Isomers of Aryldiazenido Complexes of Re
Organometallics, Vol. 15, No. 1, 1996 335
methyl ligands are probably small and comparable with
the interaction with the carbonyl ligand. For this
reason, the populations of the two conformers a and b
(Scheme 2a) are roughly the same. We do not know
whether a or b corresponds to the slightly more popu-
lated conformer in reality. For the phosphorus com-
plexes 4 and 5, we consider that the phosphine and
phosphite ligands are sterically more bulky than the
chloride ligand and that the preferred conformer is a
in each case (Scheme 2b). Finally, for the bulky PPh3
ligand complex, the only observed conformer is expected
to be 7a .
cationic complexes by comparison with the neutral
chloro compounds.
Because the complexes used in this study are neutral,
and as a result complexes 1 and 3 are soluble in alkanes
and therefore, in principle, could give quite well-resolved
IR solution spectra in hexane, it was anticipated that
the conformers a and b might be observable by indi-
vidual ν(NN) or ν(CO) absorptions in IR spectroscopy
in view of its shorter time scale. This is not the case.
The NdN stretch is typically broad as is usual for
aryldiazenido complexes. The ν(CO) mode gives a very
sharp absorption band, which on close inspection is
somewhat asymmetric but certainly is not resolved into
separate bands.
Activation parameters for the isomerization of com-
pounds 1 and 3-5 are in Table 2 and may be compared
with those of the compounds studied previously.1 En-
tropy of activation values are small and positive as
observed before.1 The greatest values of ∆Gq and ∆Hq
are those for the hydride and methyl complexes 1 and
3, which have ∆Gq values in the range 53-56 kJ mol-1
and ∆Hq values of 60-63 kJ mol-1. By comparison, the
corresponding cationic complexes [Cp*Re(CO)(PR3)(p-
N2C6H4OMe)][BF4] obtained when H or CH3 in 1 and 3
is replaced by PR3 ) PMe3, P(OMe)3, or P(OCH2)3CH3
were observed to have ∆Gq values in the region 38-43
Con clu sion
It is clear from these additional examples that in the
spectroscopic characterization of aryldiazenido com-
1
plexes by H NMR, as well as by heteronuclear NMR
spectroscopy, the possibility of observing broad or
multiple signals and temperature-dependent behavior
resulting from conformational isomerization of the aryl-
diazenido ligand can be anticipated in favorable situa-
tions. More still needs to be understood regarding the
mechanism of isomerization and the factors influencing
the size of the barriers and the stereochemical prefer-
ences. We are endeavoring to clarify the situation by
means of suitable calculations and a continued experi-
mental investigation of appropriate complexes.
kJ mol-1 and ∆Hq values of 41-47 kJ mol-1 1
Since
.
steric effects are the least for the H and CH3 ligands,
this seems to point to an electronically-driven increase
in the activation free energy and enthalpy in going from
the ground state conformers a or b to the transition
state in which the NNC skeleton is linear (if inversion
alone is the mechanism) or the plane of the NNAr group
has been rotated by 90° (if rotation alone is the mech-
anism). This is understandable when it is considered
that good σ donor ligands like H or Me are expected to
stabilize the ground state conformations a or b relative
to the transition state in these neutral complexes by
increasing the electron density on Re and, thus, the
degree of back-bonding from the rhenium fragment to
the singly-bent N2Ar+ fragment. In the frontier orbital
analysis this interaction has maximum overlap when
the filled Re dπ orbital and in-plane empty p orbital on
