Powdered KOH in OMSO: An efficient base for asymmetric cyclization via memory of chirality at ambient temperature

Article abstract of DOI:10.1021/ja077684w

Enolate chemistry has been extensively used for stereoselective C-C bond formation, in which metal amide bases are frequently employed in strictly anhydrous solvents at low temperatures. However, we found that asymmetric intramolecular C-C bond formation

Full text of DOI:10.1021/ja077684w

Published on Web 02/28/2008
Powdered KOH in DMSO: An Efficient Base for Asymmetric
Cyclization via Memory of Chirality at Ambient Temperature
Takeo Kawabata,* Katsuhiko Moriyama, Shimpei Kawakami, and Kazunori Tsubaki
Institute for Chemical Research, Kyoto UniVersity Uji, Kyoto 611- 0011, Japan
Received October 5, 2007; E-mail: kawabata@scl.kyoto-u.ac.jp
Abstract: Enolate chemistry has been extensively used for stereoselective C-C bond formation, in which
metal amide bases are frequently employed in strictly anhydrous solvents at low temperatures. However,
we found that asymmetric intramolecular C-C bond formation via axially chiral enolate intermediates
proceeded in up to 99% ee at 20 °C using powdered KOH in dry or wet DMSO as a base. The
enantioselectivity was even higher than that of the corresponding reactions with potassium hexamethyl-
disilazide in DMF at -60 °C. The racemization barrier of the axially chiral enolate intermediate was estimated
to be ∼15.5 kcal/mol. On the basis of the barrier, the chiral enolate intermediate was supposed to undergo
cyclization within ∼10-3 sec at 20 °C after it is generated to give the product in g99% ee. Thus, enolates
generated with powdered KOH in DMSO were expected to be extremely reactive.
Scheme 1. Asymmetric Cyclization at Ambient Temperature via
Axially Chiral Enolate Intermediate A
Introduction
Enantioselective construction of a chiral tetrasubstituted
stereocenter is one of the most challenging tasks in current
synthetic organic chemistry.1 We have developed a direct
method for asymmetric alkylation of R-amino acid derivatives
without the aid of external chiral sources such as chiral
auxiliaries or chiral catalysts, that is, memory of chirality.2-4
Inter- and intramolecular alkylation of R-amino acid derivatives
proceeded in up to 98% ee via axially chiral enolate intermedi-
ates, where enolate formation was performed typically at -78
to -60 °C to maintain enantiomeric purity of the chiral
enolates.4,5 However, we found that asymmetric cyclization via
axially chiral enolate intermediate A proceeded in up to 99%
ee at 20 °C using powdered KOH in DMSO as a base (Scheme
1). Surprisingly, some of the asymmetric cyclization reactions
with KOH/DMSO at 20 °C proceeded with greater enantiose-
lectivity than those with KHMDS in DMF at -60 °C. Another
intriguing feature is that four-membered cyclization proceeded
faster than the corresponding six-membered cyclization.
Results and Discussion
We have developed a method for the straightforward synthesis
of cyclic amino acids with a tetrasubstituted stereocenter from
readily available R-amino acids via memory of chirality.3i,4c,e
Treatment of N-ω-bromoalkyl-N-tert-butoxycarbonyl(Boc)-R-
amino acid derivatives with KHMDS in DMF at -60 °C gave
cyclic amino acid derivatives in up to 98% ee with retention of
configuration.4c The asymmetric cyclization was thought to
proceed through an axially chiral enolate intermediate A (X )
Br). We had believed that asymmetric reactions via memory of
chirality would not take place highly enantioselectively at
ambient temperature because the axially chiral enolate inter-
mediates suffer from temperature-dependent racemization. For
example, R-methylation of N-Boc-N-methoxymethyl(MOM)-
R-amino acid derivatives proceeded in up to 93% ee at -78 °C
via axially chiral enolate intermediate B, which has a half-life
(1) For reviews, see (a) Corey, E. J.; Guzman-Perez, A. Angew. Chem., Int.
Ed. 1998, 37, 388-401. (b) Christoffers, J.; Mann, A. Angew. Chem., Int.
Ed. 2001, 40, 4591-4597.
(2) For reviews on memory of chirality, see (a) Kawabata, T.; Fuji, K. Top.
Stereochem. 2003, 23, 175-205. (b) Zhao, H.; Hsu, D.; Carlier, P. R.
