Synthesis and vibrational spectroscopy of 1,1,2,2-tetrafluoroethane and its13C2 and d2 isotopomers

Article abstract of DOI:10.1021/jp0004967

The ¹³C₂ and d₂ isotopomers of 1,1,2,2-tetrafluoroethane (TFEA) have been synthesized. Raman spectra of these new species have been recorded, and infrared spectra of all three isotopomers, including some regions with high-resolution at -100°C, have also been recorded. Guided by recently published calculations of frequencies and infrared intensities and the new spectra, we have revised the previous assignments of fundamentals for the two rotamers of the normal species of TFEA. Assignments of the fundamentals for both rotamers of the ¹³C₂ and d₂ isotopomers are proposed. The anti rotamer is the more abundant species in the gas phase and, to a lesser extent, in the liquid phase and the only species in the crystal phase. Thus, the assignments of the anti rotamer of all three isotopic species are complete and supported by isotope product rules, but the assignments for the gauche rotamers are incomplete. Estimates of the missing frequencies for the gauche rotamer of the normal species are supplied.

Full text of DOI:10.1021/jp0004967

10092
J. Phys. Chem. A 2000, 104, 10092-10103
13
Synthesis and Vibrational Spectroscopy of 1,1,2,2-Tetrafluoroethane and Its C2 and d2
Isotopomers†
Norman C. Craig,* Jessica I. Chuang, Christiana C. Nwofor, and Catherine M. Oertel
Department of Chemistry, Oberlin College, Oberlin, Ohio 44074
ReceiVed: February 7, 2000; In Final Form: March 16, 2000
The 13C2 and d2 isotopomers of 1,1,2,2-tetrafluoroethane (TFEA) have been synthesized. Raman spectra of
these new species have been recorded, and infrared spectra of all three isotopomers, including some regions
with high-resolution at -100 °C, have also been recorded. Guided by recently published calculations of
frequencies and infrared intensities and the new spectra, we have revised the previous assignments of
fundamentals for the two rotamers of the normal species of TFEA. Assignments of the fundamentals for both
rotamers of the 13C2 and d2 isotopomers are proposed. The anti rotamer is the more abundant species in the
gas phase and, to a lesser extent, in the liquid phase and the only species in the crystal phase. Thus, the
assignments of the anti rotamer of all three isotopic species are complete and supported by isotope product
rules, but the assignments for the gauche rotamers are incomplete. Estimates of the missing frequencies for
the gauche rotamer of the normal species are supplied.
Introduction
potential for evaluating qualitative theories of the gauche effect
and for evaluating the quality of ab initio calculations.
The long-range goal of the present investigation is to find
the complete structures of both rotamers of TFEA. A fine start
was made at NIST and PNL by Stone, Philips, Fraser, Lovas,
Xu, and Sharpe, who obtained ground-state rotational constants
for both rotamers.10 The rotational constants for the anti rotamer
came from the analysis of the rotational structure in a band in
the high-resolution infrared spectrum, observed in a jet-cooled
beam with a diode laser. The rotational constants of the gauche
rotamer came from a microwave study. To complete the
structural analysis, high-resolution spectra of isotopic modifica-
tions of TFEA are needed. The present study is on the synthesis
of the needed isotopomers and the analysis of their medium-
resolution vibrational spectra as a prelude to the high-resolution
investigations.
Two studies have been made of the vibrational fundamentals
of the two rotamers of the normal form of TFEA. The first was
by Klaboe and Nielsen.11 A more extensive study was done by
Kalasinsky, Anjaria, and Little with improved instrumentation.6
The second investigation led to a revised assignment of the
fundamentals of the two rotamers. In the meantime, two ab initio
studies of the vibrational frequencies have been published.7,12
The work by Papasavva, Illinger, and Kenny,7 which used
higher-level theory, provides a good basis for reviewing the
assignment of the vibrational fundamentals. As part of their
study, they computed infrared intensities as well as frequencies.
In general, these calculations are consistent with the Kalasinsky
assignment but, coupled with information from the high-
resolution infrared spectrum and from the investigation of two
isotopomers, we propose revisions in the assignments of
vibrational fundamentals of both rotamers.
The two rotamers of 1,1,2,2-tetrafluoroethane (TFEA) are of
interest as part of the study of the puzzling gauche effect. The
rotamers of the less-substituted 1,2-difluoroethane exhibit the
gauche effect. Thus, the gauche rotamer of 1,2-difluoroethane
has a lower energy than the anti rotamer despite the closer
approach of the two highly electronegative fluorine atoms in
the gauche rotamer.1,2 The lower energy of the gauche rotamer
of 1,2-difluoroethane is affirmed in high-level ab initio calcula-
tions by Muir and Baker.3 Wiberg has provided a qualitative
rationalization for this result in terms of a destabilization of
the anti rotamer by adverse bond bending in the CC σ bond.4
Engkvist, Kalstrom, and Widmark have used what amounts to
a four-center interaction involving the FCCF framework to
rationalize the gauche effect through a stabilization of the gauche
rotamer.5 Applying either of these qualitative arguments predicts
that the rotamers of TFEA should be an even more striking
example of the gauche effect with the gauche rotamer having
the lower energy. Yet, experimental evidence shows that the
anti rotamer has a lower energy by about 4.9 kJ/mol in the gas
phase.6 Ab initio calculations by the ACM-density functional
method and by the MP2 method concur with the sign and
magnitude of this experimental result.3,7 They give 5.4 and 6.9
kJ/mol, respectively, for the energy of the gauche rotamer
relative to the anti rotamer. In addition, Epiotis has used a
valence bond argument to rationalize the lower energies of the
gauche rotamer in 1,2-difluoroethane and the anti rotamer in
TFEA.8
We have previously applied the analysis of the rotational
structure in high-resolution infrared spectra to finding the
complete structure of the nonpolar anti rotamer of 1,2-
difluoroethane.2 This structure shows significant changes in bond
lengths and bond angles in comparison with the structure of
the gauche rotamer, which was found by microwave spectros-
copy.9 Such geometric differences between rotamers have the
A method had to be devised to synthesize the isotopomers
of TFEA. We thought that the reaction of silver difluoride with
1,1,2,2-tetrabromoethane (TBrEA) might succeed, as had a
similar reaction for making 1,2-difluoroethane from 1,2-dibro-
moethane.2 Indeed, AgF2 works but at a temperature well above
the room-temperature reaction conditions that were used in the
† Part of the special issue “C. Bradley Moore Festschrift”.
10.1021/jp0004967 CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/11/2000

