Kinetic studies of the Cl + HI reaction using three techniques

Article abstract of DOI:10.1021/jp047411c

The reaction Cl + HI → HCl + I was investigated using the laser-photolysis/resonance fluorescence method to monitor (i) the disappearance of atomic Cl and (ii) the growth of atomic I. The results were similar, and together yield the rate constant k = (3.4 ± 1.4) × 10^-11 exp (+2.8±1.1 kJ mol^-1/RT) cm³ molecule ^-1s^-1 (1σerrors) over 297-390 K. The 95% confidence limits for k are ±19%. Another approach, discharge/fast-flow (iii), with monitoring of the consumption of Cl in the presence of excess HI, yields k = 1.2 × 10^-10 cm³ molecule^-1 s^-1 at 298 K, with 95% confidence limits of ±17%. The results are in accord with each other and with most literature data. This indicates that monitoring of the formation of the atomic halogen product in this example of a radical abstraction of H from a hydrogen halide is reliable.

Full text of DOI:10.1021/jp047411c

J. Phys. Chem. A 2004, 108, 6857-6862
6857
Kinetic Studies of the Cl + HI Reaction Using Three Techniques
Jessie Yuan,† Ashutosh Misra,‡ A. Goumri, Diane D. Shao, and Paul Marshall*
Department of Chemistry, UniVersity of North Texas, P.O. Box 305070, Denton, Texas 76203-5070
ReceiVed: June 15, 2004
The reaction Cl + HI f HCl + I was investigated using the laser-photolysis/resonance fluorescence method
to monitor (i) the disappearance of atomic Cl and (ii) the growth of atomic I. The results were similar, and
together yield the rate constant k ) (3.4 ( 1.4) x 10-11 exp (+2.8 ( 1.1 kJ mol-1/RT) cm3 molecule-1 s-1
(1σ errors) over 297-390 K. The 95% confidence limits for k are (19%. Another approach, discharge/fast-
flow (iii), with monitoring of the consumption of Cl in the presence of excess HI, yields k ) 1.2 × 10-10 cm3
molecule-1 s-1 at 298 K, with 95% confidence limits of (17%. The results are in accord with each other and
with most literature data. This indicates that monitoring of the formation of the atomic halogen product in
this example of a radical abstraction of H from a hydrogen halide is reliable.
1. Introduction
The fast reaction
I) radicals, and check whether the same kinetics are obtained
by monitoring I or Cl during reaction 1. As a further check, Cl
was monitored both in time-resolved and fast-flow systems.
Cl + HI f HCl + I
(1)
2. Methodology
is a well-known infrared chemiluminescent (IRCL) process that
has been important in the development of reaction dynamics
and investigations of energy disposal.1,2 We have investigated
the thermal rate constant k1 via three techniques. The temper-
ature dependence of k1 was evaluated by laser-photolysis/
resonance fluorescence, with monitoring either of (i) the
disappearance of Cl, or (ii) the appearance of a product, atomic
I. (iii) Consumption of Cl atoms in the presence of excess HI
at room temperature was also studied in a discharge flow system.
The rate constants obtained here are compared with previous
data derived via IRCL3-6 and relative rate measurements.7 The
main motivation for the present work is to test the reliability of
kinetic monitoring of the atomic halogen product X in reactions
of the general type
2.1 Flash-photolysis Experiments. General operation of our
pulsed photolysis reactor has been described elsewhere.12,13 In
this study, Cl2 was photolyzed to produce Cl radicals by excimer
laser (Questek 2010) radiation at 308 nm, and the pulse energy
E was measured with a Molectron J25 pyroelectric joulemeter.
This chlorine photolysis wavelength takes advantage of the
modest UV absorbance of HI at 308 nm,14 and no atomic iodine
was detected when HI alone was subjected to the photolysis
beam. The Cl atoms reacted with excess HI to form HCl and I.
During the course of the reaction either the decay of [Cl] or the
growth of [I] was monitored by time-resolved resonance
fluorescence. Emitted radiation was focused through a calcium
fluoride lens and captured by a solar-blind photomultiplier tube
orthogonal to both the resonance lamp and the excimer laser.
