 914, 696; mass spectrum (CI) m/z 190.1591
[C13H20N1 (M + 1) requires 190.1596] 190 (base), 122.
Procedure for the [Rh(CO)2Cl]2-Catalyzed Tandem Allylic
Alkylation/[5+2] Cycloaddition of Allylic Trifluoroacetates 58,
59, and 61. The allylic trifluoroacetate 58, 59, or 61 (0.1 mmol)
was added to a solution of [Rh(CO)2Cl]2 (5 mol %) in degassed
MeCN (0.5 mL), and the solution was stirred for 10 min at room
temperature. In a separate flask, malonate 25a (1.5 mmol) was
added to a slurry of NaH (1.4 mmol, 60% w/w in mineral oil) in
degassed MeCN (0.5 mL) at room temperature, and this mixture
was stirred for 20 min. The resulting sodium enolate was then added
via syringe to the solution of the allylic trifluoroacetate and [Rh-
(CO)2Cl]2 at room temperature. The mixture was sealed in a screw
cap vial under an atmosphere of argon, and stirring was continued
at room temperature until the starting material was consumed (6-8
h, as indicated by TLC analysis). The reaction was then heated to
80 °C (bath temperature) and stirring was continued until intermedi-
ate enyne was consumed (as indicated by TLC analysis). The
reaction was then filtered through a short plug of silica gel eluting
with Et2O (50 mL), and the combined filtrate was concentrated
under reduced pressure. The crude residue was purified by flash
chromatography, eluting with pentane/Et2O (5:1 or 10:1) to furnish
57, 60, or 62 that gave spectral characteristics consistent with those
reported in the literature.14a
Procedure for the [Rh(CO)2Cl]2-Catalyzed Tandem Allylic
Alkylation/Pauson-Khand Annulation of 63. Malonate 25a, 25b,
or 25d (0.26 mmol) was added to a suspension of NaH (60%
dispersion in mineral oil, 0.22 mmol) in dry, degassed THF (1 mL)
and the reaction was stirred for 15 min at room temperature. The
solvent was removed under reduced pressure, and the resulting
yellow oil was azeotroped with toluene (2 × 3 mL) under reduced
pressure. The residue was dissolved in THF (1 mL) and transferred
via syringe to a solution of 63 (0.2 mmol) and [Rh(CO)2Cl]2 (10
mol %) in THF (1 mL). The reaction was stirred at room
temperature for 1 h, and then the solution was heated under reflux
until the starting material was consumed (12-24 h) (as indicated
by TLC analysis). The solvent was removed under reduced pressure
and the residue was purified by flash chromatography eluting with
hexanes/EtOAc (3:1) to furnish 64a-c that gave spectral charac-
teristics consistent with those reported in the literature.46
Procedure for the [Rh(CO)2Cl]2-Catalyzed Tandem Allylic
Alkylation/Cycloisomerization of Allylic Trifluoroacetate 65 To
Furnish (3E)-Dimethyl 3-Ethylidene-4-((E)-prop-1-enyl)cyclo-
pentane-1,1-dicarboxylate (66). 65 (169 mg, 0.67 mmol) was
added to a solution of [Rh(CO)2Cl]2 (10 mol %) in degassed MeCN
(0.75 mL), and the solution was stirred for 10 min at -40 °C. In
a separate flask, 25a (50 mg, 0.27 mmol) was added to a slurry of
NaH (16 mg, 0.41 mmol, 60% dispersion in mineral oil) in MeCN
(2.0 mL) at room temperature, and the mixture was stirred for 30
min. The resulting solution of sodium enolate was then added via
syringe to the solution of 65 and [Rh(CO)2Cl]2 at -40 °C. The
reaction mixture was stirred for 6 h at -40 °C, whereupon it was
transferred to a screw cap vial that was sealed and heated at
110 °C (bath temperature) until the intermediate enyne 26b was
consumed (30 min). The reaction mixture was diluted with H2O (2
mL) and Et2O (2 mL), and the layers were separated. The aqueous
phase was extracted with Et2O (5 × 2 mL), and the combined
(E)-N-Benzyl-4-methyl-N-(pent-2-enyl)benzenamine (34e). Sul-
fonamide 34e was obtained in 71% yield (0.34 mmol scale) after
24 h at room temperature as a clear, colorless oil after chromatog-
1
raphy (pentane/Et2O ) 5:1) in a 82:18 regioisomeric ratio: H NMR
(500 MHz) δ 7.73 (d, J ) 6.8 Hz, 2 H), 7.57-7.54 (comp, 7 H),
5.42 (dddt, J ) 12.4, 6.4, 5.6, 1.6 Hz, 1 H), 5.06 (dddt, J ) 12.4,
6.8, 5.6, 1.2 Hz, 1 H), 4.32 (s, 2 H), 3.69 (br d, J ) 5.6 Hz, 2 H),
2.44 (s, 3 H), 1.91-1.86 (m, 2 H), 0.84 (t, J ) 6.0 Hz, 3 H); 13C
NMR (125 MHz) δ 143.1, 137.7, 137.6, 136.3, 129.6, 128.4, 128.4,
127.2, 122.0, 50.0, 48.9, 25.1, 21.5, 13.1; mass spectrum (CI) m/z
330.1531 [C19H24NO2S (M + 1) requires 330.1528] 330 (base),
274.
