Improved enantioselectivity by using novel bulk bases in chiral lithium amide catalysed deprotonations: Mixed dimers as reagents and catalysts

Article abstract of DOI:10.1016/S0040-4020(02)00373-3

Novel bulk bases have been developed yielding improved enantioselectivity of chiral lithium amide catalysed deprotonations as compared to using the bulk base lithium diisopropylamide (LDA). The new bulk bases are 2-lithio-1-methylimidazole, 2-(lithiomethyl)-1-methylimidazole, 2-lithio-furan and 1,8-diazabicyclo-6-lithio[5.4.0]undec-7-ene which have been used together with chiral lithium amides in deprotonations of cyclohexene oxide. Using the chiral lithium amides enhanced stereoselectivities (96% ee) have been reached. The reactivity change has been traced to the formation of novel reagents - mixed dimers - formed from a bulk base molecule and a molecule of a chiral lithium amide. The results also show that DBU, which has commonly been used as an additive to alter reactivity and enantioselectivity in deprotonations, has another important role. DBU is lithiated under the conditions used and becomes a bulk base, which forms catalytic mixed dimers with the chiral lithium amides.

Full text of DOI:10.1016/S0040-4020(02)00373-3

TETRAHEDRON
Pergamon
Tetrahedron 58 '2002) 4669±4673
Improved enantioselectivity by using novel bulk bases in chiral
lithium amide catalysed deprotonations: mixed dimers as reagents
and catalysts
Daniel Pettersen, Mohamed Amedjkouhp and Per Ahlbergp
È È
Department of Organic Chemistry, Goteborg University, SE-412 96Go teborg, Sweden
Received 4 January 2002; accepted 18 February 2002
AbstractÐNovel bulk bases have been developed yielding improved enantioselectivity of chiral lithium amide catalysed deprotonations as
compared to using the bulk base lithium diisopropylamide 'LDA). The new bulk bases are 2-lithio-1-methylimidazole, 2-'lithiomethyl)-1-
methylimidazole, 2-lithio-furan and 1,8-diazabicyclo-6-lithio[5.4.0]undec-7-ene which have been used together with chiral lithium amides
in deprotonations of cyclohexene oxide. Using the chiral lithium amides enhanced stereoselectivities '96% ee) have been reached. The
reactivity change has been traced to the formation of novel reagentsÐmixed dimersÐformed from a bulk base molecule and a molecule of a
chiral lithium amide. The results also show that DBU, which has commonly been used as an additive to alter reactivity and enantioselectivity
in deprotonations, has another important role. DBU is lithiated under the conditions used and becomes a bulk base, which forms catalytic
mixed dimers with the chiral lithium amides. q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
Chiral lithium amides are being developed for enantio-
selective deprotonation reactions of, for example, epoxides
and ketones.1 Such reactions are synthetically important
since the chiral products are useful intermediates in organic
synthesis.2 The chiral lithium amides are often used in
stoichiometric amounts or in excess. Alternatively, there
have also been a number of attempts to run the enantio-
selective deprotonation reactions under catalytic con-
ditions.3 The chiral lithium amide has been used in sub-
stoichiometric amounts 'typically 5±20 mol%) and the
less reactive lithium diisopropylamide 'LDA) has been
used as a bulk base, i.e. as a catalyst-regenerating base, to
obtain a catalytic cycle.4 However, non-enantioselective
deprotonation by LDA yielding racemic product competes
with that of the chiral base, resulting in lower enantio-
selectivity than in the stoichiometric reaction.
Thus, in order to improve the degree of enantioselectivity in
catalytic deprotonations access to bulk bases with lower
kinetic basicity than LDA, but comparable thermodynamic
basicity, are required. In our search for such bases we have
used 1-methylimidazole and 1,2-dimethylimidazole as
precursors.5 These azoles undergo carbon deprotonation at
the C2-position and the C2-methyl by n-BuLi to yield the
Keywords: allylic alcohols; asymmetric catalysis; epoxides; mixed dimers;
rearrangement.
p
Corresponding authors. Tel.: 146-31-772-2899; fax: 146-31-772-2908;
e-mail: per.ahlberg@oc.chalmers.se
Figure 1.