The signals were processed through a multichannel scaler with
photon counting and signal averaging. Microwave-powered (90
W) discharge lamps generated resonance radiation; for Cl (λ )
134-140 nm),14 a dilution of 0.3% of CCl4 in Ar flowed
continuously through the lamp, while a 0.5% CH3I/Ar mixture
was used for I (λ ) 178-206 nm).14
Ar (99.998%, Big Three) bath gas and diluent was taken
directly from the cylinder. HI and Cl2 were synthesized in our
laboratory. HI was made by the action of concentrated phos-
phoric acid on potassium iodide at 343-363 K.15 The resulting
HI was trapped with a methanol slush at about 175 K, and
purified several times by low-temperature distillation from 240
K to remove any I2 impurities, until the final condensed HI was
white. The HI was stored in a blackened bulb to prevent
photodecomposition. Cl2 was produced by mixing concentrated
HCl and solid KMnO4 at room temperature,16 and was purified
in the same way as HI. The purity of the resulting Cl2 was
checked through its UV-visible spectrum.
R + HX f RH + X
(2)
where R is a radical and where X is readily detectable by
resonance fluorescence. Such reactions have been used in the
cases of R ) alkyl and silyl radicals to obtain kinetic information
which, together with data for the reverse reaction, establishes
C-H and Si-H bond dissociation enthalpies.8-10 There have
been criticisms of this kind of experiment.11 These center on
the possible influence of excess energy in R when produced
photolytically by photons with energies well above the threshold
for dissociation of a precursor, the possibility of photochemistry
arising from the monitoring radiation, and the need to correct
for diffusion effects. Some work involved major (up to 80%)
consumption of R by an unknown process that does not produce
X, which is speculated to arise from oxygen contamination.9
This, in turn, raises the possibility of O and O2 photochemistry.11
Here, we can monitor both reactant (R ) Cl) and product (X )
* marshall@unt.edu.
HI and Cl2 were premixed with the flow of Ar before entering
the reactor, under “slow flow” conditions. Since the concentra-
tion of HI is much higher than the concentration of Cl, the
reaction of Cl with HI is pseudo-first order. The analysis of the
† Present address: Texas Instruments, Inc., 13553 Floyd Road, Dallas,
TX 75265.
‡ Present address: Air Liquide Electronics Chemicals and Services, Inc.,
13456 North Central Expressway, Dallas, TX 75243.
10.1021/jp047411c CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/27/2004

6858 J. Phys. Chem. A, Vol. 108, No. 33, 2004
Yuan et al.
Figure 1. Time profiles of (a) Cl-atom and (b) I-atom fluorescence
following 308 nm laser photolysis of Cl2/HI mixtures at 354 K.
Figure 3. Schematic diagram of the discharge/fast-flow apparatus.
background from scattered light (measured before triggering the
photolysis pulse), the four parameters in eq 6 were derived from
nonlinear fitting of the I fluorescence signals. Figure 1b shows
an example of the growth of I. The coefficient k1 was obtained
as the slope of a plot of kps1 vs [HI], as in the Cl-monitoring
experiments. When [HI] ) 0, no I is produced, and for this
datum kps1 was obtained instead from the loss of Cl. An example
is shown in Figure 2. At the highest temperatures, [I] was
observed to increase slowly following its rapid generation from
reaction 1, which corresponds to a negative value for kd4,
typically -10 s-1. The origin of this slow growth term for [I]
is unclear, but it is numerically well-separated from typical kps1
values of 500-3000 s-1
.
A potential difficulty with these experiments is the loss of
HI by adsorption on the reactor walls. This effect was
investigated by using two reactors, made of Pyrex and steel,
and varying the passivation conditions. Several parameters were
varied to check their influence on the rate constants. The average
gas residence time before photolysis, τ, was varied by a factor
of 2 and the concentration of Cl2 was changed by a factor of 4.
