Temperature Dependence of the Rate Constant for the Reaction F( 2P) + Cl2 → FCl + Cl at T = 180-360 K

Article abstract of DOI:10.1021/jp030469r

The absolute rate constant for the reaction F(²P) with Cl ₂ has been measured using the discharge flow kinetics technique coupled to mass spectrometric detection at T = 180-360 K and 1 Torr He nominal pressure. Experiments were performed at NASA Goddard Space Flight Center (GSFC) in Greenbelt, MD, and Laboratoire de Combustion et Systemes Reactifs-CNRS in Orleans, France. Results of k = (5.7 ± 0.8) × 10^-11 and (6.2 ± 0.8) × 10^-11 cm³ molecule^-1 s^-1 independent of temperature were obtained by each laboratory, respectively. When the results from both laboratories were combined into one data set, an average temperature independent value of k₁ = (6.0 ± 1.1) × 10^-11 cm³ molecule^-1 s^-1 was obtained. A very slight positive temperature dependence with k₁(T) = (6.5 ± 1.5) × 10^-11 exp{-(20 ± 60)/T} cm³ molecule^-1 s^-1 may also be derived from the combined data in the range T = 180-360 K.

Full text of DOI:10.1021/jp030469r

1726
J. Phys. Chem. A 2004, 108, 1726-1730
Temperature Dependence of the Rate Constant for the Reaction F(2P) + Cl2 f FCl + Cl at
T ) 180-360 K
Fred L. Nesbitt†
Catholic UniVersity of America, 620 Michigan AVenue NE, Washington, D.C. 20064
Regina J. Cody* and Douglas A. Dalton
Laboratory for Extraterrestrial Physics, NASA Goddard Space Flight Center, Greenbelt, Maryland 20771
Veronique Riffault, Yuri Bedjanian, and Georges Le Bras
Laboratoire de Combustion et Systemes Reactifs-CNRS, 1C AVenue de la Recherche Scientifique,
45071 Orleans Cedex 2, France
ReceiVed: April 21, 2003; In Final Form: NoVember 17, 2003
The absolute rate constant for the reaction F(2P) with Cl2 has been measured using the discharge flow kinetics
technique coupled to mass spectrometric detection at T ) 180-360 K and 1 Torr He nominal pressure.
Experiments were performed at NASA Goddard Space Flight Center (GSFC) in Greenbelt, MD, and Laboratoire
de Combustion et Systemes Reactifs-CNRS in Orleans, France. Results of k ) (5.7 ( 0.8) × 10-11 and (6.2
( 0.8) × 10-11 cm3 molecule-1 s-1 independent of temperature were obtained by each laboratory, respectively.
When the results from both laboratories were combined into one data set, an average temperature independent
value of k1 ) (6.0 ( 1.1) × 10-11 cm3 molecule-1 s-1 was obtained. A very slight positive temperature
dependence with k1(T) ) (6.5 ( 1.5) × 10-11 exp{-(20 ( 60)/T} cm3 molecule-1 s-1 may also be derived
from the combined data in the range T ) 180-360 K.
Introduction
work at GSFC covered a temperature range of T ) 180-298
K. The work at CNRS covered a temperature range of T )
230-360 K. Both studies used the technique of discharge flow
mass spectrometry (DF-MS).
The abstraction of chlorine atoms by fluorine atoms from
Cl2 has been used for years as a convenient and reliable method
to determine the concentration of F atoms.1
Experimental Section
F + Cl2 f FCl + Cl
∆Hrxn ) -2 kcal mol-1 (1)
Work at NASA/GSFC. Discharge Flow Reactor. All experi-
ments were performed in a Pyrex flow tube4 60 cm long and
2.8 cm in diameter with the inner surface of the flow tube being
lined with Teflon FEP. A diagram of the apparatus is shown in
Figure 1. The flow tube has an outer jacket for the circulation
of a controlled flow of nitrogen vapor from liquid N2 or
thermostated fluids. The flow tube was coupled via a two-stage
stainless steel collision-free sampling system to a quadrupole
mass spectrometer (ABB Extrel Corp.) that was operated at low
electron energies (typically less than 20 eV). Ions were detected
by an off-axis conversion dynode/channeltron multiplier (Detec-
tor Technology Corp.). The total pressure was 1 Torr He. The
Few investigations of this reaction have been reported in the
literature. In a study by Warnatz et al.2 that covered a
temperature range from T ) 232 to 350 K, a rate constant of
9.5 × 10-11 cm3 molecule-1 s-1 at T ) 300 K was reported.
