Development of a formal [4+1] cycloaddition: Pd(OAc)2-catalyzed intramolecular cyclopropanation of 1,3-dienyl β-keto esters and MgI 2-promoted vinylcyclopropane-Cyclopentene rearrangement

Full text of DOI:10.1021/ja809226x

Published on Web 02/04/2009
Development of a Formal [4 + 1] Cycloaddition: Pd(OAc)2-Catalyzed
Intramolecular Cyclopropanation of 1,3-Dienyl ꢀ-Keto Esters and
MgI2-Promoted Vinylcyclopropane-Cyclopentene Rearrangement
Rockford W. Coscia and Tristan H. Lambert*
Department of Chemistry, Columbia UniVersity, New York, New York 10027
Received November 25, 2008; E-mail: tl2240@columbia.edu
Cycloaddition reactions are unrivaled in their power to rapidly
construct complex molecular architectures. While classic cycload-
ditions such as the Diels-Alder reaction have long been a mainstay
of organic synthesis, the use of transition metal reagents has served
to greatly increase the scope of cycloaddition chemistry.1 Indeed,
otherwise inaccessible modalities such as [4+4],2 [5+2],3 and
[2+2+2]4 cycloadditions among many others have been achieved
by way of metal catalysis. Remarkably, however, examples of [4+1]
cycloadditions remain relatively underdeveloped, despite the obvi-
ous synthetic advantages that accompany their realization.5 To be
sure, given the conceptual analogy to [4+2] reactions and the
prevalence of five-membered rings in biologically relevant organic
architectures, the development of new [4+1] cycloadditions has
the potential for a significant impact in the area of complex molecule
synthesis.
The salient points of our cyclopropanation development studies
are shown in Table 1. To begin, we found that the use of substrates
bearing gem-dimethyl substitution in the 4-position was necessary
for optimal results due to the propensity of unsubstituted substrates
to undergo Saegusa-Ito type oxidation.9 In our first experiments,
treatment of ꢀ-keto ester 3 with 40 mol% Pd(OAc)2 and Cu(OAc)2/
O2 co-oxidant in DMSO at elevated temperatures did not produce
any of the desired cyclopropane 4 (entry 1). Hypothesizing that
the concentration of the nucleophilic enol tautomer of 3 was
insufficiently low under these conditions, we attempted to employ
base additives, but without success (entry 2).
Table 1. Optimization Studies for Intramolecular Cyclopropanation
of 1,3-Dienyl ꢀ-Keto Esters
The value of [4+1] cycloadditions has not escaped the attention of
synthetic chemists, and several strategies have been disclosed over
the past 30 years aimed at realizing this important goal. Included in
this number are direct carbene-diene cycloadditions,5a-c stepwise
mol %
Pd(OAc)2
co-oxidant
(equiv)
additive
(equiv)
temp time yield
(°C)
entry
(h)
(%)
carbenoid-ketene annulations,5d and transition metal-catalyzed
e
carbonyl-diene insertions.5f-m Perhaps the most versatile strategies
to produce [4+1] cycloadducts have been advanced by Hudlicky5n-p
and Danheiser,5q-s whereby 1,3-dienes are subjected to cyclopropa-
nation followed by vinylcyclopropane-cyclopentene (VCP-CP) rear-
rangement. This ingenious strategy allows for the leverage of powerful
cyclopropanation technologies while circumventing the mechanistic
difficulties associated with direct [4+1] reaction. Still, both the
Hudlicky and Danheiser approaches require the use of R-diazocarbo-
nyls or the generation of reactive carbenes; require harsh conditions
for rearrangement; and are not easily lent to broad generality. There
thus remains a strong need for more practical and comprehensive
solutions to the [4+1] cycloaddition problem.
Our group is pursuing the development of an alternative [4+1]
protocol based on the cyclopropanation/rearrangement paradigm
(eq 1). Specifically, we are interested in the development of
oxidative coupling6 between acidic methylene moieties (e.g., ꢀ-keto
esters) and 1,3-dienes to produce vinylcyclopropanes directly. In
conjunction with VCP-CP rearrangement, this oxidative strategy
could serve as a powerful platform for the development of a broad
range of [4+1] cycloadditions. Inspired by the work of Ba¨ckvall,7
who has pioneered oxidative nucleophilic additions to 1,3-dienes,
we decided to explore the use of Pd(II)-catalysis to achieve this
goal.8 Herein we describe the development of the first direct metal-
catalyzed cyclopropanation of 1,3-dienes and its application to a
formal [4+1] cycloaddition protocol.
