Organic Letters
Letter
in mixtures of compounds favoring the O−CF2H products.
The pyridone substrates were prefunctionalized into (2-
pyridyl)acetamides, thereby negating the possibility of forming
O-difluoromethyl products and selectively furnishing the N−
CF2H compounds. The products could then be deprotected to
yield N-(difluoromethyl)pyridones. The requisite protection
and deprotection steps and long reaction times leave room for
improvement. Furthermore, only 4- and 5-substituted sub-
strates could be converted into the desired N-difluoromethyl-
2-pyridone products.39 In 2018, Ma and co-workers performed
an O-difluoromethylation of 2-pyridones using BrCF2COOEt,
which suffers from long reaction times (12 h). Also, N-
difluoromethylation was not explored.40 Herein, we describe a
fast, effective, and tunable approach to selectively obtain N-
and O-difluoromethylated 2-pyridones in one pot.
The facile tautomerism of 2-pyridones, wherein the pyridone
form is in equilibrium with the hydroxypyridine form, has been
well-described in the literature.41−43 Evidence has shown that
temperature plays a role in the relative populations of the
tautomers.42 As shown by Forlani and co-workers, at 45 °C in
a polar solvent system, the hydroxypyridine form is favored,
and at 20 °C the pyridone form is favored.44 On the basis of
these considerations, we reasoned that elevated temperatures
would favor the hydroxypyridine form, allowing for selective O-
difluoromethylation. Likewise, lower temperatures would favor
the pyridone form, allowing for selective N-difluoromethyla-
tion. Thus, our optimization was based on the following
mechanistic hypothesis (Scheme 1). The deprotonation of
TMSCF2Br in triglyme at −15 °C using NaOtBu as the base
and activator (Table 1). It is important to note that dropwise
addition of TMSCF2Br is required for the efficient synthesis of
6a.
a
Table 1. Optimization Experiments for 5a and 6a
b
base (equiv)
Solvent
T (°C)
5a/6a (%)
c
c
c
1
2
KOH (2.2)
ACN
ACN
ACN
ACN
ACN
ACN
ACN
triglyme
60
60
60
0
0
0
68/6
44/54
99/1
68/2
46/13
2/57
2/63
5/84
K2CO3 (1.1)
Na2CO3 (1.1)
Na2CO3 (1.1)
KOtBu (2.2)
NaOtBu (2.2)
NaOtBu (2.2)
NaOtBu (2.2)
3
4
5
6
7
8
d
d
d
d
d
−15
−15
a
b
Reactions performed at 0.5 mmol scale. Yields determined by 19F
NMR using fluorobenzene as internal standard. 0.4 M concentration.
0.25 M concentration.
c
d
The model substrate 4a gave 88% of 5a and 71% of 6a under
methods A and B, respectively (Scheme 2). Alkoxy-substituted
substrates 4b and 4c afforded 5b and 5c in excellent yields.
Methyl-2-pyridones 4d and 4e were isolated in 55% and 46%
yields, respectively. Compounds 6d and 6e were also
synthesized by method B in 42% and 39% yields, respectively.
Volatility of these compounds is likely responsible for the
reduced isolated yields. 6-Methyl-2-pyridone (4f) only
generated 5f under method A and was unable to produce 6f
(see Scheme 3). The brominated 2-pyridones 4g and 4h
furnished 5g and 5h in good yields. Compound 4h afforded 6h
in 38% isolated yield under method B. The strongly electron-
withdrawing −CF3 group was also compatible with both
methods (5i and 6i). Pyridones with electron-donating
substituents (4a−4f) gave higher yields than those with
electron-withdrawing groups (4g−4i) with both methods.
Large-scale reactions (gram scale, 5 mmol) were performed on
substrate 4b with methods A and B (87% and 47% isolated
yields, respectively). A disubstituted pyridone 4j was also
explored. Related isoquinolinones 4k, 4l, and 4m were
converted to 5k, 5l, and 5m in good to excellent yields,
while 6k, 6l, and 6m were afforded in moderate yields.
Similarly, difluoromethylated quinolinones 5n and 5o were
obtained in good yields. However, both 4n and 4o gave very
low yields of the N−CF2H products likely due to the sterics of
the benzo-fused system, which may deter nucleophilic attack.
Similarly, 4s furnished 5s in 75% yield (method A), whereas
only a trace amount of N−CF2H product was formed under
method B. Quinoline-2-thiol 4p produced 5p as the major
product under both methods due to the enhanced nucleophil-
icity of S−. Method A tolerated important functional groups,
providing ester (5q) and amide (5r) (albeit in modest yield).
Neither 6q nor 6r was obtained from method B. Heteroaryl-2-
pyridones 4t−4x afforded 5t−5x in good to excellent yields,
and 6t−6x in moderate to good yields. Notably, the vinyl
group on substrate 4t was unreactive toward the difluor-
ocarbene generated under our conditions, despite their known
[2 + 1] cycloaddition.19,20
Scheme 1. Mechanistic Hypothesis
these structures produces form A (aryloxy form) and form B
(amide form), respectively. It can be expected that the −OH
form (hydroxypyridine) would have a lower pKa than the
−NH form (pyridone). This may warrant the need of a
stronger base for the formation of form B, and a milder base
for form A. Given that forms A and B are in equilibrium, strict
temperature control would be necessary to limit the
interconversion. Subsequent reaction with difluorocarbene
affords the corresponding O−CF2H and N−CF2H products.
Initial trials with TMSCF3 were unsuccessful, so TMSCF2Br
was employed as the :CF2 source. Over the course of our
optimization (see SI for a detailed discussion), we found that
5a could be produced selectively in near-quantitative yield
from 4a when reacted with 1.2 equiv of TMSCF2Br in
acetonitrile at 60 °C using Na2CO3 as the base and activator.
Subsequent changes to the reaction conditions flipped the
chemoselectivity of the reaction. Compound 6a was obtained
in 84% yield from the reaction of 4a with 1.2 equiv of
B
Org. Lett. XXXX, XXX, XXX−XXX