Synthesis of N-trifluoromethyl amides from carboxylic acids

Article abstract of DOI:10.1016/j.chempr.2021.07.005

Found in biomolecules, pharmaceuticals, and agrochemicals, amide-containing molecules are ubiquitous in nature, and their derivatization represents a significant methodological goal in fluorine chemistry. Trifluoromethyl amides have emerged as important functional groups frequently found in pharmaceutical compounds. To date, there is no strategy for synthesizing N-trifluoromethyl amides from abundant organic carboxylic acid derivatives, which are ideal starting materials in amide synthesis. Here, we report the synthesis of N-trifluoromethyl amides from carboxylic acid halides and esters under mild conditions via isothiocyanates in the presence of silver fluoride at room temperature. Through this strategy, isothiocyanates are desulfurized with AgF, and then the formed derivative is acylated to afford N-trifluoromethyl amides, including previously inaccessible structures. This method shows broad scope, provides a platform for rapidly generating N-trifluoromethyl amides by virtue of the diversity and availability of both reaction partners, and should find application in the modification of advanced intermediates.

Full text of DOI:10.1016/j.chempr.2021.07.005

ll
Article
Synthesis of N-trifluoromethyl amides from
carboxylic acids
Jianbo Liu, Matthew F.L. Parker,
Sinan Wang, Robert R. Flavell, F.
Dean Toste, David M. Wilson
fdtoste@berkeley.edu (F.D.T.)
david.m.wilson@ucsf.edu (D.M.W.)
Highlights
Synthesis from carboxylic acids
enables access to a great variety
of N–CF3 amides
The process tolerates alkyl
bromide, a-heteroaryl, and
a-carbonyl functional groups
The use of the ligand 2,4,6-
collidine or pyridine is key to the
reaction
Amide-bond formation implies
the simplest retrosynthetic
disconnection of N–CF3 amides
Amide bonds in pharmaceuticals and other applications are generally formed
through amination of carboxylic acid derivatives. However, access to N-
trifluoromethyl amides via a procedure analogous to this most common strategy
for amide-bond formation has yet to be described. Here, the synthesis of N-
trifluoromethyl amides from carboxylic acid derivatives and isothiocyanates is
achieved and markedly expands the diversity of N-trifluoromethyl amides and their
biological and industrial applications, affording previously inaccessible structures
containing alkyl bromide, alpha-heteroaryl, and alpha-carbonyl functional groups.
Liu et al., Chem 7, 2245–2255
August 12, 2021 ª 2021 Published by Elsevier
Inc.
https://doi.org/10.1016/j.chempr.2021.07.005

ll
Article
Synthesis of N-trifluoromethyl
amides from carboxylic acids
Jianbo Liu,1,2 Matthew F.L. Parker,2 Sinan Wang,2 Robert R. Flavell,2 F. Dean Toste,3,
*
and David M. Wilson2,4,
*
SUMMARY
The bigger picture
Found in biomolecules,
pharmaceuticals, and
Found in biomolecules, pharmaceuticals, and agrochemicals, amide-
containing molecules are ubiquitous in nature, and their derivatiza-
tion represents a significant methodological goal in fluorine chemis-
try. Trifluoromethyl amides have emerged as important functional
groups frequently found in pharmaceutical compounds. To date,
there is no strategy for synthesizing N-trifluoromethyl amides
from abundant organic carboxylic acid derivatives, which are ideal
starting materials in amide synthesis. Here, we report the synthesis
of N-trifluoromethyl amides from carboxylic acid halides and esters
under mild conditions via isothiocyanates in the presence of silver
fluoride at room temperature. Through this strategy, isothiocya-
nates are desulfurized with AgF, and then the formed derivative is
acylated to afford N-trifluoromethyl amides, including previously
inaccessible structures. This method shows broad scope, provides
a platform for rapidly generating N-trifluoromethyl amides by virtue
of the diversity and availability of both reaction partners, and should
find application in the modification of advanced intermediates.
agrochemicals, amide-containing
molecules are ubiquitous in
nature. The trifluoromethyl group
has emerged as an important
fluorine-containing functional
group frequently found in
pharmaceutical compounds.
Although elegant methods of
synthesizing N–CF3 amides have
been previously reported, none
have used the simplest
retrosynthetic disconnection,
namely amide-bond formation
through amination of carboxylic
acid derivatives. The new
approach presented here is based
on this disconnection and enables
access to a greater variety of N–
CF3 amides from readily available
carboxylic acids. This strategy can
achieve successive desulfurative
fluorination and acylation of
isothiocyanates in one pot to
obtain N–CF3 amides. Given the
prevalence of amides and the
diversity of available carboxylic
acid derivatives and
INTRODUCTION
The development of efficient and convenient methods for the preparation of amides
has been a central goal in synthetic chemistry because of their prevalence in syn-
thetic chemistry and the life sciences.1–5 The introduction of the trifluoromethyl
group into organic molecules has also been of major interest because this moiety
can significantly alter physical and chemical properties (solubility, lipophilicity,
conformation, pKa, and membrane permeability), metabolic stability, and bioavail-
ability of these molecules in numerous applications.6–8 As relatively new chemical
building blocks, N-trifluoromethyl amides combining these two important structures
have naturally attracted the attention of scientists.9–11 Compared with their N–Me
and N–H amide counterparts, N-trifluoromethyl amides appear to be especially sta-
ble in that they possess the longest C–N bond with minimal restriction on rotation.11
Many N-trifluoromethyl-amide-derived molecules show potential value as materials,
pharmaceuticals, and agrochemicals.12,13 The interest in fluorinated molecules has
also been driven by the unique availability to detect fluorine either via 19F nuclear
magnetic resonance (NMR) or via 18F positron emission tomography. However, until
recently, only a few minimally functionalized N–CF3 amides had been reported by
Rozen and coworkers,9 motivating subsequent interest in wider N–CF3 carbonyl
families.
isothiocyanates, the synthesis of
N–CF3 amides directly from these
components is likely to inspire
numerous applications.
Synthetic access to N-trifluoromethyl amides is presumably limited by the instability
of the parent N–CF3 amine resulting from the n(N)/s*(C–F) electron donation
Chem 7, 2245–2255, August 12, 2021 ª 2021 Published by Elsevier Inc. 2245