The photolysis energy E and [Cl2] were varied to give different
[Cl]0. The chlorine concentration [Cl]0 was placed on an absolute
basis through the absorbed photon flux (calculated from E), the
laser beam cross section, and the absorption cross-section of
Cl2 (1.85 × 10-19 cm2,17) combined with a quantum yield of 2
for atomic chlorine.14
Figure 2. Pseudo-first-order loss rate of Cl measured directly (circles)
and from the formation rate of I (squares) in the reaction Cl + HI f
HCl + I. The filled symbols correspond to the data of Figure 1.
signals (Cl or I) was carried out with different algorithms, both
of them based on the following kinetic scheme, of reaction 1
plus two loss processes:
Cl f pseudo first-order loss, kd3
I f pseudo first-order loss, kd4
(3)
(4)
2.2 Discharge Flow Experiments. A schematic diagram of
the flow tube apparatus is shown in Figure 3. The apparatus
consists of a 70-cm-long Pyrex flow tube (radius r ) 1.05 cm).
Cl atoms are produced in the sidearm of the flow tube by a
microwave discharge (Opthos Instruments, 2450 MHz) through
a dilution of Cl2 in Ar. The second reagent is introduced into
the flow tube by means of a moveable injector (o.d. 1.5 cm).
The inside of the flow tube is coated with halocarbon wax
(Series 1500-Wax, Halocarbon Products Corp.) and the outside
of the injector is coated with phosphoric acid to minimize
heterogeneous loss of atoms at the surface.18 About half of the
flow of Ar is diverted through the sidearm to push the Cl atoms
through the flow tube. Low pressures, around 3.5 mbar, are
maintained in the flow tube by a rotary pump (Sargent Welch
1397, 500 L min-1).
The decay of Cl is analyzed with a simple exponential function
because its concentration is expected to vary as
[Cl])[Cl]0 exp(-kps1 t)
(5)
where [Cl]o is the initial [Cl]. The pseudo-first-order decay
coefficient kps1 is equal to k1[HI] + kd3. The coefficient kd3
accounts for the slow loss of Cl other than by reaction 1, which
is mainly through diffusion to the reactor walls, and was
typically around 15-35 s-1. The second-order rate constant k1
is derived as the slope of plots of kps1 vs [HI]. Figure 1a shows
a typical Cl decay, and Figure 2 is an example plot of kps1 vs
[HI].
Solution of the rate eqs 1, 3, and 4 leads to a double
exponential function for the growth and decay of [I]:
The detection region consists of a six-arm cell made of Pyrex
glass. One arm of this cell is connected to the flow tube and
the opposite arm is connected to the pump. Atomic Cl
concentrations are monitored by resonance fluorescence. Vacuum
UV fluorescence at λ ) 135 nm is excited by a microwave-
[I] ) B exp (-kd4 t) - C exp (-kps1 t)
(6)
where B ) ([I]0 (kps1 - kd4) + k1[HI][Cl]0)/(kps1 - kd4) and C
) k1[HI][Cl]0 /(kps1 - kd4). After subtraction of the constant

Kinetic Studies of the Cl + HI Reaction
J. Phys. Chem. A, Vol. 108, No. 33, 2004 6859
powered discharge lamp connected to one of the arms at right
angles to the flow tube, through which a 1% mixture of Cl2 in
Ar is passed. Fluorescence is detected with a solar blind
photomultiplier tube (Thorn EMI, model 9423) optically coupled
to the six-arm cell via a calcium fluoride lens (f.l. 5.0 cm).
Output from the photomultiplier tube is fed to an amplifier-
discriminator (MIT, model F-100T), which sends TTL pulses
to the computer-controlled multichannel scaler (EG&G Ortec,
model ACE-MCS) for photon counting. A Woods horn opposite
the discharge lamp minimizes scattered light. The pressure in
the flow tube is measured with a capacitance manometer (MKS
Instruments, model 122A or 222C) opposite the arm to which
the PMT is connected.
Reaction 1 was investigated by monitoring the decay of Cl
atoms in the presence of excess molecular reagent on a time
scale of 2 to 20 ms. The measurements of k1 were carried out
under approximately pseudo-first-order conditions with [Cl] <
[HI] so that [HI] was effectively constant. Cl atoms are lost via
reaction 1 and heterogeneous loss on the exposed surface of
the injector, characterized by kinj. There is also a constant
contribution from the loss of Cl at the surface of the flow tube.
