Stereoselective synthesis of amino acid-derived β-lactams. Experimental evidence for TADDOL as a memory of chirality enhancer

Article abstract of DOI:10.1016/j.tet.2005.09.125

In model experiments, concerning the modulation of memory of chirality during the intramolecular alkylation of N-(p-methoxy)benzyl-N-chloroacetyl-Phe- O^tBu derivative to the corresponding β-lactam, TADDOL was selected from a panel of different

Full text of DOI:10.1016/j.tet.2005.09.125

Tetrahedron 62 (2006) 130–138
Stereoselective synthesis of amino acid-derived b-lactams.
Experimental evidence for TADDOL as a memory
of chirality enhancer
a
M. Angeles Bonache, Pilar L o´ pez, Mercedes Mart ´ı n-Mart ´ı nez, M. Teresa Garc ´ı a-L o´ pez,
b
a
a
b
Carlos Cativiela and Rosario Gonz a´ lez-Mu n˜ iz
a,
*
a
Instituto de Qu ´ı mica M e´ dica-CSIC, Juan de la Cierva, 3, 28006 Madrid, Spain
b
Departamento de Qu ´ı mica Org a´ nica, ICMA, Universidad de Zaragoza-CSIC, 50009 Zaragoza, Spain
Received 30 June 2005; revised 1 August 2005; accepted 29 September 2005
Available online 24 October 2005
Abstract—In model experiments, concerning the modulation of memory of chirality during the intramolecular alkylation of N-(p-
t
methoxy)benzyl-N-chloroacetyl-Phe-O Bu derivative to the corresponding b-lactam, TADDOL was selected from a panel of different phase-
transfer catalysts as the best chiral additive (ee up to 82%). We have demonstrated here that the degree and sign of the cyclization selectivity
was virtually independent on TADDOL configuration, providing experimental proof that in this reaction this additive works as a memory of
chirality enhancer and not as a real catalyst. The effect of TADDOL is strongly dependent on the starting amino acid derivative, the increase
of selectivity being only observed for those with aromatic side chains.
q 2005 Elsevier Ltd. All rights reserved.
3
–17
18–22
1. Introduction
reactions
and different photocyclization processes.
Among them, the intermolecular a-alkylation of a number
of a-amino acid derivatives, extensively studied by Fuji and
Kawabata group, proceeded with a high degree of
In recent years, seminal research from various laboratories
has confirmed the memory of chirality phenomenon as a
conceptually novel approach for asymmetric induction.
1
3–8
asymmetric induction.
version of this process allowed the enantioselective
Similarly, the intramolecular
Memory of chirality occurs in processes where an initial
stereogenic center is destroyed during the generation of the
corresponding reactive intermediate, but this intermediate is
able to ‘remember’ the configuration of its precursor to
transfer the chirality to the final compound, without using
any external chiral source. This intriguing phenomenon has
been observed in a variety of reactions, involving different
reactive intermediates, such as carbanions, carbenium
cations, and different types of radicals. Except for a few
9
,10
preparation of valuable azacyclic amino acid derivatives.
The enantioselectivity in these reactions was attributed to
the formation of chiral nonracemic enolates, with restricted
rotation around the C–N axis. On the other hand, the
sequential deprotonation/alkylation of amino acid-derived
1,4-benzodiazepine-2-ones, reported by Carlier and co-
workers, permitted the direct preparation of the correspond-
ing C-3-quaternary derivatives with high enantiomeric
11,12
2
examples, most reactions occurring with memory of
chirality involve chemical transformations at the a-carbon
purity.
In this case, a conformationally chiral enolate
was assumed to be the origin of the observed asymmetric
3
of a-amino acid derivatives. Thus, this type of sp –sp
3
12
induction. Stoodley’s group found that the intramolecular
transfer of chirality has made an important contribution to
the stereochemical course of some a-C-substitution
aldol cyclization of 1-(3-oxobutyryl) derivatives of Pro, and
their 4-oxa and 4-thia analogues, as well as other related
cyclization of thioproline derivatives, afforded the corre-
sponding bicyclic derivatives with complete retention of the
Keywords: Memory of chirality; Enantioselective synthesis; b-Lactams;
TADDOL.
Abbreviations: BEMP, 2-tert-butylimino-2-diethylamino-1,3-dimethylper-
1
3–15
configuration.
0
hydro1,3,2- diazaphosphorine; BINOL, 1,1 -bi-2-naphtol; BTPP, tert-butyl-
imino-tri(pyrrolidino)phosphorane; DCM, dichloromethane; MeCN,
acetonitrile; TADDOL, 2,3-O-isopropylidene-1,1,4,4-tetraphenyl-L-
threitol.
In this context, we have reported that memory of chirality
applies in the base-promoted cyclization of some N-benzyl-
N-chloroacetyl amino acids 1 to the corresponding 1,4,4-
*
Corresponding author. Tel.:C34 915622900; fax: C34 915644853;
e-mail: iqmg313@iqm.csic.es
2
3,24
trisubstituted 2-azetidinones 2.
In this asymmetric
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.09.125

M. A. Bonache et al. / Tetrahedron 62 (2006) 130–138
131
a-alkylation reaction, for which it was demonstrated the
importance of the reaction conditions and of the amino
acid substituents, only moderate enantioselectivities were
obtained (ee up to 58%).