ll
Article
facilitating fluorine elimination,14 thus limiting access to coupling methods tradition-
ally employed in amide syntheses. As a result, alternative synthetic strategies have
been developed for N-acyl analogs that circumvent the intermediacy of the N–CF3
amide (Figure 1A). A strategy recently reported by Fang and Li provides an efficient
and chemoselective approach to N–CF3 amides that involves the silver-mediated
electrophilic N-trifluoromethylation of N–H amides.10 This method enabled the che-
moselective N-trifluoromethylation of di- and tri-peptides with retention of configu-
ration. The Schoenebeck group reported a major breakthrough in the field of N–CF3
amide synthesis in the discovery that isothiocyanates can be transformed in a one-
pot procedure with AgF and a carbonyl electrophile to the corresponding N–CF3
carbamoyl fluorides, which subsequently were reacted with Grignard reagents to
furnish the corresponding N-trifluoromethyl amides.11 Because of the reactivity of
Grignard reagents, especially the alkyl ones, the introduction of functional groups
in the carboxylic acid side of amides (R1) was partially limited.
Despite these advances, access to N–CF3 amides via a procedure analogous to the
most common strategy for amide-bond formation, namely coupling an amine with a
carboxylic acid derivative, has yet to be described.15 Amide bonds in pharmaceuti-
cals and other applications are generally formed through amination of carboxylic
acid derivatives (acyl chloride, acid anhydride, activated ester, etc.).1 This discon-
nection would enable access to a greater variety of N–CF3 amides from readily avail-
able carboxylic acids, many of which are found in nature, straightforward to obtain,
stable, and often highly functionalized such that many are produced industrially on a
large scale. Therefore, the use of organic carboxylic acid derivatives to directly
synthesize N–CF3 amide could markedly expand the diversity of N-trifluoromethyl
amides and their biologic and industrial applications.
RESULTS AND DISCUSSION
Investigation of reaction conditions
Initially, we investigated the synthesis of N–CF3 amides through the reactions of an
acyl chloride (PhCH2CH2COCl or PhCOCl) with a stable trifluoromethylated amine,
PhNH(CF3)16,17 (Figure S1A). Whereas the reagent PhNH(CF3) was consumed,
N-trifluoromethyl amides were not detected. Presumably, H–N(R)(CF3) possesses
a much lower activity than non-alpha-fluoro-substituted amines as a result of the
strong electron-withdrawing nature and steric bulk of the trifluoromethyl moiety;
therefore, we surmised that because of the lower reactivity, isolated N-trifluorome-
thylated secondary amines were unlikely to provide general access to N–CF3 amides
via standard amide-bond-forming protocols.
Inspired by the work of the Schoenebeck group, we generated the proposed
Ag–N(R)(CF3) E11 from isothiocyanates R-NCS B and silver fluoride and reacted it
with acyl halide A (prepared directly from the organic carboxylic acid in situ) with the
goal of producing N-trifluoromethyl amide C (Figure 1B). Initially, hydrocinnamoyl
chloride 1A was added to trap the proposed Ag–N(R)(CF3) E generated from phe-
nethyl isothiocyanates 1B in various solvents (Table S1, entries 1–4). Unfortunately,
N–CF3 amide 1 was not detected under any conditions, including acetonitrile, which
had been previously identified as the ideal solvent for the formation of N–CF3 amino-
formyl fluoride.11 An extensive evaluation of the reaction conditions (Table S1, entries
5–10) revealed that the addition of a base that could serve as a ligand to activate the
proposed Ag–N(R)(CF3) E accelerated the desired coupling. Notably, compared with
the use of pyridine (Table S1, entry 5), 2,4,6-collidine (Table S1, entry 10) was found
to nearly double the yield of the desired amide. Moreover, neither strongly basic
1Department of Nuclear Medicine, First Affiliated
Hospital of Sun Yat-sen University, Guangzhou,
Guangdong 510080, P.R. China
2Department of Radiology and Biomedical
Imaging, University of California, San Francisco,
San Francisco, CA 94158, USA
3Department of Chemistry, University of
California, Berkeley, Berkeley, CA 94720, USA
4Lead contact
*Correspondence:
fdtoste@berkeley.edu (F.D.T.),
david.m.wilson@ucsf.edu (D.M.W.)
https://doi.org/10.1016/j.chempr.2021.07.005
2246 Chem 7, 2245–2255, August 12, 2021

