Selenium-catalyzed halolactonization: Nucleophilic activation of electrophilic halogenating reagents

Article abstract of DOI:10.1021/jo048460o

Diphenyl diselenide catalyzes the halolactonization of unsaturated acids with N-halosuccinimides under mild conditions. The diselenide not only accelerates the reactions, but in some cases affords regiocontrol in favor of γ-lactone products. Experiments s

Full text of DOI:10.1021/jo048460o

quite useful, the catalytic oxidation produces freely
diffusing electrophilic bromine species and thus leaves
little hope for control of selectivity. With the goal of
catalytically halogenating organic substrates through
nucleophilic activation of typical electrophilic halogen
sources, we have turned to the use of similar selenium
catalysts with “preoxidized” halogenating agents. Herein
we report that simple arylselenides catalyze the halolac-
tonization of unsaturated acids in the presence of N-
halosuccinimide oxidants. Furthermore, in some cases
the use of selenium catalysts allows kinetic control of the
regioselectivity and thus the lactone ring size.
Selenium-Catalyzed Halolactonization:
Nucleophilic Activation of Electrophilic
Halogenating Reagents
Shelli R. Mellegaard and Jon A. Tunge*
Department of Chemistry, University of Kansas,
Lawrence, Kansas 66045
tunge@ku.edu
Received September 1, 2004
In an effort to develop catalytic halogenation reactions,
we recently investigated the selenium-catalyzed allylic
halogenation of â,γ-unsaturated acids (eq 1).9 While the
Abstract: Diphenyl diselenide catalyzes the halolactoniza-
tion of unsaturated acids with N-halosuccinimides under
mild conditions. The diselenide not only accelerates the
reactions, but in some cases affords regiocontrol in favor of
γ-lactone products. Experiments show that the regioselec-
tivity in favor of γ-lactones is a result of kinetic rather than
thermodynamic control.
In the years since its discovery in the early 1900s,
halolactonization has proven to be a versatile reaction
in organic synthesis, allowing facile formation of small
or medium ring size lactones.1 The utility of the product
halolactones has been repeatedly demonstrated by their
use in total syntheses.2 The first reports of halolacton-
ization utilized a weak base, KI, and molecular iodine1
for the cyclofunctionalization of unsaturated acids. Simi-
larly, reagents such as N-bromosuccinimide,3 Br2,4 H2O2/
reaction was apparently general for alkyl-substituted
olefins, upon switching to styrylacetic acid we discovered
the ability of selenium reagents to catalyze halolacton-
izations. Specifically, trans-styrylacetic acid (R ) Ph)
undergoes cyclization in the presence of phenylselenyl
chloride and N-chlorosuccinimide to afford only the
â-chloro-γ-lactone product (2). In the absence of PhSeCl,
<1% of 2 is formed under the same conditions. In this
instance, no allylic halogenation was observed. Attempts
to favor chlorolactonization of alkyl-substituted unsatur-
ated acids by addition of bases (K2CO3, pyridine, Et3N,
and i-Pr2NEt) invariably led to mixtures of allyl chlorides
and chlorolactones. Furthermore, during the slower
reactions, the base-initiated dehydrohalogenation of the
chlorolactones resulted in the formation of butenolides
as well.10
8
NaBr,5 CuBr2/Al2O3,6 Tl2CO3/Br2,7 and ZnBr2/Pb(OAc)4
have been shown to effect bromolactonization of unsatur-
ated acids or carboxylates. Despite the utility of these
reactions, control of the regio- and stereochemistry of the
product bromolactones has not been properly addressed.
With this in mind, it is particularly surprising that there
are no reports of catalytic halolactonization. In this
regard, it is noteworthy that Detty has developed a
selenium-catalyzed protocol for halolactonization of un-
saturated acids with H2O2/NaX;5 however, this method
is best described as a catalytic oxidation of halides. While
The lack of generality of the catalytic chlorolactoniza-
tion led us to investigate the analogous catalytic bromo-
lactonizations. Specifically, trans-styrylacetic acid was
treated with 5 mol % of PhSeSePh and 1.1 equiv of
N-bromosuccinimide (NBS) in CH3CN at -30 °C. After
3 h, the â-bromo-γ-lactone was isolated in 90% yield.
