Angewandte
Communications
Chemie
diazoethers in Pd-catalysis.[10] To assess the feasibility of such
a species affecting the reaction outcome, we separately
prepared the corresponding diazoether 3 and studied its
cross-coupling reactions under both Stille and Suzuki con-
ditions in MeOH, MeCN, and toluene (Scheme 2c). Regard-
less of the solvent employed, the Pd-catalyzed cross coupling
with phenylstannane was found to be ineffective. Interest-
ingly, however, the diazoether showed high efficiency in the
coupling with phenylboronic acid, giving the coupling product
in 77% yield in acetonitrile and 47% yield in MeOH.
On the other hand, when we subjected diazoether 3 to
a stoichiometric amount of Pd2(dba)3 (in toluene), we did not
observe any consumption of 3, which suggests that direct
reaction of the Pd-catalyst with the diazoether may not
actually be feasible.
We hypothesized that the efficient reaction under Suzuki
conditions might therefore stem from the acidic nature of the
boronic acid and a potential in situ “deprotection” to the
diazonium salt.[9c] To examine this possibility, we repeated the
experiment of diazoether 3 under Stille coupling conditions,
but this time added a Lewis acid to the reaction (Scheme 2d).
We hypothesized that a Lewis acid might enable a similar
“deprotection” to release the “free” diazonium salt, which
should then react even with the stannane. Pleasingly, the use
of the Lewis acid additive, such as BF3·OEt2, indeed
promoted the Stille cross-coupling and led to notable
conversion of the diazoether 3 to the cross-coupled product
2 (see Scheme 2d). Similarly, our separate subjection of
diazoether 3 to HBF4·OEt2 (1.0 equiv) in acetonitrile resulted
in clean conversion to diazonium salt 1 within the time
required to record a quantitative 1H NMR spectrum (i.e.
6 minutes) at room temperature (Scheme 2d). These data
strongly indicate that the diazoether is not reactive directly in
Pd2(dba)3-catalyzed couplings at room temperature; it rather
functions as a reservoir of the aryldiazonium salt when
exposed to acidic (e.g. arylboronic acid) or Lewis acidic (e.g.
BF3·OEt2 or HBF4·OEt2) conditions.[11]
than oxidative addition of a diazoether at the required Ar N
bond (DDG° > 28 kcalmolÀ1).
À
These data suggest that aryldiazonium salts are highly
reactive with Pd-catalysts via a four-membered s-bond meta-
thesis mechanism. They may dynamically interconvert to
diazoethers in alcoholic solvents under basic conditions, and
were shown above to be unreactive towards the Pd-catalyst.
Their mechanistic role is hence that of an inhibitor under
basic conditions, which would be consistent with the non-
productive effect of base in the cross-coupling reactions of
aryl diazonium salts in MeOH. On the other hand, we
identified that diazoethers can also readily release the
diazonium salt in the presence of acid or Lewis acid additives.
While these results could potentially explain the detri-
mental effects of base in MeOH,[16] the formation of
a diazoether is less likely under “neutral” and base-free
couplings, and therefore does not provide a rationale for the
ineffectiveness of the Stille coupling of 1 in MeOH (see
Scheme 2a). Indeed, when we monitored a solution of
aryldiazonium salt 1 in MeOH at room temperature in the
presence or absence of Bu3SnPh over 24 hours, we saw no
indication of diazoether formation.
We hypothesized that following oxidative addition of the
diazonium salt to Pd0, and subsequent extrusion of N2,
a putative cationic PdII intermediate might more rapidly
react with the solvent MeOH than with the transmetalating
agent. The solvent induced reactivity difference could thus
potentially also arise from differences in reactivities of the
transmetalating agents with a resulting PdII–OMe species. For
Pd-catalyzed couplings of specialized ortho-substituted and in
situ made aryldiazonium salts, Felpin and co-workers have
previously proposed that PdII–OMe might form as an
intermediate in reactions in MeOH.[6a] However, these
suggestions relied primarily on mass spectrometry data;
there was no direct demonstration of a transmetalation
being possible. Instead, computational data was presented
that suggest a prohibitively high barrier for transmetalation of
PdII–OMe by ArB(OH)2 of DG° > 40 kcalmolÀ1,[6a] which
would suggest that PdII–OMe, if indeed formed, should be
unreactive.
To gain unambiguous insight, we set out to synthesize
PdII–alkoxy complex 4 (Scheme 3), which was chosen as it
contains no b hydrogens, and therefore excludes the possi-
bility for the b-H elimination side-reaction to occur.
The complex had previously been reported by Hartwig
and co-workers, who showed that the bulky phosphine helps
to avoid dimerization and maintain a mononuclear T-shaped
complex as confirmed by X-ray crystallography.[17] It was
shown that the complex does not undergo reductive elimi-
nation until its decomposition at 808C; it is therefore well
suited for our intention to study transmetalation.
With complex 4 in hand, we subsequently tested the
transmetalation with PhB(OH)2 and PhSnBu3 in toluene and
acetonitrile (MeOH would lead to exchange in 4 and
ambiguous side-reactions). In the context of transmetalation
of PdII complexes, recent years have seen a significant
increase in understanding, especially in the context of
Suzuki couplings: the key pre-transmetalation PdII intermedi-
ate was identified as a boronic acid adduct, containing a Pd-
In line with the conclusion that the diazoether is likely
primarily an inhibitor under neutral or basic conditions, the
À
oxidative addition to the C N bond of a diazoether was
calculated to proceed with rather high activation free energy
barrier (DG° = 31.0 and 29.7 kcalmolÀ1, respectively for the
E- and Z-isomer, see Scheme 2e, right).[12] In fact, we found
À
that the alternative initial addition to the distant N O bond
would be energetically preferred (DG° = 20.7 and 22.8 kcal
molÀ1, respectively for the E- and Z-isomer).[13]
To put these calculated barriers into perspective, we also
examined the activation barriers for the oxidative addition to
the “free” aryl diazonium salt 1. The vast majority of
literature quotes a three membered transition state, similar
to the geometry commonly considered for oxidative addition
of aryl halides to Pd0,[14] as preferred mechanism for oxidative
addition of diazonium salts to Pd0.[15] However, we found
a four-membered TS significantly favored (DDG° = 11.4 kcal
molÀ1, see Scheme 2e, left). It should be noted that this is
formally a s-bond metathesis-like process, and thus no change
À
in oxidation state occurs during C N bond breaking. Overall,
the calculations suggest that oxidative addition of a diazonium
salt to Pd0 occurs much more readily (DG° = 8.9 kcalmolÀ1)
Angew. Chem. Int. Ed. 2021, 60, 7007 –7012
ꢀ 2021 Wiley-VCH GmbH