Depletion of CO in a cold dense cloud core of IC 5146

We have used the IRAM 30m telescope to map the C(18O) (1->0) and (2->1) transitions in a region of 0.5\times0.5 pc^2 (200''\times200'') in the nearby (460 pc distance) molecular cloud IC 5146 at resolutions of down to 0.025 pc. Additionally, we have observed the corresponding {C¹⁷O} transitions towards 24 positions in the central region of the cloud as a check on optical depth effects. On the basis of these observations, we conclude that the C(18O) emission observed by us is optically thin and that the observed {C¹⁸O} (2->1)/(1->0) ratio is consistent with gas at 10 K and density in the range 10(4) to 10(5) cm(-3) . We have estimated C(18O) column densities over the region mapped, using an LVG program to correct for excitation effects, and have compared them, at the same spatial resolution of 30'', with the dust extinction derived from the NIR images of Lada et al. (cite {clada1998}). From this, we find a roughly constant, though with considerable dispersion, ratio of {C¹⁸O} column density to visual extinction (A_V) in directions with A_V less than ~ 10 magnitudes. The ratio is compatible with previous determinations of the [C(18O) ]/[{H_2}] abundance ratio, e.g. 1.7 10^{-7} as obtained by Frerking et al. (cite {frerking1982}). At extinctions higher than \sim10 magnitudes, however, there is a systematic fall-off in the ratio N( {C¹⁸O})/A_V which we attribute to depletion of C(18O) onto dust ice mantles within the dense, n>~10^4 {cm^{-3}}, and cold, T_dust<~15 K, T_gas\sim10 K, core interior. This fall-off is also seen, though with less statistical significance, in our observed dependence of N({C¹⁷O})/A_V with A_V. We thus think it unlikely that optical depth effects are influencing our conclusions. We have additionally made pointed C$^18O (1->0) and (2->1) observations towards 94 background stars with extinction greater than 10 magnitudes from the study of Lada et al. (\cite{clada1994}). These in general confirm our mapping results.