We report on the detection of a narrow feature at 262816.73 MHz towards Orion and the cold prestellar core B1-bS, that we attribute to the 1 0 -0 0 line of the deuterated Ammonium ion, NH 3 D + . The observations were performed with the IRAM 30m radio telescope. The carrier has to be a light molecular species as it is the only feature detected over 3.6 GHz of bandwidth. The hyperfine structure is not resolved indicating a very low value for the electric quadrupolar coupling constant of Nitrogen which is expected for NH 3 D + as the electric field over the N nucleus is practically zero. Moreover, the feature is right at the predicted frequency for the 1 0 -0 0 transition of the Ammonium ion, 262817±6 MHz (3σ), using rotational constants derived from new infrared data obtained in our laboratory in Madrid. The estimated column density is (1.1±0.2)×10 12 cm −2 . Assuming a deuterium enhancement similar to that of NH 2 D, we derive N(NH + 4 )≃ 2.6×10 13 cm −2 , i.e., an abundance for Ammonium of a few 10 −11 .

Line-mixing effects have been studied in the 3 band of CH 4 perturbed by N 2 at room temperature. New measurements have been made and a model is proposed which is not, for the first time, strictly empirical. Three different experimental set ups have been used in order to measure absorption in the 2800-3200 cm Ϫ1 spectral region for total pressures in the 0.25-2 and 25-80 atm ranges. Analysis of the spectra demonstrates the significant influence of line mixing on the shape of the Q branch and of the P and R manifolds. A model is proposed which is based on state-to-state collisional transfer rates calculated from the intermolecular potential surface with a semiclassical approach. The line-coupling relaxation matrix is constructed from these data and two additional parameters which are fitted on measured absorption. Comparisons between measurements and spectra computed accounting for and neglecting line mixing are made. They prove the quality of the approach which satisfactory accounts for the effects of pressure and of rotational quantum numbers on the spectral shape under conditions where modifications introduced by line mixing are important. For high rotational quantum number lines, the main features induced by collisions are predicted but some discrepancies remain; the latter may be due to improper line-coupling elements but there is strong evidence that the use of inaccurate line broadening parameters also contributes to errors in calculated spectra.

No abstract