Thus, an exponential decay of [Cl] is expected
where D is the diffusion coefficient of atomic Cl in Ar. The
first correction term in eq 9 may be identified with axial
diffusion,20 which arises when Cl atoms diffuse down the
concentration gradient along the flow tube. The second correc-
tion term accounts for radial diffusion of Cl across the flow
tube.20 The value of D is calculated21 from the Lennard-Jones
parameters σ and ꢀ/k for Cl (3.36 Å, 328 K) and Ar (3.418 Å,
124 K).22 Straight-line plots of the corrected kps1 values vs [HI]
would have a slope equal to an effective value of the rate
constant for reaction 1, and have kinj as the intercept. However,
further corrections were made to the kps1 values before obtaining
k1, as outlined below.
A separate set of titration experiments was carried out to
determine the absolute Cl atom concentration. The titrations
employed reaction 1, with the injector tip kept at the middle of
the range of l used in the kinetic runs. Under the titration
conditions of long residence time and high [HI], reaction 1 is
essentially complete before fluorescence is measured, and the
intensity falls to zero when [HI]end point ) [Cl]o. These absolute
concentrations were employed in simulations of [Cl] vs time
profiles taking into account reaction 1, surface loss, and the
potentially interfering secondary chemistry
I + Cl2 f Cl + ICl k ) 8.5 × 10-17 cm3 molecule-1 s-1
[Cl] ) [Cl]o exp(-kps1 t - c)
(7)
(10)
where the residence time t of the mixed gases is l/V; V is the
mean gas velocity and l is the distance from the moveable
reagent injector tip to the observation window. The loss of Cl
at the reactor walls is accounted for by c, which is constant
during each set of measurements carried out at fixed V and
variable l.
I + I + Ar f I2 + Ar k ) 8.2 × 10-33 cm6 molecule-2 s-1
(11)
Cl + I2 f ICl + I k ) 2.4 × 10-11 cm3 molecule-1 s-1
The pseudo-first-order decay constant kps1 is related to the
rate constant k1 via
(12)
with literature values for the 298-K rate constants.6,23,24 The
conditions (concentrations of species, residence time) were the
same as those used for the kinetic runs, and simulation of each
([Cl], l) point is carried out with the ACUCHEM program25 to
yield a time profile of atomic chlorine. The slope of a ln([Cl]/
[Cl]0) vs t plot provides a model kps1 value for the particular
measurement, which was then divided by the initial concentra-
kps1 ) k1[HI] + kinj
(8)
Graphs of ln [Cl]/[Cl]o vs l are linear with a slope equal to the
observed decay coefficient, kps1,obs, at a given [HI].
In each experiment, l was varied by 35 cm. The effective
first-order rate constant for loss of Cl atoms at the exterior
surface of the injector, kinj, was measured at the beginning of
each experiment by introducing Cl atoms in the flow tube in
the absence of molecular reagent and monitoring the change in
[Cl] as the injector was gradually withdrawn. Growth of atomic
Cl is expected because of the decreased surface area of the
injector exposed to Cl atoms as the injector is moved out of
the flow tube. Thus, as the reaction time increases, [Cl]
increases, and kinj in eq 8 is negative. The measurement of kinj
before each run was followed by passivation of the flow tube
using a flow of diluted (∼1%) HI over a period of 45 min. The
passivation procedure helps to reduce the scatter between kinetic
runs. The coefficient kinj is small compared to kps1,obs and,
because it is constant, is separable from kps1,obs. Observation of
an increase in kinj is a sign that the injector and wall coatings
need to be replaced.
tion of HI to yield an effective second-order rate constant, kmodel
.
The correction factor for each experimental measurement can
then be written as kinput/kmodel, where kinput is the value of the
rate constant put into the simulation. We used kinput ) 9.5 ×
10-11 cm3 molecule-1 s-1. Each experimental kps1 value (already
corrected for diffusion via eq 9) was multiplied by the
appropriate correction factor, and then plotted against [HI] to
yield the final corrected rate constant k1. Figure 4 is an example.