These results provide an activation energy of 1400 cal mol-1
.
A study by Clyne et al.1 at T ) 298 K yielded a rate constant
of 1.1 × 10-10 cm3 molecule-1 s-1. A subsequent study from
the same laboratory by Appleman et al.3 at T ) 298 K gave a
rate constant of 1.6 × 10-10 cm3 molecule-1 s-1. Thus at T )
298 K a variation of more than 50% is observed in the
measurements of the rate constant for reaction 1.
linear flow velocity ranged from 2700 to 3000 cm s-1
.
In some instances, for example to correct for undertitration
at low [F],4 the absolute value of k1 is required as a function of
temperature. In light of the limited and conflicting data available,
reaction 1 was examined as a function of temperature in two
different laboratories. One study was performed at NASA
Goddard Space Flight Center (GSFC) in Greenbelt, MD, and
the other study was performed at Laboratoire de Combustion
et Systemes Reactifs-CNRS in Orleans, France (CNRS). The
The Cl2 reactant was introduced into the flow tube via a Pyrex
movable injector, which could be changed from a distance of 2
to 44 cm from the sampling pinhole. The helium carrier gas
was flowed at 600 sccm at T ) 298 K and 1100 sccm at T )
180 K, into the reaction flow tube through ports at the rear of
the flow tube. All gas flows were measured and controlled by
mass flow controllers (MKS Instruments). The mass flow
controllers were calibrated using the pressure drop method. In
this method, the gas flow originates from a bulb of known
volume, and the pressure of the gas in the bulb is recorded as
a function of time. Using the pressure change with time (∆P/
∆t), the known volume (V0), and the temperature (T), the gas
* Corresponding author: E-mail: regina.cody@nasa.gov. Tel: (301)286-
3782. Fax: (301)286-1683.
† Also at Department of Natural Sciences, Coppin State College,
Baltimore, MD 21216.
10.1021/jp030469r CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/13/2004

Rate Constant for the Reaction F(2P) + Cl2 f FCl + Cl
J. Phys. Chem. A, Vol. 108, No. 10, 2004 1727
Figure 1. Diagram of the apparatus at NASA/GSFC.
flow (F) can be calculated from F ) (V0/RT)(∆P/∆t). This
calculated flow is compared with the flow reading of the flow
controller. This calibration is done for several gas flows covering
the range of the flow controller to yield a calibration factor.
Atomic Fluorine Production and Titration. Fluorine atoms
were generated by passing molecular fluorine (ca. 5% diluted
in helium) through a sidearm at the upstream end of the flow
tube that contained a microwave discharge (∼50 W, 2450 MHz
3
Opthos Instruments). The discharge section consisted of a /8
in. ceramic discharge tube coupled to a glass arm.
Figure 2. Diagram of the apparatus at CNRS.
The concentration of fluorine atoms in the kinetic studies was
determined by measuring the Cl2 consumption in the fast
titration reaction 1. With Cl2 in excess, the F atom concentration
in a discharge flow reactor using a modulated molecular beam
mass spectrometer5 as the detection method. The reactor
consisted of a Pyrex tube (45 cm length and 2.4 cm i.d.) with
a jacket for the circulation of the thermostated liquid (ethanol
or water). The configuration of the movable double injector used
for the introduction of the reactants into the reactor is shown in
Figure 2. To reduce the wall loss of F atoms, the inner surfaces
of the reactor and the injector were coated with halocarbon wax.
F atoms were generated in a microwave discharge of F2 diluted
in He, which was used as the carrier gas in all the experiments.
