Photodissociation dynamics of the methyl radical 3s Rydberg state

Article abstract of DOI:10.1063/1.469193

The photodissociation dynamics of methyl radical have been investigated at 193.3 nm using photofragment translational spectroscopy.The formation of CH2 and H(²S) was the only dissociation pathway observed.Although it is not possible to assign the spin state of the methylene unambiguously, we believe the methylene is produced predominately in the <*> ¹A₁ excited state.The translational energy distribution of the products is peaked at ca. 13 kcal/mole which is consistent with the magnitude of the exit barrier on the excited state potential energy surface.The breadth of the distribution suggests that the methyl radicals dissociate from a wide range of geometries.From the photofragment angular distribution an anisotropy parameter of β = -0.9+/-0.1 was determined.

Full text of DOI:10.1063/1.469193

Photodissociation dynamics of the methyl radical 3s Rydberg state
Simon W. North, David A. Blank, Pamela M. Chu, and Yuan T. Lee
Citation: The Journal of Chemical Physics 102, 792 (1995); doi: 10.1063/1.469193
View online: http://dx.doi.org/10.1063/1.469193
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/102/2?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Translational spectroscopy studies of the photodissociation dynamics of O− 4
J. Chem. Phys. 105, 10803 (1996); 10.1063/1.472888
Photodissociation dynamics of OClO
J. Chem. Phys. 105, 8142 (1996); 10.1063/1.472700
Statespecific neutral timeofflight of CO from ketene photodissociation at 351 nm: The internal energy distribution
of CH2(X̃ 3 B 1)
J. Chem. Phys. 105, 4550 (1996); 10.1063/1.472538
Photochemistry of phosgene in the solid phase: Stateresolved dissociation dynamics
J. Chem. Phys. 104, 7904 (1996); 10.1063/1.471506
Photodissociation dynamics of doubly excited Rydberg states of molecular hydrogen
J. Chem. Phys. 96, 5843 (1992); 10.1063/1.462682
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.88.90.110 On: Sat, 20 Dec 2014 15:35:04

Photodissociation dynamics of the methyl radical 3s Rydberg state
Simon W. North, David A. Blank, Pamela M. Chu,a) and Yuan T. Lee
Chemical Sciences Division, Lawrence Berkeley Laboratory, University of California, Berkeley, California
94720 and Department of Chemistry, University of California, Berkeley, California 94720
͑Received 1 August 1994; accepted 3 October 1994͒
The photodissociation dynamics of methyl radical have been investigated at 193.3 nm using
photofragment translational spectroscopy. The formation of CH2 and H͑2S͒ was the only dissociation
pathway observed. Although it is not possible to assign the spin state of the methylene
1
˜
unambiguously, we believe that methylene is produced predominately in the a A1 excited state.
The translational energy distribution of the products is peaked at ϳ13 kcal/mole which is consistent
with the magnitude of the exit barrier on the excited state potential energy surface. The breadth of
the distribution suggests that the methyl radicals dissociate from a wide range of geometries. From
the photofragment angular distribution an anisotropy parameter of ␤ϭϪ0.9Ϯ0.1 was
determined. © 1995 American Institute of Physics.
I. INTRODUCTION
which are predicted to occur in comparable yields at vibra-
tional energies corresponding to ϳ216 nm excitation fol-
Although the ground state reactivity and decomposition
of methyl radical have been the subject of numerous studies,1
there is little information regarding the photochemistry of
even the lowest electronically excited state. In fact, there are
only two reports which have directly detected a dissociation
lowed by internal conversion.2 Although the elimination of
2
˜
molecular hydrogen does not correlate with the CH3͑B A1Ј͒
state along either C2v or Cs pyramidal pathways, Cs planar
or C1 pathways are allowed. At 216 nm reaction ͑3͒ was in-
ferred from OH fluorescence produced via the reaction of
product arising from B state photochemistry.2,3 This is some-
˜
2
CH͑X ⌸͒ with O2.11 However, Ye et al. found that
what surprising in light of the considerable attention given to
CH͑A,B–X) fluorescence resulting from methyl radical pho-
todissociation at 193.3 nm was quadratically dependent on
the laser power.8 The authors concluded that CH originated
from the secondary dissociation of methylene implying that
at 193.3 nm H2 elimination does not compete efficiently with
simple C–H bond cleavage. Only H-atom elimination from
methyl radical photodissociation was considered by the au-
thors of Refs. 9 and 10. Ab initio calculations by Yu et al.
predicted a barrier to dissociation of 11.5 kcal/mole at
4
˜
the analogous A states of NH3 and H2O.