3. Results and Discussion
Table 1 summarizes the laser photolysis measurements for
CI + HI under different experimental conditions between
temperatures of 296 and 390 K. The listed standard deviations
reflect the 1σ statistical uncertainties in the slopes of plots such
as Figure 2, combined in quadrature with a 5% 1σ allowance
for instrumental errors. It took at least 20 min for a flow of HI
to saturate the surface of the steel reactor. Alternatively, a few
mbar of pure HI was incubated in the reactor for an hour before
starting the measurements. Under these two conditions, the
measured rate constants were close to the values obtained with
the Pyrex reactor. There was no significant change in the rate
constant with the photolysis energy or [Cl]0, which indicates
that secondary reactions with products or other photolysis
fragments are negligible. As may be seen in Figure 2 and in
Our experiments were carried out with Reynolds numbers,
Re, typically around 100. The length required to develop laminar
flow is 0.12 r Re,19 that is, about 12 cm, so that experiments
were performed in regions where laminar flow was close to
completely developed. The kps1,obs values are corrected for effects
of diffusion using the relation20
kps1 ) kps1,obs(1 + kps1,obs D/V2 + kps1,obs r2/48D) (9)

6860 J. Phys. Chem. A, Vol. 108, No. 33, 2004
Yuan et al.
TABLE 1: Summary of Laser Photolysis Measurements on HI + Cl
[HI],
[Cl2],
[Cl]0,
monitored
atom
k1 ( σk1,
T, K P, mbar
τ
res, s 1013 molecule cm-3 1013 molecule cm-3 laser E, mJ 1012 molecule cm-3
10-10 molecule-1 cm3 s-1
296
296
297
297
297
297
297
297
297
297
319
319
319
319
344
344
344
353
353
354
354
354
381
381
386
386
390
390
390
386
45.9
31.1
38.7
38.7
39.1
38.1
38.1
44.4
44.5
0.4
0.7-3.3
0.5-1.8
0.8-2.8
0.6-2.2
1.9-6.9
0.9-5.4
0.9-5.4
1.1-5.3
1.1-5.3
22.0
15.6
5.1
12.1
11.8
14.8
14.8
12.2
24.0
43
43
58
58
47
43
43
40
40
1.8
1.3
0.6
1.3
1.0
1.2
1.2
0.9
1.8
Cl
Cl
Cl
Cl
I
I
I
1.07 ( 0.07
1.01 ( 0.05
0.93 ( 0.06b
0.96 ( 0.05b
1.04 ( 0.09b
1.05 ( 0.07b
1.05 ( 0.09b
1.03 ( 0.07b
1.14 ( 0.12b
1.01 ( 0.02a
1.13 ( 0.06
0.82 ( 0.05
1.22 ( 0.10
0.98 ( 0.04a
1.22 ( 0.13
1.07 ( 0.07
1.10 ( 0.06a
1.10 ( 0.06
1.04 ( 0.06
0.88 ( 0.05
0.79 ( 0.05
0.93 ( 0.03a
0.60 ( 0.04
0.71 ( 0.05
0.70 ( 0.05
0.77 ( 0.06
0.86 ( 0.05
0.90 ( 0.08
0.99 ( 0.07
0.75 ( 0.02a
0.5
0.6
0.7
0.8
0.8
0.8
0.7
0.7
I
I
39.9
39.7
39.9
0.8
0.4
0.4
1.0-5.7
0.8-4.3
0.8-3.6
16.6
8.1
16.1
50
50
50
1.6
0.8
1.6
I
I
I
40.0
40.9
0.8
0.4
0.7-3.3
0.5-3.1
14.2
7.2
43
43
1.1
0.6
I
I
41.3
41.3
31.6
46.3
0.4
0.4
0.4
0.3
0.7-3.3
0.4-3.0
0.5-4.0
0.7-4.0
19.6
19.6
8.8
36
36
58
58
1.3
1.3
1.0
0.7
Cl
Cl
I
6.6
I
38.0
37.7
41.3
41.3
41.2
41.3
41.3
0.3
0.3
0.4
0.4
0.3
0.7
0.7
0.6-3.0
0.7-3.4
0.7-4.2
1.1-5.1
0.9-4.8
0.9-4.0
0.9-4.0
13.3
13.2
7.0
14.9
14.6
14.6
28.7
40
40
54
54
40
40
40
1.0
1.0
0.7
1.5
1.1
1.1
2.1
I
I
I
I
I
I
I
a Weighted average for each temperature (see text). b Measurements were made in the Pyrex reactor.