Synthesis and Spectra of 1,1,2,2-Tetrafluoroethane
J. Phys. Chem. A, Vol. 104, No. 45, 2000 10093
reaction with 1,2-dibromoethane. At an intermediate temperature
of 80 °C, AgF2 converts TBrEA into 1,1-dibromo-2,2-difluo-
roethane. A temperature of 150 °C is needed to complete the
formation of TFEA. TFEA was first made in the 1930s by
Henne and Midgley by reaction of HgF2 with TBrEA in a metal
reactor at elevated temperatures and pressures.13 We regard AgF2
as a more easily used and safer reagent for small-scale synthesis.
AgF2 is an underutilized reagent for fluorinations. Appropriate
commercially available acetylenes serve as starting materials
for making the d2 and 13C2 isotopic species. Bromine addition
to acetylene-d2 and to acetylene-13C2 provides the needed
TBrEA isotopomers.
adapter, 0.23 mL (2.0 mmol) of TBrEA-d2 was injected with a
syringe, and the stopcock adapter was replaced. Nitrogen was
pumped away on the vacuum system. The mixture was shaken
and put into an oven at 150 °C but with the stopcock kept
outside the heated region. Every 15 min over a 3-hr period the
flask was removed from the oven and shaken. After the heating
period, the bottom of the flask was cooled in liquid nitrogen,
and noncondensable gases were pumped away. To remove
byproduct CO2, Br2, and SiF4, the crude product was passed
through a column packed with Ascarite II (Thomas Scientific)
which had been dampened by drawing room air through the
column before use. The sample was dried by passing it through
a column containing P2O5 dispersed on glass wool. TFEA-d2
was isolated by preparative gas chromatography in two stages.
The first stage was on a 1-m column packed with 10% (w/w)
tricresyl phosphate on Fluoropak 80 (powdered Teflon from
Analabs) to get a fraction that was largely TFEA-d2 mixed with
some other volatiles. Partly fluorinated TBrEA fractions were
eluted later. The second stage was on a 5-m column with the
same packing at 0 °C. Finally, the TFEA-d2 was dried again by
passing it through the P2O5 column. The identity of TFEA-d2
was confirmed by 19F NMR (complex multiplet at δ ) -137.27
ppm) and deuterium NMR (multiplet at δ ) 5.61 ppm).
TFEA is also of interest as a possible replacement for Freons
in refrigeration systems. Having good spectroscopic information
about this substance is important for monitoring its presence
and fate in the atmosphere.14
Experimental Section
Synthesis and Characterization. Normal TFEA (bp -20
°C) was purchased (PCR) and used in spectroscopic studies
without purification.
TBrEA-d2 was made photochemically with an essentially 1:2
molar ratio of acetylene and bromine in a 1.1-L flask equipped
with a standard-taper joint and a vacuum stopcock on the top
end and a 19/38 standard-taper joint with a mated tube on the
bottom end. The flask was greased with Krytox grease (Dupont)
and contained some short lengths of Teflon tubing to aid in
mixing gases. 175 Torr of acetylene-d2 (Cambridge Isotope
Labs, 99% D) was measured into the evacuated 1.1-L flask and
then condensed in a holding flask. 173 Torr (fixed by the v.p.
at 20 °C) of bromine (Fisher, ACS reagent) was measured into
the 1.1-L flask and stored in a third flask. A second 173-Torr
of bromine was measured into the 1.1-L flask. The two portions
of bromine were condensed at liquid-nitrogen temperature in
the very bottom of the tube at the bottom of the evacuated flask,
and the acetylene-d2 sample was condensed higher up in the
tube out of contact with the bromine. The flask, with its contents
still condensed in the bottom, was taken into near darkness,
slowly warmed and shaken well. Initially a small flashlight was
used for illumination as shaking was continued. Then, over a
period of 35 min, the flask was slowly brought toward full
fluorescent lighting. To complete the reaction, the flask was
left for several hours near a daylit window. The rather involatile
TBrEA-d2 product, which collected as droplets on the walls of
the flask, was chased into the detachable tube at the bottom of
the flask by cooling the tube in liquid nitrogen and using a heat
gun. After warming the product to room temperature, air was
admitted to the reaction flask. The identity of TBrEA-d2 was
confirmed by 13C NMR (δ ) 46.56 ppm, equal intensity triplet
with JCD ) 27.6 Hz) and deuterium NMR (singlet, δ ) 5.92
ppm).
TFEA-d2 was prepared by reaction of silver difluoride with
TBrEA-d2. The reaction was run in a 100-mL flask equipped
with some short pieces of Teflon tubing and a 19/38 inner
ground joint, which was attached through this joint to a stopcock
and another ground joint. Silicone grease, which withstands
elevated temperatures, was used. In a reaction, 3.50 g (24 mmol)
of AgF2 (Aldrich) was transferred into the reaction flask in a
nitrogen-purged dry bag. On the vacuum system the nitrogen
was pumped away, and the AgF2 was heat-gunned repeatedly
until a new heating gave a pressure rise of only a few microns.
This rigorous “drying” of AgF2 is key to controlled reactions
with this reagent. With nitrogen purge techniques a serum bottle
cap was put on the reaction flask in place of the stopcock
TFEA-13C2 was prepared in a similar manner from acetylene-
13C2 (CDN Isotopes, Canada. 99% 13C) as the starting material.
For the intermediate TBrEA-13C2 the proton (δ ) 5.98 ppm)
and the proton-coupled 13C (δ ) 46.88 ppm) NMR spectra were
identical AA′XX′ multiplets.15 The absolute values of the
coupling constants were JCH ) 180.4 Hz, JCH′ ) 1.2 Hz, JCC′
) 40.7 Hz, and JHH′ ) 2.8 Hz. For the final product, TFEA-
13C2, the proton spectrum (δ ) 5.72 ppm) is a complex multiplet
with two prominent peaks split by 193.0 Hz. The 19F spectrum
(δ ) -136.58 ppm) is a complex multiplet with four prominent
peaks split by 211.2 and 56.2 Hz. The proton-decoupled 13C
spectrum (δ ) 109.21 ppm) is a complex multiplet with a
prominent triplet structure split by 211.2 Hz.
For normal TFEA, the natural abundance 13C spectrum with
proton decoupling has δ ) 109.22 ppm, JCF(gem) ) 246.4 Hz
and JCF(vic) ) 35.0 Hz. For the complex proton and 19F spectra
the shifts are 5.72 ppm and -136.46 ppm. These latter two
spectra have been fully analyzed before.16
Proton NMR spectra of the partly fluorinated TBrEA products
are useful in identifying these new intermediates and following
the course of the reaction. Proton-decoupled 13C and proton-
coupled 19F spectra were also recorded but are not reported here.
Also unreported are spectra of deuterium- and 13C-substituted
intermediates.
For 1,1,2-tribromo-2-fluoroethane, the multiplet centered at
5.79 ppm is for the CBr2H end and is a doublet of doublets
with JHF(vic) ) 13.0 Hz and JHH ) 3.8 Hz. The multiplet centered
at 6.56 ppm is for the CBrFH end and is a doublet of doublets
with JHF(gem) ) 48.9 Hz and JHH ) 3.8 Hz.
For 1,1-dibromo-2,2-difluoroethane, the multiplet centered at
5.57 ppm is for the CBr2H end and is a triplet of doublets with
JHF(vic) ) 9.8 Hz and JHH ) 3.5 Hz. The multiplet centered at
5.86 ppm is for the CF2H end and is a triplet of doublets with
JHF(gem) ) 55.6 Hz and JHH ) 3.5 Hz.
For 1-bromo-1,2,2-trifluoroethane, the multiplet centered at
5.89 ppm is for the CF2H end and is a triplet of triplets (partly
resolved splitting in the center triplet) with JHF(gem) ) 54.5 Hz,
JHF(vic) ≈ 4.0 Hz and JHH ) 4.0 Hz. The multiplet centered at
6.35 ppm is for the CBrFH end and is a doublet of triplets of
doublets with JHF(gem) ) 48.4 Hz, JHF(vic) ) 6.8 Hz and JHH
4.0 Hz.
)

