One- and two-color infrared multiphoton dissociation of C3F6

Article abstract of DOI:10.1021/jp961454e

Multiphoton dissociation of C₃F₆ induced by one and two infrared laser beams is studied. The contribution of the different collisional processes to the yield is quantified using the recently extended McRae et al. model. It has been shown that the increase of the fluence in single irradiation and the change to the red of the second infrared pulse in double-beam experiments can have equivalent effects on the dissociation. These effects are an important increment in the yield and the extinction of the influence of the homogeneous collisions between C₃F₆ molecules on the process. In double-irradiation experiments, the introduction of time delays between the two laser pulses also leads to large enhancements of the dissociation, giving rise to two yield maxima when the delays are around 2 and 65 μs. A change in the final proportion of the dissociation products induced by modifying the frequency of the second pulse has been detected in the double-irradiation experiments.

Full text of DOI:10.1021/jp961454e

J. Phys. Chem. A 1997, 101, 2221-2228
2221
ARTICLES
One- and Two-Color Infrared Multiphoton Dissociation of C3F6
J. A. Torresano* and M. Santos
Instituto de Estructura de la Materia, Centro de F ´ı sica “Miguel A. Catal a´ n”, C.S.I.C.,
Serrano 121, 28006 Madrid, Spain
X
ReceiVed: May 21, 1996; In Final Form: January 3, 1997
3 6
Multiphoton dissociation of C F induced by one and two infrared laser beams is studied. The contribution
of the different collisional processes to the yield is quantified using the recently extended McRae et al. model.
It has been shown that the increase of the fluence in single irradiation and the change to the red of the second
infrared pulse in double-beam experiments can have equivalent effects on the dissociation. These effects are
an important increment in the yield and the extinction of the influence of the homogeneous collisions between
3 6
C F molecules on the process. In double-irradiation experiments, the introduction of time delays between
the two laser pulses also leads to large enhancements of the dissociation, giving rise to two yield maxima
when the delays are around 2 and 65 µs. A change in the final proportion of the dissociation products induced
by modifying the frequency of the second pulse has been detected in the double-irradiation experiments.
1
. Introduction
anharmonic bottleneck through V-V′ energy transfer.8 These
two processes are effective only if the collisions take place in
Irradiation of a molecule by two (or in general, several)
the presence of IR photons. Other collisional mechanisms,
acting both during and after the laser pulse, are vibrational
1-4
infrared (IR) beams is a technique that is often employed in
multiphoton dissociation (MPD) studies. This method allows
a better understanding of the process, introducing new experi-
mental variables such as the fluence and the frequency of the
second beam and the time delay between both laser fields. The
first pulse, nearly resonant with an IR fundamental absorption
band of the parent compound, excites the molecules through
the low-lying discrete levels. The second beam, usually fixed
to a red-shifted wavelength with respect to the former to account
for the anharmonicity, drives the absorption through the higher
discrete levels and the quasicontinuum until the dissociation limit
is overcome. Therefore, by using this dichromatic double
irradiation, the bottleneck effects that prevent the multiphoton
excitation can be circumvented, and a dissociation yield larger
than in single-irradiation experiments is usually obtained. In
these experiments, it can be considered that a change in the
9
energy pooling, which takes place when two molecules in the
quasicontinuum collide, resulting in one highly excited partner
being capable of dissociation, and collisional induced vibrational
quenching of excited molecules through vibrational to transla-
1
tional energy transfer (V-T).
6
Recently, we have extended the method previously developed
10-12
by McRae et al.
for the analysis of the contribution of the
different collisional processes to IR multiphoton dissociation.
Our extension allows the distinction, in the collisional sequence
corresponding to one homogeneous collision, of those processes
that affect the dissociation yield and that need the presence of
IR photons to be effective (e.g., overcoming of the anharmonic
bottleneck) from those that have not this requirement (e.g.,
vibrational energy pooling).
Multiphoton dissociation of C3F6 was first studied by Nip et
5
effective size of the molecule in relation to the MPD process
is induced; that is, the strength of the anharmonic barrier is
modified. We have defined the anharmonic barrier as the
al.,13 who proposed the following dissociation reaction
6
C F + nhν f C F + CF
2
(1)
3
6
2
4
ensemble of all quantum characteristics of a molecule that inhibit
the multiphoton absorption process. It consists of two main
components: (a) the vibrational barrier, originated by the
progressive anharmonic detuning of the absorption of the
infrared photons, initially resonant with the first vibrational
transition; (b) the rotational barrier, originated by the existence
of the rotational selection rules, which prevents a large propor-
tion of molecules from interacting with the radiation field.
Intermolecular energy transfers induced by collisions also
modify the strength of the anharmonic barrier: A mechanism
that increases the extent of the molecular excitation by reducing
the rotational barrier consists in replenishing of laser-induced
rotational hole burning through translational to rotational energy
where recombination reactions between the formed CF2 radicals
give rise to the formation of more C2F4 and, in a minor
proportion, to C2F6. We have found another product, poly-
14
(tetrafluoroethylene), (CF2)n, which under certain conditions
could represent up to 30% of the dissociated C3F6. We have
also shown that the strength of the anharmonic barrier for this
molecule is strongly dependent on the irradiation fluence. As
a consequence, the contribution of the different collisional
mechanisms to the absorption and dissociation processes is also
largely modified by the pulse fluence. This behavior suggests
that C F could be a suitable system to be investigated with the
3
6
extended analysis method cited above.