Synthesis 2005, 1-16.
(3) For reactions related to memory of chirality, see (a) Beagley, B.; Betts, M.
J.; Pritchard, R. G.; Schofield, A.; Stoodley, R. J.; Vohra, S. J. Chem. Soc.,
Chem. Comm. 1991, 924-925. (b) Kawabata, T.; Yahiro, K.; Fuji, K. J.
Am. Chem. Soc. 1991, 113, 9694-9696. (c) Betts, M. J.; Pritchard, R. G.;
Schofield, A.; Stoodley, R. J.; Vohra, S. J. Chem. Soc., Perkin Trans. 1
1999, 1067-1072. (d) Gerona-Navarro, G.; Bonache, M. A.; Herranz, R.;
Garc´ıa-Lo´pez, M. T.; Gonza´lez-Mun˜iz, R. J. Org. Chem. 2001, 66, 3538-
3547. (e) Bonache, M. A.; Gerona-Navarro, G.; Mart´ın-Mart´ınez, M.;
Garc´ıa-Lo´pez, M. T.; Lo´pez, P.; Cativiela, C.; Gonza´lez-Mun˜iz, R. Synlett
2003, 1007-1011. (f) Carlier, P. R.; Zhao, H.; DeGuzman, J.; Lam, Polo,
C.-H. J. Am. Chem. Soc. 2003, 125, 11482-11483. (g) Kawabata, T.;
Ozturk, O.; Suzuki, H.; Fuji, K. Synthesis 2003, 505-508. (h) Kawabata,
T.; Ozturk, O.; Chen, J.; Fuji, K. Chem. Commun. 2003, 162-163. (i)
Kawabata, T.; Matsuda, S.; Kawakami, S.; Monguchi, D.; Moriyama, K.
J. Am. Chem. Soc. 2006, 128, 15394-15395.
(4) (a) Kawabata, T.; Suzuki, H.; Nagae, Y.; Fuji, K. Angew. Chem., Int. Ed.
2000, 39, 2155-2157. (b) Kawabata, T.; Chen, J.; Suzuki, H.; Nagae, Y.;
Kinoshita, T.; Chancharunee, S.; Fuji, K. Org. Lett. 2000, 2, 3883-3885.
(c) Kawabata, T.; Kawakami, S.; Majumdar, S. J. Am. Chem. Soc. 2003,
125, 13012-13013. (d) Kawabata, T.; Kawakami, S.; Shimada, S.; Fuji,
K. Tetrahedron 2003, 59, 965-974. (e) Kawabata, T.; Majumdar, S.;
Tsubaki, K.; Monguchi, D. Org. Biomol. Chem. 2005, 3, 1609-1611. (f)
Kawabata, T.; Chen, J.; Suzuki, H.; Fuji, K. Synthesis 2005, 5, 1368-
1377.
(5) For reviews of stereoselective C-C bond formation based on enolate
chemistry, see (a) Morrison, J. D. Ed. Asymmetric Synthesis; Academic
Press, Inc.: New York, 1984; Vol.3, pp 1-341. (b) Koga, K. Yakugaku
Zasshi, 1997, 117, 800-816. (c) Krause, N.; Ebert, S.; Haubrich, A. Liebigs
Ann. Chem. 1997, 2409-2418. (d) Fringuelli, F.; Piermatti, O.; Pizzo, F.
Recent Res. Dev. Org. Chem. 1997, 1, 123-136. (e) Wirth T. Organic
Synthesis Highlights IV; Wiley-VCH: New York, 2000; pp 26-33.
9
10.1021/ja077684w CCC: $40.75 © 2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 4153-4157
4153

A R T I C L E S
Kawabata et al.
Scheme 2. R-Methylation via Axially Chiral Enolate Intermediate
When 1 was treated with potassium tert-butoxide or potassium
hydride in DMF at 0∼20 °C, 2 was obtained in 87 or 89% ee,
respectively (entries 3 and 4). To our surprise, the treatment of
1 with powdered KOH (prepared by grinding commercial 85%
KOH pellets with a mortar in a glove box, see Supporting
Information) in DMF at 20 °C gave 2 in 98% ee (entry 5). We
further found that powdered KOH in DMSO (KOH/DMSO) was
an excellent base for the transformation and gave 2 in 99% ee
and 91% yield (entry 6). KOH/DMSO was a suspension even
after vigorous stirring at 20 °C for 5 min at the concentration
(0.3 M) employed for the reaction. No difference in the
efficiency of asymmetric cyclization was observed between
powdered KOH prepared from commercial 85% KOH pellets
and that from commercial >99% KOH pellets (entries 6 vs 7).
