with AgBF4 provided a precatalyst that catalyzed the cyclo-
isomerization at elevated temperature (39%, 1a:3a =
5:95), however, in favor of the formation of the undesired
3a (entry 14). The Pd(II) complex generated from (dppe)-
PdCl2 and AgBF4 was not competent to catalyze the
desired cycloisomerization (entry 15). Finally, and assum-
ing a hydrido palladium(II) species as actual catalyst,2,6
Pd(dba)2 and HBF4 were pooled with the intent of pre-
paring a LnPdH complex (entry 15);9 in the event, cyclo-
isomerization was indeed observed but with an insignif-
icant selectivity in favor for 3a (30%, 1a:3a = 58:42).
Therefore, the siloxy-substituted hexadiene 2b was synthe-
sized and treated with [Pd(MeCN)4](BF4)2 in CH2Cl2
(Scheme 3). Disappointingly, however, even after varying
catalyst loading, temperature profile, reaction time, pro-
tecting group, and relative configuration of the substrates
2bꢀe, the cycloisomerization process was largely ineffi-
cient in providing synthetically useful yields of the desired
methylene cyclopentanes 1bꢀe.
Scheme 3
In light of this disappointing result, we opted for the
introduction of a silyl group as a synthetic equivalent for
the hydroxyl group (Scheme 4). Accordingly, the allylic
alcohol 512 was converted to the allyloxy actetate 6 by
etherification and esterification; a subsequent aldol con-
densation then delivered the allyl vinyl ether 7.13,14 Upon
treatment with LDA, 7 is converted to the corresponding
ester dienolate which undergoes a moderately diastereose-
lective [2,3]-Wittig rearrangement to afford the 1,5-hexa-
diene 2f in very good yield (89%, syn:anti = 86:14).15 The
mixture of diastereomers of 2f was used for the subsequent
cycloisomerization reaction which, compared to the cy-
cloisomerization of 2a (vide supra), required elevated
reaction temperatures and an increased reaction time. In
the event, we were delighted to obtain the desired methy-
lene cyclopentane cis-1f in up to 58% yield;16 additionally,
small amounts oftrans-1f and 3f weredetectedand isolated
as an inseparable mixture. Disappointingly, subsequent
Figure 1. 1H NMR kinetic profile for the cycloisomerization of
(()-2a with [Pd(MeCN)4](BF4)2 (0.05 equiv). Experiments were
performed using a Shigemi NMR tube in CDCl3 (c = 0.1 M) at
27 °C.
We next studied the course of the cycloisomerization of
diastereomerically pure 2a with [Pd(MeCN)4](BF4)2 in
CDCl3 by 1H NMR spectroscopy (Figure 1). The kinetic
profile nicely illustrates the consumption of the starting
diene with concomitant formation of the desired methy-
lene cyclopentane 1a.10 The result emphasizes the impact
of reaction duration on yield and selectivity because
initially formed 1a is slowly but steadily consumed and
converted to the undesired methyl cyclopentene 3a and, by
subsequent elimination, to the cyclopentadiene 4a.11
Having identified a manageable and effective catalyst,
we set out to adapt the structure of the hexadiene substrate
to the requirements of our retrosynthesis (Scheme 1).
(12) Le Menez, P.; Fargeas, V.; Berque, I.; Poisson, J.; Ardisson, J.;
Lallemand, J.-Y.; Pancrazi, A. J. Org. Chem. 1995, 60, 3592–3599.
(13) Hiersemann, M. Synthesis 2000, 1279–1290.
(14) 2f is prone to undergo a GosteliꢀClaisen rearrangement if stored
at ambient temperature. Apparently, the comparatively low barrier for
the rearrangement of 2f is caused by a remarkable rate-accelerating
effect of the PhMe2SiCH2 substituent. See: (a) Rehbein, J.; Leick, S.;
Hiersemann, M. J. Org. Chem. 2009, 74, 1531–1540. (b) Rehbein, J.;
Hiersemann, M. J. Org. Chem. 2009, 74, 4336–4342.
(8) Schramm, R. F.; Wayland, B. B. Chem. Commun. 1968, 898–899.
In our hands, different commercial and self-synthetized batches of
[Pd(MeCN)4](BF4)2 exerted a discrete phenotype and reactivity (see
Supporting Information).
(9) For a review on the chemistry of hydrido complexes of palladium,
see: Grushin, V. V. Chem. Rev. 1996, 96, 2011–2034.
(10) Notably, the rate of conversion is significantly slower under the
(15) The relative configuration of the major diastereomer syn-2f was
expected based on the proposed stereochemical model for ester dienolate
[2,3]-Wittig rearrangements. See: (a) Hiersemann, M. Tetrahedron 1999,
55, 2625–2638. (b) Hiersemann, M.; Lauterbach, C.; Pollex, A. Eur. J.
Org. Chem. 1999, 2713–2724. (c) Hiersemann, M.; Abraham, L.; Pollex,
A. Synlett 2003, 1088–1095. Later corroborated by X-ray crystallogra-
conditions of the NMR experiment.
€
phy. See: Nelson, B.; Schurmann, M.; Preut, H.; Hiersemann, M. Acta
(11) The incomplete mass balance is due to the formation of dimers
by Heck reaction of the Pd-σ-complex that emerges from the hydro-
palladation/intramolecular carbopalladation sequence. Details will be
published in a full paper.
Crystallogr. 2010, E66, o3102.
(16) The assignment of the relative configuration of cis-1f is sup-
ported by NOE experiments. See the SupportingInformation for details.
4440
Org. Lett., Vol. 13, No. 16, 2011
Nelson, Bjoern