10094 J. Phys. Chem. A, Vol. 104, No. 45, 2000
Craig et al.
TABLE 1: Moments of Inertia for Three Isotopomers of the
Anti Rotamer and for the Gauche Rotamer of TFEA (in
amu Å2)
Gas-phase infrared spectra can also be used to follow the
progress of fluorination. The tribromomonofluoro intermediate
is, however, too involatile to sample in this way. Distinguishing
features in cm-1 are as follows. The spectrum of the dibromo-
difluoro species has Q branches at 2990 and 2972, medium
intensity features at 1375, 1164, 744, and a very strong triplet
at 1098. The spectrum of the trifluoromonobromo species has
a Q branch at 3006, a very strong doublet at 1128, 1121, and a
medium triplet at 763.
species
Ia
Ib
Ic
anti rotamer
normal speciesa
13C2 speciesb
98.338
98.538
102.333
160.378
161.514
165.092
245.068
245.999
245.782
d2 speciesb
gauche rotamer
normal speciesa
96.191
178.076
198.832
Spectroscopy. Midrange- and far-infrared spectra were
recorded on Perkin-Elmer FT 1760 and 1700X spectrometers,
respectively. Gas-phase samples were contained in stopcock-
equipped, 10-cm Wilmad mini cells with 25-mm diameter
cesium iodide or polyethylene windows. For low-temperature,
crystal-phase samples, a Hornig-type cold cell with cesium
iodide windows was used. A thin sample was sprayed onto the
liquid nitrogen cooled interior window and then annealed by
raising the temperature incrementally and checking for increased
crystallinity in the spectrum recorded at 77 K. Typically, 25
scans at 0.5 cm-1 resolution were accumulated for midrange
spectra, and 200 scans at 1.0 cm-1 resolution were recorded
for the far-infrared spectra. For all spectra that included regions
with water bands the instruments were purged with dry nitrogen.
Raman spectra were recorded with a Spex Ramalog 5
instrument and its associated Nicolet 1180 minicomputer. The
514.5-nm exciting line was produced by a Coherent Innova 70
argon-ion laser. Plasma lines were removed with a home-built
Amici prism system or with a Kaiser Optical corner cube.
Samples were contained in 1.8-mm capillaries and mounted in
a Harney-Miller-type cold cell cooled by a regulated stream of
cold nitrogen gas. The cell temperature was adjusted to -80
°C for liquid-phase samples and -105 °C for crystal-phase
samples. The instrument has 90° sample geometry with a
polarization filter in the scattered beam. For liquid-phase
samples, both perpendicular and parallel positions of the filter
were used, and four scans were accumulated for each setting.
For crystal-phase samples, the polarization filter was removed,
and a Kaiser Optical notch filter was used to remove the exciting
line from the scattered light. All spectra were recorded with
about 2.5 cm-1 resolution, and frequency accuracy is (2 cm-1
based on calibration with an argon pencil lamp. Reported spectra
were treated with a nine-point smoothing function.
a From ref 10. b Computed from geometric parameters in ref 10.
10 a modes and 8 b modes, all of which are both infrared- and
Raman-active. The a modes give polarized bands, and the b
modes give depolarized bands in the Raman spectrum. The
symmetry axis is also the c principal rotation axis. Thus, modes
of the a symmetry species have C-type shapes, and the modes
of the b symmetry species have hybrid A/B-type shapes in the
gas-phase infrared spectrum.
Table 1 gives the moments of inertia of the two rotamers of
TFEA and of the two isotopic modifications of the anti rotamer.
The moments of inertia of the normal species are observed
values.10 The moments of inertia of the 13C2 and d2 isotopomers
were computed from the provisional geometric parameters
reported by Stone et al.10 The moments of inertia of the
isotopomers of the anti rotamer are used in the product rule
calculations reported below. Both rotamers are rather asym-
metric tops with κ ) -0.295 and -0.756 for the normal anti
and gauche rotamers, respectively.
Kalasinsky and co-workers observed two different solid forms
in the Raman spectrum of the normal species, forms I and II.6
Although we did not repeat a study of the Raman spectrum of
the normal form in the condensed phase, we saw no evidence
of differences in the Raman spectra of the 13C2 or d2 species
that could be attributed to two different solid forms. Of course,
some differences occur that can be attributed to annealing.
A puzzling feature of the spectra of tetrafluoroethane is the
decided increase in the CH stretching frequency in going from
the gas phase to the crystal phase. This behavior is present but
unremarked upon in the spectra of the normal species as reported
before.6 The up-frequency shifts in the infrared spectrum are
49, 45, and 5 cm-1 for the normal, 13C2, and d2 species,
respectively. For the normal species in the Raman spectrum,
the corresponding up-frequency shift is 31 cm-1.6 Such large,
up-frequency shifts in going from the gas phase and the liquid
phase to the crystal phase are unexpected. A possible, but
questionable, explanation is hydrogen bonding. In going from
the gas phase to the crystal phase, one expects stretching modes
to decrease in frequency and bending modes to increase in
frequency with hydrogen bonding.17 In addition, although the
shift due to hydrogen bonding should be smaller for CD bonds
than for CH bonds, the difference observed for TFEA isoto-
pomers is too large. The bending modes do increase in frequency
some. These up-frequency shifts are 8, 7 cm-1 and 8, 8 cm-1
for the infrared-active modes of the normal and 13C2 species,
respectively. Small shifts downward occur, however, for the d2
species. Another possibility for the unusual frequency shifts is
a marked crystal splitting with the other component being
inactive in each spectrum. Thus, only the upward frequency
shift is apparent. Neither explanation is fully satisfactory.
In the discussion of assignments that follows, we consider
first the assignments of the anti rotamers of all three isoto-
pomers. These assignments are more secure and prepare the
way for the gauche assignments. Because the lower-energy, anti
form is the only rotamer present in the crystal phase, examina-
NMR spectra were recorded in CDCl3 solutions on a Bruker
AC 200 instrument and referenced to external TMS or CFCl3.
Results and Discussion
Selection Rules and General Remarks. The two rotamers
of TFEA have different symmetries and therefore different
selection rules. For the normal species as well as for both
isotopomers, the anti rotamer has C2h symmetry. The vibrational
fundamentals comprise six ag modes, four au modes, three bg
modes, and five bu modes. Mutual exclusion applies to this
centrosymmetric molecule, and only the au and bu modes are
infrared-active. The a and c principal rotation axes lie in the
plane of symmetry with the a axis passing through the center
of the CC bond and close to bisecting the line between the two
fluorine atoms on each end. The b principal axis is perpendicular
to the plane of symmetry and coincident with the 2-fold
symmetry axis. Thus, in the gas-phase spectrum, the au modes
are associated with B-type bands, and the bu modes are
associated with hybrid A/C-type bands. Only the ag and bg
modes are Raman-active. The ag modes give polarized bands,
and the bg modes give depolarized bands.
For the normal species as well as both isotopomers, the
gauche rotamer has C2 symmetry. Its fundamentals consist of

Synthesis and Spectra of 1,1,2,2-Tetrafluoroethane
J. Phys. Chem. A, Vol. 104, No. 45, 2000 10095
TABLE 2: Vibrational Fundamentals of the Anti Rotamers of TFEA and Its 13C2 and d2 Isotopomers (in cm-1
)
normal species
selection rules and
approximate descript.
13C2 species
obs.
d2 species
obs.c
obs.a
3004 m
1443 w, dp
1145 m, p
1097 s, p
623 vs, p
361 vs, p
calc.b
ag: Raman-active, pol.
ν1 sym. C-H(D) stretch
ν2 sym. C-H(D) bend
ν3 sym. CF2 stretch
ν4 C-C stretch
ν5 sym. CF2 bend
ν6 sym. CF2 rock
3202
1532
1205
1148
628
2999 m, p
1420d w, p?
1131 m, p
1061 s, p
619 vs, p
361 s, p
2231d m, p
914 m, p
1333 w, p?
1062 s, p
608 vs, p
362 s, p
363
au: IR-active, B-type
ν7 asym. C-H(D) flap
ν8 asym. CF2 stretch
ν9 asym. CF2 twist
1331 s, B
1144.3 vs, B
210 w, B
1416 s
1203 vs
203 w
91 w
1325 s, B
1117 vs, B
209 w, B
82 w, B
956 m, B
1172.1 vs, B
210 w, B?
82 w, B?
ν
10 torsion
82 m?, B
bg: Raman-active, depol.
ν
ν
ν
11 asym.′ C-H(D) flap
12 asym.′ CF2 stretch
13 asym.′ CF2 twist
1362 m, dp
1083 sh, dp
479 m, dp
1446
1171
495
1355 m, dp
1053e m
478 w, dp
956 m, dp
1159 w, dp
472e w
bu: IR-active, A/C-type
ν
ν
ν
ν
ν
14 sym.′ C-H(D) str.
15 sym.′ C-H(D) bend
16 sym.′ CF2 stretch
17 sym.′ CF2 bend
2995 s, C
1309 s, C
1127.4 vs, A/C
542 m, A
3212 s
1358 s
1166 vs
535 m
418 s
2987 s, C
1299 s, C
1101 vs, A/C
539 s, A
2218 s, C
1015 s, C
1129.5 vs, A/C
539 m, A
18 sym.′ CF2 rock
413f s, ?
405f s, ?
372f s, A/C
a Reference 6 and additional infrared spectroscopy. Here, as elsewhere in the table, frequencies are for the liquid phase in Raman spectra and the
gas phase in infrared spectra. Abbreviations: vs, very strong; s, strong; m, medium; w, weak; vw, very weak; sh, shoulder; p, polarized; dp,
depolarized; band shapes: A, A-type; B, B-type; C, C-type; A/C, hybrid. b Reference 7. Numerical values for calculated absorptivities are for au
modes, 60, 337, 2, 2, respectively, and for bu modes, 54, 37, 204, 10, 50, respectively. c CD-rich modes are listed according to the approximate
description rather than in numerical order. d Fermi resonance multiplet. e Crystal-phase value. f Extensive structure.
tion of the crystal-phase infrared and Raman spectra gives direct
evidence for bands due to the anti species. At room temperature,
the gauche rotamer contributes only about 25% in the gas
phase,18 and many of its bands are close to those of the anti
rotamer. The amount of the gauche rotamer increases to about
44% in the liquid phase at -80 °C due to its polarity. Thus, the
contribution of the gauche rotamer is more apparent in the
Raman spectrum for the liquid phase even at low temperature.
Overall, the evidence for the assignments for the gauche
rotamers is weaker than for the anti rotamers.
In the far-infrared spectra many lines due to residual water
vapor in the spectrometer or from evaporation from the walls
of the glass sample cell are present and should be overlooked
in interpreting the band shapes.
Assignment for the Anti Rotamer of Normal TFEA. The
assignment for the normal species of TFEA agrees for the most
part with the work of Kalasinsky and co-workers.6 Thus, we
comment only on those assignments which are changed from
the previous ones or which remain in doubt. Table 2 summarizes
the fundamentals for the anti rotamers of TFEA and its two
isotopomers.
We reexamined the gas-phase and crystal-phase infrared
spectra of the normal species of TFEA except in the far-infrared
region. In addition, we had available the high-resolution (0.002
cm-1) infrared spectrum of the 900-1250-cm-1 region. This
gas-phase infrared spectrum for the very intense CF2 stretching
region was recorded on a Bruker IFS 120HR instrument in a
3-m cell at -100 °C (pressure about 0.15 Torr) by Dr. Michael
Lock at Justus Liebig University in Giessen, Germany. The full
analysis of the rotational structure in this spectrum is the subject
of a future paper, in which the spectrum will be shown. Because
the infrared spectra of the 13C2 species, which are part of the
present report, are close to the spectra of the normal species
and because the infrared spectra of the normal form have been
reported before,6 the infrared spectra of the normal species are
not included in this paper.
In general, our infrared spectra agreed with the previous
report.6 The exceptions were that the crystal-phase spectrum
showed a distinct medium-weak doublet near 1300 cm-1, the
medium-resolution gas-phase spectrum showed a small Q-
branch feature in the midst of the overlap region of the B-type
and C-type bands in the 1320 cm-1 region, and the high-
resolution infrared spectrum gave definitive information about
the CF2 stretching region. At the low temperature of the high-
resolution infrared experiment, the gauche rotamer makes a
small contribution of about 12% to the gas-phase mixture. From
the high-resolution spectrum it is certain that two intense bands
due to the anti rotamer overlap to give added strength to the
central feature. Furthermore, the higher frequency band centered
at 1144.34 cm-1 is of B-type shape, and the lower frequency
band centered at 1127.43 cm-1 is of A/C-type shape with the
A-component dominant. A definite Q branch is present in the
1127-cm-1 band in the spectrum at -100 °C. In addition, Stone
et al. investigated the lower frequency band with a diode-laser,
cold-jet-expansion technique and observed the stronger A-type
and weaker C-type character.10 In the previous report, the B-type
center at 1144 cm-1 in the low-resolution infrared spectrum was
assigned to the gauche rotamer, the strong central feature at
1136 cm-1 was assigned to ν8 of the anti species, and the feature
at 1125 cm-1 was assigned to the ν16 of the anti rotamer.6 The
B-type band is now confidently assigned to ν8 of the anti species,
and the A/C-type band remains assigned to ν16 of the anti species
but with an adjustment in frequency to the band center.
A second revision in the assignment for the anti rotamer of
the normal form is in the interpretation of the CH bending region
of the infrared spectrum. Kalasinsky and co-workers chose the
common central branch of the B-type and C-type bands for ν15
for the anti rotamer. We consider the strong Q branch of the
C-type band at 1309 cm-1, which they assigned to the gauche
rotamer, to be a better choice for the anti rotamer. This band
correlates well with a medium-weak doublet, which is regarded
as crystal splitting, in the infrared spectrum of the crystal phase.

