On the H₂ line emission from NGC 6240 and other starburst galaxies

We examine several possibilities for the production of the extremely powerful H₂ line emission from the starburst galaxy NGC 6240. We review observations of NGC 6240 and argue that neither UV pumping by starlight nor collisional excitation in shock waves appears to be capable of producing the observed emission. We propose instead that most of the H₂ line emission originates in molecular gas which is heated by transient X-ray irradiation; only ∼20%-30% of the H₂ υ = 1 → OS(1) emission is attributed to UV pumping in photodissociation fronts. We deduce the physical conditions under which reprocessing of X-rays can account for the observed H₂ line emission. Under suitable conditions, the observed H₂ line emission could be produced if-∼ 1 × 10¹⁰ L_⊙ of soft X-rays is being absorbed by molecular material in NGC 6240. The required X-rays could be produced by either supernova explosions (at a rate ∼1.5 yr-1) or high-velocity shock waves associated with the galaxy merger. The required supernova rate does not appear to be incompatible with the likely stellar content of NGC 6240, if it underwent a major burst of star formation ∼ 1-2 × 10⁷ yr ago. Our scenario appears able to account naturally for the reported weakness of the υ= 2 → 1S(3) line and may provide a favorable environment for the pumping of H₂O masers. Our model predicts strong emission in unobserved H₂ lines such as υ = 0 → OS(9) λ4.6946 μm, and in fine-structure lines such as [O I] 63 μm. Rotational and vibrational transitions of H₃⁺ provide a unique spectral signature of X-ray-heated molecular gas; we predict that some of these transitions, including lines at 16.33 μm and 4.35 μm, may be detectable.