7
transfer, T-R (rotational hole filling ). Another mechanism
that lowers the vibrational barrier is the overcoming of the
In this paper, we study and quantify the contribution of the
different collisional mechanisms to the multiphoton dissociation
process of C3F6, induced by single (MPD) and double (MPD2)
infrared laser irradiation, using the extended McRae analysis
X
Abstract published in AdVance ACS Abstracts, February 15, 1997.
S1089-5639(96)01454-5 CCC: $14.00 © 1997 American Chemical Society

2222 J. Phys. Chem. A, Vol. 101, No. 12, 1997
Torresano and Santos
method. Although McRae et al. applied their method to two-
1
5
color two-beam dissociation of 1-bromo-2-fluoroethane, this
is the first time that such a kind of method is applied to double-
irradiation experiments with a variable temporal delay between
the two laser beams. We have carried out these experiments
with both beams tuned at the same or at different wavelengths.
Hereafter, we will refer to these situations as monochromatic
and dichromatic double irradiation, respectively. We also
investigate the influence of the time delay introduced between
the two laser fields on the process and the dependence of the
dissociation and recombination reactions on the different
irradiation parameters.
2
. Experimental Section
Figure 1. Experimental setup used in double-irradiation experiments:
A, CaF
lenses; P
2
attenuators; D, diaphragms; BS, Ge beam splitter; L, NaCl
i
, pyroelectric detectors.
Two TEA CO2 lasers, Lumonics K-101 and K-103, were
employed for the irradiation of the C3F6 samples. They were
equipped with frontal Ge multimode optics (85% reflection for
K-101 and 35% for K-103) and rear diffraction gratings with
the absolute values of the fluence was around 8% (5% being a
constant error coming from the nominal detector calibration
factor).
Perfluoropropene gas samples (99.0%) were kindly provided
by Prof. H. van der Bergh (Ecole Polytechnique F e´ d e´ rale de
Lausanne, Switzerland) and used without further purification.
Sample pressures in the cell were measured with two (0-1 and
135 lines/mm blazed at 10.6 µm. They were also equipped with
low jitter trigger devices (Lumonics 524) which provided a jitter
between the two pulses lower than 100 ns in the double-
irradiation experiments. The lasers operated with a CO2:N2:
He mixture in the proportion 8:8:84.
0
-10 mbar) MKS Baratron gauges. The dissociation yield was
In single-irradiation experiments, we have used the CO2
K-103 laser tuned at the 9P(26) line at 9.603 µm, which is nearly
resonant with the C-F stretching mode of C3F6.16 In the double-
irradiation case, the frequency of the K-101 laser was fixed at
the same 9P(26) laser line and provided the leading pulse in
delayed experiments. We employed the pulse originated in the
K-103 laser as the second IR field. Its frequency was also tuned
at the 9P(26) line in the experiments of monochromatic double
irradiation, while in the dichromatic case it was fixed at the
determined by monitoring the change in the absorbance of the
-1
16
1036 cm C3F6 band, with a FTIR spectrophotometer (Perkin-
Elmer Model 1725X). To study the influence of the different
experimental variables on the dissociation reaction, we have
used the two dimensionless parameters
[
C F ]
2 4
δ1 )
(2)
(3)
1
.5∆[C F ]
3 6
1
0R(36) line at 10.148 µm. This frequency corresponds to the
[C F ]
2 6
δ2 )
maximum of the multiphoton dissociation spectrum obtained
under double irradiation with the first pulse frequency fixed to
∆
[C F ]
3 6
9
.603 µm.17
3 6
where ∆[C F ] is the change in the concentration of hexafluo-
Laser pulse temporal profiles were observed with a photon
ropropene due to dissociation.
drag detector (Rofin Sinar 7080) and consisted of a spike (110
ns (fwhm) for K-101 and in the case of K-103, 65 ns for the
According to eq 1, δ1 ) 1 would mean that all the formed
CF2 radicals would be transformed into C2F4. The produced
amounts of C2F4 and C2F6 were determined by FTIR spectros-
copy, using measurements of their absorption bands at 1189
9P(26) line, and 140 ns for the 10R(36)), followed by a tail
approximately 1.5 µs long in all cases. These temporal profiles
were constant throughout the study.
-
1
18,19
and 1250 cm , respectively.
Two cylindrical Pyrex cells, 8.8 and 9.5 cm long, 0.95 and
.5 cm radius, and 30.5 and 71.0 cm total volume, respectively,
3
3. Results and Discussion
1
fitted with NaCl windows, contained the gas during the
irradiation. The largest cell was only used in the single-
irradiation experiments at a fluence of 6.8 J cm . The laser
beams were directed into the absorption cell in a near-parallel
irradiation geometry in which the laser beams were slightly
focused by 1.5 and 2 m focal length NaCl lenses.
