Polycyclic aromatic hydrocarbon and emission line ratios in active galactic nuclei and starburst galaxies

Abstract

We study the polycyclic aromatic hydrocarbon (PAH) bands, ionic emission lines, and mid-infrared continuum properties in a sample of 171 emission line galaxies taken from the literature plus 15 new active galactic nucleus (AGN) Spitzer spectra. We normalize the spectra at λ = 23μm and grouped them according to the type of nuclear activity. The continuum shape steeply rises for longer wavelengths and can be fitted with a warm blackbody distribution of T ∼ 150–300 K. The brightest PAH spectral bands (6.2, 7.7, 8.6, 11.3, and 12.7 μm) and the forbidden emission lines of [Si II] 34.8 μm, [Ar II] 6.9 μm, [S III] 18.7 and 33.4 μm were detected in all the starbursts and in ∼80% of the Seyfert 2. Taking under consideration only the PAH bands at 7.7μm, 11.3μm, and 12.7μm, we find that they are present in ∼80% of the Seyfert 1, while only half of this type of activity show the 6.2μm and 8.6μm PAH bands. The observed intensity ratios for neutral and ionized PAHs (6.2μm/7.7μm × 11.3μm/7.7μm) were compared to theoretical intensity ratios, showing that AGNs have higher ionization fraction and larger PAH molecules (≥80 carbon atoms) than SB galaxies. The ratio between the ionized (7.7μm) and the neutral PAH bands (8.6μm and 11.3μm) are distributed over different ranges for AGNs and SB galaxies, suggesting that these ratios could depend on the ionization fraction, as well as on the hardness of the radiation field. The ratio between the 7.7μm and 11.3μm bands is nearly constant with the increase of [Ne III]15.5μm/[Ne II] 12.8μm, indicating that the fraction of ionized to neutral PAH bands does not depend on the hardness of the radiation field. The equivalent width of both PAH features show the same dependence (strongly decreasing) with [Ne III]/[Ne II], suggesting that the PAH molecules, emitting either ionized (7.7μm) or neutral (11.3μm) bands, may be destroyed with the increase of the hardness of the radiation field.