The influence of imine structure, catalyst structure and reaction conditions on the enantioselectivity of the alkylation of alanine methyl ester imines catalyzed by Cu(ch-salen)

Article abstract of DOI:10.1016/S0040-4039(01)01718-X

Systematic variation of the substrate structure has shown that the most effective substrates for Cu(ch-salen)-catalyzed asymmetric enolate alkylation reactions carried out under phase-transfer conditions are the para-chlorophenyl imines of amino esters. T

Full text of DOI:10.1016/S0040-4039(01)01718-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 8093–8096
The influence of imine structure, catalyst structure and reaction
conditions on the enantioselectivity of the alkylation of alanine
methyl ester imines catalyzed by Cu(ch-salen)
b
a
a
a,
a
Yuri N. Belokon’, R. Gareth Davies, Jose A. Fuentes, Michael North * and Teresa Parsons
a
Department of Chemistry, King’s College London, Strand, London WC2R 2LS, UK
b
A.N. Nesmeyanov Institute of Organo-Element Compounds, Russian Academy of Sciences, 117813 Moscow,
Vavilov 28, Russian Federation
Received 1 August 2001; revised 3 September 2001; accepted 12 September 2001
Abstract—Systematic variation of the substrate structure has shown that the most effective substrates for Cu(ch-salen)-catalyzed
asymmetric enolate alkylation reactions carried out under phase-transfer conditions are the para-chlorophenyl imines of amino
esters. The other reaction parameters (solvent and stirring speed) have also been optimized. The introduction of substituents onto
the aryl rings of the salen ligand was found not to have a beneficial effect on the enantioselectivity of the reaction. © 2001 Elsevier
Science Ltd. All rights reserved.
a,a-Disubstituted amino acids are important compo-
chiral environment during the alkylation reaction. The
latter process is attractive in terms of reducing the
number of synthetic steps and cost of the process, and
has been achieved by conducting the reaction under
phase-transfer conditions in the presence of a chiral
phase-transfer catalyst. Most commonly, the chiral
phase-transfer reagent is an ammonium salt derivative
nents of a number of pharmaceuticals and potential
1
pharmaceuticals. The most attractive route to the syn-
thesis of these compounds is to start from a naturally
occurring a-monosubstituted amino acid and to intro-
duce the second side-chain through alkylation of an
enolate (Scheme 1). In the simplest case, the stereo-
chemical information contained in the original amino
acid will be lost during formation of the enolate, lead-
4
of a cinchona alkaloid, though recently, excellent
results have been obtained with a synthetic quaternary
2
5
ing to the formation of a racemic product. This can be
ammonium salt.
avoided by the use of a chiral auxiliary attached to the3
amino or acid component of the original amino acid,
or by the use of a chiral catalyst/base system to ensure
that the amino acid enolate is coordinated within a
In recent papers, we have shown that a variety of metal
complexes can also be used as chiral phase-transfer
reagents. Initially, sodium taddolate was found to be
6
7
effective, and more recently sodium-NOBIN and a
8
copper salen complex 1a were found to give excellent
results when applied to the benzylidene imine of alanine
esters in the presence of solid sodium hydroxide in
toluene at room temperature (Scheme 2). The great
advantage of the salen based system is the ease with
which the structure of the ligand can be varied by the
introduction of substituents onto the aromatic rings to
Scheme 1.
*
0
Corresponding author.
Scheme 2.
040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01718-X

8
094
Y. N. Belokon’ et al. / Tetrahedron Letters 42 (2001) 8093–8096
Table 1. Influence of the structure of the aryl group of 2
of the substrate, the influence of electronic effects was
investigated. The introduction of an electron-withdraw-
ing substituent in the para-position of the benzylidene
imine should increase the acidity of the alanine a-pro-
ton, and an electron-donating substituent should
decrease the acidity of this proton. This is important
since a control experiment had shown that even in the
absence of catalyst 1a, N-benzylidene alanine methyl
ester was converted into racemic N-benzylidene-a-
methyl phenylalanine methyl ester in 56% yield under
our standard conditions. Hence, it was anticipated that
an electron-rich imine might be beneficial by suppress-
ing the uncatalyzed reaction. In the event however,
both an electron-donating (OMe) and an electron-with-
on the enantioselective alkylation reaction
Structure of aryl group
Chemical yield (%)
ee (%)a
Ph
91
89
62
79
50
44
71
95
67
43
53
81
79
77
71
65
84
92
81
86
70
81
1
2
4
4
4
4
4
4
2
3
-Naphthyl
-Naphthyl
-MeOC H4
-O NC H
-FC H4
-ClC H4
6
2
6
4
6
6
6
-BrC H4
6
-IC H4
-ClC H4
-ClC H4
6
6
drawing (NO ) substituent were detrimental to the
2
a
enantioselectivity of the reaction.
