Kinetics, mechanism, and thermochemistry of the gas-phase reaction of atomic chlorine with pyridine

Article abstract of DOI:10.1039/b707017a

A laser flash photolysis-resonance fluorescence technique has been employed to study the kinetics of the reaction of atomic chlorine with pyridine (C ₅H₅N) as a function of temperature (215-435 K) and pressure (25-250 Torr) in nitrogen bath gas. At T ≥ 299 K, measured rate coefficients are pressure independent and a significant H/D kinetic isotope effect is observed, suggesting that hydrogen abstraction is the dominant reaction pathway. The following Arrhenius expression adequately describes all kinetic data at 299-435 K for C₅H₅N: k_1a = (2.08 ± 0.47) × 10^-11 exp[-(1410 ± 80)/T] cm ³ molecule^-1 s^-1 (uncertainties are 2σ, precision only). At 216 K ≤ T ≤ 270 K, measured rate coefficients are pressure dependent and are much faster than computed from the above Arrhenius expression for the H-abstraction pathway, suggesting that the dominant reaction pathway at low temperature is formation of a stable adduct. Over the ranges of temperature, pressure, and pyridine concentration investigated, the adduct undergoes dissociation on the time scale of our experiments (10 ^-5-10^-2 s) and establishes an equilibrium with Cl and pyridine. Equilibrium constants for adduct formation and dissociation are determined from the forward and reverse rate coefficients. Second- and third-law analyses of the equilibrium data lead to the following thermochemical parameters for the addition reaction: Δ_rH°₂₉₈ = -47.2 ± 2.8 kJ mol^-1, Δ_rH°₀ = -46.7 ± 3.2 kJ mol^-1, and Δ_rS° ₂₉₈ = -98.7 ± 6.5 J mol^-1 K^-1. The enthalpy changes derived from our data are in good agreement with ab initio calculations reported in the literature (which suggest that the adduct structure is planar and involves formation of an N-Cl σ-bond). In conjunction with the well-known heats of formation of atomic chlorine and pyridine, the above Δ_rH values lead to the following heats of formation for C ₅H₅N-Cl at 298 K and 0 K: Δ_fH° ₂₉₈ = 216.0 ± 4.1 kJ mol^-1, Δ _fH°₀ = 233.4 ± 4.6 kJ mol^-1. Addition of Cl to pyridine could be an important atmospheric loss process for pyridine if the C₅H₅N-Cl product is chemically degraded by processes that do not regenerate pyridine with high yield. the Owner Societies.

Full text of DOI:10.1039/b707017a

View Article Online / Journal Homepage / Table of Contents for this issue
This paper is published as part of a PCCP Themed Issue on:
Synergies between Experimental and Theoretical
Studies of Gas Phase Reactions
Guest Edited by: Paul Seakins (Leeds) and Struan Robertson (Accelrys,
Cambridge and Leeds)
Published in issue 31, 2007 of PCCP
Image reproduced by permission of Dr Taatjes and Linda A.Houston for Sandia National
Laboratories from Phys.Chem.Chem.Phys., 2007, 9, 4315-4331
Other papers published in this issue include:
Direct detection of polyynes formation from the reaction of ethynyl radical (C2H) with propyne (CH3–
C
CH) and allene (CH2 C CH2)
Fabien Goulay, David L. Osborn, Craig A. Taatjes, Peng Zou, Giovanni Meloni and
Stephen R. Leone, Phys. Chem. Chem. Phys., 2007
DOI: 10.1039/b614502g
Master equation methods for multiple well systems: application to the 1-,2-pentyl system
Struan H. Robertson, Michael J. Pilling, Luminita C. Jitariu and Ian H. Hillier, Phys. Chem. Chem. Phys.,
2007
DOI: 10.1039/b704736c
Ab initio methods for reactive potential surfaces
Lawrence B. Harding, Stephen J. Klippenstein and Ahren W. Jasper, Phys. Chem. Chem. Phys., 2007
DOI: 10.1039/b705390h
Visit the website for both cutting edge research papers and authoritative
review articles by leaders in a range of fields
www.rsc.org/pccp

PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
Kinetics, mechanism, and thermochemistry of the gas-phase reaction of
atomic chlorine with pyridine
Z. Zhao,a D. T. Huskey,b K. J. Olsen,b J. M. Nicovich,b M. L. McKeec and
P. H. Wine*ab
Received 10th May 2007, Accepted 1st June 2007
First published as an Advance Article on the web 28th June 2007
DOI: 10.1039/b707017a
A laser flash photolysis-resonance fluorescence technique has been employed to study the kinetics
of the reaction of atomic chlorine with pyridine (C5H5N) as a function of temperature (215–435 K)
and pressure (25–250 Torr) in nitrogen bath gas. At T Z 299 K, measured rate coefficients are
pressure independent and a significant H/D kinetic isotope effect is observed, suggesting that
hydrogen abstraction is the dominant reaction pathway. The following Arrhenius expression
adequately describes all kinetic data at 299–435 K for C5H5N: k1a = (2.08 ꢀ 0.47) ꢁ 10ꢂ11
exp[–(1410 ꢀ 80)/T] cm3 moleculeꢂ1 sꢂ1 (uncertainties are 2s, precision only). At 216 K r T r
270 K, measured rate coefficients are pressure dependent and are much faster than computed
from the above Arrhenius expression for the H-abstraction pathway, suggesting that the
dominant reaction pathway at low temperature is formation of a stable adduct. Over the ranges
of temperature, pressure, and pyridine concentration investigated, the adduct undergoes
dissociation on the time scale of our experiments (10ꢂ5–10ꢂ2 s) and establishes an equilibrium
with Cl and pyridine. Equilibrium constants for adduct formation and dissociation are
determined from the forward and reverse rate coefficients. Second- and third-law analyses of the
equilibrium data lead to the following thermochemical parameters for the addition reaction:
DrH1298 = ꢂ47.2 ꢀ 2.8 kJ molꢂ1, DrH01 = ꢂ46.7 ꢀ 3.2 kJ molꢂ1, and DrS1298 = ꢂ98.7 ꢀ 6.5 J
molꢂ1
K
ꢂ1. The enthalpy changes derived from our data are in good agreement with ab initio
calculations reported in the literature (which suggest that the adduct structure is planar and
involves formation of an N–Cl s-bond). In conjunction with the well-known heats of formation
of atomic chlorine and pyridine, the above DrH values lead to the following heats of formation
for C5H5N–Cl at 298 K and 0 K: DfH1298 = 216.0 ꢀ 4.1 kJ molꢂ1, DfH10 = 233.4 ꢀ 4.6 kJ molꢂ1
Addition of Cl to pyridine could be an important atmospheric loss process for pyridine if the
.
C5H5N–Cl product is chemically degraded by processes that do not regenerate pyridine with high
yield.
sion budgets and concentration distributions, as well as their
Introduction
lifetimes toward chemical loss in the atmosphere. The primary
atmospheric sink for pyridine is generally assumed to be
reaction with OH, although this is a relatively slow reaction
with a rate coefficient at 298 K of B5 ꢁ 10ꢂ13 cm3 moleculeꢂ1
Pyridine is an important heterocyclic aromatic compound that
is widely used as an intermediate in the production of insecti-
cides and herbicides. It is emitted into the atmosphere as a
result of industrial activity and also as a result of coal
combustion.1 The protonated ions of pyridine and its deriva-
tives have been observed at significant levels in measurements
of the background ion composition of the lower tropo-
sphere.2–5 The ion content of the atmosphere is of current
interest because of its influence on atmospheric electrical
conductivity and as a trigger for the formation of ultrafine
aerosol particles via ion-induced nucleation, which can impact
global climate.6,7 Quantification of the role of ion precursors
in global climate requires improved knowledge of their emis-
sꢂ1 8,9
Reaction with HNO3 may also be an important atmo-
.
spheric removal process for pyridine, although only a lower
limit for the pyridine + HNO3 rate coefficient is reported in
the literature (1.5 ꢁ 10ꢂ16 cm3 moleculeꢂ1 sꢂ1 at 298 K).8
While OH is believed to be the major oxidizing agent of
organic compounds in the troposphere, there is increasing
evidence that atomic chlorine may also play an important role
in oxidizing pollutants, particularly in the marine and coastal
atmosphere, where the Cl number density may be as high as
104–105 atoms cmꢂ3 10–12
i.e., only 1–2 orders of magnitude
,
smaller than typical OH number densities.
a School of Earth & Atmospheric Sciences, Georgia Institute of
Technology, Atlanta, GA 30332-0340, USA
In addition to its potential impact in atmospheric chemistry,
the reaction of atomic chlorine with pyridine is also of interest
to physical organic chemists. It has been known for fifty years
that the selectivity for tertiary versus primary hydrogen ab-
straction in photo-chlorination reactions can be dramatically
b School of Chemistry & Biochemistry, Georgia Institute of
Technology, Atlanta, GA 30332-0400, USA
c Department of Chemistry & Biochemistry, Auburn University, AL
36849, USA
ꢃc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4383–4394 | 4383

