Indium-catalyzed homo-nazarov cyclizations of alkenyl cyclopropyl ketones

Article abstract of DOI:10.1021/ol102497w

Herein, an efficient Lewis acid catalyzed protocol for the homo-Nazarov cyclizations of alkenyl cyclopropyl ketones is reported. Alkenes bearing β-hydrogens (or silyl groups) provide 1.5:1 mixtures of methylene cyclohexenols and cyclohexenones. When no β-

Full text of DOI:10.1021/ol102497w

ORGANIC
LETTERS
2010
Vol. 12, No. 24
5684-5687
Indium-Catalyzed Homo-Nazarov
Cyclizations of Alkenyl Cyclopropyl Ketones
Dadasaheb V. Patil, Lien H. Phun, and Stefan France*
School of Chemistry and Biochemistry, Georgia Institute of Technology,
901 Atlantic DriVe, Atlanta, Georgia 30332, United States
stefan.france@chemistry.gatech.edu
Received October 15, 2010
ABSTRACT
Herein, an efficient Lewis acid catalyzed protocol for the homo-Nazarov cyclizations of alkenyl cyclopropyl ketones is reported. Alkenes
bearing ꢀ-hydrogens (or silyl groups) provide 1.5:1 mixtures of methylene cyclohexenols and cyclohexenones. When no ꢀ-hydrogens (or silyl
groups) are present, only cyclohexenones are observed. Products are rapidly formed in good to high yields (up to 93%) under mild conditions
and could be readily derivatized.
Cyclopropanes play a unique role in organic synthesis due
to their inherent ring strain. Consequently, they can partici-
pate in a variety of reactions that are not accessible to linear
or larger cyclic molecules.1 Moreover, the appropriate choice
of substituents about the cyclopropane ring has a direct effect
on the observed reactivity. For instance, one of the C-C
bonds of a cyclopropyl ring can be polarized when an
electron-accepting group is located on one carbon and an
electron-donating substituent is located on another. These
vicinal donor-acceptor (D-A) cyclopropanes have gener-
ated a tremendous amount of interest from organic chemists
for their predictable reactivity and can often be considered
as “masked” 1,3-dipoles and homologues of Michael accep-
tors.2 Furthermore, their application in [3 + 2] and [3 + 3]
cycloadditions has been fully recognized.3 Along these lines,
a viable but underexplored approach to six-membered carbo-
and heterocycles has been exemplified by the homo-Nazarov
cyclization (Scheme 1). In this reaction, cyclopropyl vinyl
ketones are converted into R,ꢀ-unsaturated cyclohexenones.
Although mechanistically distinct, this reaction is homolo-
Scheme 1. Homo-Nazarov Cyclization
(1) (a) Reissig, H.-U. In Small Ring Compounds in Organic Synthesis
III; Meijere, A. de, Ed.; Topics in Current Chemistry; Springer-Verlag:
Berlin, Heidelberg, Germany, 1988; Vol. 144, p 1. (b) Hudlicky, T. Reed,
J. W. In ComprehensiVe Organic Synthesis; Trost, B. M., Fleming, I.,
Paquette, L. A., Eds.; Pergamon Press: New York, 1991; Vol. 5, p 899.
(2) (a) Rubin, M.; Rubia, M.; Gevorgyan, V. Chem. ReV. 2007, 107,
3117. (b) Yu, M.; Pagenkopf, B. L. Tetrahedron 2005, 61, 321. (c) Reissig,
H.-U.; Zimmer, R. Chem. ReV. 2003, 103, 1151.
10.1021/ol102497w  2010 American Chemical Society
Published on Web 11/23/2010