Enantiomeric excesses are accurate to ±3%.
More success was obtained when the use of halo-
genated aryl imines was investigated. A para-fluoro
substituent had a small positive effect on the enantiose-
lectivity, and this improved significantly when the
fluorine atom was changed to chlorine. Curiously, a
para-bromo substituent completely reversed the benefi-
cial effect of the fluoro- or chloro-substituent, giving
identical results to the benzylidene imine. Finally, an
iodo substituent again increased the enantioselectivity,
giving results comparable to a fluoro substituent. Since
a chloro substituent seemed to be optimal, the effect of
its location was investigated. If the chloro substituent
was located in the meta-position of the aryl ring, then
the enantioselectivity was identical to the benzylidene
imine, whilst locating the chloro group in the ortho-
position had even more detrimental effect on the enan-
tioselectivity. These studies indicate that the optimal
structure for the aryl group is para-chlorophenyl.
produce the optimal catalytic activity. However, before
embarking on a study to optimize the catalyst, we
elected to ensure that the structure of the substrate was
optimal and that the reaction conditions were opti-
mized. The use of a readily prepared and inexpensive
methyl ester was highly desirable, and the amino acid
side-chain had to be left variable, so the scope for
optimization was limited to the structure of the imine.
In this paper, the influence of the structure of the imine
group on the enantioselectivity of the reaction shown in
Scheme 2 with copper complex 1a as catalyst is dis-
cussed, along with the influence of the solvent and
stirring speed on the enantioselectivity. Then, the effect
of introducing substituents onto the aromatic rings of
the salen ligand is reported.
At present, it is not clear why 4-halo substituents
should be beneficial in the imine. The halo group is
inductively electron withdrawing which should increase
the acidity of the alanine a-proton, but is also
mesomerically electron donating. It may be that a
combination of these two electronic effects and other
factors can explain the order of effectiveness as Cl>I=
F>Br.
A range of aryl imines 2 of alanine methyl ester were8
prepared and alkylated under standard conditions
using 2 mol% of complex 1a as an asymmetric phase-
transfer catalyst, finely powdered sodium hydroxide
Having optimized the substrate when benzyl bromide
was used as the electrophile, the effect of this optimiza-
tion on other electrophiles was investigated (Table 2).
Allyl and propargyl bromides were chosen because of
the relatively low enantioselectivity observed using
these alkylating agents with the benzylidene imine. To
further demonstrate the utility of the chemistry, and to
avoid solubility problems associated with purifying the
methyl esters of these rather hydrophilic amino acids,
they were converted directly into the amino acids by
modifying the work-up as previously described (Scheme
(
3.5 equiv.) as base, and benzyl bromide as alkylating
agent. All reactions were carried out in toluene at
ambient temperature in an argon atmosphere. Reac-
tions were worked up with silica gel to cleave the imine,
giving a-methyl phenylalanine methyl ester as the iso-
lated product. The enantiomeric excess was determined
by treatment of this amino ester with an excess of
1
-phenylethylisocyanate followed by NMR analysis as
8
8
previously described. The results of this study are
given in Table 1.
3). A sample was then re-esterified using methanolic
HCl for enantiomeric excess determination. In contrast,
1
-bromomethylnaphthalene was chosen since it had
Changing the benzylidene imine to a larger imine,
previously been found to be a good substrate for the
N-benzylidene imine, and produced a hydrophobic
amino ester. The data in Table 2 show that the para-
chlorophenyl imine substrate is always as effective or
1
-naphthyl or 2-naphthyl did not have a beneficial
effect on the enantiomeric excess of the product. Since
steric effects did not seem to be important in this region

Y. N. Belokon’ et al. / Tetrahedron Letters 42 (2001) 8093–8096
8095
Table 2. Alkylation of imine 2 (Ar=4-chlorophenyl) with various electrophiles
Alkylating agent
Yield (%)
ee (%)a
ee obtained with benzylidene imine
1
-Bromomethylnaphthalene
93
67
82
84
69
58
86
72
43
Allyl bromide
Propargyl bromide
a
Enantiomeric excesses were determined by 1H NMR spectroscopy after reaction of the amino acid methyl ester with excess 1-phenylethyliso-
cyanate and are accurate to ±3%.
Table 3. Influence of solvent on the asymmetric alkylation
reaction
Solvent
Yield (%)
ee (%)a
Toluene
THF
Dichloromethane
Petroleum ether (40–60)
Hexane
71
66
45
0
92
39
44
74
20
a
Enantiomeric excesses were determined by 1H NMR spectroscopy
after reaction of the amino acid methyl ester with excess 1-
phenylethylisocyanate and are accurate to ±3%.
Scheme 3.
more effective (within the limits of experimental error)
than the benzylidene imine containing substrate.