Figure 5. Arrhenius plot of rate constants for the Cl + HI reaction.
Solid circles and line, laser photolysis, present work (error bars are
2σ); open downward pointing triangle, flow tube, present work; open
square, Wodarczyk and Moore;3 open upward pointing triangle, Mei
and Moore;4 asterisks, Mei and Moore;5 open circle, Dolson and Leone;6
solid triangle, Iyers et al.7
Figure 4. Plot of pseudo first-order loss rate of Cl in the presence of
excess HI in the fast-flow reactor.
photolysis/resonance fluorescence data is
Table 1, similar rate constants are obtained via monitoring of
either Cl or I. The agreement is excellent at room temperature,
while the deviations are up to around 20% at 354 K. Given the
overall scatter in the data and potential systematic errors,
discussed below, this is good agreement. Therefore, all data were
combined at each temperature.
k1 ) (3.4 ( 1.4) x 10-11 exp (+2.8 (
1.1 kJ mol-1/RT) cm3 molecule-1 s-1 (13)
where the errors are 1σ. Since the negative activation energy is
not far from zero, an alternative summary of the data over 296-
390 K is the weighted mean of the measurements, k1 ) 0.90 ×
10-10 cm3 molecule-1 s-1. Consideration of the covariance and
the 1σ errors in eq 13 gives an estimated 1σ statistical
uncertainty of up to 8% in k1. With allowance for possible
systematic errors in temperature, pressure, and flow, we suggest
95% confidence limits of ( 19% for k1.
The Arrhenius plot of the Cl + HI reaction obtained in this
study and previous measurements by other techniques are shown
in Figure 5. Each point in the graph from this study is the
weighted mean for that temperature. As outlined by Bevington,26
we weight each value as wi ) 1/σi2, and obtain the mean µ and
its standard deviation σµ via the relations µ ) Σwixi/Σwi and
2
σµ ) 1/Σwi. The error bars in Figure 5 are 2σ, and it may be
seen that this purely statistical quantity accounts for much of
the scatter in the Arrhenius plot. The best fit of the laser-
Table 2 summarizes the discharge-flow measurements. The
column headed k1,obs indicates what would be obtained from a

Kinetic Studies of the Cl + HI Reaction
J. Phys. Chem. A, Vol. 108, No. 33, 2004 6861
TABLE 2: Summary of Discharge-flow Kinetic Data for HI + Cl
a
v,
[HI],
initial [Cl2],
[Cl],
kinj,
k1,obs( σk1,obs
,
k1 ( σk1,
T, K P, mbar cm s-1 1012 molecule cm-3 1012 molecule cm-3 1012 molecule cm-3 s-1 10-11 cm3 molecule-1 s-1 10-11 cm3 molecule-1 s-1
298
298
297
297
297
297
297
297
3.5
3.5
3.5
3.6
3.6
3.6
3.6
3.6
1946
1960
1915
1860
1843
1843
1850
1850
1.9-5.3
1.5-4.4
2.5-7.2
1.2-5.4
1.4-4.3
3.3-6.7
1.8-6.7
2.7-8.0
0.9
1.1
1.0
1.5
1.4
2.3
1.1
2.1
0.6
0.8
0.7
1.0
0.9
1.5
0.7
1.4
-46
-23
-28
-36
-35
-35
-13
-13
8.9 ( 0.5
9.4 ( 0.2
9.1 ( 0.2
9.0 ( 0.3
8.4 ( 1.0
8.7 ( 1.2
7.0 ( 0.7
8.4 ( 0.6
11.8 ( 1.4
13.2 ( 3.4
12.9 ( 3.0
13.2 ( 1.3
10.5 ( 1.8
13.8 ( 2.7
11.0 ( 1.3
13.3 ( 1.8
a k1,obs is the apparent value of k1 that would be obtained if kps1,obs (uncorrected for diffusion or stoichiometry) was plotted vs [HI].
plot of uncorrected kps1,obs values vs [HI]. The (1σ values
quoted are the statistical uncertainties in the slope of this kind
of plot. The calculated Cl-atom diffusion coefficient is D )
0.157(1 bar/P) cm2 s-1. The axial term of eq 9 yields minor
corrections from 0.3 to 6% to individual kps1 values. The radial
term has a larger impact, up to 35% on the largest kps1 values.
