Fast reactions of atom substitution from polyatomic molecules and solid salts by thermally equilibrated atomic reactants

Article abstract of DOI:10.1023/A:1015338507684

A new class of fast thermal gas-phase reactions of the direct substitution of atoms in polyatomic molecules by atomic reactants is discovered. The Arrhenius parameters are determined for the reactions of atomic deuterium with a number of hydrogen-containing compounds, occurring via the direct substitution of hydrogen atoms and via the abstraction of hydrogen atoms. Fast substitution of alkali metal atoms from the crystals of their salts for the reaction chain carriers of hydrogen flames is found. The importance of reactions of these types in chain combustion is demonstrated. The kinetic isotope effects of hydrogen atom abstraction from hydrocarbon molecules by hydrogen and deuterium atoms are studied. A method for the kinetic studies of free atoms and radicals is developed, which takes into account the role of longitudinal diffusion in the jet and does not require the knowledge of the concentrations of atoms and radicals or the rate constants of other reactions.

Full text of DOI:10.1023/A:1015338507684

Kinetics and Catalysis, Vol. 43, No. 2, 2002, pp. 149–163. Translated from Kinetika i Kataliz, Vol. 43, No. 2, 2002, pp. 165–181.
Original Russian Text Copyright © 2002 by Azatyan.
Fast Reactions of Atom Substitution from Polyatomic Molecules
and Solid Salts by Thermally Equilibrated Atomic Reactants
V. V. Azatyan
Institute of Structural Macrokinetics and Problems of Material Science, Russian Academy of Sciences,
Chernogolovka, Moscow oblast, 142432 Russia
Received April 9, 2001
Abstract—A new class of fast thermal gas-phase reactions of the direct substitution of atoms in polyatomic
molecules by atomic reactants is discovered. The Arrhenius parameters are determined for the reactions of
atomic deuterium with a number of hydrogen-containing compounds, occurring via the direct substitution of
hydrogen atoms and via the abstraction of hydrogen atoms. Fast substitution of alkali metal atoms from the
crystals of their salts for the reaction chain carriers of hydrogen flames is found. The importance of reactions
of these types in chain combustion is demonstrated. The kinetic isotope effects of hydrogen atom abstraction
from hydrocarbon molecules by hydrogen and deuterium atoms are studied. A method for the kinetic studies of
free atoms and radicals is developed, which takes into account the role of longitudinal diffusion in the jet and does
not require the knowledge of the concentrations of atoms and radicals or the rate constants of other reactions.
INTRODUCTION
tution of atoms in polyatomic molecules, which were
discovered by us:
Any homolytic gas-phase reaction of thermally
equilibrated free atoms with valence-saturated mole-
cules were associated until recently with one of the fol-
lowing processes:
Y + RX = RY + X.
(A)
The possibility of reactions of this type in atomic–
(
a) The abstraction of an atom from a molecule and molecular processes was not taken into account earlier
its addition to the attacking atom:
because it was generally assumed that these reactions
virtually do not occur, unlike liquid-phase nucleophilic
(I) substitution reactions. For instance, monographs [1, 2, 5]
.
i
Y + R X = YX + R ;
i
point out that the activation energies of reactions (A)
for hydrocarbons with the participation of hydrogen
and deuterium atoms are higher than 145 kJ/mol, which
is more than 100 kJ/mol higher than the activation ener-
(
b) The abstraction of a radical with the formation of
another radical and a new molecule:
.
2
Y + R R = R Y + R ;
(II)
1
2
1
gies of reactions (I) [11]. This high activation energy
means that, at temperatures below 700 K, the substitu-
tion of an atom in a molecule is more than seven orders
of magnitude slower than the reaction of atom abstrac-
(
c) The addition to a multiple bond:
.
Y + R X(+M) = R XY (+M).
(III)
3
3
tion (I) and five to six orders of magnitude slower than
.
.
reactions (II). Obviously, in this case, reactions (A)
In these reactions, Y and X are atoms; R , R , and
i
2
cannot occur because initial reactants are consumed by
other reactions, including (I) and (II). The value of the
activation energy mentioned in [1, 5] was later used by
several other authors for interpreting their experimental
data (see below).
.
R XY are polyatomic radicals; R X and R R are
valence-saturated molecules; R X is a molecule with a
multiple bond (C=C); and M is a third species or a solid
surface which accepts excess energy of a molecule
formed in reaction (III) (see, for instance, [1–11]).
3
i
1
2
3
Obviously, the reactions of direct substitution of
Only these types of reactions are considered in the atoms considered here differ substantially from step-
studies of elementary reactions and mechanisms of
complex processes by the method of isotopes, in the
study of kinetic isotope effect, and by other methods
wise substitution both in the dynamics of an elementary
act and in kinetics. Early studies neglected this fact. For
instance, Ogg and Polanyi [12] proceeded from the
studies of racemization of optically active 2-iodobutane
[
2–11].
This work reports the results of kinetic studies of in the gas-phase reaction with atomic iodine and draw
fast homolytic gas-phase reactions of the direct substi- the conclusion that the substitution reaction occurs with
0
023-1584/02/4302-0149$27.00 © 2002 MAIK “Nauka /Interperiodica”

