Microwave, millimeterwave, and high-resolution FTIR study of the ν2=1 state of SO2F2

Article abstract of DOI:10.1016/j.molstruc.2006.02.034

Microwave and millimeterwave spectra of SO₂F₂ in The v₂=1 state up to 450 GHz have been assigned and 174 pure rotational lines measured. The high-resolution infrared spectrum of the ν₂ band (ν_sym SF₂) centered at 849.5 cm^-1 has been recorded with a resolution of 0.0024 cm^-1. About 1900 rovibrational transitions with ΔK_a=0 and ±2 have been assigned. Rotational and rovibrational data have been combined, and excited state parameters up to sextic centrifugal distortion constants have been obtained using a Watson type Hamiltonian in S-reduction. No perturbation has been observed, although weak B-type Coriolis interactions with ν₈ may be deduced from the effective state parameters.

Full text of DOI:10.1016/j.molstruc.2006.02.034

Journal of Molecular Structure 795 (2006) 185–189
www.elsevier.com/locate/molstruc
Microwave, millimeterwave, and high-resolution FTIR
study of the n2Z1 state of SO2F2
a
b
c
c
d
I. Merke a, , N. Heineking , F. Hegelund , J. Demaison , L. Margules , H. Bu¨rger
*
`
a
¨
Institut fur Physikalische Chemie, RWTH Aachen, D-52056 Aachen, Germany
b Department of Chemistry, University of Aarhus, DK-8000 Aarhus C, Denmark
´ ´
Laboratoire de Physique des Lasers, Atomes et Molecules, UMR CNRS 8523, Universite de Lille 1, F-59655 Villeneuve d’Ascq, France
c
d
¨
Anorganische Chemie, FB C, Universitat–GH Wuppertal, D-42097 Wuppertal, Germany
Received 9 December 2005; received in revised form 10 February 2006; accepted 10 February 2006
Available online 19 April 2006
Dedicated to Professor Gisbert Winnewisser.
Abstract
Microwave and millimeterwave spectra of SO2F2 in The v2Z1 state up to 450 GHz have been assigned and 174 pure rotational lines measured.
The high-resolution infrared spectrum of the n2 band (nsym SF2) centered at 849.5 cmK1 has been recorded with a resolution of 0.0024 cmK1
.
About 1900 rovibrational transitions with DKaZ0 and G2 have been assigned. Rotational and rovibrational data have been combined, and
excited state parameters up to sextic centrifugal distortion constants have been obtained using a Watson type Hamiltonian in S-reduction. No
perturbation has been observed, although weak B-type Coriolis interactions with n8 may be deduced from the effective state parameters.
q 2006 Elsevier B.V. All rights reserved.
Keywords: Sulfuryl fluoride; Microwave spectra; Millimeterwave spectra; Infrared spectrum
1. Introduction
using a new tensorial approach in the O(3)ITdIC2v group
chain was successfully performed [8].
Because of the near-spherical character of the molecule and
with the aim of obtaining equilibrium moments of inertia, we
initiated a study directed towards the determination of the
molecular parameters of all nine fundamental vibrations. Up to
now, detailed rovibrational analyses have been performed for
the ground vibrational state [7,8], for n8 (the apparently
unperturbed asymmetric SF2 stretching vibration) [9], and for
the bending motions n3, n7, and n9 [10,11] forming a triad
which is perturbed by mutual A-, B-, and C-Coriolis
resonances. In all of these recent studies as well as in the
present one the molecular parameters were determined by a
global analysis of transitions from microwave-, millimeter-
wave-, and high-resolution IR-spectra.
Sulfuryl fluoride SO2F2 was studied in the past by low and
medium resolution infrared (IR) and Raman spectroscopy [1–
4] as well as microwave (MW) spectroscopy [2,5,6]. The
microwave studies dealt with the ground state spectrum and the
spectra of fundamental bending modes. The symmetry and
ordering of these bending modes have not been evaluated with
certainty neither from pure rotational nor from vibrational
studies, because they occur in crowded regions of the infrared
spectrum. In these earlier studies no microwave spectrum, or
high-resolution IR spectrum involving these vibrational
stretching states was investigated.