In the method proposed by McRae et al. for the analysis of
the contribution of the different collisional mechanisms to the
-
2
10
dissociation process, the number of parent molecules, N , left
n
in the cell after n irradiation pulses is given by
n
N ) N
[1 - (V /V )f (φ ,Rk-1,âk-1)]
(4)
n
0
∏
I
c
k
k
In Figure 1, we show the experimental arrangement used in
the double-irradiation experiments. The time delay between the
two IR pulses was controlled by means of a digital delay
generator, Berkeley Nucleonics BNC 7036A, which allows a
delay between 0 and 100 µs with a step of 1 ns.
The laser fluence was calculated as the ratio of the pulse
energy measured with a Lumonics 20D pyroelectric detector,
and the “1/e” cross-sectional beam area was measured at the
cell position with a pyroelectric detector array Delta Develop-
ment Mark 4. (These beam areas, also used for computing the
irradiation volumes, VI’s, are given in the captions of Figures 2
and 3). The laser energy was changed by placing CaF2 plates
in the optical paths and controlling on the high-voltage settings
of the lasers for a finer adjust. The estimated uncertainty in
k)1
where N0 is the initial number of parent molecules, VI and Vc
are the irradiation and the total cell volumes, and f (φ,R,â) is
the single-pulse decomposition probability. This is a function
of the fluence, φ, and the partial pressures of parent molecule,
R, and dissociation products or any other buffer gas, â.
They assume a power dependence of f, on the partial pressures
in the cell given by
i-1
j
f(φ,R,â) )
hij(φ)R
â
(5)
∑
∑
i)1 j)0
where the coefficients hij depend only on the irradiation fluence
(φ).

Multiphoton Dissociation of C3F6
J. Phys. Chem. A, Vol. 101, No. 12, 1997 2223
given by the slope m. The value of P′ corresponds to the
pressure at which the number of collisions taking place in the
time of the pulse becomes large enough to start the deactivation
of the anharmonic barrier.
In the application of the extended McRae method to the MPD
of the C3F6 molecule, we have splitted the term corresponding
to one single heterogeneous collision to separate the contribution
due to collisions between the parent molecule and the two
gaseous products C2F4 and C2F6. These terms are written as
h (X) S(X)(C - C )
(7)
1
1
0
n
where X represents either C2F4 or C2F6. Cn is concentration of
C3F6 in the reaction cell after n IR pulses, C0 being the initial
value. The parameters S(X) are the number of C2F4 or C2F6
molecules that are produced from a dissociated molecule of C3F6
and are directly determined from the measured values of δ1 and
δ2 (eqs 2 and 3): S(C2F4) ) 1.5δ1 and S(C2F6) ) δ2.
The inclusion of other parameters that represent processes
involving more than one molecular collision has not been
necessary in any of the fits of the experimental data.
Single Irradiation. We have used the fraction of C3F6
molecules dissociated per laser pulse in the whole cell, given
by
Y ) 1 - N /N
(8)
n
0
for quantifying the dissociation yield.
In Figure 2, we give the variation of the yield Y versus the
number of laser pulses, n, for two different values of the laser
fluence. The calculated yields obtained by applying the method
outlined at the beginning of this section appear as continuous
lines in the figure. In the fits, each experimental point has been
weighted by the inverse square of the associated error. Table
1 shows the relevant parameters obtained from the fits together
3 6
Figure 2. Fraction of dissociated C F molecules, Y, versus the number
of laser pulses for different values of the initial pressure. Solid lines
represent the fits of the experimental points to the extended method
described in the text (section 3). (a) The irradiation fluence was φ )
-
2
2
2
.3 J cm , and the beam area was 0.32 ( 0.01 cm . (b) φ ) 6.8 J
2 2
-
20
2
cm , and a beam area of 0.54 ( 0.01 cm .