The ultraviolet spectroscopy of the methyl radical was
first examined by Herzberg and Shoosmith5 and the transi-
tion originating at 216 nm was assigned to an excitation of
the unpaired 2pz electron to a 3s Rydberg orbital. The
broadening of the CH3 transition was attributed to a rapid
predissociation of the excited state due to H-atom tunneling.6
Callear and Metcalfe subsequently resolved vibrational fea-
tures between 216–204 nm beyond which the spectrum
evolves into a broad continuum.7 The transition at 193.3 nm,
therefore, is an excitation, not to a discrete band, but an
absorption to the dissociative continuum. Methyl radical
photodissociation has been observed at this wavelength as
both a primary process8 and as a secondary process in the
dissociation of acetone9 and azomethane.10
R
C–Hϭ1.55 Å for reaction 1. It was noted, however, that the
calculated barrier may be reduced to 4.6–9.2 kcal/mole with
a more substantial basis set and increased configuration in-
teraction. Resonance Raman experiments carried out by
Kelly and co-workers have addressed the ro-vibronic depen-
12,13
˜
dence of the predissociation lifetimes of the B state.
˜
J-dependent tunneling rates of the B-state origin band were
At 193.3 nm there are three energetically allowed disso-
ciation pathways, two involving the loss of an H atom and
one involving the elimination of H2. Orbital correlations in
the dissociation of the methyl radical were first examined
theoretically by Ye et al.4 In agreement with the earlier pos-
modelled with a one-dimensional cubic potential and a best
fit to the experimental data yielded a barrier of 6.29 kcal/
mole at RC–Hϭ1.38 Å.13 Recent high level ab initio calcula-
tions performed by Botschwina et al. were in accord with the
results of Kelly.14 Figure 1 shows the energy level diagram
for the photodissociation of methyl radical including corre-
lations for a HCH angle of 120 °.
˜
tulation of Herzberg, the B state was found to correlate with
1
˜
the formation of singlet methylene, CH2͑a A1͒,
2
1
2
˜
˜
CH B AЈ͒ → CH a A ͒ϩH S͒
͑1͒
͑
͑
͑
3
2
1
Despite the theoretical work and spectroscopic measure-
ments there have been few experimental investigations in-
1
and the ground state methyl radicals correlate to two asymp-
totic dissociation channels,
˜
volving the photodissociation of the B state. Chen et al. in-
vestigated the photochemistry of methyl radical at 216 nm
and detected the CH2 ͑m/eϭ14) photofragment by photo-
ionization at 10.5 eV although the spin state of the methylene
could not be assigned.2 No m/eϭ13 ͑CHϩ͒ product was
observed but since the 10.5 eV photon lies below the CH͑2⌸,
2
3
2
˜
˜
CH X AЉ͒ → CH X B ͒ϩH S͒,
͑2͒
͑3͒
͑
͑
͑
3
2
1
2
ϩ
2
2
CH X AЉ͒ → CH X ⌸͒ϩH 1⌺g ͒,
˜
͑
͑
͑
2
3
2
a͒
ϭ0) adiabatic ionization potential, determination of the
branching ratio between H atom and H2 elimination was not
v
Present address: National Institute of Standards and Technology, Gaithers-
burg, Maryland.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
792 J. Chem. Phys. 102 (2), 8 January 1995 0021-9606/95/102(2)/792/7/$6.00 © 1995 American Institute of Physics
130.88.90.110 On: Sat, 20 Dec 2014 15:35:04

North et al.: Photodissociation of the methyl radical
793
FIG. 1. Energy level diagram for methyl radical showing the energetically
accessible dissociation channels for 193.3 nm excitation.
FIG. 2. Mass spectra of the molecular beam of Ͻ1% di-tertbutyl peroxide in
He taken with the pyrolytic nozzle at ͑a͒ room temperature ͑heat off͒ and ͑b͒
1400 K ͑heat on͒.
possible. Recently, Dixon and co-workers measured H-atom
time of flight from methyl radical photodissociation at 216
nm following CH3SH photolysis.3 Contrary to all theoretical
predictions ground state methylene, CH2͑X B1͒, was deter-
mined to be the momentum matched fragment. This conclu-
sion was based on energetic considerations, i.e., there was
3
˜
1
˜
CH2 photoproduct is formed predominantly in the a A1 ex-
cited state.
insufficient energy available concomitant with the formation
II. EXPERIMENT
1
˜
of CH2͑a A1͒ to account for the short arrival times of the
observed H atoms. It was postulated that the formation of
triplet methylene resulted from a coupling between the
ground and excited state potential energy surfaces in the exit
channel at small CH2 angles ͑ϳ104°͒. The structure in the
time of flight, originally thought to be CH2 bending excita-
tion, was later found to be inconsistent with the dissociation
of CH3.15,16 These features have since been attributed to the
photolysis of CH3S arising from the dissociation of
Two experimental configurations were used in this work.
All of the photofragments, with the exception of H and H2,
were measured using a fixed-source rotatable-detector appa-
ratus as described previously.18,19 The pulsed beam of radi-
cals was generated from the pyrolysis of a suitable precursor.