(E)-4-Methyl-N-(pent-2-enyl)-N-(prop-2-ynyl)benzeneamine
(34f). Sulfonamide 34f was obtained in 42% yield (0.14 mmol scale)
after 2 h at room temperature as a clear, colorless oil after
chromatography (pentane/Et2O ) 5:1) in a 82:12 regioisomeric
ratio. Spectral data are consistent with those reported in the
literature.45
General Procedure for the [Rh(CO)2Cl]2-Catalyzed Allylic
Amination of Unsymmetrical Allylic Carbonates with Secondary
Amines. Pyrrolidine (36 mg, 0.50 mmol) was added to a solution
of 23 (52 mg, 0.25 mmol), TBAI (19 mg, 0.050 mmol), and [Rh-
(CO)2Cl]2 (10 mg, 0.025 mmol) in dichloroethane (1 mL). The
reaction was stirred for 12 h at room temperature. The reaction
was concentrated under reduced pressure, and hexane (1 mL) was
added. The heterogeneous mixture was filtered through Celite
washing with hexane, and the combined filtrate and washings were
concentrated under reduced pressure. The residue was purified by
flash chromatography by using silica deactivated with 10% Et3N
eluting with the indicated solvent.
1-(1-Methyl-3-phenylallyl)pyrrolidine (34h). Amine 34h was
obtained in 99% yield (0.25 mmol scale) as a yellow oil after
chromatography (hexanes/EtOAc ) 1:1) in g95:5 regioisomeric
ratio: 1H NMR (400 MHz) δ 7.40-7.00 (comp, 5 H), 6.45 (d, J )
15.6 Hz, 1 H), 6.22 (dd, J ) 15.6, 7.0 Hz, 1 H), 2.88 (dt, J ) 14.8,
6.4 Hz, 1 H), 2.56 (comp, 4 H), 1.77 (comp, 4 H), 1.27 (d, J )
7.0, 3 H); 13C NMR (100 MHz) δ 137.2, 134.0, 129.6, 128.5, 127.2,
126.2, 63.1, 52.2, 23.3, 21.0; IR (neat) 2967, 2780, 1494, 1446,
(45) Boehmer, J.; Grigg, R.; Marchbank, J. D. Chem. Commun. 2002,
768.
(46) McDonald, F. E.; Wei, X. Org. Lett. 2002, 4, 593.
9030 J. Org. Chem., Vol. 72, No. 24, 2007

[Rh(CO)2Cl]2-Catalyzed Allylic Alkylation
organic fractions were dried (MgSO4) and concentrated under
reduced pressure. The crude residue was purified by flash chro-
matography eluting with pentane/Et2O (5:1) to furnish 49 mg (72%)
of 66 as a pale yellow oil: 1H NMR (500 MHz) δ 5.50-5.41 (app
ddq, J ) 15.7, 6.4, 0.6 Hz, 1 H), 5.52-5.22 (comp, 2 H), 3.74 (s,
3 H), 3.73 (s, 3 H), 3.04-2.98 (comp, 2 H), 2.85-2.80 (m, 1 H),
2.52-2.48 (ddd, J ) 12.9, 7.4, 1.8 Hz, 1 H), 1.92-1.87 (dd, J )
12.9, 11.6 Hz, 1 H), 1.69-1.67 (dd, J ) 6.4, 1.6 Hz, 3 H), 1.61-
1.55 (comp, 3 H); 13C NMR (500 MHz) δ 172.5, 172.4, 141.4,
132.3, 126.8, 117.5, 58.5, 52.8, 52.7, 46.8, 40.9, 36.8, 17.9, 14.6;
mass spectrum (CI) m/z 252.1361 [C14H20O4 requires 252.1362]
253 (base), 252, 221, 193, 192.
Acknowledgment. We are grateful to the National Instituted
of General Medicine Sciences (GM 31077), the Robert A. Welch
Foundation, Pfizer, Inc., and Merck Research Laboratories for
their generous support of this research. We also thank Professor
Michael J. Krische (The University of Texas) for helpful
discussions.
Supporting Information Available: Specific experimental
procedures, spectral characterization, and copies of 1H NMR spectra
for all new compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
JO701290B
J. Org. Chem, Vol. 72, No. 24, 2007 9031