Triethylborane-induced bromine atom-transfer radical addition in aqueous media: Study of the solvent effect on radical addition reactions

Article abstract of DOI:10.1021/jo010652l

A mixture of ethyl bromoacetate and 1-octene was treated with triethylborane in water at ambient temperature to provide ethyl 4-bromodecanoate in good yield. The bromine atom-transfer radical addition in benzene was not satisfactory. The addition proceede

Full text of DOI:10.1021/jo010652l

7776
J . Org. Chem. 2001, 66, 7776-7785
Tr ieth ylbor a n e-In d u ced Br om in e Atom -Tr a n sfer Ra d ica l Ad d ition
in Aqu eou s Med ia : Stu d y of th e Solven t Effect on Ra d ica l
Ad d ition Rea ction s
Hideki Yorimitsu, Hiroshi Shinokubo, Seijiro Matsubara, and Koichiro Oshima*
Department of Material Chemistry, Graduate School of Engineering, Kyoto University, Yoshida,
Sakyo-ku, Kyoto 606-8501, J apan
Kiyoyuki Omoto and Hiroshi Fujimoto*
Division of Molecular Engineering, Graduate School of Engineering, Kyoto University, Yoshida,
Sakyo-ku, Kyoto 606-8501, J apan
oshima@fm1.kuic.kyoto-u.ac.jp
Received J une 26, 2001
A mixture of ethyl bromoacetate and 1-octene was treated with triethylborane in water at ambient
temperature to provide ethyl 4-bromodecanoate in good yield. The bromine atom-transfer radical
addition in benzene was not satisfactory. The addition proceeded smoothly in polar solvents such
as DMF and DMSO, protic solvents such as 2,2,2-trifluoroethanol and 1,1,1,3,3,3-hexafluoro-2-
propanol, and aqueous media. Ab initio calculations were conducted to reveal the origin of the
solvent effect of water in the addition reaction. The polar effect of solvents, which is judged by the
dielectric constant, on the transition states in the bromine atom-transfer and radical addition steps
is moderately important. Calculations show that a polar solvent tends to lower the relative energies
of the transition states. The coordination of a carbonyl group to a proton in a protic solvent, like a
Lewis acid, would also increase the efficiency of the propagation.
Sch em e 1
In tr od u ction
Halogen atom-transfer addition has been extensively
studied and widely used in organic synthesis.1 The
addition of a C-X bond across a double bond was
pioneered by Kharasch2 (Scheme 1) and provides new
C-C and C-X bonds in a single operation. The choice of
the halogen that transfers in the reaction is crucial for
the success of atom-transfer additions. In the case of
R-iodo esters employed as the halide, the reaction pro-
ceeded smoothly to give the corresponding adducts under
mild conditions.3 Fast iodine atom transfer from an R-iodo
ester to a radical generated by the addition suppresses
side reactions and sustains the radical chain. However,
the addition of R-bromo esters usually requires high
reaction temperatures.4 The addition of ethyl bromo-
acetate to 1-octene was performed at 90 °C without
solvent in the presence of highly explosive peracetic
acid.2a Very recently, it was reported that bromine atom-
transfer radical addition of N-bromoacetyl-2-oxazoli-
dinone proceeded smoothly in the presence of a Lewis
acid at 25 °C.5 The addition resulted in poor conversion
without Lewis acids such as Yb(OTf)3 and Sc(OTf)3. The
use of Lewis acids in radical reactions often enhances a
reaction rate and controls stereoselectivity.6,7
During the last two decades, organic reactions in
aqueous media have been attracting increasing atten-
tion.8 On the other hand, radical reactions in water are
immature and have not been studied extensively.9 We
have been investigating radical reactions in aqueous
media and disclosed the remarkable solvent effects of
water.10 Here, we wish to report the bromine atom-
transfer radical addition of R-bromoacetate with 1-alkene
in water. The reaction in water could be performed at
ambient temperature in the presence of triethylborane11
to yield 4-bromoalkanoate, which will be converted into
γ-lactone2a,12 or γ-lactam.13 We have also investigated the
* To whom correspondence should be addressed. Phone: +81-75-
753-5523. Fax: +81-75-753-4863.
(1) (a) Curran, D. P. Synthesis 1988, 489-513. (b) J asperse, C. P.;
Curran, D. P.; Fevig, T. L. Chem. Rev. 1991, 91, 1237-1286. (c) Byers,
J . In Radicals in Organic Synthesis; Renaud, P., Sibi, M. P., Eds.;
Wiley-VCH: Weinheim, 2001; Vol. 1, Chapter 1.5.
(2) (a) Kharasch, M. S.; Skell, P. S.; Fisher, P. J . Am. Chem. Soc.
1948, 70, 1055-1059. (b) Kharasch, M. S.; J ensen, E. V.; Urry, W. H.
Science 1945, 102, 128.
(3) (a) Curran, D. P.; Chen, M.-H.; Splezter, E.; Seong, C. M.; Chang,
C.-T. J . Am. Chem. Soc. 1989, 111, 8872-8878. (b) Curran, D. P.;
Seong, C. M. J . Am. Chem. Soc. 1990, 112, 9401-9403. (c) Baciocchi,
E.; Muraglia, E. Tetrahedron Lett. 1994, 35, 2763-2766. (d) Nakamura,
T.; Yorimitsu, H.; Shinokubo, H.; Oshima, K. Synlett 1998, 1351-1352.
(4) The addition of diethyl bromomalonate or bromomalononitrile
proceeded smoothly: (a) Giese, B.; Horler, H.; Leising, M. Chem. Ber.
1986, 119, 444-452. (b) Riemenschneider, K.; Bartels, H. M.; Dornow,
R.; Drechsel-Grau, E.; Eichel, W.; Luthe, H.; Matter, Y. M.; Michaelis,
W.; Boldt, P. J . Org. Chem. 1987, 52, 205-212. (c) Yoshida, J .;
Yamamoto, M.; Kawabata, N. Tetrahedron Lett. 1985, 26, 6217-6220.
(5) Mero, C. L.; Porter, N. A. J . Am. Chem. Soc. 1999, 121, 5155-
5160.
(6) For a review of Lewis acid-promoted radical reactions, see: (a)
Renaud, P.; Gerster, M. Angew. Chem., Int. Ed. 1998, 37, 2562-2579.
(b) Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry of Radical
Reactions; VCH: Weinheim, 1996. (c) Reference 1c, Chapter 4. (d)
Murakata, M.; Hoshino, O. J . Synth. Org. Chem. J pn. 2001, 59, 560-
568.
10.1021/jo010652l CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/26/2001

Solvent Effect on Br Atom-Transfer Radical Addition
J . Org. Chem., Vol. 66, No. 23, 2001 7777
Sch em e 2
solvent effect in detail by performing the reaction in
various solvents and found that a polar solvent as well
as a protic solvent gave better yields. Furthermore, ab
initio calculations have revealed the origin of the solvent
effect on bromine atom-transfer addition.