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020'02)00373-3

4670
D. Pettersen et al. / Tetrahedron 58 +2002) 4669±4673
In this paper we present the results obtained using the two
new bulk bases 2-lithio-furan 3 and N-lithio-imidazole
together with and and 1,8-diazabicyclo-6-
4
1
2
lithio[5.4.0]undec-7-ene 9 and the chiral lithium amide
catalysts lithium '1R,2S)-N-methyl-1-phenyl-2-pyrrolidi-
nylpropanamide 5 'independently developed by O'Brien
et al.8 and ourselves7), lithium 'S)-N-methyl-1-phenyl-2-
pyrrolidinoethylamide 6,2 lithium 'S)-2-'pyrrolidin-1-yl-
methyl)pyrrolidide 74a,b,9and lithium 'S)-2-'piperidin-1-yl-
methyl)pyrrolidide 84a,b,9 'Fig. 1) in enantioselective
deprotonations of cyclohexene oxide 10 'Scheme 1).
Scheme 1. Enantioselective chiral lithium amide 'pCLi) catalysed deproto-
nation in presence of bulk base 'BLi) of cyclohexene oxide 10.
carbenoid compound 2-lithio-1-methylimidazole 1 and
2-'lithiomethyl)-1-methylimidazole 2, respectively 'Fig. 1).5,6
In THF, we have shown by NMR spectroscopy that 1 and 2
have basicities comparable to LDA. The deprotonating
ability of bases 1 and 2 has been studied using cyclohexene
oxide 10 as a substrate. In contrast to LDA, compounds 1
and 2 did not measurably yield any deprotonation of the
epoxide. This is presumably due to the fact that proton
transfers to carbon are usually slower than that to a more
electronegative atom like nitrogen. These properties
indicated that 1 and 2 should be useful as bulk bases in
catalytic asymmetric deprotonations.
2. Results and discussion
2.1. Epoxide rearrangement
All the bulk bases except N-lithio-imidazole 4 were found to
be strong enough bases to generate the chiral lithium amide.
The results with the regenerating bulk bases are collected in
Table 1. In the absence of any lithium amide catalyst, the
bulk bases 1±3 were found to be non-reactive towards the
Table 1. Stoichiometric and catalytic enantioselective deprotonation of cyclohexene oxide '0.1 M) in THF at 20.08C using the chiral lithium amides 5±8
Entry
[Catalyst]a 'M)
Bulk base/conc.a 'M21
)
[DBU]a 'M)
Reaction timeb 'h)
Yieldc '%)
eed '%)
1
2
3
4
5
6
7
8
1/0.2
2/0.2
3/0.2
9/0.2
9/0.2
252
149
209
220
232
25
25
32
198
13
14
56
240
60
28
23
16
32
69
42
36
22
8
12
28
1
8
7
7
6
6
8
12
12
5
12
5
0
0
0
4
±
±
±
0
0.5
2.5
95
96
90
96
96
91
93
57
88
96
96
70
86
58
65
63
46
78
63
84
81
78
89
81
82
73
76
66
80
88
71
80
86
88
0
5/0.1
5/0.2
93
93
22
93
93
93
69
89
94
96
96
77
19
73
73
72
13
80
68
76
76
80
74
79
76
78
72
59
75
75
75
74
75
75
5/0.02
5/0.02
5/0.02
5/0.02
5/0.02
5/0.1
5/0.1
5/0.1
5/0.1
6/0.1
6/0.02
6/0.02
6/0.02
6/0.1
6/0.1
7/0.1
7/0.02
7/0.02
7/0.1
7/0.02
7/0.02
7/0.02
7/0.1
LDA/0.2
1/0.2
2/0.2
3/0.2
9/0.2
9/0.2
9/0.2
2/0.1
2/0.1
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26e
27
28
29
30
31
32
33
34
35e
36
37e
38
39
0.5
0.5
LDA/0.2
1/0.2
2/0.2
1/0.1
2/0.1
LDA/0.2
1/0.2
1/0.6
2/0.2
LDA/0.2
9/0.2
1/0.1
0.5
0.5
7/0.1
8/0.1
8/0.02
8/0.02
8/0.1
8/0.02
8/0.1
8/0.1
2/0.1
LDA/0.2
1/0.2
1/0.6
2/0.2
2/0.6
1/0.1
2/0.1
4
4
8/0.1
a
Total initial concentration.