with their standard deviations, the chi-square (ø ) and the
number of degrees of freedom (ν). For the higher fluence value,
-
2
This method takes into account the variations in the pulse-
to-pulse yield induced by the change of the reactant and product
partial pressures during the course of the experiment. The
magnitude and sign of the parameters hij provide information
about the contribution to the dissociation coming from the
6.8 J cm (Figure 2b), a good fit is obtained by taking into
account just the collisionless dissociation, h10, and the hetero-
geneous collisions C3F6-C2F4 and C3F6-C2F6 represented by
the parameters h11(C2F4) and h11(C2F6), respectively. The
opposite signs of both parameters imply that, while the collisions
of C3F6 with C2F4 inhibit the dissociation process, those with
C2F6 are favorable to it. A possible explanation to this could
6
different collisional sequences. Our modification to the method
consists basically in the introduction of an initial parent pressure
dependence in the parameter corresponding to one homogeneous
collision, h20. This dependence is of the form
-
1
-1
be the existence of an absorption band at 1025 cm (16 cm
to the red of the laser frequency) in the C2F6 spectrum,19 which
would allow this compound to act as a “sensitizer”,1
0,21
by
a
2
a
20
absorbing energy from the IR field and transferring it to the
C3F6 molecule in the collision, through V-V′ exchange. The
absence of the h20 parameter in the final fits means that, at these
fluence conditions, homogeneous collisions between C3F6
molecules do not play an important role in the MPD process,
in agreement with the existence of the weak anharmonic
h ,
if P < P′
0
h ) h + m(P - P′), if P′ < P < P′′
(6)
2
0
{
a
b
20
h + h ,
if P > P′′
2
0
a
2
The parameter h is associated with those processes taking
0
place independently of the presence of the laser field, that is,
energy pooling and quenching, while h is related to those
operating only in the presence of infrared photons. Thus, h20
14
bottleneck already detected in this molecule and suggesting a
b
2
0
low influence of the V-V′ energy transfer process between
parent molecules excited in the quasicontinuum. It is also worth
noting the high value obtained for the collisionless parameter
h10. This means that a large part of the irradiated molecules
breaks without collisional assistance in accordance, not only
with the presence of a small anharmonic barrier but also with
the observed trend of the dissociation probability at the low
b
can be interpreted as a measurement of the maximum weakening
that can be induced in the anharmonic barrier by means of
homogeneous collisions at some given irradiation conditions.
a
2
The sign of h is determined by the competition between
0
energy pooling and vibrational quenching, being positive if the
b
1
4
former dominates and negative otherwise. The parameter h20
3 6
initial C F concentration limit.
-
2
is always positive. P′ and P′′ are respectively the initial pressure
When the fluence is fixed to 2.3 J cm , the analysis method
provides a different picture of the MPD process. The role of
heterogeneous collisions does not change with respect to that
obtained for the highest fluence. However, in this case, we
of the parent molecule at which the collisional processes
b
2
associated with h begin to be active and that at which they
0
saturate. The rate at which this saturation value is reached is

2224 J. Phys. Chem. A, Vol. 101, No. 12, 1997
Torresano and Santos
pressure we obtain the largest deactivation of the bottleneck
effect induced by homogeneous collisions.
The collisionless parameter h10 is around 20 times smaller
-
2
than the one obtained for the higher fluence, φ ) 6.8 J cm .
Under similar experimental conditions, a change in the laser
-
2
fluence from 2.6 to 5.9 J cm gives rise to only a 4-fold
6
increase in h10 in the MPD of CF2HCl. This difference supports
the idea that the anharmonic barrier can be circumvented more
easily in C3F6 than in CF2HCl.
The dissociation products obtained for the experiments of
1
4
Figure 2 are the same that were found previously (i.e., C2F4,
C2F6, and (CF2)n). For the two fluence values that we used,
and for a given initial pressure, the parameters δ1 and δ2 defined
in section 2 are nearly constant when the number of pulses is
increased. For the highest pressures, we saw that δ1 < 0.66,
which is, in principle, the minimum value expected from
reaction 1 (corresponding to the situation in which the formed
CF2 radicals do not dimerize to C2F4). This result could mean
that reaction 1 is not the only dissociation path in the MPD of
C3F6, at least under high-pressure conditions, or that the directly
formed C2F4 from C3F6 is involved in some recombination
reactions.
Double Irradiation. The dependence of the MPD2 yield,
Y, on the time delay between the two laser pulses, ∆τ, is shown
in Figure 3 for monochromatic and dichromatic double irradia-
tion. In both cases, a first maximum is obtained corresponding
to a ∆τ of a few microseconds. This maximum arises as a result
of the competition between the different mechanisms favoring
and inhibiting the dissociation process, such as V-V′ and V-T
intermolecular energy transfer, and the diffusion of the mol-
ecules irradiated by the first pulse outside the volume irradiated
by the second one. The finite temporal width of the laser pulse
also contributes to the appearance of this maximum.
Figure 3. Dependence of the fraction of dissociated molecules, Y, on
the time delay, ∆τ, between the two laser fields. The initial C
pressure was P ) 0.9 mbar, and 50 pairs of laser pulses were used. In
each pair, the wavelength, fluence, and beam area of the first pulse
3 6
F
0
-
2
were kept constant to λ
1
) 9.603 µm, φ
1
) 0.8 ( 0.1 J cm , and A
1
2
)
0.38 ( 0.01 cm , respectively. (a) Monochromatic irradiation, λ )
2
-
2
2
λ
1
, φ
2
) 1.2 ( 0.1 J cm , and A
2
) 0.32 ( 0.01 cm . (b) Dichromatic
2
-
irradiation, λ
2
) 10.148 µm, φ ) 1.6 ( 0.2 J cm , and A ) 0.25 (
2
2
2
0
.01 cm .
TABLE 1: Model Parameters and Their Standard
Deviations Obtained for the Fits Corresponding to the
a
Experiments of Figure 2
Figure 2a
Figure 2b
parameter
(φ ) 2.3 J cm-2)
(φ ) 6.8 J cm-2)
h
h
h
h2
P′
10
0.035 ( 0.001
-0.35 ( 0.06
0.85 ( 0.30
0.025 ( 0.002
0.35 ( 0.07
0.8 ( 0.2
0.72 ( 0.01
-1.2 ( 0.2
3.5 ( 0.9
11(C
1b 1(C
2
F
F
4
)
)
2
6
0
P′′
m
The second maximum appearing in both curves can be
associated with the reflection of the energy wave originated by
the V-T intermolecular energy transfer on the cell walls.
