Two- and three-body photodissociation of gas phase I3 -

Article abstract of DOI:10.1063/1.1691017

The photodissociation dynamics of gas phase I^-₃ by using a fast beam photofragment translational spectrometer was examined. It was stated that the photofragment translational spectrometer was coupled to a coincidence imaging detector that enabled the direct detection and analysis of two and three neutron or anion fragments from single dissociation events. The three-body dissociation yielding I^-+2I(²P_3/2) photofragments was also seen throughout the energy range probed. Analysis shows that the three-body decay dynamics was dominated by synchronous concerted dissociation.

Full text of DOI:10.1063/1.1691017

Two- and three-body photodissociation of gas phase I 3 −
Alexandra A. Hoops, Jason R. Gascooke, Ann Elise Faulhaber, Kathryn E. Kautzman, and Daniel M. Neumark
Citation: The Journal of Chemical Physics 120, 7901 (2004); doi: 10.1063/1.1691017
View online: http://dx.doi.org/10.1063/1.1691017
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/120/17?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Two- and three-body photodissociation dynamics of diiodobromide (I2Br−) anion
J. Chem. Phys. 134, 134306 (2011); 10.1063/1.3571474
Site-dependent photodissociation of vibronically excited CD 3 NH 2 molecules
J. Chem. Phys. 132, 244310 (2010); 10.1063/1.3447383
Photodissociation dynamics of acetoxime in gas phase
J. Chem. Phys. 122, 184322 (2005); 10.1063/1.1897369
A velocity map ion-imaging study on ketene photodissociation at 208 and 213 nm : Rotational dependence of
product angular anisotropy
J. Chem. Phys. 122, 104309 (2005); 10.1063/1.1858435
Photodissociation dynamics of IBr − ( CO 2 ) n , n
J. Chem. Phys. 122, 054307 (2005); 10.1063/1.1839178
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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57

JOURNAL OF CHEMICAL PHYSICS
VOLUME 120, NUMBER 17
1 MAY 2004
Two- and three-body photodissociation of gas phase I3À
Alexandra A. Hoops, Jason R. Gascooke,a) Ann Elise Faulhaber, Kathryn E. Kautzman,
and Daniel M. Neumarkb)
Department of Chemistry, University of California, Berkeley, California, 94720 and Chemical Sciences
Division, Lawrence Berkeley National Laboratories, Berkeley, California 94720
͑Received 2 September 2003; accepted 9 February 2004͒
The photodissociation dynamics of IϪ3 from 390 to 290 nm ͑3.18 to 4.28 eV͒ have been investigated
using fast beam photofragment translational spectroscopy in which the products are detected and
analyzed with coincidence imaging. At photon energies р3.87 eV, two-body dissociation that
generates IϪϩI2(A 3⌸1u) and vibrationally excited I2Ϫ(X 2⌺ϩu )ϩI(2P3/2) is observed, while at
Ϫ
ϩ
2
energies у3.87 eV, I ( P1/2)ϩI2 (X 2⌺u ) is the primary two-body dissociation channel. In
addition, three-body dissociation yielding IϪϩ2I(2P3/2) photofragments is seen throughout the
energy range probed; this is the dominant channel at all but the lowest photon energy. Analysis of
the three-body dissociation events indicates that this channel results primarily from a synchronous
concerted decay mechanism. © 2004 American Institute of Physics. ͓DOI: 10.1063/1.1691017͔
*
I. INTRODUCTION
few hundred fs in a coherent superposition of vibrational
states with
Ϸ12. Additionally, evidence of IϪ production
v
͗ ͘
The tri-iodide anion, IϪ3 , is a rare example of a molecule
whose photodissociation dynamics have been studied in de-
tail both in solution and in the gas phase, thereby presenting
a unique opportunity for understanding how isolated, gas
phase molecular photophysics can be modified by solvation
effects. While much of the original work on IϪ3 was per-
formed in condensed phases, several recent gas phase experi-
ments have provided significant insight into the energetics,
spectroscopy, and photodissociation dynamics of this linear,
closed-shell anion. However, experiments performed in dif-
ferent laboratories1,2 suggest that significant questions re-
main regarding the primary photochemistry of the tri-iodide
anion, particularly concerning the branching ratios among
available product channels. In this paper, the photodissocia-
tion dynamics of IϪ3 are investigated with an upgraded instru-
ment in order to resolve this and other issues.
via three-body dissociation was observed at 266 nm.17 These
experiments have stimulated several theoretical studies on
the electronic states and photodissociation dynamics of
IϪ3 .18–24
In the gas phase, Do et al.25 used collision-induced dis-
sociation of IϪ3 in a tandem mass spectrometer to determine a
dissociation energy of 1.31Ϯ0.06 eV for decay to IϪ
ϩI2(X 1⌺ϩg ). From this value, the spin–orbit coupling con-
stant of the iodine atom, the dissociation energies of I2 and
IϪ2 ,26,27 and the term energies of the A and B states of I2 ,28,29
one can construct the energy level diagram for IϪ3 and its
possible photofragments, Fig. 1. The lower and upper bands
of IϪ3 are also depicted in Fig. 1. There are numerous decay
pathways energetically accessible following excitation in the
UV:
IϪ3 →IϪ 1S͒ϩI X 1⌺ϩ͒ D0ϭ1.31Ϯ0.06 eV
͑1͒
͑2͒
͑3͒
͑4͒
͑5͒
͑6͒
͑
͑
2
g
The spectroscopy and decay dynamics of IϪ3 in solution
have been examined extensively. Its absorption spectrum3
comprises two broad bands centered at approximately 360
and 290 nm, henceforth referred to as the lower and upper
bands, that have been assigned to transitions from the ⌺g
ground state to admixtures of
→I P ϩIϪ X 2⌺
͒
͒
D0ϭ1.85Ϯ0.06 eV
ϩ
2
͑
͑
3/2
2
u
→IϪ 1S͒ϩI A 3⌸
͒
D0ϭ2.66Ϯ0.06 eV
→I P1/2͒ϩIϪ X 2⌺ ͒ D0ϭ2.79Ϯ0.06 eV
͑
͑
2
1u
ϩ
u
2
*
͑
͑
ϩ
1
2
2
→IϪ 1S͒ϩ2I P
͒
D0ϭ2.85Ϯ0.06 eV
3
1
͑
͑
⌸
and
⌺
excited
3/2
0ϩu
0ϩu
states.4–6 Raman spectra at laser frequencies resonant with
either band show extended progressions in the IϪ3 symmetric
stretch,7–10 indicating that initial nuclear motion in the upper
states occurs along this coordinate. Ruhman and
co-workers11–14 and, more recently, Vohringer and
co-workers15–17 have used femtosecond laser techniques to
probe the photodissociation of IϪ3 in ethanol and other sol-
vents via excitation of both electronic bands. Their findings
revealed that the IϪ2 photoproduct was generated within a
→IϪ 1S͒ϩI B 3⌸ϩ
͒
D0ϭ3.26Ϯ0.06 eV
͑
͑
2
0u
→IϪ 1S͒ϩI P ͒ϩI P1/2
͒
D0ϭ3.80Ϯ0.06 eV
2
2
*
͑
͑
͑
3/2
͑7͒
Time- and frequency-domain experiments on the photo-
dissociation and photoelectron spectroscopy of IϪ3 have been
carried out in our laboratories.1,30,31 The time-resolved pho-
todissociation of gas phase IϪ3 at 390 nm was studied by
Zanni et al.30 using femtosecond photoelectron ͑FPE͒ spec-
troscopy. Excitation of IϪ3 at this wavelength resulted in the
production of IϪ and I2Ϫ with approximately equal yields.