Powdered LiOH was not effective as a base, whereas powdered
NaOH and powdered CsOH showed reactivity similar to
powdered KOH (entries 8-10). The solvent effects in KOH-
promoted asymmetric cyclization were investigated (entries 11-
16). Whereas the existence of 1% water in DMSO did not affect
the efficiency of the asymmetric transformation (entries 6 vs
11), an increase in the amount of water in DMSO decreased
both yield and enantioselectivity of the cyclization (entries 12
and 13). Powdered KOH in CH2Cl2 was totally unreactive as a
base, and powdered KOH in THF was much less effective than
KOH/DMSO for the transformation of 1 into 2 (entries 14 and
15). Powdered KOH in EtOH (KOH/EtOH) did not promote
cyclization but rather promoted hydrolysis to give carboxylic
acid 3 as a major product in 55% yield (entry 16). The difference
in reactivity between KOH/DMSO and KOH/EtOH could be
ascribed to the difference in the pKa’s of H2O in these solvents.
The pKa’s of H2O in H2O and in DMSO are known to be 16
and 31, respectively.7 This means that KOH in DMSO is a
strong base, which can abstract the R proton of esters (pKa
18∼307,8), whereas KOH in protic solvents is not.
B
Table 1. Effects of Bases, Temperature, and Solvents on
Asymmetric Cyclization of 1a
entry base (mol equiv)
solventb
temp, time(h) 2, yieldc (%) 2, eed (%)
1e KHMDS(1.2) DMF
-60 °C, 0.5 94
98
93
87
89
98
99
99
2
3
4
5
6
7
8
9
KHMDS(1.2) DMF
t-BuOK(1.5) DMF
0 °C, 0.2
0∼20 °C, 1
0 °C, 0.2
20 °C, 2
20 °C, 2
20 °C, 2
20 °C, 17
20 °C, 2
20 °C, 2
97
67
76
89
91
KH(2.0)
DMF
DMF
KOHf,g(3.0)
KOHf,g(3.0)
KOHf,h(3.0)
LiOHf(3.0)
NaOHf(3.0)
DMSO
DMSO
DMSO
DMSO
DMSO
90
∼0 (75)
81
97
99
99
97
93
10 CsOHf(3.0)
11 KOHf,g(3.0)
12 KOHf,g(3.0)
13 KOHf,g(3.0)
14 KOHf,g(3.0)
15 KOHf,g(3.0)
16 KOHf,g(3.0)
64
1% H2O in DMSO 20 °C, 2
10% H2O in DMSO 20 °C, 1
20% H2O in DMSO 20 °C, 1
98
76
12
CH2Cl2
THF
EtOH
20 °C, 2
20 °C, 2
20 °C, 23
∼0 (97)
45 (45)
∼0i (8)
71
a Reactions in entries 1-2 and reactions in 3-15 were carried out with
a substrate concentration of 0.08 and 0.1 M, respectively. b DMF, distilled
from P2O5 at reduced pressure; DMSO, anhydrous DMSO (H2O < 0.005%)
from Aldrich, CH2Cl2, dehydrated CH2Cl2 (H2O < 0.002%) from Kanto
Kagaku; THF, distilled from sodium ketyl of benzophenone; EtOH, 99.5%
ethanol from nakalai tesque. c Numbers in the parentheses indicate the
percent recovery of 1. d ee of the corresponding N-benzoate determined by
HPLC analysis. The (S)-isomer was obtained in each run. For the
determination of the absolute configuration, see ref 4c. e Data quoted from
ref 4c. f Powdered metal hydroxide was used. g Prepared from commercial
85% KOH pellets from Nakalai Tesque. h Prepared from commercial >99%
KOH pellets from Kojundo Chemical Laboratory, Co., Ltd. i Carboxylic
acid 3 was obtained in 55% yield.
