Development of a Nazarov cyclization/Wagner-Meerwein rearrangement sequence for the stereoselective synthesis of spirocycles

Article abstract of DOI:10.1021/ja0716148

A stereoselective Nazarov cyclization/Wagner-Meerwein rearrangement sequence for the synthesis of spirocyclic compounds was developed. While a range of different substrate types engaged in the cyclization/rearrangement sequence, it was found that differen

Full text of DOI:10.1021/ja0716148

Published on Web 06/08/2007
Development of a Nazarov Cyclization/Wagner-Meerwein Rearrangement
Sequence for the Stereoselective Synthesis of Spirocycles
Jie Huang and Alison J. Frontier*
Department of Chemistry, UniVersity of Rochester, Rochester, New York 14627
Received March 7, 2007; E-mail: frontier@chem.rochester.edu
Table 1. Optimized Conditions for Spirocycle Formation
The stereodetermining step of the Nazarov cyclization is the
conrotatory 4π-electrocyclization of a 3-oxopentadienyl cation to
an oxyallyl cation.1 While the reaction is typically terminated by
elimination of a proton to give an enone derivative, it is possible
to intercept the oxyallyl cation intermediate if a suitable trapping
agent is present. A number of synthetically useful “interrupted”
Nazarov cyclizations of this type have been reported.2 Wagner-
Meerwein rearrangement processes3 have also been observed during
Nazarov cyclization.4 These studies have provided valuable insight
into the avenues available to the oxyallyl cation intermediate but
are typically characterized by multicomponent product mixtures
reflecting competing rearrangement pathways. In this communica-
tion, we describe a Nazarov cyclization/Wagner-Meerwein re-
arrangement pathway that produces unusual spirocyclic products.
The transformations are highly stereoselective, efficient for a range
of substrate types, and capable of creating adjacent quaternary
centers.
Ratio 1c/1ba
entry
promoter
10 mol %
50 mol %
100 mol %
1
2
3
4
Cu(OTf)2
I
AgSbF6
Cu(SbF6)2
<5/95
20/80
9/91
8/92
67/33b
50/50
69/31
13/87
>95/5
70/30c
>95/5
20/80d
1
a Ratio was determined by H NMR of crude mixture. b The ratio was
91/9 when substrate was slowly added to promoter. c Ratio increased to
84/16 with 200 mol % of promoter. d With 5 mol % of promoter.
experiments showed that Cu(SbF6)29 was just as effective as chiral
complex I (entry 4).
The study became even more intriguing when the cyclization
and rearrangement of p-methoxyphenyl derivative 2a did not give
spirocycle 2c as expected, but instead led to the formation of
spirocycle 2d, which contained adjacent quaternary centers (eq 3).8
In an earlier study, we found that compound 1a (as a mixture of
olefin isomers) underwent Nazarov cyclization when treated with
a catalytic amount of copper(II) triflate (eq 1).5 Bicyclic ketone 1b
was isolated as a single diastereomer.6 Given this result, it was
quite interesting to find that, when 1a was treated with a
stoichiometric amount of copper bisoxazoline complex I,7 the only
product obtained was spirocycle 1c.8 This unexpected product was
formed in 40% ee (eq 2).
Cyclizations on different substrates were carried out to explore
the scope of the process (Table 2). Depending upon the substituent
on the alkylidene â-ketoester of dienone a, the sequence selectively
produced spirocycles of either type c or type d. Substrates with
cinnamyl substitution rearranged to products of type d (entries 3
and 5), while alkyl-substituted substrate 4a gave spirocycle 4c (entry
4). The protocol also allowed smooth contraction of a seven-
membered ring to give spirocycle 5d in good yield (entry 5).
Since none of the cases initially studied addressed the question
of the stereochemistry of the spirocyclic quaternary center relative
to the adjacent chiral center, cyclohexadiene substrates 6a or 7a
were synthesized and subjected to the reaction conditions. In both
cases, smooth cyclization/ring contraction occurred as expected,
to give spirocycles 6c and 7d, respectively (entries 6 and 7).8 The
stereochemistry of 6c was assigned by X-ray crystallographic
analysis.