The FPE spectra showed that the IϪ2 photoproduct was vi-
a͒
Current address: School of Chemistry and Physics, University of Adelaide,
SA, 5005, Australia.
b͒
Author to whom correspondence should be addressed. Electronic mail:
brating coherently with
ϭ67, a considerably higher value
v
͗ ͘
dan@radon.cchem.berkeley.edu
than observed in solution. Zanni et al. constructed a collinear
0021-9606/2004/120(17)/7901/9/$22.00
7901
© 2004 American Institute of Physics
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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57

7902
J. Chem. Phys., Vol. 120, No. 17, 1 May 2004
Hoops et al.
FIG. 1. Energy level diagram for IϪ3 indicating the
ground state of IϪ3 relative to the asymptotic energies of
the fragment channels. Two-body dissociation channels
are presented to the left of IϪ3 and three-body dissocia-
tion channels to the right. The shaded area represents
the photofragment yield spectrum of gas-phase IϪ3 as
reported in Ref. 2.
was a combination of IϪϩI2 and IϩIϪ2 products. This inter-
pretation is supported indirectly by the recent theoretical
work of Vala et al.21 on the ground and excited state poten-
tials of IϪ3 . These calculations show that there are extensive
curve-crossings among the numerous excited state potentials,
implying that all energetically allowed dissociation channels
should occur to some extent.
excited state potential energy surface that reproduced the ob-
served level of IϪ2 excitation in quantum wave packet simu-
lations. These simulations also indicated that the IϪ was gen-
erated via three-body dissociation, rather than from an I2
ϩIϪ product channel.
Choi et al.1 carried out further investigations of the pho-
todissociation spectroscopy and dynamics of gas phase IϪ3 in
the range of 240–420 nm. The photofragment yield ͑PFY͒
spectrum of IϪ3 revealed two intense features with maxima at
290 and 360 nm, in accordance with solution phase work.
Additionally, coincidence detection of the photofragments
using two-particle time- and position-sensing ͑TPS͒ ͑Ref.
32͒ at several wavelengths yielded product branching ratios
In this paper, we report results on the photodissociation
of IϪ3 obtained with a recently installed photofragment coin-
cidence imaging detector.33 The new design presents several
advantages, described in more detail in the following section,
over that used in the work of Choi et al.1 In particular, the
new detector is not restricted to the detection of only two
fragments per laser shot, and charged fragments can be de-
tected directly. These features result in better characterization
of both the two- and three-body dissociation channels of IϪ3 .
The new results provide evidence for IϪϩI2(A 3⌸1u) pro-
duction following excitation of the lower band, which ac-
counts for some of the discrepancies in the branching ratios
between the two earlier photodissociation studies.1,2 In addi-
and translational energy P(E ) distributions. These results
͓
͔
T
demonstrated that both two- and three-body dissociation oc-
curred over the entire wavelength range, as evidenced by
photoproduct mass ratios of 1:2 and 1:1. The two-body prod-
ucts from both the lower and upper bands were assigned to a
Ϫ
ϩ
Ϫ
ϩ
2
2
*
combination of IϩI2 (X ⌺u ) and I ϩI2 (X ⌺u ). How-
ever, only the mass ratio of these products ͑and not the
charge state͒ was unambiguously identified in these experi-
ments. Three-body dissociation was shown to occur through
synchronous concerted dissociation of IϪ3 , with the central
iodine atom having little momentum in the center-of-mass
frame. Finally, branching ratios between the two- and three-
body channels indicated that the two-body dissociation frac-
tion varied from 45% to 67%, depending on photon energy.
Zhu et al.2 have carried out a complementary study of I3Ϫ
photodissociation, in which the anion products from the dis-
sociation of IϪ3 were monitored by photofragment mass spec-
trometry, yielding the IϪ and IϪ2 branching fractions from
270 to 540 nm. Their PFY spectrum was quite similar to that
obtained by Choi et al.,1 but the IϪ2 :IϪ ratio for the lower
band was considerably smaller than the two-body:three-body
ratio found by Choi. As a result, Zhu et al. proposed that the
two-body channel seen by Choi et al. from the lower band
Ϫ
ϩ
2
*
tion, we find no evidence for production of I ϩI2 (X ⌺u )
from excitation of the lower band, nor of IϩIϪ2 (X 2⌺uϩ) from
the upper band, in contrast to our earlier work. Finally, we
obtain considerably more information on the decay dynamics
of the three-body channels.
II. EXPERIMENT
The current photodissociation experiments were per-
formed on our fast beam photofragment translational spec-
trometer. The ion beam, mass-spectrometer, and laser inter-
action components of this instrument32,34,35 remain
unchanged from our earlier study of IϪ3 ,1 but the TPS
detector33 is quite different.