10096 J. Phys. Chem. A, Vol. 104, No. 45, 2000
Craig et al.
Figure 1. Raman spectra of 1,1,2,2-tetrafluoroethane-13C2. Upper trace, liquid at -80 °C with parallel and perpendicular settings of the polarization
analyzer; lower trace, crystal-phase spectrum at -105 °C.
No binary combination tone of Bu symmetry for the anti rotamer
is predicted near this frequency, which could otherwise explain
the doublet.
The evidence for the Raman-active modes comes entirely
from the work of Kalasinsky et al., who examined the Raman
spectrum in the gas phase as well as in the liquid and crystal
phases. For only one Raman-active mode is the spectral evidence
questionable. That mode is ν12 of the bg symmetry species. The
evidence for this fundamental is a shoulder in the spectra of
gas and liquid phases which becomes a distinct feature in the
crystal-phase spectrum at 1080 cm-1. A similar feature is seen
in the crystal-phase Raman spectrum of the 13C2 isotopomer.
Strong evidence for this assignment comes, moreover, from the
spectra of the d2 isotopomer, as discussed below.
One other issue relates to the assignment of the infrared-
active fundamentals for the anti rotamer of the normal form.
This issue is the strength of the evidence for ν10(au). Kalasinsky
et al. assigned the band of B-type shape in the gas-phase infrared
spectrum at 82 cm-1 to the ν10 fundamental. They did not,
however, investigate the spectrum of the crystal phase in this
region. Thus, this band could be due to the gauche rotamer.
However, not only is the anti rotamer three times more abundant
in the gas phase, but the recent calculations of infrared intensities
by Kenny and co-workers,7 described more fully below, predict
a very weak intensity for the corresponding band for the gauche
rotamer compared to a weak intensity for this band of the anti
rotamer. Thus, it seems certain that the observed band is due to
the anti rotamer.
In the gas-phase infrared spectrum of TFEA, Kalasinsky et
al. observed extensive structure going to higher frequency
associated with the band at about 413 cm-1. They assign this
structure to hot bands involving various levels of excitation of
the torsion, ν10(au), interacting with ν18. We offer a different
interpretation below in the section on TFEA-d2.
In summary, the assignment of the eighteen vibrational
fundamentals of the anti rotamer of the normal species is, with
the possible exception of ν12, now on a firm experimental
footing. In Table 2 the experimental evidence for these
assignments is given, and the frequencies are compared with
those computed by Kenny et al. with an MP2/6-31G** model
without scaling.7 With the exception of two low frequencies
and the strongly anharmonic CH stretching frequencies, the
experimental frequencies are lower with an average difference
of -39 cm-1. The largest difference is -89 cm-1 for the CH
bending mode at 1443 cm-1. The questionable assignment of
the ν12 mode differs from the calculated value by -88 cm-1
.

Synthesis and Spectra of 1,1,2,2-Tetrafluoroethane
J. Phys. Chem. A, Vol. 104, No. 45, 2000 10097
Figure 2. Infrared spectra of 1,1,2,2-tetrafluoroethane-13C2. Upper trace, gas phase; lower trace, crystal phase at -196 °C after annealing at higher
temperatures.
In addition, as seen in Table 2, the calculated intensities for
infrared transitions agree qualitatively with the observed ones.
Overall, the comparison with the calculated frequencies supports
the assignment made from experimental data. Assignments
which differ from the work of Kalasinsky and co-workers are
in boldface in Table 2. The increased accuracy for ν8 and ν16 is
based on an analysis of the rotational structure in the high-
resolution infrared spectrum. In supplementary Table S1, all of
the features in the infrared and Raman spectra, including the
weak ones due to combination tones, are assigned.
Assignment for the Anti Rotamer of TFEA-13C2. In general,
the assignment of fundamentals for the anti rotamer of the 13C2
isotopomer proceeds directly from the assignments for the anti
rotamer of the normal species. Of course, decreases in frequency
are expected for the substitution of heavier isotopic masses,
although these shifts will be relatively small for 13C substitution.
Frequencies that depend strongly on motions of the CH bonds
will change little, whereas frequencies that depend on heavy
atom motions will change more.
Figure 1 gives the Raman spectra of TFEA-13C2 in the liquid
and crystal phases. From the crystal-phase Raman spectra clear
evidence exists for all of the ag and bg fundamentals except
two. A weak doublet is found near 1420 cm-1. This doublet
must be due to ν2, which also has a weak band in the Raman
spectrum of the normal species. The doublet can be explained
as Fermi resonance with ν4 + ν6 (Ag), which is predicted at
1422 cm-1. The most uncertain assignment is for ν12, which
was also uncertain for the normal species. Only eight distinct
bands are seen for nine Raman-active modes in the crystal-
phase spectrum. A shoulder is found, however, at 1053 cm-1
in the crystal-phase spectrum in Figure 1. We adopt this feature
for ν12 in accord with the corresponding assignment for the
normal species. Depolarization information gives further support
to the distinction between ag and bg modes. Table 2 gives details
of the spectroscopic features of bands assigned to the ag and bg
fundamentals in comparison with those for the normal species.
Figure 2 gives the infrared spectra of TFEA-13C2 in the gas
and crystal phases. The high-resolution gas-phase spectrum has
been recorded from 880 to 1200 cm-1 at -100 °C at Giessen,
but the rotational structure is yet to be analyzed. The crystal-
phase spectrum contains good evidence for eight of the nine
infrared-active fundamentals when each of the two features in
the very intense CF2 stretching region between 1000 and 1100
cm-1 is taken to represent a fundamental. Evidence for the ninth
infrared-active fundamental is found in the gas-phase spectrum
at 82 cm-1 below the accessible range of the low-temperature
cell used for crystal-phase spectra. This assignment of ν10 is
confidently made on the same basis as was used for the
equivalent mode in the normal species. Band shapes, including
those now available from the high-resolution spectrum of the
13C2 species in the CF stretching region of the gas-phase
spectrum, provide further support to the assignments.
A striking difference between the infrared spectra of the 13C2
species and the normal species is the two-times greater intensity
of the bands due to CF2 stretching in the spectrum of the 13C2
species. Because the intensity of the bands due to CH bending
are comparably weaker in the spectrum of the 13C2 species, the
mixing of CF2 stretching with the bending modes must be
greater in the normal species. This diminished transfer of
intensity from the CF2 stretching modes to the CH bending