ll
Article
Figure 1. Direct synthesis of N-trifluoromethyl amides from carboxylic acid derivatives
(A) Retrosynthetic disconnection of trifluoromethylamides: electrophilic trifluoromethylation of
amides, carboxyamination of nucleophile equivalents, and amide-bond formation (this
manuscript).
(B) Putative mechanism of CF3-amide-bond synthesis via successive fluorination and acylation of
isothiocyanates.
See also Figure 2.
(4-dimethylaminopyridine and Et3N; Table S1, entries 7 and 8) nor weakly basic
(2-fluoropyridine; Table S1, entry 6) ligands were effective. DCM (dichloromethane;
Table S1, entry 13) was identified as a more effective solvent than THF (tetrahydro-
furan; Table S1, entry 10) and MeCN (acetonitrile; Table S1, entry 14). Finally, employ-
ing 5 equiv rather than 4 equiv of AgF (silver fluoride) nearly doubled the yield of amide
1 (96%; Table S1, entry 16), most likely because of the increase in the rate of fluoride
intercepting the proposed intermediate, R–N=CF2 D (Figure 1B).
The amide-bond formation is expected to be affected by the steric hindrance and
the low nucleophilicity of the proposed Ag–N(R)(CF3) E. As a result, when different
acyl halides or esters and isothiocyanates were used as substrates, the optimum con-
ditions for amide-bond formation were somewhat different. Therefore, we devel-
oped several general procedures (reaction conditions a–d, f, h, and i in Schemes
1, 2, and 3) to access the N-trifluoromethyl amides at room temperature (RT) in di-
chloromethane. Ideally, the activity of the carboxylate-derived electrophile and
the proposed Ag–N(R)(CF3) E must be matched to facilitate amide-bond formation.
There are two main variables for the optimization of the latter conditions: (1) the
choice of additives such as pyridine or 2,4,6-collidine and (2) the activity of acyl
Chem 7, 2245–2255, August 12, 2021 2247

ll
Article
Scheme 1. Scope of alkyl isothiocyanates for the synthesis of N–CF3 amides
Reaction conditions: carboxylic acid derivative A (x mmol), alkyl isothiocyanate B (y mmol), AgF (z mmol), 2,4,6-collidine or pyridine (p mmol or p mL),
DCM (s mL), RT, 24 h.
2248 Chem 7, 2245–2255, August 12, 2021