While the addition of catalytic PhSeSePh provided for
an increase in yield relative to the reported 54% with
NBS alone,3 styrylacetic acid has a strong preference for
formation of γ-lactones so regiochemistry was not an
issue. In the interest of investigating catalyst-controlled
regiochemistry, we employed trans-3-hexenoic acid as a
substrate since it is known to react with poor regiochemi-
cal selectivity. The reaction of trans-3-hexenoic acid (1b)
with NBS in acetonitrile at room temperature affords a
2:1 mixture of γ- and â-lactones (3b and 4b, eq 2).
Addition of 5 mol % of PhSeSePh to the reaction mixture
(1) (a) Bougault, J. Ann. Chim. Phys. 1908, 14, 145. (b) Bougault,
J. Ann. Chim. Phys. 1908, 15, 296. (c) House, H. O. Modern Synthetic
Reactions; W. A. Benjamin, Inc.: New York, 1972; p 441. (d) Dowle,
M. D.; Davies, D. I.; Chem. Soc. Rev. 1979, 8, 171-197. (e) Cardillo,
G.; Orena, M. Tetrahedron 1990, 46, 3321-3408.
(2) (a) Murata, Y.; Kamino, T.; Aoki, T.; Hosokawa, S.; Kobayashi,
S. Angew. Chem., Int. Ed. 2004, 43, 3175-3177. (b) Bodkin, J. A.;
Humphries, E. J.; McLeod, M. D. Tetrahedron Lett. 2003, 44, 2869-
2872. (c) Jung, M.; Ham, J.; Song, J. Org. Lett 2002, 4, 2763-2765.
(d) Schultz, A. G.; Kirincich, S. J. J. Org. Chem. 1996, 61, 5626-5630.
(e) Corey, E. J.; Trybulski, E. J.; Melvin, L. S., Jr.; Nicolaou, K. C.;
Secrist, J. A.; Lett, R.; Sheldrake, P. W.; Falck, J. R.; Brunelle, D. J.;
Haslanger, D. F.; Kim, S.; Yoo, S.-e. J. Am. Chem. Soc. 1978, 100,
4618-4620.
(3) (a) Terashima, S.; Jew, S. Tetrahedron Lett. 1977, 11, 1005-
1008. (b) Jew, S. Arch. Pharm. Res. 1982, 5, 97-101. (c) Cook, C.; Cho,
Y.; Jew, S.; Suh, Y.; Kang, E. Arch. Pharm. Res. 1983, 6, 45-53.
(4) Berti, B. Tetrahedron 1958, 4, 393-402.
(5) (a) Drake, M. D.; Bateman, M. A.; Detty, M. R. Organometallics
2003, 22, 4158-4162. (b) Goodman, M. A.; Detty, M. R. Organome-
tallics 2004, 23, 3016-3020.
(6) Rood, G. A.; DeHaan, J. M.; Zibuck, R. Tetrahedron Lett. 1996,
37, 157-158.
(7) Cambie, R. C.; Rutledge, P. S.; Somerville, R. F.; Woodgate, P.
D. Synthesis 1988, 1009-1011.
(9) Tunge, J. A.; Mellegaard, S. R. Org. Lett. 2004, 6, 1205-1207.
(10) Direct selenium-catalyzed formation of butenolides is also
possible: Tiecco, M. Top. Curr. Chem. 2000, 208, 7-54.
(8) Motohashi, S.; Satomi, M. Heterocycles 1985, 23, 2035-2039.
10.1021/jo048460o CCC: $27.50 © 2004 American Chemical Society
Published on Web 11/11/2004
J. Org. Chem. 2004, 69, 8979-8981
8979

TABLE 1. Yields and Selectivities of
Selenium-Catalyzed Bromolactonizations
minal â,γ-unsaturated acids, which are known to strongly
favor formation of â-lactones.11
In an effort to establish the nature of the regiocontrol
by PhSeSePh, we tested whether the product ratio was
the result of kinetic or thermodynamic control. We
thought it was possible that PhSeSePh/NBS simply
catalyzed the equilibration of â-lactone to the thermo-
dynamically favored γ-lactone. To address this issue a
2:1 mixture of γ- and â-lactones was generated by
uncatalyzed bromolactonization of trans-3-hexenoic acid
followed by treatment with either 5 mol % of PhSeSePh
or with the combination of PhSeSePh and 1 equiv of NBS.
After 2 h at room temperature, no equilibration of
â-lactone to the γ-lactone was observed in either case.
