Ro-vibrational excitation of an organic molecule (HCN) in protoplanetary disks

(Abridged) Organic molecules are important constituents of protoplanetary disks. Their ro-vibrational lines observed in the near- and mid-infrared are commonly detected toward T Tauri disks. These lines are the only way to probe the chemistry in the inner few au where terrestrial planets form. To understand this chemistry, accurate molecular abundances have to be determined. This is complicated by excitation effects. Most analyses so far have made the assumption of local thermal equilibrium (LTE). Starting from estimates for the collisional rate coefficients of HCN, non-LTE slab models of the HCN emission were calculated to study the importance of different excitation mechanisms. Using a new radiative transfer model, the HCN emission from a full two-dimensional disk was then modeled to study the effect of the non-LTE excitation, together with the line formation. We ran models tailored to the T Tauri disk AS 205 (N) where HCN lines in both the 3 {mu}m and 14 {mu}m bands have been observed by VLT-CRIRES and the Spitzer Space Telescope. Reproducing the observed 3 {mu}m / 14 {mu}m flux ratios requires very high densities and kinetic temperatures ($n > 10^{14}$ cm$^{-3}$ and $T > 750$ K), if only collisional excitation is accounted for. Radiative pumping can, however, excite the lines easily out to considerable radii $sim$ 10 au. Consequently, abundances derived from LTE and non-LTE models do not differ by more than a factor of about 3. Models with both a strongly enhanced abundance within $sim$ 1 au (jump abundance) and constant abundance can reproduce the current observations, but future observations with the MIRI instrument on JWST and METIS on the E-ELT can easily distinguish between the scenarios and test chemical models. Depending on the scenario, ALMA can detect rotational lines within vibrationally excited levels.