Rotational relaxation of molecular nitrogen in a freely expanding jet

Abstract

In connection with the assumption about the possibility of producing a gasdynamic laser at rotational transitions of diatomic molecules (1, 2] and a gasdynamic condense-laser by using the phenomenon of condensation to produce the population inversion [3–5], a quantitative description of the kinetics of the rotational relaxation of the simplest diatomic molecules is necessary. In contrast to the traditional approach in gasdynamics and the theory of transport processes, when one parameter (the rotational relaxation time) is used in the description of rotational relaxation, a substantially more detailed description at the population level of the individual rotational states is required in solving spectroscopy and laser physics problems. This paper is devoted to a theoretical and experimental investigation of the rotational relaxation of nitrogen in a free, low density jet under conditions when a substantial nonequilibrium holds in the rotational level populations. On the basis of representations developed earlier [1, 6], a model is constructed for a relaxing gas which yields the magnitude of the population of individual rotational levels. The selection of a molecular nitrogen free jet as a subject for investigation is explained by the fact that the gasdynamics of such a flow has been studied well [7]. Moreover, at this time diagnostic methods have been developed to determine the molecule concentrations at many (k≤20) rotational levels [8]; hence, a jet is a good object on which a detailed comparison between theory and experiment can be made, as is done in this paper. Separation of the translational and rotational relaxation processes is allowed in the theoretical description of the flow with relaxation studied in this paper on the basis of the fact that the buildup of a Maxwell molecule velocity distribution because of elastic collisions occurs more rapidly than the redistribution of molecules at the rotational levels because of inelastic collisions. Such an approach is apparently valid for molecular hydrogen with a large value of the rotational quantum [9]. A model with separation of the processes is hypothetical for nitrogen molecules whose rotational constant is ≈ 1/20 that for hydrogen, and is confirmed in this paper by comparing computations with experiment. It is hence assumed that rotational relaxation proceeds in an N₂ molecule stream with a known temperature, density, and velocity distribution obtained from measurements.