Ra d ica l Ad d ition of r-Br om o Ca r bon yl
Ta ble 1. Tr ieth ylbor a n e-In d u ced Ra d ica l Ad d ition of
Eth yl Br om oa ceta te to Alk en es in Wa ter
Com p ou n d s to Alk en es
Triethylborane (1.0 M ethanol solution,14 0.50 mL, 0.50
mmol) was added to a suspension of ethyl bromoacetate
(1a , 5.0 mmol) and 1-octene (2a , 1.0 mmol) in water (5
mL) under argon. Air (10 mL) was then introduced to
the reaction flask by a syringe with vigorous stirring. Air
was injected every 30 min. The reaction mixture was
heterogeneous yet clear. After 1.5 h of reaction, extractive
entry
2
R
3
yield (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
2a
2b
2c
2d
2e
2f
2g
2h
2i
n-C6H13
n-C10H21
n-C20H41
3a a
3a b
3a c
3a d
3a e
3a f
3a g
3a h
3a i
80c
65c
(7) Recent examples of radical reactions employing a Lewis acid:
(a) Sibi, M. P.; J asperse, C. P.; J i, J . J . Am. Chem. Soc. 1995, 117,
10779-10780. (b) Guindon, Y.; Lavalle´e, J .-F.; Llinas-Brunet, M.;
Horner, G.; Rancourt, J . J . Am. Chem. Soc. 1991, 113, 9701-9702. (c)
Yang, D.; Ye, X.-Y.; Gu, S.; Xu, M. J . Am. Chem. Soc. 1999, 121, 5579-
5580. (d) Yang, D.; Ye, X.-Y.; Xu, M.; Pang, K.-W.; Cheung, K. K. J .
Am. Chem. Soc. 2000, 122, 1658-1663. (e) Renaud, P.; Moufid, N.;
Kuo, L. H.; Curran, D. P. J . Org. Chem. 1994, 59, 3547-3552. (f)
Murakata, M.; J ono, T.; Mizuno, Y.; Hoshino, O. J . Am. Chem. Soc.
1997, 119, 11713-11714. (g) Murakata, M.; Tsutsui, H.; Hoshino, O.
Org. Lett. 2001, 3, 299-302. (h) Friestad, G. K.; Qin, J . J . Am. Chem.
Soc. 2000, 122, 8329-8330. (i) Sibi, M. P.; J i, J .; Sausker, J . B.;
J asperse, C. P. J . Am. Chem. Soc. 1999, 121, 7517-7526. (j) Ishibashi,
H.; Matsukida, H.; Toyao, A.; Tamura, O.; Takeda, Y. Synlett 2000,
1497-1499. (k) Porter, N. A.; Zhang, G.; Reed, A. D. Tetrahedron Lett.
2000, 41, 5773-5777. (l) Porter, N. A.; Feng, H.; Kavrakaova, I. K.
Tetrahedron Lett. 1999, 40, 6713-6716. (m) Yamamoto, Y.; Onuki, S.;
Yumoto, M.; Asano, N. J . Am. Chem. Soc. 1994, 116, 421-422. (n)
Guindon, Y.; J ung, G.; Guerin, B.; Ogilvie, W. W. Synlett 1998, 213-
220. (o) Mase, N.; Watanabe, Y.; Toru, T. Tetrahedron Lett. 1999, 40,
2797-2800. (p) Nishida, M.; Nishida, A.; Kawahara, N. J . Org. Chem.
1996, 61, 3574-3575. (q) Hayen, A.; Koch, R.; Saak, W.; Haase, D.;
Metzger, J . O. J . Am. Chem. Soc. 2000, 122, 12458-12468. (r) Enholm,
E. J .; Cottone, J . S.; Allais, F. Org. Lett. 2001, 3, 145-147. (s) Sibi, M.
P.; Rheault, T. R. J . Am. Chem. Soc. 2000, 122, 8873-8879. (t) Sibi,
M. P.; Shay, J . J .; J i, J . Tetrahedron Lett. 1997, 38, 5955-5958. (u)
Urabe, H.; Yamashita, K.; Suzuki, K.; Kobayashi, K.; Sato, F. J . Org.
Chem. 1995, 60, 3576-3577. (v) Kim, K.; Okamoto, S.; Sato, F. Org.
Lett. 2001, 3, 67-69.
(8) For a review of reactions in aqueous media: (a) Li, C.-J .; Chan,
T.-H. Organic Reactions in Aqueous Media; J ohn Wiley & Sons: New
York, 1997. (b) Grieco, P. A. Organic Synthesis in Water; Blackie
Academic & Professional: London, 1998. (c) Lubineau, A.; Auge, J . In
Modern Solvents in Organic Synthesis; Knochel, P., Ed.; Springer-
Verlag: Berlin, 1999.
(9) (a) Minisci, F. Synthesis 1973, 1-24. (b) Yamazaki, O.; Togo,
H.; Nogami, G.; Yokoyama, M. Bull. Chem. Soc. J pn. 1997, 70, 2519-
2523. (c) Breslow, R.; Light, J . Tetrahedron Lett. 1990, 31, 2957-2958.
(d) J ang, D. O. Tetrahedron Lett. 1998, 39, 2957-2958. (e) Maitra, U.;
Sarma, K. D. Tetrahedron Lett. 1994, 35, 7861-7862. (f) Bietti, M.;
Baciocchi, E.; Engberts, J . B. F. N. J . Chem. Soc., Chem. Commun.
1996, 1307-1308. (g) Miyabe, H.; Ueda, M.; Naito, T. J . Org. Chem.
2000, 65, 5043-5047. (h) Petrier, C.; Dupuy, C.; Luche, J . L. Tetra-
hedron Lett. 1986, 27, 3149-3152. (i) Giese, B.; Damm, W.; Roth, M.;
Zehnder, M. Synlett 1992, 441-443 (see also ref 7i).
(10) (a) Yorimitsu, H.; Nakamura, T.; Shinokubo, H.; Oshima, K.
J . Org. Chem. 1998, 63, 8604-8605. (b) Yorimitsu, H.; Nakamura, T.;
Shinokubo, H.; Oshima, K.; Omoto, K.; Fujimoto, H. J . Am. Chem. Soc.
2000, 122, 11041-11047. (c) Yorimitsu, H.; Shinokubo, H.; Oshima,
K. Chem. Lett. 2000, 104-105. (d) Wakabayashi, K.; Yorimitsu, H.;
Shinokubo, H.; Oshima, K. Bull. Chem. Soc. J pn. 2000, 73, 2377-
2378. (e) Yorimitsu, H.; Shinokubo, H.; Oshima, K. Bull. Chem. Soc.
J pn. 2001, 74, 225-235 (see also ref 3d).
(11) (a)Nozaki, K.; Oshima, K.; Utimoto, K. J . Am. Chem. Soc. 1987,
109, 2547-2548. (b) Oshima, K.; Utimoto, K. J . Synth. Org. Chem.
J pn. 1989, 47, 40-52. (c) Yorimitsu, H.; Oshima, K. In Radicals in
Organic Synthesis; Renaud, P., Sibi, M. P., Eds.; Wiley-VCH: Wein-
heim, 2001; Vol. 1, Chapter 1.2.
(12) (a) Anzalone, L.; Hirsch, J . A. J . Org. Chem. 1985, 50, 2128-
2133. (b) Tsuboi, S.; Muranaka, K.; Sakai, T.; Takeda, A. J . Org. Chem.
