as the excitation energy of the^3 MLCT state (E 00 ). However, var-
iation of the phosphorous ligand, which should give larger effects
on the d–d splitting, had an effect on E 00 þDG6¼ similar to or
larger than that onE 00. These results clearly indicate that the
photochemical ligand substitution reactions proceed via the^3 LF
excited state thermally produced from the^3 MLCT state.
Scheme 3 illustrates the potential energies of the ground state,
(^3) MLCT state, and (^3) LF state versus the bond distance Re CO.
In the cases of most rhenium(I) diimine tricarbonyl complexes
with a phosphorous ligand, the activation energies (DG6¼) are
enough small to be overcome even at room temperature
(kBT¼207.1 cm^1 at 298 K). The evaluated activation energy of
3awas 3810 cm^1 and the quantum yield of the photochemical
ligand substitution reaction was 0.089 using 366-nm light
(Tables III and IV). However, the complexes which do not have
a phosphorous ligand, such as fac-Re{(MeO) 2 bpy}(CO) 3 Cl (1c)
andfac-[Re(bpy)(CO) 3 (py)]þ(2a), do not react at all under the
same conditions using 366-nm light. Activation energies (DG6¼)
of these complexes, as evaluated from the temperature depen-
dence of the emission lifetime and the emission quantum yield,
are drastically smaller. For instance, the DG6¼ of 1c was
estimated as 252 cm^1. This is the same order of the magnitude
(^3) LF
(a) (b) (c)
(^3) MLCT
Energy
Ground
state
Reaction
Re--CO distance
Δdd^1
ΔGπ 1
ΔGπ 2
ΔGπ 3
ΔGπ 2 ′
Δdd^2 Δdd^3
SCHEME 3. Energy versus ReCO distance for rhenium(I)
complexes. Illustrating the three lowest-lying electronic states: (a)fac-
[Re{(CF 3 ) 2 bpy}(CO) 3 {P(OEt) 3 }]þ (3c), (b) fac-[Re(bpy)(CO) 3 (py)]þ (2a)
andfac-Re{(MeO) 2 bpy}(CO) 3 Cl (1c) for the case of a nonreactive^3 LF
state, and (c)2aand1cfor the case of a large activation energy from
the^3 MLCT to^3 LF states. Copyright 2002 American Chemical Society.
158 HIROYUKI TAKEDAet al.