Origin of diffuse CII 158 micron and SiII 35 micron emission in the Carina nebula

We present the results of mapping observations with ISO of [O I] 63 mum, 145 mum, [N II] 122 mum, [C II] 158 mum, [Si II] 35 mum, and H-2 9.66 mum line emissions for the Carina nebula, an active star-forming region in the Galactic plane. The observations were made for the central 40' x 20' area of the nebula, including the optically bright H II region and molecular cloud lying in front of the ionized gas. Around the center of the observed area is the interface between the H II region and the molecular cloud which creates a typical photodissociation region (PDR). The [C II] 158 mum emission shows a good correlation with the [O I] 63 mum emission and peaks around the H II-molecular region interface. The correlated component has the ratio of [C II] 158 mum to [O I] 63 mum of about 2.8. We estimate from the correlation that about 80% of [C II] emission comes from the PDR in the Carina nebula. The photoelectric heating efficiency estimated from the ratio of the ([C II] 158 mum + [O I] 63 mum) intensity to the total far-infrared intensity ranges from 0.06 to 1.2%. [O I] 145 mum is detected marginally at 10 positions. The average ratio of [O I] 145 mum to [O I] 63 mum of these positions is about 0.09 +/- 0.01 and is larger than model predictions. The observed [C II] 158 mum to [O I] 63 mum ratio indicates a relatively low temperature (<500 K) of the gas, while the large [O I] 145 mu m to 63 mu m ratio suggests a high temperature (similar to 1000 K). This discrepancy cannot be accounted for consistently by the latest PDR model with the efficient photoelectric heating via polycyclic aromatic hydrocarbons (PAHs) even if absorption of [O I] 63 mu m by foreground cold gas is taken into account. We suggest that absorption of [C II] 158 mu m together with [O I] 63 mu m by overlapping PDRs, in which the heating via PAHs is suppressed due to the charge-up effect, may resolve the discrepancy. Quite strong [Si II] 35 mu m emission has been detected over the observed area. It shows a good correlation with [N II] 122 mu m, but the correlation with [O I] 63 mu m is very weak, indicating that [Si II] 35 mu m comes mainly from the diffuse ionized gas rather than the PDR. The ratio of [Si II] 35 mu m to [N II] 122 mu m is about 8 and Si of about 50% of the solar abundance relative to N should be present in the gas phase. The present results suggest that efficient dust destruction takes place and a large fraction of Si returns to the gas in the Carina star-forming region.