gous to the well-established Nazarov cyclization,4 which
involves the cyclization of divinyl ketones to make cyclo-
pentenones via a similar cyclic oxyallyl cation.
Several groups have reported examples of this reaction
where either aryl or heteroaryl groups replace the vinyl group,
known as the (hetero)aromatic homo-Nazarov cyclization.5
Unfortunately, in each of these earlier reports, stoichiometric
amounts of acids (Bronsted and Lewis) or high temperatures
are required for cyclization.
In contrast, the reactions of alkenyl cyclopropyl ketones
have been less studied. In a seminal report, Tsuge reported
the treatment of various alkenyl cyclopropyl ketones with
polyphosphoric acid in hopes of obtaining cyclohexenone
products.6 However, the reaction had three major drawbacks:
(1) lack of generalitysonly 5 out of 16 substrates provided
the desired cyclohexenones; (2) poor product yields (15-63%);
and (3) harsh reaction conditionssexcess PPA in refluxing
benzene for >24 h. All of these limitations have resulted in
sparse application of the alkenyl homo-Nazarov protocol.
More recently, Waser reported a Bronsted acid catalyzed
homo-Nazarov cyclization.7 This study has provided invalu-
able insight into the potential of the alkenyl reaction with
one limitation: simple alkenes did not work (an R-heteroatom
was required to activate the alkenyl group). In this com-
munication, we describe an efficient Lewis acid catalyzed
homo-Nazarov cyclization of donor-acceptor cyclopropyl
ketones bearing substituted alkenes.
similarly demonstrated the benefit of using esters as acceptor
groups to polarize the classic Nazarov cyclization.8 Finally,
for these donor-acceptor-acceptor (D-A-A) cyclopro-
panes, the C-C bond is weaker as compared to the parent
donor-acceptor cyclopropanes, allowing for milder condi-
tions for bond cleavage, more variety of functionality about
the cyclopropanes, and greater stabilization of the resulting
dipole.9
To test the rationale of using D-A-A cyclopropanes to
promote catalysis, we set out to probe the reactivity of
alkenyl substrates. Concerned about ensuring the stability
of the resulting oxyallyl cation as compared to that of the
benzylic cation, we chose to synthesize a substrate bearing
an R-alkyl substituent to offer further stabilization (Scheme
2).10 Thus, alkenyl substrate 4a was synthesized from the
malonate derived Weinreb amide 1 via sequential diazo
transfer, cyclopropanation11 (with 4-methoxystyrene), and
ketone formation.
Scheme 2
.
Synthesis of a Model D-A-A Homo-Nazarov
Substrate 4a
In an attempt to facilitate cyclization and, ultimately,
expand the scope and applicability of the reaction, we
envisioned using donor-acceptor cyclopropyl vinyl ketones
bearing a secondary electron acceptor (such as an ester) in
the R′-position that would coordinate with Lewis acids, as
in A. The secondary acceptor group would serve to further
“polarize” the resulting cyclic oxyallyl cation by localizing
the charge density. This polarization allows for a predictable
reaction outcome for the oxyallyl cation. Frontier has
In hopes of finding the best catalyst for the reaction, we
screened substoichiometric amounts (30 mol %) of various
metal catalysts, primarily focusing on readily available triflate
salts. The initial reaction with 4a was conducted in dichlo-
romethane at room temperature in the presence of 30 mol
% catalyst (Scheme 3). While we anticipated that cyclization
would afford cyclohexenone 6a based on a standard Nazarov-
type eliminative pathway, we were intrigued to find that
while we did form 6a, another putative homo-Nazarov
product 5a, a cross-conjugated enol system with an exocyclic
alkene, was observed. Beyond these two products, dihydro-
furan 7a, which presumably arises from the enolate attack
upon the acyclic benzylic cation, was observed, as well as
unreacted 4a (Scheme 4).
(3) For recent representative examples, see: (a) Carson, C. A.; Young,
I. S.; Kerr, M. A. Synthesis 2008, 485. (b) Ivanova, O. A.; Budynina, E. M.;
Grishin, Y. K.; Trushkov, I. V.; Verteletskii, P. V. Angew. Chem. 2008,
120, 1123. (c) Pohlhaus, P. D.; Johnson, J. S. J. Am. Chem. Soc. 2005,
127, 16014. (d) Sugita, Y.; Yamadoi, S.; Hosoya, H.; Yokoe, I. Chem.
Pharm. Bull. 2001, 49, 657.
(4) For reviews, see: (a) Nakanishi, W.; West, F. G. Curr. Opin. Drug
DiscoVery DeV. 2009, 12, 732. (b) Tius, M. A. Eur. J. Org. Chem. 2005,
2193. (c) Frontier, A. J.; Collison, C Tetrahedron 2005, 61, 7577. (d)
Pellissier, H. Tetrahedron 2005, 61, 6479.
(5) (a) Murphy, W. S.; Wattansin, S. Tetrahedron Lett. 1980, 21, 1887.
(b) Murphy, W. S.; Wattansin, S. J. Chem. Soc., Perkins Trans. 1 1981,
2920. (c) Murphy, W. S.; Wattanasin, S. J. Chem. Soc., Perkin Trans. 1
1982, 1029. (d) Yadav, V. K.; Kumar, N. V. Chem. Commun. 2008, 3774.
(6) Tsuge, O.; Kanemasa, S.; Otsuka, T.; Suzuki, T. Bull. Chem. Soc.
Jpn. 1988, 61, 2897.
(8) He, W.; Herrick, I. R.; Atesin, T. A.; Caruana, P. A.; Kellenberger,
C. A.; Frontier, A. J. J. Am. Chem. Soc. 2008, 130, 1003.
(9) Pohlhaus, P. D.; Sanders, S. D.; Parsons, A. T.; Li, W.; Johnson,
J. S. J. Am. Chem. Soc. 2008, 130, 8642.
(10) This hypothesis was confirmed later when no desired cyclizations
were observed with substrates bearing an R-hydrogen.
(11) Marcoux, D.; Charette, A. B. Angew. Chem., Int. Ed. 2008, 47,
10155.
(7) De Simone, F.; Andres, J.; Torosantucci, R.; Waser, J. Org. Lett.
2009, 11, 1023.
Org. Lett., Vol. 12, No. 24, 2010
5685

