O.F. Mohammed, A.A.O. Sarhan / Chemical Physics 372 (2010) 17–21
19
*
assigned to metal–acceptor (d–
*
is Cp (ligand)-acceptor (p–p ) transition. As shown in Fig. 1B, the
p
) and the high energy transition
Femtosecond time-resolved fluorescence of Fc-ph-DCV was
measured at a series of detection wavelengths ranging from 480
to 590 nm after 400 nm excitation (the S2 S0 excitation (see
Fig. 2). The time-resolved fluorescence was recorded under the ma-
gic-angle conditions to avoid the effect of rotational motion. Fig. 2
displays the time-resolved fluorescence traces measured in ACN in
the time region up to 14 ps (A) and up to 250 ps (B). The represen-
tative time-resolved fluorescence traces in DCM at 480 and 590 nm
are also depicted in Fig. 2C. It can be clearly seen that the spike-like
fluorescence signal appears with photoexcitation in the whole
wavelength region. This extremely fast emission is attributable to
the S2 fluorescence. The long-lived fluorescence component was
seen in the wavelength region longer than 530 nm, which is
assignable to the fluorescence from the S1 state that is generated
by the internal conversion. The time-resolved fluorescence signals
were well fitted with three exponential functions with characteris-
low energy transition is strongly affected by replacing dicyanovinyl
with a weaker electron acceptor, CHO. This fact disagrees with the
first model because the d–d transition should not be very sensitive
to the acceptor strength. Therefore, we consider that the low en-
ergy transition is assignable to the metal–acceptor (d–
*
p ) CT band.
Both low and high energy bands of Fc-ph-DCV exhibit red-shifts
in dichrolomethane (DCM, n = 1.424) and chloroform (CHCl3,
n = 1.445) compared to that in acetonitrile (ACN, n = 1.344), n is
the solvent refractive index. This suggests the dipole–dipole inter-
action is not the predominant factor to determine the transition
energy because the absorption bands should be red-shifted in polar
ACN in such a case. The transition energy decreases with increasing
polarizability of the solvent molecules. This suggests that the Fc-
ph-DCV has only small dipole moment in the ground state and
the dipole–dipole interaction does not influence the transition en-
ergy. The observed solvatochromism is probably dominated by dis-
persion interactions and by dipole-induced-dipole interactions
[15,16] for which the polarizability of the solvent determines the
stabilization energy. A similar type of the solvatochromic shifts
has already been suggested for other CT systems [17].
tic times of 110 30 fs (
s1), 590 100 fs (
s
2), 49 3 ps (s3) in ACN
and 120 40 fs ( 1), 900 150 fs (
s
s2), 23 3 ps (s
3) in DCM. The s1
component was S2 fluorescence and the relevant time constant is
the S2 lifetime determined by the S2 ? S1 internal conversion.
The evaluated lifetime of 110–120 fs is a typical lifetime of the
highly excited state of large polyatomic molecules of this size
[18]. The slower component (s3) is assigned to the lifetime of the
relaxed S1 state. Importantly, the lifetime of the S1 state is signifi-
cantly longer in a more polar solvent than that in less polar ones,
indicating that the S1 state having CT character which is stabilized
in the polar medium. It is known that, in the conventional intra-
molecular CT dynamics, the CT state is more stabilized in solvents
having higher polarity because of higher barrier of the solvent reor-
ganization energy for back electron transfer reaction [19]. We note
that the torsional motion of the vinyl moiety may also contribute
to relatively short lifetime of S1/CT state. The effect of the twisting
motion on the non-radiative decay of the CT state has been dis-
cussed and it was suggested that it reduces the lifetime of the CT
state [19,20].
In regard to the s2 component, it is very likely that the solvation
process plays a very crucial role in the excited-state dynamics be-
cause the S1 state has CT character due to a substantial charge
transfer from the electron donor Fc to the acceptor (DCV). The sol-
vation process occurs on the subpicosecond to picosecond time
scales, depending on the solvent. In fact, the time scale of the dif-
fusive solvation process is reported to be 600 fs in ACN and 1 ps in
DCM [21], which are in good agreement with the time scale of the
s2 component (see Table 1). In this process, the solvent molecules
start relaxing and reorienting themselves to adapt the excited-
state molecules to achieve the equilibrium condition from the ini-
tial non-equilibrated one. Alternative possible assignment of the s2
component is intra-molecular vibrational relaxation process in the
S1 state following the ultrafast S2 ? S1 internal conversion. How-
ever, decisive assignment of the s2 component is difficult with
the available data, at the moment.
We note that the vibrational cooling (intermolecular vibrational
relaxation by energy transfer to the solvent molecules) does not
contribute to this subpicosecond component because the vibra-
tional cooling typically occurs on much longer time scales, in the
temporal range 5–30 ps [22–24].
We also carried out femtosecond time-resolved transient
absorption measurements with excitation at 390 nm and
Table 1
Fitted decay times of up-conversion profiles of Fc-ph-DCV in ACN and DCM at 590 nm
(limit of error 20%).
Fig. 2. Femtosecond time-resolved fluorescence signals of Fc-ph-DCV (cꢀ10ꢁ3 M)
in ACN measured at various wavelengths after 400 nm excitation in the short time
range (A) and in the longer time range (B), inset is the zoom of the ultrafast
component (lifetime of S2 which is about 100 fs). Time-resolved fluorescence
signals in DCM at the longer time range are also shown for comparison (C).
Solvent
s1 (fs)
s2 (fs)
s3 (ps)
ACN
DCM
110 30
120 40
590 100
900 150
49
23
3
3