The observed transitions were taken from
(1) F. C. de Lucia and P. Helminger,
1976, J. Mol. Spectrosc. 54, 200;
(2) E. A. Cohen and H. M. Pickett,
1982, J. Mol. Spectrosc. 93, 83;
and from
(3) L. Fusina, D. di Lonardo, J. W. C. Johns, and L. Halonen,
1988, J. Mol. Spectrosc. 127, 240.
Using three additional higher order centrifugal distortion
parameters compared to the ones from (3),
it was possible to fit the transition frequencies
within experimental uncertainties on the average.
NH_{2}D tunnels between two equivalent positions
as does the main isotopomer NH_{3}.
The strong ctype transitions occur between the
tunneling substates whereas atype transitions
occur within the states.
The antisymmetric, J = 0 state, in the catalog
with the state number 1, is higher than
the symmetric, J = 0 state by
0.4059 cm^{–1} or 12169.4 MHz.
The rotational constants are average values.
In addition, one has to distinguish between ortho and
para levels with a spinstatistical weight ratio of
3 : 1.
In the symmetric substate, the ortho and para
levels are described by K_{a} odd and even,
respectively, while it is reversed for the antisymmetric
substate. The 1_{11} level is the lowest ortho
and para level in the symmetric and antisymmetric
substate, respectively. It is 14.3725 and
14.7761 cm^{–1} above the symmetric
J = 0 level, respectively.
Since ^{14}N hyperfine splitting may be resolved
for low values of J, a
separate calculation is provided for J up to 10.
The partition function takes into account the spin multiplicity
g_{I} = 3 of the ^{14}N nucleus !
July 2017: separate
para and
ortho predictions with hyperfine splitting are available
up to J = 4 and 2600 GHz along with separate
para and
ortho partition function values.
The dipole moment was taken from (2). The negative value of
the interaction constant F_{ac} affords a
negative value for μ_{a}.