The measured absolute [Cl] shows that around a third of the
Cl2 introduced through the microwave discharge was dissociated.
The kinetic simulation indicates that secondary chemistry had
a negligible impact, but that stoichiometric correction to
individual kps1 values obtained at the lowest [HI]:[Cl] ratios
could be as large as 60% (although usually much smaller).
Because these corrections are largest for the lowest kps1 values,
the impact on k1 was moderate, up to 16%. The dominant
correction term is the radial component of eq 9. The final
column of Table 2 lists k1 values derived by incorporating all
these corrections, and the (1σ values quoted are the statistical
uncertainties combined in quadrature with a 10% (1σ) allowance
for uncertainty in the corrections. The weighted mean rate
constant at 298 K is k1 ) (1.22 ( 0.06) × 10-10 cm3 molecule-1
s-1, where the quoted error is (1σ in the mean. Allowance for
potential systematic errors yields 95% confidence limits of (
0.2 × 10-10 cm3 molecule-1 s-1. This flow tube determination
of k1 agrees with our photolysis measurements.
by the study of Iyer et al.7 They measured the rate constant for
removal of Cl by HI relative to C2H6 using a competitive
radiochemical technique, and derived k1 ) (1.26 ( 0.14) ×
10-10 cm3 molecule-1 s-1
.
With respect to the issues raised by Dobis and Benson11 noted
in the Introduction, small values for Cl diffusive loss, typically
up to a few percent of the overall consumption (see Figure 2),
mean that any uncertainty here has a negligible impact, and
that in this system there was no major loss pathway for Cl that
did not lead to I formation. The accord to within roughly 20%
between the three approaches applied here, approaches that vary
as to monitored species and bath gas pressure regime, is
consistent with experimental uncertainties. The general accord
with other prior determinations further suggests that all methods
are consistent for reaction 1. This implies that any excess
translational or electronic excitation in the photolytically
produced Cl atoms does not appreciably change the kinetics
relative to Cl atoms generated in the discharge/fast-flow system.
Further, in the flow system the reacting mixture is not exposed
to actinic radiation from a resonance lamp until the detection
region, unlike the pulsed photolysis system.
Conclusions
Three different methods to determine the rate constant k1 for
Cl + HI f HCl + I are consistent, and agree with most previous
literature data. This indicates that, at least for this system, the
approach of monitoring the growth of the atomic halogen
product in reactions of the type R + HX f RH + X appears
to be reliable.
At room temperature, hard-sphere collision theory combined
with standard diameters for Cl and HI of 3.4 and 4.1 Å,27
respectively, yields a collision rate of 2.1 × 10-10 cm3
molecule-1 s-1. Comparison with the measured k1 implies that
reaction occurs very readily, at about 1 out of 2 collisions, and
therefore there is little or no energy barrier, consistent with the
observed lack of temperature dependence.
Acknowledgment. This work was supported by the Robert
A. Welch Foundation (Grant B-1174) and the UNT Faculty
Research Fund.
In 1974 Wodarczyk and Moore3 determined the absolute rate
constant of this reaction at 295 K by photolyzing Cl2 with laser
pulses and monitoring the reaction via the time-resolved IRCL
of the vibrationally excited product, HCl. The average value of
the rate constant obtained for Cl + HI was (9.6 ( 2.4) × 10-11
cm3 molecule-1 s-1, in good agreement with this study. Three
years later, a new rate constant, 1.64 × 10-10 cm3 molecule-1
s-1, for the Cl + HI reaction was reported by the same group4
with supposedly more accurate [HI] determination and less wall-
catalyzed interference, although this value seems somewhat high.
The only published temperature dependence of k1 was obtained
by Mei and Moore,5 and was also based on IRCL measurements.