Rotationally and vibrationally cooled IϪ3 is produced by
flowing argon gas ͑14 psi͒ over iodine crystals at room tem-
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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57

Three-body dissociation of I3Ϫ
7903
J. Chem. Phys., Vol. 120, No. 17, 1 May 2004
FIG. 2. Schematic diagram of the coincidence imaging
detector. The paths of the photofragments and parent
anion are represented by dotted and dashed lines, re-
spectively.
perature and supersonically expanding the resulting gas mix-
ture through a pulsed piezoelectric valve. A 1 keV electron
beam intersects the free jet expansion downstream of the
valve orifice. The resulting anions are accelerated to a known
laboratory beam energy between 7 and 9 keV and mass-
selected using a Bakker time-of-flight mass spectrometer,
which imparts negligible kinetic energy spread to the ion
beam.36,37 An excimer-pumped dye laser with a bandwidth of
0.3 cmϪ1 is used to dissociate mass-selected ions at a specific
photon energy. The photofragments then strike a microchan-
nel plate ͑MCP͒ detector assembly with TPS capabilities for
each fragment. A movable beam block in front of the detec-
tor prevents undissociated ions from impacting the MCPs.
Consequently, any observed signal is due to the recoiling
photoproducts.
detected. The three-body distributions presented are the raw
P(ET) distributions.
The new detector presents several advantages with re-
spect to the IϪ3 experiment. First, the front MCP of the pre-
vious TPS detector,32 which employed a dual wedge-and-
strip anode for time- and position-sensing, was biased at a
high negative potential, precluding the detection of anions.
Therefore, coincidence detection of all photofragments re-
sulting from anion photodissociation required that any an-
ionic fragments be photodetached by absorption of an addi-
tional photon. In the case of IϪ3 , all of the photodissociation
studies1 were carried out at photon energies above the elec-
tron affinities of I and I2 ͑3.0591 eV and 2.524 eV,
respectively39͒, so that detachment of the anionic fragments
was energetically feasible. Nonetheless, this requirement was
an undesirable feature because one had to assume in con-
structing the P(ET) distributions that the detachment cross
sections for the IϪ2 photofragments were independent of in-
ternal excitation. In the new detector design, the front MCP
is at ground potential, which allows anionic ͑and neutral͒
fragments to be detected directly.
In addition, the previous TPS detector was restricted to
the detection of two photofragments per laser shot. Synchro-
nous concerted three-body dissociation of IϪ3 was inferred
from events in which the photofragment mass ratio was 1:1.
For such an event, the two detected fragments corresponding
to the outer I atoms were approximately equally displaced
from the center of the detector, while the third fragment,
corresponding to the central I atom, impacted the beam
block. The new detector permits the detection of three pho-
tofragments per dissociation event, as was demonstrated in
our recent study of IϪ2 •Ar photodissociation,33 yielding a
more complete picture of the three-body decay channel.
The TPS detector in use, Fig. 2, is a coincidence imaging
detector based on the design by Amitay and Zajfman.38 In
this detection scheme, the MCP assembly is coupled to a
phosphor screen. Each photofragment that impacts the MCP
assembly produces a spot on the phosphor screen that is then
detected by a charge-coupled device ͑CCD͒ camera and a
multianode photomultiplier tube ͑PMT͒. The positions of the
fragments are obtained from the CCD camera, while the
PMT provides relative timing information. For each dissocia-
tion event, the masses of the photoproducts, center-of-mass
translational energy release ET , and, for two-body dissocia-
tion, recoil angle ␪ are obtained from the fragments’ posi-
tions and arrival times. For three-body dissociation, the
center-of-mass momenta of all three particles are determined.
Based on the P(ET) distribution resulting from the dissocia-
tion of O2 B 3⌺Ϫu
(
ϭ7), the current translational energy
v
Ј
resolution of this arrangement is ⌬ET /ETϷ0.7%.33
The beam block reduces the detection efficiency of those
fragments that have low translational energy or values of ␪
close to 90° with vertically polarized laser light, while frag-
ments of high translational energy with ␪ close to 0° or 180°
miss the detector entirely. These effects are accounted for by
normalizing the raw translational energy distributions with a
detector acceptance function ͑DAF͒ for decay yielding two
fragments.32 The P(ET) distributions from two-body disso-
ciation reported here have been corrected with the DAF. Ow-
ing to the increased detector area and reduced size of the
beam block as compared to the prior TPS detector, the DAF
is relatively constant over the energy range for both two- and
three-body dissociation of IϪ3 for which two fragments are
III. RESULTS
The photodissociation of IϪ3 was investigated at photon
energies ranging from 3.18 to 4.28 eV ͑390–290 nm͒. There
are three possible types of valid events that can result from
the dissociation of IϪ3 : ͑i͒ two-body dissociation to IϪϩI2 or
IϩIϪ2 , which will give rise to a photofragment mass ratio of
1:2, ͑ii͒ three-body dissociation with all three fragments de-
tected, and ͑iii͒ three-body decay in which one of the frag-
ments either strikes the beam block or is not detected due to
less than unit detection efficiency, yielding a photofragment
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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57

7904
J. Chem. Phys., Vol. 120, No. 17, 1 May 2004
Hoops et al.
FIG. 3. Photofragment mass ratio of two-fragment valid events following
excitation of IϪ3 at 3.64 eV.
mass ratio of 1:1. From this point forward, the terms ‘‘two-
fragment’’ and ‘‘three-fragment’’ will refer to the detection of
two and three photoproducts from a dissociation event, while
the ‘‘two-body’’ and ‘‘three-body’’ dissociation channels are
those generating two and three iodine fragments regardless
of the number detected. At each excitation energy, all three
types of events are observed. Figure 3 presents an example
͑at h␯ϭ3.64 eV) of photoproduct mass ratios for valid coin-
cidence events resulting in the detection of two fragments,
and demonstrates well-separated contributions from mass ra-
tios 1:2 and 1:1.
The photofragment P(ET) distributions for the 1:2 mass
ratio products are presented in Fig. 4; Figs. 4͑a͒–4͑d͒ corre-
spond to the lower band, Figs. 4͑f͒–4͑g͒ to the upper band,
and Fig. 4͑e͒ to the minimum between the bands ͑see Fig. 1͒.
The brackets in each panel represent the allowed transla-
tional energies between the maximum and minimum, ETmax
and ETmin , for channels ͑1͒–͑4͒ and ͑6͒. The values for EmT ax
are determined from the dissociation energies presented in
Sec. I, while EmT in corresponds to an internal energy that is
equal to the bond dissociation energy of the diatomic frag-
ment. Within each band, the distributions shift towards
higher ET with increasing photon energy. However, there is a
qualitative change in the contours between the two bands. In
the lower band, no signal is seen below EmT in for channels
͑1͒–͑3͒, while in the upper band, no signal is seen above
EmT ax for channel ͑4͒. Hence, excitation of the two bands
leads to markedly different two-body channel distributions.