10098 J. Phys. Chem. A, Vol. 104, No. 45, 2000
Craig et al.
Figure 3. Raman spectra of 1,1,2,2-tetrafluoroethane-d2. Upper trace, liquid at -80 °C with parallel and perpendicular settings of the polarization
analyzer; lower trace, crystal-phase spectrum at -105 °C.
TABLE 3: Teller-Redlich Product Ratios for the Anti
Assignment for the Anti Rotamer of TFEA-d2. Figure 3
contains the Raman spectra of TFEA-d2 in the liquid and crystal
phases. In the Raman spectrum of the crystal phase, nine bands
with some complications due to crystal splitting and Fermi
resonance are found for the anti rotamer. Reinforced with
polarization data from the liquid-phase Raman spectrum, the
assignment of the nine Raman-active fundamentals of the anti
rotamer is straightforward. Selection of the correct frequency
for ν13 of bg symmetry is, however, a problem. The obvious
peak in the liquid-phase Raman spectrum in the 480-cm-1 region
gives a frequency that is larger than the value for the normal
isotopomer. The corresponding peak in the crystal-phase
spectrum is at 472 cm-1, which is a reasonable value. We use
this frequency in Table 2. The change in intensity pattern
between the liquid phase and the crystal phase either reflects
the loss of an overlapping band due to the gauche rotamer or
Fermi resonance. A combination tone for the gauche rotamer
is predicted as ν8 + ν10 ) 483 cm-1 (A) but is unlikely to have
sufficient intensity. A possible partner for Fermi resonance in
the anti rotamer is ν10 + ν18 (Bg), which has a predicted gas-
phase frequency of 454 cm-1. If the frequencies of the
contributing fundamentals increase between the gas phase and
the liquid phase, as is observed for some low-frequency modes,
then Fermi resonance is possible. We favor this latter explana-
tion. Strong spectral evidence for ν12, which was a doubtful
assignment for the other two isotopomers, is the depolarized
band at 1159 cm-1 in the Raman spectrum. All the frequencies
of the Raman-active fundamentals and their spectral evidence
are summarized in Table 2 in comparison with the assignments
for the normal and 13C2 isotopomers.
Rotamer of TFEA
molecules
sym. species obs. ratio calc. ratio difference
13C2/normal
ag
au
bg
bu
ag
au
bg
bu
0.932
0.967
0.965
0.943
0.519
0.736
0.740
0.516
0.926
0.970
0.965
0.941
0.508
0.714
0.723
0.510
0.6%
0.3%
0.2%
0.2%
2.2%
3.1%
2.3%
1.2%
d2/normal
modes in the spectrum of the 13C2 species is consistent with
the greater frequency difference between the two types of modes
for the 13C2 species. The greater frequency difference is a
consequence of the greater 13C isotope effect on the CF2 modes
than on the CH bending modes.
The assignments of all eighteen fundamentals of TFEA-13C2
are given in Table 2 in comparison with those of the normal
species. Except for ν10, which appears unchanged, all funda-
mentals of the 13C2 species have lower frequencies than their
counterparts for the normal species. Thus, the Rayleigh Rule
for isotopic substitution is satisfied. In addition, the frequency
shifts are largest for modes in the central part of the infrared
region for which substitution of 13C is expected to have the
largest effect on normal modes. A quantitative check on the
assignment for the 13C2 species in relation to the assignment
for the normal species is provided by the Teller-Redlich product
rules. The satisfactory comparison of the observed frequency
product ratios for each symmetry species with the predicted ones
is shown in Table 3. Supplementary Table S2 gives the details
of the spectra of the 13C2 species and their assignments.
Figure 4 contains the infrared spectra of TFEA-d2 in the gas

Synthesis and Spectra of 1,1,2,2-Tetrafluoroethane
J. Phys. Chem. A, Vol. 104, No. 45, 2000 10099
Figure 4. Infrared spectra of 1,1,2,2-tetrafluoroethane-d2. Upper trace, gas phase; lower trace, crystal phase at -196 °C before annealing.
these features are so prominent, widely spaced by about 4 cm-1
and crystal phases. The high-resolution infrared spectrum in the
gas phase was recorded from 930 to 1230 cm-1 at -100 °C in
Giessen. With an allowance for crystal splitting, eight bands
for the anti rotamer are evident in the crystal-phase spectrum.
The remaining fundamental is the band at 82 cm-1 of probably
B-type shape in the gas-phase infrared spectrum. Elsewhere in
the gas-phase infrared spectrum, band shapes support the
assignments to the au and bu symmetry species. Reasonable
frequency shifts are found for modes that are rich in CD motion.
Notable changes are observed in the two antisymmetric CF2
stretching frequencies compared to the spectrum of the normal
species. These bands have higher frequencies than for the normal
species. This adjustment is consistent with the Rayleigh Rule
because the frequencies of CD bending modes are below those
of the CF2 stretching modes, whereas the frequencies of CH
bending modes are above. In the summary of frequencies of
fundamentals of the d2 species in Table 2, the CD-rich bending
modes are aligned with the approximate descriptions and not
in the conventional frequency order. As given in Table 3, the
application of the Teller-Redlich rule shows that the assign-
ments of the fundamentals for the normal isotopomer are
consistent with those of the d2 isotopomer.
,
of comparable intensity, and on the high-frequency side of the
band raises doubt about this interpretation. The significant
displacement of these features to the high-frequency side of the
ν18 band in the spectrum of the d2 species is strong evidence
against the simple hot band interpretation. A more satisfactory
interpretation attributes these features to Coriolis coupling
between the upper state of the ν18 transition and the series of
upper states for the transitions ν6 + (n + 1)ν10 - nν10. For n of
1, ν6 + ν10 (Au) is within one cm-1 of 444 cm-1 for all three
isotopomers. Coriolis coupling is allowed because the symmetry
product Au × Bu is Bg, which is the symmetry species of two
rotations. Frequencies of the three series of sharp lines are in
footnotes to Tables S1, S2, and S3.
All of the assignments of fundamentals of the anti rotamer
of the d2 isotopomer are firmly based on experimental evidence.
Since these assignments are consistent with those for the normal
species as reflected in the satisfactory product rule ratios in Table
3, we have strong evidence in support of the one doubtful
assignment for the normal species, namely, for ν12. In turn, the
corresponding assignment of ν12 for the 13C2 species is sup-
ported. The increased accuracy in the values for ν8 and ν16
comes from the analysis of the rotational structure in the high-
resolution infrared spectrum. Supplementary Table S3 gives all
of the observed bands for the d2 species, including the weaker
ones, and their assignments.
As noted in the discussion of the spectrum of the normal
species, unusual and extensive structure extends to the high-
frequency side of the band for ν18(bu) near 418 cm-1 in the
infrared spectrum. For the d2 species this structure is less
pronounced and further from the band center for ν18. In addition,
the band shape for ν18 of the d2 isotopomer is less distorted.
For the normal species, Kalasinsky and co-workers assigned
this structure to hot bands of the type ν18 + nν10 - nν10.6 That
Probable Impurity. A series of seven sharp, weak features
begins at 1089.1 cm-1 and runs to 1101.7 cm-1 in the infrared
spectrum of the gas phase, Figure 4, of TFEA-d2. These lines
are spaced almost equally by about 2 cm-1. Since these features