2. Results and discussions
2
5
2
6
2.1. Effect of different phase-transfer catalysts on the
cyclization of Phe derivatives
Alternatively, the asymmetric phase-transfer alkylation of
amino acid aldimines is a powerful tool for the construction
of a wide range of natural and unnatural a-amino acids,
including a,a-dialkyl derivatives. Up to three generations
of cynchona-derived catalysts have been successfully
To explore the usefulness of different phase-transfer
catalysts in the synthesis of the amino acid-derived
b-lactams 2, we studied the base-assisted cyclization of
2
7
t
the Phe-O Bu derivative 3 in the presence of the two
2
8–31
cinchonidine related compounds (K)-7 and (K)-8, (K)-
TADDOL (9), (K)-BINOL (10), and the biphenyl catalyst
11 (Scheme 2, Table 1).
exploited in this transformation.
efforts in this field led to the development of C symmetric
Further research
2
3
quaternary ammonium salts, tartaric acid derivatives
2
3
3,34
35
aminophenols (NOBIN), and chiral
(TADDOL),
salen-metal complexes as excellent catalysts to stereo-
selectively prepare a-amino acids (Scheme 1).
Table 1. Influence of the phase-transfer catalyst on the selectivity of the
cyclization of L-Phe-O Bu derivative 3 (MeCN, BTPP)
3
6
t
Entry
Catalyst (10%)
Yield
a
5a:5b
er
ee
(config.)
b
(
%)
1
—
73
76:24
74:26
52 (S)
48 (S)
(
(
K)-7
2
3
85
Scheme 1.
K)-8
69
71
70:30
85:15
40 (S)
70 (S)
Considering that the generation of the quaternary stereo-
genic center in compounds 2 occurs through the intra-
molecular alkylation of an a-amino acid, it would be
attractive to investigate whether the use of the above
indicated phase-transfer catalysts could serve to get and/or
to improve the stereoselectivity in the synthesis of b-lactams
(
K)-9
4
5
(K)-10
79
55
77:23
76:24
54 (S)
52 (S)
2
. Moreover, in those cases where the azetidinone
derivatives are obtained in an enantioselective manner due
to memory of chirality, this study could give an idea if the
memory event might be modulated using such chiral
catalysts. In this paper, we describe the first experimental
evidence for TADDOL as a helpful additive to improve
memory of chirality during the synthesis of our amino acid-
derived b-lactams.
(
K)-11
6
a
Isolated yield.
Measured by chiral HPLC (Column: OL-389, hexane–acetone (96/4),
.5 mL/min).
b
1
Chiral additives 7–9 were selected by their out-
standing results in the synthesis of optically pure amino
acids,
2
8–31,33,34
while BINOL has become among the most
widely used ligands for both stoichiometric and catalytic
asymmetric reaction, including different types of C–C bond
3
7
formation. On the other hand, the L-Phe-containing
biphenyl derivative 11 has recently been introduced as an
effective catalyst for the stereoselective preparation of
3
8
cyanhydrines and the addition of diethylzinc to aldehydes.
As shown in Table 1, the cinchonidine derivatives and the
biphenyl compound 11 proved to be ineffective for
improving the enantioselectivity obtained in the formation
of compound 5, compared to the reaction without chiral
additive. In fact, a 4 and 12% drop in the S-ee was observed
after the addition of catalysts 7 and 8, respectively. It is well
documented that the cinchona-derived catalysts preferably
work at low temperatures, but in our case the reaction
should be carried out above 0 8C to proceed. Only a
marginal increase in the asymmetric induction was achieved
in the reaction with (K)-BINOL, while the tartaric acid
derivative (K)-TADDOL (9), that operates on a wider
Scheme 2.

132
M. A. Bonache et al. / Tetrahedron 62 (2006) 130–138
Table 2. Influence of (K)-TADDOL on the selectivity of the base-promoted cyclization of compounds 3 and 4
1
a
b
Entry
R
Base
Solvent
Additive
10% mol)
Yield (%)
a:b er
ee (S)
(
t
1
2
3
Bu
Bu
Bu
Bu
Bu
Bu
Bu
Bu
BTPP
BTPP
BEMP
BEMP
BTPP
BTPP
BEMP
BEMP
BTPP
BTPP
BEMP
BEMP
BTPP
BTPP
BEMP
BEMP
DCM
DCM
DCM
DCM
MeCN
MeCN
MeCN
MeCN
DCM
DCM
DCM
DCM
MeCN
MeCN
MeCN
MeCN
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
58
73
65
70
73
71
81
76
65
83
69
83
68
87
59
72
74:26
82:18
75:25
77:23
76:24
85:15
76:24
81:19
68:32
80:20
68:32
70:30
67:33
85:15
69:31
79:21
48
64
50
56
52
70
52
62
36
60
36
46
34
70
38
58
t
t
t
t
t
t
t
4
5
6
7
8
9
1
1
1
1
1
1
1
(*)
(*)
(*)
(*)
Me
Me
Me
Me
Me
Me
Me
Me
0
1 (*)
2
3
4
5
6(*)
(
*) Repeated experiences for the regression analysis.
Isolated 2-azetidinone 5 or 6.
Measured by chiral HPLC (Ref. 26).
a
b
temperature range, was able to produce an important
increase in the enantioselectivity of the final b-lactam
interactions, such as additive/ester (x x ), additive/base
3 1
(x x ), and additive/solvent (x x ). These results demon-
3 2 3 4
(
Dee, 18%). Although these improved enantioselectivities
strate that description of the effect of (K)-TADDOL would
not be completed without consideration of its interactions
with these others variables.
could in principle be due to double asymmetric induction,
experiments with (G)-TADDOL and (C)TADDOL
strongly argue against this explanation (vide infra). Based
on these results, further research efforts were concentrated
on studying TADDOL, which showed the best potential.
