Collisional relaxation of highly excited vibrational levels of I2(X)

The relaxation dynamics of vibrationally excited levels of I2(X) are of relevance to the chemical oxygen iodine laser. We have investigated relaxation of 12(vM>20) by pulsed laser techniques. Individual ro-vibrational levels were populated by stimulated emission pumping. Collisionally populated levels were monitored via laser excitation of the D-X transition. Preliminary rate constants for vibrational relaxation of 12(vH>20) by Ar and Hz0 are reported. Improved spectroscopic constants for the D state are also presented. Introduction. At present, the Chemical Oxygen Iodine Laser (COIL) is the only demonstrated example of a chemically pumped electronic transition laserl2. In this device chemically generated 0 2 1A dissociates molecular iodine and excites the atoms to lasing by sequential processes. The mechanism by which I2 is dissociated in this laser has been the subject of many investigations3-10, and some controversy. Even now, the mechanism is not fully understood. As 02(lA) does not posses sufficient energy to dissociate I2 in a single transfer event, multiple collisions must be involved. Current theories identify, as a possible first step, the electronic to vibrational energy transfer (E-V) process4$ I2(X) + 02(1A) -+ 12(X, v"e33-44) + 02(3Z-) (1) Subsequent collisions of 02(1A) with I2(X, vV>20) can result in direct dissociation, or electronic excitation of I2 to the A'(2,), A(lU), or B(O,+) stated-11. There are then several channels, generalized by the equation 02(1A) + 12* + 21(2P3/2) + 02(3Z-) (2) that can contribute to the dissociation4-11. Here, 12* denotes vibrationally ( ~ 5 2 0 ) or electronically excited (A', A, or B state) 12. Once atomic 1 is liberated it is rapidly excited by the process U2p3j2) + 02(lA) -+ I (2P i~ ) + 02(3Z-) (3) In addition to being the species responsible for lasing, there is good evidence that I(2pln) participates in the dissociation via the reactions4+5 I(2PlD) + I2(X) -+ 12(XI ~"033-44) + I(2P3/2) (4) and I(2p1/2) + I2* -j 3IPp312) (5) Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jp4:19944202 04-730 JOURNAL DE PHYSIQUE IV The presence of 12* in 12/02(1A) mixtures has been confirmed by recording chemiluminescence4~6~~~ and laser induced fluorescence (LIF) spectra6,8,9,11. Populations in the B and A states were readily observed via the B+X and A+X transitions. For example, Cerny et al.10 recently used chernilurninescence to characterize the steady-state rovibrational population distributions in the A and B states. The presence of population in the non-fluorescing I ~ ( A ' ) ~ and I2@, vn>20)6,8911 states has been detected using LIF. In probing the dissociation of I2 by 02(lA), it is of obvious value to show that I2(X,vU>20), 12(A'), I2(A), and I;?(B) are present in 1~/0~(a lA) mixtures. Unfortunately, this is not sufficient to determine the mechanism. For this purpose, measurements of the rates at which these states are populated and depleted are needed. The present work is primarily concerned with the relaxation dynamics of I2(X, vW>20). A data base for the rates at which various species deactivate I2(X, vM>20) is an essential pre-requisite for realistic models of the dissociation mechanism and laser performance2.7. Beyond mechanistic elucidation, measurements of energy transfer rate constants can be used to identify undesirable species in the laser, and facilitate the development of strategies for their minimization or removal. At this time, little is known about the relaxation rates for highly excited levels of I2(X). Hall et al.5 used reaction (4) to generate vibrationally excited 12, and they obtained rate constants for vibrational relaxation of vW=40 by He and Ar. Abramson et a1.12 demonstrated that the relaxation dynamics could be studied using stimulated emission pumping (SEP) techniques, but they did not determine rate constants. More recently, the authors13 used SEP state preparation, and LIF probing techniques to examine the collisional relaxation dynamics of Iz(X, vU=42, Jw=17). We now report measurements made by selectively populating the levels vS=38, J"=49, and v"=23, J"=57. As in the flow-tube studies of Crozet et al.8, and our previous SEPLIF measurementsl3, we have used the D-X transition to detect population in ro-vibrationally excited levels. There are definite advantages to using the D-X transition, rather than the more familiar B-X system, for this purpose. The D-X transition is much more intense than the B-X, the shorter D state lifetime (-10 ns) renders it less susceptible to collisional quenching, and the fluorescence is spectrally well-removed from B-X and A-X emissions. However, in the course of the present work we found that the existing electronic term energy and vibrational constantsl418 for the D state did not accurately predict the band origins of transitions terminating on the D, v1<50 levels. Consequently, we also report a re-analysis of the low energy levels of