The effect of solvent on yield and enantioselectivity was
also investigated using the para-chlorophenyl imine of
alanine methyl ester with benzyl bromide as the alkylat-
ing agent. However, as Table 3 illustrates, whilst good
chemical yields were obtained in both THF and hexane,
only toluene gave a high level of asymmetric induction.
Hence, it appears that a non-polar aromatic solvent is
necessary to obtain high enantiomeric excesses.
Scheme 4.
stituents with different steric and electronic properties
were chosen, giving complexes 1b–f. The copper com-
plexes were all prepared from the corresponding salen
ligands and copper bromide as previously reported for
8
complex 1a. The salen ligands were prepared from the
appropriate para-substituted salicylaldehyde and cyclo-
Finally, the influence of stirring rate on the chemical
yield and enantioselectivity was studied, again using the
reaction of the para-chlorophenyl imine of alanine
methyl ester with benzyl bromide. The enantiomeric
excess was found not to vary significantly with stirring
speed, though the chemical yield was highest at a
stirring rate of about 500 rpm. A lower rate of stirring
8
hexane diamine. The only salicylaldehyde that was not
a known compound was 2-hydroxy-5-trifluoromethyl
benzaldehyde 3, which was prepared from 4-tri-
fluoromethyl phenol as shown in Scheme 4. Thus,
bromination of 4-trifluoromethyl phenol gave 2-bromo-
9
4
-trifluoromethyl phenol which could be lithiated and
reacted with dimethylformamide to give the desired
(
100 rpm) gave a very low chemical yield, presumably
10
aldehyde.
due to inadequate agitation of the heterogeneous reac-
tion mixture, whilst a higher rate of stirring (1,000 rpm)
gave a low yield due to solid being expelled from the
reaction solution.
Complex 1d was found to be highly insoluble and so
was used without purification. The other copper com-
plexes could all be purified by size exclusion chro-
matography as previously described. The results
obtained when complexes 1b–f were used as asymmetric
phase-transfer catalysts for the alkylation of N-benzyli-
dene alanine methyl ester with benzyl bromide are
presented in Table 4. Complex 1b gave the same level of
asymmetric induction as the unsubstituted complex 1a,
confirming that there was no detrimental (or advanta-
geous) steric effect associated with these complexes.
Surprisingly however, both electron-donating 1c and
electron-withdrawing 1d–f substituents had a severely
detrimental effect on the enantioselectivity of the cata-
lyst. It should be noted that even in the absence of any
Having optimized the substrate and reaction condi-
tions, the effect of substituents on the aryl rings of the
salen ligand was investigated. Previous work had
already shown that the introduction of a tert-butyl
group into the ortho- (relative to phenol) position of
the aryl ring had a seriously detrimental effect on the
enantioselectivity of the catalyst. Hence, the sub-
stituents were located in the para-position of the aryl
rings where they would be able to exhibit an electronic
effect but would not be expected to interfere with the
coordination of substrates to the copper ion. Five sub-

8
096
Y. N. Belokon’ et al. / Tetrahedron Letters 42 (2001) 8093–8096
Table 4. Influence of ligand structure on the enantioselec-
tivity of the alkylation of N-benzylidene alanine methyl
ester
tively) and King’s College London for financial
support.
Complex
Yield (%)
ee (%)a
References
1
1
1
1
1
b
c
d
e
f
39
78
60
54
68
80
45
0
42
25
1
. For example, see: (a) Saari, W. S.; Halczenko, W.;
Cochran, D. W.; Dobrinska, M. R.; Vincek, W. C.; Titus,
D. C.; Gaul, S. L.; Sweet, C. S. J. Med. Chem. 1984, 27,
713; (b) Fenteany, G.; Standeart, R. F.; Lane, W. S.;
Choi, S.; Corey, E. J.; Schreiber, S. L. Science 1995, 268,
a
Enantiomeric excesses were determined by 1H NMR spectroscopy
after reaction of the amino acid methyl ester with excess 1-
phenylethylisocyanate and are accurate to ±3%.
726; (c) Hanessian, S.; Haskell, T. H. Tetrahedron Lett.
1964, 2451; (d) Jung, G.; Beck-Sickinger, A. G. Angew.
Chem., Int. Ed. Engl. 1992, 31, 367; (e) Veber, D. F.;
Freidinger, R. M. Trends Neuorosci. 1995, 8, 392.
catalyst, a 56% yield of racemic product is obtained.
Hence, the unpurified complex 1d may not exert any
catalytic influence on the reaction. The enantioselec-
tivities observed for the other complexes may also be
being affected to various extents (depending on the
relative reaction rates) by the competing uncatalyzed
reaction, thus masking the true influence of the sub-
stituent. Hence, it appears that the optimal electronic
properties of the ligand are met by an unsubstituted
or alkyl substituted salen complex.
