Solute-solvent interactions in liquid noble gases as probed by the conformational equilibrium of some 1,2-disubstituted ethanes

Article abstract of DOI:10.1021/jp953713q

Using variable-temperature infrared spectroscopy, the enthalpy difference between the trans and gauche conformers of 1-chloro-2-fluoroethane, 1,2-dichloroethane, and 1-chloropropane has been determined in the gas phase and in solutions of liquefied krypton and xenon. Using a simple reaction field model, the changes in the enthalpy difference between vapor phase and solution are quantitatively explained as the result of weak dipole-induced dipole interactions appearing between the solute molecules and the surrounding noble gas atoms.

Full text of DOI:10.1021/jp953713q

J. Phys. Chem. 1996, 100, 9671-9677
9671
Solute-Solvent Interactions in Liquid Noble Gases As Probed by the Conformational
Equilibrium of Some 1,2-Disubstituted Ethanes
W. A. Herrebout and B. J. van der Veken*
Department of Chemistry, UniVersitair Centrum Antwerpen, Groenenborgerlaan 171, B2020-Antwerp, Belgium
ReceiVed: December 13, 1995; In Final Form: March 18, 1996X
Using variable-temperature infrared spectroscopy, the enthalpy difference between the trans and gauche
conformers of 1-chloro-2-fluoroethane, 1,2-dichloroethane, and 1-chloropropane has been determined in the
gas phase and in solutions of liquefied krypton and xenon. Using a simple reaction field model, the changes
in the enthalpy differences between vapor phase and solution are quantitatively explained as the result of
weak dipole-induced dipole interactions appearing between the solute molecules and the surrounding noble
gas atoms.
Introduction
selected were 1,2-dichloroethane, 1-chloro-2-fluoroethane, and
1-chloropropane. From this list, for instance, the potentially
In cryogenic solvents such as liquefied xenon, krypton, and
argon, the vibrational frequencies of solute molecules are
substantially closer to the vapor phase values than those
observed in the corresponding solid matrices. Therefore, these
solutions have been described as a pseudo gas phase,1 and over
the years this description has persisted.2-6 The extensive
rotational fine structure observed in these solutions for diatomic
molecules such as HF and HCl7,8 appears to bear out this notion.
An implication of the pseudo gas phase hypothesis is that in
such solutions the thermodynamics of a conformational equi-
librium should be similar to that in the vapor phase. Although
only a very limited number of conformational equilibria have
been quantified in cryogenic solutions, this conclusion in at least
one case has been contradicted: the enthalpy difference between
trans- and gauche-1,2-dichloroethane in liquefied xenon was
determined by Turner et al.9 to be 3.87 ( 0.09 kJ mol-1, while
the accepted vapor phase value is 4.90 kJ mol-1. More recently,
the enthalpy differences between the gauche and trans conform-
ers of n-butane in gas, liquefied argon, krypton, and xenon have
been reported,10 and in this case no significant change of ∆H°
has been found. These two studies do not necessarily contradict
each other, however. If contributions from higher multipole
moments are neglected, the interaction between solute and
solvent depends on the dipole moments. For the trans
conformer of 1,2-dichloroethane, the moment is zero, while for
the gauche conformer it is 2.55 D. Therefore, the two
conformers will be stabilized to a different extent by the solvent.
For n-butane, on the other hand, both conformers have a very
small, and, therefore, similar, dipole moment and no differential
stabilization is expected.
interesting 1,2-difluoroethane is missing on account of the very
limited population of its second conformer under cryogenic
circumstances. Each of the selected molecules occurs as a
mixture of gauche and trans conformers, defined in Figure 1.
For 1,2-dichloroethane and 1-chloro-2-fluoroethane the dipole
moments of the two conformers are sufficiently different, so
that significant differences in their enthalpy difference for the
different phases studied can be anticipated. For 1-chloropro-
pane, on the other hand, both conformers have a substantial
but very similar dipole moment, so that the solvent influences
on its conformational equilibrium must be expected to be
different from those for the other two compounds.
The rationalization of the experimental data calls for an
underlying quantitative theory. For this, we have chosen the
low road. Recently, the reaction field theory of Onsager11 and
Kirkwood12 has been revived because it can be incorporated
into quantum chemical calculations.13 The success of this
approach is witnessed by its implementation in the Gaussian
92 suite of programs,14 and the ability of reaction field theory
combined with ab initio calculations in the rationalization of
solvent effects has been recently demonstrated.15 In spite of
its simplicity, it will be shown that this theory satisfactorily
supports the present data.
Experimental Section
The sample of 1,2-dichloroethane (Jansen Chimica 11.336.84;
99.8+%) was obtained commercially. The sample was dried
over 4 Å molecular sieves and used without further purification.