1
2
3
4
5
6
7
8
40
40
40
20
20
10
10
10
Cu(OAc)2 (0.2)/O2
-
65
65
rt
48
48
18 35
48 25
18 39
12 50
-
-
Cu(OAc)2 (0.2)/O2 K2CO3 (2)
Cu(OAc)2 (0.2)/O2 Mg(ClO4)2 (1)
Cu(OAc)2 (0.2)/O2 Mg(ClO4)2 (1)
Cu(OAc)2 (0.2)/O2 Mg(ClO4)2 (1)
Cu(OAc)2 (0.1)/O2 Mg(ClO4)2 (1)
Cu(OAc)2 (2.5)
Cu(O2CiPr)2 (2.5) Mg(ClO4)2 (1) 65
rt
40
65
65
Mg(ClO4)2 (1)
8
52
12 92
a
1
Diastereomeric ratios were determined by GC or H NMR analysis.
On the other hand, with the addition of 1 equiv of Mg(ClO4)2,10
we observed the formation of the desired vinylcyclopropane 4 in
35% yield after 18 h at room temperature (entry 3). Decreasing
the loading of Pd(OAc)2 to 20 mol% resulted in increased reaction
time and decreased yield (entry 4), which could be remedied by
raising the reaction temperature to 40 °C (entry 5). A further
increase of temperature to 65 °C allowed for lowering of the Pd
catalyst loading to 10 mol% (with either sub- or superstoichiometric
copper co-oxidant) while shortening the reaction time and increasing
the yield to ∼50% (entries 6 and 7). Finally, by simply changing
the co-oxidant from Cu(OAc)2 to 2.5 equiv of Cu(O2CiPr)2, we
were able to achieve highly efficient formation of 4 in 92% yield
with greater than 10:1 diastereoselectivity (entry 8).11 In this case,
substoichiometric Cu(O2CiPr)2/O2 was less efficient.
The dramatic improvement in yield that followed from alteration
of the carboxylate ligands was due to suppression of formation of
a significant (g20% yield) side product identified to be dienyl
acetate 8 (Figure 1). Our rationale for the formation of 8 is shown
as part of our overall mechanistic hypothesis.12 Thus we propose
that keto ester substrate 3 coordinates to both the Pd(II) catalyst
and Mg(ClO4)2 to generate intermediate 5. Intramolecular attack
9
2496 J. AM. CHEM. SOC. 2009, 131, 2496–2498
10.1021/ja809226x CCC: $40.75
2009 American Chemical Society

C O M M U N I C A T I O N S
Table 2. Scope Studies of Pd(OAc)2-Catalyzed Intramolecular
Cyclopropanation of 1,3-Dienyl ꢀ-Keto Estersa
of the enol moiety to the Pd-bound diene would then, with loss of
RCO2H, produce π-allyl intermediate 6. A second nucleophilic
attack of the enol function to the π-allyl moiety produces the desired
4, with concomitant reduction of Pd (thus requiring reoxidation to
complete the catalytic cycle). When the carboxylate ligand of
intermediate 6 is acetate, reductive elimination to form allyl ester
7 is a competitive process and leads to formation of the diene 8
upon further oxidation. On the other hand, when the carboxylate
ligand of 6 is isobutyrate, the reductive elimination pathway is
inhibited, allowing for exclusive formation of the desired 4. The
reason for the divergent behavior of these two ligands is at present
unclear.
a See Supporting Information for details on substrate synthesis.
Reactions were run in the presence of 10 mol% Pd(OAc)2, 2.5 equiv of
Cu(O2CiPr)2, and 1 equiv of Mg(ClO4)2 at 0.1 M in DMSO at 65 °C.
b Diastereomeric ratios were determined by GC. c Diastereomeric ratio
determined by 1H NMR analysis; the minor diastereomer is epimeric at
all cyclopropyl carbons. d 50% conversion.