enhanced by conducting the reactions in certain solvents
including pyridine, a number of other aromatics, and CS2,
an effect that has been attributed to formation of weakly-
bound Cl–solvent complexes.13–17 By studying the Cl +
pyridine system both experimentally and theoretically, Bre-
slow et al.18 suggested that the complex involves a loose s*-
type bond between Cl and the pyridine nitrogen atom. Further
experimental19 and theoretical20 work supports the Breslow
et al. interpretation, and provides improved knowledge of the
geometry and thermochemistry of C5H5N–Cl.
Chlorine atoms were produced by 355 nm laser flash
photolysis of Cl2
Cl2 þ hnð355 nmÞ ! Cl þ Cl
ð2Þ
Third harmonic radiation from a Quanta Ray Model DCR-3
Nd:YAG laser served as the 355 nm light source. The photo-
lysis laser can deliver up to 60 mJ pulseꢂ1 at a repetition rate of
up to 10 Hz; the pulse width was 6 ns. Fluences employed in
this study ranged from 0.5 to 36 mJ cmꢂ2 pulseꢂ1. The use of
Cl2 as the photolyte for producing Cl atoms has the disadvan-
tage that the carbon-centered radical generated as a product of
the H-abstraction pathway is likely to react rapidly with Cl2 to
regenerate Cl. Our approach for obtaining quantitative kinetic
information in spite of this secondary chemistry complication
is discussed in detail in a later section of the paper. Secondary
regeneration of Cl can often be avoided by using 248 or 266 nm
photolysis of Cl2CO as the Cl atom source.21,22,25 However,
strong absorption by pyridine in the 240–270 nm wavelength
region27 makes use of Cl2CO as the photolyte unfeasible in this
study.
In this paper we report the first experimental study of the
gas-phase kinetics of reaction (1). Our approach couples laser
flash photolysis of Cl2–pyridine–N2 mixtures with time-
resolved detection of Cl by atomic resonance fluorescence
spectroscopy. Pressure independent kinetics are observed at
high temperature and pressure-dependent adduct 2 Cl equi-
libration kinetics are observed at low temperature. The experi-
mental results allow rate coefficients to be evaluated for the
pressure independent pathway 1a (presumably hydrogen ab-
straction), the adduct formation pathway 1b, and the adduct
dissociation pathway ꢂ1b:
An atomic resonance lamp situated perpendicular to the
photolysis laser excited resonance fluorescence in the photo-
lytically-generated Cl atoms. About 1 Torr of helium contain-
ing a trace of Cl2 flowed into and was pumped out of the lamp
through two side arms. The flows of a 0.1% Cl2 in He mixture
and pure He into the lamp were controlled by separate needle
valves, thus allowing the total pressure and Cl2 concentration to
be independently adjusted for optimum signal-to-noise. Radia-
tion was coupled out of the lamp through a magnesium fluoride
window and into the reaction cell through a magnesium fluoride
lens. The region between the resonance lamp and the reaction
cell was purged with N2 to allow transmission of Cl resonance
lines in the 135–140 nm range. Fluorescence intensities were
found to vary linearly with atom concentration up to levels
several times higher than any employed in the kinetics experi-
ments ([Cl]0 o 4 ꢁ 1011 atoms cmꢂ3 in all experiments).
Fluorescence was collected by a calcium fluoride lens on an
axis orthogonal to both the photolysis laser beam and the
resonance lamp beam and imaged onto the photocathode of a
solar blind photomultiplier tube. The calcium fluoride lens
eliminates detection of emissions at wavelengths shorter than
125 nm (Lyman-a emission, for example). The region between
the reaction cell and the photomultiplier was purged with N2.
Signals were processed using photon counting techniques in
conjunction with multi-channel scaling. For each chlorine atom
decay temporal profile measured, signals from a large number
of laser shots were averaged to obtain a temporal profile with
adequate signal to noise for quantitative kinetic analysis. The
multi-channel scalar sweep was triggered prior to the photolysis
laser to allow a pre-trigger baseline to be obtained.
Cl þ C5H5N ! C5H4N þ HCl
ð1aÞ
Cl þ C5H5N þ M $ C5H5NꢂCl þ M
ð1b; --1bÞ
In addition, approximate rate coefficients for the reaction of
Cl2 with o-pyridinyl radical (the H-abstraction product) are
obtained as are, of course, equilibrium constants for adduct
formation and dissociation. From the temperature dependence
of the equilibrium constant, we obtain an adduct bond
strength that agrees well with the best available theoretical
value.20
Experimental technique
Chlorine atom kinetics as a function of temperature and
pressure of N2 bath gas were studied using the laser flash
photolysis (LFP)-resonance fluorescence (RF) technique. The
LFP-RF apparatus was similar to those we have employed in
several previous studies of chlorine atom kinetics.21–25
A
schematic diagram of the apparatus is published elsewhere.26
Important features of the apparatus and experimental details
that are specific to this study are described below.
A jacketed Pyrex reaction cell with an internal volume of
approximately 150 cm3 was used in all experiments. The cell
was maintained at a constant temperature by circulating
ethylene glycol (for T Z 299 K) or a 2 : 1 ethanol–methanol
mixture (for T o 299 K) from a thermostated bath through
the outer jacket. A copper–constantan thermocouple was
inserted into the reaction zone through a vacuum seal, thus
allowing measurement of the gas temperature under the pre-
cise pressure and flow rate conditions of each kinetics experi-
ment. The temperature variation in the reaction volume, i.e.,
the volume from which resonance fluorescence could be
detected, was less than 2 K at all temperatures employed in
this study.
In order to avoid accumulation of photochemically gener-
ated reactive species, all experiments were carried out under
‘‘slow flow’’ conditions. The linear flow rate through the
reactor was B3 cm sꢂ1 while the laser repetition rate was in
the range 2–10 Hz. Since the direction of flow was perpendi-
cular to the photolysis laser beam, no volume element of the
reaction mixture was subjected to more than a few laser shots.
Molecular chlorine, pyridine and SF6 were introduced into the
reaction cell from 12 L Pyrex bulbs containing dilute mixtures
ꢃc
This journal is the Owner Societies 2007
4384 | Phys. Chem. Chem. Phys., 2007, 9, 4383–4394