It was verified by mass spectrometry that more than 90% of F2
was dissociated in the microwave discharge. The fluorine atoms
were detected at their parent peak (F+, m/e ) 19) and also as
FBr at m/e ) 100, after scavenging by Br2 at the end of the
reactor through the fast reaction:6
+
was determined by measuring the decrease in the Cl2 signal
(m/z ) 70) at an electron energy of ∼14 eV when the discharge
was initiated. The diluted Cl2/He mixture was admitted via the
movable injector. The position of the injector was chosen to
ensure that reaction 1 went to completion during the titration,
and that this position was close to the middle of the decay range
for Cl2 under reaction conditions for the rate constant measure-
ment. The absolute F atom concentration is given by
[F] ) [Cl2]DiscOff - [Cl2]DiscOn ≡ (∆Cl2 signal)[Cl2]DiscOff
(2)
where (∆Cl2 signal) is the fractional decrease in the Cl2+ mass
spectrometric signal (S) when the microwave discharge is turned
on and is calculated from
F + Br2 f FBr + Br
k3 ) 2.2 × 10-10 cm3
molecule-1 s-1 (3)
(∆Cl2 signal) )
[S(discharge off) - S(discharge on)]/S(discharge off)
Br2 was added 5 cm upstream of the sampling cone.
Typically, 65-75% of the F2 was dissociated and initial F atom
concentrations were (1.3-10.7) × 1013 molecule cm-3 for the
kinetic studies. [Cl2] was 3-4 times greater than [F] to yield
fractional reductions in [Cl2] of 0.2-0.3. The reaction time for
the titration was typically 1-4 ms, which was less than the
typical flow time from the probe position to the sampling hole
to the mass spectrometer.
Materials. Helium (99.9995%, Air Products) and F2 (4.92%
in helium, Air Products) were used as obtained. Cl2 (VLSI grade,
Air Products) was degassed at liquid nitrogen temperature.
Work at Laboratoire de Combustion et Systemes Reactifs-
CNRS. Discharge Flow Reactor. Experiments were carried out
This last method of F atom detection was preferred to the
direct detection at m/e ) 19 (F+), when working with low
concentrations of F atoms (kinetics of F consumption in excess
of Cl2, see below), because detection sensitivity of the mass
spectrometer for FBr was better than for F atoms. In addition,
in this case one does not need to make corrections on the
possible contribution at m/e ) 19 of FCl, product of reaction
1. All the other relevant species were detected at their parent
peaks: m/e ) 38 (F2+), 70 (Cl2+), 160 (Br2+).
Absolute concentration of F atoms was measured from the
titration reaction 3 using an excess of Br2. In this case, [F] )
∆[Br2] ) [FBr]. Another method of the absolute calibration of

1728 J. Phys. Chem. A, Vol. 108, No. 10, 2004
Nesbitt et al.
TABLE 1: Experimental Conditions and Results for the
Study of the F + Cl2 Reaction: Kinetics of Cl2 Consumption
in Excess of F Atoms
a
b
no. of decays
T (K)
[F]mean
k1
lab.
9
8
8
8
8
7
9
4
8
8
360
320
298
294
273
250
230
220
200
180
1.8-13.4
1.6-9.9
6.1 ( 0.8
CNRS
CNRS
GSFC
CNRS
CNRS
CNRS
CNRS
GSFC
GSFC
GSFC
6.0 ( 0.9
1.21-10.3
0.9-9.6
5.68 ( 0.54
6.0 ( 0.7
1.3-9.1
6.2 ( 0.8
0.8-6.7
7.0 ( 0.8
0.4-7.5
6.5 ( 0.8
2.76-8.51
2.32-9.15
2.54-8.63
6.53 ( 1.18
5.81 ( 1.16
4.87 ( 0.47
Figure 3. Examples of the kinetic runs of Cl2 consumption in reaction
with excess F atoms: T ) 298 K and T ) 180 K.
a Units of 1012 molecules cm-3 b Units of 10-11 cm3 molecule-1 s-1
. ;
errors are 1σ plus 10%.
the F signal consisted of titration of F atoms by an excess of
Cl2; [F] ) ∆[Cl2]. Results obtained by these two approaches
were always consistent within a few percent. The concentrations
of the stable species in the reactor were calculated from their
flow rates obtained from the measurements of pressure drop of
mixtures of the species with helium in calibrated volume flasks.
Materials. The purities of the gases were as follows: He >
99.9995% (Alphagaz) was passed through liquid nitrogen traps;
Cl2 > 99% (Ucar); Br2 > 99.99% (Aldrich); F2 (5% in helium,
Alphagaz).
in He was estimated from that of Ar in He according to the
method of Lewis et al.7 to be D ) 197 cm2 s-1 at T ) 180 K
and D ) 367 cm2 s-1 at T ) 298 K. A T3/2 dependence of D on
T was assumed in estimating D at other temperatures. A small
(8%) stoichiometric correction to [F]0 was made to allow for
the depletion of F during the reaction.
