Organometallics
Article
In particular, a mechanistic model should explain how
bidentate ligands can accommodate the formation of the key
[Pd−O−B] and [Pd−O−B]′ species, both of which are
necessary for the TM to occur according to the mechanism
reported in Scheme 1. This issue has not been addressed
systematically in the literature. Denmark and co-workers
observed a slightly reduced transmetalation rate when dppf
was employed instead of PiPr3 and conjectured the need of
generating a coordinatively unsaturated intermediate,5o whose
nature could not be firmly established. A theoretical study
focusing on the S-M reaction promoted by complexes of the
model ligand H2PCH2CH2PH2 investigated only an associative
mechanism, for which very high energy barriers were
computed.6f
These two reports and the frequent use of diphosphine
ligands in S-M cross-couplings prompted us to investigate the
B-to-Pd transmetalation in the case of diphosphines. We
observed that the oxidation of dppf, a diphosphine widely used
in S-M reactions applied to the total synthesis of natural
products,1 could take place under conditions mimicking a
typical catalytic reaction, yielding the monoxide dppfO. Our
results highlight the potentially crucial role of this in situ
generated species in diphosphine-mediated S-M reactions
(Scheme 2).8
which corresponds to a difference in activation energy of
approximately 3 kcal mol−1. These observations indicate that
dppf strongly inhibits B-to-Pd transmetalation. The TM turns
out to be the second elementary step of the S-M reaction to be
inhibited by excess dppf, as it has been shown that extra
diphosphine also inhibits OA by hampering the formation of
the reactive 14-electron complex [Pd0(dppf)].11
To rationalize these kinetic results, we estimated the energy
barriers of the TM by DFT calculations (see the computational
S6). The slower TM rate with dppf in comparison to that with
PPh3 could possibly be due to a dissociative mechanism, in
contrast with the working hypothesis previously formulated by
Huang et al.6f Indeed, as first demonstrated by Goossen and
Thiel,6c the TM with PPh3-ligated Pd(II) requires partial
phosphine decoordination and takes place via a four-centered
transition state involving the concerted formation of a Pd−C
bond and cleavage of Pd−B bonds. Similar behavior is
predicted for dppf (Figure 1C and Figure S6).
With the complex [Pd−O−B] as the starting point (A), the
cleavage of one P−Pd bond can be assisted by one OH of the
boronate moiety to form complex B. This release is
endothermic (+12.8 kcal mol−1) and almost entropically
neutral (+9.8 cal mol−1 K−1), leading to an overall endergonic
process. For comparison, the same process involving PPh3 lies
2.6 kcal mol−1 lower in energy (+7.3 kcal mol−1), driven by the
strong positive entropic contribution (+59.3 cal mol−1 K−1).
Nonetheless, complex B cannot directly take part in TM since
the phenyl moiety on the boron center is too far from the Pd
center (dC−Pd = 3.42 Å). Therefore, prior to TM a ligand
exchange between the OH and Ph linked to the boron atom is
required, leading to the formation of complex C. The two pre-
TM complexes C-cis (+12.8 kcal mol−1) and C-trans (+15.7
kcal mol−1) can be formed depending on the relative position
of the two aromatic rings with respect to the Pd center. For
clarity, in the main text and figures we will refer only to the
most stable cis conformer, while all data corresponding to the
the case of dppf, both isomers are 4−5 kcal mol−1 higher in
free energy in comparison to the PPh3 analogues. Finally, both
cis and trans transition states were optimized, lying at 24.5 and
27.5 kcal mol−1, respectively. The energy barrier for phosphine
decoordination directly affects these transition states. In the
case of PPh3, the most favorable TS-cis was localized at +21.9
kcal mol−1, i.e. about 3 kcal mol−1 lower in comparison to dppf,
corresponding roughly to a factor of 102 on the kinetics of the
reaction.13 Both experimental and theoretical studies thus
point toward a slower transmetalation rate when diphosphine
ligands are used. However, when the formation of the coupling
product Ar-Ph was monitored in the absence of added dppf,
the reaction proceeded more quickly, and it was essentially
complete after 30 min (Figure 1B, black curve). In the latter
case, the kinetic curve of formation of Ar-Ph displayed an
induction period, which is either typical of an autocatalytic
reaction or hints at the in situ generation of an active species
from a less reactive precursor.10 The induction period varies
from nearly 1 h at low base concentration to a few seconds at
high base concentration (Figure S23). This is in agreement
with the instantaneous reaction reported by Denmark and co-
workers,5n as the TM was studied starting from complex 3 with
1 equiv of boronic acid (corresponding to [OH−]/[PhB-
(OH)2] = 1). Consistent with the concentration profiles
(Figure 2C), this induction period probably results from the
Scheme 2. Hypothesis on the Tole of dppfO in S-M
Reactions Proposed in This Paper
RESULTS AND DISCUSSION
■
The oxidative addition complex cis-[PdII(Ar)Br(dppf)] (2, Ar
= 4-F-C6H4) was prepared by ligand exchange between trans-
[PdII(Ar)Br(PPh3)2] (1) and dppf (Figure 1A) to study TM
involving PhB(OH)2 using tetrabutylammonium hydroxide
(TBAOH) as a base.
Inhibiting Effect of Extra Diphosphine on the TM. We
first investigated the effect of excess ligand on the B-to-Pd TM
starting from either the OA complex 1 or 2 and different
amounts of added PPh3 or dppf. The ratio [OH−]/[PhB-
(OH)2] ≈ 0.6 (6 equiv/10 equiv) previously reported to
maximize the TM rate in the case of PPh3-ligated complexes
was used.5g The formation of the coupling product Ar-Ph was
monitored by 19F{1H} NMR spectroscopy (Figure 1B).9
In the case of complex 1 in the presence of additional PPh3
(2 equiv), the coupling reaction was almost complete after 10
min (Figure S5a). Formation of Ar-Ph followed a first-order
law with an apparent rate constant of kapp = 3.3 × 10−3 s−1.5g
No intermediate could be detected in this case, in agreement
with TM being rate determining.5g In stark contrast, in the case
of complex 2, in the presence of 1 equiv of dppf (Figure 1B,
green curve), TM was very slow and only 20% conversion was
observed after 90 min. Under these conditions, the kinetics
could be fitted by a first-order rate law and the apparent rate
constant was estimated to be kapp = 3.2 × 10−5 s−1 (Figure
S5d). The ratio between the two rate constants is about 100,
1121
Organometallics 2021, 40, 1120−1128