These newly developed mild reaction conditions (KOH/
DMSO/20 °C) were applied to asymmetric four-, five-, and six-
membered cyclization of phenylalanine, valine, and methionine
derivatives. The results are summarized in Table 2. Asymmetric
cyclization of phenylalanine derivative 4 with KOH/DMSO
proceeded within 2 h at 20 °C to give azetidine 5 with a
tetrasubstituted carbon center in 99% ee and 82% yield with
retention of configuration (entry 1). Similarly, four-membered
cyclization of valine and methionine derivatives 6 and 8 gave
azetidines 7 and 9, respectively, in 99% ee (entries 2 and 3).
Five-membered cyclization of 1, 10, and 12 with KOH/DMSO
proceeded smoothly to give pyrrolidines 2, 11, and 13,
respectively, in 98∼99% ee and 91∼94% yield (entries 4-6).
We previously reported that asymmetric four-membered cy-
clization of 4 and 8 and five-membered cyclization of 1, 10,
and 12 took place in 94∼98% ee upon treatment with KHMDS
in DMF at -60 °C (brackets in entries 1 and 3-6).3i,4c
Surprisingly, asymmetric four- and five-membered cyclizations
with KOH/DMSO at 20 °C proceeded with higher enantiose-
lectivity than those with KHMDS in DMF at -60 °C. This could
be ascribed to the high reactivity of the axially chiral enolate
intermediate A generated with KOH/DMSO, which immediately
undergoes cyclization before suffering from noticeable racem-
of racemization of 22 h at -78 °C,4a whereas that at 20 °C was
calculated to be less than 0.1 s from the racemization barrier
(Scheme 2).6 Accordingly, it did not seem feasible to develop
highly enantioselective intermolecular reactions via the memory
of chirality at ambient temperature. On the other hand, it seemed
possible to develop highly enantioselective intramolecular
reactions at ambient temperature if the chiral enolate intermedi-
ates react very rapidly within the time-scale of their racemiza-
tion.
We chose the transformation of 1 into 2 as an intramolecular
reaction via memory of chirality and examined the temperature
dependence of its enantioselectivity. Treatment of 1 with
KHMDS at -60 °C in DMF gave 2 in 98% ee,4c whereas, at 0
°C, 2 was formed in only slightly diminished enantioselectivity
of 93% ee (Table 1, entries 1 vs 2). This finding prompted us
to further investigate bases and solvents for asymmetric reactions
near ambient temperature. Bases that contained a potassium
cation were screened for the reactions in DMF (entries 3-5).
(7) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456-463.
(6) The racemization barrier at 20 °C was roughly estimated from that at -78
°C based on the assumption that ∆Sq of the unimolecular racemization
process (bond rotation along the chiral C-N axis) is nearly zero.
(8) Mashchenko, N. V.; Matveeva, A. G.; Odinets, I. L.; Matrosov, E. I.; Petrov,
E. S.; Terekhova, M. I.; Matveev, A. K.; Mastryukova, T. A.; Kabachnik,
M. I. Zh. Obshch. Khim. 1988, 58, 1973-1979.
9
4154 J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008

Asymmetric Cyclization via Memory of Chirality
A R T I C L E S
Table 2. Asymmetric Cyclization of R-Amino Acid Derivatives with
posed that axially chiral enolate A (X ) Br) is the key
intermediate for asymmetric cyclization with KHMDS in
DMF.4c Whereas a mechanism via intermediate A is also
expected for asymmetric cyclization with KOH/DMSO, an
alternative route may involve a concerted SEi process, as shown
in C. To investigate the validity of A and C, cyclization of 26
with two identical substituents on the nitrogen was examined
because the enolate generated from 26 cannot be axially chiral
along the C-N axis (Scheme 4). Treatment of 26 with KOH/
DMSO at 20 °C for 5 h gave racemic 27 in 27% yield together
with 44% recovery of 26. This indicates that asymmetric
cyclization with KOH/DMSO also proceeds via axially chiral
enolate intermediate A.