ll
Article
Scheme 1. Continued
aAcyl chloride A and 2,4,6-collidine (3.0 mmol), x:y:z = 0.5:0.7:2.5, s = 1.6.
bAcyl chloride A and pyridine (1.0 mL), x:y:z = 0.7:0.5:2.0, s = 1.0.
cPentafluorophenyl ester A and pyridine (0.4 mL), x:y:z = 0.5:1.0:2.5, s = 1.6.
dAcyl chloride A and pyridine (0.8 mL), x:y:z = 0.9:0.5:2.5, s = 0.4.
eAcyl fluoride A and 2,4,6-collidine (3.0 mmol), x:y:z = 0.5:0.8:2.5, s = 1.6.
fAcyl chloride A and 2,4,6-collidine (3.0 mmol), x:y:z = 0.7:0.5:2.0, s = 1.6.
gAcid chloride A and pyridine (1.0 mL), x:y:z = 0.5:0.7:2.5, s = 1.0.
h2-chloro-pyrimidine-5-carbonyl chloride was added as carboxylic acid derivative A.
halide or esters and the proposed Ag–N(R)(CF3) E nucleophile. A more in-depth dis-
cussion of notable trends observed for these conditions is presented in Table S3.
More specifically, reaction conditions (a) and (b) were consolidated into two general
ones and used for the synthesis of up to 43 and 31 amides, respectively. On one
hand, the typical reaction conditions when acyl halides were used as reactants
were (a), the optimization process of which is shown in Table S1. For pentafluoro-
phenyl esters as reactants, on the other hand, conditions (c) were typically employed
(Table S2).
Substrate scope for the synthesis of N-trifluoromethyl amides
Applying the optimized conditions, we first evaluated the scope of the reaction by
exploring the combination of the alkyl carboxylic acid derivative component with
the more active alkyl isothiocyanate partner (Scheme 1). A wide range of acyl halides
A were prepared in situ from commercially available organic carboxylic acids, re-
flecting their high availability, and used directly in the N-trifluoromethyl amide for-
mation. Aliphatic acyl halide substrates (1–5, 11, 12, and 14–25), including cycloalkyl
acyl halides (15–19), generally performed well, although the yields of the cyclic var-
iants tended to decrease with ring size. Although a range of primary isothiocyanates
performed well in the amide-bond formation, secondary isothiocyanates were
incompatible with the reaction conditions to afford N–CF3 amide, presumably
because of the decreased reactivity resulting from larger R-groups in the proposed
substituted (Ag–N(R)(CF3) E intermediate. The incorporation of an array of functional
groups on both partners, including various halo groups (Cl [2 and 21], Br [3]), a car-
boalkoxy (4 and 20) and an alkenyl group (5 and 22), was well tolerated despite the
possibility that these groups can react with organic silver salts. Moreover, the
method provided access to N-trifluoromethyl alpha-ketoamides (7 and 8),18–20 an
oxalamide (9), an alpha-esteramide (10), and an aryldifluoroamide (11), the latter
of which is a structural unit present in several biologically active compounds.21
The activated acyl derivatives of alpha-ether acids (12 and 13) and trans-cinnamic
acid (6) were also smoothly converted into the corresponding N–CF3 amides.
In order to further evaluate the scope of the reaction beyond aliphatic carboxylic acyl
halides, we examined the various aryl and heteroaryl carboxylic acyl halides as
coupling partners in amide-bond formation. In particular, amides derived from het-
eroaryl carboxylic acids are prevalent in pharmaceuticals, such as avanafil, bortezo-
mib, meloxicam, and raltitrexed,22 and many routes rely on the use of carboxylic acid
starting materials;23,24 however, reports related to the systematic preparation of
N-trifluoromethyl amides are rare. Thus, a general procedure for forming N-trifluor-
omethyl heteroaryl amides—by leveraging activated acyl donors that can be
accessed from heteroaryl carboxylic acids—under mild conditions with simple re-
agents would provide an important tool for synthetic and medicinal chemistry. Initial
studies with benzoyl chlorides produced N-trifluoromethyl amides by using the sil-
ver-fluoride-promoted coupling with alkyl isothiocyanate, although substituting pyr-
idine for 2,4,6-collidine was required for improved yields in some cases. In general,
Chem 7, 2245–2255, August 12, 2021 2249

ll
Article
Scheme 2. Scope of aryl isothiocyanates for the synthesis of N–CF3 amides
Reaction conditions: carboxylic acid derivative A (x mmol), alkyl isothiocyanate B (y mmol), AgF (z mmol), 2,4,6-collidine or pyridine (p mmol or p mL),
DCM (s mL), RT, 24 h.
aAcyl chloride A and 2,4,6-collidine (3.0 mmol), x:y:z = 0.5:0.7:2.5, s = 1.6.
bAcyl chloride A and 2,4,6-collidine (0.6 mL), x:y:z = 0.5:0.7:2.5, s = 0.6.
electron-deficient (hetero)aroyl chlorides gave better yields than those that were
substituted with electron-donating groups, most likely because of the faster relative
rate of reaction between the proposed (Ag–N(R)(CF3) E and the more electrophilic
acyl chlorides. The transformation proved to be very general (Scheme 1), and
numerous functional groups were tolerated, including various halo (27–30 and 42),
carboalkoxy (31 and 71), nitro (32), nitrile (33), carbonyl (34), vinyl (35), sulfonyl
(36), trifluoromethyl (37), trifluoromethylthio (38), trifluoromethoxy (39), and ether
(40 and 41) groups. Multi-substituted arenes, which include halogen-substituted
benzoyl chlorides, were also effective coupling partners (45–50). Importantly, a num-
ber of heteroaroyl chlorides, such as those derived from furan (51 and 52), thiophene
(53 and 54), thiazole (55), pyridine (56), pyrazine (57), pyrimidine (58 and 59), pyrida-
zine (60), benzo[b]furan (61), and quinoline (62–64), were well tolerated (40%~95%
yields). The related acyl chlorides containing halogen substituents at different posi-
tions on the ring maintained excellent results (65–70). Additionally, secondary iso-
thiocyanate (74–76) underwent the desired reaction in reasonable chemical yields
when reacting with electron-deficient aroyl chlorides. The smooth reaction with
Lewis basic heterocycles is particularly noteworthy because the potential of these
heterocycles to act as ligands for the proposed Ag–N(R)(CF3) E, in analogy to pyri-
dine or 2,4,6-collidine, did not impede the reaction.
Having established the scope of the reaction with respect to the acyl halide coupling
component, we turned our efforts toward the thiocyanate coupling partner. More
specifically, in addition to the alkyl thiocyanates described above, aryl isothiocya-
nates reacted smoothly with hydrocinnamoyl chloride to afford the corresponding
N–CF3 amides (Scheme 2). A variety of functional groups, including various halo
(78–81, 85, and 87–89), ether (82 and 88), trifluoromethyl (83), nitro (84), and pyrimi-
dinyl (86) groups, were tolerated. Whereas the carbamimidic fluoride F1 was iso-
lated in 41% yield from the silver-mediated reaction of 4-iodophenyl isothiocyanate
with benzoyl chloride, arylthiocyanates underwent successful amide coupling with
2250 Chem 7, 2245–2255, August 12, 2021

