Synthesis of bridged medium-sized rings through the intramolecular Pauson-Khand reaction

Article abstract of DOI:10.1021/ol016308s

(Equation presented) A series of aromatic enynes containing steric buttressing elements were prepared and evaluated in the NMO-mediated Pauson-Khand cyclization. O-Allyl systems led to the expected angularly fused products, whereas the O-butenyl and O-pentenyl derivatives afforded the unprecedented bridge systems.

Full text of DOI:10.1021/ol016308s

ORGANIC
LETTERS
2001
Vol. 3, No. 16
2607-2610
Synthesis of Bridged Medium-Sized
Rings through the Intramolecular
Pauson−Khand Reaction
,†
Carl J. Lovely,* Hemalatha Seshadri,† Brian R. Wayland,† and
A. Wallace Cordes‡,§
Department of Chemistry & Biochemistry, UniVersity of Texas at Arlington,
Arlington, Texas 76019, and Department of Chemistry & Biochemistry,
UniVersity of Arkansas, FayetteVille, Arkansas 72701
loVely@uta.edu
Received June 20, 2001
ABSTRACT
A series of aromatic enynes containing steric buttressing elements were prepared and evaluated in the NMO-mediated Pauson−Khand cyclization.
O-Allyl systems led to the expected angularly fused products, whereas the O-butenyl and O-pentenyl derivatives afforded the unprecedented
bridge systems.
The Pauson-Khand reaction (PKR) is a powerful method
for the construction of cyclopentenones.1 Over the past few
years this transformation has been investigated extensively,
resulting in the expansion of the scope of the reaction and
the development of catalytic variants.1 However, despite its
obvious synthetic potential, until recently the intramolecular
variant of this reaction has been largely restricted to the
construction of bicyclo[3.3.0]octenones and bicyclo[4.3.0]-
nonenones.2,3 This limitation would seem to be extremely
unfortunate given the large number of biologically active
natural products that possess a medium-sized ring annulated
to a cyclopentyl moiety, which in principle might be
accessible in a direct fashion via an intramolecular PKR.4 It
is well-known that the construction of medium rings is more
difficult than the formation of normal rings due to a
combination of unfavorable entropic and enthalpic factors.5
In the absence of a detailed understanding of the mechanism
of the PKR, it appeared to us that the incorporation of
structural features into the cyclization precursors that would
reduce the entropic contribution to the free energy of
activation might favor medium ring formation. The gem-
dimethyl effect has already been shown to be effective in
the PKR and other cyclizations by increasing the reactive
rotamer population.6 It was proposed that the construction
of substrates such as 1 and 2 might serve a similar purpose
(Figure 1). It was envisioned that the introduction of an
† University of Texas at Arlington.
‡ University of Arkansas.
§ To whom correspondence regarding X-ray structures should be ad-
dressed.
(1) For a recent review of this area, see: Brummond, K. M.; Kent, J. L.
Tetrahedron 2000, 56, 3263.
(2) Krafft, M. E.; Fu, Z.; Bonaga, V. R. Tetrahedron Lett. 2001, 42,
1427.
(3) Cazes has reported the low yield synthesis of a bicyclo[5.3.0]decene
from an allenyne. See: Ahmar, M.; Locatelli, C.; Colombier, D.; Cazes,
B. Tetrahedron Lett. 1997, 38, 5281.
Figure 1. Conformationally restricted aryl enynes.
10.1021/ol016308s CCC: $20.00 © 2001 American Chemical Society
Published on Web 07/10/2001