KOH/DMSO at Ambient Temperaturea
entry substrate
n
R
X
timeb (h) product yield(%)
eec (%)
1
2
3
4
5
6
7
8
9
4
6
8
2
2
2
3
3
3
4
4
4
4
4
4
PhCH2
Me2CH
MeSCH2CH2 Br
PhCH2
Me2CH
MeSCH2CH2 Br
PhCH2
Me2CH
MeSCH2CH2 Br
PhCH2
Me2CH
Br
Br
2
3
1
2
2
5
7
9
82
79
85
91
94
91
73
74
86
97
90
89
99 (R) [95d]
99
99 (S) [97e]
99 (S) [98d]
98 [94d]
98 (S) [97d]
90 [97d]
94
1
Br
Br
2
10
12
14
16
18
20
21
22
11
13
15
17
19
15
17
19
2
Br
Br
12
17
4
2
8
88
97
98
97
10
11
12
I
I
I
MeSCH2CH2
3
a A mixture of a substrate (0.25 mmol) and powdered KOH (prepared
from 85% commercial KOH pellets from Nacalai Tesque, 0.75 mmol) in
dry DMSO (anhydrous DMSO (H2O < 0.005%) from Aldrich, 2.5 mL)
was stirred vigorously at 20 °C for the time indicated in the table. b Time
required for the consumption of the starting material checked by TLC. c ee
of the corresponding N-benzoate determined by HPLC analysis. A letter in
the parenthesis indicates the absolute configuration. For the determination
of the absolute configuration, see refs 3i and 4c. Numbers in the brackets
indicate percent ee of the product obtained by the reaction with KHMDS
in DMF at -60 °C. d Data quoted from ref 4c. e Data quoted from reference
3i.
The racemization behavior of A was investigated next. We
previously determined the barrier to racemization of the axially
chiral enolate B by periodic quenching of the enolate with
methyl iodide (Scheme 2).4a However, this protocol cannot be
applied to enolate A because it undergoes cyclization im-
mediately after it is generated. 28, an analogue of 1, was used
to estimate the racemization barrier of the axially chiral enolate
A because the enolate C generated from 28 would not undergo
cyclization (Figure 1). The barrier to racemization of a potassium
enolate generated from 28 and KHMDS in DMF-THF (1:1)
at -78 °C was determined through the periodic quenching of
the enolate with methyl iodide.13 Experiments were performed
twice at the strictly controlled temperature. The decrease in ee
(ln ee0/eet) as a function of time for the base treatment of 28
was plotted (Figure 1, ee0 and each of eet are the average of
two runs). The barrier was calculated from the slope, 2k ) 1.99
× 10-3 min-1 (r ) 0.995), to be 15.5 kcal/mol at -78 °C. On
the basis of the assumption that asymmetric cyclization of 1
with KOH/DMSO proceeds via enolate A (X ) Br, n ) 3)
whose racemization barrier is comparable to C, enolate A is
expected to undergo cyclization at 20 °C within ∼10-3 sec after
it is generated to give product 2 in 99% ee (t99/100 at 20 °C )
3.0 × 10-4 sec, calculated from ∆Gq (15.5 kcal/mol)).6 Because
the asymmetric cyclization reactions in Table 2 are also assumed
to proceed via axially chiral enolates A, enolates generated with
KOH/DMSO in these reactions are expected to undergo rapid
cyclization within ∼10-2 sec at 20 °C to give products in g88%
ee (t88/100 at 20 °C ) 3.8 × 10-3 sec, calculated from ∆Gq (15.5
kcal/mol)).6 The extremely high reactivity of the enolates
generated with KOH/DMSO could be ascribed to their amine-
free structure.14,15
Scheme 3. Asymmetric Synthesis of Fmoc-Cyclic Amino Acid 25
ization (vide infra). In contrast to the four- and five-membered
cyclization, six-membered cyclization of 14 with KOH/DMSO
at 20 °C gave piperidine 15 with enantioselectivity lower than
that observed in the reaction of 14 with KHMDS in DMF at
-60 °C (entry 7). The enantioselectivity (88∼94% ee) observed
in the six-membered cyclization of bromides 14, 16, and 18
(entries 7-9) was improved by use of the corresponding iodides.
Treatment of 20, 21, and 22 with KOH/DMSO for 2-8 h at 20
°C gave piperidines 15, 17, and 19, respectively, in 97∼98%
ee and 89∼97% yield (entries 10-12). An increase in enanti-
oselectivity of cyclization of the iodides could be ascribed to
their inreased rates of cyclization. The similar effects of leaving
groups on asymmetric alkylation via memory of chirality have
been reported by Carlier and co-workers.9 This protocol was
applied to the medium-scale synthesis of Fmoc-cyclic amino
acid 25, which is expected to be a useful building block for
conformationally restricted peptides of biological interest (Scheme
3).10 Isoleucine derivative 23 (17 mmol) was treated with KOH/
DMSO11 at 20 °C for 2 h to give 24 as a single diastereomer in
94% yield, which was then converted into 25 in 53% yield.