Recently, SO2F2 has attracted our attention because it is
very close to a spherical rotor by accident. Therefore, in spite
of an asymmetry parameter of kZK0.589 an analysis with
Watson’s A-reduction failed [7]. Either S-reduction or 6-
reduction [7] in which all six quartic centrifugal distortion
constants are determined, had to be employed. Also an analysis
In this work, we report on the analysis of the symmetric SF2
stretching vibration n2 which belongs to the symmetry species
A1 of the point group C2v.
2. Experimental details
*
Corresponding author.
Sulfuryl fluoride SO2F2 was prepared by passing SO2Cl2 at
400 8C over dried KF in a dynamic vacuum. Since the
fluorination proceeds stepwise this procedure was repeated
several times until all SO2FCl (monitored by low resolution IR
E-mail address: ilona.merke@rwth-aachen.de (I. Merke).
0022-2860/$ - see front matter q 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2006.02.034

186
I. Merke et al. / Journal of Molecular Structure 795 (2006) 185–189
spectroscopy) had been completely converted to SO2F2.
Purification of SO2F2 was achieved by fractional condensation
using a standard greaseless vacuum line.
Magnetic hyperfine structure was caused by spin–spin and
spin–rotation interactions of the fluorine nuclei. From these
hyperfine structures of the transitions, which closely resemble
those of the ground state, it was concluded that the vibrational
state to which they belong had to have A1 or A2 symmetry.
The K-pattern, which is usually mainly governed by inertia
asymmetry, was different from any of the other vibrational
satellite sets or the ground state, distinguishing this set from the
alternative assignment as a lower-energy vibrational satellite of
the 34S-isotopomer.
Intensity comparison with transitions of the 34S-isotopomer
resulted in an estimated state population of roughly one-third
that of the natural abundance of sulfur-34 (w4%), confirming
the initial assumption of population for a state with relatively
high energy. In fact, only one of the remaining fundamental
states, namely n2, could conceivably have satellite sets of such
intensity. Since the n2 state has an energy of w850 cmK1, this
assignment yields in the molecular beam a vibrational
temperature for this state of about 280G10 K, i.e. no
substantial cooling. For comparison, the satellite sets of the
dyad states n4 and n5 (w380 cmK1) had intensities correspond-
ing to an average vibrational temperature of about 170 K, those
of the triad states n3, n7 and n9 (all three w550 cmK1) of about
220 K. The rotational temperature is dramatically lower,
usually less than 10 K.
In Aachen the measurements in the frequency region
10–40 GHz were carried out using supersonic molecular
beam Fourier transform microwave (MB-FTMW) spec-
trometers which have been described elsewhere [12,13]. In
all cases, a mixture of about 1% SO2F2 in helium was used. The
stagnation pressure was typically 100 kPa. Due to only partly
resolved fluorine spin–spin and spin–rotation coupling, the
accuracy for the center frequency was limited to about 2 kHz.
Transitions in the frequency region 50–110 GHz were
recorded using an absorption-type spectrometer with source
modulation and 2f phase sensitive detection. Schottky diodes
were used as detectors. The pressure in the cell was about 2 Pa
throughout and the centers of unblended lines could be
determined with an accuracy of about 10–30 kHz depending
on their intensity.
In Lille, the millimeterwave (MMW) spectrum was
measured with a source-modulation spectrometer using as
source either harmonics of a Gunn diode between 150 and
250 GHz, or a Thomson-CSF backward-wave oscillator
between 420 and 450 GHz. In both cases a He-cooled InSb
bolometer was used as detector. The accuracy of the
measurements varies between 50 and 200 kHz for unblended
lines depending on the spectral region and the frequency step
chosen to record the spectrum.