in N2, while N2 was flowed directly from its high pressure
liquid nitrogen tank. All gas flows were controlled by cali-
brated mass flow meters. The gas mixture and additional N2
were pre-mixed before entering the reaction cell except at low
temperatures, where Cl2 was introduced through a glass tube
that could be inserted into the cell just upstream from the
reaction volume, thus preventing loss of Cl2 to the cold cell
walls. Concentrations of each component other than pyridine
in the reaction mixture were determined from the correspond-
ing bulb fractions in conjunction with measured mass flow
rates and the total pressure. The bulb fraction of Cl2 was
measured frequently by UV photometry at 366 nm using a
mercury penray lamp as the light source. Concentrations of
pyridine and pyridine-d5 were measured in situ in the slow flow
system by UV photometry at 253.7 nm using a mercury penray
lamp as the light source and a 50.8 cm long absorption cell
positioned either upstream or downstream from the reaction
cell. The absorption cross sections used to convert measured
pyridine absorbances to concentrations, in units of 10ꢂ18 cm2,
were 5.50 for C5H5N (g) and 6.06 for C5D5N (g); these cross
sections were measured during the course of this study.
The pure gases used in this study had the following stated
minimum purities: N2 99.999%; Cl2 99.9%; SF6 99.9%. For
Cl2, the stated purity refers to the liquid phase in the high-
pressure gas cylinder. The liquid sample of C5H5N had a
stated purity of 99%, while the isotopic purity of C5D5N was
99.5%. C5H5N and C5D5N were each transferred into vials
fitted with high-vacuum stopcocks and then degassed repeat-
edly at 77 K before being diluted with N2 and stored in 12 L
Pyrex bulbs.
The bimolecular rate coefficients of interest, k1(P, T), are
determined from the slopes of k0 vs. [pyridine] plots obtained
at constant T and P. However, the observed Cl temporal
profiles are not perfectly exponential due to Cl recycling from
(i) Cl2 reaction with the pyridinyl radical and (ii) dissociation
of C5H5N–Cl. Because the dominant observed reaction path-
way varies as a function of temperature, the reaction respon-
sible for a majority of Cl regeneration also varies as a function
of temperature. Hence, the following discussion of the ob-
served kinetics of reaction (1) considers the high- and low-
temperature regimes separately.
Kinetics at T Z 299 K
Most of the experiments at a temperature of 299 K and above
were carried out under conditions of relatively low Cl2 con-
centrations, where well-behaved pseudo-first-order kinetics
were usually observed over the first 1.5–2 1/e decay times.
Typical data are shown in Fig. 1 and 2, and measured
bimolecular rate coefficients, k1(P, T), are summarized in
Table 1. In this relatively high temperature regime, k1 is found
to be independent of pressure for both C5H5N and C5D5N
reactants. A single exponential fit of the uncorrected rate
coefficients k1 for Cl + C5H5N vs. 1/T gives the following
Arrhenius expression for 299 K r T r 435 K:
ku1ncorrðTÞ ¼ð3:45 ꢀ 0:41Þ ꢁ 10ꢂ11
ꢁ exp½ꢂð1630 ꢀ 40Þ=Tꢄ cm3 moleculeꢂ1 sꢂ1
Errors in the above expression are 2s and represent precision
only. The rate coefficients ku1ncorr and, therefore, the above
Results and discussion
Near UV photolysis of Cl2 can produce atomic chlorine in its
2
ground state, P3/2, and also in its spin–orbit excited state,
2P1/2. However, it is well-established that for 355 nm photo-
2
lysis the fraction of atoms produced in the P1/2 state is very
small, i.e., less than 0.01.28,29 The rate coefficient for deactiva-
tion of Cl(2P1/2) by N2 is 5.0 ꢁ 10ꢂ15 cm3 moleculeꢂ1 sꢂ1 30
,
suggesting equilibration of the spin–orbit state populations
should occur rapidly on the time scale for chemical reaction
under a majority of the experimental conditions employed. As
a further check on the assumption of spin state equilibration, a
few experiments were carried out with SF6, a very efficient
Cl(2P1/2) quencher,31,32 added to the reaction mixture. As
expected, this variation in experimental conditions had no
effect on observed kinetics.
All experiments were carried out under pseudo-first order
conditions with pyridine in large excess over Cl atoms. Hence,
in the absence of side reactions that remove or produce Cl
atoms, the Cl temporal profile following the laser flash would
be described by the relationship
Fig. 1 Typical Cl atom temporal profiles observed at T Z 299 K
under the conditions of relatively low [Cl2]. Experimental conditions:
T = 320 K; P = 25 Torr; [Cl2] = 1.2 ꢁ 1013 molecules cmꢂ3; [Cl]0 =
1.0 ꢁ 1011 atoms cmꢂ3; [pyridine] in units of 1015 molecules cmꢂ3
=
(a) 0, (b) 1.33, (c) 2.76, (d) 5.90. Lines are obtained from linear least-
squares analyses of data for the first 1.1 e-folding times of decay and
give the following pseudo-first order decay rates in units of sꢂ1: (a) 47,
(b) 369, (c) 664, (d) 1290. For the sake of clarity, traces (b) and (c) are
scaled by factors of 2.0 and 2.5, respectively. The number of e-folding
times over which the data were fit was determined by visual inspection
of curvature in the ln signal vs. time plots; it was smaller in the data
shown here than in a majority of experiments because [Cl2] was
relatively high.
lnf½Clꢄ0=½Clꢄg ¼ ðk1 ½pyridineꢄ þ k3Þt ¼ k0t
ðIÞ
where k3 is the rate coefficient for the following reaction
Cl ! first order loss by diffusion from the detector field of view
and=or reaction with background impurities:
ð3Þ
ꢃc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4383–4394 | 4385

Fig. 3 Arrhenius plot for the reaction Cl + C5H5N - HCl +
o-C5H4N. The open symbols represent data obtained at a total
pressure of 25 Torr while the filled symbol represents data obtained
at a total pressure of 250 Torr. The data plotted are corrected for
secondary chemistry as described in the text. The solid line is obtained
from a least squares analysis which yields the corrected Arrhenius
parameters given in the text. Error bars on data points are 2s and
represent precision only.
Fig. 2 Plots of k0, the pseudo-first order Cl atom decay rate, versus
[pyridine] at four temperatures for data obtained employing relatively
low [Cl2]. All data shown were obtained at a pressure of 25 Torr. The
solid lines are obtained from linear least-square analyses, and the
resulting bimolecular rate coefficients, i.e., the slopes of the plots, are
listed in Table 1.
Arrhenius expression, are subject to small corrections for Cl
regeneration; these corrections are discussed in detail in a
later section of the paper. For the sake of clarity, only the
corrected rate coefficients are plotted in the Arrhenius plot
shown in Fig. 3.
some experiments were carried out under conditions of rela-
tively high [Cl2], i.e., 1.5–20 times larger than employed in
the experiments used to obtain values for ku1ncorr. The data
obtained in these ‘‘high Cl2’’ experiments were analyzed
assuming a reaction scheme that includes reactions (1), (3),
(4) and (5)
Comparison of the measured values of k1 for C5H5N and
C5D5N reactants (see Table 1) shows a significant H/D kinetic
isotope effect, i.e., kH1 /k1D 4 3. Both the observed kinetic isotope
effect and the temperature and pressure dependencies of k1
strongly support hydrogen abstraction as the dominant observed
reaction pathway in this relatively high temperature regime.
In order to investigate the impact on observed kinetics of Cl
regeneration via reaction of the H-abstraction product
(pyridinyl radical) with Cl2,
C5H4N !removal by processes that do not produce
ð5Þ
chlorine atoms
The rate equations for the expanded reaction scheme can be
solved analytically, and yield a double exponential functional
form for the Cl decay
C5H4N þ Cl2 ! C5H4ClN þ Cl;
ð4Þ
St=S0 ¼ ½ðQ þ l1Þ expðl1tÞ ꢂ ðQ þ l2Þ expðl2tÞꢄ=ðl1 ꢂ l2Þ ðIIÞ
Table 1 Summary of kinetic data for the reaction of Cl with pyridine at T Z 299 K
Concentration/1011 molecules cmꢂ3
No.
k1a ꢀ 2scd/10ꢂ13 cm3 moleculeꢂ1 sꢂ1
[Pyridine]b
[Cl2]
[Cl]0
A
B
C
T/K P/Torr
of exptsa
k0max/sꢂ1
C5H5N
299
300
320
335
354
387 250
392
413
25
25
25
25
25
5
6
6
7
6
6
7
5
6
9
44 600
49 500
59 000
43 700
39 300
45 000
52 000
37 200
36 800
15 700
107
79
120
64
180
67
83
63
0.9
0.9
1.0
0.6
1.8
0.9
0.7
0.4
0.7
0.7
710
810
1.44 ꢀ 0.14
1.49 ꢀ 0.28
2.10 ꢀ 0.10
2.67 ꢀ 0.20
3.31 ꢀ 0.13
5.22 ꢀ 0.16
5.52 ꢀ 0.23
6.33 ꢀ 0.16
7.93 ꢀ 0.26
8.10 ꢀ 0.15
1.81 ꢀ 0.09
1.88 ꢀ 0.23
2.31 ꢀ 0.09
2.76 ꢀ 0.19
3.35 ꢀ 0.13
5.22 ꢀ 0.16
5.52 ꢀ 0.23
6.33 ꢀ 0.16
7.93 ꢀ 0.26
8.10 ꢀ 0.15
1.91 ꢀ 0.15
2.10 ꢀ 0.29
2.44 ꢀ 0.12
2.99 ꢀ 0.18
3.73 ꢀ 0.19
5.44 ꢀ 0.14
5.85 ꢀ 0.39
6.84 ꢀ 0.07
8.23 ꢀ 0.17
8.66 ꢀ 0.41
1280
1230
1380
2380
2970
2410
2970
1410
25
25
25
25
434
435e
65
97
C5D5N
339
393
433
25
25
25
6
7
9
51 600
56 700
37 900
64
56
62
0.5
0.4
0.8
630
1050
1120
1.07 ꢀ 0.14
1.66 ꢀ 0.12
2.67 ꢀ 0.17
1.11 ꢀ 0.13
1.66 ꢀ 0.12
2.67 ꢀ 0.17
1.17 ꢀ 0.17
1.75 ꢀ 0.14
2.82 ꢀ 0.22
a
b
Expt ꢅ measurement of a single pseudo-first-order Cl decay rate. The concentration of pyridine was measured in situ in the slow flow reactor by
c
UV photometry. Uncertainties represent precision only. A: uncorrected data; B: corrected for equilibration of Cl with C5H5N–Cl; C: corrected
for both equilibration of Cl with C5H5N–Cl and Cl regeneration via the o-C5H4N + Cl2 reaction. B1.7 ꢁ 1015 molecule cmꢂ3 SF6 was added in
d
e
some runs to promote more rapid equilibration of the Cl spin–orbit states.
ꢃc
This journal is the Owner Societies 2007
4386 | Phys. Chem. Chem. Phys., 2007, 9, 4383–4394