1986, 51, 4944-4946. (c) Curran, D. P.; Chang, C.-T. J . Org. Chem.
1989, 54, 3140-3157.
(13) Fujita, M.; Kitagawa, O.; Yamada, Y.; Izawa, H.; Hasegawa,
H.; Taguchi, T. J . Org. Chem. 2000, 65, 1108-1114.
(14) An ethanol solution of triethylborane was prepared according
to the reported procedure. See ref 10b.
79c
(CH2)4OH
81
26
84
63
79
69
82
CH2OH
CH2CH2COCH3
(CH2)3Br
(CH2)8COOCH3
(CH2)4OTHP
(CH2)3NPHd
cyclohexene
trans-4-octene
2-methyl-1-dodecene
2j
2k
2l
3a j
3a k
3a l
22 (54/46)c,e
10 (1/1)c,e
77c
2m
3a m
a
b
Ethanol solution (1.0 M). Air was added every 30 min.
c Additional triethylborane (0.50 mmol) was added 1 h after the
reaction started. NPH ) phthalimidyl group. e Diastereomer
d
ratios are in parentheses.
workup followed by silica gel column purification pro-
vided ethyl 4-bromodecanoate (3a a ) in 74% yield (Scheme
2). In contrast to the reaction in water, the reaction
proceeded sluggishly in benzene and dichloromethane (12
and 9%, respectively) under similar reaction conditions.
When the reaction mixture in water was kept strictly
under argon until concentration, poor conversion was
seen (<7%). An excess of 1a was crucial to obtain 3a a in
satisfactory yield.15 The reaction of stoichiometric 1a with
2a gave 3a a in only a 10% yield with stirring for 1.5 h.
Employing 5 equimolar amounts of 2a also resulted in a
miserable yield (19% based on 1a ). However, 3a a was
satisfactorily obtained in 71% yield upon treatment of a
mixture of 2a and 1.3 mol of 1a with triethylborane (0.50
mmol × 2) for 3 h. These observations suggest that the
rate-determining step in the chain propagation would be
the bromine atom-transfer step.
The results of the addition of 1a (5 equiv) to various
olefins in water are summarized in Table 1. A wide range
of functionalities in 2d -2j could survive under the
reaction conditions. Interestingly, waxy 1-docosene (2c)
also reacted with 1a in an aqueous medium, although
the mixture was cloudy during the reaction. Allyl alcohol
(2e) was not a good substrate. Internal double bonds 2k
and 2l were less reactive (entries 11 and 12). 1,1-
Disubstituted alkene could be employed to yield the
corresponding tertiary bromide 3a m .
Not only ethyl bromoacetate but also other reactive
bromides underwent radical addition (Table 2). Bromo-
malonate 1c was so reactive4 that the addition requires
only 1.5 equimolar amounts of 1c. Similar to bromo-
(15) Iodine atom-transfer addition of R-iodo-γ-butyrolactone in water
proceeded smoothly to yield the corresponding adduct. In this case,
an excess of alkene was used. See ref 3d.

7778 J . Org. Chem., Vol. 66, No. 23, 2001
Yorimitsu et al.
Ta ble 2. Tr ieth ylbor a n e-In d u ced Ra d ica l Ad d ition of
r-Br om o Ca r bon yl Com p ou n d s to 1-Octen e in Wa ter
Sch em e 4
materials (Scheme 4). As described above, an excess of
bromide is necessary to accomplish the bromine atom-
transfer reaction. These intramolecular versions invari-
ably consist of the reaction of an R-bromo amide or ester
with 2 or 1 equiv of alkene, respectively. Therefore,
bromine atom transfer would not be sufficiently efficient
to maintain the radical chain in these cyclizations.
Ra d ica l Ad d ition of Eth yl Br om oa ceta te to
1-Octen e in Va r iou s Solven ts
We observed a significant solvent effect in the inter-
molecular bromine atom-transfer addition; that is, radical
addition of 1a to 2a was much easier in water than it
was in benzene or dichloromethane. To investigate the
solvent effect in detail, a series of solvents were surveyed
in the reaction of 1a with 2a under the otherwise same
reaction conditions. Triethylborane (0.50 mmol) was
added to a mixture of 1a (5.0 mmol) and 2a (1.0 mmol)
in a solvent (5 mL) with vigorous stirring, and 5 mL of
air was added by a syringe. After 30 min, air (5 mL) was
again introduced, and the whole mixture was vigorously
stirred for an additional 30 min. Evaporation of the
solvent or extractive workup with hexane yielded a crude
oil. The yield of 3a a was determined by fine NMR
experiments with dibenzyl ether as an internal standard.
Three different procedures for introducing triethylborane
were set up. For reactions in nonpolar solvents such as
benzene or dichloromethane, a hexane solution of trieth-
ylborane was used as an initiator (Method A). An ethanol
solution of triethylborane was added in the case where
ethanol, aqueous ethanol, aqueous THF, or water was
employed as a solvent (Method B). Neat triethylborane
was directly added dropwise when reactions were con-
ducted in other solvents (Method C).
a
b
Ethanol solution (1.0 M). Air was added every 30 min.
c Additional triethylborane (0.50 mmol) was added 1 h after the
d
reaction started. 2.5 mmol of 1b and 0.50 mmol of 2a were used.
e Compound 1b was recovered (1.82 mmol). f A 1.5 mmol sample
of 1c was used.
Sch em e 3
The results are listed in Table 3 with the dielectric
constant16 for the solvents examined. The starting ma-
terials were almost unchanged when hydrocarbons or
halogenated solvents were employed (entries 1-6). The
adduct 3a a was obtained in more than 20% yield in DME
and dioxane, although the reaction proceeded sluggishly
in ether and THF (entries 7-10). The use of N-methyl-
formamide, N-methylacetamide, and DMF gave rise to
a remarkable improvement in yield among polar solvents
having a carbonyl moiety (entries 11-19). N-Methylform-
amide, which has a much larger dielectric constant than
water, gave the best result (83% yield) among the solvent
systems examined. However, formamide itself resulted
in the marginal yield of 3a a . The reason for the difference
is not clear. The reactions in cyclic propylene carbonate
and NMP afforded 3a a in moderate yields. Acetonitrile
and HMPA also performed fairly well (entries 20 and 21).
DMSO gave a satisfactory result (entry 22, 72% yield).
In general, the more polar solvent tends to provide 3a a
in a higher yield.
acetate, bromoacetonitrile (1d ) reacted easily with 2a .
However, bromo amide 1e and N-bromoacetyl-2-oxazo-
lidinone (1f) were recovered after subjection to radical
addition reaction, probably due to the relatively stronger
carbon-bromine bond of bromo amides. Unfortunately,
the reaction of secondary bromo ester 1g and 1h gave a
poor conversion in this system. The addition of 1b with
allyltrimethylsilane yielded a mixture of the adduct
3bn and benzyl 4-pentenoate (4). The crude mixture
(3bn /4 ) 94/6) was treated with tetrabutylammonium
fluoride in THF at 25 °C to furnish 4 in 68% overall yield
(Scheme 3).
