Some empirical estimates of the H₂ formation rate in photon-dominated regions

¹ Osservatorio Astrofisico di Arcetri, INAF, Largo E. Fermi 5, 50125 Firenze, Italy
² Institut d'Astrophysique Spatiale, Université Paris-Sud, 91405 Orsay Cedex, France

Corresponding author: E. Habart, habart@arcetri.astro.it

Received: 20 December 2002
Accepted: 25 July 2003

Abstract

We combine recent ISO observations of the vibrational ground state lines of H₂ towards Photon-Dominated Regions (PDRs) with observations of vibrationally excited states made with ground–based telescopes in order to constrain the formation rate of H₂ on grain surfaces under the physical conditions in the layers responsible for H₂ emission. We briefly review the data available for five nearby PDRs. We use steady state PDR models in order to examine the sensitivity of different H₂ line ratios to the H₂ formation rate R_f. We show that the ratio of the 0–0 S(3) to the 1–0 S(1) line increases with R_f but that one requires independent estimates of the radiation field incident upon the PDR and the density in order to infer R_f from the H₂ line data. We confirm earlier work by [CITE] on the Oph W PDR which showed that an H₂ formation rate higher than the standard value of  cm³ s^-1 inferred from UV observations of diffuse clouds is needed to explain the observed H₂ excitation. From comparison of the ISO and ground-based data, we find that moderately excited PDRs such as Oph W, S140 and IC 63 require an H₂ formation rate of about five times the standard value whereas the data for PDRs with a higher incident radiation field such as NGC 2023 and the Orion Bar can be explained with the standard value of R_f. We compare also the H₂ 1–0 S(1) line intensities with the emission in PAH features and find a rough scaling of the ratio of these quantities with the ratio of local density to radiation field. This suggests but does not prove that formation of H₂ on PAHs is important in PDRs. We also consider some empirical models of the H₂ formation process with the aim of explaining these results. Here we consider both formation on classical grains of size roughly 0.1 μm and on very small (~10 Å) grains by either direct recombination from the gas phase (Eley–Rideal mechanism) or recombination of physisorbed H atoms with atoms in a chemisorbed site. We conclude that indirect chemisorption where a physisorbed H-atom scans the grain surface before recombining with a chemisorbed H-atom is most promising in PDRs. Moreover small grains which dominate the total grain surface and spend most of their time at relatively low (below 30 K for χ ≤ 3000) temperatures may be the most promising surface for forming H₂ in PDRs.