Although both azomethane and di-tert-butyl peroxide
͑DTBP͒ were employed with similar success, all of the data
presented in this paper was obtained with DTBP as the pre-
cursor. A pulsed valve20 fitted with a SiC pyrolysis nozzle
͑1.0 mm I.D.͒ was operated at 80 Hz. The source employed
is an adaptation of the one designed and characterized by
Chen et al. and a more complete description can be found
elsewhere.21 Briefly, a SiC tube was resistively heated to
ϳ1400 K, as determined from optical pyrometry and beam
time of flight, by a dc current limited power supply. DTBP,
Ͻ1%, was expanded in helium at a total stagnation pressure
of ϳ800 Torr. The thermal decomposition of DTBP results in
two methyl radicals and two acetone molecules and has been
shown to be an efficient source of CH3.22 Shown in Fig. 2
are mass spectra of the molecular beam with the nozzle un-
heated ͑upper trace͒ and heated ͑lower trace͒. The lower
spectra shows the quantitative loss of the parent ͑m/eϭ146,
m/eϭ73, and m/eϭ57, etc.͒, the appearance of the known
cracking pattern for acetone, and a large contribution from
CH3SH.17 A reinvestigation of methyl radical dissociation at
1
˜
216 nm suggests that only CH2͑a A1͒ is produced in agree-
ment with theoretical predictions.15 The measured transla-
tional energy release however, is energetically consistent
with either singlet or triplet methylene.
In this paper we present results on the photodissociation
of methyl radical excited at 193.3 nm using photofragment
translational spectroscopy. This is the only experiment which
has directly measured the H-atom/H2 elimination branching
ratio at any wavelength. The only observed products were
CH2 and H. This is extremely relevant to those interested in
the 193 nm photolysis of acetone and azomethane as a
source of CH3 for kinetics studies. Although our methodol-
ogy cannot unambiguously discriminate between the differ-
ent spin states of the methylene we believe, based on the
translational energy and photofragment anisotropy, that the
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 102, No. 2, 8 January 1995
130.88.90.110 On: Sat, 20 Dec 2014 15:35:04

794
North et al.: Photodissociation of the methyl radical
the methyl radical parent ͑m/eϭ15). The beam was mea-
sured in situ and had a velocity of ϳ3600 m/s with a full
width at half maximum ͑FWHM͒ of ϳ15%. The pulsed
beam was twice collimated to an angular divergence of 2°
and crossed at a right angle with the output of a Lambda-
Physik EMG 202 MSC excimer laser which was operated at
the ArF transition ͑193.3 nm͒. The 5–30 mJ/pulse output
was focused to a 2ϫ4 mm spot in the interaction region
using a spherical lens ͑fϭ30 cm͒. Polarization of the laser
beam was achieved by 10 quartz plates at Brewster’s angle
͑Ͼ95% polarization͒. The neutral photoproducts traveled
20.8 cm where they were ionized by electron bombardment
and mass selected using a quadrupole mass filter. The ions
were then counted as a function of time using a Daly ion
counter and a multichannel scaler interfaced to a computer
that was triggered with the laser pulse.
In order to detect H and H2 photofragments an alterna-
tive arrangement was necessary. The requirements for detec-
tion of these fragments have been thoroughly discussed else-
where and will only be summarized.23,24 The pulsed valve
with pyrolytic nozzle was mounted mutually perpendicular
to both the detector and the laser beam. The radicals were
photolyzed ϳ4.0 mm above the nozzle opening and the
H-atom photoproducts passed through two 3.0 mm skimmers
28.5 cm to the detector. Although the beam velocity distri-
bution could not be determined in this configuration, the
H-atom velocity is large compared to the beam velocity and,
as a result, the H-atom TOF spectra were found to be rela-
tively insensitive to the beam conditions. The data was fit
with the forward convolution technique. Details of the fitting
procedure and the data analysis program can be found
elsewhere.25
FIG. 3. TOF spectra at a laboratory angle of 7.5 ° for ͑a͒ m/eϭ15 ͑CHϩ3 ͒,
͑b͒ m/eϭ14 ͑CHϩ2 ͒, ͑c͒ m/eϭ13 ͑CHϩ͒, and ͑d͒ m/eϭ12 ͑Cϩ͒. The open
circles represent data. The dotted line is the contribution from acetone dis-
sociation, the dashed line is the contribution from CH3 dissociation using the
P(ET͒ in Fig. 4, and the solid line is the total fit to the data.
the P(ET͒ for acetone photodissociation. In the
m/eϭ14 ͑CHϩ2 ͒ TOF spectrum, in addition to the CH3
daughter ion from acetone, is a second feature which appears
at longer times. This is attributed to methylene originating
from the photodissociation of methyl radical. An identical
feature is observed at m/eϭ12–14 when using azomethane
The DTBP 98% was obtained from Aldrich and used
without further purification. Azomethane was synthesized by
the method of Renaud and Leitch26 and was purified by trap
to trap distillation prior to use.