b
c
d
e
The reaction was quenched at the time stated.
The yield was determined by GC-analysis.
The ee of 'S)-11 was determined by chiral GC.
The initial epoxide concentration was 0.5 M.

D. Pettersen et al. / Tetrahedron 58 +2002) 4669±4673
4671
epoxide 10 in THF gave 'S)-11 in 77% ee 'entry 17). In
contrast, using LDA as a bulk base the ee decreased to 19%
'entry 18). The catalytic reaction of chiral base 6 in the
presence of either 1 or 2 on the other hand gave improved
enantioselectivity upto 73% ee of ' S)-11 'entries 19 and
20). Using stoichiometric amounts of epoxide 10, chiral
base 6 and 1 or 2 resulted in ees of 72 and 13%, respectively
'entries 21 and 22).
Using Asami's chiral base 7 in stoichiometric amounts
produced 'S)-11 in 80% ee 'entry 23), but using sub-
stoichiometric amounts of 7 and LDA as a bulk base
lowered the ee to 68% 'entry 24). The ee increased to
76% when 1 was used as a bulk base 'entries 25 and 26).
The enantioselectivity increased further when 2 was used in
place of 1 'entry 27). In the case of chiral base 7, improved
selectivity was obtained when using 1 or 2 'entries 25±27)
as a bulk base. Interestingly, sub-stoichiometric amounts of
chiral base 7 in the presence of both LDA and DBU as an
additive gave 74% ee of 'S)-11 'entry 28), but when
employing chiral base 7 with bulk base 9, 'S)-11 was formed
in 79% ee 'entry 29). Rearrangement of epoxide 10 in a
solution with equimolar amounts of 7 and 1 or 2 resulted
in ees of 76 and 78%, respectively 'entries 30 and 31).
Figure 2. Mixed dimers built from monomers of 5 and monomers of 1, 2
and 9, respectively, have been shown by multinuclear NMR investigations
to be the major species present in THF.
epoxide 'entries 1±3). Bulk base 9 did yield only a very
slow deprotonation of the epoxide 'entries 4 and 5).
Reaction of epoxide 10 with chiral base 8 gave 'S)-11 in
72% ee 'entry 32). Similarly, reaction with sub-stoichio-
metric amounts of chiral base 8 in presence of the bulk
base LDA resulted in a lower ee of 59% 'entry 33). In
contrast when 1 or 2 were used as bulk bases 'entries
34±36) a slight increase of the ee to 75% compared with
the stoichiometric reaction with 8 was noticed. Similarly
rearrangement of epoxide 10 with equimolar amounts of
chiral base 8 and 1 or 2 resulted in a 75% ee of 'S)-11
'entries 38 and 39).
Using chiral base 5 in stoichiometric amount or in excess of
epoxide 10 'entries 6 and 7) 'S)-2-cyclohexen-1-ol 'S)-11 is
formed in 93% ee. In contrast, using a sub-stoichiometric
amount of chiral base 5 and LDA as a bulk base gave 'S)-11
in only 22% ee 'entry 8). When employing either 1, 2 or 3
as bulk base the reactions gave 93% ee of 'S)-11 'entries
9±11). However, the catalytic reaction was much slower
using 1 than the reaction with 2. Furthermore, the reaction
using the bulk base 2 is faster than that with LDA. Thus, a
96% yield was achieved within 13 h without a change of
the enantioselectivity 'entry 10). Surprisingly, chiral base
5 together with stoichiometric amounts of either 1 or 2
increased the enantioselectivity to 96% ee 'entries 15 and
16). This is further discussed below as are the results
obtained in the presence of the bulk base 9.