The samples of 1-chloro-2-fluoroethane and 1-chloropropane
were synthesized by mixing the required alcohol (2-fluoroeth-
anol, Aldrich 16.034-2 and 1-propanol, Aldrich 11.003-5) with
PCl3 at room temperature. These compounds were purified on
a low-temperature, low-pressure fractionating column. The
noble gases used in this study, argon, krypton and xenon, have
a stated purity of 99.9999, 99.998, and 99.995%, respectively,
and were used without further purification.
All infrared spectra were recorded using either a Bruker 113v
or a Bruker 66v Fourier transform spectrometer, equipped with
a Globar source, a Ge/KBr beamsplitter, and a broadband MCT
detector. The corresponding interferograms, recorded at a
resolution of 0.5 cm-1, were averaged over 200 scans, Happ-
Genzel apodized, and Fourier transformed using a zero-filling
factor of 4. The liquefied noble gas set-up was described
before.10
The strength of the interaction also depends on the polariz-
ability of the solvent atoms. From xenon to argon, the atomic
polarizability decreases, from 4.02 to 1.63 Å3. Thus, in this
sequence, the solute/solvent interaction must be expected to
decrease. By the nature of its conformers, the butane data were
bound to be inconclusive on this aspect, and no other studies
appear to have systematically addressed this phenomenon. As
a first step toward filling this gap, in this study we report on
the enthalpy difference between the conformers of three 1,2-
disubstituted ethanes dissolved in liquefied xenon and krypton.
Attempts were made to include solutions in liquefied argon,
but solubility problems have prevented this. The compounds
X Abstract published in AdVance ACS Abstracts, May 1, 1996.
S0022-3654(95)03713-0 CCC: $12.00 © 1996 American Chemical Society

9672 J. Phys. Chem., Vol. 100, No. 23, 1996
Herrebout and van der Veken
Figure 1. Stable conformers for CH2X-CH2Y: (a) antiperiplanar
(trans); (b) synclinal (gauche).
In order to be able to distinguish the spectra of dissolved
species from those of suspended solid species, the infrared
spectra of the solids were obtained by condensing a small
amount of the compound onto a liquid nitrogen cooled CsI
window, followed by annealing until no further changes were
observed in the spectra. Low-temperature spectra of the gas
phase were recorded by using the liquid noble gas cell.
All ab initio calculations described below were carried out
using the Gaussian92 program,14 as implemented on an IBM
RS/6000 workstation.
Figure 2. The 850-600 cm-1 region of the vapor phase infrared
spectrum of 1-chloropropane recorded at 293.0 and 229.1 K. The
excessive noise near 620 cm-1 is due to an absorption band in the silicon
windows of the cell.
Theoretical Framework
Conformational Enthalpy Differences. The concentration
of a species A is related to the integrated intensity IA of one of
its infrared bands via the expression
dipole moments, the experimental values, where available, were
used. For the trans conformer of 1-chloro-2-fluoroethane no
experimental value for the dipole moment appears to have been
published. For this conformer the RHF/6-31G* ab initio value
of 0.31 Debye14 has been used.
IA ) nARAl
(1)
The radius a of the solute cavity to be used has been the
subject of considerable debate.18-23 In a first approximation,
also tested in this study, it was assumed that the different
conformers occupy cavities of identical sizes. Then the radius
can be calculated from the solute molar volume Vm and
Avogadro’s number NA:
in which nA is the concentration of A, RA the absorptivity of
the spectral band, and l the path length of the absorption cell.
Substituting this expression in the van’t Hoff isochore leads to
the following relation between the intensity ratio of bands due
to conformers A and B and the enthalpy difference ∆H° and
entropy difference ∆S° between the conformers:
4πa3
3
Vm
NA
∆H°
RT
∆S°
R
)
(5)
ln(IA/IB) ) -
+ ln(RA/RB) + ln(gA/gB) +
(2)
In this expression gA and gB are the statistical weights of the
conformers. In a typical study the temperature range covered
is relatively small, so it can be assumed that ∆H°, ∆S°, RA,
and RB are constant. Then, plotting ln(IA/IB) vs (1/T) in a van’t
Hoff plot results in a straight line, the slope of which is equal
to -∆H°/R.
Reaction Field Model. In the simple reaction field model
(see, for instance, ref 16), the solute molecule, with dipole
moment µ, occupies a spherical cavity of radius a, immersed
in a continuous dielectric of dielectric constant κ. The dipole
moment induced by the solute molecule in the dielectric
stabilizes the solute molecule. The following expression is
obtained for the solvation free enthalpy ∆Gsol:
In a second approximation, the radius is derived using ab initio
calculations, as was recently suggested.13 In this approximation,
the molecular volume is taken to be the space in which the ab
initio electron density of the molecule equals or exceeds 0.001
au.24
More sophisticated expressions for ∆Gsol than the one above
have been derived. These include effects due to polarizability
and quadrupole moments.11,12,16 Also, corrections to the
dielectric continuum, to account for specific solute/solvent
interactions, have been discussed.25 Such sophistications,
however, were judged beyond the scope of the present inves-
tigation, and were not pursued.