aromatic ring into the backbone of the enyne would reduce
the conformational degrees of freedom of the side chains,
thereby preorganizing the alkyne and alkene for cyclization.
After an initial, largely unsuccessful, foray with substrates
patterned after 1,7 our attention was drawn to the alternate
series 2.8,9 This Letter details our initial findings in this latter
substrate.
Some time ago, Sammes and co-workers demonstrated that
the introduction of substituents ortho to conformationally
flexible side chains in aromatic substrates had a beneficial
effect on both the rates and yields of the [3 + 2] azide-
alkene cycloadditions under study.11 These “steric buttresses”
decreased the conformationally accessible space, and hence
the entropic change during the reaction was reduced. This
strategy appeared to be a sound one, and therefore it was
applied to the substrates in this study (Figure 2).12
The cyclization precursors were readily prepared via
alkylation of the TMS-protected 2-ethynylphenol 3,10 either
directly with the alkyl bromide or through a Mitsunobu
protocol (Scheme 1).8 Desilylation then provided the requisite
Scheme 1a
Figure 2. Steric buttressing in aryl enynes.
Initially it was decided to evaluate substrates that contained
an o-t-Bu group as the buttressing element. For the sake of
synthetic convenience, 2,4-di-tert-butylphenol was employed
as starting material, since regiochemical complications could
be avoided in the preparation of o-ethynylphenol 8.13 With
this material in hand, the hydroxyl group was alkylated either
with the alkyl halide or through a Mitsunobu reaction as
before (Scheme 2). After desilylation, the enynes were
converted into the corresponding Co2(CO)6 complex and then
treated with NMO. The influence of the t-Bu group on these
reactions was immediately apparent. In the case of enyne
10a, it was converted into the expected enone 12a in 70%
yield. However, instead of requiring 8 h for complete
consumption of starting material such as 4a, the reaction of
10a was complete after 30 min (after the addition of NMO
was complete). In a similar fashion, the consumption of the
higher homologues 10b and 10c was rapid, leading to the
formation of new compounds. However, when the products
were isolated, it was clear from the NMR data that although
the products were cyclic and they possessed carbonyls, they
were not the expected cycloadducts 13a or 16a.14 In the case
of the PKR of enyne 10b, two new compounds were isolated
in 18% and 36% yields. The determination of the identity
of the major cycloadduct proved to be quite challenging as
inter alia the mass spectrum revealed that the compound
contained one more oxygen atom than expected. Fortunately,
this compound provided crystals of suitable quality for
analysis by X-ray crystallography. The results obtained from
this determination were quite surprising (Figure 3). The
a (a) K2CO3, acetone, allyl bromide; (b) Ph3P, DEAD, 3-buten-
1-ol, THF; (c) K2CO3, NaI, acetone, 5-bromo-1-pentene; (d) K2CO3,
MeOH, THF; (e) Co2(CO)8, CH2Cl2 then NMO.
enynes 4a-c.8 After conversion to the corresponding
Co2(CO)6 complexes, the enynes were treated with excess
N-methylmorpholine oxide (NMO) to initiate the cyclization
process. In the case of enyne 4a, a smooth cyclization
reaction was observed, providing the 6,6,5-tricyclic system
5 in 75% yield.8 On the other hand, both enynes 4b and 4c
failed to undergo cyclization. While this and other experi-
ments confirmed that the aromatic ring was a suitable
scaffold for PKR’s, apparently the entropic issues were not
addressed in the higher homologues.8,9
(4) For an attempt to prepare seven-membered rings through the
intramolecular Pauson-Khand reaction, see: Mukai, C.; Sonobe, H.; Kim,
J. S.; Hanoaka, M. J. Org. Chem. 2000, 65, 6654. See also Wender, P. A.;
McDonald, F. E. Tetrahedron Lett. 1990, 31, 3691.
(5) (a) Petasis, N. A.; Patane, M. A. Tetrahedron 1992, 48, 5757. (b)
Roxburgh, C. J. Tetrahedron 1993, 49, 10749.
(6) (a) Jung, M. E. Synlett 1999, 843. (b) Jung, M. E. Synlett 1990, 186.
(7) Seshadri, H. Ph.D. Dissertation, The University of Texas at Arlington,
Arlington, TX, 2001.
(10) Arcadi, A.; Cacchi, S.; Del Rosario, M.; Fabrizi, G.; Marinelli, F.
J. Org. Chem. 1996, 61, 9280.
(11) Orlek, B. S.; Sammes, P. G.; Weller, D. J. Tetrahedron 1993, 49,
8179.
(12) For a review of this approach see Sammes, P. G.; Weller, D. J.
Synthesis 1995, 1205.
(8) Lovely, C. J.; Seshadri, H. Synth. Commun. 2001, 31, xxxx. See
also: Blanco-Urgoiti, J.; Casarrubios, L.; Dom´ınguez, G.; Pe´rez-Castells,
J. Tetrahedron Lett. 2001, 42, 3315.
(9) For studies relating to systems derived from o-vinylphenol, see:
Pe´rez-Serrano, L.; Blanco-Urgoiti, J.; Casarrubios, L.; Dom´ınguez, G.;
Pe´rez-Castells, J. J. Org. Chem. 2000, 65, 3513.
(13) The o-ethynylphenols were prepared by iodination of the corre-
sponding phenol and then Sonogashira cross-coupling of the iodide with
TMS-acetylene. See Supporting Information for details.
(14) The absence of a three-proton doublet quickly ruled out migration
of the double bond to an internal position followed by cyclization to provide
a methylcyclopentenone: see ref 4.
2608
Org. Lett., Vol. 3, No. 16, 2001