where St and S0 are the resonance fluorescence signal levels at
times t and 0, and
averages of results at 340, 392 and 433 K:
k4 ¼ ð1:3 ꢀ 0:4Þ ꢁ 10ꢂ11 cm3 moleculeꢂ1 sꢂ1
Q ¼ k4½Cl2ꢄ þ k5
Q þ k3 þ k1 ½pyridineꢄ ¼ ꢂðl1 þ l2Þ
k3Q þ k5k1 ½pyridineꢄ ¼ l1l2
ðIIIÞ
ðIVÞ
ðVÞ
k5 ¼ 650 ꢀ 290 sꢂ1
Uncertainties in the above rate coefficients are 2s and repre-
sent precision only. The rather large background decay rate of
pyridinyl (B650 sꢂ1 compared to B50 sꢂ1 for Cl) is probably
due, in large part, to the reaction with O2 that is present either
as an impurity in the bath gas or due to small leaks in the slow
flow system:
Fig. 4 shows typical Cl temporal profiles observed under
conditions of high [Cl2]. The kinetic data were fit to the
predicted double exponential eqn (II) to obtain values for
S0, Q, l1 and l2. The relevant rate coefficients k1, k4 and k5 can
be derived as follows:
C5H4N þ O2 þ M ! C5H4NO2 þ M
ð6Þ
To the best of our knowledge, no kinetic data for k6 have been
reported in the literature. However, if we assume that the rate
coefficient for reaction (6) is equal to the phenyl + O2 rate
coefficient, which is known to be 1.7 ꢁ 10ꢂ11 cm3 moleculeꢂ1 sꢂ1
at room temperature and 20–80 Torr,34 a residual O2 partial
pressure of B1.5 mTorr in the slow flow system would com-
pletely account for the observed background C5H4N removal
rate.
k1 ¼ ꢂðQ þ k3 þ l1 þ l2Þ=½pyridineꢄ
k5 ¼ ꢂðk3Q ꢂ l1l2Þ=ðQ þ k3 þ l1 þ l2Þ
k4 ¼ ðQ ꢂ k5Þ=½Cl2ꢄ
ðVIÞ
ðVIIÞ
ðVIIIÞ
The results obtained at three different temperatures are sum-
marized in Table 2. Comparison of Tables 1 and 2 shows that
values for k1 obtained from simple averages of all the runs at
the same temperature are in good agreement with the values
obtained from the low Cl2 experiments. Reported values for k5
were obtained from the arithmetic mean of results from all
experiments at a given temperature, while values for k4 were
evaluated from the slopes of plots of Q ꢂ k5 vs. [Cl2] (Fig. 5);
this approach gives improved precision and accuracy for
determination of k4 compared to using arithmetic means of
results obtained from individual experiments.33 Because the
values obtained for k4 and k5 from the data analysis are
coupled, the reported precisions are conservatively taken to
be the square roots of the quadratic sums of the k4 and k5
precisions. The temperature dependencies of k4 and k5 are not
significant and are difficult to evaluate quantitatively due to
the large uncertainties involved; hence, we choose to report
temperature-independent rate coefficients obtained from
Although this study does not provide information that
allows the hydrogen abstraction product to be identified,
information in the literature as well as new electronic structure
calculations strongly suggest that the dominant H-abstraction
product in the temperature range of this study is the
o-pyridinyl radical, i.e., the radical site is a carbon atom
adjacent to the nitrogen atom
Cl þ C5H5N ! o-C5H4N þ HCl
ð1aÞ
Enthalpy changes associated with H-abstraction from pyridine
by Cl at the ortho, meta, and para positions have been
evaluated and are summarized in Table 3. To obtain the
results in Table 3, we employ recent determinations of the
C–H bond dissociation enthalpies in pyridine from electronic
structure calculations at the B3LYP/6-311+G(2d,p) level of
theory35 as well as the well-known standard enthalpies of
formation of Cl and HCl36 The enthalpy changes in Table 3
are likely to be correct to within ꢀ8 kJ molꢂ1, and clearly
demonstrate that only o-pyridinyl is energetically accessible as
a hydrogen abstraction product over the temperature range of
our study. Similar conclusions regarding the identity of the
H-abstraction product have been reported based on theoreti-
cal and experimental studies of pyridine chlorination
reactions,37 and also theoretical studies of the H, O, and
OH + pyridine H-abstraction reactions.38
Kinetics at 216 K r T r 270 K
Over the temperature range 216–270 K, chlorine atom decay
rates observed at short times following the laser flash were
considerably faster than expected on the basis of extrapolation
of the high-temperature results (see above). Also in contrast to
the high-temperature results, for a given temperature and
pyridine concentration it was found that the rate of Cl loss
increased with increasing N2 pressure. These observations
suggest that the dominant observed reaction pathway in the
low-temperature regime (T r 270 K) is addition of Cl to
pyridine. As will be discussed in more detail below, theoretical
evidence in the literature18,20 supports the hypothesis that the
Fig. 4 Typical Cl atom temporal profiles observed at T Z 299 K
under conditions of relatively high [Cl2]. Experimental conditions:
T = 392 K; P = 25 Torr; [pyridine] = 3.9 ꢁ 1015 molecules cmꢂ3
;
[Cl]0 = 1.3 ꢁ 1011 atoms cmꢂ3; [Cl2] in units of 1013 molecules cmꢂ3
=
(a) 12.1, (b) 6.57, (c) 1.74. Solid curves are obtained from non-linear
least-squares fits to eqn (II) and give the best fit parameters shown in
Table 2. Trace (a) is scaled by a factor of 2.5 for clarity.
ꢃc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4383–4394 | 4387