Results
κ - 1 µ2
∆Gsol ) -
(3)
1. 1-Chloropropane. 1.1. Vapor Phase. A literature
survey shows that rather varying values for the vapor phase
∆H° of 1-chloropropane have been obtained. This is illustrated
by the more recent values of 1.97 ( 1.26 and 1.51 ( 0.12 kJ
mol-1 obtained from vapor phase Raman and from a study of
the internal rotation potential, respectively,26 and the value of
0.19 ( 0.29 kJ mol-1 obtained from a combined microwave
and electron diffraction study.27 In view of this, it was decided
to reinvestigate this enthalpy difference. To this end, the vapor
phase mid-infrared spectrum of 1-chloropropane was recorded
at seven different temperatures in the interval between 293 and
229 K. In Figure 2 a part of the spectra recorded at 293 and at
229 K is given. The region shown is the only one in which
trans (747 cm-1) and gauche (792 and 665 cm-1) bands occur
well separated. The relative intensities in this figure confirm
(
)
3
2κ + 1
a
The shift in conformational equilibrium upon condensing a
compound from the vapor into solution is determined by the
difference ∆(∆Gsol) in solvation free enthalpy for its conformers
A and B:
2
2
µA
µB
κ - 1
2κ + 1
∆(∆Gsol) ) -
-
(4)
3
3
(
)(
)
aA
aB
The dielectric constants of the solutions are unknown, and,
therefore, in the calculations the κ values of the pure solvents
were used. This approximation is justified by the dilute nature,
in the order of 10-3 M, of the solutions studied here. For the

Solute-Solvent Interactions in Liquid Noble Gases
J. Phys. Chem., Vol. 100, No. 23, 1996 9673
Figure 3. Van’t Hoff plot, obtained from the 792/747 cm-1 confor-
mational doublet, for 1-chloropropane in the vapor phase.
previous conclusions that in the vapor phase the gauche
conformer is the more stable.
The integrated intensities of the 792 and 747 cm-1 bands
were obtained by numerical integration. It is clear from Figure
2 that the 665 cm-1 gauche band is also well separated from
the other bands in this region. The low-frequency side of this
band, however, is severely disturbed, which is a consequence
of the infrared absorption in this region of the silicon windows
of the cell. Therefore, the intensities obtained for the 665 cm-1
band were judged unreliable and they were not used in the
analysis. The van’t Hoff plot constructed from the 792 and
747 cm-1 bands is shown in Figure 3. The enthalpy difference
obtained from the least-squares linear fit equals -0.39 ( 0.10
kJ mol-1. This result, within the error limits, equals the value
obtained from the microwave and electron diffraction study.27
Figure 4. The 825-700 cm-1 region of the infrared spectrum, at
different temperatures, of a liquefied krypton solution containing
approximately 3.0 × 10-3 M CH3CH2Cl: (A) 158.9 K, (B) 141.2 K,
and (C) 123.4 K.
1.2. Liquefied Noble Gas Solutions. In the spectra in
liquefied argon, bands due to dissolved n-C3H7Cl molecules and
also weak bands due to suspended crystalline particles were
observed. For the solutions in krypton and xenon only bands
due to dissolved molecules were observed.
Infrared spectra of n-C3H7Cl were recorded at different
temperatures, varying from 88 to 108 K for the solutions in
argon, 120 to 165 K for the solutions in krypton, and 168 to
223 K for the solutions in xenon.
Figure 5. Van’t Hoff plots for 1-chloropropane dissolved in liquefied
krypton (a) and in liquefied xenon (b).
As described above, the spectrum of gaseous 1-chloropropane
shows two well-separated C-Cl stretching bands, at 747 and
665 cm-1. In the same spectral region of an argon solution
with a concentration of approximately 3.0 × 10-3 M, corre-
sponding bands are found at 742 and 664 cm-1, respectively.
bands were observed in the spectra of solid matrices,29 where
they were assigned to the trans conformer. We adopt this
assignment.
For the krypton and xenon solutions the integrated intensities
of the 790 cm-1 gauche band and of the 745 cm-1 absorption
complex due to the trans conformer were obtained by numerical
integration. The van’t Hoff plots constructed using these
intensities are shown in Figure 5. From this, the enthalpy
difference between trans- and gauche-1-chloropropane was
determined to be -0.30 ( 0.20 kJ mol-1 in krypton and -0.50
( 0.25 kJ mol-1 in xenon.
In addition, two more bands are present, at 748 and 791 cm-1
.
In agreement with Ogawa et al.,28 these bands are assigned to
the (C)CH2(C) rocking fundamental in gauche- and trans-n-
C3H7Cl, respectively.
Upon decreasing the temperature of the solution, the relative
intensity of the 747 cm-1 band slightly decreases. This shows
that the gauche conformer is the more stable species in the argon
solution. Unfortunately, below 108 K the bands due to dissolved
molecules become too weak to allow reliable measurement of
their intensity. Therefore, no accurate value for the enthalpy
difference in argon could be obtained.