3.2. Rovibrational spectrum
The IR spectrum of n2 was obtained in Wuppertal in
connection with our previous study of the n8 band, see Ref. [9]
for experimental details. The resolution (1/maximum optical
path difference) was 0.0024 cmK1, the precision of isolated
lines of medium intensity 0.0001 cmK1. Owing to the density
The n2 band is located ca. 38 cmK1 below the slightly more
intense n8 band, and therefore the R-branch of n2 is overlapped
by the P-branch of n8 giving rise to a high line density in the
region between the band centers. The reader is referred to
Fig. 1 of Ref. [9] for a survey spectrum of the n2 and n8 bands of
SO2F2.
The Q-branch of n2 is located at 849.5 cmK1 degrading
towards smaller wavenumbers. The band shows the typical
appearance of an A-type band with P- (Fig. 1) and R-clusters
(Fig. 2) spaced (BCC), w0.338 cmK1. Nuclear spin statistical
weights are clearly observed in the clusters: lines with Ka odd
are three times stronger than lines with Ka even. The most
intense transitions are governed by the selection rule DKaZ0.
However, transitions with DKaZG2 are also present as
indicated by weaker cluster structures partly overlapping the
DKaZ0 clusters.
The assignment of J-quantum numbers to the P- and
R-branch cluster structures (Figs. 1 and 2) was straightforward.
However, the assignment to Ka of the individual cluster lines
could not be done just by inspection since in A-type bands the
unsplit KaZJ or JK1 lines at the onset of the cluster are
relatively weak. Overlap from DKaZG2 clusters also
complicated the assignment. We finally used ground state
combination differences (GSCDs) based on the ground state
constants [9] and assigned in this way ca. 1100 DKaZ0 and ca.
800 DKaZG2 P- and R-branch lines. The maximum upper
state quantum numbers are JZ68 and KaZ59. For KaO47 only
lines with odd Ka values were assigned. Low-Ka lines are weak
and we were not able to assign with confidence lines where Ka
of the spectrum the effective precision is about 0.0002 cmK1
3. Description of the spectra and assignments
3.1. Rotational spectrum
.
The initial MW experiments, in the frequency range around
30 GHz (w1 cmK1), had to take into account the observation
that in the previously recorded MB-FTMW spectra of SO2F2,
the vibrational satellite transitions of the known states nx, ny, nz,
nw, nu [2] (now assigned as n4, n5, n9, n7, n3) had intensities
which indicated distinctly different vibrational temperatures.
This is due to the fact that in a beam molecules are not in
thermodynamic equilibrium. Therefore formally different
vibrational and rotational temperatures have to be assigned to
different states. In particular, higher energy states appeared to
reveal higher vibrational temperature, i.e. less cooling and
hence enhanced relative intensity in the supersonic expansion.
Searches for the SF stretching state satellites were made on the
low-frequency sides of the transitions of the ground state, and
weak line patterns, characteristic for a near-prolate top, and
belonging to a hitherto unknown species eventually emerged.
These findings were then confirmed by the assignment of
further transitions in lower (w10 and w20 GHz) and higher
frequency ranges (50–435 GHz).

I. Merke et al. / Journal of Molecular Structure 795 (2006) 185–189
187
OP(22)
SP(22) OP(21)
SP(21)
OP(20)
SP(20)
0.80
0.60
0.40
0.20
0.00
10 15 20
QP(22)
10 15 20
QP(21)
10 15
QP(20)
842.00
842.50
Wavenumber/cm–1
Fig. 1. Assignment of a part of the P-wing of the A-type band n2 of SO2F2. Individual Ka-components of the QP(20)–QP(22) clusters are assigned. The location of the
DKaZC2 and K2 clusters SP(J) and OP(J), respectively, are indicated for JZ20–22.
was below 5. No asymmetry split lines were assigned.
Q-branch lines are seriously blended, and we did not include
such transitions in the assignment. In Figs. 1 and 2, a part of our
assignments is shown.