III. RESULTS AND ANALYSIS
Time-of-flight spectra were taken at m/eϭ12–15, m/e
ϭ2, and m/eϭ1 with unpolarized light. TOF spectra of m/e
ϭ28 ͑COϩ͒ were also collected in order to confirm the pho-
todissociation of acetone, a pyrolysis bi-product. Acetone is
known to have an appreciable absorption cross section ͑ϳ2.7
ϫ10Ϫ18 cm2͒ at 193.3 nm due to a 3s Rydberg transition.27
The dynamics of acetone photodissociation, particularly the
resulting translational energy distributions of the CH3 and
CO photofragments, have been well characterized.28,29 All of
the observed signals exhibited a linear dependence upon the
laser power demonstrating that the photofragments resulted
exclusively from single photon processes at the laser flu-
ences employed. Time-dependent background from the
pulsed valve was collected with the laser off and subse-
quently subtracted from the TOF spectra. Figure 3 shows
TOF spectra taken at ⌰labϭ7.5 ° at m/eϭ15–12. The m/e
ϭ15 ͑CHϩ3 ͒ ͓Fig. 3͑a͔͒ cannot arise from methyl radical
photodissociation and was successfully fit, as shown by the
solid line, with the previously determined translational en-
ergy distribution, P(ET͒, for the photodissociation of acetone
at 193.3 nm.28 TOF spectra at m/eϭ28 were also fit using
FIG. 4. Center-of-mass translational energy distribution used to fit the data
in Figs. 3, 5, and 6. Arrows indicate the thermodynamic maximum available
energies for formation of singlet and triplet methylene.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 102, No. 2, 8 January 1995
130.88.90.110 On: Sat, 20 Dec 2014 15:35:04

North et al.: Photodissociation of the methyl radical
795
FIG. 6. TOF spectrum of m/eϭ1 ͑Hϩ͒ at a laboratory angle of 90°. The
open circles represent data with the time-dependent background subtracted.
The solid line represents the forward convolution fit using the P(ET͒ in
Fig. 4.
FIG. 5. TOF spectra of m/eϭ12 ͑Cϩ͒ at laboratory angles of ͑a͒ 10 ° and
͑b͒ 14 °. The open circles represent data. The dotted line is the contribution
from acetone dissociation, the dashed line is the contribution from CH3
dissociation using the P(ET͒ in Fig. 4, and the solid line is the total fit to the
data.
on the heavy fragment ͑m/eϭ15–12͒ and m/eϭ2 data, we
estimate that if the CHϩH2 channel exists it certainly repre-
sents Ͻ5% of the dissociation products.
The center-of-mass P(ET͒ shown in Fig. 4 is peaked
away from zero, ͗ET͘ϳ15 kcal/mole, with a FWHM of 12
kcal/mole. The available energy for the formation of ground
state methylene is given by
as the radical precursor and, therefore, the assignment of this
peak can be made with some confidence.30 The dashed line is
the fit to the methylene contribution using the P(ET͒ shown
in Fig. 4. Any CH͑2⌸͒ produced from the elimination of H2
would appear at m/eϭ13 ͑CHϩ͒ and m/eϭ12 ͑Cϩ͒. Both of
these spectra can be fit without including the CHϩH2 chan-
nel and therefore it appears that formation of CH2 and H is
the only active dissociation channel. Additional TOF spectra
for m/eϭ12 at laboratory angles of 10° and 14° are shown in
Fig. 5.
E
availϭh␯
ϪD0 H C–H͒,
͑4͒
͑
2
193 nm
A representative H-atom ͑m/eϭ1) TOF spectrum is
shown in Fig. 6 with the time-dependent background sub-
tracted. The solid line fit to the data is using the P(ET͒ ob-
tained from fitting the CH2 ͑m/eϭ14–12͒ TOF spectra dem-
onstrating that this is the momentum matched fragment.
Since H-atom elimination from both acetone9 and
azomethane10 is thought to occur in minor yields at 193 nm,
the importance of measuring the momentum matching part-
ner cannot be overemphasized in unambiguous determina-
tion of the dissociation channels. Similar m/eϭ1 TOF spec-
tra were obtained in the pyrolysis of azomethane.
Furthermore, lowering the temperature of the SiC nozzle
Ͼ100 °C eliminated the H-atom peak entirely. Shown in
Fig. 7 are laser on and laser off spectra at m/eϭ2, each
accumulated for 2 000 000 shots. Within the signal-to-noise
ratio of the data we conclude that there is no H2 produced
from the photodissociation of CH3 at 193.3 nm. This is in
agreement with the analysis of the heavy fragments. Based
FIG. 7. TOF spectrum of m/eϭ2 ͑Hϩ2 ͒ at 90 °. The open circles are the data
with the laser on and the solid line is the data taken with the laser off.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 102, No. 2, 8 January 1995
130.88.90.110 On: Sat, 20 Dec 2014 15:35:04

796
North et al.: Photodissociation of the methyl radical
The laboratory angular distribution for the CH2 frag-
ments, monitored at m/eϭ12, is shown in Fig. 8. Each point
represents several individual measurements corrected for the
laser power and the error bars are one standard deviation of
the data. The low time-dependent background from the
pulsed valve at m/eϭ12 ͑Cϩ͒ made it the preferable mass to
detect. The laser was polarized in the plane of the detector
and along the molecular beam axis. For each measurement
the total TOF profile was integrated with a scaler and the
laser off signal was subtracted from the laser on signal at
alternate shots. Since the TOF spectrum at m/eϭ12 involves
contributions from both methyl radical and acetone photodis-
sociation, the acetone signal needed to be carefully taken
into account. In order to do this counts were taken at angles
⌰
lab Ͼ20 ° where only the acetone contribution was present.