ll
Article
Scheme 3. Scopes of aryl amino acids and chloroformates for the synthesis of N–CF3 amides
Reaction conditions: carboxylic acid derivative A (x mmol), alkyl isothiocyanate B (y mmol), AgF (z mmol), 2,4,6-collidine or pyridine (p mmol or p mL),
DCM (s mL), RT, 24 h.
aAcyl chloride A and 2,4,6-collidine (3.0 mmol), x:y:z = 0.5:0.7:2.5, s = 1.6.
bAcyl chloride A and pyridine (1.0 mL), x:y:z = 0.7:0.5:2.0, s = 1.0.
cPentafluorophenyl ester A and pyridine (0.4 mL), x:y:z = 0.5:1.0:2.5, s = 1.6.
dAcyl chloride A and 2,4,6-collidine (0.6 mL), x:y:z = 0.5:0.7:2.5, s = 0.6.
ePentafluorophenyl ester A and pyridine (0.4 mL), x:y:z = 0.5:1.0:2.5, s = 0.8.
f97 (0.2 mmol), dry EtOH (3.0 mL), NH2NH2 (g, 0.5 mmol, 2.5 equiv), RT, N2, 16 h.
electro-deficient aryl- and heteroaroyl halides. In light of the successful amide cou-
plings with aliphatic acyl chlorides, we were initially surprised by the failed coupling
with benzoyl chloride; however, further studies (see below) suggest that the interme-
diacy of ketene intermediates could account for this unexpected observation.
Pentafluorophenyl esters of N-protecting amino acids, including glycine (90–92),
sarcosine (93–95), beta-alanine (carboxylic acyl chloride as reaction reagent; 96),
alanine (97 and 98), and phenylalanine (99), were also converted to the correspond-
ing N-trifluoromethyl amides (Scheme 3). General protecting groups including
t-butoxycarbonyl (Boc), benzyloxycarbonyl (Cbz), and phthalyl (Pht) were tolerated,
allowing for the generation of alpha-amino amide (100) after deprotection. Reaction
of the generated proposed Ag–N(R)(CF3) E with chloroformates instead of the acyl
chloride enabled the syntheses of the corresponding carbamates. Benzyl (101 and
106), methyl (102), allyl (103), and phenyl chloroformate (105) underwent efficient
transformation (Scheme 3) at RT with both alkyl and aryl isothiocyanates (106).
Mechanistic insights and applications
We note that amide 97 was produced as a racemate (50:50 enantiomeric ratio [er])
from the silver-mediated coupling with the protected enantiopure amino acid pen-
tafluorophenyl ester. Moreover, a crossover experiment between L-alanine deriva-
tive-d 107A and L-phenylalanine derivative 99A showed that deuterium and
hydrogen exchanged between the pentafluorophenyl esters 107A and 99A (Fig-
ure 2A), consistent with racemization of L-alanine amide (97). This information
Chem 7, 2245–2255, August 12, 2021 2251