The obtained coefficients indicate that x (presence of the
3
additive, b Z8.2) is the factor that greatly affects
3
enantioselectivity. Thus, the addition of a 10% mol of
catalyst (K)-9 to the reaction of the L-Phe derivatives 3 and
2.2. Effect of TADDOL on the cyclization of Phe
derivatives
4
resulted always in an increase of the enantiocontrol. This
enhancement is higher for L-Phe-OMe derivative 4 than for
tert-butyl ester analogue 3. This fact is a consequence of the
We have previously shown that the enantioselectivity of the
cyclization of N-chloroacetyl-L-Phe esters is highly depen-
ester influence (b Z4.7) in the enantioselectivity, since in
2
5
dent on the base and solvent used and also on the nature of
1
absence of TADDOL, the tert-butyl derivative 3 cyclizes in
a higher ee than the corresponding methyl ester 4 (Table 2,
compare entries 1, 3, 5, and 7 to entries 9, 11, 13, and 15). It
seems that a maximum limit is reached in both cases in the
presence of the chiral additive (eez60% in DCM, and 70%
in MeCN).
2
6
the ester group. A realistic study on the influence of the
chiral additive TADDOL in this enantioselective reaction
can not ignore these other factors to choose the model
experiments to evaluate. Consequently, we decided to
investigate the effect of (K)-TADDOL on the cyclization
of the L-Phe tert-butyl and methyl ester derivatives 3 and 4
3
9
(
Scheme 2), in the presence of BTPP or BEMP as base and
DCM or MeCN as solvent.
The role of the solvent (x ) is also important, although less
4
than x and x , being the enantioselectivity in MeCN higher
3
1
than in DCM. Nevertheless, its interaction with the additive
x x ) is the minor factor, and no different tendencies were
observed with or without (K)-9.
In order to know the influence of each variable, a careful
experimental study was undertaken using a classical
(
3
4
4
0
factorial experimental design by selection of the appro-
priate 16 reactions for the four desired variables: the nature
of the ester (x ), the base (x ), the presence of catalyst (x )
A rather surprising result is the fact that the base is less
important than their corresponding interaction with the
1
2
3
4
1
and the solvent (x ), being the ee the measured response.
4
additive (b Ob ). Thus, in the presence of (K)-TADDOL,
The experiments and the results obtained are collected in
Table 2. The coefficients of the main effects (b ) and those
32
2
the degree of stereocontrol was reliant on the base (BTPPO
BEMP), while without the chiral additive the influence of
this variable was negligible. Taking into account that the
pK of both bases is similar, it is possible to consider their
a
difference in volume and, hence, a different interaction of
each base with the additive.
i
all second- (b ) and third-order (b ) interactions were
ij
ijk
calculated, and those significant are gathered in Table 3. The
inspection of these coefficients firstly shows that the four
individual variables are statistically important and, sec-
ondly, that there are some noteworthy second-order
Table 3. Significant coefficients of main effects and second-order interactions calculated with the factorial design (22 experiments)
Factor
—
x
1
x
2
x
3
x
4
x
3
x
1
x
3
x
2
3 4
x x
Coefficient
52.2
4.7
K1.7
8.2
2.3
K2.8
K3.0
1.5

M. A. Bonache et al. / Tetrahedron 62 (2006) 130–138
133
Table 4. Importance of the configuration of Phe residue and TADDOL on the selectivity (BTPP)
1
R
a
b
Entry
Starting Phe
Solvent
TADDOL
10% mol)
Yield (%)
a:b er
ee
(
1
2
3
4
5
6
7
8
9
D-4
D-4
D,L-4
D,L-4
L-4
D-4
D,L-4
L-4
L-4
L-4
L-3
L-3
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
MeCN
MeCN
DCM
DCM
MeCN
MeCN
—
(K)-9
—
73
75
52
71
74
85
61
72
71
60
78
79
67
81
31:69
19:81
51:49
51:49
81:19
20:80
51:49
79:21
88:12
86:14
78:22
77:23
81:19
79:21
38 (R)
62 (R)
2 (S)
(K)-9
(C)-9
(C)-9
(C)-9
(G)-9
(C)-9
(G)-9
(C)-9
(G)-9
(C)-9
(G)-9
2 (S)
62 (S)
60 (R)
2 (S)
58 (S)
76 (S)
72 (S)
56 (S)
54 (S)
62 (S)
58 (S)
10
11
12
13
14
t
Bu
t
Bu
Bu
t
L-3
L-3
t
Bu
a
Yield of isolated product.
Measured by chiral HPLC (Ref. 26).
b
2
.3. Influence of Phe and TADDOL configurations
or the racemic TADDOL, especially when acetonitrile was
used as solvent. This confirms again the strong inter-
relationships of the reaction variables and seems to indicate
that the alkyl group of the ester could be, to some extent,
implicated in the interaction with the chiral additive.
After the initial stimulating results, we next explored the
consequence of the use of (K)-TADDOL in the cyclization
of compounds 4 derived from D- and D,L-Phe-OMe
(
Table 4). As expected, in the absence of the chiral additive,
the D-Phe derivative led to the main formation of the
R-enriched b-lactam 6b (entry 1), while the D,L-analogue
afforded compound 6 in almost racemic form (entry 3). Both
results are proofs that the memory of chirality phenomenon
is responsible for the observed selectivity.