Figure 5 shows the P(ET) distributions for the three-
body dissociation products when either two or three frag-
ments are detected ͑dotted and solid lines, respectively͒ and
reflects the relative number of two- and three-fragment
events at each photon energy. The vertical arrows denote the
expected kinetic energy release based on the dissociation en-
FIG. 4. Photofragment translational energy distributions for two-body dis-
sociation of IϪ3 ͑solid curves͒. The allowed translational energies for chan-
nels ͑1͒–͑4͒ and ͑6͒ are bracketed. The dotted lines represent the fit of the
observed P(ET) distributions to Gaussian distributions for channel ͑3͒ that
have been truncated at EmT ax for channel ͑3͒. Photon energies in eV are ͑a͒
3.18, ͑b͒ 3.31, ͑c͒ 3.44, ͑d͒ 3.64, ͑e͒ 3.87, ͑f͒ 4.15, and ͑g͒ 4.28.
ergies for channels ͑5͒ and ͑7͒. Channel ͑7͒, which is acces-
sible at the two highest photon energies, is evident as a very
small amount of two-fragment signal. As with the two-body
dissociation channel, the P(ET) distributions shift to higher
translational energy with increasing photon energy. The
three-fragment distributions are generally broader than the
two-fragment distributions, but they become narrower with
increasing excitation energy. Furthermore, the fraction of
three-body dissociation events for which all three fragments
are detected generally increases with photon energy, as
shown in Table I.
The P(ET) distributions presented in Figs. 4 and 5 differ
from those previously reported by Choi et al.1 In that work,
the P(ET) distributions for the photofragments with a mass
TABLE I. Branching fractions for detection of two and three fragments
from the three-body dissociation of I3Ϫ
.
Photon energy ͑eV͒
2 fragments
3 fragments
3.18
3.31
3.44
3.64
3.87
4.15
4.28
0.74
0.81
0.58
0.52
0.48
0.36
0.39
0.26
0.19
0.42
0.48
0.52
0.64
0.61
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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57

Three-body dissociation of I3Ϫ
7905
J. Chem. Phys., Vol. 120, No. 17, 1 May 2004
ratio of 1:2 exhibited two features throughout the upper and
lower bands, labeled A and B, that were assigned to the I
Ϫ
ϩ
Ϫ
ϩ
2
2
*
ϩI2 (X ⌺u ) and I ϩI2 (X ⌺u ) channels, respectively. In
the current study, the distributions in Figs. 4͑a͒–4͑d͒ are
quite similar to feature A, but feature B formerly seen at low
ET is absent. In the upper band, we now find feature A at
high energy to be missing, and the minimum energy at which
signal is observed is considerably higher than in our earlier
study. In addition, the new distributions for the 1:1 mass ratio
events are considerably cleaner; the earlier results showed a
continuum rising toward low ET that is not present in the
current data.
Finally, analysis of the photofragment angular distribu-
tions yields an average anisotropy parameter40 of ␤ϭ1.4
Ϯ0.2 for both the 1:2 and 1:1 mass channels throughout the
photon energy range probed. This value is consistent with
previous results indicating that transitions to the upper and
lower bands are parallel.1,5
IV. ANALYSIS
A. Two-body dissociation
The photofragment translational energy distributions
originating from excitation of the lower band of IϪ3 , Figs.
4͑a͒–4͑d͒, are peaked in the region of EmT in for the IϪ
ϩI2(X 1⌺ϩg ), IϩI2Ϫ(X 2⌺uϩ), and IϪϩI2(A 3⌸1u) channels,
indicating an internally excited diatomic fragment. The sharp
decrease in intensity below EmT in signifies that one or more of
these channels contribute to the distributions, and that the
FIG. 5. Photofragment translational energy distributions for three-body dis-
sociation of IϪ3 . The expected kinetic energy release for channels ͑5͒ and ͑7͒
are indicated by vertical arrows. The solid and dotted lines represent the
distributions resulting from the detection of all three and two fragments,
respectively. Photon energies in eV are ͑a͒ 3.18, ͑b͒ 3.31, ͑c͒ 3.44, ͑d͒ 3.64,
͑e͒ 3.87, ͑f͒ 4.15, and ͑g͒ 4.28.
Ϫ
ϩ
ϩ
2
Ϫ
3
*
I ϩI2 (X ⌺u ) and I ϩI2(B ⌸0u) channels are absent. At
a photon energy of 3.18 eV ͓Fig. 4͑a͔͒, the P(ET) distribu-
tion has two partially resolved features; one between ETmin
and ETmax for channel ͑3͒, IϪϩI2(A 3⌸1u), and a second fea-
ture that extends to considerably higher energy, but not be-
yond ETmax , for channel ͑2͒, IϩI2Ϫ(X 2⌺ϩu ). Hence, we assign
the lower energy feature to channel ͑3͒ and the higher energy
feature to vibrationally and rotationally excited IϪ2 ϩI. These
two contributions are apparent but less resolved at 3.31 eV
͓Fig. 4͑b͔͒. In Figs. 4͑c͒ and 4͑d͒, the two channels can no
longer be resolved and the distributions appear as a single
feature that is likely a combination of channels ͑2͒ and ͑3͒.
There is no evidence for channel ͑1͒, IϪϩI2(X 1⌺gϩ), in the
form of signal at ETϾEmT ax for channel ͑2͒, although there
could be a contribution from this channel with very highly
excited I2 .
photon energy that lies between the two bands results in a
P(ET) distribution, Fig. 4͑e͒, that appears to have contribu-
tions from channels ͑2͒, ͑3͒, and ͑4͒.
B. Three-body dissociation
At all excitation energies, channel ͑5͒, IϪϩ2I(2P3/2), is
the dominant three-body dissociation pathway. The P(ET)
distributions resulting from this channel should appear at a
single energy, but the results in Fig. 5 clearly show that the
two- and three-fragment distributions extend asymmetrically
towards values of ET that are higher than expected. This
observation is attributed to internal excitation of the IϪ3 in the
ion beam. However, one would also like to understand why
the three-fragment signal becomes more dominant at higher
photon energies ͑Table I͒, why the three-fragment distribu-
tions generally extend to higher ET , and why the two- and
three-fragment distributions differ the most at the lowest
photon energies.
