AB-INITIO CONFIGURATION-INTERACTION CALCULATIONS OF THE PREDISSOCIATION OF ROVIBRATIONAL LEVELS OF THE C(3)PI(G) AND D(1)PI(G)3S-SIGMA RYDBERG STATES OF THE OXYGEN MOLECULE
Y. Li et al., AB-INITIO CONFIGURATION-INTERACTION CALCULATIONS OF THE PREDISSOCIATION OF ROVIBRATIONAL LEVELS OF THE C(3)PI(G) AND D(1)PI(G)3S-SIGMA RYDBERG STATES OF THE OXYGEN MOLECULE, The Journal of chemical physics, 106(3), 1997, pp. 1123-1133
Ab initio configuration interaction calculations have been carried out
for seven low-lying states of the oxygen molecule, Three different ty
pes of nonadiabatic couplings have been considered: spin-orbit, radial
, and rotational. The complex scaling method has been employed to comp
ute rovibrational level locations and predissociation linewidths with
a basis of 200 Hermite polynomials for each of 13 different Omega elec
tronic states. The calculations correctly predict that the nu = 2 leve
l has the narrowest linewidth for the O-16(2) C (3) Pi(g) state, while
nu = 4 is narrowest for O-18(2). Marked variations in the linewidths
of the different Omega components of the C state are explained by the
fact that the pi --> 3s sigma Rydberg and sigma --> pi* valence (3) P
i(g) states have different occupations of the pi orbital, causing opp
osite orderings of their respective Omega levels. Rotational, coupling
is found to be important for high J values of the C state. The d (1)
Pi(g) 3s sigma state shows ever more unusual effects by virtue of the
fact that there is a sharply avoided crossing between the. correspondi
ng Rydberg diabatic state with a bound sigma --> pi (1) Pi(g) valence
state. The calculations find irregular spacings in the d-state vibrat
ional manifold, wide variations in linewidth for different nu,J levels
, and a large change in the rotational constant in successive vibratio
nal levels, all of which effects have been earlier demonstrated in exp
erimental work. Satellite lines are indicated for both the nu = 2 and
3 levels as a result of the interaction with the bound (1) Pi(g) valen
ce state, whereby experimental verification exists only for nu = 2. Th
e nu = 3 state has not yet been successfully identified due to the bro
adness of the d-X spectrum in the energy range of interest. The observ
ed temperature dependence of the linewidths of the two features near t
he expected location of the nu = 2 level can also be understood on the
basis of these calculations. Finally, the change in the predissociati
on mechanism for the d state from spin-orbit to radial as nu changes f
rom 0 to 2 which has been deduced experimentally is also verified in t
he present theoretical treatment.