2.2. Mixed dimers as reagents or catalysts
The sometimes increased enantioselectivity obtained with
the novel bulk bases over the corresponding reactions using
only the chiral lithium amides suggested to us that new
reagents were present in the former cases. Interestingly,
6
we found by NMR spectroscopy using the [15N, Li]-iso-
The stoichiometric reaction of Singh's chiral base 6 with
topologue of chiral base 5 and bulk bases 1 and 2, that mixed
dimers 12 and 13, respectively, were formed in THF 'Fig.
2).5 These species were found to be more stable than the
homodimer of chiral base 5, which has been shown to be the
reagent in some ether solutions in absence of bulk bases 1
and 2.5 Therefore the new reagents in the enantioselective
deprotonations when using bulk bases 1 and 2 with chiral
lithium amide 5 are the mixed dimers.
Scheme 2 shows a catalytic cycle involving the chiral amide
5, the mixed dimer 12 and the bulk base 1.
The above ®ndings have led us to explore the potential
of mixed dimers in both catalytic and stoichiometric
enantioselective deprotonations.
We have previously reported that the rate limiting transition
state in the deprotonation of epoxide 10 by chiral base 5 is
built from one dimer of the lithium amide and one epoxide
molecule.7 It is possible that deprotonation by the mixed
Scheme 2. Catalytic cycle using the mixed dimer catalyst 12 in presence of
the bulk base 1 producing 'S)-11 and the precursor of 1, 1-methylimidazole.

4672
D. Pettersen et al. / Tetrahedron 58 +2002) 4669±4673
dimers is making use of a rate-limiting transition state built
from a mixed dimer and an epoxide.
distilled from sodium and benzophenone. The concentration
of the commercially available n-BuLi 'ca. 2.5 M in hexanes,
Acros) was determined by titration.7 Cyclohexene oxide,
1-methylimidazole, DBU, and furan 'Sigma-Aldrich) were
distilled from CaH2. 1,2-Dimethylimidazole and imidazole
'Aldrich) were puri®ed by distillation. All precursors were
by NMR spectroscopy and GC found to be not less than
99.3% chemically pure. The diamine precursor of 5 was
prepared in 99.2% ee as found by chiral GC.7 The diamine
precursor of 6 was prepared as described earlier by Bhuniya
et al.2f They reported an enantiomeric excess of at least 97%
based on optical rotation. The diamine precursors of 7 and 8
were synthesised using a method recently developed in our
laboratory.9 The enantiomeric purity was determined
2.3. On the role of DBU as an additive
It has been reported that addition of amidines like DBU
'1,8-diazabicyclo[5.4.0]undec-7-ene) alters the reactivity
and the enantioselectivity in epoxide rearrangements.3a,4
In light of our ®ndings, together with earlier reports, we
envisioned that DBU under the conditions used, may be
lithiated and working as a bulk base besides being a strongly
solvating agent.10 Indeed, NMR studies have shown that
DBU is deprotonated by either n-BuLi or LDA to give
lithiated DBU 9 'Fig. 1).10 LDA appears not to be strong
enough to completely deprotonate DBU. An equilibrium
mixture composed of 9, DBU, LDA and diisopropylamine
is obtained. These ®ndings led us to investigate the useful-
ness of 9 as a bulk base. Using multinuclear NMR and
isotopically labelled 5 together with 9 it has been shown
that a mixed dimer 14 is formed 'Scheme 2) similar to
those obtained with 1 and 2 and chiral lithium amide in
THF.10
20
by optical rotation, [a]D ˆ18.87 'c 2.24, EtOH) and
20
[a]D ˆ115 'c 7.75, EtOH), respectively, comparing with
literature data.11 For all chiral base reactions the reported
ees for the deprotonation of epoxide 10 in the literature has
been reproduced.4a,b,8 All GC analyses were run on a chiral
stationary phase column 'CP-Chirasil-DEX CB, 25 m,
0.32 mm) from Chrompack. The column was held at 908C
'injector 2258C, detector 2508C) using helium '2 mL min21
)
as a carrier gas. tR'10)ˆ3.25 min, tR''S)-11)ˆ7.45 min,
tR''R)-11)ˆ7.90 min.