These trends can be explained as follows. Two-fragment
events with photofragment mass ratios of 1:1 result from
dissociation trajectories in which the center-of-mass mo-
menta of the two outer I atoms are nearly equal and opposite;
the central iodine has very little momentum and strikes the
beam block. In a three-fragment event, the central I atom
must have enough momentum to scatter out of the beam,
miss the beam block, and impact the detector. For ions with
The P(ET) distributions from the upper band, Figs. 4͑f͒
and 4͑g͒, exhibit broader features that extend to EmT ax for
Ϫ
ϩ
2
*
channel ͑4͒, I ϩI2 (X ⌺u ). The relatively sharp onset to
dissociation in the region of maximum kinetic energy release
for this channel and the near lack of signal in the region
corresponding to channels ͑1͒–͑3͒ indicate that channel ͑4͒
has become the principal dissociation pathway. The photo-
fragment translational energy distributions reveal greater in-
tensity near EmT ax for channel ͑4͒, signifying less vibrational
excitation in the diatomic fragment than for channel ͑2͒ from
the lower band. There is intensity in the energetically al-
lowed region for channel ͑6͒, IϪϩI2(B 3⌸0ϩu); thus, this
channel may contribute as well. Finally, excitation at 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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57

7906
J. Chem. Phys., Vol. 120, No. 17, 1 May 2004
Hoops et al.
TABLE II. Branching fractions for two- and three-body dissociation of I3Ϫ
and IϪ2 :IϪ branching ratio.
IϪ2 :IϪ
Photon energy ͑eV͒
Two-body
Three-body
3.18
3.31
3.44
3.64
3.87
4.15
4.28
0.64
0.60
0.49
0.37
0.30
0.45
0.45
0.36
0.40
0.51
0.63
0.70
0.55
0.55
1:1.7
1:3.2
1:3.6
1:4.9
1:2.3
1:1.2
1:1.2
ciation of IϪ3 are indistinguishable in the current experiment,
the momentum partitioning exhibits sixfold symmetry,44–46
and all events can be plotted in one-sixth of the inscribed
circle. Figure 6͑b͒ shows the relevant sector of the Dalitz
plots at higher magnification for four photon energies. Point
A represents synchronous concerted dissociation from a lin-
ear geometry in which both I–I bonds break simultaneously.
This produces two fragments with equal and opposite mo-
menta and a third that has zero momentum. There is intensity
not only in the region of point A, but also extending slightly
away from point A, which corresponds to events in which
one of the fragments has very low momentum and there is
unequal momentum partitioning between the remaining two
iodine atoms. The implications of the Dalitz plots will be
discussed in Sec. V.
FIG. 6. Dalitz plots of momentum partitioning among the three fragments in
the three-body dissociation of I3Ϫ . ͑a͒ Due to conservation of momentum
and energy, all points must lie within the circle inscribed in the triangle. The
notation I(i) in the axis labels indicates the relative momenta of the three
fragments, with ‘‘1’’ being the highest and ‘‘3’’ being the lowest. Point A
corresponds to those events with zero momentum for I(3) and equal but
opposite momenta in the remaining two I fragments. The photon energy is
3.18 eV. ͑b͒ Enlarged view of the Dalitz plots at selected photon energies.
C. Branching fractions
The branching fractions for the two- and three-body dis-
sociation channels were calculated at all photon energies in-
vestigated, with the results presented in Table II. The two-
body dissociation channel is represented by all two-fragment
events with a 1:2 mass ratio, while the three-body channel is
the sum of the three-fragment events and the 1:1 mass ratio
two-fragment events. At all but the lowest two photon ener-
gies, three-body dissociation is the favored decay pathway.
Additionally, the results in Table II reveal a decrease in the
fraction of two-body dissociation signal with increasing ex-
citation energy within the lower band. The last column, list-
ing the IϪ2 :IϪ branching ratios, is obtained using the proce-
dure described in Sec. V.
no initial internal energy, three-fragment events become
more likely at higher photon energies, since more transla-
tional energy is available to all three photofragments. At a
fixed photon energy, photodissociation of internally excited
ions should also favor three-fragment over two-fragment
events, especially if the bending and antisymmetric stretch-
ing modes of the ions are excited. Thus, at the lowest photon
energies in Fig. 5, three-body photodissociation of cold ions
leads preferentially to two-fragment events, while only those
ions that are vibrationally hot yield three-fragment events.
Consequently, the P(ET) distributions for two- and three-
fragment events differ the most at low photon energy. As the
photon energy is raised, three-fragment events become pos-
sible for a larger fraction of the total ensemble of ions, re-
sulting in the observation of a higher fraction of three-
fragment events and a reduction in the differences between
the P(ET) distributions from two- and three-fragment
events.
D. Comparison to earlier results
The results reported in this section differ in several re-
spects from those previously reported.1 We formerly asserted
Ϫ
ϩ
Ϫ
ϩ
2
2
*
that both IϩI2 (X ⌺u ) and I ϩI2 (X ⌺u ) were produced
regardless of which band was excited, in contrast to the as-
signments in Sec. IV A. The source of this discrepancy lies in
the detection method employed in the previous study. As
discussed in Sec. II, the TPS detector used in the earlier work
precluded the direct detection and analysis of anion photo-
products; photodetachment of an electron from the anion
fragment was required for detection. If a single dissociation
event occurred without detachment, then the detection of a
lone fragment would not constitute a coincidence event.
However, in the instance that two dissociation events re-
sulted from a single laser shot with detection of only the
The three-fragment events can be analyzed with the aid
of the Dalitz plots41–43 presented in Fig. 6. These plots illus-
trate the momentum partitioning among the three fragments
for each dissociation event. Conservation of momentum and
energy dictate that all events lie within the inscribed circle in
Fig. 6͑a͒. Moreover, as the three fragments from the disso-
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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57

Three-body dissociation of I3Ϫ
7907
J. Chem. Phys., Vol. 120, No. 17, 1 May 2004
lower band, their IϪ2 branching fractions were significantly
lower than our previously reported two-body fractions,1 and
proposed that this discrepancy resulted from a contribution
to the two-body signal from a channel yielding IϪ. The im-
proved two-body branching fractions in Table II of this paper
are closer to the IϪ2 branching fractions of Zhu et al. but
remain higher for all photon energies of the lower band. With
this in mind, we can examine the composition of the two-
body channel more closely and try to reconcile the various
measurements.