The addition proceeded moderately in alcohols to afford
3a a (entries 23-29). It is worth noting that 2,2,2-
Radical cyclization of N,N-diallyl-2-bromoacetamide or
allyl bromoacetate resulted in recovery of the starting
(16) Riddick, J . A.; Bunger, W. B.; Sakano, T. K. Organic Solvent;
J ohn Wiley & Sons: New York, 1986.

Solvent Effect on Br Atom-Transfer Radical Addition
Ta ble 3. Tr ieth ylbor a n e-In d u ced Ra d ica l Ad d ition of 1a to 2a in Va r iou s Solven tsa
yield of dielectric
J . Org. Chem., Vol. 66, No. 23, 2001 7779
yield of
dielectric
entry
solvent
methodb 3a a (%)c constantd
entry
solvent
methodb 3a a (%)c constantd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
hexane
benzene
toluene
chlorobenzene
dichloromethane
chloroform
ether
THF
DME
A
A
A
A
A
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
B
C
C
10
12
6
5
9
3
10
4
23
24
15
18
30
39
16
83
65
67
42
44
48
72
36
35
35
48
1.88
2.27
2.38
5.62
8.93
4.81
4.34
7.58
7.20
2.21
20.56
6.02
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
ethylene glycol
2,2,2-TFEe
C
C
C
B
C
C
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
C
C
16
73
64
70
50
12
63
50
8
37.7
26.67
1,1,1,3,3,3-HFPf
water
79.39
water
no solvent
0.10 M HCl
0.10 M NaOH
waterg
dioxane
acetone
waterh
2
0.10 M NaOHi
14
57
77
81
49
47
48
50
43
53
61
43
46
33
43
ethyl acetate
dimethyl carbonate
propylene carbonate
formamide
N-methylformamide
N-methylacetamide
DMF
NMP
acetonitrile
HMPA
DMSO
methanol
ethanol
isopropyl alcohol
t-butyl alcohol
ethanol/water ) 4.5/0.5
ethanol/water ) 4.0/1.0
ethanol/water ) 3.5/1.5
ethanol/water ) 3.0/2.0
ethanol/water ) 2.5/2.5
ethanol/water ) 2.0/3.0
ethanol/water ) 1.5/3.5
ethanol/water ) 1.0/4.0
ethanol/water ) 0.5/4.5
THF/water ) 4.0/1.0
THF/water ) 2.5/2.5
THF/water ) 1.0/4.0
C6F14
64.92
111.0
182.4
191.3
36.71
33.2
35.94
29.30
46.45
32.66
24.55
19.92
12.47
FC-75j
a
Compounds 1a (5.0 mmol) and 2a (1.0 mmol), triethylborane (0.50 mmol), and solvent (5 mL) were employed. Air (5 mL × 2) was
b
added every 30 min. Reaction time was 1 h. Method A: a hexane solution of triethylborane (1.0 M) was added. Method B: an ethanol
solution of triethylborane (1.0 M) was used. Method C: neat triethylborane was added. c Yields were determined by 1H NMR measurement
with dibenzyl ether as an internal standard. Reproduced from ref 16. e 2,2,2-Trifluoroethanol. f 1,1,1,3,3,3-Hexafluoro-2-propanol.
d
g
h
i
j
Galvinoxyl (10 mol %) was used. TEMPO (10 mol %) was used. 4-CarboxyTEMPO (10 mol %) was employed. Perfluoro-2-
butyltetrahydrofuran.
trifluoroethanol17 was highly effective for this addition
reaction, whereas 3a a was obtained in only 35% yield
when ethanol was used. Thus, the highest yield would
be attributed to the acidity of 2,2,2-trifluoroethanol
(pKa ) 12.8). Moreover, 1,1,1,3,3,3-hexafluoro-2-propanol
(pKa ) 9.3) is also superior to 2-propanol.
ethanol/water ratio. The 3.5/1.5 ethanol/water ratio gave
the highest yield. Interestingly, the reaction in perfluoro-
hexane or perfluoro-2-butyltetrahydrofuran (FC-75) fur-
nished 3a a in a moderate yield.18 Although the reaction
systems are heterogeneous, a higher yield was achieved
in a fluorinated solvent than in hexane or THF. The
reason for the effectiveness of fluorous solvents is not
clear because two possibilities, namely, the acceleration
by fluorophobic interaction19 or the high solubility of
oxygen in a perfluorinated solvent to improve the ef-
ficiency of an initiator,18,20 could not be distinguished.
Radical addition of bromotrichloromethane to 1-octene
was also examined to attain more information about the
solvent effect of bromine atom-transfer radical addition.
The reaction employed a smaller amount of triethylbo-
rane because bromotrichloromethane adds to alkenes
very easily. The results are listed in Table 4. The yields
of 7 depend on the solvent polarity. This tendency is
similar to the result in Table 3. However, two consider-
able differences were observed. One is the lower yield in
the reaction in 2,2,2-trifluoroethanol (60%). The reaction
in ethanol yielded the adduct 7 in 80% yield. These
results indicate that 2,2,2-trifluoroethanol would promote
the radical addition of ethyl bromoacetate through hy-
drogen bonding between the oxygen of the carbonyl group
and the hydrogen of the fluorous alcohol, while bromo-
trichloromethane cannot form such a hydrogen bond that
In water, treatment of a mixture of 1a and 2a with an
ethanol solution of triethylborane provided 3a a in 70%
yield under the same reaction conditions. Utilization of
neat triethylborane, instead of an ethanol solution,
resulted in a lower yield (50%). It is therefore likely that
a small amount of ethanol may increase the solubility of
the reactants in water. Moreover, the reaction without
any solvent led to a poor conversion. Thus, the addition
in water, which always forms a heterogeneous system,
is not a mere solvent-free reaction in the organic phase,
and water clearly plays a critical role as a solvent.
Lipophilic galvinoxyl and TEMPO (10 mol %) inhibited
the reactions. In addition, the reaction in the presence
of 4-carboxyTEMPO in 0.10 M NaOH furnished 3a a in
a very poor yield. In this reaction, 4-carboxyTEMPO was
mostly converted to the corresponding anion and would
trap the radicals in the aqueous phase. These results
suggest that the free radicals that are involved in this
reaction would easily migrate from the aqueous phase
to the organic phase and vice versa. Thus, the reaction
might involve an interfacial process.
Mixed solvent systems, ethanol/water and THF/water,
improved the yield in comparison to ethanol and THF
themselves. Notably, the yields were dependent on an
(18) For a review of reactions in fluorous media: (a) Betzemeier,
B.; Knochel, P. In Modern Solvents in Organic Synthesis; Knochel, P.,
Ed.; Springer-Verlag: Berlin, 1999. (b) Hiyama, T.; Kanie, K.; Kusu-
moto, T.; Morizawa, Y.; Shimizu, M. Organofluorine Compounds;
Yamamoto, H., Ed.; Springer-Verlag: Berlin, 2000; Chapter 7.
(19) Myers, K. E.; Kumar, K. J . Am. Chem. Soc. 2000, 122, 12025-
12026.
(17) A radical reaction in 2,2,2-trifluoroethanol is reported. See ref
7c,d. Radical polymerization in perfluorinated alcohols is also re-
ported: Isobe, Y.; Yamada, K.; Nakano, T.; Okamoto, Y. Macromol-
ecules 1999, 32, 5979-5981.