4. Analysis and results
We have been able to assign 174 pure rotational transitions
with JmaxZ43 and Ka,maxZ39, and 1873 rovibrational
0.80
0.60
0.40
0.20
0.00
SR(20)
OR(21)
SR(21) OR(22)
SR(22) OR(23)
10 15 20
QR(22)
10 15 20
QR(21)
10 15 20
QR(20)
10 15 20
QR(23)
856.50
857.00
Wavenumber/cm–1
Fig. 2. Assignment of a part of the R-wing of the A-type band n2 of SO2F2. Individual Ka-components of the QR(20)–QR(23) clusters are assigned. The location of
some of the DKaZC2 and K2 clusters SR(J) and OR(J), respectively, are indicated for JZ20–23.

188
I. Merke et al. / Journal of Molecular Structure 795 (2006) 185–189
transitions with maximum quantum numbers J0Z68 and Ka’Z
59 in the upper state. For the analysis a Watson-type
Hamiltonian in S-reduction was applied using the program
written by H.M. Pickett [14]. The transition frequencies were
weighted proportional to 1/s2, with s as their experimental
uncertainty. The uncertainties of the pure rotational lines range
from 2 to 300 kHz in correspondence to their frequency range
and quality. For all IR-lines s was taken to be 0.0001 cmK1
irrespective of the above-mentioned poorer precision in case of
blended lines. This is however without major relevance for the
derived upper state parameters. In the data fit the ground state
parameters were fixed to the values given in [9]. We were able to
determine the bandcenter, rotational constants, all quartic, and
four of the sextic centrifugal distortion constants, three sextic
constants were fixed to ground state values. The results are
presented in Table 1, in which for comparison also the employed
ground state values [9] and the parameters of the n8Z1 state [9]
are shown. For pure rotational lines a standard deviation of
61 kHz was found, a value which is in good agreement with the
quality of the data set. For IR-transitions the standard deviation
is 0.00040 cmK1, which also seems reasonable as compared to
the actual precision of the present IR data set. We notice that the
IR data of the n8 band had been fitted with an rms deviation of
0.000144 cmK1. No indication of systematic differences
between observed and calculated line positions was found.
Lists of observed and calculated line positions, their weights,
and their assignments, are deposited as Supplementary Material.
As in n8, the asymmetric SF2 stretching mode, the rotational
constants of the n2 state differ less than 0.5% from the ground
state values, and also the quartic centrifugal distortion
constants agree with the exception of d1 within 13% with
those of the ground state. Like in n3 and n7 [10] as well as in n8
[9] d1 shows the largest difference from the ground state values,
in the present case almost a factor of 8. However, it should be
pointed out that in n2 the d1 value is very small.
5. Discussion and conclusion
The obtained parameters show that n2 can be treated as an
isolated state, although the value of d1, which is rather sensitive
to Coriolis effects, and the values of the four determined sextic
parameters, which are typical fit parameters at the present level
of significance, differ considerably from those of the ground
state. Therefore, we cannot exclude that some interactions with
perturber states still to be identified are present.
There is no binary or ternary combination of vibrational
states closer than ca. 80 cmK1 to n2, and therefore strong
interactions, in particular such with level crossings, can hardly
occur. However, ab initio calculations [10] show that there is a
medium-to-strong B-type Coriolis interaction between n2 and
n8, with z(B)ZK0.370. This interaction is supposedly
responsible for the opposite deviations of the d1 values of the
interacting states from that of the ground state. It is particularly
pleasing to note that the average of the two excited state d1
values, K0.10251!10K8 cmK1, is close to that of the ground
vibrational state, K0.10039!10K8 cmK1 (Table 1).
In conclusion we have successfully determined molecular
parameters of the v2Z1 state of SO2F2 by analyzing and jointly
fitting molecular beam Fourier transform microwave, milli-
meterwave, and high-resolution infrared data. The rms
deviations of the different data blocs are in reasonable
agreement with the uncertainties of the measurements. The
merged data have been fitted for an unperturbed isolated
rovibrational state, but there is some evidence for global
Coriolis interactions with the near-by n8 state.