Similar experiments carried out in CFHCl2 with cells of different
0.066 ( 0.015
ν
ø
48
37
52
55
2
2
4
a
b
2
diameter support this interpretation.
Together with this
P′′ has been computed from the values of h , P′, and m through
0
-
1
thermoacoustic wave, there may exist another wave associated
with the V-V′ intermolecular transfer. This vibrothermal wave
is reflected on the cell walls and returns to the central core,
resulting in an overexcitation of the molecules therein. The
value of ∆τ for the second maximum, ≈65 µs, corresponds to
the time at which this wave arrives to the central core
simultaneously with the second IR field. For a given cell, this
characteristic time only depends on the parameters defining the
first IR pulse and on the nature of the molecule. In this sense,
we can say that the second pulse acts like a “probe” beam of
the situation created in the cell by the first one.
2 4 2 6
eq 5. Units are mbar for P′ and P′′, mbar for h11(C F ), h11(C F ),
b
-2
2
and h , and mbar for m. The chi-square (ø ) and the number of
2
0
degrees of freedom (ν) of each fit are also given.
have to take into account the homogeneous collisions that take
place during the laser pulse. They induce a weakening of the
anharmonic barrier and produce an increment in the dissociation
b
2
yield as it is indicated by the positive sign of h . On the other
0
a
2
hand, it is not necessary to include the parameter h in the fits,
0
confirming in this way that collisions between C3F6 molecules
excited in the quasicontinuum do not contribute to the MPD
process. In addition, a study of the MPD of C3F6 carried out
in the presence of argon22 strongly suggests that the nature of
the anharmonic barrier in C3F6 is essentially vibrational.
In Figure 3, the continuous lines represent the fits of the
experimental points to a simple model that we have developed
2
4
for the double-irradiation process, which takes into account
the existence of the thermoacoustic wave and relates the energy
absorption from both beams with the effective size of the
The value of the pressure P′ enables us to calculate the
minimum number of collisions in the time of the pulse that are
required to weaken effectively the anharmonic barrier. If we
suppose that the mean molecular speed is the given by the
5
molecule.
In Figures 4 and 5, we give the dependence of the dissociation
yield, Y, on the initial C3F6 pressure in the cell, P0, for two
different delay values: ∆τ ) 0 and ∆τ ) 3 µs, respectively.
Figures 4a and 5a correspond to monochromatic irradiation,
whereas Figures 4b and 5b present the behavior for pulses of
different frequencies.
The results for monochromatic irradiation suggest that the
effect of the anharmonic barrier is relevant in this MPD2
process. For simultaneous irradiation and rather small initial
C3F6 pressure, the number of homogeneous collisions that take
2
3
kinetic theory and we take a collisional cross section σ ) 1.1
-
14
2
×
10
approximately two, for an effective temporal pulse width around
00 ns. The obtained value of P′ is close to the pressure found
cm , we obtain the number of collisions to be
5
1
4
in our previous work, 0.3 mbar, which corresponds to the
minimum of the dissociation yield versus the initial pressure of
C3F6. The saturation value of h20 (equal in this case to h ) is
b
0
2
reached for an initial C3F6 pressure P′′ ) 0.8 mbar; at this

Multiphoton Dissociation of C3F6
J. Phys. Chem. A, Vol. 101, No. 12, 1997 2225
Figure 4. Dependence of the fraction of dissociated molecules, Y, after
Figure 5. Same as in Figure 4 for a time delay, ∆τ ) 3 µs, between
the two IR pulses.
5
0 pairs of simultaneous laser pulses on the initial C
F
3 6
pressure in the
cell: (a) monochromatic irradiation and (b) dichromatic irradiation.
All other conditions as in Figure 3.
TABLE 2: Model Parameters and Their Standard
Deviations Obtained for the Fits Corresponding to the
Experiments of Figure 6a
place during the effective time of the pulses is not enough to
induce the overcoming of the anharmonic barrier. The only
relevant collisional processes are those inhibiting the dissociation
through V-T and V-V′ intermolecular transfers, justifying the
initial decrease of the yield as P0 increases. As the C3F6 pressure
grows, the number of homogeneous collisions also increases,
and the mechanism of weakening of the anharmonic barrier
becomes effective, giving rise to the increment in the yield as
monochromatic
parameter ∆τ ) 0 µs ∆τ ) 3 µs
0.034 ( 0.001 0.030 ( 0.002 0.074 ( 0.001 0.284 ( 0.005
dichromatic
∆τ ) 0 µs ∆τ ) 3 µs
h10
h11(C2F4) -0.51 ( 0.03 -0.46 ( 0.04 -0.38 ( 0.03 -1.89 ( 0.07
h11(C2F6) 1.6 ( 0.1
1.4 ( 0.2
0.82 ( 0.15
6.4 ( 0.3
b
0.035 ( 0.001 0.047 ( 0.002
h
20
P′
0
0
P0 grows that is observed in Figure 4a. The curve “yield/P0”
P′′
m
ν
0.90 ( 0.05
0.52 ( 0.08
found in the MPD study of C3F614 for a fluence of 2.4 J cm-2
0.039 ( 0.001 0.09 ( 0.01
50
11
49
60
51
63
53
86
is very much coincident with the one shown in Figure 4a,
including the position of the yield minimum around 0.2 mbar.
