Moreover, the characteristic timescales of action vary from the tens of seconds up to
the tens of femtoseconds upon one-photon absorption throughout the S 0 !S 1
spectral range. One has to bear in mind, that IC is a dominant channel at low
excitation energies, which is one order of magnitude faster than electron emission
out of S 1 , whereas prompt PD completely surpasses IC at higher energies. The
problem arises in the red part of the spectrum, since only a small fraction of the
signal, which comes from the prompt PD, might have been registered [32], whereas
a larger fraction coming from IC has to be registered in the delayed channel,
corresponding to the decay of the anion in the hot ground state. Upon one-photon
absorption, the delayed action occurs on a timescale of seconds (see Fig. 5.10),
when IR radiative cooling sets in and inhibits the action.
Precautions should apply, however, when trying to interpret the blue shift of the
maximum in the one-photon absorption profile compared to those in the action
spectra. The peak at 482 nm, registered in the action spectra, may erroneously be
attributed solely to the multiple-photon absorption [32]. As discussed above, this
cannot be the case, when photons are absorbed sequentially. Moreover, the calculated FC factors reveal that the 0–0 transition has the largest relative strength, and it
is located at 492 nm close to the maximum in the action spectra at 482 nm (see
Fig. 5.18). It must be stressed that the single-photon cross section does not necessarily peak at the wavelength of the main transition with the largest FC strength.
This is well-known, but it is often overlooked, since such shifts are usually small
(~300–500 cm
À1 ) [45] and originate from a set of transitions to the states, where
low-frequency active modes are excited in addition to the main transition. Even
though the relative strengths of such transitions are smaller, but their density is
obviously higher. However, the shift disclosed here is much larger than that
predicted in [45]. The maximum rather refers to the transitions, where the highfrequency modes are excited. This is usually not expected, since each subsequent
higher-energy transition, also accompanied by the same low-frequency satellites as
the main transition, has a smaller relative strength according to the corresponding
FC progression. The non-trivial aspect here is that there is an accidental degeneracy
of two sets of active modes, breathing at 849 cm
À1 and stretches at
1,600–1,700 cm
À1 . The former breathing mode has a long FC progression, such
that its 0–2 excitation coincides in energy and in the corresponding strengths with
those of the 0–1 excitation for the high-frequency modes. This causes a peak in the
density of states of the active modes and a maximum in the absorption at 452 nm.
The sequential two-photon contribution in the delayed peak, as well as the multiplephoton contribution to the prompt peak at 482 nm, on the other hand, are indeed
relevant, as they peak in the region of the vertical excitation that is close to the
adiabatic transition in this case.
The experimental low-temperature excitation spectrum of the emission [66],
which is solely attributed to the anionic form of the chromophore in the wild-type
GFP, also perfectly matches the calculated spectrum (see Fig. 5.18). The shoulder at
446 nm in the low-temperature spectrum refers to the excitation of high-frequency
stretching modes that can unambiguously be correlated with those excited in the gas
phase (see Fig. 5.18).
5 Photo-initiated Dynamics and Spectroscopy of the Deprotonated Green. . .
97
the tens of femtoseconds upon one-photon absorption throughout the S 0 !S 1
spectral range. One has to bear in mind, that IC is a dominant channel at low
excitation energies, which is one order of magnitude faster than electron emission
out of S 1 , whereas prompt PD completely surpasses IC at higher energies. The
problem arises in the red part of the spectrum, since only a small fraction of the
signal, which comes from the prompt PD, might have been registered [32], whereas
a larger fraction coming from IC has to be registered in the delayed channel,
corresponding to the decay of the anion in the hot ground state. Upon one-photon
absorption, the delayed action occurs on a timescale of seconds (see Fig. 5.10),
when IR radiative cooling sets in and inhibits the action.
Precautions should apply, however, when trying to interpret the blue shift of the
maximum in the one-photon absorption profile compared to those in the action
spectra. The peak at 482 nm, registered in the action spectra, may erroneously be
attributed solely to the multiple-photon absorption [32]. As discussed above, this
cannot be the case, when photons are absorbed sequentially. Moreover, the calculated FC factors reveal that the 0–0 transition has the largest relative strength, and it
is located at 492 nm close to the maximum in the action spectra at 482 nm (see
Fig. 5.18). It must be stressed that the single-photon cross section does not necessarily peak at the wavelength of the main transition with the largest FC strength.
This is well-known, but it is often overlooked, since such shifts are usually small
(~300–500 cm
À1 ) [45] and originate from a set of transitions to the states, where
low-frequency active modes are excited in addition to the main transition. Even
though the relative strengths of such transitions are smaller, but their density is
obviously higher. However, the shift disclosed here is much larger than that
predicted in [45]. The maximum rather refers to the transitions, where the highfrequency modes are excited. This is usually not expected, since each subsequent
higher-energy transition, also accompanied by the same low-frequency satellites as
the main transition, has a smaller relative strength according to the corresponding
FC progression. The non-trivial aspect here is that there is an accidental degeneracy
of two sets of active modes, breathing at 849 cm
À1 and stretches at
1,600–1,700 cm
À1 . The former breathing mode has a long FC progression, such
that its 0–2 excitation coincides in energy and in the corresponding strengths with
those of the 0–1 excitation for the high-frequency modes. This causes a peak in the
density of states of the active modes and a maximum in the absorption at 452 nm.
The sequential two-photon contribution in the delayed peak, as well as the multiplephoton contribution to the prompt peak at 482 nm, on the other hand, are indeed
relevant, as they peak in the region of the vertical excitation that is close to the
adiabatic transition in this case.
The experimental low-temperature excitation spectrum of the emission [66],
which is solely attributed to the anionic form of the chromophore in the wild-type
GFP, also perfectly matches the calculated spectrum (see Fig. 5.18). The shoulder at
446 nm in the low-temperature spectrum refers to the excitation of high-frequency
stretching modes that can unambiguously be correlated with those excited in the gas
phase (see Fig. 5.18).
5 Photo-initiated Dynamics and Spectroscopy of the Deprotonated Green. . .
97