M. Yu et al. / Tetrahedron Letters 55 (2014) 5591–5594
5593
Table 3
Table 4
Scope of heterocyclic ring substitution pattern
Scope on the alkyl groups at the arylic position
O
Ar
O
Ar
NaH (1.5 equiv)
DMF, rt, 3.0 hr
O
H
N
R
N
N
NaH (1.5 equiv)
DMF, rt, 3.0 hr
O
N
R
H
Br
2d, 2m-s
+
N
N
Br
1b
3d, 3m-s
1b
2d, t-x
3d, t-x
Entry
1
Heterobenzylic Bromide
Ar
Product/yielda (%)
a
Entry
Heterobenzylic bromide
R
Product/yield (%)
2d
N
3d/85
1
2
3
4
5
2t
2d
2u
2v
–Me
–Et
3t/89
3d/85
3u/69
3v/<2b
3w/37
–nPr
–iPr
–Ph
H3C
N
2
3
4
2m
2n
2o
3m/82
2w
F3C
N
N c
Br
6
2x
3x/62
3n/78
3o/87
a
b
c
Br
N
Isolated yield.
Yield based on GC–MS analysis of crude reaction mixture.
The complete structure (instead of R) of 2x is as shown.
CF3
Table 4 summarizes the effects from the alkyl groups (–R) at the
benzylic position of the heterobenzylic bromides. It had been clear
that the reaction efficiency gradually declined with increased steric
bulkiness of these groups (2t < 2d < 2u < 2v). Reactions with the
bromide bearing isopropyl group at this position (2v) led only to
a trace amount of 3v. The phenyl substituted bromide 2w provided
the desired product 3w in 37% yield. Also interesting is the fact that
the constrained 8-bromo-5,6,7,8-tetrahydroquinoline bromide 2x
was fairly effective as well (62%).
The scope of lactam is illustrated by commercially available
substrates 1a–g as listed in Table 5. Lactams in 5- to 7-member
ring size (1a, 1b, and 1c) demonstrated equally strong reactivity
toward 2d. Azetidin-2-one, however, has not been successful for
5
2p
3p/77
N
N
6
7
2q
2r
3q/31
3r/22
CH3
N
Br
iPr
8
2s
3s/84
N
S
2d so far under the established conditions. The
v-lactam substrate
was also tested in multiple substitution contexts. Introducing a
methyl group immediately next to the amide motif at either side
(1d and 1e) did not markedly impair the reactivity. Moreover, as
indicated by the high yields from 1d and 1e (82% and 85%, respec-
tively), increased steric bulkiness adjacent to the carbonyl element
did not result in a negative impact on the reaction either. These
data therefore forecast that a comprehensive collection of lactams
would be suitable for the discussed transformation.
a
Isolated yield.
these species as well. These results suggested that electron defi-
ciency on the aromatic ring benefits the electrophilicity of the bro-
mides in general, and that there is no obvious correlation of
reactivity versus heteroatom position.
Studies on the bromides were extended to the substitution
pattern around the heterocycles (Table 3). As exemplified by
2m–p, small alkyl (e.g., –CH3) and electron withdrawing groups
(e.g., –CF3, and –Br) were introduced at multiple positions around
the 2-pyridyl ring. These modifications demonstrated good com-
patibility (71–87%) to the reaction. Electron donating groups
(e.g., –OMe) at the 5- or 6-position of the 2-pyridyl ring were also
tolerated in reaction with 4 (data not included). Ring substitution
adjacent to the alkylation site, as illustrated by 2q and 2r, was,
however, significantly unfavorable. We attribute it to the excessive
steric congestion that resulted from these ortho-modifications. It is
noteworthy that the methoxy group at these positions led to no
desired product but complete decomposition of the bromides
under the reaction condition. It is likely that the methoxy group
under this setting engages in a neighboring group participation
effect on the benzylic position, thus disrupting the N-alkylation
process. Finally, reaction of 1b with 2-(4-isopropyl-2-thiazol-
yl)propyl bromide 2s resulted in 84% yield. This result, compared
to the data on 2j in Table 2, predicts good toleration of ring mod-
ifications for secondary five-membered heterobenzylic bromides
as well.
The secondary heterobenzylic bromides in this Letter were syn-
thesized following either of the two general procedures as outlined
in Scheme 3. The bromides 2d, 2i, and 2j (Table 2) were prepared
through direct bromination of the commercially available ArCH2Et
with N-bromosuccinimide (NBS) in the presence of azobisisobuty-
ronitrile (AIBN). Similarly, reactions of 5,6,7,8-tetrahydroquinoline
provided 2x. Bromides 2e–h and 2k–w (Tables 2–4) were synthe-
sized from the appropriate heterobenzylic carboxaldehydes via
Grignard reaction, followed by bromination of the resulting alco-
hols with CBr4 and PPh3.
A general experimental procedure for this reaction is as follows:
To an oven dried 25 mL round flask charged with stir bar under N2
atmosphere was added 1b (3.0 mmol, 1.0 equiv) followed by DMF
(6.0 mL). The reaction flask was placed into a 0 °C ice bath. To the
solution under N2 was added NaH (3.9 mmol, 1.3 equiv, 60%
dispersed in mineral oil). After 20 min, to the solution was added
2d (3.6 equiv, 1.2 equiv). The reaction flask was then removed from
the ice bath and allowed to stir at ambient temperature for
1.0–15.0 h. Reaction aliquot samples were taken for LCMS analysis
to monitor the reaction to completion. The reaction mixture was
slowly poured into a saturated aqueous NH4Cl solution (5 mL)