Organic Letters
Letter
a
(CF3SO3CF3) is a good source of fluorophosgene, generated in
situ by fragmentation of OCF3 anions.11 Most recently, the
Schoenebeck group reported a novel method for the
preparation of N-trifluoromethyl carbamoyl fluorides using
isothiocyanates as substrates and AgOCF3 generated in situ
from the reaction between silver fluoride and bis-
(trichloromethyl) carbonate.12 One year later, the same
group reported a practical approach to access AgOCF3 and
its reaction with secondary amines to afford aryl or
alkylcarbamoyl fluorides.13 Although the reagents are still
based on the in situ generation of poisonous difluorophosgene,
these two methods improved the convenience of access to
carbamoyl fluorides with a broad substrate scope and mild
reaction conditions. In 2019, Onida and Tlili developed an
exceptionally practical method using CO2 activation strategy
that does not rely on OCF2 precursors, in which the N−
C(O)−F bonds were constructed via a three component
reaction using secondary amines, carbon dioxide, and
(diethylamino)sulfur trifluoride (DAST) in the presence of
N,N-dimethylaminopyridine (DMAP).14
Table 1. Reaction Optimization
Herein, we propose a new, cost-efficient, and reliable
synthetic route to provide carbamoyl fluorides from α-
oximinoamides 3. We recently demonstrated the concept of
fluorinative C−C bond cleavage of activated ketones; the key
to this operation was the use of DAST, which plays a dual role
as an oxime activator and nucleophilic fluoride donor.15,16 We
thus hypothesized that the fluorinative C−C bond cleavage of
activated amides 3 enables the synthesis of carbamoyl fluorides
4 (Scheme 1b). However, the decreased electrophilicity of
amides compared to ketones was a concern in this trans-
formation; there were no existing reports of a reaction between
tertiary amides and DAST derivatives. In this respect, the low
reactivity of the amide functionality might be a selection factor
between the fluorinative Beckmann fragmentation (route A)
and Beckmann rearrangement (route B). While not obvious,
we believe that the pathway selection is likely dependent on
the geometry of the oximes and the structural and electronic
properties of the substrates 3.
To prepare α-oximinoamides 3, three types of pyruvate
oximes (2a−c) were considered as available coupling partners
with amines 1. When employed in amide bond formation,
compound 2a requires additional coupling reagents to activate
the inert acid functionality, which usually accompanies
unpleasant byproducts. In contrast, the enhanced reactivity
of activated esters such as 2b and 2c was expected to allow a
direct coupling reaction with amines. Although 2c has a better
leaving group, its use has been limited because it requires
multiple steps for preparation. Moreover, the phenol by-
product of the amidation step is less atom economical and less
environmentally benign compared to methanol, which is
generated when using 2b. Therefore, 2b was chosen for the
amide synthesis. Note that it is bench-stable and available on a
multigram scale from inexpensive methyl pyruvate. During the
amidation reaction tests using piperidine 1a and methyl
pyruvate oxime 2b as model substrates,17 the use of Lewis acid
catalysts including ZrCl4, Nb(OEt)5, Ti(OiPr)4, and Zr-
(OtBu)4 was found to lead to excellent conversion into α-
oximinoamide 3a.18
a
Determined by 1H NMR using 1,3,5-trimethoxybenzene as the
b
c
internal standard. Isolated yield. Piperidine-1-carboxylic anhydride
was obtained in 33% yield.
comparable to that of DCM (entries 2, 3, and 4). In addition,
no Beckmann rearrangement product 5a was observed.
However, when a polar aprotic solvent such as DMF was
used, the desired product was not obtained (entry 5). Next,
deoxyfluorination reagents were evaluated. Interestingly, sulfur-
based reagents such as DAST and bis(2-methoxyethyl)-
aminosulfur trifluoride (Deoxo-Fluor) were highly efficient
with good selectivity, whereas a reagent system employing
XtalFluor-E and 3HF-Et3N exhibited no conversion (entries 6
and 7). Consequently, HF-Et3N was considered not to be a
relevant nucleophilic fluoride source for the substrate 3a,
indicating that amide-based substrates require a more reactive
fluoride source than that used in the activated ketones.19 When
3HF-Et3N was replaced with sodium fluoride,20 4a was
obtained in 50% yield along with piperidine-1-carboxylic
anhydride as a side product (entry 8). Non-sulfur-based
reagent such as Ishikawa’s reagent also worked smoothly to
provide 4a in excellent yield (entry 9). The reaction was
completed with similar efficiency in 10 min even at a lower
concentration (entry 10).
Having completed the reaction optimization, we surveyed
the scope of the amines (Scheme 2). Piperidine derivatives
3a−c were smoothly converted to the corresponding
carbamoyl fluorides 4a−c in good yields. The morpholine-
containing substrate 3d also afforded the desired product 4d in
high yield. However, piperazine substrates, 3e (with an acid-
sensitive Boc group) and 3f (with a basic tertiary amine/acid-
sensitive aryl olefin) exhibited reduced efficiency. However, the
yields of the two substrates could be increased by lowering the
concentration. This indicates that the intermolecular reaction
After determining a simple and scalable synthesis for 3a, we
commenced optimization of the designed reaction (Table 1).
To our delight, simple addition of DAST into a solution of 3a
in DCM gave the desired product 4a in high yield (entry 1).
The use of other solvents such as DCE, THF, and CH3CN was
5395
Org. Lett. 2021, 23, 5394−5399
Lim, Hee Nam