Table 2 Results from experiments employing relatively high Cl2 concentrations
[Pyridine]a/1011 [Cl2]/1011
[Cl]0/1011
k1a/10ꢂ13 cm3
k4/10ꢂ13 cm3
T/K P/Torr Q/sꢂ1 ꢂl1/sꢂ1 ꢂl2/sꢂ1 molecules cmꢂ3 molecules cmꢂ3 molecules cmꢂ3 moleculeꢂ1 sꢂ1 moleculeꢂ1 sꢂ1 k5/sꢂ1
340
25
992
691
482
395
288
307
246
1770
2180
3050
3860
1440
2210
37 300
36 700
37 400
37 000
18 800
18 400
149
436
886
1580
425
853
1.2
0.85
1.0
3.80
3.17
3.24
2.99
3.31
3.33
3.31
0.54
112
140
145
135
107
125
134
36
826
833
893
852
614
724
790
211
1440
2180
2980
1070
1790
0.72
0.78
1.2
Meanb
2sc
392d 25d
709
708
485
396
390
339
292
273
243
213
2500
2490
2890
3330
3480
3930
2130
3300
38 500
37 900
38 400
38 800
39 300
39 300
19 200
19 600
108
174
347
657
1190
1210
650
1.1
1.3
0.71
1.2
1.5
1.5
5.74
5.58
5.55
5.23
4.22
4.65
4.94
5.09
5.13
1.03
169
151
171
163
128
149
140
157
147
23
528
445
497
506
532
505
454
546
501
79
1090
1570
2050
2310
1360
2450
0.67
1.1
1210
Meanb
2sc
433
25
768
789
571
566
531
580
403
335
529
2970
3080
3590
3590
2210
2820
4010
35 900
34 800
35 000
35 100
18 400
17 200
35 600
158
160
493
496
558
1110
1120
1.1
1.1
1.4
0.91
1.0
7.52
8.01
7.97
7.86
7.53
7.47
7.26
7.66
0.57
102
115
123
125
104
105
100
106
23
607
606
649
718
583
634
766
652
143
1260
1340
1160
1800
1880
1.9
3.9
Meanb
2sc
a
b
The concentrations of pyridine were measured in situ in the slow flow reactor by UV photometry. For k1a and k5, arithmetic means were
c
d
determined; values for k4 were obtained from slopes of (Q ꢂ k5) vs. [Cl2] plots (see text). Precisions of k4 and k5 are coupled (see text). Temporal
profiles of Cl atoms and best fit temporal profiles are shown in Fig. 4.
adduct is formed by addition of Cl to the nitrogen atom in
found to be independent of both Cl2 and pyridine concentra-
tions and, at a given temperature, was found to increase with
increasing N2 pressure. These observations strongly suggest
that Cl regeneration results from unimolecular decomposition
of the product of reaction (1b):
pyridine (rather than by addition to the ring p-system):
Cl þ C5H5N þ M ! C5H5NꢂCl þ M
ð1bÞ
As observed in the high-temperature experiments (particularly
those at high [Cl2]), Cl temporal profiles were non-exponential
in a manner that is consistent with Cl regeneration via a
secondary reaction. Unlike the high-temperature experiments,
however, the rate of Cl regeneration was much too fast to be
attributable to a product + Cl2 reaction, was experimentally
C5H5NꢂCl þ M ! C5H5N þ Cl þ M
ð--1bÞ
Within experimental uncertainty, observed kinetics at T r 270
K were unaffected by substitution of C5D5N for C5H5N in the
reaction mixture, as would be expected if the dominant reac-
tions involved formation and dissociation of an N–Cl bond.
The relevant kinetic scheme for analysis of the low-tem-
perature data includes reactions (1a), (1b), (ꢂ1b), (3), and (7).
C5H5NꢂCl
ð7Þ
! loss by processes that do not regenerate Cl
Table 3 Bond dissociation enthalpies and the enthalpy changes for
reaction 1a for different H abstraction products
BDEa/kJ molꢂ1
DrH/kJ molꢂ1
Radical product
0 K
298 K
0 K
298 K
Fig. 5 Plot of Q ꢂ k5 vs. [Cl2] for the results in Table 2 at T = 392 K.
The open symbols represent data obtained from the fits shown in Fig
4. The solid line is obtained from a linear least squares analysis and its
slope gives the bimolecular rate coefficient k4 = 1.47 ꢁ 10ꢂ11 cm3
2-Pyridinyl
3-Pyridinyl
4-Pyridinyl
432
458
452
438
464
458
4
30
24
6
32
26
a
Calculated at the B3LYP/6-311+G(2d,p) level of theory.
moleculeꢂ1 sꢂ1
.
ꢃc
This journal is the Owner Societies 2007
4388 | Phys. Chem. Chem. Phys., 2007, 9, 4383–4394

The rate equations can be solved analytically, and predict a
double exponential Cl decay, i.e., the same functional form as
the one that describes the high-temperature Cl temporal
profiles in the presence of non-negligible levels of Cl2. Hence,
the temporal evolution of St/S0 is described by eqn (II). In the
case of interest here, i.e., reversible adduct formation, the fit
parameters Q, l1, and l2 are related to the elementary rate
coefficients as follows:
Experimental conditions and results of the low-temperature
equilibration kinetics experiments are summarized in Table 4.
The values of the equilibrium constant KP given in Table 4 are
derived from the following relationship
KP ¼ k1b=ðkꢂ1bRTÞ ¼ KC=ðRTÞ
ðXVÞ
The results at 245 ꢀ 1 K show that k1b and kꢂ1b have strong
pressure dependencies while KP has none, as expected for an
addition/dissociation reaction; this result improves confidence
that the non-linear least squares fitting procedure is relatively
free of systematic errors in the extracted rate coefficients. Since
Q ¼ kꢂ1b þ k7
ðIXÞ
ðXÞ
Q þ k3 þ ðk1a þ k1bÞ ½pyridineꢄ ¼ ꢂðl1 þ l2Þ
Qðk3 þ k1a ½pyridineꢄÞ þ k7k1b½pyridineꢄ ¼ l1l2
the precision of multiple determinations of k1b and k
are
ꢂ1b
ðXIÞ
quite good, we estimate that the absolute accuracies of
reported k1b, kꢂ1b and KP values are ꢀ30% over the full range
of temperature and pressure spanned by the results given in
Table 4.
Observed Cl temporal profiles (Fig. 6) were fit to the double
exponential eqn (II) to obtain values for the fit parameters.
Rearrangement of eqn (IX)–(XI) gives relationships for the
rate coefficients k1b, kꢂ1b, and k7:
A plot of ln KP vs. 1/T, i.e., a van’t Hoff plot, is shown in
Fig. 7. A linear least-squares analysis of the data gives the
following expression:
k1b ¼ ꢂ ðQ þ k3 þ k1a½pyridineꢄ þ l1 þ l2Þ=
ðXIIÞ
½pyridineꢄ
ln KP ¼ ꢂð11:87 ꢀ 0:96Þ þ ð5700 ꢀ 130Þ=T
k7 ¼fl1l2 ꢂ Qðk3 þ k1a½pyridineꢄÞg=
Errors in the above expression are 2s and represent precision
only. As discussed below, the slope of the van’t Hoff plot is
related to DrH for reaction (1b) while the intercept is related
to DrS.
ðXIIIÞ
ðk1b½pyridineꢄÞ
k
¼ Q ꢂ k7
ðXIVÞ
ꢂ1b
Secondary chemistry and corrections to rate coefficients
The background Cl atom loss rate (k3) was directly measured
at each temperature and pressure, while temperature-depen-
dent rate coefficients for the H-abstraction pathway (k1a) were
obtained by extrapolation of the (corrected as described in a
later section) high-temperature kinetic data assuming Arrhe-
nius behavior, i.e., a linear ln k1a vs. 1/T dependence from
299 K down to 216 K.
The radical concentrations employed in this study were suffi-
ciently low that radical–radical side reactions exerted a negli-
gible influence on observed chlorine atom kinetics. Hence, the
only important secondary chemistry complications result from
regeneration of chlorine atoms by reactions (ꢂ1b) and/or (4).
As described below, these complications can be treated differ-
ently in the low and high temperature regimes.
Under the experimental conditions employed in the equili-
bration kinetics experiments at T r 270 K, the branching ratio
for H-abstraction, i.e., k1a/(k1a + k1b), was less than 0.02 in all
experiments, and the first order rate of decomposition of
C5H5N–Cl was faster than the pseudo-first order rate of
reaction of C5H4N with Cl2 in all experiments. Hence, the
only important source of regenerated Cl was reaction (ꢂ1b),
i.e., the influence of reaction (4) on observed Cl kinetics was
negligible and can be ignored without introducing significant
systematic error into reported kinetic and thermochemical
parameters.
At temperatures of 299 K and above, reaction (ꢂ1b) was
sufficiently fast that the kinetics of equilibration between Cl
and C5H5N–Cl could not be directly observed with the
experimental time resolution of a few microseconds. However,
over the range of pyridine concentrations employed to follow
the kinetics of the H-abstraction reaction (at temperatures
around 299 K), C5H5N–Cl represented a significant Cl reser-
voir, and its influence on observed kinetics had to be taken
into account. Furthermore, since Cl decayed via the relatively
slow H-abstraction reaction, regeneration of Cl via reaction (4)
occurred at a non-negligible rate even though very low Cl2
concentrations were employed.
Fig. 6 Typical Cl atom temporal profiles observed at 216 K r T r
270 K. Experimental conditions: T = 252 K, P = 25 Torr; [Cl2] = 1.4
ꢁ 1013 molecules cmꢂ3; [Cl]0 = (1.9–2.7) ꢁ 1011 atoms cmꢂ3
;
[pyridine] in units of 1015 molecules cmꢂ3 = (a) 0.636, (b) 1.04, (c)
2.72. Solid lines are obtained from non-linear least-square fits to eqn
(II) and give the best fit parameters in Table 4. Trace (a) is scaled by a
factor of 1.5 for clarity. The upper plot shows trace (b) with the signal
plotted on a linear (rather than a logarithmic) scale.
ꢃc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4383–4394 | 4389