(20) Klement, I.; Knochel, P. Synlett 1995, 1113-1114.

7780 J . Org. Chem., Vol. 66, No. 23, 2001
Yorimitsu et al.
Sch em e 5
Ta ble 4. Tr ieth ylbor a n e-In d u ced Ra d ica l Ad d ition of
Br om otr ich lor om eth a n e to 2a in Va r iou s Solven tsa
bromine atom transfer and radical addition using the
Gaussian 98 program.21 All structures were optimized
with Becke’s three-parameter hybrid exchange functional
and the Lee-Yang-Parr correlation functional (B3LYP)22
using the 6-31G* basis set. Single-point calculations of
the total energies were performed at second-order Mφller-
Presset perturbation theory23 with the 6-31G** basis set
at B3LYP/6-31G*-optimized geometries. Zero point en-
ergy and thermal energy corrections were made at the
B3LYP/6-31G* level of theory, and the correction was
added to the energy obtained at the PMP2/6-31G**//
B3LYP/6-31G* level of theory. The enthalpic corrections
were made at 298.150 K. The transition states were
obtained with hand guesses. These transition states gave
single imaginary frequencies, and IRC calculations sup-
ported the transition structures. Orbital alteration around
the frontier orbitals did not affect the result of calcula-
tions. A solvent effect that comes from its permittivity
was estimated by polarized-continuum-model calcula-
tions using the polarizable conductor calculation model
based on the self-consistent reaction field theory (SCRF/
CPCM)24,25 at the B3LYP/6-31G* level. The total energy
yield of
dielectric
constantd
entry
solvent
hexane
methodb
7 (%)c
1
2
A
A
A
C
C
B
C
B
B
A
A
45
32
67
84
77
1.88
7.58
6.02
35.94
46.45
24.55
26.67
79.39
79.39
THF
3
4
5
6
7
8
9
10
11
ethyl acetate
acetonitrile
DMSO
ethanol
80
60
2,2,2-TFEe
water
99f
39g
100f
28g
water
no solvent
no solvent
a
Bromotrichloromethane (3.3 mmol), 2a (3.0 mmol), triethylbo-
rane (0.20 mmol), and solvent (5 mL) were employed unless
otherwise noted. Air (2 mL × 2) was added every 30 min. Reaction
b
time was 2 h. Method A: a hexane solution of triethylborane (1.0
M) was added. Method B: an ethanol solution of triethylborane
(1.0 M) was used. Method C: neat triethylborane was added.
c Yields were determined by 1H NMR measurement with dibenzyl
d
ether as an internal standard. Reproduced from ref 16. e 2,2,2-
Trifluoroethanol. f Triethylborane (0.10 mmol) was used. g Trieth-
ylborane (0.03 mmol) was used.
(21) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,
J r.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J . L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J . A. Gaussian 98, revision A.6; Gaussian, Inc.: Pittsburgh, PA,
1998.
lowers the activation energy of the radical reaction (vide
infra). The other difference is the quantitative formation
of 7 in the reaction without solvent and in water. Under
the highly concentrated conditions, the addition of bromo-
trichloromethane proceeded efficiently enough to be com-
pleted. We suppose that the reaction in water would
proceed in the organic phase consisting of bromotrichloro-
methane and 1-octene. Interestingly, comparing entry 9
with entry 11 reveals that the reaction in water was more
efficient than the reaction with no solvent.
(22) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098-3100. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785-789.
(23) Mφller, C.; Plesset, M. S. Phys. Rev. 1934, 46, 618-622.
(24) Barone, V.; Cossi, M. J . Phys. Chem. A 1998, 102, 1995-2001.
(25) Density functional calculations of the addition of a methyl
radical including solvent effects by the polarizable-continuum model
are reported: Arnaud, R.; Bugaud, N.; Vetere, V.; Barone, V. J . Am.
Chem. Soc. 1998, 120, 5733-5740.
Ab In itio Ca lcu la tion s a n d Discu ssion
To explain these results, we carried out ab initio
calculations on the chain-propagation steps consisting of

Solvent Effect on Br Atom-Transfer Radical Addition
J . Org. Chem., Vol. 66, No. 23, 2001 7781
Ta ble 5. Ca lcu la ted En er gies for th e Sim p lified Mod el in Sch em e 5
TS-1 + E )
TS-2 + Da
TS-2′ + Db
∆Gq for step 2a
∆Gq for step 2′b
F + Da
F ′ + Db
A + B + E
∆Gq for step 1
C + D + E
Absolute Energy (au)c
-2854.948663
-2854.922758
-2854.957493
-2854.934276a
-2854.976258a
-2854.977179b
-2854.934426b
Relative Energy (kcal/mol)d
under vacuum
in cyclohexane
in benzene
in acetone
0.00
16.26
-5.54
9.03a
14.57a
14.47b
15.04a
15.06b
14.69a
14.67b
14.89a
14.81b
14.83a
14.78b
14.79a
14.68b
-17.32a
8.93b
-17.89b
0.00 (-1.15)
15.96 (-1.45)
15.60 (1.15)
15.71 (-1.39)
15.49 (0.35)
15.43 (-0.86)
-6.75 (-2.36)
-6.81 (0.54)
8.29 (-1.89)a
8.31 (-1.77)b
7.88 (0.66)a
7.86 (0.74)b
8.02 (-1.85)a
7.94 (-1.83)b
7.69 (-0.22)a
7.64 (-0.17)b
7.84 (-1.22)a
7.73 (-1.23)b
-19.57 (-3.40)
-20.17 (-3.43)b
-19.66 (-0.53)a
-20.25 (0.55)b
-19.64 (-3.16)a
-20.18 (-3.13)b
-20.47 (-2.03)a
-21.05 (-2.04)b
-20.61 (-3.32)a
-21.22 (-3.36)b
0.00 (1.81)
0.00 (-0.84)
0.00 (1.12)
0.00 (-0.03)
-6.87 (-2.17)
-7.14 (-0.48)
-6.95 (-1.44)
in DMSO
in water
Data for step 2. Data for step 2′. c Absolute energies are those under vacuum. Solvation energies are those in parentheses.
a
b
d
F igu r e 2. Transition structures of the radical addition step
F igu r e 1. Transition structure of the bromine atom-transfer
(steps 2 and 2′).
step (step 1).
computed activation energies for step 2 and step 2′ are
value in a solution was obtained as the sum of the energy
14.57 and 14.47 kcal/mol, respectively. The former, the
at the PMP2/6-31G**//B3LYP/6-31G* level, the solvation
bromine atom-transfer step, is suggested to be rate-
energy that was obtained at the B3LYP/6-31G* level, and
determining, which is consistent with the experiment.
thermodynamic corrections that were calculated at the
We then considered the effect of solvation on the activa-
B3LYP/6-31G* level. No structures were reoptimized
tion energy. As a whole, the CPCM method could account
with the continuum solvent model, and the solvent-phase
for the results in Table 3. The energy barriers to bromine
energies were estimated via single-point calculations of
atom transfer are calculated to be 15.96, 15.60, 15.71,
the gas-phase structure.