2
ø
a
b
20
P′′ has been computed from the values of h , P′ and m through
(This fluence is roughly the sum of the fluences of the two
2
eq 5. The chi-square (ø ) and the number of degrees of freedom (ν)
of each fit are also given. Units are the same indicated in Table 1.
pulses used in MPD2.)
When a time delay is introduced between the two-equal
frequency IR pulses, the number of homogeneous collisions
suffered by a C3F6 molecule interacting with the first field before
the arrival of the second is larger for a given value of P0. This
allows a proportion of initially bottlenecked molecules to
overcome the barrier and thus to get a higher excitation level
above the dissociation limit by absorbing photons from the
second laser field. In this way, the molecular RRKM rate
becomes faster, giving rise to the larger dissociation yield found
in delayed irradiation with respect to the simultaneous case. This
also explains the shift to lower pressure values in the position
of the minimum of Figure 4a. This shift leads to the progressive
reduction of the initial decreasing yield region, until it becomes
unobservable for sufficient long delays, as it is observed in
Figure 5a.
a suitable value of the second IR wavelength. In this case, due
to the red-shifted frequency, the molecules excited to the
bottleneck region by the first pulse can absorb more efficiently
the second nonresonant field, so that the effect of the anharmonic
barrier is largely suppressed. In this way, the introduction of a
time delay between the two pulses would not produce an
increment of the proportion of molecules absorbing the second
IR pulse. Nevertheless, such an increment actually occurs when
a suitable short delay is established because, in this case, the
molecules can absorb the largest part of the energy delivered
by the first laser pulse and so get a higher excitation level before
the arrival of the second one. This time delay must be nearly
coincident with the total temporal width of the first pulse. In
our case, we see in Figure 3b that, for 1.5 µs (approximately
the first pulse width), the maximum has been reached. Because
of the inhibitory character to the excitation of the homo- and
2
5
Substantial changes take place in the above scheme if the
dissociation process is induced by dichromatic irradiation with

2226 J. Phys. Chem. A, Vol. 101, No. 12, 1997
Torresano and Santos
3 6
Figure 6. Fraction of dissociated C F molecules, Y, versus the number of pairs of laser pulses for different values of the initial pressure: (a)
monochromatic simultaneous irradiation, (b) monochromatic irradiation with a delay, ∆τ, of 3 µs between the two IR pulses, (c) dichromatic
simultaneous irradiation, and (d) dichromatic irradiation and ∆τ ) 3 µs. Solid lines represent the fits of the experimental points to the extended
method described in the text. In each pair the wavelength and fluence of the first pulse were kept constant to λ
1
) 9.603 µm and φ
1
) 1.0 ( 0.1
-
2
-2
-2
J cm . For mochromatic irradiation: λ
2
) λ
1
and φ
2
) 1.4 ( 0.1 J cm . For dichromatic irradiation: λ
2
) 10.148 µm and φ ) 1.7 ( 0.2 J cm .
2
Other conditions as in Figure 3.
heterogeneous collisions, an increment in the value of the initial
C3F6 pressure must induce a decrease of the yield for any value
of the time delay, ∆τ, as is observed in Figures 4b and 5b. This
situation is analogous to the one obtained in single irradiation
Finally, we see that heterogeneous collisions have the same
effect on the MPD2 process as in single MPD, i.e., inhibiting
the dissociation in the C3F6-C2F4 case and enhancing it in the
C3F6-C2F6 one.
The value of the limiting pressure P′ is so small that the
method cannot determine it with enough accuracy, and it has
been set to zero in the fits. The difference of this situation with
respect to the one obtained in single irradiation (P′ ) 0.35 mbar)
is probably due to the larger effective temporal width of the
pulse employed in MPD2 experiments.
-
2
for a fluence value of 6.8 J cm .
The dissociation yield of C3F6 versus the number of pairs of
laser pulses, n, is shown in the Figure 6a-d for monochromatic
and dichromatic irradiation and for the two values of ∆τ used
in Figures 4 and 5. By applying the extended McRae model to
the results of this figure, we can analyze the different dissocia-
tion schemes described above. In Table 2, we give the values
of the relevant parameters and their associated standard devia-
tions obtained in the fits. Note that, for delayed irradiation,
the results obtained from the application of the method refer to
the effect of the second infrared pulse, which acts on molecules
affected by both the effect of the first laser and the time
evolution during the established time delay. As can be expected
for simultaneous monochromatic irradiation, the determined
parameters are close to those obtained in single MPD for a
The result of the fits obtained for ∆τ ) 3 µs in monochro-
matic irradiation is also in accord with our previous discussion.