In Figure 4, the νC-Cl region of the spectra of 1-chloropropane
dissolved in krypton, recorded at 127.5, 146.6, and 163.5 K,
are given. The relative intensity of the 745 cm-1 doublet,
assigned to the trans conformer, can be seen to slightly decrease
with temperature. This shows that also in krypton the gauche
conformer is the more stable species. A similar phenomenon
occurs in xenon, which leads to an identical conclusion for this
solvent.
2. 1,2-Dichloroethane. 2.1. Vapor Phase. The confor-
mational equilibrium in 1,2-dichloroethane has been reviewed
recently by Wiberg et al.15 These authors propose the value of
4.6 ( 0.4 kJ mol-1 for ∆H°, which is an average over results
obtained by several techniques. We have redetermined the
value, using infrared spectra recorded at several temperatures
between 290.6 and 249.5 K. In Figure 6, the CH2 wagging
region of the mid-infrared spectrum recorded at 290.6 and at
249.5 K is shown. In both spectra a strong band belonging to
the trans conformer30,31 can be observed at 1232 cm-1. On
the high-frequency side of this band, two gauche bands are
present, at 1286 and 1314 cm-1
.
In the same spectral region, both the krypton and xenon
solutions show two weaker bands, at 730 and 721 cm-1. Similar
Numeric integration yielded the intensity of the 1232 cm-1
band. The bands at 1286 and 1314 cm-1 strongly overlap, and

9674 J. Phys. Chem., Vol. 100, No. 23, 1996
Herrebout and van der Veken
Figure 6. The 1350-1200 cm-1 region of the vapor phase infrared
spectrum of 1,2-dichloroethane recorded at 291 and 250 K.
Figure 8. The 1500-1400 cm-1 region of the infrared spectrum, at
different temperatures, of a liquefied krypton solution containing
approximately 3.0 × 10-3 M CH2ClCH2Cl: (A) 161.5 K, (B) 147.0
K, and (C) 132.1 K.
Figure 7. Van’t Hoff plot, obtained from the 1232/(1286,1314)
conformational multiplet, for 1,2-dichloroethane in the vapor phase.
the intensity of the individual bands cannot be obtained using
simple numerical integration. Because both bands are due to
the same conformer, however, the integrated intensity of the
complete doublet is a correct measure of the population of the
gauche conformer. Therefore, in this study the numerically
integrated intensity of the 1314/1286 cm-1 doublet was used
as intensity for the gauche conformer. From the resulting van’t
Hoff plot, shown in Figure 7, the enthalpy difference between
trans- and gauche-CH2ClCH2Cl was calculated to be 4.90 (
0.15 kJ mol-1. This value agrees, within the error limits, with
the value proposed by Wiberg et al.15
2.2. Liquefied Noble Gas Solutions. For CH2ClCH2Cl,
substantial information has been published16 on the relative
stability of the conformers in the pure liquid and in polar and
nonpolar solvents. These results include the enthalpy difference,
3.87 ( 0.09 kJ mol-1, between the conformers in liquid xenon.9
In this study, attempts were made to expand the data with values
for the solutions in krypton and argon. Additionally, the mid-
infrared spectrum of CH2ClCH2Cl dissolved in liquefied xenon
was reinvestigated.
Figure 9. Van’t Hoff plots for CH2ClCH2Cl dissolved in cryosolu-
tions: (a) solution in liquefied krypton, using the 1456/1432 cm-1
conformational doublet; (b) solution in liquefied xenon, using the 1292/
1232 cm-1 doublet; (c) solution in liquefied xenon, using the 1315/
1232 cm-1 doublet.
tion on the argon solutions could be obtained. In the spectra
of krypton and xenon solutions, no bands due to crystalline
particles were observed.
For the temperature study, spectra were recorded at temper-
atures ranging from 132.0 to 161.6 K for the solutions in krypton
and from 170.5 to 214.9 K for the solutions in xenon. For the
krypton solutions, the presence of impurities in the solvent
prevents the use of the 1300-1200 cm-1 region, which was
used for the vapor phase. As shown in Figure 8, however, two
well-separated bands belonging to trans- and gauche-CH2ClCH2-
Cl can be observed at 1456 and 1432 cm-1, respectively. The
integrated intensities of these bands were obtained using
numerical integration. With these intensities, a van’t Hoff plot
was constructed. From this plot, shown in Figure 9a, the
enthalpy difference ∆H° between both conformers was calcu-
In the spectra of the argon solutions only bands due to
crystalline particles were observed, and no absorption bands
due to dissolved molecules were observed. Hence, no informa-
lated to be 3.81 ( 0.30 kJ mol-1
.

Solute-Solvent Interactions in Liquid Noble Gases
J. Phys. Chem., Vol. 100, No. 23, 1996 9675
Figure 11. Van’t Hoff plot, obtained from the 777/685 cm-1
conformational doublet, for 1-chloro-2-fluoroethane in the vapor phase.