the N2Ar group are coplanar (cf. Scheme 1b), and this
corresponds to the conformations a or b when L1 and
L2 differ substantially in π donor-acceptor ability, as
here. However, upon rotation to the perpendicular
conformation, this overlap is lost as the orbitals are now
orthogonal; in a different manner, linearization of the
NNC skeleton also decreases the π acceptor propensity
of the N2Ar ligand.1, 2 Of course, a synchronous combi-
nation of these processes is also possible. The neutral
compounds 4 and 5 are also related to the cationic
complexes [Cp*Re(CO)(PR3)(p-N2C6H4OMe)][BF4] (where
R ) Me or OMe) in the sense that the Cl ligand in 4 or
5 has been replaced by CO. The ∆Gq values for 4 and
5 are 40-47 kJ mol-1, which are quite close to those of
the cationic complexes, but the ∆Hq values of 51-53 kJ
mol-1 are higher than those of the cationic complexes
(41-47 kJ mol-1). A consistent interpretation would be
that the presence of the π electron-withdrawing CO
ligand and the positive charge deplete the π electron
density in the rhenium dπ orbital in the cationic
complexes by comparison with neutral 4 and 5. This
makes the π bond to the aryldiazenido ligand (Scheme
1b) less effective and raises the ground state conforma-
tion enthalpies relative to the transition state in the
Exp er im en ta l Section
Gen er a l Meth od s. The general synthetic procedures and
the spectroscopic instrumentation and techniques used in this
work were similar to those described previously, except where
otherwise noted.1 Hydride complex 1 was prepared by a minor
modification of the published method, as indicated in the
Results section, and gave IR and NMR spectroscopic data in
agreement with the literature.3
Cp *Re(CH3)(CO)(p-N2C6H4OMe) (3). An excess of CH3-
MgBr (ca. 3 mL of 3.0 M in ether, Aldrich) was added to a
stirred solution of Cp*ReCl(CO)(p-N2C6H4OMe) (2)4 (50 mg,
0.010 mmol) in freshly distilled ether. The reaction was
followed by IR and was complete in 36 h. Two drops of water
were added to destroy excess Grignard reagent, and the
mixture was extracted with hexane-ether (1:1) and filtered
through Celite. The solvent was removed under vacuum to
yield an orange-yellow solid that was recrystallized from
hexane at -78 °C. Yield: 38.9 mg (0.078 mmol, 81%). IR
(CH2Cl2, cm-1): ν(CO) 1896, ν(NN) 1613. IR (hexane, cm-1):
ν(CO) 1915, ν(NN) 1618. 1H NMR (298 K, 400 MHz, CDCl3):
δ 1.16 (s, 3H, CH3), 1.98 (s, 15H, Cp*), 3.83 (s, 3H, OMe), 6.89
(d, 2H, J ) 9 Hz, C6H4), 7.19 (d, 2H, J ) 9 Hz, C6H4). EIMS:
m/ z 500 (M+), 472 (M+ - CO). Anal. Calcd: C, 45.68; H, 5.04;
N, 5.61. Found: C, 46.04; H, 5.23; N, 5.47.
Cp *ReCl(P Me3)(p-N2C6H4OMe) (4). A large excess of
finely ground KCl was added to a stirred solution of [Cp*Re-
(CO)(PMe3)(p-N2C6H4OMe)][BF4]8 (50 mg, 0.080 mmol) in
CHCl3 (5 mL). Then, NaOH in ethanol (4 M, 3 mL) was added
slowly. The solution was stirred overnight and excess KCl was
removed by filtration through Celite. The solvent was removed
under vacuum, and the remaining solid was extracted with
hexane; removal of the hexane yielded a yellow oily product.
This was recrystallized from hexane to give 4 as a brown solid
in 78% yield (38 mg, 0.067 mmol). IR (CH2Cl2 or CHCl3, cm-1):
ν(NN) 1605. 1H NMR (298 K, 400 MHz, CDCl3): δ 1.54 (d,
9H, PMe3), 1.93 (s, 15H, Cp*), 3.79 (s, 3H, OMe), 6.87 (d, 2H,
J ) 9 Hz, C6H4), 7.26 (d, 2H, J ) 9 Hz, C6H4). 13C{1H} NMR
(CDCl3, 100.6 MHz): δ 10.96 (s, C5Me5), 16.68 (d, PMe3, J )
34 Hz), 55.52 (s, OMe), 100.76 (s, C5Me5), 114.34, 120.85,
(8) Klahn, A. H.; Sutton, D. Organometallics 1989, 8, 198.
Garcia-Minsal, Ariana