neutral fragments, then these detected fragments would be
treated as a coincidence event in the experiment. While many
such occurrences were discarded as false coincidences, a sig-
nificant number were accepted by the discrimination routine,
particularly at low ET where the displacements of the frag-
ments from the center of the detector are small. Thus, we
attribute the differences between the previously reported
two-body P(ET) distributions1 and those currently presented
to these false coincidences in which in which two tri-iodide
anions were dissociated but the anion fragments were not
detected. Consequently, the previous conclusion that the I
As discussed in Sec. IV A, there is strong evidence that
channel ͑3͒, IϪϩI2(A 3⌸1u), contributes to the two-body
signal, particularly at an excitation energy of 3.18 eV where
there are two distinct contributions to the P(ET) distribution,
Fig. 4͑a͒. If 22% of the total two-body signal in Fig. 4͑a͒ is
from channel ͑3͒, with the remainder from channel ͑2͒, then
the IϪ2 :IϪ ratio would be unity, in agreement with the results
of Zanni et al., whereas 69% of this signal would have to
originate from channel ͑3͒ to match the results of Zhu et al.
at 3.18 eV. By fitting the P(ET) distribution in Fig. 4͑a͒ with
the assumptions that the contribution from channel ͑3͒ is a
truncated Gaussian whose contour matches the P(ET) distri-
bution at low ET , and that the remainder of the photofrag-
ment translational energy distribution arises from channel
͑2͒, we find that 43% of the signal is due to channel ͑3͒. The
resulting IϪ2 :IϪ ratio, 1:1.7, lies between the values obtained
by Zanni et al. and Zhu et al. A similar analysis of the data in
Figs. 4͑b͒–4͑d͒ yields channel ͑3͒ contributions of 60%,
56%, and 54%, respectively. The curves used to obtain these
values are depicted as dotted lines in Figs. 4͑a͒–4͑d͒ and
have been truncated at ETmax for channel ͑3͒. The resulting
IϪ2 :IϪ ratios, presented in Table II, are now close to ͑al-
though systematically slightly higher than͒ the ratios ob-
tained by Zhu et al. Therefore, the two sets of measurements
are largely consistent provided that channel ͑3͒ is the two-
body source of IϪ. The remaining differences may result
from either the uncertainty associated with this analysis, as
the contributions from channels ͑2͒ and ͑3͒ are not cleanly
resolved in Figs. 4͑b͒–4͑d͒, or from a small contribution to
the two-body signal from channel ͑1͒ with highly excited I2 .
In the upper band, our branching fractions for two- and
three-body decay are in accordance with the IϪ2 and IϪ
branching fractions measured by Zhu et al. This is consistent
Ϫ
ϩ
Ϫ
ϩ
2
2
*
ϩI2 (X ⌺u ) and I ϩI2 (X ⌺u ) channels were present at
all photon energies is incorrect. False coincidences of this
type also gave rise to the intensity at low ET in the 1:1 mass
ratio P(ET) distributions of Choi et al.1 that does not appear
in the current results.
The improved analysis procedure afforded by the new
detector results in different branching fractions for the two-
and three-body decay channels than those previously
reported.1 While we originally claimed that two-body disso-
ciation was more prevalent from 3.44 to 4.15 eV, we now
find that three-body decay is the dominant channel in this
energy range.
V. DISCUSSION
The current experiments demonstrate that two- and
three-body dissociation occur in both the upper and lower
bands of gas phase IϪ3 . In the lower band, the observed
P(ET) distributions for two-body dissociation comprise con-
tributions from the IϩIϪ2 (X 2⌺uϩ) and IϪϩI2(A 3⌸1u) chan-
nels, while in the upper band, the primary two-body disso-
Ϫ
ϩ
2
*
ciation channel is the I ϩI2 (X ⌺u ) channel with the
possibility of a contribution from the IϪϩI2(B 3⌸0ϩu) chan-
nel. The three-body decay process, IϪϩ2I(2P3/2), is present
at all excitation energies; it is the predominant channel
throughout the upper band and at higher energies within the
lower band. The P(ET) distributions within the lower band
extend down to EmT in for channels ͑2͒ and ͑3͒, indicating that
the I2 or IϪ2 fragments are produced with internal energies
from zero to their dissociation energies, after which point
three-body dissociation occurs. Hence, there appears to be a
smooth transition from two- to three-body dissociation
within the lower band. The P(ET) distributions for the upper
band, on the other hand, are most intense near EmT ax for chan-
nel ͑4͒, indicating significantly less internal excitation of the
IϪ2 product than for channel ͑2͒ from the lower band.
*
with our assignment of the two-body channel as I
ϩIϪ2 (X 1⌺ϩg ). A small contribution to the two-body signal
from channel ͑6͒ would yield slightly better agreement, but
the discrepancies are not large enough to confirm the pres-
ence of channel ͑6͒.
It is of interest to compare the branching fractions for the
two- and three-body decay channels in Table II to the IϪ and
IϪ2 branching fractions of Zanni et al.30 and Zhu et al.2 If one
were to assume that three-body dissociation is the sole
source of IϪ, then the IϪ2 :IϪ ratio would be identical to the
two-body:three-body branching ratio. At 3.18 eV ͑390 nm͒,
the two-body:three-body ratio reported in Table II is 1.8:1.
Zanni et al. found approximately equal production of IϪ2 and
IϪ, while Zhu et al. determined the IϪ2 :IϪ ratio to be ap-
proximately 1:4. Zhu et al. also reported a decrease in the
IϪ2 :IϪ ratio across the lower band profile, reaching values as
low as 1:10 around 3.7 eV. They pointed out that for the
Looking more globally at the two-body channels, our
finding that excitation of the lower band yields ground state
iodine atoms while excitation of the upper band produces
2
*
I ( P1/2) atoms is consistent with some of the earliest cor-
relations proposed for I3Ϫ photodissociation,47 as well as the
recent adiabatic potentials calculated by Vala et al.21 using
the diatomics-in-molecules ͑DIM͒ approximation including
spin–orbit coupling. However, as pointed out by those au-
thors, there are significant problems with the DIM curves for
Ϫ
3
I3 , including an inverted ordering of the excited ⌸u and
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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57

7908
J. Chem. Phys., Vol. 120, No. 17, 1 May 2004
Hoops et al.