Figure 10. The 800-650 cm-1 region of the vapor phase mid-infrared
spectra of 1-chloro-2-fluoroethane, recorded at 295 and 230 K.
In the spectra of xenon solutions, similar bands as those in
krypton were observed, at 1452 and 1430 cm-1. At the higher
temperatures of xenon solutions, the increased bandwidths cause
the bands to overlap too strongly to allow a simple numerical
integration. Attempts were made to obtain the intensities from
a least-squares band profile analysis using a set of two Gauss-
Lorentz sum profiles. It was found, however, that this leads to
a poor reproduction of the spectrum, owing to the unusually
high wings of the 1452 cm-1 band. Therefore, the 1452/1430
cm-1 conformer doublet was not used in the temperature study.
In the xenon solutions, the 1300-1200 cm-1 region is not
disturbed by impurities, and this region was used instead. The
experimental spectra in this region were analyzed using a set
of three Gauss-Lorentz sum profiles. Subsequently, for the
1229/1286 cm-1 and for the 1229/1310 cm-1 conformer
doublets a van’t Hoff plot was constructed. From these plots,
shown in Figure 9, b and c, the enthalpy difference between
both conformers was calculated to be 3.30 ( 0.10 and 3.21 (
0.15 kJ mol-1, respectively. As value for ∆H°, we propose to
Figure 12. The 825-650 cm-1 region of the infrared spectrum, at
different temperatures, of a liquefied xenon solution containing
approximately 3.0 × 10-3 M CH2ClCH2F. From top to bottom the
temperatures are 211.5, 190.8, and 170.7 K.
take the average of these two values, 3.26 ( 0.13 kJ mol-1
.
This value is somewhat smaller than the value of 3.87 ( 0.09
kJ mol-1 obtained by Turner et al.9
3. 1-Chloro-2-fluoroethane. 3.1. Vapor Phase. From the
vibrational spectra,17,33 the trans form has been determined to
be the more stable in the gas phase while in the pure liquid
phase the gauche form is the preferred conformer. A problem
with the literature values for ∆H° of 1-chloro-2-fluoroethane
has been pointed out by Wiberg et al.15 Therefore, the enthalpy
difference in the vapor phase was redetermined. Mid-infrared
spectra of the gaseous compound were recorded at several
temperatures, varying from 295 to 230 K. In Figure 10, the
νC-Cl region recorded at 295 and 230 K is shown. This is the
only region of the mid-infrared spectrum where trans (777 cm-1
)
Figure 13. Van’t Hoff plots for CH2ClCH2F, obtained from the 774/
682 cm-1 conformational doublet, dissolved in cryogenic solutions: (a)
solution in liquefied krypton; (b) solution in liquefied xenon.
and gauche (685 cm-1) bands appear well separated. The
intensities of the bands were determined by numerical integra-
tion, and a van’t Hoff plot was constructed. From this plot,
shown in Figure 11, the enthalpy difference between the more
stable trans and the gauche conformer was calculated to be 1.51
( 0.09 kJ mol-1. This value is somewhat smaller than the value
of 2.0 kJ mol-1, at 295 K, derived by Wiberg et al. using ab
initio calculations at the G2/MP2 level.15
the solutions in xenon. In the spectra of the xenon solution,
shown in Figure 12, the isotopic νC-Cl doublets are observed
around 773 and 682 cm-1. It can be seen that the relative
intensities of these doublets hardly vary with temperature,
showing that the enthalpy difference is very small. The
intensities of the 770 and 680 cm-1 doublets were determined
by numerical integration. From the van’t Hoff plots, shown in
Figure 13, the enthalpy difference in liquefied krypton and
liquefied xenon was calculated to be 0.20 ( 0.10 and -0.24 (
0.12 kJ mol-1, respectively. These enthalpy differences are very
small indeed. Also, it is noteworthy that in krypton the stability
order of the conformers is the same as in the vapor phase, while
in xenon solutions the order is reversed.
3.2. Liquefied Noble Gas Solutions. In the spectra of the
argon solutions, only bands due to crystalline particles were
observed. Hence, no information about the relative stability of
the conformers in argon solutions could be obtained. For the
krypton and xenon solutions, no bands due to crystalline
particles were observed in the spectra.
Spectra were recorded at temperatures ranging from 127 to
164 K for the solutions in krypton and from 170 to 212 K for

9676 J. Phys. Chem., Vol. 100, No. 23, 1996
Herrebout and van der Veken
TABLE 1: Enthalpy Differences (kJ mol-1)a for
1-Chloropropane, 1,2-Dichloroethane, and
1-Chloro-2-fluoroethane in Cryogenic Solutions
solution
T/K gas phase
xenon
krypton
argon ∆(∆H°sol)
CH3CH2CH2Cl
261 -0.39 ( 0.10
192 -0.32 ( 0.10 -0.50 ( 0.25
142 -0.25 ( 0.10
-0.18 ( 0.27
-0.05 ( 0.22
-0.30 ( 0.20
3.81 ( 0.30
0.20 ( 0.10
CH2ClCH2Cl
270 4.90 ( 0.15
192 5.02 ( 0.15 3.27 ( 0.16
147 5.11 ( 0.15
-1.75 ( 0.22
-1.30 ( 0.36
CH2ClCH2F
Figure 14. Trans-gauche enthalpy difference for 1,2-dichloroethane
in the gas phase, in liquefied krypton and in liquefied xenon as a
function of the Kirkwood function.
