S. Malhotra 1, G. Helou 1, M. Kaufman , E. Valjavec , J. Brauher , D.J. Hollenbach , S. Lord 1 N.A. Silbermann , H. Dinerstein , G.J. Stacey , R.H. Rubin , M.W. Werner , D.A. Hunter , K.Y. Lo , C.A. Beichman 1 N. Lu 1 & H. Thronson
IPAC, Caltech, USA
One of the major surprises from ISO-LWS observations has been the
severe deficiency of CII in many normal and ultraluminous galaxies
[1,2]. We show that this is not an isolated phenomenon: there is a
smooth decline in [CII]/FIR with increasing dust temperature and
star-forming activity in galaxies, independent of their luminosity or
morphology. In a sample of 59 normal galaxies, this trend spans a
factor of 100 in [CII]/FIR with [CII] deficient galaxies at its
extreme end. Of the numerous explanations proposed for this variation
the leading ones are (A) optical depth and extinction and (B)
inefficient heating by charged grains when 
is high.
We use other cooling lines to understand the variation in [CII]/FIR and to infer the physical conditions in the ISM. We find that:
(1) We can rule out hypothesis (A) with the argument that [OI] should
be more severely effected by both optical depth and extinction, and we
see an increase in [OI]/[CII] line ratios for galaxies with higher
F60/F100. So, in galaxies with warmer dust, there is less cooling via
the [CII] line, while [OI] at 63 
remains a major coolant. This
trend is qualitatively explained in PDR models by an increase in
radiation field 
,
which raises the dust temperature and the
[OI]/[CII] line ratio. However, PDR models do not explain the actual
[OI]/[CII] ratios observed, only its increase with F60/F100. That is
because of contribution from HII regions, neutral and ionized diffuse
medium to total [CII] emission from a galaxy.
(2) [NII] at 122 microns, which arises from ionized regions, shows behavior similar to [CII], i.e. the ratio [NII]/FIR declines with F60/F100 and L(FIR)/L(B). This suggests that some fraction of [CII] comes from ionized gas.
We use measurements of [NII] (122 ), [OI] (63 )
and other
lines to estimate contributions from the different phases of the
ISM. The observed variations in FIR-normalized [CII], [NII] and [OI]
line fluxes may be due to different proportions of ISM phases in
galaxies.
We expect little variation in [CII]/FIR ratio because it is a measure of the fraction of energy absorbed by dust that goes into heating the gas via photoelectron ejection. It has been shown theoretically that very small grains responsible for aromatic features are more efficient at heating the gas than the large grains[3]. Thus, [CII] emission should at least as well correlated with PAH emission than with classical grain emission at 60 and 100 microns. We find that the ratio of [CII] to PAH emission in the ISOCAM LW2 band (5-8.5 microns) is constant to within a factor of 10 in our sample of normal galaxies and shows no trends with L(FIR)/L(B) and F60/F100.