In a few cases, the Nazarov cyclization/1,2-sigmatropic rearrange-
ment sequence was examined using Cu(SbF6)2 as promoter (entries
1, 2, and 6). Under these conditions, both the yield and reaction
rates increased relative to the reaction with chiral complex I,
confirming this protocol as an inexpensive and superior alternative
for carrying out the reaction sequence.
The proposed mechanism of the Nazarov cyclization/rearrange-
ment sequence is shown in Scheme 1. After the 4π conrotatory
electrocyclic process, oxyallyl cation intermediate A is generated.
Elimination of a proton gives the normal Nazarov product of type
b (path 1), whereas ring contraction leads to spirocyclic cation B
(path 2). Then, depending on the migratory ability of the substituent
Further investigation of this process revealed that product
distribution is dependent on both the structure of the catalyst
complex and the amount of catalyst present in solution (Table 1).
Copper triflate gave predominantly the Nazarov cyclization product
1b at both low and high catalyst loadings (entry 1). Catalytic
amounts (10 mol %) of I gave mostly 1b, but as the loading of I
was increased, the proportion of spirocyclic product 1c in the
product mixture also increased (entry 2). In the presence of 100
mol % of I, 1c was obtained exclusively. High selectivity for 1c
over 1b could also be achieved by adding 1a slowly to 50 mol %
of the promoter I.
The spirocyclization could also be carried out using 1 equiv of
AgSbF6, but selectivity was moderate (entry 3). This result seemed
to indicate that the counterion was an important factor and suggested
that the ligand might not be necessary. Indeed, the next set of
9
8060
J. AM. CHEM. SOC. 2007, 129, 8060-8061
10.1021/ja0716148 CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
Table 2. Spirocycle Formation from Dienones 1a-7aa
explanation is that free carbonyl groups can act as Lewis bases,
accelerating the elimination step (path 1) that leads to products b.
Several lines of evidence support this idea. First, ratios of 1c/1b
were highest when enough promoter was used to provide one
coordination site for each carbonyl oxygen in the reaction mixture
(i.e., 1 equiv of a Cu(II) salt or 2 equiv of a Ag(I) salt; see Table
1). Second, reactions carried out in noncoordinating solvents
(dichloromethane or nitroethane) favored products c and d, while
product b dominated when the Lewis basic solvent THF was used
(with 100 mol % of Cu(SbF6)2, 1a gave a 7:1 ratio of 1b to 1c in
94% combined yield).12 Third, promoters containing the noncoor-
dinating counterion SbF6- were optimal, suggesting that a reaction
medium devoid of Lewis basic species allows rearrangement (path
2) to compete with elimination (path 1).
In summary, a 4π-electrocyclization/1,2-sigmatropic rearrange-
ment sequence has been developed. The efficient synthesis of richly
functionalized spirocycles using an inexpensive promoter was
demonstrated. Most promising for potential synthetic application
was the finding that adjacent stereocenters, including adjacent
quaternary centers, could be installed with high selectivity using
this protocol. Future efforts in our laboratory will focus on
applications of the method to natural product synthesis, as well as
development of a catalytic, enantioselective version of the reaction.
Acknowledgment. We are grateful to the NSF (CAREER award
for A.J.F., CHE-0349045) and the University of Rochester for
generous financial support.
a Reaction conditions: substrate a in CH2Cl2 (0.03 M); 100 mol % of I;
25 °C. b Values in parentheses were obtained using 100 mol % Cu(SbF6)2.
c The enantiomeric excess of products cyclized using I ranged from 29 to
64% (unoptimized). d Heated to 100 °C in dichloroethane.
Supporting Information Available: Experimental procedures,
spectroscopic data for all new compounds, and X-ray structure data of
1c and 6c (CIF). This material is available free of charge via the Internet
at http://pubs.acs.org.
Scheme 1. Proposed Mechanism of Spirocycle Formation
References
(1) (a) Habermas, K. L.; Denmark, S. E.; Jones, T. K. Org. React. 1994, 45,
1. (b) Frontier, A. J.; Collison, C. Tetrahedron 2005, 61, 7577. (c)
Pellissier, H. Tetrahedron 2005, 61, 6479.
(2) For selected examples, see: (a) Bender, J. A.; Arif, A. M.; West, F. G. J.