ꢃc
This journal is the Owner Societies 2007
4390 | Phys. Chem. Chem. Phys., 2007, 9, 4383–4394

tial (due to reaction (4)), leading to small underestimates of k0,
especially at 299 K where experimental conditions were such
that the ratio of the rate of Cl decay to the rate of Cl
regeneration was smaller than at other temperatures. To
correct the pseudo-first-order decay data, we increased the
initial k1a by 10% at room temperature and 5% at other
temperatures as the first guess for the corrected k1a values, and
carried out the simulations using these elevated rate coeffi-
cients and the same conditions as in the low [Cl2] experiments
with and without inclusion of Cl regeneration via reaction (4)
in the simulated mechanism; the simulated decay rates were
obtained from linear least-squares analyses of the simulated ln
[Cl]t vs. time data for the same fitting time as in the experi-
mental runs, i.e., typically 1.5–2.0 e-folding times but as low as
1.0 and as high as 2.5 e-folding times depending on experi-
mental conditions. Ratios of the simulated pseudo-first-order
decay rate with reaction (4) included and excluded from the
mechanism were adopted to correct the experimental decay
data, thus yielding the final values of k1a, which are labeled C
in Table 1. Because the values obtained for the corrected rate
coefficients are very close to our initial guesses, a second
iteration of the process was not necessary.
Fig. 7 van’t Hoff plot for the reaction Cl + C5X5N 2 C5X5N–Cl
(X ꢅ H or D). The solid line is obtained from a least squares analysis
of ln KP vs. 1/T data for C5H5N and gives the second-law thermo-
chemical parameters for the reaction (see Table 6). Error bars on data
points are 2s and represent precision only.
The first step in correcting the data at T 4 299 K for
secondary chemistry complications involved accounting for
the fraction of Cl atoms that were tied up as C5H5N–Cl.
Because of its stability, C5H5N–Cl cannot react with pyridine
by H-abstraction at the relatively low temperatures employed
in this study. Under the assumptions that (i) C5H5N–Cl is an
unreactive reservoir for Cl and (ii) equilibrium is maintained
between Cl and C5H5N–Cl throughout the course of the
reaction, it can be shown that the correct pseudo-first order
rate of reaction of Cl with pyridine, k00(corr), is related to the
observed pseudo first order rate of reaction, k00(obs), as
follows:
The corrected kinetic data are plotted in Arrhenius form in
Fig. 3. The corrected Arrhenius expression for data obtained
over the temperature range 299–435 K is
kc1orrðTÞ ¼k1aðTÞ ¼ ð2:08 ꢀ 0:47Þ ꢁ 10ꢂ11
ꢁ exp½ꢂð1410 ꢀ 80Þ=Tꢄ cm3 moleculeꢂ1 sꢂ1
Errors in the above expression are 2s and represent precision
only.
As mentioned above, radical–radical reactions such as
reactions (8) and (9) are not expected to contribute signifi-
cantly to Cl atom removal because of the low radical concen-
trations employed.
k00ðcorrÞ ¼ ð1 þ aÞk00ðobsÞ
ðXVIÞ
In eqn (XVI), a is the equilibrium adduct-to-Cl concentration
ratio and k00 is related to the measured pseudo-first order decay
rate k0 (see eqn (I)) by eqn (XVII):
Cl þ C5H4N ! products
ð8Þ
ð9Þ
k00 ¼ k0 ꢂ k3
ðXVIIÞ
Cl þ C5H5NꢂCl ! products
Even if reactions (8) and (9) occurred on every reactant
encounter, these processes would not be fast enough to
compete with reaction (1) as a Cl removal pathway under
the experimental conditions employed.
a ¼ ½C5H5N ꢂ Clꢄeq=½Clꢄeq ¼ KC½C5H5Nꢄ
¼ KP½C5H5Nꢄ=RT
ðXVIIIÞ
For every experimental run at T Z 299 K, the parameter a
was evaluated using the measured pyridine concentration in
conjunction with a KP value obtained by extrapolation of the
van’t Hoff expression (eqn (XV)) to the appropriate tempera-
ture. Comparison of the k1a values labeled A and B in Table 1
shows that correction for the presence of the adduct as a
reservoir for Cl increased the measured k1a by about 25% at
299 K. The magnitude of the correction drops with increasing
temperature; it is B1% at 354 K and is negligible at 387 K and
above.
Adduct thermochemistry
A second-law analysis was employed to evaluate the enthalpy
and entropy changes associated with reaction (1b) using the
van’t Hoff plot (Fig. 7). Since
ln KP ¼ ðDrS=RÞ ꢂ ðDrH=RTÞ;
ðXIXÞ
the enthalpy change is obtained from the slope of the van’t
Hoff plot while the entropy change is obtained from the
intercept. At 240 K, the midpoint of the experimental 1/T
range, this ‘‘second-law analysis’’ gives the results DrH =
The corrected rate coefficients labeled B in Table 1 were
further corrected to account for Cl regeneration via the C5H4N
+ Cl2 reaction, i.e., reaction (4). Preliminary kinetic simula-
tions that included Cl loss via reactions (1) and (3) and Cl
regeneration via reaction (4) demonstrated that, over the range
of Cl2 and C5H5N concentrations employed in our studies, the
observed Cl temporal profiles should be slightly non-exponen-
ꢂ47.2 ꢀ 1.9 kJ molꢂ1 and DrS = ꢂ98.7 ꢀ 8.0 J molꢂ1 Kꢂ1
where the errors are 2s and represent precision only.
,
In addition to the second-law analysis, we have carried out a
third-law analysis, where the value of KP at 240 K from the
van’t Hoff analysis, (1.35 ꢀ 0.40) ꢁ 105 atmꢂ1, has been
ꢃc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4383–4394 | 4391

