Vibrational Dynamics of the I
3
Radical: A Semiempirical Potential Surface, and
Semiclassical Calculation of the Anion Photoelectron Spectrum
C. J. Margulis, D. A. Horner, S. Bonella, and D. F. Coker*
Department of Chemistry, Boston UniVersity, 590 Commonwealth AVenue, Boston, Massachusetts 02215
ReceiVed: July 26, 1999; In Final Form: September 24, 1999
The semiempirical diatomics in molecules (DIM) approach is used to model the potential surface for ground-
state vibration of a linear I
3
molecule. We use this system to explore semiclassical methods for treating
quantal nuclear vibrations by computing the photoelectron spectrum of I
3
-
which produces vibrationally excited
I
3
. We compare semiclassical results with full quantum calculations and experimental results recently reported
by Neumark and co-workers. (Taylor, T. R.; Asmis, K.R.; Zanni, M. T.; Neumark, D. M. J. Chem. Phys.
1999, 110, 7607.)
1. Introduction
Since the 1950s the existence of the triiodide radical has been
proposed as an intermediate species to explain how iodine atoms,
produced by photodissociating I
2
, can recombine in the gas phase
to reproduce the diatomic species. The accepted mechanism
involves an I
•
radical first colliding I
2
to form stable I
3
which
subsequently undergoes collision with another I
•
radical, then
this collision complex breaks apart to give two I
2
molecules.
Despite its proposed importance in this most fundamental
reaction of gas-phase chemical kinetics, direct experimental
observation of the I
3
molecule has only very recently been
accomplished in the high-resolution photoelectron spectroscopy
studies of Neumark and co-workers.
1
The first goal of this paper is to demonstrate that a very simple
description of the I
3
molecule offered by the semiempirical
diatomics-in-molecules (DIM) approach is actually capable of
providing a very reliable representation of this molecule,
reproducing the recently measured binding energies, and
vibrational frequencies with surprising accuracy.
Next we summarize how time dependent perturbation theory
can be used to compute the distribution of ejected photoelectron
kinetic energy P
(ǫ) in the thermal equilibrium photoelectron
spectrum of I
3
-
in which vibrationally hot I
3
is produced
according to the following process:
Finally we will compare the results of fully quantum dynamical
calculations of this photoelectron spectrum, with classical and
semiclassical calculations of the I
3
vibrational dynamics probed
by these measurements, and we compare our theoretical results
with the experimental photoelectron spectrum of Neumark and
co-workers.
1
2. Methods
2.1. A DIM Potential Model for Ground Electronic State
Intramolecular Vibrations of I
3
. The model assumes that the
I
3
molecule is linear, hence the projection of the total angular
momentum of the system into the molecular axis is a good
quantum number. Following the same scheme used in previous
work
2
we write the basis set as Hund’s case C kets
where J
k
and m
jk
are the total angular momentum and it’s
projection on the molecular axis for each of the iodine atoms.
Our purpose in this paper is to focus on the properties of the
ground electronic state of I
3
, hence we will reduce our basis
set to a minimum subspace that will include only the necessary
kets mixing to form this lowest energy eigenstate. Thus the
approach we present here will not produce all the subsequent
excited states, but only some of them.
We suppose that the ground-state eigenket has total angular
momentum projection
so only kets satisfying this condition will combine to generate
this assumed lowest energy eigenstate.
3
Our knowledge of the
ground-state dissociation limits of the I
3
radical can be used to
further limit the basis set. Thus if one of the bonds is stretched,
the molecule should dissociate to ground-state I
2
and an I
•
radical
species. The ground state of I
2
in our representation is
and the I
•
radical ground state is
These considerations thus enable us to limit the basis kets to
only those having all the J
k
)
3
/
2
and the m
jk
)(
1
/
2
. This
procedure is only valid in the gas phase, in solution other states
that we are not including in the calculations described here will
be coupled by anisotropic interactions with the solvent and thus
make contributions to the lowest energy solution phase eigenket,
but this is beyond the scope of the current paper.
The DIM Hamiltonian operator has the form
4,5
and we choose the zero in energy to be that of the isolated I
•
radicals; therefore, the second sum in the above expression can
be disregarded. For convenience in the notation we will drop
the J index in the angular momentum expression since it is the
I
3
-
(therm) + pν f I
3
(vib) + e
-
(ǫ) (1)
|J
1
m
j1
〉|J
2
m
j2
〉|J
3
m
j3
〉 ) |J
1
m
j1
, J
2
m
j2
, J
3
m
j3
〉
M
J
)
∑
k)1
3
m
jk
)(
1
/
2
1/ 2(|
3
/
2
,
1
/
2
〉|
3
/
2
,-
1
/
2
〉 - |
3
/
2
,-
1
/
2
〉|
3
/
2
,
1
/
2
〉)
|
3
/
2
(
1
/
2
〉
∑
i<j
H ˆ
i, j
- n
∑
i
H ˆ
i
(2)
9552 J. Phys. Chem. A 1999, 103, 9552-9563
10.1021/jp992596m CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/06/1999