Similarly, treatment of dibrominated product 5b under
the conditions of catalysis does not result in formation
of γ-lactone 3b. Collectively, these reactions demonstrate
that the product selectivity is a result of kinetic rather
than thermodynamic control.
Next, we felt it was necessary to substantiate the
catalytic effect of diphenyl diselenide despite the causal
evidence in the change of kinetic product ratios. Toward
this end, variable-temperature NMR spectroscopy al-
lowed us to directly observe the reaction under the
conditions of the reaction. Two reactions, one without
PhSeSePh and one with PhSeSePh, were run with trans-
3-hexenoic acid under otherwise identical conditions.
Specifically, a 0.32 M trans-3-hexenoic acid solution in
CD3CN was treated with 5 mol % of PhSeSePh then
cooled to -30 °C. NBS (1.1 equiv) was added, and the
reaction mixture was placed in a precooled NMR probe
at -30 °C. While the reaction mixture containing Ph-
SeSePh had completely consumed the soluble NBS before
the first spectrum was obtained, the control reaction had
consumed <5% of the dissolved NBS. Although the
soluble NBS was rapidly consumed to give halolactone
when PhSeSePh was present, the hexenoic acid was not
completely converted to product (30% conversion). Fur-
ther reaction was more sluggish as evidenced by the low
conversion (6%) of the remaining hexenoic acid over the
following 10 min. This is a consequence of the low
solubility of NBS in CD3CN at -30 °C. From this
information, it was surmised that the catalyzed reaction
rate is limited by the rate of dissolution of NBS at low
temperature.12 To test this idea, the reaction mixture was
warmed to -15 °C. At this temperature, product lactone
began to form rapidly in the mixture containing Ph-
SeSePh. Once again, free NBS was not observed, indicat-
ing mass transport limited kinetics. In contrast, the
uncatalyzed reaction did not ensue despite the observable
increase in NBS concentration. These experiments un-
equivocally establish that PhSeSePh is a catalyst for
bromolactonization.
a γ vs â or γ vs δ. b Reactions were run with 5 mol % of
PhSeSePh, 1.1 equiv of NBS, CH3CN, -30 °C. c Combined yield.
d Mixture of diastereomers reflecting the E/Z ratio of starting
material. e 5:1 mixture of diastereomers. f 3:1 mixture of diaster-
eomers.
significantly improved the selectivity, yielding a 17:1
mixture of γ- and â-lactones. The selenium-catalyzed
reaction also afforded a small amount (7%) of dibromi-
nated acid (5b), which was not present in the uncatalyzed
product mixture.
Other substrates were treated under our established
conditions and in most cases the bromolactones were
afforded in moderate to good yields (Table 1). For
substrates that generally favor formation of γ-lactones,
the addition of PhSeSePh only affects the time and/or
conditions required for cyclization. However, for sub-
strates where regiochemistry is an issue, catalytic quan-
tities of PhSeSePh also favor formation of γ-lactones. For
example, cyclohexenylacetic acid provides â-lactone ex-
clusively when treated with NBS alone, but with 5 mol
% of PhSeSePh γ-lactone is favored. Attempts to lower
the catalyst loading to 1 mol % resulted in total loss of
regiocontrol. Furthermore, PhSeSePh did not provide any
regiocontrol in the catalytic bromolactonization of ter-
The observation of small amounts of dibrominated
product 5b indicated that it was possible that molecular
bromine was being formed and was responsible for the
observed bromolactonization as well. Thus, trans-3-
hexenoic acid was treated with Br2 at -30 °C in CH3CN
in both the presence and absence of PhSeSePh. In both
(11) Shibata, I.; Toyota, M.; Baba, A.; Matsuda, H. J. Org. Chem.
1990, 55, 2487-2491.
(12) The rate of NBS dissolution is much higher in larger scale
reactions where vigorous stirring is possible.
8980 J. Org. Chem., Vol. 69, No. 25, 2004

SCHEME 1
cases, the reaction cleanly afforded a 1:1 ratio of γ-lactone
3b and dibromination product 5b. Thus, the 17:2 ratio
of 3b:5b in the selenium-catalyzed reaction is inconsis-
tent with the formation of molecular bromine. On the
basis of this evidence, and the fact that selenium provides
for kinetic control of the product ratio of halolactoniza-
tion, we suggest that the reaction occurs through a
selenium-bound electrophilic bromine species. We cannot
rule out the possibility that dibrominated acid 5b is
formed from a low concentration of freely diffusing
bromine.