The fact that, for a given starting material, (K)-9, (C)-9 and
(G)-9 generated quite similar asymmetric inductions, both
in percentage and sign, is indicative that these chiral
additives are not simply acting as chiral catalysts, for if they
were, matched and mismatched behavior would be
expected. Instead, remarkably, these chiral additives,
independent of configuration and enantiopurity, enhance
the chirality transfer from the starting material to the
product. This particular behavior of TADDOL was also
reproduced by (K)- and (C)-BINOL in the cyclization of
compound L-3, although the total increase in selectivity was
much less important than with TADDOL (Dee (K)-BINOL,
Cyclization of D-4 in the presence of (K)-9 gave a 24% rise
in the 4R enantioselectivity. This increment in the ee, due to
the addition of the chiral additive, was totally comparable to
that obtained for the L-Phe-OMe analogue (in absolute
value, see Table 2, entry 10). On the contrary, the addition
of (K)-TADDOL to the reaction of the racemic D,L-4 had
no influence on the estereoselectivity, indicating that
TADDOL is only effective when the chiral memory event
is operative.
2%; Dee (C)-BINOL, 8%).
2.4. Importance of the quantity of TADDOL and its
structure
These intriguing results prompted us to study the influence
of TADDOL configuration on the cyclization of compounds
An increase of the (K)-9 catalyst led to a significant
enhancement in the ee value (up to 82%), but accompanied
with lower yields of the b-lactam 5, due to significant
O-monoalkylation of (K)-9 with the N-chloroacetyl
derivative 3, to give compound (K)-12 (Table 5). This
latter compound did not contribute in any extent to the high
enantioselectivity observed in the reactions where it is
produced, as subsequently demonstrated by using it as the
only additive (Table 5, entry 5). In standard trials, the
transformation of a 10% (K)-TADDOL into compound
(K)-12 was complete within 2 h, as quantified by HPLC.
Therefore, when the cyclization of compound 3 in the
presence of (K)-9 was stopped after 2 h, the ee of the
b-lactam 5 (80%, Table 5, entry 6) was very close to that
obtained with an equimolecular amount of the additive,
corroborating that catalyst (K)-9 was the only enhancer of
the memory of chirality. The formation of the O-alkylation
product (K)-12 could suggest that TADDOL, in the
presence of the strong organic bases, could work as a chiral
TADDOLate base, but the extent and trend of the
3
and 4 (Table 4). The addition of a 10% mol of catalyst
(
C)-9 to the reaction of the L-4 derivative in DCM produced
a slightly higher asymmetric induction, also favoring the 4S
enantiomer 6a, than that obtained upon addition of the
enantiomeric catalyst (K)-9 (compare entry 10 in Table 2
with entry 5 in Table 4). These outcomes were corroborated
during cyclization of the corresponding D-Phe analogue in
the presence of (K)-9 and (C)-9, providing quite similar
ees in favor of the R-enriched b-lactam 6 (Table 4, entries 2
and 6). More interestingly, an analogous improvement in
selectivity was also achieved when racemic (G)-TADDOL
was used as additive (entry 8), suggesting a minor role for
the additive configuration. Cyclization of L-4 in MeCN and
in the presence of (C)-9 gave a 6% higher increase in the ee
value than reaction with (K)-9, while (G)-9 led to an
intermediate situation (compare entry 14 in Table 2 with
entries 9 and 10 in Table 4). On the contrary to the
cyclization of Phe-OMe derivatives 4, the increase in
t
selectivity during the cyclization of the L-Phe-O Bu
derivative 3 was higher using (K)-9 than its enantiomer

134
M. A. Bonache et al. / Tetrahedron 62 (2006) 130–138
Table 5. Influence of the concentration of (K)-TADDOL on the selectivity of the BTPP-promoted cyclization of compound 3
1
R
a
Entry
Solvent
Additive (% mol)
Yield (%)
a:b er
ee
t
1
2
3
4
5
6
Bu
Bu
Bu
Bu
Bu
Bu
DCM
DCM
MeCN
MeCN
MeCN
MeCN
(K)-9 (50)
(K)-9 (100)
(K)-9 (50)
(K)-9 (100)
(K)-12 (10)
60
46
68
53
84
47
85:15
86:14
91:9
91:9
73:27
90:10
70
72
82
82
46
80
t
t
t
t
t
b
(K)-9 (10)
a
Measured by chiral HPLC (Ref. 26).
Reaction processed after 2 h.
b
asymmetric induction promoted by (K)-9 and (C)-9 seems
to discard this possibility.
described that TADDOLates bearing four 1-naphtyl groups
often show a complete absence of catalytic activity due to
4
3
high steric hindrance. These results seem to suggest that the
phenyl (or 2-naphtyl) groups of TADDOL could be
implicated in direct interactions either with the starting
amino acid or with the enolate intermediate.
Of further interest were the experiments performed with
differently substituted TADDOLs, intended for establishing
the importance of the OH and Ph groups of this catalyst on
the selectivity (Table 6).
Since the stereochemistry of TADDOL is not a determining
factor for the observed selectivity, a structurally related
achiral derivative, diphenylcarbinol 16, was also explored
as additive. However, the cyclization of compound 3 to 5 in
the presence of diphenylcarbinol, far from amplify the
asymmetric induction, led to a dramatic reduction in the
selectivity, suggesting a role for the whole three-dimen-
sional structure of TADDOL.
Dimethylated TADDOL derivatives (K)-13 and (C)-13
failed to induce any additional asymmetry over the usually
obtained by the memory of chirality, demonstrating that
the 1,4-hydroxyl groups were necessary features for the
asymmetric inducing power of these catalysts. The incor-
poration of bulky 1-naphtyl substituents instead of the phenyl
groups of TADDOL, as in (K)-14, resulted in a negligible
increase of the selectivity with respect to the reaction without
additive. In contrast, catalyst (K)-15, with a phenyl to
2-naphtyl interchange, was a quite efficient additive, yielding
to similar ee values than (K)-9 (Table 6, compare entries 1
and 2 with 9 and 10). The disparity between the behavior of
2.5. Effect of TADDOL on the cyclization of different
amino acid derivatives
The capacity of TADDOL to enhance memory of chirality
in the cyclization of other amino acid derivatives was next
examined (Scheme 3).