The observed reaction rates showed almost no temperature
dependence around a temperature range comparable to the
present study, although a significant variation of k1 was seen at
lower temperatures. Dolson and Leone6 studied the kinetics of
the slow chain of reactions 1 and 10 by real-time detection of
IRCL from vibrationally excited HCl. Cascading fluorescence
was observed from individual vibrational levels. A rate constant
of (1.4 ( 0.3) × 10-10 cm3 molecule-1 s-1 was obtained, which
agrees with the present study. A different approach is illustrated
References and Notes
(1) Maylotte, D. H.; Polanyi, J. C.; Woodall, K. B. J. Chem. Phys.
1972, 57, 1547.
(2) Levine, R. D.; Bernstein, R. B. Molecular Reaction Dynamics and
Chemical ReactiVity; Oxford University Press: New York, 1987, Ch. 1.
(3) Wodarczyk, F. J.; Moore, C. B. Chem. Phys. Lett. 1974, 26, 484.
(4) Mei, C. C.; Moore, C. B. J. Chem. Phys. 1977, 67, 3936.
(5) Mei, C. C.; Moore, C. B. J. Chem. Phys. 1979, 70, 1759.
(6) Dolson, D. A.; Leone, S. R. J. Chem. Phys. 1982, 77, 4009.
(7) Iyer, R. S.; Rogers, P. J.; Rowland, F. S. J. Phys. Chem. 1983, 87,
3799.
(8) Seakins, P. W.; Pilling, M. J. J. Phys. Chem. 1991, 95, 9874.
(9) Nicovich, J. M.; Van Dijk, C. A.; Kreutter, K. D.; Wine, P. H. J.
Phys. Chem. 1991, 95, 9890.
(10) Kalinovski, I. J.; Gutman, D.; Krasnoperov, L. N.; Goumri, A.;
Yuan, W.-J.; Marshall, P. J. Phys. Chem. 1994, 98, 9551.
(11) Dobis, O.; Benson, S. W. J. Am. Chem. Soc. 1995, 117, 8171.
(12) Shi, Y.; Marshall, P. J. Phys. Chem. 1991, 95, 1654.
(13) Ding, L.; Marshall, P. J. Phys. Chem. 1992, 96, 2197.
(14) Okabe, H. Photochemistry of Small Molecules; Wiley: New York,
1978.

6862 J. Phys. Chem. A, Vol. 108, No. 33, 2004
Yuan et al.
(15) Vogel, A. I. Textbook of Macro and Semimicro QualitatiVe
Inorganic Analysis, 4th ed.; Longmans, Green, and Co.: London, 1954.
(16) Mellor’s Modern Inorganic Chemistry; Parkes, G. D., Ed.; Wiley:
New York, 1967.
(17) DeMore, W. B.; Sander, S. P.; Golden, D. M.; Hampson, R. F.;
Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R.; Kolb, C. E.; Molina,
M. J. Chemical Kinetics and Photochemical data for Use in Stratospheric
Modeling. EValuation Number 12; JPL: Pasadena, 1997.
(18) Clyne, M. A. A.; MacRafert, A. J.; Murrells, T. P.; Stief, L. J. J.
Chem. Soc., Faraday Trans. 2 1984, 80, 877.
(21) Reid, R. C.; Sherwood, T. K. The Properties of Gases and Liquids;
McGraw-Hill: New York, 1958.
(22) Vandooren, V.; Fristrom, R. M.; Tiggelen, P. J. V. Bull. Soc. Chim.
Belg. 1992, 101, 901.
(23) Baulch, D. L.; Duxbury, J.; Grant, S. J.; Montague, D. C. J. Phys.
Chem. Ref. Data 1981, 10.
(24) Kuznetsova, S. V.; Maslov, A. I. Khim. Fiz. 1987, 6, 1554.
(25) Braun, W.; Herron, J. T.; Kahaner, D. Int. J. Chem. Kinet. 1988,
22, 51.
(26) Bevington, P. R. Data Reduction and Error Analysis for the
Physical Sciences; McGraw-Hill: New York, 1969, Ch. 5.
(27) Hirschfelder, J. O.; Curtis, C. F.; Bird, R. B. Molecular Theory of
Gases and Liquids; Wiley: New York, 1954.
(19) Padet, J. Fluides en Encoulement. Methodes et Modeles.; Masson:
Paris, 1991.
(20) Keyser, L. F. J. Phys. Chem. 1984, 88, 4750.