In Table 1, entry 13 the result of the reaction of the epoxide
10 with equimolar amounts of 5 and twice the equimolar
amount of 9 is shown. Under these conditions the rearrange-
ment was much slower and the ee of 'S)-11 was lowered to
89% compared to 93% 'entry 6). Using a reaction mixture
with the same composition as in entry 13 but with the
addition of DBU 'entry 14) resulted in shorter reaction
times and an increase of the ee to 94%. This demonstrates
that DBU also plays an important solvation role in which it
in¯uences both the enantioselectivity and rates. In entry 12
the amide 5 is used in sub-stoichiometric amounts and the
initial concentration of bulk base 9 is ten times that of 5 and
the concentration of DBU is the same as in entry 14. Under
these conditions the reaction time was short but the ee was
lowered to 69%. The latter result is possibly due to a
competing racemic reaction of 9 with 10 and or racemi-
sation of 11 by 9.
4.2. Catalytic deprotonation of cyclohexene oxide 10
The bulk bases were prepared from n-BuLi and the appro-
priate precursor, i.e. 1-methylimidazole '1), 1,2-dimethyl-
imidazole '2), furan '3), imidazole '4) or diisopropylamine
'LDA). In the case of the crystalline precursors 1,2-di-
methylimidazole and imidazole, 2.0 M stock solutions in
THF were prepared.
Amine 5 '4.4 mL, 0.02 mmol) and 1-methylimidazole
'16 mL, 0.20 mmol) were added to THF '881 mL) in a reac-
tion vessel in the glove box. After transfer out of the glove
box n-BuLi '89 mL, 2.47 M in hexanes, 0.22 mmol) was
added to the reaction vessel in a nitrogen atmosphere. The
yellow reaction solution was allowed to equilibrate at
20.08C for 10 min in a thermostat. The reaction was started
by addition of cyclohexene oxide 3 '10 mL, 0.10 mmol) to
the reaction mixture. To follow the reaction, samples
'50 mL) were withdrawn at different times and diluted
with diethyl ether '500 mL). The solutions were quenched
in saturated NH4Cl '250 mL) and washed with brine
'250 mL). The samples were analysed by chiral GC and
the reaction yield of 11 was determined using a standard
added after the quenching as previously described.7
3. Conclusion
Novel bulk bases have been developed yielding improved
enantioselectivity of chiral lithium amide catalysed depro-
tonation as compared to using the bulk base LDA. The
reactivity change has been traced to the formation of
novel reagensÐmixed dimersÐformed from a bulk base
molecule and molecule of a chiral lithium amide.
Acknowledgements
4. Experimental
4.1. General
We thank the Swedish Natural Science Research Council
for ®nancial support and Dr Sten. O. Nilsson Lill for fruitful
discussions.
All syringes and glass vessels used were dried overnight in a
vacuum oven '508C) before being transferred into a glove
box 'Mecaplex GB 80 equipped with a gas puri®cation
system that removes oxygen and moisture) containing a
nitrogen atmosphere. Typical moisture content was less
than 0.5 ppm. All handling of the compounds was carried
out with gas-tight syringes. The solvent THF used was
References
1. For reviews see: 'a) Jones, S. J. Chem. Soc., Perkin Trans. 1
2002, 1. 'b) O'Brien, P. J. Chem. Soc., Perkin Trans. 1 2001,
95. 'c) O'Brien, P. J. Chem. Soc., Perkin Trans. 1 1998, 1439.