We then investigated the mechanistic aspects of asymmetric
cyclization promoted by KOH/DMSO.12 We previously pro-
Another interesting observation in asymmetric cyclization
with KOH/DMSO is that the four-membered cyclization pro-
Scheme 4. Cyclization of 26 with Powdered KOH in DMSO via an Achiral Enolate Intermediate
9
J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008 4155

A R T I C L E S
Kawabata et al.
Figure 1. Plot of ln (ee0/eet) versus time (t) for the base treatment of 28. ee0 ) the ee value of 29 obtained by the reaction of the enolate immediately after
its generation from 28 and KHMDS with methyl iodide. eet ) the ee value of 29 obtained by the treatment of 28 with KHMDS for the time indicated
followed by addition of methyl iodide. Experiments were performed twice at the strictly controlled temperature. ee0 and each of eet are the average of two
runs. The barrier to racemization was determined to be 15.5 kcal/mol at -78 °C from the slope, 2k ) 1.99 × 10-3 min-1 (r ) 0.995). See the Supporting
Information for the experimental detail for the determination of the racemization barrier.
ceeds faster than the six-membered one. This tendency was
generally observed in the asymmetric cyclization of amino acid
derivatives prepared from phenylalanine, valine, and methionine
(Table 2, entries 1-3 vs 7-9). As a typical example, the
progress of the four-membered cyclization of 4 and six-
membered cyclization of 14 was monitored (Figure 2). The four-
membered cyclization of 4 was 2∼3 times faster than the six-
membered cyclization of 14. The rate-determining step for the
cyclization with KOH/DMSO must be the enolate-formation
step because the half-lives of racemization of the chiral enolate
intermediates generated from 4 and 14 are supposed to be much
(9) (a) MacQuarrie-Hunter, S.; Carlier, P. R. Org. Lett. 2005, 7, 5305-5308.
(b) Carlier, P. R.; Lam, P. C.-H.; DeGuzman, J. C.; Zhao, H. Tetrahedron:
Asymmetry 2005, 16, 2998-3002.
(10) For recent reviews on the importance and preparation of cyclic amino acids
with a tetrasubstituted stereocenter, see (a) Cativiela, C.; D´ıaz-de-Villegas,
M. D. Tetrahedron: Asymmetry 2000, 11, 645-732. (b) Park, K.-H.; Kurth,
M. J. Tetrahedron 2002, 58, 8629-8659. (c) Kotha, S. Acc. Chem. Res.
2003, 36, 342-351.
(11) Commercial (not dry) DMSO was used without further purification.
(12) One might suppose that the active base in these reactions is a dimsyl anion
generated in a small quantity from KOH and DMSO (pKa’s of H2O and
Figure 2. Plot of percent conversion of four-membered cyclization of 4
and six-membered cyclization of 14 versus time (min).
DMSO in DMSO are 31 and 35, respectively; ref 7). However, this seems
not feasible because the dimsyl anion prepared from KH in DMSO showed
a different reactivity in the reaction of 1. Treatment of 1 with 3.0 mol
equiv of potassium dimsylate in DMSO at 20 °C for 2 h gave 2 in 99% ee
and 68% yield together with the corresponding carboxylic acid of 2 in
25% yield.
shorter (<0.1 s)16 than the time required for the reactions to be
complete (2∼12 h). Thus, axially chiral enolates A (X ) Br, n
) 2 or 4) would form gradually, and once formed, would
immediately undergo asymmetric cyclization. Because enolates
A suffer from time-dependent racemization, ee’s of the products
would correlate with the rate of cyclization of enolate intermedi-
ates; that is, the faster the cyclization, the higher the enanti-
oselectivity. On the basis of this assumption, the four-membered
cyclization of enolates A (X ) Br, n ) 2) (99% ee, Table 2,
entries 1-3) is expected to proceed faster than the corresponding
six-membered cyclization of enolates A (X ) Br, n ) 4) (∼88-
(13) Whereas determination of the racemization barrier of the enolate generated
from 28 with KOH/DMSO is desirable, it was not possible. This is because
(1) KOH/DMSO cannot be used at low temperatures suitable for the
measurement of enolate racemization, and (2) KOH/DMSO cannot generate
the enolate from 28 in a quantitative manner. Quantitative generation of
the enolate in a much shorter period than the half-life of racemization of
the enolate is indispensable for the measurement of the racemization barrier.