employed in conjunction with a calculated entropy change to
determine DrH(240 K) for reaction (1b) (using eqn (XIX)).
To evaluate DrS for reaction (1b), absolute entropies as a
function of temperature were obtained from the JANAF
tables36 for Cl, and calculated using ab initio vibrational
frequencies and moments of inertia for C5H5N and
C5H5N–Cl. Relevant parameters used in the calculations of
absolute entropies and heat capacity corrections are summar-
ized in Table 5. At 240 K, our analysis gives the results DrS =
(1b), at the G2(MP2, SVP) level of theory, and obtained a
result that is in excellent agreement with the experimental
value reported in this study (see Table 6).
As mentioned in the Introduction, this is the first gas-phase
kinetic studies of the Cl + pyridine reaction to be reported.
However, a number of studies reported in the literature are
relevant to this work. Hadad and co-workers38 have calculated
rate constants for reactions of H, O, and OH radicals with
pyridine at 298–2000 K using a combination of density func-
tional theory and transition state theory. They found that for
the reactions of pyridine with O and OH, the H-atom abstrac-
tion pathway is dominant over addition to the carbon atoms at
all temperatures investigated, while for the reaction of pyridine
with H, addition to the carbon atoms is preferred at low
temperatures. Without exploring reactions at the nitrogen
atom site, they concluded that H-atom abstraction preferen-
tially occurs at ortho positions and that addition preferentially
occurs at meta positions. Turecek and co-workers41 studied
the N-hydroxypyridinyl radical (C5H5N–OH) and suggested
that this radical was unlikely to be generated via OH addition
to the pyridine nitrogen because of a large energy barrier of
49 kJ molꢂ1 for the process. ESR42 and pulse radiolysis43
investigations of the liquid-phase reaction of OH with pyridine
also suggest that OH addition predominantly (B80%) takes
place at the meta carbon. Hence, it appears that the mechan-
isms are different for the reactions of pyridine with OH and Cl,
particularly at low temperature where addition pathways
become more important.
ꢂ99.6 ꢀ 8.0 J molꢂ1 Kꢂ1 and DrH = ꢂ47.5 ꢀ 2.5 kJ molꢂ1
;
the uncertainties we report reflect an estimate of our imperfect
knowledge of the input data needed to calculate absolute
entropies (particularly the low-frequency C5H5N–Cl vibra-
tions) as well as the estimated uncertainty in the experimental
value for KP at 240 K. In arriving at the above uncertainty in
DrS, we assume that the frequencies of the three lowest
frequency C5H5N–Cl vibrations could differ from the values
given in Table 5 by ꢀ25 cmꢂ1. Appropriate heat capacity
corrections have been employed to obtain DrH1 values at 298
and 0 K; the results are given in Table 6. The thermochemical
parameters determined from the second- and third-law ana-
lyses are in excellent agreement, and lead to the following
reported enthalpy and entropy changes for reaction (1b) at 0
and 298 K: DrH10 = ꢂ46.7 ꢀ 3.2 kJ molꢂ1, DrH1298 = ꢂ47.2 ꢀ
2.8 kJ molꢂ1, and DrS2198 = ꢂ98.7 ꢀ 6.5 J molꢂ1 Kꢂ1
;
uncertainties are accuracy estimates at the 95% confidence
level.
The values for DrH1 obtained above can be used in con-
junction with literature values for the standard enthalpies of
formation of Cl36 and pyridine39,40 to deduce the following
values for the standard enthalpy of formation of C5H5N–Cl
(in units of kJ molꢂ1) at 298 and 0 K: DfH2198 = 216.0 ꢀ 4.1
and DfH10 = 233.4 ꢀ 4.6.
Implications for atmospheric chemistry
The gas phase rate coefficients measured in this study can be
employed to compare the potential importance of Cl as an
atmospheric sink for pyridine with contributions from two
other atmospheric species that are known to also be reactive
toward pyridine, i.e., OH and HNO3. Since kinetic data for the
OH and HNO3 reactions with pyridine are available only at
298 K, we assume this temperature to make the comparison
Comparison with previous studies
Both experimental18 and theoretical20 studies have been re-
ported where it is suggested that generation of C5H5N–Cl
involves formation of a two-center—three-electron (2c–3e)
s-bond between the chlorine atom and the nitrogen atom in
pyridine. According to McKee et al.,20 the pyridine fragment
of the N-chloropyridinyl radical (C5H5N–Cl) has a very
similar geometry to that of isolated pyridine, and the chlorine
atom is in the plane of the ring. McKee et al.20 also calculated
the Cl–N bond dissociation enthalpy, i.e., DrH1 for reaction
summarized below. Assuming an average tropospheric OH
44
concentration of 9 ꢁ 105 molecule cmꢂ3
and a rate
coefficient for OH reaction with pyridine of 5 ꢁ 10ꢂ13 cm3
moleculeꢂ1 sꢂ1 7,8 the average lifetime of pyridine toward loss
,
via reaction with OH can be estimated to be B25 days. The
rate coefficient for the pyridine + HNO3 reaction has been
shown by Atkinson et al.8 to be at least 1.5 ꢁ 10ꢂ16 cm3
Table 5 Summary of parameters used in calculations of absolute entropies and heat capacity corrections
Cl
C5H5N
1
C5H5N–Cl
2
g0
4
g1
2
882.36
De/cmꢂ1a
s
2 2
6.07, 5.85, 2.98 5.98, 1.19, 0.99
383, 419, 619, 671, 715, 762, 890, 957, 1001, 1010, 53, 88, 190, 391, 421, 634, 661, 704, 759, 881, 960,
Rot. constants/GHzb
Vib.frequencies/cmꢂ1b
1011, 1050, 1077, 1094, 1171, 1242, 1277, 1391,
1475, 1517, 1617, 1622, 3145, 3148, 3169, 3183,
3191
997, 1015, 1018, 1038, 1084, 1094, 1175, 1225, 1295,
1374, 1482, 1511, 1618, 1622, 3176, 3180, 3186, 3197,
3202
a
De ꢅ energy splitting between the lowest two electronic states; pyridine has no low- energy excited states and the C5H5N–Cl adduct is assumed to
b
have none. Calculated values at the B3LYP/6-311+G(2d,p) level of theory.
ꢃc
This journal is the Owner Societies 2007
4392 | Phys. Chem. Chem. Phys., 2007, 9, 4383–4394