D. Pettersen et al. / Tetrahedron 58 +2002) 4669±4673
4673
'd). Hodgson, D. M.; Gibbs, A. R.; Lee, G. P. Tetrahedron
1996, 52, 14361. 'e) Cox, P. J.; Simpkins, N. S. Tetrahedron:
Asymmetry 1991, 2, 1.
5. 'a) Amedjkouh, M.; Pettersen, D.; Nilsson Lill, S. O.;
È
Davidsson, O.; Ahlberg, P. Chem. Eur. J. 2001, 7, 4368±
4377. 'b) Nilsson Lill, S. O.; Pettersen, D.; Amedjkouh, M.;
Ahlberg, P. J. Chem. Soc., Perkin Trans. 1 2001, 3054.
6. Hilf, C.; Bosold, F.; Harms, K.; Marsch, M.; Boche, G. Chem.
Ber. 1997, 130, 1213.
2. 'a) Hodgson, D. M.; Gibbs, A. R. Synlett 1997, 657.
'b) Hodgson, D. M.; Witherington, J.; Moloney, B. A.
J. Chem. Soc., Perkin Trans. 1 1994, 3373. 'c) Asami, M.;
Takahashi, J.; Inoue, S. Tetrahedron: Asymmetry 1994, 5,
1649. 'd) Kasai, T.; Watanabe, H.; Mori, K. Biorg. Med.
Chem. 1993, 1, 67. 'e) Mori, K.; Murata, N. Liebigs Ann.
Chem. 1995, 2089. 'f) Bhuniya, D.; DattaGupta, A.; Singh,
V. K. J. Org. Chem. 1996, 61, 6108. 'g) Asami, M.; Inoue, S.
Tetrahedron 1995, 51, 11725. 'h) de Sousa, S. E.; O'Brien, P.;
Pilgrim, C. D. Tetrahedron Lett. 2001, 42, 8081.
Â
7. Pettersen, D.; Amedjkouh, M.; Nilsson Lill, S. O.; Dahlen, K.;
Ahlberg, P. J. Chem. Soc., Perkin Trans. 2 2001, 1654.
8. Colman, B.; de Sousa, S. E.; O'Brien, P.; Towers, T. D.;
Watson, W. Tetrahedron: Asymmetry 1999, 10, 4175.
9. Diamine precursors of 7 and 8 were synthesised according to a
coupling reaction developed in our group. Amedjkouh, M.;
Ahlberg, P., submitted for publication.
10. 'a) Pettersen, D.; Amedjkouh, M.; Nilsson Lill, S.O.; Ahlberg,
P., submitted for publication. 'b) Matsumura, N.; Ohba, T.;
Yoneda, S. Chem. Lett. 1983, 317. 'c) Matsumura, N.;
Nishiguchi, H.; Okada, M.; Yoneda, S. Heterocycles 1985,
885.
3. 'a) Tierney, J. P.; Alexakis, A.; Mangeney, P. Tetrahedron:
Asymmetry 1997, 8, 1019. 'b) Amadji, M.; Vadecard, J.;
Plaquevent, J.-C.; Duhamel, L.; Duhamel, P. J. Am. Chem.
Soc. 1996, 118, 12483. 'c) Yamashita, T.; Sato, D.; Kiyoto,
T.; Kumar, A.; Koga, K. Tetrahedron 1997, 53, 16987.
4. 'a) Asami, M.; Ishizaki, T.; Inoue, S. Tetrahedron: Asymmetry
1994, 5, 793. 'b) Asami, M.; Suga, T.; Honda, K.; Inoue, S.
11. 'a) Asami, M. Bull. Chem. Soc. Jpn 1990, 63, 721.
'b) Harrison, J. R.; O'Brien, P. Synth. Commun. 2001, 31,
1155.
È
Tetrahedron Lett. 1997, 38, 6425. 'c) Sodergren, M. J.;
Andersson, P. G. J. Am. Chem. Soc. 1998, 120, 10760.
È
'd) Sodergren, M. J.; Bertilsson, S. K.; Andersson, P. G.
J. Am. Chem. Soc. 2000, 122, 6610.