1
50
AZATYAN
the Walden inversion of configuration, that is, in one a substitution reaction, e.g.,
step:
D + NH = NH D + H
3
2
I + d-sec-C H I = l-sec-C H I + I.
4
9
4
9
and a reaction between H and D2
H + D = HD + D
However, the possibility of a stepwise reaction was not
excluded; that is, the preliminary abstraction of an
iodine atom from the molecule and further reaction of
2
alternate, the number of deuterated ammonium mole-
cules per one absorbed quantum can be higher than
unity. However, the fact that the quantum yield is higher
than unity does not prove that the substitution reaction
occurs. Indeed, in the above reactions, this may also
happen when the mechanism does not involve the sub-
stitution step at all and when the following reactions
alternate:
the radical with I or recombination with an iodine atom
2
via the following sequence of reactions:
I + d-sec-C H I
4
9
.
sec-C H + I
I + l-sec-C H I.
4 9
4
9
2
.
.
These reactions also lead to racemization. Data on the
inhibition of the exchange process by oxygen support
the stepwise mechanism [13]. The authors of that work
reasonably explain this by the fact that racemization
occurs with the participation of butyl radicals, which
are consumed in the competing reaction in the presence
of oxygen and form butylperoxy radicals.
D + NH = DH + NH , NH + D = NH D + D.
3 2 2 2
2
The rate constant of the first reaction here is lower than
the rate constant of the reverse reaction [18], and the
latter is equal to the rate constant of the second reaction
with an error equal to the kinetic isotope effect. There-
.
fore, the NH radicals formed in the first reaction also
2
rapidly react with D and form atomic deuterium. In
2
In monograph [4] devoted to the kinetics of vari-
ous atomic–molecular reactions, the possibility of
reactions (A) is completely ignored. At the same
time, the monograph provides an estimate of the acti-
vation energy for reactions (II) with the cleavage of
the C–C bond (58–60 kJ/mol). The monograph points
out that hydrogen atom abstraction from a hydrocarbon
molecule is a faster reaction than radical abstraction no
this chain mechanism, the quantum yield is equal to the
chain length and is higher than unity. An analogous
mechanism accounts for the data reported in [15] on
methane deuteration. Thus, even if the reported quan-
tum yields [15] were higher than unity, they could not
prove the substitution reaction.
Reactions (A), including those with water vapors,
matter what the attacking species is, a free atom or a were initially tolerated in [16, 17]. However, in further
free radical.
papers ([19, 20]) one of the authors (Steacie) inter-
preted the results without taking into account this reac-
tion, and he asserts in his monograph [3] that the sub-
stitution reactions are unreal.
The authors of [14, 19, 20], who studied the reac-
tions of atomic deuterium with alkanes, explained the
formation of deuterated hydrocarbons by sole hydrogen
abstraction followed by atomic cracking and recombi-
nation (the mechanism of stepwise substitution). For
instance, the reaction with ethane is described as fol-
lows:
The possibility of gas-phase reactions of type (A)
was considered in some very early papers on the reac-
tions of deuterium atoms [14–17]. It was pointed out
that the amounts of deuterated products formed in these
experiments are large. Farkas and Melville [15] studied
the formation of molecular products of the reaction
between free deuterium atoms, obtained by the dissoci-
ation of D photosensitized by mercury, in the mixtures
2
of NH , CH , and water vapor. The formation of deuter-
3
4
ated ammonium and methane observed above a certain
temperature was explained by the substitution of hydro-
gen atoms in the initial molecules. The authors moti-
vated this by the fact that the quantum yield was higher
than unity at T > 500 K. However, they noted that the
quantum yield was only slightly higher than unity, and
one cannot be sure that the formation of a deuterated
water molecule is due to the direct substitution reaction
rather than to the reaction between an excited mercury
.
D + C H = DH + C H ,
(ID)
(IID)
2
6
2
5
.
.
.
3
D + C H = CH D + CH ,
2
5
2
.
D + CH (+M) = CH D(+M),
(IIID)
3
3
.
D + CH D (+M) = CH D (+M),
(IVD)
(VD)
2
2
2
atom and H O and further steps. Steacie [3] pointed to
2
.
the fact that the values of quantum yields were overes-
timated in [15] and cannot provide evidence for the
substitution reaction. Note also the following circum-
stances. Indeed, the reaction of hydrogen atom abstrac-
tion from a molecule, as well as the recombination of
D + C H (+M) = C H D(+M),
2
5
2
5
.
.
4
D + C H = (C H D)* = C H D + H. (VID)
2
5
2
5
2
Thus, the formation of a hydrogen atom in the above
the radical formed with deuterium atoms, cannot them- scheme (reaction (VID)) is only the result of the
selves lead to quantum yields higher than unity. When decomposition of a quasimolecule, which was formed
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