The processes involved in the dissociation do not change with
respect to the simultaneous case. The weakening of the
anharmonic barrier is made clear in the higher value of m
b
0
(h2 gets its maximum for a pressure P′′ ) 0.5 mbar, while this
pressure is 0.9 mbar for ∆τ ) 0).
The results for dichromatic irradiation resemble those obtained
for single irradiation under high-fluence conditions. A large
weakening of the anharmonic barrier, and therefore a substantial
change in the dissociation process, is manifested in the obtained
fit parameters: The homogeneous collisions do not have a
noticeable influence on the dissociation, and only three param-
eters, those corresponding to the pressure-independent dissocia-
tion and the heterogeneous collisions, are sufficient to describe
-2
fluence of 2.3 J cm . The existence of a noticeable anharmonic
barrier is supported by the small value of the collisionless
parameter, h10, and the dependence of h20 on the initial pressure
of C3F6. In addition, the zero value obtained for the parameter
a
2
h
shows that there is no significant contribution to the
0
dissociation yield coming from the energy pooling mechanism.

Multiphoton Dissociation of C3F6
J. Phys. Chem. A, Vol. 101, No. 12, 1997 2227
Figure 7. Proportions of C
Figure 3, represented by the parameters δ
a) monochromatic irradiation; (b) dichromatic irradiation.
2
F
4
and C
2
F
6
formed in the experiments of
Figure 8. Same as in Figure 7 for the experiments of Figure 5.
1
and δ defined in the text:
2
(
reasonable to assume that this product is the polymer (CF2)n
that we have obtained in a significant amount because we have
not detected any other dissociation product.
the results. For simultaneous irradiation, the value of h10 is
twice that found in the monochromatic case. When ∆τ is fixed
to 3 µs, the situation does not change qualitatively; however,
the obtained yield is much larger than for ∆τ ) 0, with a large
increment in the pressure-independent contribution to the yield.
The dissociation products that we have obtained in the double-
irradiation experiments of C3F6 are the same that we found for
the single-irradiation case. The dependence of the parameters
δ1 and δ2 on the time delay between the laser pulses is given in
Figure 7 for the same experiments of Figure 3. We can see
that the proportion of formed C2F6 is correlated with the obtained
dissociation yield: The maxima (minima) of δ2 are placed at
those values of ∆τ corresponding to the maxima (minima) of
the yield, approximately. This clearly occurs for dichromatic
irradiation and seems to appear in the monochromatic case as
well. On the other hand, the proportion of formed C2F4 is nearly
constant, within the limit of our experimental error.
For the experiments shown in Figures 4 and 5, we obtain
that an increment in the C3F6 pressure, P0, leads to a larger
proportion of formed C2F6 and to a smaller proportion of C2F4,
in agreement with the results obtained in single-irradiation
experiments. Although both parameters tend to saturation, the
limiting value of δ1 remains constant as the wavelength
associated with the second IR field changes, whereas δ2
increases. This increment is small in the case of simultaneous
irradiation and noticeable when the delay is 3 µs. In Figure 8,
we show the dependence of δ1 and δ2 on P0, for the delayed
experiments.
The different energy content of the CF2 radicals created in
the dissociation of C3F6 could provide a possible explanation
for this competition, if one admits that the less energetic CF2’s
are more susceptible to engage in polymer formation. The larger
amount of energy in the fragments formed by dichromatic
irradiation is due to the weaker anharmonic barrier that C3F6
presents at these conditions with respect to the monochromatic
case. In this way, the frequency of the second pulse would
control the ratio between the production of polymer and C2F6
because it determines the energy content in the formed species.
4
. Conclusions
In this work, we have analyzed and quantified the different
collisional mechanisms involved in the infrared multiphoton
dissociation of hexafluoropropene by using the extended McRae
et al. method. For double-monochromatic irradiation, a rather
small contribution from collisionless dissociation is found. We
have established that homogeneous collisions play an important
role in circumventing the bottleneck effect, but they do not give
rise to any energy pooling transfer between the molecules
excited in the quasicontinuum. These results are similar to those
obtained from single-irradiation experiments at low fluence and
support the existence of a significant anharmonic barrier in the
MPD of C3F6 under these conditions. On the other hand, when
the frequency of the second IR field is red-shifted with respect
to that of the first pulse, the bottleneck effect is largely removed
and leads to an important increase in the collisionless contribu-
tion to the process. In this case, the influence of homogeneous
collisions disappears. The same conclusions are obtained from
single-irradiation results at high fluences. In double-irradiation
These results suggest that, for some given values of ∆τ, P0
and the wavelength of the second field, there exists a competition
between the production of C2F6 and another product, different
from C2F4, in the overall MPD2 process of C3F6. It seems

2228 J. Phys. Chem. A, Vol. 101, No. 12, 1997
Torresano and Santos
experiments, a time delay of a few microseconds between the
two IR fields induces a large increment in the dissociation yield,
giving rise to the occurrence of a maximum of the dissociation
yield. For monochromatic irradiation, this effect is due to the
large number of homogeneous collisions that take place before
the arrival of the second pulse, weakening the strength of the
anharmonic barrier. Nevertheless, for the dichromatic case, the
yield increment is mainly due to the temporal distribution of
the first pulse energy. In both cases, heterogeneous collisions
between C3F6 and the two gaseous dissociation products have
opposite effects: favoring the dissociation in the case of C2F6
and quenching it in the case of C2F4. We also point out the
appearance of a second maximum in the dependence of the yield
on the time delay for rather long values of this variable (≈65
µs). We have related this maximum to the reflection on the
cell walls of the energy wave created by the V-T intermolecular
relaxation.