Using the P(ET͒ for CH3 from acetone and an isotropic dis-
sociation as has been observed previously, the angular distri-
bution for these fragments28,29 was simulated using the for-
ward convolution program. The simulation was then
normalized to the measured signal at large laboratory angles
and then the acetone contribution subtracted from the m/e
ϭ12 signal. TOF spectra taken at m/eϭ15 provided confir-
mation of the acetone angular distribution. It should be noted
that given the large isotropic recoil of the CH3 fragments
from acetone photodissociation, the angular distribution is
almost invariant over the range of interest. The center-of-
mass product angular distribution is given by
FIG. 8. Laboratory angular distribution of the CH2 photofragments detected
at m/eϭ12 ͑Cϩ͒ with the laser polarized at 0 °. The lines are simulated
angular distributions assuming three different values of the anisotropy pa-
rameter ͑see text͒.
where h␯
is the photon energy ͑147.8 kcal/mole͒ and
193 nm
D0͑H2C–H͒ is the bond dissociation energy. From the
3
˜
known heats of formation for CH2͑X B1͒ ͑92.6
2
kcal/mole͒,31 H͑2S) ͑52.1 kcal/mole͒,32 and CH3͑X AЉ2͒
˜
͑34.0 kcal/mole͒,32 the C–H bond dissociation energy in CH3
1
P ␪͒ϭ
͑
1ϩ␤P cos ␪͒ ,
͑5͒
͓
͑
͔
2
is 110.7 kcal/mole. Therefore, the available energy for pro-
4␲
3
˜
duction of CH2͑X B1͒ϩH is 37.1 kcal/mole. The singlet-
triplet splitting in CH2 has been the subject of a voluminous
literature.33 Adopting the accepted best value of 9.08 kcal/
where ␤ is the anisotropy parameter and is equal to 2 if the
transition moment and the dissociating bond are parallel and
Ϫ1 if they are perpendicular in the limit of prompt recoil.35
Shown in Fig. 8 are three simulated laboratory angular dis-
tributions using the determined P(ET͒ and anisotropy param-
eters of ␤ϭ0.0, Ϫ0.5, and Ϫ1.0 corresponding to isotropic
and increasingly perpendicular dissociations. From a best fit
to the experimental angular distributions an anisotropy pa-
rameter of ␤ϭϪ0.9Ϯ0.1 has been determined. A negative ␤
parameter was confirmed independently by rotating the po-
larization of the laser and observing the m/eϭ1 signal al-
though the low signal-to-noise ratio precluded an accurate
quantitative measurement.
mole the available energy for the formation of
1
˜
CH2͑a A1͒ϩH is then, 147.8Ϫ110.7Ϫ9.08ϭ28.0 kcal/
mole.
Equation ͑4͒ has neglected any internal excitation of the
parent. Although we have no way of ascertaining the internal
excitation in the CH3 at 1400 K, we estimate that the methyl
radicals should contain ϳ6 kcal/mole of vibrational energy
prior to supersonic expansion, with most of this residing in
the low frequency ␯ umbrella mode. Chen et al. have re-
2
ported the complete vibrational relaxation of allyl radical
produced under similar conditions34 and even partial relax-
ation would leave the nascent methyl radicals with only 1–2
kcal/mole of vibrational energy on average. The maximum of
the translational energy distribution, 28 kcal/mole, extends to
the allowed thermodynamic limit for the formation of
IV. DISCUSSION
A. Laser polarization dependence
1
2
˜
CH2͑a A1͒ϩH͑ S͒. Although the fastest CH2 fragments
overlap with the slowest CH3 fragments from acetone in the
m/eϭ12–14 spectra, the H-atom TOF spectrum and the an-
gular distribution ͑Fig. 8͒ allow sensitivity to the P(ET͒
maximum. We are confident, therefore, that the P(ET͒ does
not have a significant contribution beyond 28 kcal/mole. Ow-
ing to high time dependent background at small laboratory
angles, ⌰labϭ6.0 was the minimum angle at which TOF data
was recorded. Based on the kinematics, CH2 photofragments
with less than 7 kcal/mole of translational energy could not
be detected. Fortunately, the H-atom TOF spectra provides
reliable information about the P(ET͒ down to 1.5 kcal/mole.
At 193.3 nm methyl radical is excited to the dissociative
continuum above the barrier along the C–H bond dissocia-
tion coordinate. Even at excitation energies below this bar-
rier the CH3 lifetime is subpicosecond. Kelly and co-workers
have determined predissociation lifetimes of the CH3 origin
band to be Ͻ82 fs ͑Ref. 12͒ and the ͓1000͔ state was found
to have a lifetime of ϳ13 fs which is comparable to the C–H
stretching frequency.13 By nature of the continuum excitation
at 193.3 nm the dissociation should occur with extreme ra-
pidity. It is therefore unlikely that internal conversion could
˜
compete effectively with dissociation from the B state and
the observed anisotropy indicates that the dissociation is di-
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 102, No. 2, 8 January 1995
130.88.90.110 On: Sat, 20 Dec 2014 15:35:04

North et al.: Photodissociation of the methyl radical
797
1
3
˜
˜
rect. The measured ␤ϭϪ0.9Ϯ0.1 indicates that the transition
dipole moment lies perpendicular to the C–H bond which is
consistent with the initial excitation of the out-of-plane pz
orbital on the carbon to a 3s Rydberg orbital. This represents
the first measurement of the photofragment anisotropy for
CH2͑a A1͒ ␪HCHϭ102.4 ° and for CH2͑X B1͒ ␪HCHϭ134 °.