ll
Article
Figure 2. Crossover experiment, proposed mechanism of ketene intermediates, and applications
(A) Crossover experiment: pentafluorophenyl ester 109A (0.25 mmol, 1.0 equiv) and 100A (0.25 mmol, 1.0 equiv), alkyl isothiocyanate B (1.0 mmol), AgF
(2.5 mmol), pyridine (0.4 mL), DCM (0.8 mL), RT, 24 h.
(B) Proposed mechanism of carboxylic acid derivatives A with alpha-H through the addition of ketene and amine.
(C) Applications include two pharmaceutically relevant molecules, melatonin and flecainide, and N-trifluoromethyl peptide synthesis.
aPentafluorophenyl ester A (0.25 mmol), alkyl isothiocyanate B (2.0 equiv), AgF (5.0 equiv), pyridine (0.2 mL), DCM (0.4 mL), RT, 20 h.
suggests that the a-proton of alkyl carboxylate ester derivatives is easily abstracted
by Ag–N(R)(CF3) E to form intermediate I, which could be reprotonated to return the
racemized starting ester (Figure 2B). However, although base-promoted racemiza-
tion of the chiral pentafluorophenyl esters has been previously reported, rate
constants for this process are low.25 Alternatively, intermediate I could proceed to
products via ketene J.26,27 This pathway accounts for the formation of racemic prod-
uct, and the increased electrophilicity of the ketene intermediate is most likely
responsible for the observed higher reactivity of alkyl acyl chlorides than of benzoyl
chlorides, especially when reacting with weakly nucleophilic Ag–N(R)(CF3) E derived
from aryl isothiocyanates. In light of these findings, we surmise that the proposed
Ag–N(R)(CF3) E might undergo side reactions (Figure S1) other than the targeted
amide-bond formation with the carboxylic acid derivative (Figure 1B). As discussed,
it can act as a base to generate inert H–N(R)(CF3) (Figure S1C) and, in some cases,
concomitant formation of ketene intermediates. Additionally, it can react with its
proposed active precursor R–N=CF2 D to generate byproduct F11 (Figure S1B), as
2252 Chem 7, 2245–2255, August 12, 2021

ll
Article
observed in the attempted coupling of benzoyl chloride. Finally, the proposed Ag–
N(R)(CF3) E could add to the electrophilic C=S bond of the starting isothiocyanates
B, although this pathway seems to be mitigated under the current reaction condi-
tions (Figure S1D).28
In order to further explore the utility of this synthetic methodology, we synthesized
two pharmaceutically relevant molecules by using these reaction conditions (Fig-
ure 2C). Melatonin (5-methoxy-N-acetyl tryptamine) has been widely used in health
supplements and cosmetics.29 The acetylation of the primary amine in the synthesis
of melatonin can be replaced by successive desulfurative fluorination and acylation
of isothiocyanate 109B to successfully yield the N–CF3 amide (109) in one pot.
Similarly, a N–CF3 amide derivative (110) of flecainide,30 a membrane-stabilizing
antiarrhythmic agent, was prepared in 53% yield by desulfurative fluorination
and acylation of isothiocyanate 110B. These results show the potential of this mild
procedure for the complex N–CF3 amides synthesis from carboxylic acid derivatives
and isothiocyanates via amide-bond formation.
Comparison of our method with prior methods
Finally, we provide a clearer and more direct comparison of our method of synthe-
sizing N-trifluoromethyl amides with prior methods9–11 on a sufficient number of
representative examples for better assessment of methodological diversity (Table
S4). The Schoenebeck group reported a strategy11 wherein carbamoyl fluorides
are readily diversified to the corresponding N–CF3 amides in high yields and excel-
lent functional-group tolerance at R2 (Figure 1A); however, R1 is introduced via a
Grignard reagent. Here, we report the direct synthesis of N-trifluoromethyl amides
from carboxylic acid derivatives. The process allows for the synthesis of 20 heteroaryl
examples (51, 52, 55–70, 86, and 87); tolerates alkyl halides (2, 3, and 21), esters (4),
and amides (90–96); and is amenable to the synthesis of products bearing functional
groups at the alpha position (6–13). These examples suggest that our reported sys-
tem complements previous systems by enabling access to N-trifluoromethyl amides
whose synthesis might be challenging via Grignard introduction of R1. However, it is
worth noting that although Schoenenbeck reported excellent retention of stereo-
chemistry at R2, the racemization of the alpha-stereocenter of natural amino acid es-
ters (at R1) is a limitation of our process (97). With respect to the Fang-Li method, our
results also expand the scope of N–CF3 amide structures especially at R2. When R2 is
an aryl group, only one example with 34% yield was synthesized, and compatibility
of heteroaromatic rings might also be limited given that one example is provided.
The product-type distribution of the N–CF3 amides with various strategies is
described in Table S4.
Conclusion
In conclusion, we have described a method for the direct transformation of simple
available acyl halides (or activated esters) into their corresponding N–CF3 amides
via successive fluorination and acylation of isothiocyanates. Key to the success of
the silver-fluoride-mediated amide-bond-forming reaction is the use of the ligand
2,4,6-collidine or pyridine. This approach provides access to a wide range of
different N–CF3 amides with various substituents at both the amine and carboxylic
acid components. Many of these structures are likely to be inaccessible through
previously described methods, especially those products containing alkyl bromide,
alpha-heteroaryl, and alpha-carbonyl functional groups. Moreover, the syntheses
of N-trifluoromethyl derivatives of several medicinally relevant compounds, such
as melatonin and flecainide, were successfully achieved. Most importantly,
this platform enables the formation of N–CF3 amides through the traditional
Chem 7, 2245–2255, August 12, 2021 2253