15.49, and 15.43 kcal/mol in cyclohexane, benzene,
acetone, DMSO, and water, respectively. Polar solvents
tend to lower the barrier. This tendency is also observed
in the addition step via TS-2 or TS-2′. The calculations
in benzene show exceptionally lower energies of the three
transition states than those anticipated, judging from its
permittivity. The reason for the unexpected results of
To make ab initio MO calculations feasible, we set up
simplified reaction models as shown in Scheme 5. The
transition structures of bromine atom transfer (TS-1 in
step 1) and those of radical addition, which consists of
two possible modes with regard to the orientation of the
approaching radical to ethylene (TS-2 and TS-2′ in step
2 and step 2′, respectively),26 are shown in Figures 1 and
calculations in benzene is not clear. The energy differ-
2. The energy of each structure in a vacuum, cyclohexane,
ences shown in Table 5 are not sufficient to explain the
acetone, methanol, DMSO, and water is given in Table
solvent effect. The difference, for example, between the
5. At the PMP2/6-31G**//B3LYP/6-31G* level of theory,
relative energy of TS-1 in cyclohexane and that in water
the energy difference between reactants A + B and the
transition state TS-1 in the bromine atom-transfer step
under vacuum is 16.26 kcal/mol. On the other hand, the
is calculated to be 0.53 kcal/mol. The energy difference
(∆Gq
- ∆Gqcyclohexane) corresponds to kwater/kcyclohexane
)
water
2.44, according to the equation k ) (kBT/h) exp(-∆Gq/
RT), where k, kB, h, R, T, and ∆Gq are the rate constant,
Boltzmann constant, Planck constant, gas constant,
(26) Zipse, H.; He, J .; Houk, K. N.; Giese, B. J . Am. Chem. Soc. 1991,
113, 4324-4325.

7782 J . Org. Chem., Vol. 66, No. 23, 2001
Ta ble 6. Ca lcu la ted En er gies w ith a n Exp licit Wa ter
Yorimitsu et al.
TS-1(W) + E )
TS-2(W) + Da
TS-2(W)′ + Db
∆Gq for step 2(W)a
∆Gq for step 2′(W)b
F (W) + Da
F ′(W) + Db
A(W) + B + E
-3031.160997
∆Gq for step 1(W)
-3031.136158
C(W) + D + E
Absolute Energy (au)c
-3031.170732
-3031.148611a
-3031.149201b
Relative Energy (kcal/mol)d
-3031.190085a
-3031.189851b
under vacuum
in water
0.00
15.59
-6.11
7.77a
13.88a
13.51b
14.58a
14.17b
-18.26a
7.40b
-18.11b
0.00 (-0.67)
14.73 (-1.53)
-7.80 (-2.36)
6.78 (-1.66)a
6.37 (-1.70)b
-21.76 (-4.17)a
-21.47 (-4.03)b
Data for step 2. Data for step 2′. c Absolute energies are those under vacuum. Solvation energies are those in parentheses.
a
b
d
F igu r e 3. Transition structure of step 1 with an explicit
water.
temperature (298.15 K here), and activation energy,
respectively. However, qualitative discussion would be
acceptable.
There remains another problem: the remarkable sol-
vent effect of protic solvents, especially fluorinated alco-
hols. Protic solvents usually have large values of a
dielectric constant due to their hydroxy groups. However,
F igu r e 4. Transition structures of steps 2 and 2′ with an
it is difficult to explain the extraordinary reactivity in
fluorinated alcohols. Fluorous alcohols are highly acidic.
Accordingly, a carbonyl group in reactants would coor-
dinate to proton in a fluorous solvent more strongly than
in usual alcohols. We then considered the importance of
the coordination of a carbonyl oxygen to hydrogen in a
solvent molecule in the present radical reaction. In fact,
Lewis acid-promoted radical reactions were reported,6,7
and it is therefore of significance to conduct calculations
including coordination.
For the case where one explicit water molecule is
arranged to be coordinated by a carbonyl group, the
structures were optimized again in a similar way.27 The
results were summarized in Figures 3 and 4 and Table
6. Comparison of the activation energies for step 1 (16.26
kcal/mol), step 2 (14.57 kcal/mol), and step 2′ (14.47 kcal/
mol) with those for step 1(W) (15.59 kcal/mol), step 2(W)
(13.88 kcal/mol), and step 2′(W) (13.51 kcal/mol), respec-
tively, showed that the coordination lowered the activa-
tion energy in each case.
explicit water.
Ta ble 7. En er gies of th e HOMO a n d LUMO of A a n d
th e SOMO of C
HOMO
HOMO - 1
LUMO
A (au)
A(W) (au)
∆E[A(W)-A] (eV)
-0.40919
-0.40629
0.0790
-0.40953
-0.40858
0.0260
0.11173
0.10680
-0.134
SOMO
C (au)
C(W) (au)
∆E[C(W)-C] (eV)
-0.41157
-0.41444
-0.0781
Ta ble 8. Mu llik en Ch a r ges on th e
(Meth oxyca r bon yl)m eth yl Moiety
charge
A
TS-1
C
TS-2
TS-2′
+0.0996
+0.0087
0
-0.0761
-0.0777
The HOMO, HOMO - 1, and LUMO of A and the
SOMO of a (methoxycarbonyl)methyl radical C are shown
in Figure 5. The water-coordinated form also has similar
orbitals. The energies of these orbitals are listed in Table
7. The HOMO and the HOMO - 1 of A exist as the lone
pairs of bromine, and these are almost orthogonal to the
SOMO of an approaching methyl radical. Thus, the
HOMO of A and the SOMO of B do not overlap with each
other. On the other hand, the LUMO of A, the σ* orbital
of the C-Br bond, can interact closely with the SOMO
of B. The lower the energy of LUMO becomes, the faster
the atom transfer proceeds. This is indeed the case, and
the LUMO of A(W) is calculated to be lower than that of
A by 0.134 eV. Additionally, we focused on the electron
populations of A and TS-1 (Table 8). The Mulliken
charges on the (methoxycarbonyl)methyl moiety of A and
TS-1 are +0.0996 and +0.0087, respectively. Thus, the
electron population of the (methoxycarbonyl)methyl moi-
ety becomes larger as the reaction proceeds to the
transition state. Furthermore, the length of the hydrogen
bond of TS-1(W) optimized at the B3LYP/6-31G* level
(27) The solvent effect of water in the Diels-Alder reaction was
studied by ab initio computation including calculations on acrolein with
explicit water: Evanseck, J . D.; Kong, S. J . Am. Chem. Soc. 2000, 122,
10418-10427 and references therein.

Solvent Effect on Br Atom-Transfer Radical Addition
J . Org. Chem., Vol. 66, No. 23, 2001 7783
F igu r e 5. Molecular orbitals of A and C. (a) HOMO of A, (b) HOMO - 1 of A, (c) LUMO of A, and (d) SOMO of C.