Figure 1. Pd(OAc)2-Catalyzed Intramolecular Cyclopropanation of 1,3-
Dienyl ꢀ-Keto Esters
In anticipation of our [4+1] process, we have probed the scope of
this intramolecular cyclopropanation method (Table 2). In addition to
monosubstituted dienyl substrate 3 (entry 1), we have found that
substitution at the 2 or 4 position of the diene is readily accommodated,
and the corresponding cyclopropyl adducts can be prepared in high
yields and with excellent diastereoselectivities (entries 2 and 3). The
facile production of the cyclopropane product in entry 3 is especially
noteworthy given that two of the three cyclopropyl carbons are
quaternary. Notably, we have found that incorporation of a useful
functional handle in the form of a silyloxymethyl substituent could be
achieved in high yield and with excellent stereoselectivity (entry 4).
Furthermore, cyclopropanation could be effected in reasonable yield
with gem-dimethyl substitution in the δ-position to block undesired
substrate oxidation (entry 5). We suggest the lower efficiency of this
substrate is due to 1,3-diaxial strain between a methyl substituent and
the π-allyl Pd group in the transition state for the cyclopropane-forming
event. Interestingly, efficient cyclization could be achieved even in a
relatively complex setting, such as with a substrate derived from estrone
(entry 6), although the diastereoselectivity in this case was relatively
poor. Finally, we found that our method was viable for the production
of even highly congested cyclopropanes such as that shown in entry
7, albeit with a yield significantly lower than that with other substrates.
The poor efficiency of this cyclization is perhaps not surprising,
considering the steric strain experienced by the tricyclic product.
With cyclopropanation conditions in hand, we next sought to
transform our vinylcyclopropanes into the corresponding formal
[4+1] cycloadducts. To our dismay, none of the standard condi-
tions13 (pyrolysis, transition metal reagents, or Lewis acids) for
vinylcyclopropane-cyclopentene rearrangement were effective for
our substrates. We thus sought to identify a workable new protocol.
Because of the mildness and simplicity such a protocol would
offer, we were especially intrigued by the notion of employing metal
iodide reagents to effect VCP-CP rearrangement.14 A screen of
metal iodide salts under various conditions revealed that MgI2 in
CH3CN at 40 °C efficiently promoted the rearrangement of a
number of our vinylcyclopropane adducts (Table 3).15 Despite the
potential sensitivity to the iodide ion, these conditions were
sufficiently mild so as to leave the ꢀ-keto ester moiety intact (entries
1-5). Vinyl substitution was well tolerated (entries 2 and 4), as
was an allylic silyl ether moiety (entry 4). This protocol was also
effective at promoting rearrangement of the complex estrone-derived
substrate (entry 5). It should be noted that these conditions were
not productive for the rearrangement of the products shown in
entries 3 and 7 in Table 2, indicating further development of this
protocol is required.
The mechanism by which we presume magnesium iodide effects
VCP-CP rearrangement is shown in eq 2. Thus the iodide ion
undergoes homoconjugate addition to the vinylcyclopropane sub-
strate 9, producing an allyl iodide enolate intermediate 10. The only
alkylative pathway available to the (E)-10 isomer would be that of
the SN2′ reaction, reforming the starting vinylcyclopropane. To the
extent that (Z)-10 is formed, however, this isomer can undergo
intramolecular SN2 alkylation to produce the cyclopentene [4+1]
cycloadduct 11.
9
J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009 2497

C O M M U N I C A T I O N S
Table 3. Formal [4 + 1] Products via MgI2-Promoted
Vinylcyclopropane-Cyclopentene Rearrangementa
1987, 28, 2221. (c) Wender, P. A.; Ihle, N. C. Tetrahedron Lett. 1987, 28,
2451. (d) Wender, P. A.; Ihle, N. C.; Correia, C. R. D. J. Am. Chem. Soc.
1988, 110, 5904. (e) Wender, P. A.; Tebbe, M. J. Synthesis 1991, 1089.
Pd: (f) Murakami, M.; Itami, K.; Ito, Y. Synlett 1999, 951. Ru: Itoh, K.;
Masuda, K.; Fukahori, T.; Nakano, K.; Aoki, K.; Nagashima, H. Organo-
metallics 1994, 13, 1020. See also: (g) Kaupp, G. Angew. Chem., Int. Ed.
1992, 31, 422.