Although Dixon and co-workers have now concluded15 that
vibrationless singlet methylene is produced from methyl
3
˜
radical photodissociation at 216 nm ͓rather than CH2 ͑X B1͒
with 0-3 quanta in ␯2 excitation͔,3 the present results indicate
that singlet or triplet methylene are formed with copious in-
ternal energy. According to the measured translational energy
˜
methyl radical dissociation from the B state. Although the
dissociation is much faster than rotation of the parent mol-
ecule, any motion that contributes a velocity component per-
pendicular to the C–H bond will diminish the observed an-
isotropy. This effect would be particularly salient if the
methyl radical dissociates from a range of nonplanar geom-
etries. Using the expression, ␤ϭ2P2͑⌰͒, where ⌰ represents
the angle between the C–H bond and the dipole moment
operator this range can be calculated. On average, the methyl
would have to dissociate from geometries 10 ° from planarity
with 75°Ͻ⌰Ͻ90° coinciding to the spread in the measured
anisotropy ͑Ϫ0.8 to Ϫ1.0͒. Any umbrella motion originating
from either the transition at 193.3 nm or a hot band absorp-
tion of radicals already possessing ␯2 excitation would result
in a range of dissociations from nonplanar geometries.
3
˜
distribution there would be no CH2͑X B1͒ containing less
3
˜
than 8 kcal/mole of internal energy, with most CH2͑X B1͒
fragments possessing 23.1 kcal/mole of internal energy on
average. If, however, singlet methylene were the dissociation
product it would contain ϳ13 kcal/mole of internal energy
on average. Since the P(ET͒ extends to the thermodynamic
limit there is a finite probability of forming vibrationless
1
1
˜
˜
CH2͑a A1͒. Although the slope of the CH2͑a A1͒ bending
coordinate in the exit channel is not as steep as for
3
˜
CH2͑X B1͒ the breadth of the P(ET͒ suggests that the me-
thyl radical samples a wide range of dissociative geometries.
The absorption to the dissociative continuum is sufficiently
far above the barrier to be relatively insensitive to the least
motion pathway. Therefore, the observed P(ET͒ is certainly
consistent with excitation in the other vibrational modes of
B. Product translational energy distribution
the methylene. We conclude that the most likely dissociation
˜
The barrier on the B state arises from a ‘‘de-
1
˜
product is CH2͑a A1͒ which is consistent with theoretical
Rydbergization’’ of the 3s orbital initially localized on the
prediction.2 In addition, wave packet calculations on empiri-
cal potential energy surfaces find that formation of triplet
methylene requires an anomalously large coupling between
*
carbon as it evolves into a ␴ orbital on the C–H bond, and
eventually to a 1s orbital on the departing hydrogen atom.
The disposal of available energy into translation of the
photofragments should be dominated by the repulsive inter-
the ground and excited state surfaces.38
1
˜
In light of the vibrationless CH2͑a A1͒ observed by
*
action as the excited state develops ␴ character in the C–H
Dixon and co-workers, one must ask why the resulting en-
ergy distribution at 193.3 nm is so different. The apparent
discrepancy can be explained by considering the fundamen-
tal differences between the dynamics imposed by a con-
tinuum compared to a tunneling dissociation. Since methyl
radical at 216 nm is excited to the band origin only zero
point motion will contribute to the range of dissociative ge-
ometries. In addition, the dissociation barrier depends on the
HCH angle and to an extent, so should the tunneling rate.
Therefore, the methyl radical excited at 216 nm dissociates
from an extremely narrow range of geometries which are
bond. Given the 6.29 kcal/mole barrier height13 and the 12.7
kcal/mole exoergicity of the reaction to form
CH2͑a A1)ϩH, the total exit barrier is ϳ19 kcal/mole. If
1
˜
ground state methylene is produced then the exit barrier is
increased by the magnitude of the singlet triplet splitting and
is therefore ϳ28 kcal/mole. For a dissociation involving the
cleavage of a single bond the fraction of the exit barrier
appearing in translation can be estimated by using an impul-
sive model.36 With the soft fragment impulsive model ϳ98%
of the exit barrier would be expected to appear in product
translation.37 In fact, this partitioning was predicted earlier
by Yu et al.2 and is intuitively reasonable since the light H
atom should not effectively couple its impulse into the CH2
vibrational degrees of freedom. Furthermore, the near planar-
biased toward HCH angles Ͻ120°. These out going trajecto-
1
˜
ries run close to the minimum on the CH2͑a A1͒ bending
potential ͑␪HCHϭ102 °͒ and therefore result in predominantly
vibrationless nascent products. At 193.3 nm there is no a
priori reason to believe that the dissociation should be con-
strained to a small dispersion about the CH3 equilibrium ge-
ometry. The range of dissociation geometries should be de-
termined, in part, by the initial motion of the nuclei upon
excitation to the continuum. Therefore, there is no preference
for outgoing trajectories to sample that part of the potential
energy surface that corresponds to a least motion pathway. If
the dissociation proceeds through a broad range of coordi-
nate space a high degree of internal excitation in the CH2
products should result.
