74
R. de Nalda et al.
100 fs. The minor channel through the non-adiabatic surface crossing, yielding
CH 3 (ν = 0) + I( 2 P 3/2 ), is complete earlier, in just 60 fs. This is expected from the
higher available kinetic energy for this channel, which is translated into a larger terminal velocity and thus earlier completion. More surprisingly, the channel yielding
vibrationally excited CH 3 in its symmetric stretch mode, CH 3 (ν 1 = 1) + I( 2 P 3/2 ),
is severely delayed, with a clocking time of around 150 fs.
It is interesting to note here that, as was noted by Zewail and others in seminal
femtochemistry papers [2], the temporal resolution in this type of experiments is not
limited by the duration of the pump and probe pulses (through their linear or nonlinear cross-correlation), but rather by the signal-to-noise ratio, and the reproducibility
of the pulse temporal profiles. Sub-pulse duration temporal resolution is therefore
not only possible, but routinely achieved with stable systems.
The results shown above were obtained with a probe laser centered at 333.5 nm
causing a (2 + 1) REMPI transition in CH 3 , corresponding to the Q branch of the
3p z ( 2 A
2 ← 2 A
2 )0 0
0 transition. Vibrationless and symmetric stretch excited methyl
fragments were detected. It is well known that, upon CH 3 I photolysis, the CH 3 moiety goes through a severe change in geometry, from pyramidal in the CH 3 I molecule,
to planar as a free fragment. As a consequence of this, an umbrella-mode (ν 2 ) wave
packet is created in the UV photodissociation. The issue of whether C–I bond elongation precedes vibrational excitation of the methyl radical, or vice versa, could be
in principle examined through the measurement of “clocking” times for different
vibrational components of this wave packet. In this case, contrarily to symmetric
stretch excitation, that has similar frequencies for the ground and excited 3p z Rydberg state of methyl, causing the 1 1
1 transition to lie spectrally close to the 0 0
0
transition, the umbrella mode frequency is almost twice as large for the 3p z Rydberg state than for the ground state. This causes a considerable blue shift in the resonance wavelength, which is beyond the probe laser bandwidth. By tuning the probe
laser to 329.4 nm (2 1
1 transition), or to 325.8 nm (2 2
2 transition), vibrationally excited methyl with one (ν 2 = 1) or two quanta (ν 2 = 2) in the umbrella mode can be
probed. Such an experiment was performed in the Madrid laboratory and was reported in detail in Ref. [39]. Indeed, vibrationally excited CH 3 was detected, both in
pure umbrella overtones and combination bands with symmetric stretch excitation.
Figure 4.5 also depicts the transients measured when detecting vibrationally excited
methyl fragments.
The values obtained for the clocking of all the channels explored are shown in
Table 4.1. Some of the values are consistent with arguments based solely on total available energy and final relative velocity. However, they contain an intriguing
result: whereas for the adiabatic channel yielding CH 3 + I ∗ ( 2 P 1/2 ) the choice of
the vibrational component probed does not provoke any variation of the measured
clocking time, this is noticeable not the case for the non-adiabatic channel yielding
CH 3 + I( 2 P 3/2 ), where, systematically, higher-lying vibrational components seem
to show a “delayed” appearance time.
These experimental results have been confronted with state-of-art theoretical calculations for the CH 3 I in the A-band [39]. A wave packet model including four
degrees of freedom, namely the C–I dissociation coordinate, the I–CH 3 bending
R. de Nalda et al.
100 fs. The minor channel through the non-adiabatic surface crossing, yielding
CH 3 (ν = 0) + I( 2 P 3/2 ), is complete earlier, in just 60 fs. This is expected from the
higher available kinetic energy for this channel, which is translated into a larger terminal velocity and thus earlier completion. More surprisingly, the channel yielding
vibrationally excited CH 3 in its symmetric stretch mode, CH 3 (ν 1 = 1) + I( 2 P 3/2 ),
is severely delayed, with a clocking time of around 150 fs.
It is interesting to note here that, as was noted by Zewail and others in seminal
femtochemistry papers [2], the temporal resolution in this type of experiments is not
limited by the duration of the pump and probe pulses (through their linear or nonlinear cross-correlation), but rather by the signal-to-noise ratio, and the reproducibility
of the pulse temporal profiles. Sub-pulse duration temporal resolution is therefore
not only possible, but routinely achieved with stable systems.
The results shown above were obtained with a probe laser centered at 333.5 nm
causing a (2 + 1) REMPI transition in CH 3 , corresponding to the Q branch of the
3p z ( 2 A
2 ← 2 A
2 )0 0
0 transition. Vibrationless and symmetric stretch excited methyl
fragments were detected. It is well known that, upon CH 3 I photolysis, the CH 3 moiety goes through a severe change in geometry, from pyramidal in the CH 3 I molecule,
to planar as a free fragment. As a consequence of this, an umbrella-mode (ν 2 ) wave
packet is created in the UV photodissociation. The issue of whether C–I bond elongation precedes vibrational excitation of the methyl radical, or vice versa, could be
in principle examined through the measurement of “clocking” times for different
vibrational components of this wave packet. In this case, contrarily to symmetric
stretch excitation, that has similar frequencies for the ground and excited 3p z Rydberg state of methyl, causing the 1 1
1 transition to lie spectrally close to the 0 0
0
transition, the umbrella mode frequency is almost twice as large for the 3p z Rydberg state than for the ground state. This causes a considerable blue shift in the resonance wavelength, which is beyond the probe laser bandwidth. By tuning the probe
laser to 329.4 nm (2 1
1 transition), or to 325.8 nm (2 2
2 transition), vibrationally excited methyl with one (ν 2 = 1) or two quanta (ν 2 = 2) in the umbrella mode can be
probed. Such an experiment was performed in the Madrid laboratory and was reported in detail in Ref. [39]. Indeed, vibrationally excited CH 3 was detected, both in
pure umbrella overtones and combination bands with symmetric stretch excitation.
Figure 4.5 also depicts the transients measured when detecting vibrationally excited
methyl fragments.
The values obtained for the clocking of all the channels explored are shown in
Table 4.1. Some of the values are consistent with arguments based solely on total available energy and final relative velocity. However, they contain an intriguing
result: whereas for the adiabatic channel yielding CH 3 + I ∗ ( 2 P 1/2 ) the choice of
the vibrational component probed does not provoke any variation of the measured
clocking time, this is noticeable not the case for the non-adiabatic channel yielding
CH 3 + I( 2 P 3/2 ), where, systematically, higher-lying vibrational components seem
to show a “delayed” appearance time.
These experimental results have been confronted with state-of-art theoretical calculations for the CH 3 I in the A-band [39]. A wave packet model including four
degrees of freedom, namely the C–I dissociation coordinate, the I–CH 3 bending