The assignment of the transitions was Ðrst performed in the
P- and R-branch regions where regular strong absorptions
due to eP(1,[1)/oP([1,[1) and eR(1,1)/oR([1,1) subbranches
are evident. In particular the assignment started from lines
correlation between the constants and to improve the stan-
dard deviation. The results achieved were not successful since
the quality of the Ðt remained invariable and the identical
correlation between d and r constants was obtained as in
J
J
with medium J (D20) and low K values. The identiÐed tran-
the A-reduced Hamiltonian.
a
sitions were Ðtted to obtain, keeping the ground state param-
eters of ref. 3 Ðxed, preliminary values for band origin and
upper state constants that were used to predict further P- and
R-transitions and to extend the assignment to the Q-branch
series. The analysis proceeded iteratively with identiÐcation of
new features and reÐnement of constants.
3.3. Upper state analysis
From the present analysis a total of 3613 lines were identiÐed
in the P, Q and R branches; the assignments comprise the
even transitions and only the odd split ones since no improve-
ment was obtained in the determination of the molecular
parameters by taking into account the unsplit lines. In the
reÐnement procedure blended or scarcely resolved lines were
weighted 1/10 with respect to apparently single lines. Band
origin and upper state constants, reported in column 3 of
Table 1, were obtained with a standard deviation of about
As higher J and K values were assigned, the set of ground
a
state constants became inadequate and the Ðt showed, for
transitions with higher quantum numbers, systematic devi-
ations between observed and calculated wavenumbers. Since
the maximum J and K values reached in the present analysis
a
(70 and 30, respectively) are higher than those employed to
obtain the previous ground state parameters3 a new Ðt was
performed in order to obtain a more complete set of ground
state constants including the sextic distortion terms. Therefore
the 2640 combination di†erences from the present study were
added to those from l 4 and l 3 and used, together with the
8 ] 10~4 cm~1 employing transitions with J@ O 70, K@ O 30,
a
K@ O 69. Some larger deviations in the Ðnal Ðt are produced
c
by weak features subject to overlapping or blended lines due
to the occurring of di†erent transitions at close wavenumbers.
Concerning the sextic distortion constants only the four
parameters U , U , U , r were determined with physical
4
10
available microwave data,5 in a simultaneous Ðt. Since the
microwave measurements have much higher precision than
the combination data, adopting the same choice of ref. 3, the
weights of all the combination di†erences were multiplied by a
factor of 10~5. The present analysis produced a well deter-
mined set of ground state constants (column 2 of Table 1)
obtained with a standard deviation of about 7 ] 10~5 cm~1
from the Ðt of 6254 data. Concerning the sextic distortion
parameters only four terms (U , U , U , r ) could be
J
KJ
K
J
meaning while the remaining terms were Ðxed to zero.
During the analysis some lines with medium J and low K
a
values could not be found in the calculated positions, being
the deviations probably due to weak interactions. Three
nearby states l ] 2l (D1741 cm~1) A , 2l ] l (D1744
3
5
1
7
12
cm~1) B and l ] 2l (D1749 cm~1) A are the possible
2
5
12
1
perturbers of the l level; the reported band centers were cal-
2
culated as harmonic values from vibrational data.7,9 A
J
KJ
K
J
obtained with physical meaning whereas the remaining three
coefficients led to poorly determined values which did not
a†ect the quality of the Ðt and therefore they were constrained
to zero.
detailed discussion on the possible resonances is not reliable
at this stage since the l perturbed features could not be
2
clearly identiÐed due to the high density of the lines in the
spectrum also arising from ““hotÏÏ band contributions.
A comparison between the present constants and those
obtained in ref. 3, included for completeness in Table 1, shows
On the whole, the values of the upper state parameters are
close to their ground state values thus indicating that the tran-
sitions used in the Ðt are almost unperturbed. To allow a
direct comparison with the observed details, simulations of
di†erent spectral regions were performed using the molecular
constants of Table 1. A small portion of the R branch around
a good agreement even if the d value has a larger uncertainty.
J
Owing to the high correlation of d with r the former term
J
J
could be obtained with a smaller uncertainty only by Ðxing r
J
to zero. Since in any case the quality of the Ðt is substantialy
invariable, the more complete set of constants was preferred.
It is worthwhile to note that, although the asymmetry
parameter (i \ [0.841) is rather far from the value of the
symmetric top, the data have also been Ðtted using the S-
reduced Hamiltonian in order to test a somewhat smaller
1750 cm~1 is shown in Fig. 6, where the distinct RR (J) fea-
K
tures with K \ 18 and 20 are also indicated. As can be seen,
a
the calculated spectrum reproduces quite faithfully the
observed one thus indicating the reliability of the obtained
molecular parameters; small discrepancies are due to weak
Table 1 Molecular parameters (cm~1) for the ground and the v \ 1 states of cis-CHF2CHFa
2
Ground state
Ref. 3
Present work
v \ 1
2
l
1718.838 91(5)
0
A
B
C
0.703 933 2(13)
0.703 934 54(75)
0.700 559 45(56)
0.197 808 96(28)
0.154 180 26(29)
0.243 41(20)
0.197 809 31(23)
0.154 180 69(26)
0.245 16(28)
[0.160 67(18)
0.488 31(36)
0.7299(11)
0.4373(48)
0.560(74)
[0.487(33)
0.1375(44)
0.228(38)
0.066
0.197 563 83(30)
0.153 893 36(25)
0.246 83(12)
[0.159 632(55)
0.479 65(16)
0.737 75(98)
0.4496(13)
0.472(32)
[0.5320(99)
0.1435(14)
0.218(23)
0.841
3613
D ] 106
J
JK
K
J
K
J
KJ
K
D
] 105
[0.158 14(23)
D ] 105
0.484(12)
d ] 107
0.722 56(77)
0.4675(86)
d ] 106
U ] 1012
U
] 1010
U ] 109
r ] 1012
J
p ] 103
0.025
799
No. of data
6254
J
59
25
70
31
70
30
max
K
a, max
a Uncertainties given are one standard deviation in units of the last signiÐcant digit.
Phys. Chem. Chem. Phys., 2001, 3, 4242È4246
4245