FAST REACTIONS OF ATOM SUBSTITUTION FROM POLYATOMIC MOLECULES
151
by the recombination and had no time for stabilization was not taken into account either. Moreover, judging
by collision with a third species or a solid surface. from the text of the cited article, T formation in the
2
Although a partially deuterated ethyl radical and a deu- above reaction was also neglected. Thus, the overall
terated product are formed in this reaction scheme amount of molecules containing tritium (HT and T2)
along with a hydrogen atom, the substitution reaction is
not taken into account.
was only assigned to HT, and this was an additional fac-
tor in overestimating the fraction of the abstraction
In connection with the above, note that handbook reaction.
21] provides the activation energies and the rate con-
[
Menzinger and Wolfgang [26] assumed that the sub-
stants that were mistakenly assigned to the hydrogen
substitution reactions in hydrocarbon molecules on
behalf of the authors of [14, 19, 20, 22], whereas it was
noted in the cited papers that these activation energies
and the rate constants are calculated from the overall
rate of the formation of all deuterated products via all
reactions considered there, and reaction (A) was not
among them. Accordingly, the authors of the cited
papers stressed that they assign these kinetic parame-
ters to the overall process rather than to separate reac-
tions. For instance, the authors of [14, 22] took into
account only the following parallel reactions of deuter-
ated ethane formation, which are radical recombination
and disproportionation:
stitution of a hydrogen atom in a hydrocarbon molecule
is only observed if the energy of atomic tritium
obtained by the β-decay of T is higher than 1 eV. How-
2
ever, this conclusion can be a result of the very short
residence time of the reaction mixture in the reaction
zone. This residence time was not specified and it can-
not be estimated from other data reported in [26].
The authors of [27–29] obtained hot deuterium
atoms via tritium β-decay and found that the reaction of
deuterium and methane forms deuterated hydrocar-
bons. They explain their finding by the abstraction of a
hydrogen atom and further recombination of the methyl
radical and atomic deuterium rather than by hydrogen
atom substitution. The authors of [27] found in a simi-
lar experiment that other deuterated methanes are
formed in addition to monodeuterated methane. They
suggested that ionizing radiation might lead to the for-
mation of deuterated methanes via the formation of
.
D + C H (+M) = C H D(+M),
2
5
2
5
.
.
.
3
D + C H = CDH + CH ,
2
5
2
.
.
3
+
+
CDH + CH = C H D.
2
2
5
CH and CH . They also assumed that the addition of
4 5
ammonium removed complexities associated with the
ionization under the action of β-rays. Deuterated meth-
anes found in the absence of ionization are explained by
the abstraction of a hydrogen atom and further recom-
bination of deuterium atoms and a hydrocarbon radical.
The substitution reaction was not considered.
Moreover, the authors of [3, 19, 20] specially noted that
reactions of type (A) do not occur.
The possibility of the direct substitution of atoms
from organic compounds by thermally equilibrated
atomic reagents was negated in [23–25]. These papers
were devoted to the studies of reactions of atomic tri-
tium that have an increased kinetic energy (>2.7 eV).
In most studies where tritium was used, the authors
paid attention to the fact that data are hard to interpret
because of the presence of charged species. Martin and
Willard [30] noted that it is almost impossible to distin-
guish between the substitution and abstraction reac-
tions when hot atoms are used. When authors claim that
the substitution of an atom from a molecule is impossi-
ble, most of them make references to [1, 3, 5], where
the occurrence of reactions (A) with a noticeable rate is
negated.
To obtain these “hot” atoms, a mixture of TBr, CH , and
4
Br was irradiated with UV. The ratio of the rates of
2
hydrogen atom abstraction from the hydrocarbon mol-
ecule and the substitution of these atoms was judged
from the value of the ratio of concentrations of two final
products (HT and CH T) determined by chromatogra-
3
phy. Based on these data and referring to monographs
[1, 3], the authors of [23–25] claimed that the substitu-
tion of hydrogen atoms is much slower than their
abstraction. The reaction rate constants were not deter-
mined. Obviously, the ratio of the rates of substitution
and abstraction with the participation of hot atoms does
not provide evidence for the value of this ratio in the
case of thermally equilibrated atoms. Besides that, the
authors of [23–25] overestimated the amount of HT
assigned to hydrogen abstraction, because they did not
take into account that these molecules are also formed
in the reaction between hydrogen atoms substituted
from a hydrocarbon molecule and TBr, which is the ini-
tial reactant. HT formation by the reaction of H with T2
following the step
Shatz et al. [31] reported a very high activation
energy for reaction (A) when they consider the interac-
tion of atomic deuterium with H O molecules in the gas
2
phase and neglect substitution. Ingold and Roberts [32]
also assumed that the substitution of a hydrogen atom
from CH is very slow and proposed to search for reac-
4
tions of this type in more exothermic processes than
those considered earlier. Monographs devoted to the
theory of elementary reactions and chemical kinetics,
including those that consider transition state theory
[
7, 9, 33–35], did not consider the possibility of reac-
tions (A). Only Kondrat’ev and Nikitin [35] provide a
reference to [16, 17], although the authors of these
T + TBr = T + Br
2
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