Am. Chem. Soc. 1999, 121, 7443. (b) Wang, Y.; Arif, A. M.; West, F. G.
J. Am. Chem. Soc. 1999, 121, 876. (c) Giese, S.; West, F. G. Tetrahedron
2000, 56, 10221. (d) Browder, C. C.; Marmsater, F. P.; West, F. G. Org.
Lett. 2001, 3, 3033. (e) Nair, V.; Bindu, S.; Sreekumar, V.; Chiaroni, A.
Org. Lett. 2002, 4, 2821. (f) White, T. D.; West, F. G. Tetrahedron Lett.
2005, 46, 5629. (g) Dhoro, F.; Tius, M. A. J. Am. Chem. Soc. 2005, 127,
12472.
(3) For a recent review, see: Hanson, J. R. In ComprehensiVe Organic
Synthesis; Trost, B. M., Fleming, I., Pattenden, G., Eds.; Pergamon
Press: Oxford, 1991; Vol. 3, Chapter 3.1, pp 705-720.
R,10 either a hydride shift (path 2c) or an sp2 carbon shift (path 2d)
occurs. Finally, loss of Lewis acid gives products c and d,
respectively. Our results indicate that the alkyl shift (path 2d) is
favored when the R group is electron-rich and can stabilize the
adjacent tertiary cation (Table 2, entries 2, 3, 5, and 7).11 The
governing principle behind the selective formation of products c is
not as clear. R groups that cannot provide strong stabilization for
the cation intermediate B, for either steric (entry 1) or electronic
reasons (entries 4 and 6), may favor the hydride shift (path 2c).
The stereochemical results from the rearrangements of substrates
6a and 7a were consistent with the proposed mechanism. Supra-
facial shift of the hydride in intermediate B would be expected to
place the hydrogen syn to the vinyl group, and the only diastereomer
observed in the rearrangement of 6a was 6c. Similarly, suprafacial
shift of carbon would be expected to place the preexisting methyl
group syn to the vinyl group, as seen in the rearrangement of 7a to
7d.
(4) (a) Motoyoshiya, J.; Yazaki, T.; Hayashi, S. J. Org. Chem. 1991, 56, 735.
(b) Paquette, L.; Fristad, W.; Dime, D.; Bailey, T. J. Org. Chem. 1980,
45, 3017. (c) Ohloff, G.; Schulte-Elte, K.; Demole, E. HelV. Chim. Acta
1971, 54, 2913. (d) Gruhn, A. G.; Reusch, W. Tetrahedron 1993, 49,
8159. (e) Chiu, P.; Li, S. L. Org. Lett. 2004, 6, 613. See also: (f) Denmark,
S.; Hite, G. HelV. Chim. Acta 1988, 71, 195.
(5) He, W.; Sun, X.; Frontier, A. J. J. Am. Chem. Soc. 2003, 125, 14278;
2004, 126, 10493 (addition/correction).
(6) In studies on E/Z mixtures of alkylidene â-ketoesters a, we found that
only the Z isomer cyclizes to give product. The E isomer does not cyclize
under the reaction conditions, but it does isomerize to the Z isomer. Thus,
the product is derived from the cyclization of the Z isomer only.
(7) Davies, I. W.; Senanayake, C. H.; Larsen, R. D.; Verhoeven, T. R.; Reider,
P. J. Tetrahedron Lett. 1996, 37, 1725.
(8) For descriptions of the spectroscopic data and analyses that led to structure
assignments, see the Supporting Information.
(9) For the preparation of Cu(SbF6)2, see the Supporting Information.
(10) Collins, C. J. J. Am. Chem. Soc. 1955, 77, 5517.
(11) The presence of bridged cationic intermediates seems likely; see: Cram,
D. J. J. Am. Chem. Soc. 1964, 86, 3767.
The highest selectivity is observed when 1 equiv of a Cu(II)
complex is present. This would seem to indicate that both carbonyl
oxygens must be bound to the promoter before efficient rearrange-
ment to products c and d can occur (see A, Scheme 1). One
(12) These findings are consistent with those of Chiu and co-workers, who
also propose a role for Lewis basic solvent as a control element in a similar
rearrangement sequence (ref 4e).
JA0716148
9
J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007 8061