[F]mean ) [F]0 - 1/2[Cl2]0
(6)
Results
The bimolecular rate constant k1 is related to the corrected
pseudo-first-order rate constant kcorr through the expression
Work at NASA/GSFC. The rate constant measurements for
reaction 1 were performed under pseudo-first-order conditions
with [F]0 > [Cl2]0 and [F]0/[Cl2]0 ranging from 6 to 41 at a
total pressure of 1 Torr He. The initial concentrations of Cl2
were (2.0-9.9) × 1011 molecules cm-3. The first-order decay
of Cl2 is given by the expression
kcorr ) k1[F]mean + kloss
(7)
where [F]mean is calculated from eq 6 and kloss is the first-order
rate constant for the loss of Cl2 on the walls or other first-order
processes. Figure 4 shows a plot of kcorr versus [F]mean at T )
298 K. The slope of this plot gives k1(298K) ) (5.68 ( 0.54)
× 10-11 cm3 molecule-1 s-1. Similar plots give k1(220K) )
ln[Cl2]t ) -kobs(d/V) + ln[Cl2]0
(4)
where kobs is the measured pseudo-first-order decay constant, d
is the distance from the tip of the movable injector to the
sampling pinhole, and V is the linear velocity. Linear least-
squares analysis of plots of ln(Cl2 signal) at m/z ) 70 versus
contact time yielded the observed pseudo-first-order rate
constant, kobs. Figure 3 shows two examples of such plots. The
experimental first-order decay constants were corrected (2.5%)
for axial diffusion of the Cl2 in helium gas with the relationship
(6.53 ( 1.18) × 10-11, k1(200K) ) (5.81 ( 1.16) × 10-11
,
and k1(180K) ) (4.87 ( 0.47) × 10-11, all in units of cm3
molecule-1 s-1. The errors are (1σ + 10% for systemic error.
The results are summarized in Table 1.
Work at Laboratoire de Combustion et Systemes Reactifs-
CNRS. All the experiments were carried out at 1 Torr total
pressure of helium. Two series of experiments were performed
to measure the rate constant of the reaction F + Cl2: one by
monitoring Cl2 consumption kinetics in excess of F atoms and
another by monitoring of F decays in excess of Cl2 molecules.
Cl2 Kinetics in an Excess of F Atoms. In this series of
experiments reaction 1 was studied under pseudo-first-order
kcorr ) kobs(1 + kobsD/V2)
(5)
where V is the linear flow velocity and D is the diffusion
coefficient of Cl2 in helium. The diffusion coefficient for Cl2
Figure 4. Plot of pseudo-first-order rate constants versus the mean fluorine atom concentration at T ) 298 K.

Rate Constant for the Reaction F(2P) + Cl2 f FCl + Cl
J. Phys. Chem. A, Vol. 108, No. 10, 2004 1729
TABLE 2: Experimental Conditions and Results for the
Study of the F + Cl2 Reaction, Kinetics of F Consumption in
Excess of Cl2 (CNRS)
a
b
no. of decays
T (K)
[Cl2]mean
k1
8
7
8
360
335
240
0.5-8.1
0.4-7.2
0.5-6.5
5.8 ( 0.7
5.8 ( 0.7
6.1 ( 0.8
a Units of 1012 molecule cm-3 b Units of 10-11 cm3molecule-1s-1
. ;
errors are 1σ plus 10%.
Figure 5. Example of the kinetic runs of Cl2 consumption in reaction
with excess F atoms: T ) 294 K
Figure 7. Example of pseudo-first-order plots of F atoms consumption
in reaction with excess Cl2.
atom decay kinetics were monitored in the presence of excess
of Cl2. Because both F atoms and Cl2 were introduced through
the double movable injector (see Figure 2), the F atom signal
was monitored as a function of position and, therefore, time.
Experiments were carried out at three temperatures, T ) 360,
335, and 240 K. The initial concentration of F atoms (detected
as FBr, see Experimental Section) was (3.5-5.0) × 1011
molecule cm-3, and the initial concentration of Cl2 was varied
in the range (0.4-8.1) × 1012 molecule cm-3 (see Table 2).