ϩ
u
1
⌺
states responsible for the lower and upper bands, re-
synchronous concerted decay mechanism previously inferred
from the two-fragment measurements is largely valid, but
also more accurately and completely reflect the dissociation
dynamics. The plots result from a single decay mechanism
with the internal motion of IϪ3 causing the distribution to be
spread about point A. There is no evidence of a competing
mechanism, such as sequential or asynchronous concerted
decay, as was observed in the collisional dissociation of
Naϩ3 .45 A secondary decay mechanism of this type would
yield an additional region of intensity on the diameter of the
inscribed circle removed from point A. In addition, the rela-
tively smooth evolution of the Dalitz plots from the lower
band to the upper band suggests similar decay dynamics for
channel ͑5͒ regardless of which electronic state is excited.
Thus, it appears that the initial symmetric stretch motion on
the two excited states, as evidenced by the symmetric stretch
progressions in the resonance Raman spectra of IϪ3 ,7–9 maps
onto the asymptotic three-body dissociation dynamics.
spectively.
Assignment of the IϪϩI2(A 3⌸1u) channel as a signifi-
cant decay pathway from the lower band represents a new
wrinkle on the overall dynamics. The potential energy curves
in the Franck–Condon region for excitation of IϪ3 calculated
by Vala et al.21 demonstrate that the lower optically active
excited state lies very close to two potential energy curves
that correlate to the IϪϩI2(A 3⌸1u) channel asymptote.
Thus, nonadiabatic coupling to these states may be respon-
sible for the production of channel ͑3͒. In contrast, the upper
state dissociation dynamics appear to be more adiabatic from
*
I
the perspective of two-body dissociation, with
ϩIϪ2 (X 1⌺ϩg ) being the dominant products.
Turning to three-body dissociation, the current results, in
which both two- and three-fragment events are detected,
yield a more complete picture of the overall dynamics than
was possible in our previous study1 that only permitted the
detection of two-fragment events. The previous work was
strongly biased towards detection of synchronous concerted
dissociation of IϪ3 with equal momentum partitioning be-
tween the observed iodine fragments. One can now deter-
mine, by means of the Dalitz plots in Fig. 6, the extent to
which this type of dissociation dominates the three-body
channel.
The interpretation of Dalitz plots for the case of three
particles of equal mass has been discussed at length by
Muller et al.,44 Babikov et al.,45 and others.42,46,48 Each point
within the inscribed circle represents a mapping of the
center-of-mass momentum vectors for the three photofrag-
ments, subject to the constraint that the sum of the vectors is
zero. As mentioned in Sec. IV B, point A in Fig. 6͑a͒ corre-
sponds to two atoms having equal and opposite momentum
and the third having zero momentum, the result expected for
synchronous concerted dissociation of a linear triatomic mol-
ecule. As one moves from point A towards the center of the
circle, the three-body dissociation becomes progressively
less collinear; the tips of the three vectors evolve from a line
to an obtuse isosceles triangle, becoming an equilateral tri-
angle at the center. On the other hand, moving along the
diameter of the circle away from point A corresponds to all
three vectors being collinear but increasingly unequal in
magnitude, representing progressively more asymmetric col-
linear dissociation.
Figure 6 shows that at 3.18 eV, events are scattered
throughout the sector of the inscribed circle but are most
dense in the region near point A, indicating the presence of
synchronous concerted dissociation from linear IϪ3 . Addi-
tionally, Fig. 6͑b͒ illustrates that the concentration of events
in this region of low momentum for the lowest momentum
fragment increases with photon energy, corresponding to a
greater proportion of three-fragment events originating from
cold IϪ3 as was discussed in Sec. IV B. The Dalitz plots also
reveal a number of events corresponding to noncollinear and
asymmetric collinear dissociation. Thus, the bending and an-
tisymmetric stretching vibrations of IϪ3 are influencing the
decay dynamics through both zero point motion and initial
population of excited vibrational levels.
VI. CONCLUSIONS
The photodissociation dynamics of gas phase IϪ3 have
been examined using a fast beam photofragment translational
spectrometer coupled to a coincidence imaging detector that
enables the direct detection and analysis of two and three
neutral or anion fragments from single dissociation events.
The new detection scheme results in a more complete and
accurate picture of the dissociation dynamics than we ob-
tained in an earlier study.1
Throughout the entire energy range investigated, 3.18–
4.28 eV, two- and three-body dissociation channels are de-
tected, with three-body dissociation dominating at all but the
lowest photon energy. Although many decay pathways are
energetically accessible at the excitation energies employed
in this study, the two-body fragmentation from the lower
band leads mainly to the production of I(2P3/2
ϩIϪ2 (X 2⌺ϩu ) and IϪϩI2(A 3⌸1u), and the upper band re-
)
Ϫ
ϩ
2
sults in I ( P1/2)ϩI2 (X 2⌺u ) photofragments. The assign-
ment of a significant fraction of the two-body products from
the lower band to IϪϩI2(A 3⌸1u) yields branching fractions
for IϪ2 and IϪ that are consistent with those measured by Zhu
et al.2 by photofragment tandem mass spectrometry.
Finally, the observed three-body decay at all photon
energies arises primarily from IϪϩ2I(2P3/2) photoproducts.
Analysis of the three-body events using Dalitz plots il-
lustrates that the three-body decay dynamics are dominated
by synchronous concerted dissociation but are influenced by
the reactant bend and antisymmetric stretch vibrational
motions.
*
ACKNOWLEDGMENTS
This research was supported by the Director, Office of
Basic Energy Science, Chemical Sciences Division of the
U.S. Department of Energy under Contract No. DE-AC03-
76SF00098. K.E.K. is a National Science Foundation Pre-
doctoral Fellow.
Overall, the Dalitz plots not only demonstrate 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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57

Three-body dissociation of I3Ϫ
7909
J. Chem. Phys., Vol. 120, No. 17, 1 May 2004
1 H. Choi, R. T. Bise, A. A. Hoops, and D. M. Neumark, J. Chem. Phys.
113, 2255 ͑2000͒.
28 D. R. T. Appadoo, R. J. LeRoy, P. F. Bernath, S. Gerstenkorn, P. Luc, J.
Verges, J. Sinzelle, J. Chevillard, and Y. DAignaux, J. Chem. Phys. 104,
903 ͑1996͒.
2 L. Zhu, K. Takahashi, M. Saeki, T. Tsukuda, and T. Nagata, Chem. Phys.
Lett. 350, 233 ͑2001͒.