Concerning the dissociation and recombination reactions, we
have found a competition between the formation of C2F6 and
the polymer (CF2)n, induced by changing the frequency of the
second IR pulse.
Finally, it has to be remarked that hexafluoropropene has
revealed a molecular system whose behavior, in relation to
multiphoton dissociation, is modified in a highly noticeable way
by rather small changes in the irradiation conditions, suggesting
that this molecule is particularly appropiate for the study of the
process of multiphoton excitation and dissociation.
(2) Letokhov, V. S., Ed. Laser Spectroscopy of Highly Vibrationally
Excited Molecules; Adam Hilger: Bristol, 1989.
(3) Quack, M. Infrared Phys. 1989, 29, 441.
(
4) Ivanco, M.; Evans, D. K.; McAlpine R. D. J. Phys. Chem. 1989,
9
3, 2383.
(5) Simpson, T. B.; Black, J. G.; Burak, I.; Yablonovitch, E.;
Bloembergen, N. J. Chem. Phys. 1985, 83, 628.
(6) Torresano, J. A.; Santos, M. J. Phys. Chem. 1996, 100, 9726.
(7) Stephenson, J. C.; King, D. S.; Goodman, M. F.; Stone, J. J. Chem.
Phys. 1979, 70, 4496.
(8) Chou, J. J.; Grant, E. R. J. Chem. Phys. 1981, 74, 384.
(9) Gauthier, M.; Hackett, P. A.; Willis, C. Chem. Phys. 1980, 45, 39.
(
10) McRae, G. A.; Yamashita, A. B.; Goodale, J. W. J. Chem. Phys.
990, 92, 5997.
11) McRae, G. A.; Evans, D. K.; Goodale, J. W. J. Chem. Phys. 1990,
93, 1689.
(12) McRae, G. A.; Lee, P. E.; McAlpine, R. D. J. Phys. Chem. 1991,
5, 9332.
13) Nip, W. S.; Hackett, P. A.; Willis, C. J. Phys. Chem. 1980, 84,
32.
14) Torresano, J. A.; Santos, M.; Gonz a´ lez-D ´ı az, P. F. Laser Chem.
1
(
9
(
9
(
1994, 14, 217. For comparison and due to the different method used to
determine the beam area, the fluences given in this reference have to be
multiplied by 1.5. Due to a computational error, the given values for the
parameter R2, corresponding to δ1 in the present work, have to be divided
by 1.18.
(15) McRae, G. A.; Ivanco, M.; Goodale, J. W. Int. J. Chem. Kinet.
1994, 26, 147.
(16) Nielsen, J. R.; Claassen, H. H.; Smith, D. C. J. Chem. Phys. 1952,
2
0, 1916.
17) Santos, M.; Sig u¨ enza, C. L.; Torresano, J. A.; Gonz a´ lez-D ´ı az, P.
F. Spectrochim. Acta 1990, 46A, 455.
18) Nielsen, J. R.; Claassen, H. H.; Smith, D. C. J. Chem. Phys. 1950,
(
(
Acknowledgment. The authors acknowledge to L. J. Garay,
C. L. Sig u¨ enza, P. F. Gonz a´ lez-D ´ı az, and L. D ´ı az for useful
discussions. We are largely grateful to the referees for their
valuable comments and suggestions. This work has been carried
out with financial support provided by the Spanish DGICYT
under Project PB93-0145-C02-02.
1
8, 812.
(19) Nielsen, J. R.; Richards, C. M.; McMurry, H. L. J. Chem. Phys.
948, 16, 67.
1
(20) Press, W. H.; Flannery, B. P.; Teukolsky, S. A.; Vetterling, W. T.
Numerical Recipes; Cambridge University Press: Cambridge, 1986.
(
(
(
21) Kramer, J. J. Phys. Chem. 1983, 87, 3563.
22) Torresano, J. A.; Santos, M. Manuscript in preparation.
23) Ireton, R. C.; Ko, A. N.; Rabinovitch, B. S. J. Phys. Chem. 1974,
References and Notes
7
8, 1984.
(24) Torresano, J. A.; Santos, M. Manuscript in preparation.
(1) Cantrell, C. D., Ed. Multiple Photon Excitation and Dissociation
of Polyatomic Molecules; Topics in Current Physics 35; Springer-Verlag:
Berlin, 1986.
(25) Pochon, E.; Weston Jr., R. E.; Flynn, G. W. J. Phys. Chem. 1985,
89, 86.