Ta ble 9. Sta biliza tion En er gies Gen er a ted by th e Coor d in a tion
energy/aua
methyl bromoacetate of A(W)
water of A(W)
sum of the above two energies E[A(W)*]
E[A(W)]
-2837.023988
-76.219529
-2913.243518
-2913.256630
∆Ecoord[A(W)] ) E[A(W)] - E[A(W)*]
-0.01311 ) -8.23 kcal/mol
TS-1(W) without water
water of TS-1(W)
sum of the above two energies E[TS-1(W)*]
E[TS-1(W)]
∆Ecoord[TS-1(W)] ) E[TS-1(W)] - E[TS-1(W)*]
-2876.708310
-76.219504
-2952.927814
-2952.941483
-0.01367 ) -8.58 kcal/mol
(methoxycarbonyl)methyl moiety of C(W)
water of C(W)
sum of the above two energies E[C(W)*]
E[C(W)]
-266.953391b
-76.220755b
-343.174146b
-343.194289
∆Ecoord[C(W)] ) E[C(W)] - E[C(W)*]
-0.02014 ) -12.64 kcal/molb
TS-2(W) without water
water of TS-2(W)
sum of the above two energies E[TS-2(W)*]
E[TS-2(W)]
∆Ecoord [TS-2(W)] ) E[TS-2(W)] - E[TS-2(W)*]
-345.263045b
-76.221187b
-421.484232b
-421.508705
-0.024473 ) -15.36 kcal/molb
(methoxycarbonyl)methyl moiety of TS-2′(W)
water of TS-2′(W)
sum of the above two energies E[TS-2′(W)*]
E[TS-2′(W)]
-345.263324b
-76.221127b
-421.484451b
-421.509167
∆Ecoord [TS-2′(W)] ) E[TS-2′(W)] - E[TS-2′(W)*]
-0.024716 ) -15.51 kcal/molb
a
b
None of the energies include thermal energy corrections. Basis set superposition errors are considered.
is shorter than that of A(W) by 0.024 Å. Some of the
increasing electron population is probably delocalized
around the water molecule, and TS-1 would be stabilized.
To clarify the assumption, the stabilization energies
generated by the coordination to a water molecule were
estimated at A(W) and TS-1(W) (Table 9) as follows. We
first obtained the sum of the energy given E[A(W)*] by
the single-point calculations at the PMP2/6-31G** level
on the structure obtained by lifting water from A(W) and
the energy of water extracted from A(W). Next, we

7784 J . Org. Chem., Vol. 66, No. 23, 2001
Yorimitsu et al.
Ta ble 10. Resu lts of Ra d ica l Ad d ition in a
P r otic Solven t
yield
yield
entry
1
2
solventa (%) entry
1
2
solventa (%)
1
2
3
4
5
6
7
8
9
1a 2a
1a 2a
1a 2d
1a 2d
1a 2e
1a 2e
1a 2f
1a 2f
1a 2g
A
B
A
B
A
B
A
B
A
64
73
91
64
18
41
51
53
65
10
11
12
13
14
15
16
17
18
1a 2g
1a 2h
1a 2h
1a 2a
1a 2a
1a 2a
1a 2a
1a 2a
1d 2a
B
A
B
72
72
70
80
50
81
94
76
79
Bb
Bc
Cd,e
Cb,d
Cc,d
B
a
Otherwise, conditions are the same as those in Table 1. Solvent
A: a mixed solvent of ethanol and water (3.5/1.5). Solvent B: 2,2,2-
b
trifluoroethanol. Solvent C: water. A 3 mL volume of a solvent.
F igu r e 6. Optimized structures of A(W) and C(W).
d
c A 1 mL volume of a solvent. Triethylborane was added twice.
e A 10 mL volume of a solvent.
prescribed the difference between the total energy
E[A(W)*] and the energy of A(W) E[A(W)] as the
bonyl)methyl moiety that comes from the electron trans-
fer from the olefin would be stabilized by the hydrogen
bond. The water molecule stabilizes the transition states
of the addition step by having a stronger hydrogen bond
than it does the initial state. Consequently, water
enhances electron transfer from ethylene to the radical.
The coordination to water has proved to lower the
activation barriers in both steps. Furthermore, we ap-
plied the CPCM option to the structures with water
coordination. The relative energy of the transition struc-
ture of the bromine atom transfer in water is calculated
to be 14.73 kcal/mol (Table 6), lower than that in
cyclohexane without explicit water (15.96 kcal/mol, Table
5) by 1.23 kcal/mol. The energy difference means that
the reaction in water with coordination proceeds 7.9
times faster than that in cyclohexane without coordina-
tion. In the addition step, the computed activation
energies for step 2(W) and step 2′(W) in water are 14.58
and 14.17 kcal/mol, respectively. Compared with the
result in cyclohexane shown in Table 5, the activation
energies decrease by 0.46 kcal/mol in step 2(W) and 0.89
kcal/mol in step 2′(W). Hence, not only a polar effect but
also the coordination of a carbonyl group to a protic
solvent plays a significant role in the enhancement of
intermolecular bromine atom-transfer addition reactions.
The result of the calculations on the explicit water
model would be regarded as the model of a Lewis acid-
promoted radical reaction. The coordination of a Lewis
acid would accelerate propagation in a similar fashion.
Consequently, a catalytic amount of Lewis acid7a-7g as
well as a stoichiometric Lewis acid could control radical
reactions.
stabilization energy provided by the coordination (∆Ecoord
)
at A(W). Similar calculations were done on TS-1(W). The
coordination of the transition states to water affords the
stabilization energies at TS-1(W) as 8.58 kcal/mol. In
contrast, ∆Ecoord at A(W) is calculated to be 8.23 kcal/
mol. Thus, water binds more firmly at TS-1(W) than at
A(W), and the coordination would promote electron
transfer to methyl bromoacetate.
Next, the SOMO of C is discussed to examine the
addition step. It is known that radicals with electron-
withdrawing substituents at the radical center have low-
level SOMOs, and the interaction between the SOMO of
the radical and the HOMO of the olefin therefore
dominates, with an electron in the HOMO being trans-
ferred to the SOMO.28 In fact, the Mulliken charges of
the (methoxycarbonyl)methyl moieties of TS-2 and TS-
2′ are calculated to be -0.0761 and -0.0777, respectively
(Table 8), which indicates the transfer of an electron from
E to C during the uphill stage toward the transition
states. Lowering the level of the SOMO of C would
thereby give rise to easier electron transfer to the SOMO
and promotes steps 2 and 2′. Actually, the SOMO energy
is found to be lower by 0.0781 eV as a result of the
coordination to water as shown in Table 7. Furthermore,
as illustrated in Figures 4 and 6, the length of the
hydrogen bond between water and the carbonyl group
becomes shorter as the reaction progresses, which seems
to help the electron transfer. We calculated the stabiliza-
tion energy ∆Ecoord as performed in the bromine atom-
transfer step, applying corrections due to the basis set
superposition error.29 The coordinations of the transition
states to water afford values of 15.36 and 15.51 kcal/mol
for the stabilization energies at TS-2(W) and TS-2′(W),
respectively. In contrast, ∆Ecoord at C(W) is calculated to
be 12.64 kcal/mol. Thus, the hydrogen bond is stronger
at TS-2(W) and TS-2′(W) than at C(W). This fact
indicates that the negative charge on the (methoxycar-
Oth er Resu lts
Other successful results in a protic solvent are sum-
marized in Table 10. Similar to the reaction in water,
comparable results were given in aqueous ethanol and
in 2,2,2-trifluoroethanol. The addition of 1a to allyl
alcohol (2e) is more effective in 2,2,2-trifluoroethanol
(entry 6, Table 10) than in water (entry 5, Table 1) or in
aqueous ethanol (entry 5, Table 10). Carbonyl groups in
2f and 2h also survived under these conditions. The yield
of the adducts is dependent on the amount of solvent
employed (entries 13-17). The reactions in 3 and 1 mL
(28) (a) Fossey, J .; Lefort, D.; Sorba, J . Free Radicals in Organic
Synthesis; J ohn Wiley & Sons: Chichester, England, 1995; Chapter
6. (b) Giese, B. Radicals in Organic Synthesis; Pergamon Press:
Oxford, 1986; Chapter 2. (c) Fujimoto, H.; Yamabe, S.; Minato, T.;
Fukui, K. J . Am. Chem. Soc. 1972, 94, 9205-9210. (d) Fischer, H.;
Radom, L. Angew. Chem., Int. Ed. 2001, 40, 1340-1371. (e) Giese, B.