ll
Article
amide-bond-forming disconnection. Given the prevalence of amides in pharma-
ceutical molecules and the life sciences and the diversity of available carboxylic
acid derivatives and isothiocyanates, the synthesis of N–CF3 amides directly
from these components is likely to inspire numerous applications.
EXPERIMENTAL PROCEDURES
Resource availability
Lead contact
Further information and requests for resources should be directed to and will be ful-
filled by the lead contact, David M. Wilson (david.m.wilson@ucsf.edu).
Materials availability
This study did not generate new unique reagents.
Data and code availability
There is no dataset or code associated with the paper.
METHODS
Full experimental procedures are provided in the supplemental information.
SUPPLEMENTAL INFORMATION
Supplemental information can be found online at https://doi.org/10.1016/j.chempr.
2021.07.005.
ACKNOWLEDGMENTS
The authors acknowledge financial support from the National Institutes of Health
(NIBIB R01 EB024014, R01 EB025985, and R01 EB030897; NIGMS R35 GM118190).
J.L. acknowledges financial support from the Hundred Talents Program of Sun Yat-
sen University.
AUTHOR CONTRIBUTIONS
F.D.T. and D.M.W. supervised the project; J.L. developed the reaction system, con-
ducted the experiments, and analyzed the data; J.L., M.P., S.W., and R.R.F. conduct-
ed the mechanistic studies; and J.L., F.D.T., and D.M.W. wrote the manuscript with
input from all other authors.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: January 15, 2021
Revised: April 20, 2021
Accepted: July 12, 2021
Published: August 3, 2021
REFERENCES
1. Roughley, S.D., and Jordan, A.M. (2011).
The medicinal chemist’s toolbox: an
analysis of reactions used in the pursuit of
drug candidates. J. Med. Chem. 54, 3451–
3479.
3. Bode, J.W. (2006). Emerging methods in
amide- and peptide-bond formation. Curr.
Opin. Drug Discov. Devel. 9, 765–775.
5. Schneider, N., Lowe, D.M., Sayle, R.A., Tarselli,
M.A., and Landrum, G.A. (2016). Big data from
pharmaceutical patents: a computational
analysis of medicinal chemists’ bread and
butter. J. Med. Chem. 59, 4385–4402.
4. Cupido, T., Tulla-Puche, J., Spengler, J.,
and Albericio, F. (2007). The synthesis of
naturally occurring peptides and their
analogs. Curr. Opin. Drug Discov. Devel.
10, 768–783.
2. Pattabiraman, V.R., and Bode, J.W. (2011).
Rethinking amide bond synthesis. Nature 480,
471–479.
6. Muller, K., Faeh, C., and Diederich, F. (2007).
¨
Fluorine in pharmaceuticals: looking beyond
intuition. Science 317, 1881–1886.
2254 Chem 7, 2245–2255, August 12, 2021