261 1.51 ( 0.09
192 1.59 ( 0.09 -0.24 ( 0.12
142 1.66 ( 0.09
-1.83 ( 0.15
-1.46 ( 0.13
a ∆H° ) Hgauche - Htrans; ∆(∆H°sol) ) (∆H°)liquid - (∆H°)gas
.
TABLE 2: Dipole Moments and Solute Radii for
1-Chloropropane, 1,2-Dichloroethane, and
1-Chloro-2-fluoroethane
solute radius/Å
mol vol
RHF/6-31G*
electron density
dipole moment/D
trans
gauche
trans
gauche
CH3CH2CH2Cl
CH2ClCH2Cl
CH2ClCH2F
2.02a
0.00
1.96a
2.55b
2.71d
3.28
3.17
3.03
3.77
3.76
3.55
3.78
3.76
3.56
0.31c
a Taken from ref 27. b Taken from ref 32. c Taken from ref 17.
d Taken from ref 38.
Figure 15. Trans-gauche enthalpy difference for 1-chloro-2-fluoro-
ethane in the gas phase, in liquefied krypton and in liquefied xenon as
a function of the Kirkwood function.
Discussion
TABLE 3: Least-Squares Values of p and q
p/kJ mol-1
q/kJ mol-1
The vapor phase enthalpy differences are determined near
275 K, while the liquid phase values are obtained at substantially
lower temperatures. In spite of the assumptions made above,
the constancy of ∆H over such a wide temperature range is not
guaranteed. Therefore, corrected gas phase values were cal-
culated, using straightforward statistical thermodynamics,35 at
the temperatures at which the solutions were studied. These
temperatures were taken to be the midpoint of the experimental
interval used to obtain the ∆H°. The rotational constants and
vibrational frequencies required for the calculations were derived
from ab initio calculations at the RHF/6-31G* level. The ab
initio frequencies were rescaled using a factor of 0.89.36,37 The
vapor phase values of ∆H° thus obtained are compared with
the liquid phase values in Table 1. In Table 2 the dipole
moments for both conformers of each of the compounds, and
the cavity radii obtained from molar volumes and from ab initio
calculations, are collected.
ClCH2CH2Cl
(this study)
ClCH2CH2Cl
(other solutions)
ClCH2CH2F
-8.20 ( 1.22
4.92 ( 0.17
-8.59 ( 1.07
-9.15 ( 0.79
5.20 ( 0.77
1.52 ( 0.11
For 1,2-dichloroethane and 1-chloro-2-fluoroethane the dipole
moment of the gauche conformer is much larger than that of
the trans conformer. Hence, a pronounced solvent influence
must be anticipated. This is neatly confirmed by the data in
Table 1. For both compounds, in the vapor phase the more
stable conformer is the less polar one, so that in both cases a
lowering of the enthalpy difference in solution is predicted. Also
this is borne out by the data in Table 1.
Table 1 shows that for 1,2-dichloroethane and for 1-chloro-
2-fluoroethane, the temperature effects on the vapor phase ∆H°
are relatively small, much smaller than the solvent influences
on ∆H°. In a first approximation, therefore, these temperature
effects can be neglected, allowing the following analysis.
Relation 4 suggests that ∆(∆Hsol) should vary linearly with (κ
- 1)/(2κ + 1) of the surrounding solution. In Figures 14 and
15, ∆(∆Hsol) is plotted against this parameter for 1,2-dichloro-
ethane and 1-chloro-2-fluoroethane, respectively. In each case
a linear relationship can be seen to hold. A linear regression
analysis of the type
Relation 4 above expresses the solvent influence on the free
enthalpy. The considerations that lead to this relation take into
account the solvent influences on the enthalpy but not on the
entropy. Therefore, ∆Gsol may be replaced by ∆Hsol, so that a
comparison with the values for ∆H° obtained in this study can
be made.
For conformers of similar size, the relation shows that the
solvent influence on ∆H° is determined by the difference of
the squares of the dipole moments of the conformers. The
conformers of 1-propyl chloride are quite polar, but, as can be
seen from Table 2, the dipole moments are nearly equal.
Consequently, no change in the conformational equilibrium upon
disolution should occur. In line with these predictions, the
enthalpy differences determined in liquid krypton and xenon
are, within the experimental error, equal to the vapor phase
value.