Scheme 3
.
Initial Lewis Acid Screening
Figure 1. Lewis acid catalyst screen for alkenyl cyclopropyl ketone
cyclization. Reactions were conducted in dichloromethane at room
temperature in the presence of 30 mol % catalyst. Percentages are
based on product ratios calculated from crude 1H NMR spectra for
each individual reaction.
Scheme 4
.
Formation of Dihydrofuran 7a
to the reaction conditions, mixtures of 5 and 6 are observed.
Model substrate 4a provided alkenes 5a and 6a in 75% yield
as a 1.5:1 mixture (entry 1). For 5a, the methyl group and
phenyl were oriented in a trans relationship (5:1 trans/cis
dr) as confirmed by NMR.13 Similarly, cyclohexenyl ketone
4b also proved to be a good substrate, affording the 1.5:1
alkene mixture in 75% yield (entry 2). Using 4c, products
5c and 6c were formed in 77% yield as a 1.5:1 mixture (entry
3). Interestingly, when the aryl substituent was changed to
phenyl or 4-fluorophenyl (entries 4 and 5), 5d and 5e were
the only cyclized products observed in 46% (80% BRSM)
and 56% yields (77% BRSM), respectively. The reactions
did not go to completion after 24 h, only offering starting
material 4. These observations can be rationalized based on
the decreased cation stabilizing ability of the phenyl and
4-fluorophenyl groups in comparison to the donating 4-meth-
oxyphenyl. With hopes of influencing the product ratio, we
installed a silyl group on the R-substituent in 4f to further
stabilize the resulting oxyallyl cation. The reaction proceeded
rapidly to 5c in 92% yield within 0.5 h (entry 6). The
formation of this product agrees with the mechanistic
hypothesis of carbocation formation due to the ꢀ-silyl effect.
When the R-substituent was a heteroatom, efficient cycliza-
tion of 4g was observed to give product 6d in 93% yield
(entry 7). Finally, we looked at a substrate bearing an
R-phenyl group (entry 8). We anticipated that the phenyl
group would serve to stabilize the resulting cyclic oxyallyl
cation, if formed, allowing for increased reaction rates.
However, when 4h (bearing a R-phenyl substituent) was
subjected to the reaction conditions, only a 30% yield (50%
BRSM) of the expected cyclized product 6e was observed
after 35 h. Along with unreacted starting material, the major
component was the dihydrofuran byproduct. This marks the
1
Based on the crude H NMR spectra of the individual
screening reactions, the ratios of homo-Nazarov products 5a
and 6a were compared directly against those of 4a and 7a
and plotted the data in Figure 1. To our satisfaction, In(OTf)3
emerged as the most efficient promoter (full conversion of
4a to 5a and 6a within 3 h), while Zn(OTf)2 and Cu(OTf)2
were the next best Lewis acids, although neither gave full
conversion of 4a even after 105 h. Pd(0) and Pd(II)
complexes did not give any homo-Nazarov products. Lastly,
no conversion of 4a was observed for Ir(acac)3, which was
chosen based on Frontier’s successes with Ir(III) complexes
for polarized Nazarov substrates.
We next turned our attention to probing for any noticeable
solvent effects. When a variety of solvents (i.e., tetrahydro-
furan, benzene, dichloroethane, acetonitrile, nitromethane,
etc.) were screened, dichloromethane remained the best
choice and no observable changes in product distributions
were observed with any of the other solvents.
Having established that In(OTf)3 in dichloromethane was
the best overall catalyst system to promote the reaction,12
this method was applied to a diverse set of alkenyl cyclo-
propyl ketone substrates to determine the reaction scope and
applicability (Table 1). When R-alkyl substrates are subjected
(12) For recent reviews of In(III) catalysis, see: (a) Ghosh, R.; Maiti, S.
J. Mol. Catal. A: Chem. 2007, 264, 1. (b) Frost, C. G.; Hartley, J. P. Mini-
ReV. Org. Chem. 2004, 1, 1–7.
(13) See Supporting Information.
5686
Org. Lett., Vol. 12, No. 24, 2010