10100 J. Phys. Chem. A, Vol. 104, No. 45, 2000
Craig et al.
TABLE 4: Vibrational Fundamentals of the Gauche Rotamers of TFEA and its 13C2 and d2 Isotopomers (in cm-1
)
normal species
IRa
13C2 species
d2 speciesc
selection rules and
approximate descript.
Ramana
3004 p
calc.b
Raman
IR
Raman
IR
a: Raman, pol.; IR, C-type
ν1 sym. C-H(D) stretch
ν2 sym. C-H(D) bend
ν3 sym. C-H(D) rock
ν4 sym. CF2 stretch
ν5 sym.′ CF2 stretch
ν6 C-C stretch
3197 m
1510 w
1451 m
1249 s
1165 w
953 m
601 w
413 w
253 w
83 vw
2999 m,p
1168 m, p
2231 m,p
772 m, p
1046 s, p
1432 w, C
1363 mw, Q
1206 m, C
[1105]
905.8 m, C
598 w, C
1355 w, C
1183 s, C
773 w, ?
1338 m, C
1053 m, C
1195 m, p
1100 sh
898 m, p
595 m, p
412 m, p
257 w, p
879 m, p
591 m, p
412 w, p
259 w, p
885 m, C
595 m, C
ν7 sym. CF2 flap
588 m, p
408 m, p
258 w, p
591 w, C
ν8 sym. CF2 deform.
ν9 sym. CF2 twist
ν
10 torsion
[75]
b: Raman, depol.; IR, A/B type
ν
ν
ν
ν
ν
ν
ν
ν
11 asym. C-H(D) stretch
12 asym. C-H(D) bend
13 asym. C-H(D) rock
14 asym. CF2 stretch
15 sym.′ CF2 stretch
16 asym. CF2 flap
3003 m, Q
1393 w, B
1320 m, Q
[1135]
1075 w, ?
780 ms, A
523 w, A
∼240 w
3189 m
1477 m
1392 m
1194 vs
1161 m
803 s
2979 m, Q
1382 w, B
2225 w, A?
1204 m, B?
1390d sh
1323 w, dp
1318 w, dp
991 vw, dp
1047 w, sh
763 m, A
520 sh, A
∼240 w
775 w, dp
530 vw, br
247 w, p?
760 w, dp
519 w, dp
741 w, dp
511 w, dp
240 vw, dp
743 m, ?
510 w, A
∼240 w, ?
17 asym′ CF2 deform.
18 asym. CF2 twist
512 w
232 w
a See Table 2. Also, br, broad. b Reference 7. Numerical values of calculated absorptivities are for a modes, 47, 5, 25, 125, 9, 47, 5, 3, 2, 1,
respectively, and for b modes, 23, 38, 36, 274, 36, 74, 15, 6, respectively. c See Table 2. d Gas-phase value.
appeared with equal intensity in two different preparations of
TFEA-d2, we thought initially that these lines were due to
difference tones involving ν12(bg) and transitions between
various states in the ν10(au) manifold. However, the intensities
of the lines did not fall off as expected for difference tones,
and these lines are present in the spectrum at -100 °C without
an appreciable intensity falloff and with a width that is too
narrow for Q branches of bands. In addition, similar series are
not found for the sum tones at higher frequency in the infrared
spectrum of TFEA-d2. Nor are comparable lines for possible
difference tones found in the spectra of the other two isoto-
pomers. Thus, we conclude that this series of lines is due to an
unidentified impurity with a relatively low molecular mass.
Assignment for the Gauche Rotamer of TFEA. Several
adjustments in the previous assignment for the gauche rotamer
of the normal species of TFEA are proposed.6 Table 4 gives
details of the assignments. Columns 1 and 2 of the data give
the Raman and infrared spectral evidence, respectively, which
is drawn from the work of Kalasinsky and co-workers and from
our reexamination of the gas-phase and crystal-phase infrared
spectra. Assignments that we have revised are shown in boldface
type. Column 3 gives the predictions of frequencies and infrared
intensities from the recent calculations of Kenny et al.7
discernible rotational structure, CF2H2, CF3H, or H2FCCFH2,
accounts for the band. Support for accepting this structured band
for ν2 is the occurrence of extensive hot band structure
associated with Q branches elsewhere in the spectrum. A
combination tone predicted at 1429 cm-1 for ν6 + ν17 (B) is an
alternative assignment for the 1432-cm-1 feature. This prediction
is, however, based on gas-phase values and thus implies a yet
lower frequency than 1429 cm-1 for the combination tone. Thus,
we tentatively assign ν2 as 1432 cm-1. The C-type band at 1206
cm-1, assigned to ν4, decreases significantly in relative intensity
in the infrared spectrum at -100 °C, thereby confirming this
assignment to the gauche rotamer.
Another adjustment in the assignment for the a symmetry
species of the gauche rotamer is to reject 1146 cm-1 for ν5,
which is predicted to have only a weak intensity in the infrared
spectrum. This feature is now known to be the R branch of a
B-type band of the anti rotamer based on the high-resolution
infrared spectrum. In addition, because the CF2 stretching
frequencies for the anti rotamer are, on the average, 60 cm-1
below the calculated values, the calculated value of 1165 cm-1
suggests that this mode will be around 1105 cm-1. With the
band’s weak predicted intensity and the low concentration of
the gauche rotamer, this band is probably hidden under the very
intense CF2 stretching bands of the anti rotamer in the infrared
spectrum. Another possibility is to interpret the shoulder,
reported at 1100 cm-1 in the liquid-phase Raman spectrum and
previously assigned as ν14, as evidence for this totally symmetric
fundamental. We have done so. Stone et al. investigated ν6 with
the jet-beam, high-resolution method.10 From their analysis of
the rotational structure, they made a definitive assignment as a
C-type band with a band center of 905.834 cm-1. This band
for the gauche rotamer was observable at low temperature
because vibrational relaxation from the gauche to the anti
rotamer does not occur fully in the rapid cooling of the jet beam.
Three low-frequency modes of the a symmetry species remain
to be discussed. For ν8, which is strongly supported by a
For the modes of the a symmetry species, the previous choice
of the C-type band in the infrared spectrum at 1362 cm-1 for
ν2 was made unreasonable by the calculated frequency of 1510
cm-1. We have reassigned 1362 cm-1, as 1363 cm-1, to ν3 of
the gauche rotamer and have reassigned 1309 cm-1, which had
been assigned to ν3, to the anti rotamer as discussed above. For
ν2 the only reasonable possibility is a weak, C-type band at
1432 cm-1 with an absorbance of about one-fifth of that of the
C-type band at 1363 cm-1. In accord with this observation, the
calculations of Kenny et al. predict an absorbance ratio of five
for these two bands. This predicted absorbance ratio also implies
that the band for ν2 should be observable. The alternative
possibilities, the very weak Q branches at 1457 and 1413 cm-1
,
are too weak to be credible assignments for ν2. The band at
1432 cm-1 has extensive structure in the Q branch (about 1-cm-1
spacing) and fainter structure in the R branch (about 1.3-cm-1
spacing). This structure suggested that the band might be due
to an impurity. None of the likely impurities that would have
polarized Raman feature of medium intensity at 412 cm-1
,
Kalasinsky et al. also assigned this mode to a Q-branch feature
at 408 cm-1 in the infrared spectrum. Given the prediction of
a weak intensity for the infrared transition for ν8 and its near
coincidence with the intense and complex band for ν18 for the