Scheme 2a
a (a) K2CO3, acetone, allyl bromide; (b) Ph3P, DEAD, 3-buten-1-ol or 4-penten-1-ol, THF; (c) K2CO3, NaI, acetone, 5-bromo-1-pentene;
(d) K2CO3, MeOH, THF; (e) Co2(CO)8, CH2Cl2 then NMO.
mystery of the additional oxygen became immediately
clear: the product contained an epoxide moiety 14a.15 The
greatest surprise however came in the form of the ring size
and the mode of ring fusion. The X-ray structure revealed
that instead of cyclizing to afford a seven-membered ring,
i.e., 13a, an eight-membered ring was formed. Furthermore,
instead of the expected angular ring fusion, the eight- and
five-membered rings were bridged. This is, to the best of
our knowledge, the first example of the formation of this
type of bridged ring system in the PKR.16 This crystal-
lography result suggested that the minor product was in fact
enone 15a.17
cyclic product was formed, but there were inconsistencies
observed in the 1H NMR spectrum with regard to the enone
proton, cf. 15a. The structure of the product was secured
through X-ray analysis, and this showed again that a bridged
system was formed (Figure 4). In this case, a nine-membered
An equally unusual result was obtained from the cycliza-
tion of enyne 10c. Again the spectroscopic data suggested a
Figure 4. X-ray crystal structure of bridged enone 17a.
ring was formed bridged to a five-membered ring, 17a. No
epoxide was isolated from this cyclization reaction.
Given that the t-Bu substituent facilitated this reaction,
we wished to establish whether the steric bulk of this group
was excessive. Therefore, substrates containing the smaller
methyl group were prepared and evaluated. The necessary
precursors were prepared in a largely analogous fashion to
enynes 10a-c, from the corresponding o-ethynylphenol 9.10
Figure 3. X-ray crystal structure of epoxyenone 14a.
Org. Lett., Vol. 3, No. 16, 2001
2609