˜
ity of the B state should not facilitate any rotational excita-
tion of the methylene fragment and even if the methyl radical
dissociated from nonplanar geometries the very small exit
impact parameters when an H atom recoils from the C atom
in CH2 would still result in negligible CH2 rotation. How-
ever, only ϳ15 kcal/mole on average is observed in transla-
tion. This corresponds to only 80% and 54% of the exit
barrier for singlet and triplet methylene respectively indicat-
ing substantial internal excitation. If the dissociation is
prompt, as supported by the exhibited photofragment aniso-
tropy, and there is little interaction in the exit channel, then
the internal energy distribution may be determined by the
change in RC–H and ␪HCH between the CH3 and the free CH2.
Although there is little change in the C–H bond length, there
is a substantial change in the HCH bond angle. The methyl
radical is of D3h symmetry and ␪HCHϭ120°, while for
V. CONCLUDING REMARKS
In conclusion, the photodissociation of methyl radical at
193 nm has been studied using photofragment translational
spectroscopy. The only single photon products observed
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 102, No. 2, 8 January 1995
130.88.90.110 On: Sat, 20 Dec 2014 15:35:04

798
North et al.: Photodissociation of the methyl radical
were CH2 and H and is the first time that the branching ratio
between H atom and H2 elimination from CH3 photochemis-
try has been directly determined. Although we cannot unam-
biguously assign the spin state of the methylene product, we
12 S. G. Westre. P. B. Kelly, Y. P. Zhang, and L. D. Zeigler, J. Chem. Phys.
94͑1͒, 270 ͑1991͒.
13 S. G. Westre, T. E. Gansberg, P. B. Kelly, and L. D. Zeigler, J. Phys.
Chem. 96, 3610 ͑1992͒.
14 P. Botschwina, E. Schick, and M. Horn, J. Chem. Phys. 98, 9125 ͑1993͒.
15 S. H. S. Wilson, J. D. Howe, K. N. Rosser, M. N. R. Ashfold, and R. N.
Dixon, Chem. Phys. Lett. 227, 456 ͑1994͒.
believe, in light of the translational energy distribution and
1
˜
overwhelming theoretical work, that CH2͑a A1͒ is the pre-
16 The apparent high resolution of the TOF spectrum in Ref. 3 could not be
reproduced in this laboratory using the forward convolution technique.
17 S. H. S. Wilson, M. N. R. Ashfold, and R. N. Dixon, J. Chem. Phys. ͑to be
published͒.
dominate dissociation product. Since a definitive determina-
tion of the methylene spin-state has yet to be made at any
wavelength, any measurement to this end would represent an
important contribution to the further elucidation of methyl
radical photochemistry. The breadth of internal energy mea-
sured in this experiment is consistent with the dissociation
occurring from a wide range of dissociative geometries.
18 Y. T. Lee, J. D. McDonald, P. R. LeBreton, and D. R. Herschbach, Rev.
Sci. Instrum. 40, 1402 ͑1969͒.
19 S. W. North, C. A. Longfellow, and Y. T. Lee, J. Chem. Phys. 99, 4423
͑1993͒.
20 D. Proch and T. Trickl, Rev. Sci. Instrum. 60, 713 ͑1989͒.
21 D. W. Kohn, H. Clauberg, and P. Chen, Rev. Sci. Instrum. 63͑8͒, 4003
͑1992͒; H. Clauberg, D. W. Minsek, and P. Chen, J. Am. Chem. Soc. 114,
99 ͑1992͒; H. Clauberg and P. Chen, ibid. 113, 1445 ͑1991͒.
22 T. G. DiGiuseppe, J. W. Hudgens, and M. C. Lin, J. Phys. Chem. 86, 36
͑1982͒; G. N. Robinson, G. M. Nathanson, R. E. Continetti, and Y. T. Lee,
J. Chem. Phys. 89, 6744 ͑1988͒; C. Yamada, E. Hirota, and K. Kawaguchi,
ibid. 75, 5256 ͑1981͒.
˜
Work is currently in progress to examine the B-state photo-
chemistry in greater detail with a higher resolution TOF tech-
nique and at energies below the excited state barrier.
ACKNOWLEDGMENTS
The authors would like to thank Dr. A. Suits for many
helfpul discussions. They would also like to acknowledge Dr.
D. Stranges and Dr. M. Stemmler for their help with the
pyrolytic radical source. This work was supported by the
Director, Office of Energy Research, Office of Basic Energy
Sciences, Chemical Sciences Division of the U.S. Depart-
ment of Energy under Contract No. DE-AC03-76SF00098.