Synthesis and Spectra of 1,1,2,2-Tetrafluoroethane
J. Phys. Chem. A, Vol. 104, No. 45, 2000 10101
anti rotamer, we think it unlikely that the feature of medium
intensity is due to ν8 of the gauche rotamer. Similar reasoning
rules out observing a band for ν9 in the infrared spectrum,
although this mode is represented by a polarized band at 257
cm-1 in the Raman spectrum. A weak intensity is predicted for
the band for ν9 in the infrared spectrum, and the greater intensity
predicted for the band for ν18 should dominate this region. At
present no direct spectral evidence exists for ν10, which is
calculated at 83 cm-1. The infrared spectrum in this region is
dominated by the band for the corresponding mode of the anti
rotamer. A revised prediction of 75 cm-1 is based on the
difference between the calculated and observed frequencies for
the torsional mode of the anti rotamer. This prediction is
consistent with the previous estimate of 72 cm-1.6 This estimate
was made from the hot band structure of the infrared band at
780 cm-1 and a series of Q branches observed near 849 cm-1
at high sample pressure. We confirmed this latter observation.
this region. The weakness of the band may have contributed to
uncertain polarization information.
In summary, the assignment of fundamentals for the gauche
rotamer has been substantially revised but remains incomplete
in contrast to the complete assignment for the anti rotamer. Table
4 summarizes the assignments. Experimental support for two
of the CF2 stretching modes, ν5 and ν15, is sketchy. No direct
spectral evidence exists for one of the other CF2 stretching
modes, ν14, or for the torsional mode, ν10. The assignment for
ν18 is guided by the calculated value, although some spectral
evidence is available. Even though bands due to all of the
fundamentals of the gauche rotamer are both infrared- and
Raman-active, bands due to eight modes are observed in only
one of the two spectroscopies. Overall the agreement between
observed frequencies and calculated ones is comparable to that
for the anti rotamer. Insofar as intensities can be compared,
observations and calculations concur. Details of the assignment
of all of the features of the spectra of TFEA are in supplementary
Table S1.
Assignment for the Gauche Rotamer of TFEA-13C2. The
assignment of vibrational fundamentals for the gauche rotamer
of TFEA-13C2 follows directly from the assignment for the
gauche rotamer of the normal species. The assignments from
the Raman and infrared spectra are given in Table 4 in data
columns 4 and 5, respectively. Figure 1 shows the liquid-phase
and crystal-phase Raman spectra of TFEA-13C2. Figure 2 shows
the gas-phase and crystal-phase infrared spectra of the 13C2
species.
For the modes of the b symmetry species, several comments
apply. As discussed in the section on the anti rotamer of TFEA,
we have revised the assignments in the CH bending region. We
consider the small Q branch at 1320 cm-1 on top of the strong
absorption caused by the overlap of two bands of the anti
rotamer to be due to ν13 of the gauche rotamer. This choice is
reinforced by the earlier designation of a depolarized Raman
band of medium intensity at 1323 cm-1 for this mode.
Unfortunately, the value of 1312 cm-1 from the gas-phase
Raman spectrum does not agree with the value of 1320 cm-1
from the gas-phase infrared spectrum.6 This discrepancy is
discounted because the Raman band is reported as very weak
and is not even apparent in the published spectrum.
For the modes of the a symmetry species for the gauche
rotamer of the 13C2 species, only a few comments are needed.
A weak feature for ν2 is not found in the 1420-cm-1 region, as
would be expected from the corresponding assignment of the
normal species. However, as discussed above, weaker intensities
are found for CH bending modes of the 13C2 species in
comparison with those of the normal species. In the high-
resolution infrared spectrum the intensity of the C-type band at
1183 cm-1 decreases appreciably in going to -100 °C, thereby
confirming that this band is due to the gauche rotamer.
Frequency shifts between the two isotopomers are reasonable
except that the frequency for ν9 appears to shift up by 2 cm-1
in going to the heavier isotopomer. This apparent anomaly is
attributed to uncertainties of (2 cm-1 in measurements in the
Raman spectra.
The previous assignment of the ν14 to the shoulder at 1100
cm-1 in the Raman spectrum is unlikely because of a higher
predicted frequency. As discussed above, this shoulder has been
reassigned to a mode of the a symmetry species. An estimate
for ν14 is about 1135 cm-1 (ca 1194-60 cm-1). Despite the
prediction of a very strong band in the infrared spectrum, the
predicted frequency for the gauche rotamer falls in the region
where the band would be overlapped by the intense CF2
stretching modes of the more abundant anti rotamer. The
estimate of 1135 cm-1 for ν14 in the gas phase is used in Table
4.
Evidence for ν15 is uncertain. The weak band of indistin-
guishable shape at 1075 cm-1 in the infrared spectrum was
previously assigned to ν15. Because this band appears to be
absent from the spectrum at -100 °C, its assignment to the
gauche rotamer is supported. Furthermore, no binary combina-
For the modes of the b symmetry species, a similar agreement
is found between the observations for the 13C2 species and the
normal species. The pattern of Raman and infrared data is
familiar with one exception. No Q branch appears for ν13 in
the infrared spectrum, although a weak polarized band in the
Raman spectrum at 1318 cm-1 reveals ν13. As was the case for
the normal species, no spectral feature is available for ν14. The
evidence for ν15 is more problematic than even for the normal
species. In the infrared spectrum the feature at 1058 cm-1
persists at -100 °C and thus must be due to the anti rotamer.
A weak broadening on the P-branch side of this band in the
room-temperature spectrum appears to be absent at -100 °C.
This feature is assigned to ν15 in accord with the interpretation
for the normal species.
tion tone for the gauche rotamer is predicted near 1075 cm-1
.
A drawback to this assignment is the estimated frequency of
1100 cm-1 (1161-60 cm-1), which is 25 cm-1 above the
observation.
The biggest adjustment in the assignments for the modes of
the b symmetry species is replacing 495 cm-1 for ν18 with the
approximate value of 240 cm-1. The calculated frequency is
232 cm-1. The predictions for all of the low-frequency modes
of the anti rotamer are within 15 cm-1 of the observed values,
and a similar agreement is seen for the gauche fundamentals
that have been confidently assigned. The band for ν18 is probably
contributing to the distinct hump on the high-frequency side of
the distorted B-type band at 210 cm-1. It is also possible that
the weak liquid-phase Raman feature reported by Kalasinsky
et al. at 247 cm-1 reflects this mode. This feature and its weak
neighbor at 257 cm-1 are, however, both reported as polarized.
No reasonable explanation exists for two polarized bands in
For ν18, the evidence is the broad shoulder in the infrared
spectrum in the 240 cm-1 region. This observation not only
supports the assignment for the 13C2 species but also reinforces
the interpretation of the spectrum of the normal species. In
assessing a weak feature of this type, it is regrettable that
calculated Raman intensities are not available. All of the
observed modes of the 13C2 species for the b symmetry species