2
attributed to a completely different spatial arrangement of the
-naphtyl (phenyl-like) and the 1-naphtyl additives could be
4
aromatic groups in these derivatives. In fact, it has been
2
Table 6. Influence of TADDOL structure on the selectivity of the BTPP-promoted cyclization of compound 3
a
b
Entry
Solvent
Additive (10% mol)
Yield (%)
a:b er
ee (S)
1
2
3
4
MeCN
DCM
MeCN
DCM
(K)-9
(K)-9
(C)-9
(C)-9
71
73
67
79
85:15
82:18
81:19
78:22
70
64
62
56
(
(
K)-13
5
6
MeCN
MeCN
79
79
75:25
75:25
50
50
C)-13
(
K)-14
7
8
MeCN
DCM
65
73
79:21
76:24
58
52
(K)-14
(
K)-15
9
MeCN
70
83:17
66
1
1
0
1
DCM
MeCN
(K)-15
16
73
92
79:21
59:41
58
8
a
Isolated compounds.
Measured by chiral HPLC (Ref. 26).
b

M. A. Bonache et al. / Tetrahedron 62 (2006) 130–138
135
Figure 1. Low energy complexes between N-Pmb-N-chloroacetyl-L-Phe-
OMe 4 and (K)-TADDOL (A), and (C)-TADDOL (B).
NCO–HO H-bond and an offset stacking interaction, with
inclined ring moieties, between the phenyl ring of the Phe
side chain and one aromatic group of TADDOL. The
observed arrangements led to a correct positioning of the
Scheme 3.
As highlighted in Table 7, in addition to Phe derivatives,
only Tyr and Trp analogues 17–19 were prone to the
asymmetric influence of TADDOL. From these results, the
improvement in the memory of chirality due to this chiral
additive can be rationalized by postulating a transition-state
model in which particular conformations of either the
starting amino acid or the initially formed enolate, or both,
would be fixed or stabilized through p–p interactions with
the catalyst. This hypothetical transition state could explain
the differences in the increase of selectivity between Phe
Tyr and Trp derivatives [Dee due to TADDOL: 18–36% for
Phe, 36% for Tyr, and 8% for Trp(Boc) and Trp], since the
lower aromatic character of the indole ring in Trp could take
CH Cl substituent to allow the cyclization to take place.
2
This mode of interaction is in good agreement with the
capital importance of the free OH groups in TADDOL and
the need for amino acid side chains that favor p–p contacts.
Alternative scenarios, which postulate the interaction of
TADDOL with the enolate intermediates, could not be
discarded to participate in the production of selectivity.
3. Conclusions
In this paper, we have described the first examples of the
TADDOL-promoted enhancement of the enantioselectivity
due to memory of chirality. The degree of the selectivity
improvement in the intramolecular a-alkylation of N-ben-
zyl-N-chloroacetyl amino acid derivatives was dependent
on the solvent and base used, and most importantly on the
nature of the amino acid side chain, that should be aromatic.
However, the extent and sign of the enantioselectivity in the
b-lactam formation was practically independent on the
configuration of TADDOL. This means that the chiral
additive does not work as a true catalysts, but could be
4
4
account for the minor change in the asymmetric induction.
In agreement with the suggested importance of p–p
interactions, TADDOL was ineffective for aliphatic amino
acid derivatives, independently that the memory of chirality
phenomenon existed previously (Leu, Orn) or not (Ala).
Preliminary molecular modeling studies have revealed that
the starting amino acid derivative 4 was able to establish
very similar interactions with both enantiomers of
TADDOL (Fig. 1). Main contacts are characterized by an
Table 7. Effect of TADDOL in the cyclization of different amino acid derivatives (BTPP, MeCN)
a
b
Entry
Starting Compd.
Additive
10 mol %)
Final Compd.
Yield (%)
a:b er
ee (S)
(
c
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
3
3
4
4
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
—
(K)-9
5
5
6
6
73
71
68
87
68
73
89
90
24
30
75
77
42
27
60
65
76:24
85:15
67:33
85:15
63:37
81:19
71:29
75:25
65:35
69:31
55:45
55:45
78:22
77:23
50:50
50:50
52
70
34
70
26
62
42
50
30
38
10
10
56
54
0
c
d
d
e
e
f
17
17
18
18
19
19
20
20
21
21
22
22
23
23
24
24
25
25
26
26
27
27
28
28
f
f
f
0
1
2
3
4
5
6
g
g
h
h
h
h
0
a
b
c
d
e
f
Yield of isolated compounds.
Measured by chiral HPLC (column: OL-389).
Hexane–acetone (96/4), 1.5 mL/min.
Hexane–EtOH (95/5), 1 mL/min.
Hexane–acetone–EtOH (95/4/1), 1.2 mL/min.
Hexane–acetone (96/4), 1.2 mL/min.
Hexane–EtOH (90/10), 1 mL/min.
g
h
Measured by RP-HPLC from the corresponding dipeptide derivatives (Ref. 26).

136
M. A. Bonache et al. / Tetrahedron 62 (2006) 130–138
0
considered like a memory of chirality enhancer. An
hypothetical mechanism to rationalize the function of
TADDOL in these reactions could involve the formation
of complexes between the starting amino acid derivative, or
its derived enolate intermediate, and the chiral additive,
through H-bonds and p–p interactions.