Table 6 Thermochemical parameters for the reaction Cl + C5H5N 2 C5H5N–Cl
T/K
Method
ꢂDrH/kJ molꢂ1
ꢂDrS/J molꢂ1 Kꢂ1
DfH (C5H5N–Cl)a/kJ molꢂ1
240
Second lawb
Third lawb
47.2 ꢀ 1.9
47.5 ꢀ 2.5
46.5 ꢀ 2.2
46.8 ꢀ 2.9
47.5
47.1 ꢀ 1.9
47.3 ꢀ 2.5
47.2
98.7 ꢀ 8.0
99.6 ꢀ 8.0
0
Second lawb
Third lawb
233.4 ꢀ 4.6
G2(MP2, SVP)c
Second lawb
Third lawb
298
98.2 ꢀ 8.0
99.2 ꢀ 8.0
216.0 ꢀ 4.1
G2(MP2, SVP)c
a
b
Values are based on combined second and third law results. This work. Ref. 20.
c
moleculeꢂ1
s
ꢂ1. The concentration of HNO3 in the tropo-
8 R. Atkinson, E. C. Tuazon, T. J. Wallington, S. M. Aschmann, J.
Arey, A. M. Winer and J. N. Pitts, Jr, Environ. Sci. Technol., 1987,
21, 64.
9 L. Y. Yeung and M. J. Elrod, J. Phys. Chem. A, 2003, 107,
4470.
sphere is highly variable, ranging from o108 molecules cmꢂ3
to 41011 molecules cmꢂ3 45 Hence, reaction with HNO3 could
.
be a dominant pyridine sink or a relatively minor one,
depending on the local HNO3 mixing ratio and the poorly
characterized rate coefficient for the pyridine + HNO3 reac-
tion. The contribution of Cl to pyridine destruction is difficult
to evaluate because (i) atmospheric Cl concentrations are
highly variable in space and time and (ii) the atmospheric fate
of C5H5N–Cl is uncertain. Assuming an average Cl concen-
tration of 4 ꢁ 104 atoms cmꢂ3, which appears to be appro-
priate for the marine boundary layer (MBL),12 a value of
5 ꢁ 10ꢂ11 cm3 moleculeꢂ1 sꢂ1 for k1b at atmospheric pressure
(based on a rough extrapolation of the data in Table 4), and no
regeneration of pyridine from C5H5N–Cl, one obtains a life-
time of B6 days for loss of pyridine by reaction with Cl, i.e., a
factor of 4 shorter than the estimated lifetime of pyridine
toward reaction with OH. Since (i) Cl levels in the MBL are
thought to be the highest encountered in the troposphere and
(ii) it is likely that a significant fraction of C5H5N–Cl loss in
the atmosphere results in regeneration of pyridine, the above
scenario only represents the maximum possible contribution
from the Cl + pyridine reaction. However, it is clear from the
above analysis that further investigation of C5H5N–Cl chem-
istry under tropospheric conditions would be desirable.
10 A. A. P. Pszenny, W. C. Keene, D. J. Jacob, S. Fan, J. R. Maben,
M. P. Zetwo, M. Springer-Young and J. N. Galloway, Geophys.
Res. Lett., 1993, 20, 699.
11 H. Boudries and J. W. Bottenheim, Geophys. Res. Lett., 2000, 27,
517.
12 O. W. Wingenter, B. C. Sive, N. J. Blake, D. R. Blake and F. S.
Rowland, J. Geophys. Res., [Atmos.], 2005, 110, D20308, DOI:
10.1029/2005JD005875.
13 G. A. Russell, J. Am. Chem. Soc., 1957, 79, 2977.
14 C. Walling and M. F. Mayahi, J. Am. Chem. Soc., 1959, 81, 1485.
15 P. S. Skell, H. N. Baxter III and C. K. Taylor, J. Am. Chem. Soc.,
1983, 105, 120.
16 N. J. Bunce, R. B. Joy, J. P. Landers and J. S. Nakai, J. Org.
Chem., 1987, 52, 1155.
17 R. K. Khama, B. Armstrong, H. Cui and J. M. Tanko, J. Am.
Chem. Soc., 1992, 114, 6003.
18 R. Breslow, M. Brandl, J. Hunger, N. Turro, K. Casidy, K. Krogh-
Jespersen and J. D. Westbrook, J. Am. Chem. Soc., 1987, 109,
7204.
19 A. Abu-Raqabah and M. C. R. Symons, J. Am. Chem. Soc., 1990,
112, 8614.
20 M. L. McKee, A. Nicolaides and L. Radom, J. Am. Chem. Soc.,
1996, 118, 10571.
21 R. E. Stickel, J. M. Nicovich, S. Wang, Z. Zhao and P. H. Wine, J.
Phys. Chem., 1992, 96, 9875.
22 J. M. Nicovich, S. Wang and P. H. Wine, Int. J. Chem. Kinet.,
1995, 27, 359.
23 Y. V. Ayhens, J. M. Nicovich, M. L. McKee and P. H. Wine, J.
Phys. Chem. A, 1997, 101, 9382.
24 J. J. Orlando, C. A. Piety, J. M. Nicovich, M. L. McKee and P. H.
Wine, J. Phys. Chem. A, 2005, 109, 6659.
Acknowledgements
25 J. M. Nicovich, S. Parthasarathy, F. D. Pope, A. T. Pegus, M. L.
McKee and P. H. Wine, J. Phys. Chem. A, 2006, 110, 6874.
26 B. J. Finlayson-Pitts and J. N. Pitts, Jr, Chemistry of the Upper
and Lower Atmosphere, Academic Press, San Diego, CA, 2000,
p. 146.
This research was supported by the National Science Foun-
dation—Atmospheric Chemistry Program through grant no.
ATM-0350185 and by the National Aeronautics and Space
Administration—Upper Atmosphere Research Program
through grant no. NNG06GD90G.
27 I. Yamazaki, K. Sushida and H. Baba, J. Chem. Phys., 1979, 71,
381.
28 G. E. Busch, R. T. Mahoney, R. I. Morse and K. R. Wilson, J.
Chem. Phys., 1969, 51, 449.
29 J. Park, Y. Lee and G. W. Flynn, Chem. Phys. Lett., 1991, 186,
441.
30 G. S. Tyndall, J. J. Orlando and C. S. Kegley-Owen, J. Chem. Soc.,
Faraday Trans., 1995, 91, 3055.
31 S. A. Sotnichenko, V. Ch. Bokun and A. I. Nadkhin, Chem. Phys.
Lett., 1988, 153, 560.
32 A. I. Chichinin, Chem. Phys. Lett., 1993, 209, 459.
33 J. M. Nicovich, C. J. Shackelford and P. H. Wine, J. Photochem.
Photobiol., A, 1990, 51, 141.
34 T. Yu and M. C. Lin, J. Am. Chem. Soc., 1994, 116, 9571.
35 Calculations with the hybrid DFT (a,b) method, B3LYP (c), were
made with the G03 program package (d).(a) R. G. Parr and
W. Yang, Density-Functional Theory of Atoms and Molecules,
Oxford University Press, Oxford, 1989; (b) W. Koch and
References
1 ATSDR toxicological profile, parts 4 and 5; http://www.atsdr.cdc.-
gov/toxprofiles/tp52.html#.
2 F. L. Eisele and E. W. McDaniel, J. Geophys. Res., [Atmos.], 1986,
91, 5183.
3 F. L. Eisele, J. Geophys. Res., [Atmos.], 1988, 93, 716.
4 F. L. Eisele, J. Geophys. Res., [Atmos.], 1989, 94, 6309.
5 D. J. Tanner and F. L. Eisele, J. Geophys. Res., [Atmos.], 1991, 96,
1023.
6 R. P. Turco, J. X. Zhao and F. Yu, Geophys. Res. Lett., 1998, 25,
635.
7 F. Yu and R. P. Turco, Geophys. Res. Lett., 2000, 27, 883.
ꢃc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4383–4394 | 4393

M. C. Holthausen, A Chemist’s Guide to Density Functional
Theory, Wiley–VCH, Weinheim, 2000; (c) A. D. Becke, J. Chem.
Phys., 1993, 98, 5648; (d) M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman,
J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant,
J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci,
M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji,
M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene,
X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken,
C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. Ochterski,
P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador,
J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels,
M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul,
S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M.
W. Gill, B. G. Johnson, W. Chen, M. W. Wong, C. Gonzalez and
J. A. Pople, GAUSSIAN 03 (Revision D.01), Gaussian, Inc.,
Wallingford, CT, 2004.
36 M. W. Chase, Jr, C. A. Davies, J. R. Downey, Jr, D. J. Frurip, R.
A. McDonald and A. N. Syverud, J. Phys. Chem. Ref. Data, 1985,
14(Suppl. 1).
37 W. Wang, E. Yang, J. Hao, X. Zhao, Y. Cao and X. Han, Chem. J.
Chin. Univ., 2002, 23, 2256, and references therein.
38 C. Barckholtz, T. A. Barckholtz and C. M. Hadad, J. Phys. Chem.
A, 2001, 105, 140.
39 R. D. Chirico, W. V. Steele, A. Nguyen, T. D. Klots and S. E.
Knipmeyer, J. Chem. Thermodyn., 1996, 28, 797.
40 Theoretical heats of formation calculated using G3B3 theory;
http://chemistry.anl.gov/compmat/g3xenergies/G3SXHeatsofFor-
mation.htm.
41 S. Vivekananda, J. K. Wolken and F. Turecek, J. Phys. Chem. A,
2001, 105, 9141.
42 S. Steenken and P. O’Neil, J. Phys. Chem., 1978, 82, 372.
43 N. Selvarajan and N. V. Raghavan, J. Phys. Chem., 1980, 84, 2548.
44 R. G. Prinn, J. Huang, R. F. Weiss, D. M. Cunnold, P. J. Fraser,
P. G. Simmonds, A. McCulloch, C. Harth, P. Salameh, S. O’Doh-
erty, R. H. J. Wang, L. Porter and B. R. Miller, Science, 2001, 292,
1882.
45 See for example:L. G. Huey, Mass Spectrom. Rev., 2007, 26
166.
ꢃc
This journal is the Owner Societies 2007
4394 | Phys. Chem. Chem. Phys., 2007, 9, 4383–4394