Journal of the American Chemical Society
Communication
formations giving access to a range of useful transforma-
tions.15,16 This approach could be extended to distal γ-
C(sp3)−H functionalization reactions17 and enantioselective
transformations using strained rings.18
used for the C(sp3)−H activation of carboxylic acids (Table
1).16b−d,g,17c,18a Ligands L1−L5 gave low but measurable levels
of de-deuteration (entries 2−6). Ethylenediamine-derived
ligand L6 developed by Yu gave the first encouraging result,
with an overall de-deuteration of 11% (entry 7).18 On the basis
of this observation, we prepared a series of ethylenediamine-
based ligands, varying the structure of the amide. In related
studies ethylenediamine ligands bearing the NHAc-protecting
group proved to facilitate the CMD step in enantioselective
arylations of free carboxylic acids.18 We discovered that the
reactivity for the de-deuteration could be increased signifi-
cantly when changing this group to a more bulky isopropyl
substituent L7 (entry 8). Further systematic variation resulted
in highly active ligands L8 and L9 (entries 9 and 10).
Carboxylic acids and derivatives have also been used as
directing groups to enable the ortho-deuteration of aromatic
substrates.19 This includes a report on the ortho-deuteration of
phenylacetic acids derivatives by Yu, in which the authors also
observed the deuteration of the benzylic positions α to the
carboxylic acid,19b presumably through a deprotonation/
reprotonation mechanism. With carboxylic acids as directing
groups (DG), the only reports for the deuterations of
C(sp3)−H bonds have remained limited to heterogeneously
catalyzed protocols for simple fatty acids, enabling an
unselective deuteration under harsh conditions.8,20 Regiose-
lective C(sp3)−H deuterations of carboxylic acid derivatives
were achieved indirectly via exogenous directing groups,21 but
direct regioselective deuterations of free carboxylic acids have
remained elusive to date.
a
Table 1. Ligand Screening
Carboxylate-directed C(sp3)−H functionalizations are de-
scribed to occur via transition state TS-1 in which the ligand
acts as an internal base for the concerted metalation−
deprotonation (CMD) step (Figure 1B).22 The C−H activated
palladacycle 2 can undergo further functionalization with a
range of electrophilic coupling partners. Mechanistic inves-
tigations on such systems showed through de-deuteration
experiments that under suitable conditions the C−H activation
step is reversible.16g,17a−c
We envisioned a deuteration of aliphatic carboxylic acids
based on this mechanism. Key to such a method would be to
reversibly and efficiently form intermediate 2 while at the same
time avoiding side reactions leading to catalyst or substrate
decomposition (Figure 1C). Considering the correlation
between a deuteration of regular substrates in deuterated
solvents and the reverse de-deuteration of labeled compounds
in regular solvents,23 we reasoned that it should be possible to
optimize de-deuteration processes when ultimately aiming to
develop a deuteration method. This generally applicable
strategy offers substantial advantages. In particular, the
optimization studies could be performed with regular solvents
including 1,1,1,3,3,3-hexafluoropropan-2-ol (HFIP) which we
assumed would be required, due to its well-documented
unique properties in C−H activation processes.24 Screening
would be more robust, because the acidic proton of the
substrate and potential contaminations with proton sources
such as water would exert no detrimental effect. Another
attractive feature is that new catalytic activities could be
uncovered: Since the C−H activation step is separated from
any subsequent transformation, the discovery of catalyst
systems that activate challenging positions is not affected by
the performance of the same catalyst in the subsequent steps of
the model reaction. Such single-step screenings have been used
successfully in the discovery of new photochemical trans-
formations25 and are expected to prove useful in C−H
activation. Finally, taking the kinetic isotope effect (KIE) into
account, catalysts developed in this way should even perform
better when used for deuteration. Importantly, while de-
deuterations have been used to gain mechanistic insights and
drive mechanism-based catalyst improvement,26 to the best of
our knowledge, the proposed screening of reaction conditions
in the reverse reaction has not been reported.
b
b
entry
ligand
yield 4 (%)
de-deuteration (%)
1
2
3
4
5
6
7
8
9
10
none
L1
L2
L3
L4
L5
L6
L7
L8
L9
L9
82
87
91
87
92
91
89
90
93
93
92
<5
<5
<5
<5
<5
<5
11
32
73
80
c
11
68
a
b
Reactions were performed on a 0.1 mmol scale. Yields and degrees
1
of de-deuteration were determined by H-NMR spectroscopy using
c
1,3,5-trimethoxybenzene as internal standard. No Ag2CO3.
Interestingly, the analogous ligand with a simple benzamide
group gave no activity. Presumably, substituents in the 2- and
6-position are required to either force the arene out of the
amide plane, thereby altering the ligand properties, or to
prevent an intramolecular C−H activation leading to catalyst
deactivation. Substitution on the backbone or at the tertiary
amine of the ligands gave no further improvements.27 The
optimized protocol employs catalytic amounts of silver
carbonate. Notably, silver is not essential, although the
reaction proceeds with reduced efficiency in its absence
(entry 11). Silver can thus not be part of the catalyst, but
presumably acts as a Lewis acid accelerating the reaction and/
or as an oxidant preventing catalyst decomposition via Pd(0)
formation.28 A condition-based sensitivity assessment showed
that the method tolerates variations in parameters that
negatively affect many C−H activation methods.27,29
Encouraged by these considerations, we initiated our
investigation by screening ligands classes L1−L6 commonly
We proceeded to investigate the scope of this protocol
(Scheme 1). The carboxylic acid substrates were generally
10896
J. Am. Chem. Soc. 2021, 143, 10895−10901