4.1.1. (2R,3R,1 S)-2,3-O-Isopropyliden-1-O-[N-(p-meth-
0
0
oxybenzyl)-N-[1 -tert-butoxycarbonyl-2 -(phenyl)-ethyl]-
aminocarbonylmethyl-1,1,4,4-tetraphenyl-L-threitol
(12). Eluent: EtOAc–hexane (1/6). Solid. Mp: 86–88 8C
(EtOAc–hexane). HPLC: t Z6.81 min. (A:BZ80:20).
R
C
ESI-MS: 870.4 (MCNa) . [a] K53.29 (c 0.85, CHCl ).
D
3
1
H NMR (300 MHz, CDCl ): d 7.76–6.27 (m, 29H), 4.45 (d,
3
In spite of the new insights into the memory phenomenon
inferred from this paper, whether the use of other chiral
catalysts, and even of non-chiral additives, could result in
more general approaches to control memory of chirality
events remains an elusive goal.
1H, JZ8.0 Hz), 4.33 (m, 1H), 4.23 (d, 1H, JZ8.0 Hz), 3.74
(d, 1H, JZ14.5 Hz) 3.68 (m, 4H), 3.28 (d, 1H, JZ16.6 Hz),
3.20 (m, 2H), 3.09 (d, 1H, JZ14.5 Hz), 1.37 (s, 9H), 0.99 (s,
1
3
3H), 0.83 (s, 3H). C NMR (75 MHz, CDCl ): d 169.02,
168.53, 158.82, 146.68–108.17 (35C), 84.78, 84.70, 82.07,
3
8
2
1.72, 79.99, 63.30, 61.85, 55.19, 50.34, 35.12, 27.90,
7.04, 26.73. Anal. Calcd for C H NO : C, 76.48; H, 6.77;
5
4
57
8
4. Experimental
N, 1.65. Found: C, 76.35; H, 6.50; N, 1.44.
All reagents were of commercial quality. Solvents were
dried and purified by standard methods. Amino acid
derivatives were from commercial sources. Analytical
TLC was performed on aluminium sheets coated with a
4.2. Factorial experimental design
The factors assumed to influence the reaction studied were
chosen on the basis of prior knowledge of their weight on
the enantioselectivity. They were: the nature of the ester
(x ), the base (x ), the presence of catalyst (x ) and the
0
.2 mm layer of silica gel 60 F254. Silica gel 60 (230–400
mesh) was used for column chromatography. Analytical
HPLC uses a Novapak C18 (3.9!150 mm, 4 mm) column,
with a flow rate of 1 mL/min, using a tuneable UV detector
1
2
3
solvent (x ), being the ee the measured response. For these
4
4
four factors, at two levels, 16 experiments (2 ) are necessary
set at 214 nm. Mixtures of MeCN (solvent A) and 0.05%
TFA in H O (solvent B) were used as mobile phase. H
NMR spectra were recorded with a spectrometer operating
to determinate the coefficients of the main effects and all the
possible interactions between two or more factors. Six
experiences, randomly selected (marked in Table 2 with an
asterisk), were repeated in order to determine the
experimental error that resulted to be 1.7% ee. The results
of all reactions (22 experiments) were included in the
1
2
1
3
3
00 MHz, using TMS as internal standard. C NMR spectra
were registered at 75 MHz. Electrospray mass spectra
positive mode) were recorded with a Hewlett Packard
100SD spectrometer.
(
1
regression analysis. The coefficients of main effects (b ) and
i
those of all second- (b ) and third-order (b ) interactions
ij
ijk
Chiral stationary phase was prepared by covalent bonding of
a mixed cellulose derivative (10-undecenoate/3,5-dimethyl-
phenylcarbamate) on allylsilica gel. This procedure was
carried out by radical polymerization in the presence of
were calculated in a first run. Then, only the variables with
statistically significant coefficients (main effects and
second-order x x , x x and x x ) were kept in the model
in a second regression analysis to calculate the new
coefficients. The fitting of this regression analysis was
good (rZ0.987) and the model error (2.1% ee) was not
appreciably different from the experimental error, so we
can be quite confident with the significance of the
coefficients.
3
1
3 2
3 4
4
5
AIBN, without solvent.
Starting N-chloroacetyl derivatives 3, 4, and 16–21 were
2
4,26
prepared as described.
showed the expected analytical and spectroscopic data.
Final b-lactams 5, 6 and 22–27
24,26
Catalysts 8 and dimethylated TADDOL derivatives (K)-13
and (C)-13 were synthesized following previous reported
4.3. Molecular modeling
4
2,46
procedures.
Compound 11 was a generous gift from
Dr. Herrad o´ n. Catalysts (K)-7, (K)-TADDOL (9), (K)-
BINOL (10), (K)-14, and (K)-15 were obtained from
commercial sources.
Complexes between TADDOL derivatives and different
conformations of compound 4 were manually constructed.
The 3D structures of (K)-TADDOL and (C)-TADDOL
were taken from the Cambridge Structural Database (CSD).
The resulting complexes were submitted to simulated
annealing molecular dynamics calculations, using a con-
straint function to maintain the distance between the CO
corresponding to compound 4 and one of the hydroxyl
groups of TADDOL, and/or the distance between the center
of the aromatic ring of the Phe side chain in compound 4 and
the center of one of the aromatic rings of TADDOL.
Moreover, the coordinates of TADDOL were kept fixed.