(3) For a review, see: Wender, P. A.; Gamber, G. G.; Williams, T. J. In Modern
Rhodium-Catalyzed Organic Reactions; Evans, P. A., Ed.; Wiley-VCH:
Weinheim, Germany, 2005; Chapter 13.
(4) For reviews, see:(a) Tanaka, K. Synlett 2007, 1977. (b) Shibata, T.;
Tsuchikama, K. Org. Biomol. Chem. 2008, 6, 1317. (c) Chopade, P. R.;
Louie, J. AdV. Synth. Catal. 2006, 348, 2307. (d) Gandon, V.; Aubert, C.;
Malacria, M. Chem. Commun. 2006, 2209. (e) Robinson, J. E. In Modern
Rhodium-Catalyzed Organic Reactions; Evans, P. A., Ed.; Wiley-VCH:
Weinheim, Germany, 2005; p 129. (f) Kotha, S.; Brahmachary, E.; Lahiri,
K. Eur. J. Org. Chem. 2005, 4741.
(5) Examples of [4 + 1] cycloadditions: (a) Lilienblum, W.; Hoffman, R. W.
Chem. Ber. 1977, 110, 3405. (b) Spino, C.; Rezaei, H.; Dupont-Gaudet,
K.; Be´langer, F. J. Am. Chem. Soc. 2004, 126, 9926. (c) Boisvert, L.;
Beaumier, F.; Spino, C. Org. Lett. 2007, 9, 5361. (d) Loebach, J. L.; Bennett,
D. M.; Danheiser, R. L. J. Am. Chem. Soc. 1998, 120, 9690. (e) Rigby,
J. H.; Wang, Z. Org. Lett. 2003, 5, 263. (f) Eaton, B. E.; Rollman, B.
J. Am. Chem. Soc. 1992, 114, 6245. (g) Sigman, M. S.; Kerr, C. E.; Eaton,
B. E. J. Am. Chem. Soc. 1993, 115, 7545. (h) Sigman, M. S.; Eaton, B. E.
J. Org. Chem. 1994, 59, 7488. (i) Morimoto, T.; Chatani, N.; Murai, S.
J. Am. Chem. Soc. 1999, 121, 1758. (j) Murakami, M.; Itami, K.; Ito, Y.
J. Am. Chem. Soc. 1997, 119, 2950. (k) Murakami, M.; Itami, K.; Ito, Y.
J. Am. Chem. Soc. 1999, 121, 4130. (l) Murakami, M.; Itami, K.; Ito, Y.
Organometallics 1999, 18, 1326. (m) Oshita, M.; Yamashita, K.; Tobisu,
M.; Chatani, N. J. Am. Chem. Soc. 2005, 127, 761. (n) Hudlicky, T.;
Koszyk, F. J.; Kutchan, T. M.; Sheth, J. P. J. Org. Chem. 1980, 45, 5020.
(o) Hudlicky, T.; Koszyk, F. J.; Dochwat, D. M.; Cantrell, G. L. J. Org.
Chem. 1981, 46, 2911. (p) Short, R. P.; Revol, J.-M.; Ranu, B. C.; Hudlicky,
T. J. Org. Chem. 1983, 48, 4453. (q) Danheiser, R. L.; Bronson, J. J.;
Okano, K. J. Am. Chem. Soc. 1985, 107, 4579. (r) Danheiser, R. L.;
Martinez-Davilla, C.; Auchus, R. J.; Kadonaga, J. T. J. Am. Chem. Soc.
1981, 103, 2443. (s) Danheiser, R. L.; Martinez-Davilla, C.; Morin, J. M.,
Jr J. Org. Chem. 1980, 45, 1340.
(6) For related examples, see:(a) Pei, T.; Widenhoefer, R. A. J. Am. Chem.
Soc. 2001, 123, 11290. (b) Pei, T.; Wang, X.; Widenhoefer, R. A. J. Am.
Chem. Soc. 2003, 125, 648 See also ref 9.
(7) (a) Ba¨ckvall, J. E. In Metal-Catalyzed Cross Coupling Reactions; Stang,
P., Diederich, F., Eds.; Wiley-VCH: Weinheim, 1998; p 339. (b) Ba¨ckvall,
J. E. Acc. Chem. Res. 1983, 16, 335. (c) Nilsson, Y. I. M.; Gatti, R. G. P.;
Andersson, P. G.; Ba¨ckvall, J. E. Tetrahedron 1996, 52, 7511. (d) Lo¨fstedt,
J.; Na¨rhi, K.; Dorange, I.; Ba¨ckvall, J. E. J. Org. Chem. 2003, 68, 7243.