The flow velocity in the reactor was 1540-2160 cm s-1. The
concentrations of Cl2 and F atoms were simultaneously meas-
ured as a function of reaction time. A consumption of excess
reactant, Cl2, was also observed. In most cases, this consumption
was less then 10% and it reached 30% in a few kinetic runs,
where insufficient excess of Cl2 over the F atom concentration
was used. For the rate constant calculations the mean values of
[Cl2] over the reaction time of the F kinetics were used. The
examples of the pseudo-first-order plots measured from F atom
decay kinetics are shown in Figure 7. The pseudo-first-order
rate constant values were corrected to take into account the axial
and radial diffusion of F atoms. The diffusion coefficient of F
in He, was calculated from DO-He to be DF-He ) 2.77 ×
Figure 6. Example of pseudo-first-order plots of Cl2 consumption in
reaction with excess F atoms.
conditions using an excess of F atoms over Cl2 molecules.
Experiments were carried out in the temperature range T )
230-360 K. The initial concentrations of Cl2 were (1.2-4.0)
× 1011 molecule cm-3; the range of F concentrations is shown
in Table 1. Flow velocities in the reactor were (1500-2300)
cm s-1. The wall loss rate of Cl2 was negligible (<1 s-1). The
consumption of excess reactant, F atoms, was negligible (<10%)
in all the experiments, and was taken into account using the
mean F concentration over the whole reaction time for the
calculation of k1. An example of the kinetic runs of the
exponential decay of Cl2 measured with different concentrations
of F atoms is shown in Figure 5. Figure 6 shows an example of
the dependence of the pseudo-first-order rate constant, k′1 )
k1[F], on the concentration of F atoms. All the pseudo-first-
order rate constants, k′1, were corrected for axial and radial
diffusion of Cl2.8 The diffusion coefficient of Cl2 in He was
calculated from that of Kr in He.9 The maximum correction
was about 10%. The values of k1 were obtained from the slope
of the linear least-squares fit to the experimental data, as shown
in Figure 6. All the results obtained for k1 at the different
temperatures of the study are reported in Table 1.
10-2T1.8 10
. Maximum corrections were around 10%. The values
of k1 obtained from the slopes of the straight lines in Figure 7
are presented in Table 2. The intercepts in Figure 7, 15 ( 10
s-1, are in fair agreement with the F loss rate 8-15 s-1 measured
in absence of Cl2.
As noted above, F atoms were scavenged by Br2 at the end
of the reactor to be detected as FBr molecules. This procedure
implies that any secondary reaction could not affect the FBr
concentration after its formation and therefore could not affect
the observed kinetics of F decay. Br atoms are the only active
species present at the zone of the reactor downstream of Br2
addition point. They are formed simultaneously with FBr in
F Atom Kinetics in an Excess of Cl2. In this series of
experiments to measure the rate constant of reaction 1, k1, F

1730 J. Phys. Chem. A, Vol. 108, No. 10, 2004
Nesbitt et al.
expression with a very weak positive temperature dependence:
k1 )
(6.5 ( 1.5) × 10-11 exp{-(20 ( 60)/T} cm3 molecule-1 s-1
T ) 180-360 K (10)
The results for the rate constant for reaction 1 from this study
differ significantly from those previously published. The only
other temperature study is that of Warnatz et al.2 in a discharge
flow mass spectrometer system that covered a temperature range
from T ) 232 to 350 K. As shown by the dashed line in Figure
8, this earlier study measured a much steeper temperature
dependence for k1 than was found in this present work. Their
derived Arrhenius expression is
Figure 8. For the combined data set the temperature dependence of
the rate constant of the reaction F + Cl2 f Cl + FCl. Continuous and
dashed lines represent the exponential fit to the experimental data from
this study and ref 2, respectively.
k1 ) 9.1 × 10-10 exp(-700/T} cm3 molecule-1 s-1
T ) 232-373 K
reaction 3 and also in reaction 8 of Cl atoms (product of reaction
1) with Br2:11
where the activation energy is more than a factor of 10 greater
than that derived in this work. Figure 8 also shows the value of
1.6 × 10-10 cm3 molecule-1 s-1 measured by Appelman and
Clyne3 at T ) 298 K in a discharge flow mass spectrometer
system; this value of k1 is about a factor of 2 greater than that
measured in this present study.