10102 J. Phys. Chem. A, Vol. 104, No. 45, 2000
Craig et al.
satisfy the Rayleigh Rule. In addition, shifts in frequencies due
to substitution of the heavier isotope are reasonable.
assignment for the 13C2 species. This incompleteness is in
contrast to the anti rotamer of the d2 species for which the
strongest evidence exists for a complete assignment. Supple-
mentary Table S3 gives the details of the observations and
assignments of all the spectral features for the d2 isotopomer.
In summary, the assignment of the vibrational fundamentals
for the gauche rotamer of the 13C2 species has a few more gaps
than the incomplete assignment for the gauche rotamer of the
normal species. Evidence for two CF2 stretching modes is
absent. No experimental evidence exists for the frequency of
the torsion. Qualitatively, frequency shifts between the two
isotopomers are reasonable. Satisfying these relationships
reinforces the assignments for both species. Because of the gaps
in the assignments, the Teller-Redlich product rule is inap-
plicable in assessing the assignments of the gauche rotamers of
the d0 and 13C2 species. Details of the assignment of all of the
features of the spectra of the 13C2 species are in supplementary
Table S2.
Conclusions
An efficient method has been developed for synthesizing
small amounts of the 13C2 and d2 isotopomers of TFEA by
preparing isotopically substituted TBrEA from acetylene-13C2
and acetylene-d2 and then reacting the substituted TBrEA with
AgF2. Intermediates and products have been characterized.
The published assignments of vibrational fundamentals for
the rotamers of the normal form of TFEA have been revised.
High-resolution, gas-phase infrared spectra at low temperature
from the work of Stone et al. and from Giessen, though
fragmentary, helped considerably. For the anti rotamer two
adjustments have been made based on the new infrared spectral
evidence. For the gauche rotamer, seven adjustments have been
made in assignments of fundamentals, and estimates for the
missing ones are proposed. Several of the revisions were
dependent on the calculations of frequencies and infrared
intensities reported by Kenny et al. The assignments for the
two rotamers of the normal form are in Tables 2 and 4.
Assignments of the fundamentals of both rotamers of TFEA
have undergone a remarkably large change since the initial,
extensive investigation of Klaboe and Nielsen. Nine fundamen-
tals for the anti rotamer and eight for the gauche rotamer have
been revised.
Turning to a closer comparison of the observed and calculated
frequencies for the rotamers of the normal species, we see that,
as commonly found, the observed values are smaller than the
calculated values with some exceptions at low frequencies. A
simple scaling rule for adjusting the calculated frequencies is
insufficient. A better procedure is to recognize that the differ-
ences between calculated and observed frequencies are ap-
proximately equal for fundamentals of similar normal modes.
In this way, the expected values for ν5, ν10, and ν14 are estimated
as 1100, 75, and 1135 cm-1, respectively, for the gauche rotamer
of the normal species.
Based on infrared and Raman spectral evidence, assignments
of fundamentals are proposed for both rotamers of the 13C2 and
d2 isotopomers. For the two anti rotamers of the isotopomers
the assignments are complete. For all three isotopomers, the
occurrence of only the anti rotamer in crystalline material guides
the assignments for this rotamer, which are summarized in Table
2. The assignments for the anti rotamers of the three isotopomers
are consistent with Teller-Redlich product rule ratios.
For the less abundant gauche rotamers of the three isoto-
pomers, the identification of the vibrational fundamentals
remains incomplete. Dominance of spectral regions by bands
of the more abundant anti rotamer coupled with the absence of
a way to observe bands of the gauche rotamer alone compromise
the experimental evidence for the fundamentals of the gauche
rotamer. The assignments for the fundamentals of the gauche
rotamers of the three isotopomers are summarized in Table 4.
Frequencies in brackets for the normal species are estimates.
In the absence of ab initio calculations for the 13C2 and d2
isotopomers, good estimates cannot be made for the missing
frequencies for these species.
Assignment for the Gauche Rotamer of TFEA-d2. Spectral
data for use in the assignment of fundamentals of the gauche
rotamer of TFEA-d2 are in Figures 3 and 4. The first contains
the Raman spectra; the second the infrared spectra. The
assignments are summarized in Table 4, where data columns 6
and 7 are for the d2 species. The numbering of frequencies for
the d2 species in Table 4 conserves approximate descriptions
of normal modes rather than descending frequency order.
Compared to the assignment for the gauche rotamer of the
normal species, the assignment for the d2 species is more
problematic. A cause is the congestion of frequencies for CD
bending modes with those for the CF and CC stretching modes.
Reasonable spectral evidence, mostly as polarized bands in
the Raman spectrum, exists for six of the 10 fundamentals of
the a symmetry species. In addition, an indistinct Q-branch
feature at 2225 cm-1 is used for ν11 in accord with assignments
for the gauche rotamers of the other two isotopomers. No
spectral evidence exists for ν2, which was a weak feature in
the spectrum of the normal species, and no evidence exists for
ν6. The use of the strong Raman band at 1046 cm-1 for ν5 is
questionable. This band seems too intense for the minority
component of the rotamer mixture. Attempts to attribute this
band to Fermi resonance in the anti rotamer required an unlikely
ternary combination. In addition, the crystal-phase spectrum
shows no sign of splitting due to Fermi resonance. Supporting
this assignment is a corresponding band at 1053 cm-1 in the
infrared spectrum, which disappears at -100 °C. No evidence
was found in the liquid-phase Raman spectrum for ν10, the
elusive torsional mode, despite searching down to the exciting
line with a notch filter in front of the spectrometer slit.
For the modes of the b symmetry species, the spectral
evidence is comparably scanty to that for the totally symmetric
species. Features, albeit some questionable ones, exist for six
of the eight fundamentals of this symmetry species. Good
evidence exists for two CF2 bending modes, ν16 and ν17, in both
the Raman spectrum and the infrared spectrum. For ν18, as for
the gauche rotamers of the other two isotopomers, we interpret
the broad shoulder in the infrared spectrum at about 240 cm-1
and a weak depolarized band in the Raman spectrum at 240
cm-1 as evidence for ν18. A very weak, apparently depolarized
Raman band at about 991 cm-1 is a possibility for ν13. The
weak feature at 980 cm-1 remains at -100 °C and thus is
assigned to a combination tone of the anti rotamer. The B-like
band at 1204 cm-1 in the infrared spectrum is a reasonable
candidate for ν14 because it loses intensity as the temperature
is decreased. No spectral evidence was found for the other two
fundamentals of the b symmetry species.
Because the assignments of the fundamentals of the anti
rotamers of the three isotopomers are well supported by
experimental observations and calculations, a firm foundation
exists for the investigation of the complete structure of the anti
As shown in Table 4, the assignment of fundamentals for
the gauche rotamer of the d2 species is as incomplete as the

Synthesis and Spectra of 1,1,2,2-Tetrafluoroethane
J. Phys. Chem. A, Vol. 104, No. 45, 2000 10103
(4) Wiberg, K. B. Acct. Chem. Res. 1996, 29, 229.
(5) Engkvist, O.; Kalstrom, G.; Widmark, P.-O. Chem. Phys. Lett. 1997,
265, 19.
(6) Kalasinsky, V. F.; Anjaria, H. V.; Little, T. S. J. Phys. Chem. 1982,
86, 1351.
(7) Papasavva, S.; Illinger, K. H.; Kenny, J. E. J. Phys. Chem. 1996,
100, 10 100.
rotamer of TFEA by high-resolution infrared spectroscopy.
Further clarification of the assignment of fundamentals for the
gauche rotamer awaits a study of the high-resolution infrared
spectra for the gas phase over the whole spectral range at low
temperature.
(8) Epiotis, N. D. Deciphering the Chemical Code. Bonding Across
the Periodic Table; VCH Publishers: New York, 1996; Chapter 11.
(9) Takeo, H.; Matsumura, C.; Morino, Y. J. Chem. Phys. 1986, 84,
4205.
(10) Stone, S. C.; Philips, L. A.; Fraser, G. T.; Lovas, F. J.; Xu, L.-H.;
Sharpe, S. W. J. Mol. Spectrosc. 1998, 192, 75.
(11) Klaboe, P.; Nielsen, J. R. J. Chem. Phys. 1960, 32, 899.
(12) Chen, Y.; Paddison, S. J.; Tschuikow-Roux, E. J. Phys. Chem. 1994,
98, 1100.
Acknowledgment. This research was supported by National
Science Foundation grant CHE 9710375 and by Oberlin College.
We are grateful to Dr. Michael Lock at Justus Liebig University
in Giessen, Germany for recording the high-resolution infrared
spectra and to Joshua B. Burton for his contributions to the
development of the synthetic method.
(13) Henne, A. L.; Midgley, T. J. Am. Chem. Soc. 1936, 58, 884.
(14) Smith, K.; Newnham, D.; Page, M.; Ballard, J.; Duxbury, G. J.
Quant. Spectrosc. Radiat. Transfer 1998, 59, 437.
(15) Becker, E. D. High-Resolution NMR, Theory and Chemical
Applications, 2nd ed.; Academic Press: New York, 1980.
(16) Cavalli, L.; Abraham, R. J. Mol. Phys. 1970, 19, 265.
(17) Ackermann, M. N.; Burdge, J. J.; Craig, N. C. J. Chem. Phys. 1973,
58, 203.
Supporting Information Available: Tables S1-S3 contain
details of the spectra and assignments for each of the isotopic
species of TFEA. This material is available free of charge via
the Internet at http://pubs.acs.org.
References and Notes
(1) Durig, J. R.; Liu, J.; Little, T. S.; Kalasinsky, V. F. J. Phys. Chem.
1992, 96, 8224.
(2) Craig, N. C.; Chen, A.; Suh, K.-H.; Klee, S.; Mellau, G. C.;
Winnewisser, B. P.; Winnewisser, M. J. Phys. Chem. A 1997, 101, 9302.
(3) Muir, M.; Baker, J. Mol. Phys. 1996, 89, 211.
(18) Calculated with the energy differences reported in ref 6 and
assuming ∆rS° ) R ln 2 for the anti-to-gauche conversion. The two rotamers
have the same rotational symmetry number (2), and the gauche has weight
2 because of two forms of this rotamer. For the gas phase, ∆rH° ) 4.94
kJ/mol; for the liquid phase, ∆rH° ) 1.51 kJ/mol.