Table 1
Vibrational rotational, and centrifugal distortion parameters (cmK1) and
statistical data for the v2Z1 excited state of SO2F2, S-reduction, Ir
representation
Ground statea
n2
a
n8
E0
887.21882699
0.171039858
0.169062692
0.168295358
5.02701
K5.4492
6.47264
K0.19168
0.40983
1.952
849.4696554(58)
0.170902455(24)
0.168448698(12)
0.168275651(12)
4.86287(25)
K4.8971(15)
6.0517(22)
K0.01334(29)
0.51554(29)
0.99b
A
0.171281099
0.169219650
0.168685233
4.93029
K5.0958
6.1723
B
Acknowledgements
C
DJ!108
DJK!108
DK!108
d1!108
d2!108
HJ!1014
HJK!1014
HKJ!1014
HK!1014
h1!1014
h2!1014
h3!1014
This work has been supported by the French/German PICS
program (599). We also thank the Deutsche Forschungsge-
meinschaft, the Land Nordrhein-Westfalen, and the Fonds der
Chemie for financial support. F.H. thanks Mrs G. Nathansen for
assistance in the assignment procedure for IR spectra.
K0.10039
0.45750
0.99
3.16
6.91
K7.68(51)
27.0(13)
1.20
K9.00
2.53
K2.88
1.77
K18.84(93)
1.77b
Supplementary data
3.365
0.368
1.642
K2.187b
0.368b
Supplementary data associated with this article can be found,
in the online version, at doi:10.1016/j.molstruc.2006.02.034.
K2.187
2.47(17)
Ka,max/Jmax
59/68c
NIR/NMW
sIR!103
sMW/MMW
1873/174c
0.396
61d
References
The ground state and n8 parameters are included for comparison.
a
From Ref. [9].
[1] W.D. Perkins, M.K. Wilson, J. Chem. Phys. 20 (1952) 1791.
[2] D.R. Lide, D.E. Mann, J.J. Comeford, Spectrochim. Acta 21 (1965) 497.
[3] R.A. Suthers, T. Henshall, Z. Anorg. Allg. Chem. 388 (1972) 257.
[4] A. Kornath, J. Mol. Spectrosc. 188 (1998) 63.
b
Constrained to ground state value.
c
Dimensionless.
d
In kilohertz.

I. Merke et al. / Journal of Molecular Structure 795 (2006) 185–189
189
`
[10] H. Bu¨rger, J. Demaison, F. Hegelund, L. Margules, I. Merke, J. Mol.
[5] R.M. Fristrom, J. Chem. Phys. 20 (1952) 1.
[6] D.R. Lide, D.E. Mann, R.M. Fristrom, J. Chem. Phys. 26 (1957) 734.
`
[7] K. Sarka, J. Demaison, L. Margules, I. Merke, N. Heineking, H. Bu¨rger,
Struct. 612 (2002) 133.
`
¨
¨
[11] M. Rotger, V. Boudon, M. Loete, N. Zvereva- Loete, L. Margules,
J. Demaison, I. Merke, F. Hegelund, H. Bu¨rger, J. Mol. Spectrosc.
submitted.
H. Ruland, J. Mol. Spectrosc. 200 (2000) 55.
`
[8] M. Rotger, V. Boudon, M. Loete, L. Margules, J. Demaison,
¨
¨
¨
H. Mader, G. Winnewisser, H.S.P. Muller, J. Mol. Spectrosc. 222
[12] U. Andresen, H. Dreizler, J.-U. Grabow, W. Stahl, Rev. Sci. Instrum. 61
(1990) 3694.
(2003) 172.
`
[9] H. Bu¨rger, H. Ruland, I. Merke, K. Sarka, L. Margules, J. Demaison,
[13] I. Merke, W. Stahl, H. Dreizler, Z. Naturforsch. A 49 (1994) 490.
[14] H.M. Pickett, J. Mol. Spectrosc. 148 (1991) 371.
J. Mol. Spectrosc. 203 (2000) 268.