Astron. Astrophys. 322, L1-L4 (1997)

4. Discussion

The experimental uncertainties of our frequency measurements at 1 THz are 50 kHz in the best cases, a number which is about 10 times better than the uncertainties obtained by van den Heuvel et al. (1982 ). This frequency accuracy is sufficiently high for a radioastronomical identification of NH and the determination of velocities in molecular clouds, because the corresponding uncertainty in velocity, 0.015 km s^-1, is much smaller than the linewidths of molecular lines in quiescent clouds (typically 0.3-1.0 km s^-1).

The newly determined molecular constants are compared in Table 2 with those obtained by tunable laser-sideband spectroscopy (van den Heuvel et al. 1982 ), by FT spectroscopy (Brazier et al. 1986 ), and by a molecular beam, laser-induced fluorescence experiment (Ubachs et al. 1984 ). They agree well with each other, but an improvement of one to two orders of magnitude was achieved in the present work. Van den Heuvel et al. (1982) included results from LMR measurements (Wayne & Radford 1976 ) in the fit, since not all three fine structure components had been observed at that time. In our analysis, only the frequencies listed in Table 1 were used to determine the molecular constants presented in Table 2. Brazier et al. 1986 reported slightly different values for and , because they included the higher terms , , and in their analysis.

The value of eQq (N) obtained here is MHz. Ubachs et al. (1984) reported a value of MHz, whereas our value is found to be larger than the corresponding value, MHz, obtained for ND (Saito & Goto 1993, Saito 1996). Further measurements are necessary to obtain a more precise quadrupole coupling constant.

The value of (N) can be calculated by a second order perturbation method using the formula (Townes & Schawlow 1975 ):

where is the matrix element of the nuclear spin-orbit interaction between the electronic state concerned, , and other electronic states, . A possible electronic state which can interact with the ground state of NH is the state, which exists at 776.76 cm^-1 above the ground state (Dixon 1959). We employ a value of 138.8 MHz calculated by Morton & Preston (1978 ) for the value of . The value of is difficult to estimate, but, if we assume it to be of the order of one, the calculated value of (N) is about 0.3 MHz. This agrees roughly with our value of 0.180(27) MHz. In addition, the ratio (N)/ is comparable to the corresponding values from other molecules, e.g. NO (Saleck et al. 1994b ). This indicates that, besides the second order perturbation between the and states, the relatively large value of (N) in NH is mainly caused by the large rotational constant.

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de