Angewandte
Communications
Chemie
Table 1: Optimization for the synthesis of chiral pyroglutamic acid esters 4.[a]
and displayed good catalytic activity and excellent
selectivity in biomimetic Mannich reaction of glyci-
nate (Table 1, entry 2).[8] This is probably because the
too stable a-amino carbanion intermediate pro-
moted by 5a during the catalysis is not nucleophilic
enough for the 1,4-conjugated addition to 2a.
Replacement of the primary amide side chain of
the catalyst 5b with a secondary amide (5c) led to an
obvious improvement in reaction yield (Table 1,
entry 4 vs. 3). Further catalyst screening showed
that chiral pyridoxal 5d was the best choice among
the catalysts 5a–k examined in terms of activity and
enantioselectivity (Table 1, entries 5 and 8 vs. 2–4, 6–
7 and 9–13). Further condition investigations indi-
cated that acetonitrile was the best solvent (Table 1,
entry 8 vs. 5 and 14–15) and DBU was the base of
choice (Table 1, entry 8 vs. 19–21). Lewis acid was
a necessary additive (Table 1, entry 8 vs. 16) and
LiOTf was the most active for the reaction (Table 1,
entry 8 vs. 17–18). Interestingly, temperature influ-
enced the activity of the reaction, but displayed little
impact on the enantioselectivity of the products 4a
(Table 1, entry 8 vs. 22 and 23). The reaction was
chosen to carry out at 408C.
Under the optimal conditions, the substrate
scope was then investigated for the pyridoxal-cata-
lyzed direct synthesis of chiral pyroglutamic acid
esters 4 (Table 2). Phenyl (for 4b) and various
substituted phenyl (for 4a and 4c–m) a,b-unsatu-
rated esters all smoothly underwent asymmetric 1,4-
conjugated addition and subsequent lactamization to
give chiral pyroglutamic acid esters 4a–m in good
yields with low diastereoselectivities but high enan-
tioselectivities for both of the trans- and cis-diaste-
reomers. The 2-substituted substrates such as methyl
Entry Cat.
Conditions
Yield [%][b] trans:cis[c] ee [%][d]
1
–
LiOTf, DBU, THF, 408C
LiOTf, DBU, THF, 408C
LiOTf, DBU, THF, 408C
LiOTf, DBU, THF, 408C
LiOTf, DBU, THF, 408C
LiOTf, DBU, THF, 408C
LiOTf, DBU, THF, 408C
LiOTf, DBU, CH3CN, 408C
LiOTf, DBU, CH3CN, 408C
LiOTf, DBU, CH3CN, 408C
LiOTf, DBU, CH3CN, 408C
LiOTf, DBU, CH3CN, 408C
LiOTf, DBU, CH3CN, 408C
LiOTf, DBU, CHCl3, 408C
LiOTf, DBU, DMF, 408C
DBU, CH3CN, 408C
0
0
–
–
–
–
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
5a
5b
5c
5d
5 f
5g
5d
5e
5h
5i
47
77
85
60
88
90
58
70
89
71
46
60
26
0
1.2:1
1:1
À75/À82
77/87
82/85
74/85
80/84
86/91
84/89
86/89
À87/À90
35/41
21/À15
82/85
87/91
–
1.5:1
1.1:1
1.1:1
1.1:1
1.1:1
1.2:1
1.1:1
1.2:1
1.3:1
1.2:1
1.2:1
–
5j
5k
5d
5d
5d
5d
5d
5d
5d
5d
5d
Zn(OTf)2, DBU, CH3CN, 408C
LiBF4, DBU, CH3CN, 408C
LiOTf, CH3CN, 408C
LiOTf, Et3N, CH3CN, 408C
LiOTf, LiOH, CH3CN, 408C
LiOTf, DBU, CH3CN, 208C
0
–
–
59
0
1:1
–
86/92
–
0
–
–
trace
39
71
58
47
–
1:1
1.4:1
1.2:1
1.1:1
–
85/91
87/91
86/92
86/92
5d LiOTf, DBU, CH3CN, 508C, 48 h
5d LiOTf, DBU, CH3CN, 508C, 30 h
5d LiOTf, DBU, CH3CN, 508C, 15 h
[a] All reactions were carried out with glycinate 1a (0.15 mmol), 2a (0.10 mmol), 5
(0.010 mmol), LiOTf (0.040 mmol) and DBU (0.10 mmol) in solvent (0.30 mL) at
408C for 48 h unless otherwise stated. [b] Isolated yields based on 2a. [c] The trans/
cis ratios were determined by HPLC analysis. [d] The ee values were determined by
chiral HPLC analysis.
(E)-3-(2-fluorophenyl)acrylate (for 4k) and methyl (E)-3-(2-
bromophenyl)acrylate (for 4l) gave somewhat lower reaction
yields likely due to steric hindrance. The electronic property
of the substituted phenyl groups seemed to have little
influence on the enantioselectivity. Naphthyl (for 4n) and
heteroaromatic (for 4o–r) a,b-unsaturated esters both were
effective substrates for the transformation, providing the
corresponding chiral pyroglutamic acid esters 4n–r in 43–
96% yields with similar selectivities. Alkynyl and alkyl a,b-
unsaturated esters underwent the transformation to produce
chiral pyroglutamic acid esters 4s and 4t in 45% and 52%
yields, respectively. Disubstituted a,b-unsaturated ester
methyl (E)-2-methyl-3-phenylacrylate was less reactive for
the reaction likely due to steric hindrance, to give a pair of
diastereomers 4u with the 3-phenyl and 4-methyl groups on
the same side of the pyrrolidinone ring in a 14% yield with
excellent enantioselectivities. The trans/cis configurations of
Scheme 1. Chiral pyridoxals 5a–k examined.
as expected, producing the desired chiral pyroglutamic acid
ester 4a in a 47% yield with moderate enantioselectivities for
both of the diastereomers (Table 1, entry 3). Although the
diastereoselectivity was very low (trans/cis 1.2:1), fortunately
the big polarity difference between the two diastereomers
makes it easy to get them separated by column chromatog-
raphy. The pyridoxal catalyst is crucial for the reaction. No
desired addition products were observed in the absence of
pyridoxals 5 (Table 1, entry 1). To our surprise, N-methyl
chiral pyridoxal 5a was totally ineffective for the reaction
although it has very strong electron-withdrawing capability
1
the products 4a–u were determined by H-1H NOESY (see
Supporting Information). The absolute configurations for
trans-4a and cis-4p were, respectively assigned as (2R,3S) and
(2S,3S) based on X-ray analysis (Figure 2).[11]
While a precise mechanism awaits further studies, a plau-
sible pathway has been proposed for the reaction (Sche-
me 2a). Chiral pyridoxal 5d condenses with glycinate 1a to
Angew. Chem. Int. Ed. 2021, 60, 10588 –10592
ꢀ 2021 Wiley-VCH GmbH