Finally, we wanted to address catalyst speciation.
Sharpless has demonstrated that PhSeSePh reacts with
N-chlorosuccinimide to give two separate selenium spe-
cies, phenylselenyl chloride and phenylselenyl succinim-
ide (eq 3).13 In light of this precedent, we became curious
regiocontrol, we suggest that lactone formation proceeds
through nucleophilic displacement of a selenium-coordi-
nated bromonium ion (B). Such a cyclization is reminis-
cent of halocyclizations with bromopyridinium reagents,
where the pyridine derivative remains coordinated to
bromine during lactone ring formation.16
In conclusion, we have demonstrated that diphenyl
diselenide is an efficient catalyst for the halolactonization
of unsaturated acids. This is to our knowledge the first
example of catalytic halolactonization. Moreover, sele-
nium catalytically activates halogen oxidants toward
reaction, which is a new approach to oxidative halogena-
tion. In the future, we hope to take advantage of our
ability to alter the ligand environment of selenium to
achieve higher degrees of regioselectivity.
whether PhSeBr or phenylselenyl succinimide was singly
responsible for the catalytic bromolactonization. In an
effort to determine the active catalytic species, phenyl-
selenyl bromide and N-phenylselenopthalimide14were
separately screened as catalysts for the halolactonization
of trans-3-hexenoic acid. While both catalysts effected
halolactonization, the 3b:5b product ratios (4:1 for phen-
ylselenyl bromide, and 2:1 for N-phenylselenopthalimide)
obtained were not consistent with that observed with
PhSeSePh. The fact that neither catalyst reproduced the
experimental results given by PhSeSePh suggests that
the diselenide may not be oxidatively cleaved by NBS
during catalysis.
Although more experiments are required to elucidate
the mechanism of catalysis, we suggest the following
mechanism based on the available experimental results.
First, we propose that the NBS undergoes nucleophilic
attack by the diselenide to produce the cationic selenium
complex (A) with a succinimide counterion (Scheme 1).
This transformation is equivalent to the first step in the
well-known oxidative cleavage of diselenides with mo-
lecular bromine.15 The resulting succinimide anion is
expected to rapidly deprotonate the carboxylic acid (1)
to form the carboxylate. To rationalize the observed
Experimental Section
General Procedure for Bromolactonization of â,γ- or
γ,δ-Unsaturated Acids. Diphenyl diselenide (0.05 mmol) was
dissolved in 5 mL of CH3CN (stored over 4 Å MS), producing a
yellow solution. The unsaturated acid (1.00 mmol) was added,
and the resulting mixture was cooled to -30 °C. Next, N-
bromosuccinimide (1.1 mmol) was added and the resulting
reaction mixture was stirred for the reported time. The resulting
solution was concentrated to <1 mL, and 10 mL of diethyl ether
was added. The ether was decanted from the solid and washed
with H2O (2 × 3 mL). The resulting ether layer was dried over
MgSO4 and concentrated, and the residue was purified by flash
chromatography (100% methylene chloride).
Acknowledgment. We thank the EPSCoR program
of the National Science Foundation for funding.
Supporting Information Available: Spectroscopic data
for new compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
(13) (a) Hori, T.; Sharpless, K. B. J. Org. Chem. 1979, 44, 4208-10.
(b) Chabaud, B.; Sharpless, K. B. J. Org. Chem. 1979, 44, 4204-4208.
(14) N-Phenylselenophthalimide is a more thermally stable form of
N-phenylselenosuccinimide. Nicolaou, K. C.; Seitz, S. P.; Sipio, W. J.;
Blount, J. F. J. Am. Chem. Soc. 1979, 101, 3884-3893.
(15) (a) Sharpless, K. B.; Lauer, R. F. J. Org. Chem. 1974, 39, 429-
430. (b) Clive, D. L. J.; Chittattu, G.; Curtis, N. J.; Kiel, W. A.; Wong,
C. K. J. Chem Soc., Chem. Commun. 1977, 725-727.
JO048460O
(16) (a) Neverov, A. A.; Brown, R. S. J. Org. Chem. 1998, 63, 5977-
82. (b) Neverov, A. A.; Feng, H. X.; Hamilton, K.; Brown, R. S. J. Org.
Chem. 2003, 68, 3802-10.
J. Org. Chem, Vol. 69, No. 25, 2004 8981