The laboratory data has been taken from
(1) P. Caselli and L. Dore, 2005, Astron. Astrophys. 433, 1145.
With respect to the Oct. 2005 entry, additional high-J data were used. They were taken from
(2) V. Lattanzi, A. Walters, B. J. Drouin, and J. C. Pearson, 2007, Astrophys. J. 662, 771.
In contrast to earlier work, (1) and (2) seem to have avoided frequency shifts due to ion drift.
With respect to the Oct. 2007 entry, newly published astronomical observations for the J = 1 – 0 transition were used in the fit:
(3) F. F. S. van der Tak, H. S. P. Müller, M. E. Harding, and J. Gauss, 2009, Astron. Astrophys. 507, 347.
Predictions beyond 2 THz should be viewed with caution.
At low J, it is possible to resolve the D hyperfine splitting. Therefore, a separate hyperfine calculation up to J" = 2 is provided. Note: The partion function given below does NOT take into account this splitting ! The partition function has to be multiplied with I(D) = 3 when hyperfine splitting is considered.
The dipole moment is assumed to agree with that of the main isotopic species, see e029507.cat.