The simulated annealing strategy consisted of 100 loops of
slow cooling, each one leading to a low energy confor-
mation. Each loop begins by fixing the temperature at
1000 K, followed by 5000 steps of molecular dynamics
(MD). The temperature is then decreased in steps of 100 K
every 50,000 steps, up till the temperature of the
system corresponds approximately to 300 K. The final
4
.1. Synthesis of 2-azetidinones
General procedure. A solution of the corresponding
chloroacetyl derivative (0.4 mmol) and the additive (nor-
mally 10% mol) in the appropriate solvent (1.5 ml) was
treated, at room temperature and under Ar atmosphere, with
the selected base (0. 6 mmol). The reaction was monitored
by TLC or HPLC until complete disappearance of the
starting material. Then, the solution was evaporated,
dissolved in EtOAc, washed with H O, and dried over
2
Na SO . After evaporation, the resulting residue was
2
4
purified on a silica gel column. The obtained 2-azetidinone
was directly evaluated by chiral HPLC, or transformed into
the corresponding dipeptide derivatives, as previously
2
6
described.

M. A. Bonache et al. / Tetrahedron 62 (2006) 130–138
137
conformation obtained at the end of this process was refined
using a conjugate gradient algorithm, and after storage, was
used to start a new simulation at high temperature with a
slow cooling stage. This procedure produced samples of 100
energy-minimized conformations. From these samples, the
lowest energy structures, within 3 Kcal/mol, that were able
to explain the selectivity of the b-lactam formation were
chosen. The molecular dynamic simulations were accom-
plished using the cff91 force field within the program
package (Insight II, Discover, Homology, Biopolymer)
from Molecular Simulations Inc. (San Diego, CA).
13. Brewster, A. G.; Frampton, C. S.; Jayatissa, J.; Mitchell, M. B.;
Stoodley, R. J.; Vohra, S. Chem. Commun. 1998, 299–300.
14. Betts, M. J.; Pritchard, R. G.; Schofield, A.; Stoodley, R. J.;
Vohra, S. J. Chem. Soc., Perkin Trans. 1 1999, 1067–1072.
15. Brewster, A. G.; Jayatissa, J.; Mitchell, M. B.; Schofield, A.;
Stoodley, R. J. Tetrahedron Lett. 2002, 43, 3919–3922.
16. Matsumura, Y.; Shirakawa, Y.; Satoh, Y.; Umino, M.; Tanaka,
T.; Maki, T.; Onomura, O. Org. Lett. 2000, 2, 1689–1691.
17. Wanyoike, G. N.; Onomura, O.; Maki, T.; Matsumura, Y. Org.
Lett. 2002, 4, 1875–1877.
18. Griesbeck, A. G.; Kramer, W.; Lex, J. Angew. Chem., Int. Ed.
2
001, 40, 577–579.
9. Griesbeck, A. G.; Kramer, W.; Lex, J. Synthesis 2001,
159–1166.
1
2
2
2
1
Acknowledgements
0. Griesbeck, A. G.; Kramer, W.; Bartoschek, A.; Schmickler, H.
Org. Lett. 2001, 3, 537–539.
Authors want to express their gratitude to Dr. J. I. Garc ´ı a for
assistance with the regression analysis and for his helpful
suggestions. This work was supported by CICYT (SAF
1. Sauer, S.; Schumacher, A.; Barbosa, F.; Giese, B. Tetrahedron
Lett. 1998, 39, 3685–3688.
2. Giese, B.; Wettstein, P.; Stahelin, C.; Barbosa, F.; Neuburger,
¨
2003-07207-C02). M.A.B. thanks the financial support from
the Ram o´ n Areces Foundation.
M.; Zehnder, M.; Wessig, P. Angew. Chem., Int. Ed. 1999, 38,
2
586–2587.
2
3. Gerona-Navarro, G.; Bonache, M. A.; Herranz, R.; Garc ´ı a-
L o´ pez, M. T.; Gonz a´ lez-Mu n˜ iz, R. Synlett 2000, 1249–1252.
4. Gerona-Navarro, G.; Bonache, M. A.; Herranz, R.; Garc ´ı a-
L o´ pez, M. T.; Gonz a´ lez-Mu n˜ iz, R. J. Org. Chem. 2001, 66,
3538–3547.
2
References and notes
2
5. Bonache, M. A.; Gerona-Navarro, G.; Mart ´ı n-Mart ´ı nez, M.;
´a-L o´ pez, M. T.; L o´ pez, P.; Cativiela, C.; Gonz a´ lez-
Garcı
Mu n˜ iz, R. Synlett 2003, 1007–1011.
26. Bonache, M. A.; Gerona-Navarro, G.; Garc ´ı a-Aparicio, C.;
Al ´ı as, M.; Mart ´ı n-Mart ´ı nez, M.; Garc ´ı a-L o´ pez, M. T.; L o´ pez,
P.; Cativiela, C.; Gonz a´ lez-Mu n˜ iz, R. Tetrahedron: Asym-
metry 2003, 14, 2161–2169.
1
2
. For reviews on memory of chirality, see: (a) Fuji, K.;
Kawabata, T. Chem. Eur. J. 1998, 4, 373–376.
b) Kawabata, T.; Fuji, K. Top. Stereochem. 2003, 23,
(
1
75–205. (c) Zhao, H.; Hsu, D. C.; Carlier, P. R. Synthesis
005, 1–16.
2
3
3
3
. For examples on sp –sp , sp -axial and axial–axial transfer of
chirality, based on the memory principle, but starting from non
amino acidic materials, see: (a) Meyers, A. I.; Wettlaufer,
D. G. J. Am. Chem. Soc. 1984, 106, 1135–1136. (b) Kawabata,
T.; Yahiro, K.; Fuji, K. J. Am. Chem. Soc. 1991, 113,
27. For reviews on the enantioselective synthesis of a-amino acids
by phase-transfer catalysis see: (a) Maruoka, K.; Ooi, T. Chem.
Rev. 2003, 103, 3013–3128. (b) O’Donnell, M. J. Acc. Chem.
