9 On the Investigation of Excited State Dynamics with (Pump-)Degenerate
211
cation, filter transmission, etc.). This can be easily overcome by spectrally resolving
the signal in a spectrometer and detecting the signal with a CCD camera.
9.2.3 Role of Spectral Overlap
A central point in the investigation of excited state dynamics with (pump-)DFWM
is the spectral overlap between excitation pulses and the molecular absorption. In
this regard, there are two important aspects how DFWM and pump-DFWM can be
tuned in order to separate different signal contributions.
The first aspect can be illustrated by a comparison between DFWM and transient
absorption (TA). Both techniques are third-order time-resolved methods. DFWM,
however, uses three beams and homodyne detection, while TA uses two beams and
self-heterodyne detection by the probe beam. In principle, both methods are able to
generate vibrational coherence: In DFWM, the first two pulses excite two or more
different vibrational levels. Similarly, in transient absorption vibrational levels lying
within the bandwidth of the pump pulse are excited. Except for these similarities,
DFWM presents additional features not found for e.g. TA. One of these features
is the possibility of efficient non-resonant excitation (Fig. 9.1(c)). DFWM can efficiently generate vibrational modes with non-resonant spectra in comparison to TA,
which clearly contains only ground state modes. This can be contrasted to resonant (or even near-resonant) excitation, which potentially contains both excited and
ground state modes. By comparing both signals, it is possible to assign modes to
their electronic potential.
The second aspect is related to the detection of excited state dynamics without
any interference of ground state contributions. The spectrum tuning of initial pump
and DFWM allows detecting exclusively excited state dynamics. This is performed
by making the initial pump spectrum resonant (or near-resonant if the molecular
absorption is not perfectly within the nc-OPA wavelength range) which leads to
the excitation of the excited state as well as ground state vibrational manifold. The
suppression of ground state contributions is achieved by the correct tuning of the
DFWM spectrum: It must be tuned to the excited state absorption (ESA) or stimulated emission (SE) spectral regions. If the ESA and SE do not overlap with the
ground state absorption, this leads to a pure excited electronic state signal. This
can be easily understood since the DFWM signal is nonresonant with the ground
state absorption and the DFWM signal scales with the transition dipole moment μ 8 .
Non-resonant transition dipole moments are several orders of magnitude smaller
than resonant transition dipole moments, generating resonant signals 10 6 stronger
than non-resonant signals [32, 33].
For partially or totally overlapping ESA/SE and ground state absorption, the
pump-DFWM will intrinsically have ground state contributions, which can be,
however, strongly reduced by changing the energy relation of the initial pump
and the DFWM sequence. Third order techniques with homodyne detection like
DFWM signal provide two control knobs which are critical in this respect: The
211
cation, filter transmission, etc.). This can be easily overcome by spectrally resolving
the signal in a spectrometer and detecting the signal with a CCD camera.
9.2.3 Role of Spectral Overlap
A central point in the investigation of excited state dynamics with (pump-)DFWM
is the spectral overlap between excitation pulses and the molecular absorption. In
this regard, there are two important aspects how DFWM and pump-DFWM can be
tuned in order to separate different signal contributions.
The first aspect can be illustrated by a comparison between DFWM and transient
absorption (TA). Both techniques are third-order time-resolved methods. DFWM,
however, uses three beams and homodyne detection, while TA uses two beams and
self-heterodyne detection by the probe beam. In principle, both methods are able to
generate vibrational coherence: In DFWM, the first two pulses excite two or more
different vibrational levels. Similarly, in transient absorption vibrational levels lying
within the bandwidth of the pump pulse are excited. Except for these similarities,
DFWM presents additional features not found for e.g. TA. One of these features
is the possibility of efficient non-resonant excitation (Fig. 9.1(c)). DFWM can efficiently generate vibrational modes with non-resonant spectra in comparison to TA,
which clearly contains only ground state modes. This can be contrasted to resonant (or even near-resonant) excitation, which potentially contains both excited and
ground state modes. By comparing both signals, it is possible to assign modes to
their electronic potential.
The second aspect is related to the detection of excited state dynamics without
any interference of ground state contributions. The spectrum tuning of initial pump
and DFWM allows detecting exclusively excited state dynamics. This is performed
by making the initial pump spectrum resonant (or near-resonant if the molecular
absorption is not perfectly within the nc-OPA wavelength range) which leads to
the excitation of the excited state as well as ground state vibrational manifold. The
suppression of ground state contributions is achieved by the correct tuning of the
DFWM spectrum: It must be tuned to the excited state absorption (ESA) or stimulated emission (SE) spectral regions. If the ESA and SE do not overlap with the
ground state absorption, this leads to a pure excited electronic state signal. This
can be easily understood since the DFWM signal is nonresonant with the ground
state absorption and the DFWM signal scales with the transition dipole moment μ 8 .
Non-resonant transition dipole moments are several orders of magnitude smaller
than resonant transition dipole moments, generating resonant signals 10 6 stronger
than non-resonant signals [32, 33].
For partially or totally overlapping ESA/SE and ground state absorption, the
pump-DFWM will intrinsically have ground state contributions, which can be,
however, strongly reduced by changing the energy relation of the initial pump
and the DFWM sequence. Third order techniques with homodyne detection like
DFWM signal provide two control knobs which are critical in this respect: The