M. Reinfelds et al. / Tetrahedron Letters 56 (2015) 5882–5885
5883
Table 2
Table 3
a
Alkylation of chiral glycine equivalents by 2a
Alkylation of Williams’ morpholinone by quaternary gramines
Yieldb (%)
O
O
N
Ph
Ph
Entry
Substrate
Product
1
MeSO4-
N+
LiHMDS (2.1 eq)
Li2CuCl4 (0.1 eq)
O
N
O
O
O
R
R
O
O
N
1
4
5
40
N
THF/DMF 6:1
78 °C 1.5 h
N
H
H
-
2a-i
3a-i
NH
NH
Yielda (%)
O
O
Bn
O
O
O
Entry
Gramine
R
Product
Bn
1
2
3
4
5
6
7
8
9
2a
2b
2c
2d
2e
2f
2g
2h
2i
H
2-Me
7-Et
5-OMe
4-OBn
5-OBn
5-F
3a
3b
3c
3d
3e
3f
3g
3h
3i
78
82
75
75
76
80
76
65
45
2
3
N
9a
9b
N
10a
10b
50
22
Ph
Ph
Ph
Ph
O
O
O
O
O
N
N
NH
O
O
6-Br
5-CN
OMe
OMe
N
N
a
4
11
NH
12
0
Isolated yield.
N
N
OMe
OMe
a
Reaction conditions: LiHMDS 2.1 equiv, LiCuCl
.5 h.
Isolated yields.
4
0.1 equiv THF/DMF 6:1, À78 °C,
1
MeSO4-
O
O
b
O
O
N+
LiHMDS
Ph
Ph
N
Li2CuCl4 (0.1 eq)
Ph
Ph
N
O
+
THF/DMF,
- 78 °C, 1.5 h
N
R
N
R
O
O
O
1
2a R = H
3 R = Me
3a R = H 78 %
15 R = Me < 5%
O
O
H
N
1
N
O
O
N
N
14
N
n
NH
6
7
8
n=1
n=2
Figure 2. Reaction mechanism investigation.
Figure 1. Side products from the alkylation of Xu lactone.
The next glycine equivalent examined, Seebach’s oxazolidi-
2
0
none 9a, was subjected to the optimized reaction conditions to
give indole derivative 10a in 50% yield. To test whether the steric
achieved when LiHMDS was used in combination with DMF
Table 1, entry 6). In all cases (Table 1, entries 1–5) the conversion
bulk of the a-Bn group reduced the oxazolidinone reactivity, race-
(
mic oxazolidinone 9b was also tested, however to our disappoint-
ment alkylation of this sterically less hindered substrate proceeded
in only 22% yield.
of starting material 1 was complete and only one diastereomer 3a
1
was detected in the crude reaction mixture (confirmed by H NMR
and LC–MS). However, when only one equivalent of LiHMDS was
used, unreacted 1 remained and 3a was isolated in only 60% yield
Finally, we attempted to alkylate Schöllkopf’s dihydropy-
2
1
razine 11 using gramine 2a under the optimized conditions,
however only trace amounts of the desired product 12 were
detected (Table 2, entry 4).
(
2 4
Table 1, entry 7). In the absence of Li CuCl the yield of 3a was
16,17
slightly reduced (Table 1, entry 8).
After the successful results with Williams’ morpholinone, our
Therefore, none of the tested glycine equivalents provided bet-
ter yields than Williams’ morpholinone 1.
1
8
attention turned to the recently developed Xu lactone 4, which
was synthesized according to literature procedures.19 This was
subjected to the conditions optimized for morpholinone 1 to give
indole derivative 5 in only 40% yield (Table 2, entry 1). During
the course of the reaction, the formation of minor isomer 6 (endo)
in up to 2% yield, as well as double 7 and triple 8 alkylation prod-
ucts were observed (see Fig. 1). The ratio of mono, di, and tri alky-
lated products in the crude reaction mixture were 1:0.2:0.1,
respectively. It is difficult to explain the observed reactivity. Any
speculation regarding the formation of aggregates was avoided
since DMF was used as a co-solvent. Changing the solvent from
DMF to HMPA did not influence the product distribution. The
amounts of bis and tris alkylated products were reduced to nearly
undetectable amounts when 1.1 equiv of base was used during the
alkylation with gramine 2a. In this case, however, unreacted
starting material 4 was present in the reaction mixture. These
results suggested that bis 7 and tris 8 alkylated products originated
from overalkylation of the initial alkylation product 5, not from
pre-formation of gramine dimers (or trimers), which then react
with lactone 4.
Having evaluated various chiral enolates, we next examinated a
2
2
number of quaternary gramine derivatives for the alkylation of
2
3
Williams’ morpholinone. Compounds 3a–g were obtained in
good yields (Table 3, entries 1–8). In all cases only one diastere-
omer was detected in the crude reaction mixture (confirmed by
1
H NMR and LC–MS). The cyano substituted gramine 2i formed
large aggregates after addition to the reaction mixture and as a
result, considerable amounts of unreacted starting material 2i as
well as the dialkylated product were obtained from the reaction
2
4
mixture with 3i being obtained in 45% yield.
1
5,25
As reported,
alkylation of enolates with quaternary grami-
nes proceeds via the 3-methylene-3H-indole 14 intermediate
(Fig. 2). Indeed, when N-methyl quaternary gramine 13 that could
react only by the S 2 mechanism, was applied to the alkylation of
N
Williams’ morpholinone, only a small amount of product 15 was
detected.
The yield of 3a reached 60% when only one equivalent of base
was used (Table 1, entry 7). Conversion of the starting material
was incomplete because the initially prepared enolate could be