ll
Article
7. Purser, S., Moore, P.R., Swallow, S., and
Gouverneur, V. (2008). Fluorine in
medicinal chemistry. Chem. Soc. Rev. 37,
320–330.
16. Umemoto, T., Adachi, K., and Ishihara, S.
(2007). CF3 oxonium salts, O-(trifluoromethyl)
dibenzofuranium salts: in situ synthesis,
racemization in the fragment condensation.
Tetrahedron 65, 7105–7108.
properties, and application as a real CF3+
species reagent. J. Org. Chem. 72, 6905–6917.
24. Marsham, P.R., Hughes, L.R., Jackman, A.L.,
Hayter, A.J., Oldfield, J., Wardleworth, J.M.,
Bishop, J.A.M., O’Connor, B.M., and Calvert,
A.H. (1991). Quinazoline antifolate
8. Isanbor, C., and O’Hagan, D. (2006). Fluorine in
medicinal chemistry: a review of anti-cancer
agents. J. Fluor. Chem. 127, 303–319.
17. van der Werf, A., Hribersek, M., and Selander,
N. (2017). N-trifluoromethylation of
nitrosoarenes with sodium triflinate. Org. Lett.
19, 2374–2377.
thymidylate synthase inhibitors: heterocyclic
benzoyl ring modifications. J. Med. Chem.
34, 1594–1605.
9. Hagooly, Y., Gatenyo, J., Hagooly, A., and
Rozen, S. (2009). Toward the synthesis of the
rare N-(trifluoromethyl)amides and the
N-(difluoromethylene)-N-(trifluoromethyl)
amines [RN(CF3)CF2R0] using BrF3. J. Org.
Chem. 74, 8578–8582.
18. Tanaka, H., Kuroda, A., Marusawa, H.,
Hatanaka, H., Kino, T., Goto, T., Hashimoto, M.,
and Taga, T. (1987). Structure of FK506, a novel
immunosuppressant isolated from
Streptomyces. J. Am. Chem. Soc. 109, 5031–
5033.
25. Hui, K.Y., Holleran, E.M., and Kovacs, J. (1988).
Protected amino acid tetrafluorophenyl esters
for peptide synthesis. Int. J. Pept. Protein Res.
31, 205–211.
10. Zhang, Z., He, J., Zhu, L., Xiao, H., Fang, Y., and
Li, C. (2020). Silver-mediated N-
26. Hatchard, W.R., and Schneider, A.K. (1957).
Dehydrohalogenation products of
hexahydroterephthaloyl chloride. A
bifunctional ketene and a bicyclo[2.2.1]-
heptan-7-one derivative. J. Am. Chem. Soc. 79,
6261–6263.
trifluoromethylation of amides and peptides.
Chin. J. Chem. 38, 924–928.
19. Qian, J., Cuerrier, D., Davies, P.L., Li, Z.,
Powers, J.C., and Campbell, R.L. (2008).
Cocrystal structures of primed side-extending
a-ketoamide inhibitors reveal novel calpain-
inhibitor aromatic interactions. J. Med. Chem.
51, 5264–5270.
11. Scattolin, T., Bouayad-Gervais, S., and
Schoenebeck, F. (2019). Straightforward access
to N-trifluoromethyl amides, carbamates,
thiocarbamates and ureas. Nature 573,
102–107.
27. Schyja, W., Petersen, S., and Luning, U.
¨
(1996). Concave reagents, 21. Selective
acylations of primary and secondary alcohols
by ketenes. Liebigs Ann 1996, 2099–2105.
20. Ovat, A., Li, Z.Z., Hampton, C.Y., Asress, S.A.,
Ferna´ndez, F.M., Glass, J.D., and Powers, J.C.
(2010). Peptidyl a-ketoamides with
12. Vuettner, G., Klaude, E., Oehlmann, L., and
Kaspers, H. (1973). Fungicidal 3-
nucleobases, methylpiperazine, and
dimethylaminoalkyl substituents as calpain
inhibitors. J. Med. Chem. 53, 6326–6336.
(trifluoromethyl)benzo-pyrimidine-2,4-diones
and similar compounds. Ger. Offen., DE
2218362 19731108.
28. Bhattacharjee, J., Das, S., Kottalanka, R.K., and
Panda, T.K. (2016). Hydroamination of
carbodiimides, isocyanates, and
isothiocyanates by a bis(phosphinoselenoic
amide) supported titanium(IV) complex. Dalton
Trans 45, 17824–17832.
21. Arlow, S.I., and Hartwig, J.F. (2016). Synthesis of
aryldifluoroamides by copper-catalyzed
cross-coupling. Angew. Chem. Int. Ed. 55,
4567–4572.
13. Sahu, K.K., Ravichandran, V., Mourya, V.K., and
Agrawal, R.K. (2007). QSAR analysis of caffeoyl
naphthalene sulfonamide derivatives as HIV-1
integrase inhibitors. Med. Chem. Res. 15,
418–430.
29. Prabhakar, C., Kumar, N.V., Reddy, M.R.,
Sarma, M.R., and Reddy, G.O. (1999). Process
research and development of melatonin. Org.
Process Res. Dev. 3, 155–160.
22. Friis, S.D., Skrydstrup, T., and Buchwald,
S.L. (2014). Mild Pd-catalyzed
14. Kloeter, G., and Seppelt, K. (1979).
Trifluoromethanol (CF3OH) and
trifluoromethylamine (CF3NH2). J. Am. Chem.
Soc. 101, 347–349.
aminocarbonylation of (hetero)aryl bromides
with a palladacycle precatalyst. Org. Lett.
16, 4296–4299.
30. Paolini, E., Stronati, G., Guerra, F., and
Capucci, A. (2019). Flecainide:
15. Valeur, E., and Bradley, M. (2009). Amide bond
formation: beyond the myth of coupling
reagents. Chem. Soc. Rev. 38, 606–631.
23. Ivanov, A.S., Zhalnina, A.A., and Shishkov, S.V.
(2009). A convergent approach to synthesis of
bortezomib: the use of TBTU suppresses
electrophysiological properties,
clinical indications, and practical aspects.
Pharmacol. Res. 148, 104443–104452.
Chem 7, 2245–2255, August 12, 2021 2255