∆(∆Hsol) ) p(κ - 1)/(2κ + 1) + q
(6)
leads to the values of p and q given in Table 3. The enthalpy
difference between the conformers of 1,2-dichloroethane has
been studied in a wide variety of solvents.16 Using the published
∆H° values16 together with the vapor phase ∆H° obtained in
this study, and using the κ values of the pure solvents, an

Solute-Solvent Interactions in Liquid Noble Gases
J. Phys. Chem., Vol. 100, No. 23, 1996 9677
2
TABLE 4: Calculated and Experimental Values of (µT
µG2)/a3 for CH2ClCH2Cl and CH2ClCH2F
-
ments as Research Assistant (aspirant) and Postdoctoral Fellow.
The N.F.W.O. is also thanked for financial help toward the
spectroscopic equipment used in this study.
CH2ClCH2Cl CH2ClCH2F
using molar volume Vm
(µT2 - µG2)/a3/D2 Å-3
References and Notes
0.2047
12.06
0.2603
15.33
(µT2 - µG2)/a3/kJ mol-1
using RHF/6-31G* electron density
(µT2 - µG2)/a3/D2 Å-3
(1) Bulanin, M. O. J. Mol. Struct. 1973, 19, 59.
(2) Freund, S. M.; Maier, N. B.; Holland, F. R.; Beattie, W. H. J. Chem.
Phys. 1978, 69, 1961.
0.1263
7.44
0.1620
9.55
(µT2 - µG2)/a3/kJ mol-1
(3) MacLaughlin, J. G.; Poliakoff, M.; Turner, J. J. J. Mol. Struct. 1982,
82, 51.
(4) Tacke, M.; Sparrer, P.; Teuber, R.; Stadter, H. J.; Schuster, F. J.
Mol. Struct. 1995, 349, 251.
(5) Sennikov, P. G. J. Phys. Chem. 1994, 98, 4973.
(6) Brock, A.; Mina-Camilde, N.; Manzanares, I. C. J. Phys. Chem.
1994, 98, 4800.
analogous linear regression was performed. The values of p
and q determined are also given in Table 3. It is clear that the
latter, within their experimental uncertainty, are identical to the
corresponding values for the noble gas solutions. This shows
that the noble gases influence the conformational equilibrium
in the same way as more common solvents.
(7) Van der Veken, B. J.; De Munck, F. R. J. Chem. Phys. 1992, 97,
3060.
(8) Bulanin, M. O. J. Mol. Struct. 1995, 347, 73.
(9) Turner, J. J.; Poliakoff, M.; Howdle, M.; Jackson, S. A.; McLaugh-
lin, J. G. Faraday Discuss. Chem. Soc. 1988, 83, 271.
(10) Herrebout, W. A.; Van der Veken, B. J.; Wang, A.; Durig, J. R. J.
Phys. Chem. 1995, 99, 578.
(11) Onsager, L. J. Am. Chem. Soc. 1936, 58, 1486.
(12) Kirkwood, J. G. J. Chem. Phys. 1934, 2, 357.
(13) Wong, M. W.; Frisch, M. J.; Wiberg, K. B. J. Am. Chem. Soc.
1991, 113, 4776.
(14) Gaussian 92, Revision E.3; Frisch, M. J.; Trucks, G. W.; Head-
Gordon, M.; Gill, P. M. W.; Wong, M. W.; Foresman, J. B.; Johnson, B.
G.; Schlegel, H. B.; Robb, M. A.; Replogle, E. S.; Gomperts, R.; Andres,
J. L.; Raghavachari, K.; Binkley, J. S.; Gonzalez, C.; Martin, R. L.; Fox,
D. J.; Defrees, D. J.; Baker, J.; Stewart, J. J. P.; Pople, J. A.; Gaussian,
Inc.: Pittsburgh PA, 1992.
According to relation 4, the direction cos p for each
2
3
2
conformational equilibrium must be equal to µT /aT - µG
/
aG3. This quantity can be calculated using the data in Table 2.
Because the ab initio values of the cavity radii consistently are
larger than the values obtained from the molar volume, the above
quantity was calculated using both sets of radii. The results
are given in Table 4. From a comparison with the data in Table
3 it is clear that for both molecules the experimental value of
p agrees well with the value calculated from the ab initio cavity
radii. This agreement, of course, is corroborating evidence for
the usefulness of the definition of the radii.24
Finally, using the experimental values for p and q, the
enthalpy difference ∆H° between the conformers dissolved in
liquefied argon has been predicted, by interpolating the linear
relation 6 at the values of (κ - 1)/(2κ + 1) for argon. In this
way, the ∆H° for 1,2-dichloroethane is predicted to be 3.80 (
0.23 kJ mol-1, while for 1-chloro-2-fluoroethane it is calculated
to be 0.38 ( 0.15 kJ mol-1. These enthalpy differences are
significantly below their gas phase values, which are 4.90 (
0.15 and 1.51 ( 0.09 kJ mol-1, respectively. As the enthalpy
differences in liquefied krypton and xenon differ even more
from the vapor phase values, it is clear that, at least in terms of
conformational equilibria, the above results refute the notion
that the liquefied noble gases discussed in this paper create a
pseudo gas phase1-6 in which solvent effects are negligible.
