General discussion

Abstract

University of Bristol) began the discussion with a question to Prof. Dalgarno: When the molecub lar signals you mentioned were observed following the recent explosion of a supernova, how was it known that these signals emanated from gas associated with the supernova? Prof. A. Dalgarno (Harvard-Smithsonian Center for Astrophysics) replied: Spectra were taken in the direction of the supernova on frequent occasions during a period of years. The velocities inferred from the line profiles matched those expected from the expanding ejecta.Prof. J. P. Simons (University of Nottingham) addressed Prof. Dalgarno: You remarked on your surprise at the appearance of molecular species only 100 days after a supernova expansion. Does that imply that there are no models for this behaviour at present? Prof. Dalgarno replied: The initial temperature of the supernova was over a million degrees. That it cooled sufficiently as it expanded so that molecules could survive after 100 days was at first sight surprising. However, it can be explained by a plausible model of the thermal balance and the chemistry for molecular formation. Dr. D. Flower (University of Durham) said: It is generally believed that the value of the ratio n(C): n(CO), predicted by equilibrium models of dark molecular clouds, is much less than observed. However, this ratio can be large [n(C) : n(C0) z 0.11 under conditions in which charge transfer reactions, particularly with H + , rather than proton transfers by H3+ dominate the ion-molecule chemistry. This is the case when the rate of destruction of H3+ by dissociative recombination with electrons is greater than or equal to its rate of removal through protonation reactions, principally with 0 and CO. Then, any further increase in the electron density leads to a decrease in the fractional abundance of H; and hence of 0, and H,O, which are produced in series of ionneutral reactions which originate in the protonation of 0 by H:. As charge transfer to 0, and H,O is a significant sink for H+, the fractional abundance of H + rises when that of H3+ falls. The consequent rise in the fractional electron density pushes HZ down further, and so the transition from the 'protonation-dominated' to the 'charge transferdominated' regime occurs abruptly.'The location of this transition depends on the value of the rate coefficient for dissociative recombination of H z , a(H;).Recent measurements suggest that this rate coefficient is approximately lop7 cm3 s-', at room temperature. Earlier work of N. Adams and D. Smith yielded a much smaller value, lo-'' cm3 s-'. If the latter value is adopted in the model, the transition between the two chemical regimes still occurs, but at values of the cloud density nH which are too small to be astronomically interesting. Fig. 1 illustrates these comments.limited width of 0.3 cm-'. In the Q-transitions the opposite parity component of the A-doubled 'n state is excited. Thus we conclude that the predissociation of the L-state of CO is J-dependent and parity dependent.