342
N. Hidalgo et al.
9.3.2 Polarized heterobimetallic compounds
In some cases, polar M–M bonds exhibit reactivity that is reminiscent of FLP systems.
At variance with main group frustrated pairs, the integrity of the M–M bond may
remain virtually intact during small molecule activation events, in a manner that
could be understood as traditional heterobimetallic activation. However, it is also
possible that the monometallic fragments may coexist in solution due to the lability
of the M–M bond, thus enabling FLP-type activation pathways in the same fashion as
thermally induced FLPs [101]. The latter situation is more facile in unsupported heterobimetallic compounds, that is, those in which the M–M bond is the sole interaction
holding the two metallic fragments together. Thus, apart from some selected examples, we will here focus on unsupported polarized heterobimetallic entities, while
other bimetallic species containing bridging ligands will not be discussed. Moreover,
the presence of bridging ligands usually entails additional complications regarding
mechanistic aspects during small molecule activation that blurs the connection with
FLPs.
The analogous reactivity between FLPs and polarized heterobimetallic complexes
can be drawn by looking at early examples of unsupported systems. In a paradigmatic example, Cutler showed that complexes [Cp(CO) 2 M–Zr(Cl)Cp 2 ] (M = Fe,
Ru) react with CO 2 to yield the corresponding bimetallocarboxylates [Cp(CO) 2 M(η
1 -C:η
2 -O,O’)Zr(Cl)Cp 2 ]. The resemblance to FLP systems is obvious, although the
analogy could not be delineated at that time. The bimetallocarboxylates are stabilized by push–pull interactions derived from the Lewis basic group 8 compound and
the electrophilic zirconium fragment [155]. Although a traditional bimetallic mechanism involving the insertion of CO 2 into the M–M bond was favored, an alternative
pathway through dissociation of the bimetallic compound into monometallic fragments followed by concerted trapping of CO 2 —as a thermally induced TMFLP—
could not be ruled out. Subsequent reports further support the aforementioned
analogy of the reactivity between polar M–M bonds and FLPs [156–168].
More recently the use of metal-only Lewis adducts of [Pt(P
t Bu 3 ) 2 ] (38) has been
described in the context of bond activation (Scheme 9.22). In a first report, Jamali
demonstrated the capacity of [(P
t Bu 3 ) 2 Pt → Cu(NCMe) n ] (39) to activate an O–H
bond of water (by using wet acetone as a solvent) to generate a cationic Pt(II) hydride
and copper hydroxide (Scheme 9.22a) [169]. The origin of the hydride ligand was
corroborated by using D 2 O. Although the mechanism could not be unambiguously
determined, a cooperative pathway that implies the bimetallic adduct seems more
likely. In the same vein, the group of Campos demonstrated that a similar architecture
based on silver, more precisely [(P
t Bu 3 ) 2 Pt → AgNTf 2 ] (40), revealed a rich reactivity toward the activation of X–H (X = H, C, N, O) bonds (Scheme 9.22b) [170].
The metal-only Lewis pair 40 can be easily prepared by dissolving an equimolar
mixture of its parent monometallic precursors ([Pt(P
t Bu 3 ) 2 ] (38) and AgNTf 2 ) in
common organic solvents. The cooperative reactivity of this pair strikingly contrasts
that of its monometallic fragments. For instance, while neither [Pt(P
t Bu 3 ) 2 ] (38)
nor AgNTf 2 react with H 2 , compound 40 produces a Pt(II)/Ag(I) heterobimetallic
N. Hidalgo et al.
9.3.2 Polarized heterobimetallic compounds
In some cases, polar M–M bonds exhibit reactivity that is reminiscent of FLP systems.
At variance with main group frustrated pairs, the integrity of the M–M bond may
remain virtually intact during small molecule activation events, in a manner that
could be understood as traditional heterobimetallic activation. However, it is also
possible that the monometallic fragments may coexist in solution due to the lability
of the M–M bond, thus enabling FLP-type activation pathways in the same fashion as
thermally induced FLPs [101]. The latter situation is more facile in unsupported heterobimetallic compounds, that is, those in which the M–M bond is the sole interaction
holding the two metallic fragments together. Thus, apart from some selected examples, we will here focus on unsupported polarized heterobimetallic entities, while
other bimetallic species containing bridging ligands will not be discussed. Moreover,
the presence of bridging ligands usually entails additional complications regarding
mechanistic aspects during small molecule activation that blurs the connection with
FLPs.
The analogous reactivity between FLPs and polarized heterobimetallic complexes
can be drawn by looking at early examples of unsupported systems. In a paradigmatic example, Cutler showed that complexes [Cp(CO) 2 M–Zr(Cl)Cp 2 ] (M = Fe,
Ru) react with CO 2 to yield the corresponding bimetallocarboxylates [Cp(CO) 2 M(η
1 -C:η
2 -O,O’)Zr(Cl)Cp 2 ]. The resemblance to FLP systems is obvious, although the
analogy could not be delineated at that time. The bimetallocarboxylates are stabilized by push–pull interactions derived from the Lewis basic group 8 compound and
the electrophilic zirconium fragment [155]. Although a traditional bimetallic mechanism involving the insertion of CO 2 into the M–M bond was favored, an alternative
pathway through dissociation of the bimetallic compound into monometallic fragments followed by concerted trapping of CO 2 —as a thermally induced TMFLP—
could not be ruled out. Subsequent reports further support the aforementioned
analogy of the reactivity between polar M–M bonds and FLPs [156–168].
More recently the use of metal-only Lewis adducts of [Pt(P
t Bu 3 ) 2 ] (38) has been
described in the context of bond activation (Scheme 9.22). In a first report, Jamali
demonstrated the capacity of [(P
t Bu 3 ) 2 Pt → Cu(NCMe) n ] (39) to activate an O–H
bond of water (by using wet acetone as a solvent) to generate a cationic Pt(II) hydride
and copper hydroxide (Scheme 9.22a) [169]. The origin of the hydride ligand was
corroborated by using D 2 O. Although the mechanism could not be unambiguously
determined, a cooperative pathway that implies the bimetallic adduct seems more
likely. In the same vein, the group of Campos demonstrated that a similar architecture
based on silver, more precisely [(P
t Bu 3 ) 2 Pt → AgNTf 2 ] (40), revealed a rich reactivity toward the activation of X–H (X = H, C, N, O) bonds (Scheme 9.22b) [170].
The metal-only Lewis pair 40 can be easily prepared by dissolving an equimolar
mixture of its parent monometallic precursors ([Pt(P
t Bu 3 ) 2 ] (38) and AgNTf 2 ) in
common organic solvents. The cooperative reactivity of this pair strikingly contrasts
that of its monometallic fragments. For instance, while neither [Pt(P
t Bu 3 ) 2 ] (38)
nor AgNTf 2 react with H 2 , compound 40 produces a Pt(II)/Ag(I) heterobimetallic