(e) Ro¨nn, M.; Andersson, P. G.; Ba¨ckvall, J. E. Tetrahedron Lett. 1997,
38, 3603.
a Reactions were run in the presence of 150 mol% MgI2 at 0.2 M in
CH3CN at 40 °C.
Importantly, although the MgI2 promoted VCP-CP rearrange-
ment is not compatible with DMSO as solvent, the vinylcyclopro-
panation is operative in acetonitrile, thus raising the hope that this
two-step [4+1] protocol can be rendered into a one-pot procedure.
This possibility has not yet been confirmed experimentally.
In conclusion, we have developed a new formal [4+1] cycload-
dition protocol for the conversion of 1,3-dienyl ꢀ-keto esters to
bicyclic cyclopentene products. The key components of this strategy
are the development of a Pd(II)-catalyzed intramolecular cyclo-
propanation reaction and a new mild VCP-CP rearrangement
promoted by MgI2. Currently, we are working to (1) extend this
strategy to include other nucleophilic moieties, (2) remove the need
for a gem-dimethyl blocking group, and (3) render this sequence
into a one-pot protocol.
(8) Andersson and Ba¨ckvall have reported hetero [4 + 1] reactions using Pd(II)
catalysis. Andersson, P.; Ba¨ckvall, J.-E. J. Am. Chem. Soc. 1992, 114, 8696.
(9) Ito, Y.; Hirao, T.; Saegusa, T. J. Org. Chem. 1978, 43, 1011.
(10) Yang has reported the use of Mg(ClO4)2 in related chemistry. (a) Yang,
D.; Gao, Q.; Lee, C.-S.; Cheung, K.-K Org. Lett. 2002, 4, 3271. (b) Yang,
D.; Li, J.-H.; Gao, Q.; Yan, Y.-L. Org. Lett. 2003, 5, 2869. (c) Yip, K.-T.;
Li, J.-H.; Lee, O.-Y.; Yang, D. Org. Lett. 2005, 7, 5717.
(11) The use of catalytic Cu(O2CiPr)2 was less efficient.
(12) When the reaction show in Table 1, entry 8, was conducted without
Pd(OAc)2, no product was observed.
Acknowledgment. We thank the Parkin Group for obtaining a
crystal structure, and the NSF (CHE-0619638) is thanked for
acquisition of an X-ray diffractometer.
(13) (a) Baldwin, J. E. Chem. ReV. 2003, 103, 1197. (b) Wang, S. C.; Tantillo,
D. J. J. Organomet. Chem. 2006, 691, 4386. (c) Corey, E. J.; Myers, A. G.
J. Am. Chem. Soc. 1985, 107, 5574.
Supporting Information Available: Experimental procedures and
product characterization data. This material is available free of charge
via the Internet at http://pubs.acs.org.
(14) For a VCP-CP rearrangement involving LiI, see: Hashimoto, S.-i.; Shinoda,
T.; Ikegami, S Tetrahedron Lett. 1986, 27, 2885. This method was also
ineffective with our substrates. .
(15) Carreira and co-workers have utilized MgI2 for the ring opening of
cyclopropyl carbonyls. (a) Alper, P. B.; Meyers, C.; Siegel, D. R.; Carreira,
E. M. Angew. Chem., Int. Ed. 1999, 38, 3186. (b) Marti, C.; Carreira, E. M.
J. Am. Chem. Soc. 2005, 127, 11505. (c) Lerchner, A.; Carreira, E. M.
Chem.sEur. J. 2006, 12, 8208.
References
(1) For selected reviews, see:(a) Fruhauf, H.-W. Chem. ReV. 1997, 97, 523.
(b) Lautens, M.; Klute, W.; Tam, W. Chem. ReV. 1996, 96, 49.
(2) For selected examples, see:(a) Wender, P. A.; Ihle, N. C. J. Am. Chem.
Soc. 1986, 108, 4678. (b) Wender, P. A.; Snapper, M. L. Tetrahedron Lett.
JA809226X
9
2498 J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009