29 P. Luc, J. Mol. Spectrosc. 80, 41 ͑1980͒.
3 A. D. Awtrey and R. E. Connick, J. Am. Chem. Soc. 73, 1842 ͑1951͒.
4 T. Okada and J. Hata, Mol. Phys. 43, 1151 ͑1981͒.
5 M. Mizuno, J. Tanaka, and I. Harada, J. Phys. Chem. 85, 1789 ͑1981͒.
6 H. Isci and W. R. Mason, Inorg. Chem. 24, 271 ͑1985͒.
7 K. Kaya, N. Mikami, M. Ito, and Y. Udagawa, Chem. Phys. Lett. 16, 151
͑1972͒.
30 M. T. Zanni, B. J. Greenblatt, A. V. Davis, and D. M. Neumark, J. Chem.
Phys. 111, 2991 ͑1999͒.
31 T. R. Taylor, K. R. Asmis, M. T. Zanni, and D. M. Neumark, J. Chem.
Phys. 110, 7607 ͑1999͒.
32 R. E. Continetti, D. R. Cyr, D. L. Osborn, D. J. Leahy, and D. M. Neu-
mark, J. Chem. Phys. 99, 2616 ͑1993͒.
8 W. Kiefer and H. J. Bernstein, Chem. Phys. Lett. 16, 5 ͑1972͒.
9 A. E. Johnson and A. B. Myers, J. Phys. Chem. 100, 7778 ͑1996͒.
10 A. E. Johnson and A. B. Myers, J. Chem. Phys. 104, 2497 ͑1996͒.
11 U. Banin, A. Waldman, and S. Ruhman, J. Chem. Phys. 96, 2416 ͑1992͒.
12 U. Banin and S. Ruhman, J. Chem. Phys. 98, 4391 ͑1993͒.
13 U. Banin and S. Ruhman, J. Chem. Phys. 99, 9318 ͑1993͒.
14 E. Gershgoren, E. Gordon, and S. Ruhman, J. Chem. Phys. 106, 4806
͑1997͒.
33 A. A. Hoops, J. R. Gascooke, A. E. Faulhaber, K. E. Kautzman, and D. M.
Neumark, Chem. Phys. Lett. 374, 235 ͑2003͒.
34 D. J. Leahy, D. L. Osborn, D. R. Cyr, and D. M. Neumark, J. Chem. Phys.
103, 2495 ͑1995͒.
35 D. L. Osborn, H. Choi, D. H. Mordaunt, R. T. Bise, D. M. Neumark, and
C. M. Rohlfing, J. Chem. Phys. 106, 3049 ͑1997͒.
36 J. M. B. Bakker, J. Phys. E 6, 785 ͑1973͒.
15 T. Kuhne and P. Vohringer, J. Chem. Phys. 105, 10788 ͑1996͒.
16 T. Kuhne and P. Vohringer, J. Phys. Chem. A 102, 4177 ͑1998͒.
17 T. Kuhne, R. Kuster, and P. Vohringer, Chem. Phys. 233, 161 ͑1998͒.
18 I. Benjamin, U. Banin, and S. Ruhman, J. Chem. Phys. 98, 8337 ͑1993͒.
19 G. Ashkenazi, R. Kosloff, S. Ruhman, and H. Talezer, J. Chem. Phys. 103,
10005 ͑1995͒.
37 J. M. B. Bakker, J. Phys. E 7, 364 ͑1974͒.
38 Z. Amitay and D. Zajfman, Rev. Sci. Instrum. 68, 1387 ͑1997͒.
39 J. C. Rienstra-Kiracofe, G. S. Tschumper, H. F. Schaefer, S. Nandi, and G.
B. Ellison, Chem. Rev. 102, 231 ͑2002͒.
40 R. N. Zare, Mol. Photochem. 4, 1 ͑1972͒.
41 R. H. Dalitz, Philos. Mag. 44, 1068 ͑1953͒.
20 R. M. Lynden-Bell, R. Kosloff, S. Ruhman, D. Danovich, and A. Vala, J.
Chem. Phys. 109, 9928 ͑1998͒.
42 L. M. Wiese, O. Yenen, B. Thaden, and D. H. Jaecks, Phys. Rev. Lett. 79,
4982 ͑1997͒.
21 J. Vala, R. Kosloff, and J. N. Harvey, J. Chem. Phys. 114, 7413 ͑2001͒.
22 D. Danovich, J. Hrusak, and S. Shaik, Chem. Phys. Lett. 233, 249 ͑1995͒.
23 S. B. Sharp and G. I. Gellene, J. Phys. Chem. 101, 2192 ͑1997͒.
24 C. J. Margulis, D. F. Coker, and R. M. Lynden-Bell, J. Chem. Phys. 114,
367 ͑2001͒.
43 C. Maul and K. H. Gericke, J. Phys. Chem. 104, 2531 ͑2000͒.
44 U. Muller, T. Eckert, M. Braun, and H. Helm, Phys. Rev. Lett. 83, 2718
͑1999͒.
45 D. Babikov, E. A. Gislason, M. Sizun, F. Aguillon, V. Sidis, M. Barat, J. C.
Brenot, J. A. Fayeton, and Y. J. Picard, J. Chem. Phys. 116, 4871 ͑2002͒.
46 T. G. Clements, A. K. Luong, and R. E. Continetti, Chem. Phys. Lett. 366,
650 ͑2002͒.
25 K. Do, T. P. Klein, C. A. Pommerening, and L. S. Sunderlin, J. Am. Soc.
Mass Spectrom. 8, 688 ͑1997͒.
26 K. P. Huber and G. Herzberg, Molecular Spectra and Molecular Structure
IV Constants of Diatomic Molecules ͑Van Nostrand Reinhold, New York,
1979͒.
47 D. Meyerstein and A. Treinin, Trans. Faraday Soc. 59, 1114 ͑1962͒.
48 D. Strasser, L. Lammich, S. Krohn, M. Lange, H. Kreckel, J. Levin, D.
Schwalm, Z. Vager, R. Wester, A. Wolf, and D. Zajfman, Phys. Rev. Lett.
86, 779 ͑2001͒.
27 M. T. Zanni, T. R. Taylor, B. J. Greenblatt, B. Soep, and D. M. Neumark,
J. Chem. Phys. 107, 7613 ͑1997͒.
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:
129.24.51.181 On: Sun, 30 Nov 2014 17:18:57