The experimental data were taken from
(1) M. Winnewisser, B. P. Winnewisser, and C. Wentrup, 1984, J. Mol. Spectrosc., 105, 193.
Additional data with extensive ¹⁴N hyperfine splitting were obtained in the context of
(2) W. H. Stolze, D. H. Sutter, and C. Wentrup,
1989, Z. Naturforsch., 44A, 291.
Data of two of the four rotational transitions were used because for the other two the frequencies were apparently less certain than the HFS splitting. The HFS structure is sufficiently well contrained. Also used in the fit were data from
(3) B. Bak and H. Svanholt, 1980, Chem. Phys. Lett., 75, 528.
Predictions are probably accurate enough for astronomical observations. They should be viewed with caution once the predicted uncertainties exceed 0.5 MHz.
The ¹⁴N hyperfine splitting may be resolved in astronomical observations. Therefore, a separate hyperfine calculation is provided up to J" = 18, K_a" = 6, and up to 260 GHz along with appropriate partition function values.
The a dipole moment component was determined in
(4) B. Bak, O. J. Nielsen, and H. Svanholt, 1978, Chem. Phys. Lett., 59, 330;
the b component was estimated in (3).