(15) Wiberg, K. B.; Keith, T. O.; Frisch, M. J.; Murcko, M. J. Phys.
Chem. 1995, 99, 9072.
(16) Abraham, R. J.; Bretschneider, E. Medium Effects on Rotational
and Conformational Equilibria. In Internal Rotation in Molecules; Orville-
Thomas, W. J., Ed.; Wiley: London, 1974.
(17) Durig, J. R.; Liu, J.; Little, T. S. J. Phys. Chem. 1991, 95, 4664.
(18) Meyer, A. Y. J. Chem. Soc., Perkin Trans. 2 1981, 1161, and
references therein.
(19) Meyer, A. Y. Chem. Soc. ReV. 1986, 15, 449, and references
therein.
(20) Meyer, A. Y. J. Comput. Chem. 1986, 7, 144, and references
therein.
(21) Pascual-Ahuir, J. L.; Silla, E.; Tomasi, J.; Bonaccorsi, R. J. Comput.
Chem. 1987, 8, 778, and references therein.
(22) Silla, E.; Tunon, I.; Pascual-Ahuir, J. L. J. Comput. Chem. 1991,
12, 1077, and references therein.
(23) Pascual-Ahuir, J. L.; Silla, E. J. Comput. Chem. 1990, 11, 1047,
and references therein.
(24) Wong, M. W.; Wiberg, K. B.; Frisch, M. J. J. Comput. Chem. 1995,
16, 385.
Conclusion
(25) Enciso, E.; Alonso, J.; Almarza, N. G.; Berjemo, F. J. J. Chem.
Phys. 1989, 90, 413, and references therein.
(26) Durig, J. R.; Godbey, S. E.; Sullivan, J. F. J. Chem. Phys. 1984,
80, 5983.
(27) Yamanouchi, K.; Sugie, M.; Takeo, K.; Matsumura, C.; Kuchitsu,
To obtain more detailed information about the weak interac-
tions between solute molecules and surrounding noble gas
(argon, krypton, and xenon) atoms, in this study the influence
of noble gas solutions on the conformational equilibrium of
some simple molecules with the general formula CH2XCH2Y
was investigated.
K. J. Phys. Chem. 1984, 88, 2315.
(28) Ogawa, Y.; Imazeki, S.; Yamaguchin, H.; Matsuura, H.; Harada,
I.; Shimanouchi, T. Bull. Chem. Soc. Jpn. 1978, 51, 748.
(29) Rasanen, M. A.; Bondybey, V. E. J. Phys. Chem. 1986, 90, 5038.
(30) Mizushima, S.; Shimanouchi, T.; Harada, I.; Abe, Y.; Takeuchi,
H. Can. J. Phys. 1975, 53, 2085, and references therein.
(31) Sverdloo, L. M.; Kovner, M. A.; Krainov, E. P. Vibrational Spectra
of Polyatomic Molecules; John Wiley & Sons: New York, 1974; p 398,
and references therein.
The experimental results obtained in this study were compared
with the theoretical predictions based on the reaction field model
of Onsager11 and Kirkwood.12 The radii of the solute molecules
to be used in this model were calculated from the molar volume
and from the RHF/6-31G* electron density. It was found that
the radii calculated from the molar volume lead to substantial
overestimation of the solvent influences. When the electron
density values for the radii are used, however, the solvent
influence calculated from the reaction field agrees very well
with the experimental results.
(32) Wiberg, K. B.; Murcko, M. A. J. Phys. Chem. 1987, 91, 3616, and
references therein.
(33) El Bermani, M. F.; Jonathan, N. J. Chem. Phys. 1968, 49, 340.
(34) Huang, J.; Hedberg, K. J. Am. Chem. Soc. 1990, 112, 2070.
(35) Knox, J. H. Molecular Thermodynamics. An Introduction to
Statistical Thermodynamics for Chemists; Wiley-Interscience: London,
1971.
(36) Foresman, J. B.; Frisch, Æ. Exploring Chemistry with Electronic
Structure Methods; Gaussian Inc.: Pittsburgh, 1993.
(37) Pople, J. A.; Krishnan, R.; Schlegel, H. B.; Defrees, D. J.; Binkley,
J. S.; Frisch, M. J.; Whiteside, R. F.; Hout, R. F.; Hehre, W. J. Int. J.
Quantum Chem., Symp. 1981, 15, 269.
In contrast with what has been claimed in the literature,1-6 it
was found that the noble gas solutions do not behave as a
“pseudo gas phase”. In fact, these solvents significantly shift
the equilibrium if the conformers have a different polarity.
(38) Mukhtarov, I. A. Opt. Spektrosk. 1966, 20, 58.
Acknowledgment. W.A.H. thanks the National Fund for
Scientific Research (NFWO, Belgium) for successive appoint-
JP953713Q