Cl + Br2 f BrCl + Br
(8)
k8 ) 2.3 × 10-10 exp(-135/T) cm3 molecule-1 s-1
It is generally accepted that there is no secondary chemistry
in the F + Cl2 reaction system.1 In addition, the previous
studies1-3 as well as the present one all employed the same
experimental technique of discharge flow mass spectrometry.
Therefore, there is no obvious explanation for the discrepancy
in the values of k1 between those from the present study and
those from the earlier studies.1-3
Considering that reaction 9 of Br atoms with FBr is endothermic
by approximately 14 kcal mol-1, its possible influence in the
present experiments can be ruled out:
Br + FBr f F + Br2
(9)
Excellent agreement between the results obtained for k1 under
different experimental conditions (see Tables 1 and 2) is
additional evidence that the procedure used for F atom detection
was correct.
Acknowledgment. We thank very much Dr. Louis Stief for
all the advice and help he has given us. This work was supported
by the NASA Planetary Atmospheres Research Program and
the NASA Upper Atmosphere Research Program. Dr. Fred
Nesbitt acknowledges the support of NASA Cooperative Agree-
ment NCC5-68 to the Catholic University of America. D.A.D.
thanks the NASA Undergraduate Student Research Program at
Goddard Space Flight Center for support.
Discussion
From the GSFC laboratory the average temperature inde-
pendent rate constant is k1 ) (5.7 ( 0.8) × 10-11 cm3
molecule-1 s-1. An Arrhenius plot of these data for the
temperature range T ) 180-298 K shows a relatively flat
temperature dependence. For the results from the CNRS
laboratory the least-squares analysis of the data provides the
Arrhenius expression: k1 ) (4.8 ( 0.5) × 10-11 exp{(70 ( 60)/
T} cm3 molecule-1 s-1 for T ) 230-360 K, where the quoted
uncertainties represent 2σ for the activation energy and 1σ for
the preexponential factor. Thus, this shows no more than a small
negative temperature dependence of the rate constant of the F
+ Cl2 reaction. In fact, considering the experimental uncertainty,
the temperature independent value of k1 ) (6.2 ( 0.8) × 10-11
cm3 molecule-1 s-1 at T ) 230-360 K can be also recom-
mended from the CNRS set of results.
When the two sets of laboratory results from CNRS and
GSFC are combined and treated as one set of results, the average
temperature independent value derived for k1 is (6.0 ( 1.1) ×
10-11 cm3 molecule-1 s-1 for the temperature range 180-360
K. In Figure 8 all the results from both laboratories are plotted
according to the Arrhenius equation. The least-squares analysis
of the combined results provides the following Arrhenius
References and Notes
(1) Clyne, M. A. A.; McKenney, D. J.; Walker, R. F. Can. J. Chem.
1973, 51, 3596.
(2) Warnatz, Von J.; Wagner, H. Gg.; Zetzsch, C. Ber. Bunsen-Ges.
Phys. Chem. 1971, 75, 119.
(3) Appleman, E. H.; Clyne, M. A. A. J. Chem. Soc., Faraday Trans.
1 1975, 71, 2072.
(4) Cody, R. J.; Payne, W. A., Jr.; Thorn, R. P., Jr.; Nesbitt, F. L.;
Iannone, M. A.; Tardy, D. C.; Stief, L. J. J. Phys. Chem. A 2002, 106,
6060.
(5) Poulet, G.; Lancar, I. T.; Laverdet, G.; Le Bras, G. J. Phys. Chem.
1990, 94, 278.
(6) Bemand, P. P.; Clyne, M. A. A. J. Chem. Soc. Faraday Trans. 2
1976, 72, 191.
(7) Lewis, R. S.; Sander, S. P.; Wagner, W.; Watson, R. T. J. Phys.
Chem. 1980, 84, 2009.
(8) Kaufman, F. J. Phys. Chem. 1984, 88, 4909.
(9) Marrero, T. R.; Mason, E. A. J. Phys. Chem. Ref. Data 1972, 1, 3.
(10) Zelenov, V. V.; Kukui, A. S.; Dodonov, A. F. SoV. J. Chem. Phys.
1990, 5, 1109.
(11) Bedjanian, Y.; I. T.; Laverdet, G.; Le Bras, G. J. Phys. Chem. A
1998, 102, 953.