(14) Amine-free lithium enolates have been known to be more reactive than
the enolates in the presence of a secondary amine generated in situ by
abstraction of the R proton of the carbonyl group with the lithium amide
base; see (a) Laube, T.; Dunitz, J. D.; Seebach, D. HelV. Chim. Acta. 1985,
68, 1373-1393. (b) Seebach, D.; Boes, M.; Naef, R.; Schweizer, W. B. J.
Am. Chem. Soc. 1983, 105, 5390-5398. (c) Aebi, J. D.; Seebach, D. HelV.
Chim. Acta. 1985, 68, 1507-1518.
(15) We suppose that a water molecule generated by deprotonation of the
substrate with KOH may be excluded from the coordination sphere of the
potassium enolate by strongly coordinating DMSO molecules.
(16) T50/100 was roughly estimated to be 0.02 sec at 20 °C based on the
racemization barrier of C (15.5 kcal/mol).
9
4156 J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008

Asymmetric Cyclization via Memory of Chirality
A R T I C L E S
94% ee, Table 2, entries 7-9). Accordingly, enantioselectivities
would provide a measure of the rate of cyclization of the enolate
intermediates,17 whereas the reaction time for asymmetric
cyclization depends on the rate of enolate formation.18 Higher
enantioselectivity observed in the cyclization of the iodides than
the corresponding bromide (Table 2, entries 7-12) is also
consistent with expected higher rates of cyclization of the
iodides.
In conclusion, we have developed a highly enantioselective
cyclization via memory of chirality at ambient temperature.
Powdered KOH in DMSO was an effective base for this
purpose. Although KOH in DMSO has been used in C-O,
C-N, and C-S bond formation,19 its use in stereoselective C-C
bond formation has been limited.20 We expect that powdered
KOH in DMSO would be further applicable in enolate chemistry
for fine organic synthesis because it generates highly reactive
enolates under mild conditions and by simple operations.
Acknowledgment. This work was supported by a grant-in-
aid for Scientific Research on Priority Areas “Advanced
Molecular Transformation of Carbon Resources” from the
Ministry of Education, Culture, Sports, Science and Technology,
Japan.
Supporting Information Available: An experimental proce-
dure in Table 2. Preparation of 6, 20, 25-27, and 29.
Chracterization of 3, 6, 7, and 16-29. Experimental procedure
for the determination of the racemization barrier of the enolate
generated from 28 and KHMDS in DMF-THF (1:1) at -78
°C. This material is available free of charge via the Internet at
http://pubs.acs.org.
JA077684W
(17) For an additional example, seven-membered cyclization of a substrate
derived from L-phenylalanine gave the product in 49% ee (25% yield) by
treatment with powdered KOH in dry DMSO at 20 °C for 2 h.
(18) The rates of enolate formation from the substrates undergoing four-
membered cyclization are assumed faster than those undergoing six-
membered cyclization. Reasons are totally unclear. A possible explanation
might be due to the slight increase in the acidity of R proton in the former
substrates caused by the inductive effect of the bromo group at the N-2-
bromoethyl substituent or due to the favorable delivery of OH- via chelation
of K+ with the bromo group.
(19) For examples, see: (a) Pang, Y.-P.; Hong, F.; Quiram, P.; Jelacic, T.;
Brimijoin, S. J. Chem. Soc., Perkin Trans 1 1997, 171-176. (b) Chapman,
D. R.; Reed, C. A. Tetrahedron Lett. 1988, 29, 3033-3036. (c) Goswami,
B. N.; Rastogi, R. C. Ind. J. Chem. B. 1992, 31B, 703-704. (d) Ensch, C.;
Hesse, M. HelV. Chim. Acta. 2002, 85, 1659-1673. (e) Andreeva, O. V.;
Militsina, O. I.; Kovylyaeva, G. I.; Korochkina, M. G.; Strobykina, I. Y.;
Bakaleinik, G. A.; Al’fonsov, V. A.; Kataev, V. E.; Musin, R. Z. Russ. J.
Gen. Chem. 2007, 77, 469-473.
(20) For an example, see, Dechoux, L.; Ebel, M.; Jung, L.; Stambach, J. F.
Tetrahedron Lett. 1993, 34, 7405-7408.
9
J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008 4157