With enynes 11a-c in hand, they were converted into their
cobalt complexes and then treated with NMO. In the case
of enyne 11a, the expected normal cyclization product 12b
was obtained in 74% yield. However, the rate of this
cyclization was almost identical to that of the unsubstituted
system 4a. This did not appear to bode well for the higher
homologues. However, we were pleased to find that both
enynes 11b and 11c participated in the cyclization to again
provide bridged products. In the case of 11b, the only isolated
product was the epoxide 14b, which was obtained in 42%
yield. 11c provided the bridged enone 17b in 20% yield.
The mode of cyclization observed in these substrates is
highly unusual for the PKR. However these results can be
rationalized in terms of the alkene rotamer that undergoes
insertion; this is illustrated in Scheme 3 for the cycloaddition
complex (21, Scheme 3B). The alkene inserts into the Co-
C6 bond, and new bonds are formed between Co-C5 and
C6-C4, furnishing the eight-membered ring containing
metallocycle 22. This is followed by CO insertion and loss
of the cobalt complex, affording the bridged compound 23.18
The driving force behind the unprecedented regiochemistry
observed in this cycloaddition is not clear at this point. The
buttressing groups seemingly play a crucial role in decreasing
the entropic barrier to cycloaddition by bringing the reacting
partners closer to each other. They may also limit the space
for accommodating the linker between the phenol oxygen
and alkene, which then forces insertion through the abnormal
rotamer 21. The insertion of the carbonyl moiety at the
internal carbon of the alkyne is unusual in the intramolecular
PKR. However, this orientation is normal in the intermo-
lecular variant and so the examples reported in this paper
may simply follow typical sterically driven regiochemistry
that can be accommodated as a result of the extended tether
length. Electronic effects cannot be ruled out however since
the phenolic oxygen atom can interact with the alkyne group
in these substrates. The net effect of this interaction is to
polarize the alkyne, which in turn may influence the
regioselectivity of the insertion.19,20
Scheme 3a
In summary, the work reported herein describes the use
of steric buttressing to facilitate the formation of medium-
sized rings via the PKR. Furthermore, these cyclizations
occur to form bridged rings as a result of unusual regiose-
lectivity during alkene insertion. Currently we are investigat-
ing the scope of this interesting variant of the PKR and the
factors that control this unusual mode of insertion. We will
report on this and other aspects of this work in due course.
a CO ligands are omitted for clarity.
Acknowledgment. This work was supported by the
Robert A. Welch Foundation (Grant Y-1362) and The
University of Texas at Arlington (Start-up Funds). The NSF
(CHE-9601771) is thanked for partial funding of the purchase
of a 500 MHz NMR spectrometer (UTA). Arkansas Science
and Technology Authority is gratefully acknowledged for
support of the X-ray facility (UA). We would like to thank
Prof. Ken Turnbull and Dr. Tony Hudgens (Department of
Chemistry and Biochemistry, The University of Arkansas)
for providing the experimental protocol for the preparation
of 2-iodo-4,6-dimethylphenol.
of the butenyl substrates. The generally accepted mechanistic
hypothesis for the formation of the “normal” enone involves
the orientation of the alkene terminus away from the cobalt
alkyne complex (18, Scheme 3A). The alkene inserts into
the Co-C7 bond, and new bonds between Co-C5 and C7-
C4 are formed, furnishing the seven-membered ring contain-
ing metallocycle 19. Subsequent CO insertion followed by
loss of the complexed cobalt cluster furnishes the [5.3.0]
bicyclic system 20. To account for the observed outcome,
presumably the alkene orients itself toward the cobalt alkyne
Supporting Information Available: Experimental pro-
cedures and characterization data for all compounds de-
scribed in this paper. Details of the X-ray analysis of
compounds 14a and 17a. This material is available free of
charge via the Internet at http://pubs.acs.org.
(15) At this point we are operating under the assumption that the
epoxidation of the strained double bond is NMO promoted. However, control
experiments indicate that this does not take place by direct reaction of the
enone with NMO. For a related observation, see: Muto, R.; Ogasawara,
K. Tetrahedron Lett. 2001, 42, 4143 and references therein.
(16) Krafft has observed bridged products as a result of CO insertion at
the internal position of both the alkyne and the alkene; see ref 2. See also:
Forsyth, G. S.; Kerr, W. J.; Ladduwahetty, T. In Organometallics in Organic
Synthesis; Bateson, J. H., Mitchell, M. B., Eds.; Academic Press: London,
1994; p 239. We are grateful to Prof. Marie Krafft for bringing this reference
to our attention.
OL016308S
(18) Presumably, with the allyl-containing substrates it is geometrically
difficult to adopt a conformation related to 21 to produce a bridged adduct.
(19) Krafft, M. E.; Romero, R. H.; Scott, I. L. J. Org. Chem. 1992, 57,
5277.
(17) Particularly supportive of this assignment is the downfield signal
for the â-enone proton (δ ) 8.25 vs 6.30-6.47); see ref 8.
(20) Robert, F.; Milet, A.; Gimbert, Y.; Konya, D.; Greene, A. E. J.
Am. Chem. Soc. 2001, 123, 5396.
2610
Org. Lett., Vol. 3, No. 16, 2001