2
. Yaozhong, J.; Changyou, Z.; Shengde, W.; Daimo, C.;
Youan, M.; Guilan, L. Tetrahedron 1988, 44, 5343.
. (a) Schollkopf, U.; Tolle, R.; Egert, E.; Nieger, M.
Liebigs Ann. Chem. 1987, 399; (b) Yamada, S.-I.; Oguri,
T.; Shioiri, T. J. Chem. Soc., Chem. Commun. 1976, 136;
3
(
c) Ikegami, S.; Uchiyama, H.; Hayama, T.; Katsuki, T.;
Yamaguchi, M. Tetrahedron 1988, 44, 5333.
4
. For leading references, see: (a) O’Donnell, M. J.; Del-
gado, F.; Hostettler, C.; Schwesinger, R. Tetrahedron
Lett. 1998, 39, 8775; (b) Lygo, B. Tetrahedron Lett. 1999,
4
0, 1389; (c) Alvarez, R.; Hourdin, M.-A.; Cave, C.;
d’Angelo, J.; Chaminade, P. Tetrahedron Lett. 1999, 40,
091; (d) Horikawa, M.; Busch-Petersen, J.; Corey, E. J.
Further work is needed to understand exactly why
these reaction conditions are optimal. It is possible
that a combination of steric and electronic effects
determine the optimal structure of the imine, and that
p–p interactions between the substrate, catalyst and
solvent are also involved in optimizing the enantiose-
lectivity. The preference for no substituents to be
present on the aryl rings of the catalyst is consistent
with an oligomeric form of the catalyst being the
active species; this is also consistent with the observa-
tion of a non-linear effect during these reactions as
7
Tetrahedron Lett. 1999, 40, 3843; (e) Arai, S.; Nakayama,
K.; Ishida, T.; Shioiri, T. Tetrahedron Lett. 1999, 40,
4
. (a) Ooi, T.; Kameda, M.; Maruoka, K. J. Am. Chem.
Soc. 1999, 121, 6519; (b) Ooi, T.; Takeuchi, M.; Kameda,
M.; Maruoka, K. J. Am. Chem. Soc. 2000, 122, 5228; (c)
Ooi, T.; Takeuchi, M.; Ohara, D.; Maruoka, K. Synlett
215.
5
2
001, 1185.
6
. Belokon’, Y. N.; Kochetkov, K. A.; Churkina, T. D.;
Ikonnikov, N. S.; Chesnokov, A. A.; Larionov, O. V.;
Parmar, V. S.; Kumar, R.; Kagan, H. B. Tetrahedron:
Asymmetry 1998, 9, 851.
8
reported earlier.
In conclusion, we have shown that the structure of
the imine within the substrate has a significant influ-
ence on the enantioselectivity of asymmetric phase-
transfer alkylation reactions of alanine methyl ester
with a para-chlorophenyl imine being optimal. The
reaction solvent and rate of stirring have also been
optimized. Attempts to optimize the structure of the
salen ligand have indicated that the initial choice of
an unsubstituted aryl ring provides optimal steric and
electronic properties to the ligand under the reaction
conditions.
7
. (a) Belokon’, Y. N.; Kochetkov, K. A.; Churkina, T. D.;
Ikonnikov, N. S.; Vyskocil, S.; Kagan, H. B. Tetra-
hedron: Asymmetry 1999, 10, 1723; (b) Belokon, Y. N.;
Kochetkov, K. A.; Churkina, T. D.; Ikonnikov, N. S.;
Larionov, O. V.; Harutjunan, S. R.; Vyskocil, S.; North,
M.; Kagan, H. B. Angew. Chem., Int. Ed. 2001, 40, 1948.
. (a) Belokon’, Y. N.; North, M.; Kublitski, V. S.; Ikon-
nikov, N. S.; Krasik, P. E.; Maleev, V. I. Tetrahedron
Lett. 1999, 40, 6105; (b) Belokon’, Y. N.; Davies, R. G.;
North, M. Tetrahedron Lett. 2000, 41, 7245; (c) Belokon’,
Y. N.; North, M.; Churkina, T. D.; Ikonnikov, N. S.;
Maleev, V. I. Tetrahedron 2001, 57, 2491.
8
Acknowledgements
9. Yamamoto, S.; Hashiguchi, S.; Miki, S.; Igata, Y.;
Watanabe, T.; Shiraishi, M. Chem. Pharm. Bull. 1996, 44,
734.
The authors thank the EPSRC and King’s College
London for studentships (to T.P. and J.A.F., respec-
10. Johannes, C. W.; Visser, M. S.; Weatherhead, G. S.;
Hoveyda, A. H. J. Am. Chem. Soc. 1998, 120, 8340.