Angew. Chem., Int. Ed. Engl. 1983, 22, 753-764.
(29) Lendvay, G.; Mayer, I. Chem. Phys. Lett. 1998, 297, 365-373.

Solvent Effect on Br Atom-Transfer Radical Addition
J . Org. Chem., Vol. 66, No. 23, 2001 7785
hexane/ethyl acetate) provided ethyl 4-bromodecanoate (3a a ,
207 mg) in 74% yield. Alternatively, the following procedure
is useful. In a 20 mL reaction flask filled with argon, a mixture
of 1a (0.14 mL, 1.3 mmol) and 2a (0.16 mL, 1.0 mmol) was
treated with neat triethylborane (0.07 mL, 0.5 mmol) in water
(5.0 mL). Air (10 mL) was introduced into the reaction flask
with stirring. After 30 min, air was added again. The mixture
was stirred for an additional 30 min. Neat triethylborane was
added again, and the mixture was then exposed to air with
stirring for 2 h. Usual workup and silica gel column purifica-
tion afforded 198 mg of 3a a (71%).
Sch em e 6
of 2,2,2-trifluoroethanol provided 3a a in 80 and 50%
yields, respectively. The highest yield was achieved using
3 mL of water (entry 16).
The addition of 1b to allyltrimethylsilane shown in Scheme
3 provided a mixture of 3bn and 4 (94/6). The crude product
was dissolved in 3 mL of THF and treated with tetrabutylam-
monium fluoride (1.0 M THF solution, 2.0 mL, 2.0 mmol) at
25 °C. The temperature of the mixture rose somewhat. After
the mixture was stirred for 30 min, the reaction was quenched
with water, and the products were extracted twice with
hexane. Concentration of the organic layer followed by silica
gel column purification furnished 0.68 mmol of 4.
Interestingly, bromo oxazolidinone 1f underwent radi-
cal addition in 2,2,2-trifluoroethanol without a Lewis
acid.7 A mixture of 1-hexene (3 mmol) and 1f (1 mmol)
was treated with neat triethylborane (0.5 mmol) for 1.5
h. Evaporation of the solvent followed by silica gel column
purification provided 85 in 90% yield (Scheme 6). In
contrast, the reaction of 1f in water resulted in a poor
yield in the presence or absence of 10 mol % Yb(OTf)3.
Coordination of 1f to the solvent enhanced the reaction.
In conclusion, we have disclosed that the intermolecu-
lar bromine atom-transfer addition reaction of R-bromo
ester to an alkene, which resulted in poor conversion in
nonpolar solvents at ambient temperature, proceeded
efficiently in water. The addition reaction was examined
in various solvents at ambient temperature, and more
polar solvents as well as protic solvents, especially
perfluorinated alcohols, improved the efficiency of the
reaction. The origin of the solvent effect was revealed by
ab initio calculations. The polar nature of both the
reactants and transition states in the bromine atom
transfer and in the radical addition would promote the
reaction. Furthermore, the coordinations of the carbonyl
groups to protons in a protic solvent would also increase
the efficiency of the propagation. The choice of solvent is
worth consideration in radical reactions.
P r oced u r e for Br om in e Atom -Tr a n sfer Ad d ition in
Va r iou s Solven ts (Ta ble 3): Meth od A. The reaction in
hexane is representative. A 20 mL reaction flask was filled
with argon, and hexane (5 mL) was added to the reaction flask.
Compounds 1a (5.0 mmol) and 2a (1.0 mmol) were then added.
A 1.0 M hexane solution of triethylborane (0.50 mL, 0.50 mmol)
was introduced by a syringe, and air (5 mL) was added. After
30 min, 5 mL of air was injected again. The reaction mixture
was stirred for 1 h overall. After evaporation of hexane,
dibenzyl ether (0.3 mmol) was added to the residual oil as an
internal standard. 1H NMR experiment indicated a 10% yield
of 3a a . Meth od B. Compounds 1a and 2a were placed in a 20
mL flask, and ethanol (3.5 mL) and water (1.5 mL) were added
under argon. A solution of triethylborane in ethanol (1.0 M,
0.50 mL, 0.50 mmol) and air were sequentially added with
vigorous stirring. The reaction mixture was homogeneous.
After the mixture was treated with additional air as described
1
in Method A, extractive workup (hexane/brine) followed by H
NMR measurement yielded 3a a in an 81% yield. Meth od C.
Neat triethylborane (0.07 mL, 0.5 mmol) was added to a
solution of 1a and 2a in 2,2,2-trifluoroethanol under argon.
Air was then introduced twice as described above. Concentra-
tion in vacuo provided 3a a in a 73% yield.
Exp er im en ta l Section
Typ ica l P r oced u r e for Br om in e Atom -Tr a n sfer Ad d i-
tion of a n r-Br om o Ca r bon yl Com p ou n d to a n Alk en e.
The reaction of ethyl bromoacetate (1a ) with 1-octene (2a ) is
representative. Compounds 1a (0.55 mL, 5.0 mmol) and 2a
(0.16 mL, 1.0 mmol) and distilled water (5.0 mL) were placed
Ack n ow led gm en t. We thank Prof. Shigeki Kato
(Graduate School of Science, Kyoto University) for
helpful suggestions. This work was supported by Grants-
in-Aid for Scientific Research (12305058 and 10208208)
from the Ministry of Education, Culture, Sports, Science
and Technology, Government of J apan. H.Y. acknowl-
edges the J SPS Research Fellowship for Young Scien-
tists. H.S. also thanks Banyu Pharmaceutical Co., Ltd.
in
a 20 mL reaction flask under an argon atmosphere.
Triethylborane (1.0 M ethanol solution, 0.50 mL, 0.50 mmol)
was then added at 25 °C with vigorous stirring. Air (10 mL)
was immediately introduced into the reaction flask (not by
bubbling). CAUTION: Triethylborane may ignite spontane-
ously when exposed to air. The reaction mixture was treated
with additional air (10 mL × 2) every 30 min. The reaction
was quenched with 10 mL of hexane, and an organic layer was
dried over anhydrous sodium sulfate and concentrated in
vacuo. Silica gel column chromatography of the crude oil (20/1
Su p p or tin g In for m a tion Ava ila ble: Characterization
data and the Cartesian coordinates. This material is available
free of charge via the Internet at http://pubs.acs.org.
J O010652L