Table 1. In(OTf)3-Catalyzed Homo-Nazarov Cyclization of
Scheme 5. Examples of Facile Product Derivatization
Alkenyl Cyclopropyl Ketonesa
between the R-phenyl substituent and the oxygen lone pairs
of the adjacent ester group.14 Consequently, upon cyclopro-
pane ring opening, enolate trapping of the acyclic cation
directly outpaces the homo-Nazarov cyclization pathway due
to the difficulty in rotating into the more reactive s-trans
enone conformer.
As a further illustration of the utility of our reaction, we
sought to use our alkenyl products as synthetic building
blocks by converting them into other useful compounds
(Scheme 5). For instance, when 5c was treated with thiophe-
nol,15 thioether 8 was obtained in 58% yield as a 1:1 mixture
of diastereomers. Similarly, 6a was subjected to Krapcho16
decarbalkoxylation conditions to generate 9 in 74% yield.
In conclusion, an efficient protocol for the homo-Nazarov
cyclization of alkenyl cyclopropyl ketones has been reported.
Further studies are underway to apply this protocol to the
heteroaromatic homo-Nazarov cyclization. Other activities
include the determination of the reaction kinetics as well as
full elucidation of the effects of substituents on the reaction
mechanism. Lastly, reactions employing chiral catalyst
complexes are being explored. The results of each of these
studies will be reported in due course.
Acknowledgment. S.F. thanks the NSF FACES Program
for a Career Initiation Grant and ORAU for the Ralph E.
Powe Junior Faculty Enhancement Award.
a Reactions run with 1 equiv of substrate 4 and 30 mol % In(OTf)3 in
CH2Cl2 at 25 °C and complete within 1 h. b Isolated yields after column
chromatography. c Reaction did not go to completion over 24 h. d Yields
based on recovered starting material are as follows: 5d (80%); 5e (77%);
and 6e (50%).
Supporting Information Available: Experimental pro-
cedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
OL102497W
only instance in which this byproduct is observed when
In(OTf)3 is used as the catalyst. This can be rationalized if
4h exists in the less reactive s-cis enone conformation.
MMFF calculations revealed that the s-cis conformation is
more stable than the s-trans conformation by more than 6
kcal/mol due to the presence of stabilizing π-interactions
(14) MMFF calculations were run using Trident software available from
Wavefunction, Inc.
(15) Cabrera, S.; Alema´n, J.; Bolze, P.; Bertelsen, S.; Jorgensen, K. A.
Angew. Chem., Int. Ed. 2008, 47, 121–125.
(16) Krapcho, A. P.; Jahngen, E. G. E.; Lovey, A. J.; Short, F. W.
Tetrahedron Lett. 1974, 15, 1091.
Org. Lett., Vol. 12, No. 24, 2010
5687