23 R. E. Continetti, Ph.D. thesis, University of California, Berkeley, 1989; R.
E. Continetti, B. A. Balko, and Y. T. Lee, Chem. Phys. Lett. 184, 400
͑1991͒; J. Chem. Phys. 93, 5719 ͑1990͒.
24 B. A. Balko, Ph.D. thesis, University of California, Berkely, 1991; B. A.
Balko, J. Zhang, and Y. T. Lee, J. Chem. Phys. 94, 7958 ͑1991͒.
25 A. M. Wodtke, Ph.D. thesis, University of California, Berkeley, 1986; X.
Zhao, Ph.D. thesis, University of California, Berkeley, 1988; X. Zhao, G.
M. Nathanson, and Y. T. Lee, Acta Physico-Chim. Sinica 8, 70 ͑1992͒.
26 R. Renaud and L. C. Leitch, Can. J. Chem. 32, 545 ͑1954͒.
27 G. A. Gaines, D. J. Donaldson, S. J. Strickler, and V. Vaida, J. Phys.
Chem. 92, 2762 ͑1988͒.
1 M. T. Macpherson, M. J. Pilling, and M. J. C. Smith, J. Phys. Chem. 89,
2268 ͑1985͒; J. E. Butler, L. P. Gross, M. C. Lin, and J. W. Hudgens,
Chem. Phys. Lett. 63, 104 ͑1979͒; M. R. Berman and M. C. Lin, J. Chem.
Phys. 81, 5743 ͑1984͒; S. Zabarnick, J. W. Fleming, and M. C. Lin, ibid.
85, 4373 ͑1986͒; M. Aoyagi, R. Shepard, A. F. Wagner, T. H. Dunning, Jr.,
and F. B. Brown, J. Phys. Chem. 94, 3236 ͑1990͒; B. R. Brooks, H. F.
Schaefer, III, J. Chem. Phys. 67, 5146 ͑1977͒.
28 S. W. North, D. A. Blank, J. D. Gezelter, C. A. Longfellow, and Y. T. Lee
͑unpublished͒.
29 K. A. Trentelman, S. H. Kable, D. B. Moss, and P. L. Houston, J. Chem.
Phys. 91, 7498 ͑1989͒.
30 S. W. North, P. Chu, and Y. T. Lee, XIVth International Symposium on
Molecular Beams, Asilomar, CA, 1992 ͑unpublished͒.
31 D. R. Stull and H. Prophet et al., JANAF Thermochemical Tables, 2nd
Ed., Natl. Bur. Stand. Ref. Data Ser., Natl. Bur. Stand. ͑U.S. GPO, Wash-
ington, D.C., 1971͒; S. Zabarnick, J. W. Fleming, and M. C. Lin, J. Chem.
Phys. 85, 4373 ͑1986͒.
2 P. Chen, S. D. Colson, and W. A. Chupka, Chem. Phys. Lett. 147, 466
͑1988͒.
3 S. H. S. Wilson, M. N. R. Ashfold, and R. N. Dixon, Chem. Phys. Lett.
222, 457 ͑1994͒.
4 H. T. Yu, A. Sevin, E. Kassab, and E. M. Evleth, J. Chem. Phys. 80͑5͒,
2049 ͑1984͒.
32 S. W. Benson, Thermochemical Kinetics, 2nd Ed. ͑Wiley, New York,
1976͒.
5 G. Herzberg and J. Shoosmith, Can. J. Phys. 34, 523 ͑1956͒.
6 G. Herzberg, Proc. R. Soc. London, Ser. A 262, 291 ͑1961͒.
7 A. B. Callear and M. P. Metcalfe, Chem. Phys. 14, 275 ͑1976͒.
8 C. Ye, M. Suto, and L. C. Lee, J. Chem. Phys. 89͑5͒, 2797 ͑1988͒.
9 P. D. Lightfoot, S. P. Kirwan, and M. J. Pilling, J. Phys. Chem. 92, 4938
͑1988͒.
33 I. Shavitt, Tetrahedron, 41͑8͒, 1531 ͑1985͒, and references therein.
34 J. A. Blush, D. W. Minsek, and P. Chen, J. Phys. Chem. 96, 10150 ͑1992͒.
35 R. N. Zare, Ph.D. thesis, Harvard University, 1964; R. N. Zare, Mol.
Photochem. 4, 1 ͑1974͒.
36 G. E. Busch and K. R. Wilson, J. Chem. Phys. 56, 3626 ͑1972͒.
37 The expression for the fraction of the available energy partitioned into
10 J. E. Baggott, M. Brouard, M. A. Coles, A. Davis, P. D. Lightfoot, M. T.
Macpherson, and M. J. Pilling, J. Phys. Chem. 91, 317 ͑1987͒.
11 C. Kassner, P. Heinrich, F. Stuhl, S. Couris, and S. Haritakis, Chem. Phys.
Lett. 208, 27 ͑1993͒.
translation is given by, fTϭ͓␮C–H/ ␮ –H], where ␮ is the reduced mass.
CH
2
38 R. N. Dixon ͑unpublished͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 102, No. 2, 8 January 1995
130.88.90.110 On: Sat, 20 Dec 2014 15:35:04