Organic Letters
Letter
Scheme 6. Proposed Mechanism of the Suzuki−Miyaura Reaction of MIDA Acylboronates
were too broad to be analyzed. Instead, a series of control
reactions was conducted and monitored by 11B NMR
spectroscopy. The reaction of 1a and 2a under standard
conditions in DMF was monitored over 17 h (Table 2, entry
1). After 1 h, a 11B NMR spectrum consisted of a signal at 5.5
ppm (relative to BF3·Et2O) corresponding to 1a and a new
resonance at 0.47 ppm. The intensity of the latter continued to
rise over the course of the reaction, accompanied by the
disappearance of the resonance of 1a within 11 h. When 1a
was mixed with 1.5 equiv of Cu(OAc)2 and 3.0 equiv of
Cs2CO3 in DMF at 60 °C in the absence of Pd catalyst, the
same signal at 0.47 ppm in the 11B NMR spectrum was
observed (entry 2). Further analysis was required to probe the
intermediacy of the species with the distinct resonance at 0.47
ppm that appeared in the course of both control reactions.
Additional experiments were performed to understand the
appeared after 1 h and its intensity continued to increase, while
the peak corresponding to 1a decreased. Gratifyingly, once 1a
was consumed (∼11 h), the resonance at 0.47 ppm began to
disappear. After 16 h, complete consumption of I was observed
confirming the intermediacy of the species.
We then turned our attention to detecting the hypothesized
intermediacy of an acyl palladium complex. The progress of a
stoichiometric reaction of 1a, 2a, and Pd precatalyst was
monitored by 31P NMR spectrometry over 16 h (Scheme 5).
After 15 min, the signals at 41.9 ppm, which corresponded to
XPhos oxide, and 27.4 ppm appeared. The latter was assigned
to the oxidative addition palladium complex III, which
exhibited a chemical shift consistent with previous literature
reports.15 After 1 h, a new signal at 25.9 ppm appeared and its
intensity continued to rise while the resonance of III slowly
disappeared. This new signal, upfield of 1.5 ppm from that of
complex III, was assigned to the acyl palladium species IV. A
similar upfield shift for analogous carbonylated complexes was
observed by the Moser16 and Yamamoto17 groups in Pd-
catalyzed carbonylation reactions with carbon monoxide.
Moreover, IV was detected by HRMS (ESI+) under
stoichiometric conditions with the expected isotopic distribu-
Based on the experimental results, a plausible reaction
mechanism shown in Scheme 6 is proposed. According to this
scheme, acylboronate 1 reacts with Cu(OAc)2 and Cs2CO3 to
generate intermediate I, which was observed by 11B NMR
spectrometry, and 2 equiv of CsOAc. Further transmetalation
affords acyl copper species II. The Pd(0) active catalyst is
generated from the base-mediated reductive elimination from
the Pd(II) precatalyst. Oxidative addition with aryl bromide
affords intermediate III. Subsequent transmetalation with II or
directly with I gives acyl palladium complex IV. Further
reductive elimination provides the desired ketone 3 and
regenerates the active catalyst.
reaction mechanism. No conversion of 1a was observed by 11
B
NMR spectrometry when the reaction was conducted with
only Cu(OAc)2, demonstrating that a base was necessary for
the reaction (entry 3). Cu(II) was also found to play a key role,
as no reaction was observed in the absence of Cu(OAc)2
(entry 4), with only Cs2CO3 (entry 5) or when KOAc was
used as the acetate source (entry 6). Since CsOAc is
potentially generated in situ, we conducted two control
reactions of 1a with only CsOAc (entry 7) and with CsOAc
and Cu(OAc)2 (entry 8) as additives. No reaction was
observed in either experiment. The presence of an acetate
anion was necessary for the formation of the species with the
signal at 0.47 ppm (entries 9−11).
The 11B NMR resonance in the area between 9 and −7 ppm
is typical of tetracoordinated boron compounds bearing −OR
groups.12 A tetracoordinated acylboron intermediate with the
11B NMR chemical shift of 2.32 ppm was recently observed by
Mankad and co-workers in the copper-catalyzed carbonylative
borylation of alkyl halides, as tricoordinated pinacol-acylboron
coordinated with 1 equiv of LiOtBu.4g Bode and co-workers
originally proposed that hydrolysis of MIDA acylboronate
leads to the highly reactive boron-ligated intermediate,8a which
undergoes either ligation with O-Me hydroxylamines or rapid
protodeboronation in the presence of water.9 Based on the
results of the control reactions, we propose that a ligand
exchange between the Cu(II) additive and a base promotes
complexation13 of an acetate anion to activate 1a. The
observed intermediate with the 11B NMR signal at 0.47 ppm
likely corresponds to the tetracoordinated boron species I,
where X is either the acetate anion and/or the MIDA ligand
with a decoordinated nitrogen atom.8d,14
In summary, we have developed a new Pd-catalyzed cross-
coupling reaction between aryl (pseudo)halides and MIDA
acylboronates that serve as acyl anion equivalents. The mild
reaction conditions were found to be compatible with various
aliphatic and aromatic acylboronates. 11B and 31P NMR
mechanistic studies, as well as mass spectrometry, revealed that
the tetracoordinated boron and acyl palladium(II) species are
possible reaction intermediates. Further exploration of boron-
based acyl anion equivalents in chemical synthesis as well as
studies aimed at defining broader utility for the ArC(O)B-
(OH)2 class of molecules are underway in our laboratory.
To assess whether or not I is the reactive intermediate in the
reaction, a substoichiometric amount of 1a was subjected to
the standard conditions (Scheme 4). The signal at 0.47 ppm
3297
Org. Lett. 2021, 23, 3294−3299