Res. 2004, 37, 506–517. (c) Lygo, B.; Andrews, B. I. Acc.
Chem. Res. 2004, 37, 518–525.
9
694–9696. (c) Curran, D. P.; Liu, W.; Chen, C. H.-T. J. Am.
Chem. Soc. 1999, 121, 11012–11013. (d) Buckmelter, A. J.;
Kim, A. I.; Rychnovsky, S. D. J. Am. Chem. Soc. 2000, 122,
28. Griffith, D. L.; O’Donnell, M. J.; Pottorf, R. S.; Scott, W. L.;
Porco, J. A. Tetrahedron Lett. 1997, 38, 8821–8824.
29. Lygo, B.; Crosby, J.; Peterson, J. A. Tetrahedron Lett. 1999,
40, 8671–8674.
9
386–9390. (e) Mori, T.; Saito, H.; Inoue, Y. Chem. Commun.
003, 2302–2303. (f) Eames, J.; Suggate, M. J. Angew. Chem.,
Int. Ed. 2005, 44, 186–189.
2
30. Corey, E. J.; Xu, F.; Noe, M. C. J. Am. Chem. Soc. 1997, 119,
12414–12415.
3
4
5
6
7
8
9
. Kawabata, T.; Wirth, T.; Yahiro, K.; Suzuki, H.; Fuji, K.
J. Am. Chem. Soc. 1994, 116, 10809–10810.
31. Jew, S.-s.; Jeong, B.-S.; Lee, J.-H.; Yoo, M.-S.; Lee, Y.-J.;
Park, B.-s.; Kim, M. G.; Park, H.-g. J. Org. Chem. 2003, 68,
4514–4516.
. Kawabata, T.; Suzuki, H.; Nagae, Y.; Fuji, K. Angew. Chem.,
Int. Ed. 2000, 39, 2155–2157.
. Kawabata, T.; Chen, J.; Suzuki, H.; Nagae, Y.; Kinoshita, T.;
Chancharunee, S.; Fuji, K. Org. Lett. 2000, 2, 3883–3885.
. Kawabata, T.; Kawakami, S.; Fuji, K. Tetrahedron Lett. 2002,
32. Ooi, T.; Takeuchi, M.; Kameda, M.; Maruoka, K. J. Am.
Chem. Soc. 2000, 122, 5228–5229.
33. 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, 851–857.
4
3, 1465–1467.
. Kawabata, T.; Kawakami, S.; Shimada, S.; Fuji, K.
Tetrahedron 2003, 59, 965–974.
. Kawabata, T.; Ozt u¨ rk, O.; Suzuki, H.; Fuji, K. Synthesis 2003,
34. Belokon, Y. N.; Kochetkov, K. A.; Churkina, T. D.;
Ikonnikov, N. S.; Chesnokov, A. A.; Larionov, O. V.; Singh,
I.; Parmar, V. S.; Vyskocil, S.; Kagan, H. B. J. Org. Chem.
2000, 65, 7041–7048.
5
05–508.
. Kawabata, T.; Kawakami, S.; Majumdar, S. J. Am. Chem. Soc.
003, 125, 13012–13013.
2
1
1
1
0. Kawabata, T.; Majumdar, S.; Tsubaki, K.; Monguchi, D. Org.
Biomol. Chem. 2005, 125, 1609–1611.
35. Belokon, Y. N.; Kochetkov, K. A.; Churkina, T. D.;
Ikonnikov, N. S.; Vyskocil, S.; Kagan, H. B. Tetrahedron:
Asymmetry 1999, 10, 1723–1728.
1. Carlier, P. R.; Zhao, H.; DeGuzman, J.; Lam, P. C.-H. J. Am.
Chem. Soc. 2003, 125, 11482–11483.
2. Lam, P. C.-H.; Carlier, P. R. J. Org. Chem. 2005, 70,
36. Belokon, Y.; North, M.; Churkina, T. D.; Ikonnikov, N. S.;
Maleev, V. I. Tetrahedron 2001, 57, 2491–2498.
37. Brunel, J. M. Chem. Rev. 2005, 105, 857–897.
1
530–1538.

1
38
M. A. Bonache et al. / Tetrahedron 62 (2006) 130–138
42. Seebach, D.; Beck, A. K.; Heckle, A. Angew. Chem., Int. Ed.
3
3
8. Ana Montero, Aguado Ph.D. Thesis, Autonoma University,
Madrid, Spain 2004.
9. BTPP: tert-butylimino-tri(pyrrolidino)phosphorane. BEMP:
2001, 40, 92–138.
43. Seebach, D.; Plattner, D. A.; Beck, A. K.; Wang, I. M.; Hunziker,
D.; Petter, W. Helv. Chim. Acta 1992, 75, 2171–2209.
44. Ooi, T.; Takahashi, M.; Doda, K.; Maruoka, K. J. Am. Chem.
Soc. 2002, 124, 7640–7641.
2-tert-butylimino-2-diethylamino-1,3-dimethylperhydro-1,3,2-
diaza-phosphorine.
4
0. Box, G. E. P.; Hunter, J. S. Statistics for Experimenters; Wiley:
New York, 1978.
1. Cativiela, C.; Fraile, J. M.; Garc ´ı a, J. I.; Mayoral, J. A.;
Figueras, F.; Menorval, L. C.; Alonso, P. J. J. Catal. 1992, 137,
45. Minguill o´ n, C.; Franco, P.; Oliveros, L.; L o´ pez, P.
J. Chromatogr. A 1996, 728, 407–414.
4
46. O’Donnell, M. J.; Delgado, F.; Pottorf, R. S. Tetrahedron
1999, 55, 6347–6362.
394–407.