1
52
AZATYAN
papers started to deny the possibility of reactions (A) order reaction is determined by the equation
soon after these papers had been published.
d
dn
dn
dx


Thus, all researchers, including those who did not
exclude substitution reactions in their early papers,
have accepted that reactions (A) are very slow and
interpreted their results of studying H, D, and T reac-
tions with polyatomic molecules without taking into
account these steps.
----- D----- – v ----- – kn = 0,
(1)


dx dx
where x is the coordinate along the jet, D is the diffu-
sion coefficient of a reactant, n is the concentration, k is
the reaction rate constant, v is the linear rate of a jet
(
see, for instance, [41]). If the diffusion coefficient is
Despite current understanding, we directly detected independent of x (for instance, the mixture composition
and studied fast thermal reactions (A) of the direct sub- and the pressure along the jet change insignificantly),
stitution of atoms from polyatomic molecules in their then this equation can be described in the following
interactions with thermally equilibrated atomic reactants form:
using the ESR method [36–40]. We found that these
reactions are even faster than most reactions of atom or
radical abstraction from molecules (reactions (I) and
2
d n
dn
D------- – v ----- – kn = 0.
dx
(2)
2
dx
(
II)). ESR spectroscopy used in these studies enabled the
The general solution to this equation is
selective registration of free hydrogen and deuterium
atoms and the determination of their concentrations.
n = C exp(β x) + C exp(β x),
(3)
1
1
2
2
The data from [36–40] are discussed below together
with new results obtained in this study on the homoge-
neous and heterogeneous reactions of hydrogen and
deuterium atoms that are the direct substitution of
atoms in a molecule and the substitution of metal atoms
from solid salts. The results of the primary kinetic iso-
tope effect in the reactions of abstraction by hydrogen
and deuterium atoms from closed-shell molecules are
also presented.
where β and β are the solutions to the characteristic
equation corresponding to differential equation (2); C1
and C are constants. Taking into account the boundary
conditions nt → ∞ = 0 and nx = 0 = nt = 0, where t is time,
1
2
2
we obtain
n = nt = 0 exp(β x).
(4)
1
The value β is determined by the equation
1
2
β = (k/v )(kD/v – 1).
(5)
1
THE ROLE OF LONGITUDINAL DIFFUSION
IN KINETIC STUDIES OF FREE ATOMS
AND RADICALS IN A JET
Because v = x/t, Eq. (5) can be presented in the follow-
ing form:
2
In addition to elucidating the possibility of the reac-
tions of direct substitution of atoms from polyatomic
molecules, the goal of our studies was to determine the
kinetic parameters of these processes and compare
them with the corresponding parameters of abstraction
reactions. Reactions were carried out in jet conditions.
β x = kt(kD/v – 1).
(6)
1
It can be seen from Eqs. (4) and (6) that n =
nt = 0exp(−kt) only if
2
kD/v Ӷ 1;
(7)
Jet methods in kinetic studies offer substantial advan- that is, the longitudinal diffusion can be neglected when
tages, but they are associated with complications due to one considers reaction kinetics only if inequality (7) is
the effect of longitudinal diffusion of reactants on the fulfilled. The higher the rate constant in inequality (7),
profiles of their concentrations. This effect is consider- the higher should be the jet rate to meet this condition.
able in fast reactions, in which case the concentration When studying fast reactions, this requirement compli-
gradients along the jet are high. As a result, a change in cates the experiment and necessitates taking into
the concentration of reacting species due to their diffu- account the pressure along the flow and rate gradients
sion becomes comparable with a change due to an across the jet.
intrinsic chemical reaction. When the effect of diffu-
When the above expressions are used for taking into
sion on the concentration profile of the reaction system
account longitudinal diffusion by traditional methods,
is taken into account correctly using previously known
it is necessary to know the diffusion coefficient and the
methods, equations for the calculation are very com-
rate constants of reactions that enter the expression for
plex and require the knowledge of kinetic and diffusion
calculation, including Eq. (5). In the quantitative
parameters, while ignoring longitudinal diffusion is
kinetic studies of a reaction, the rate constant is an
correct for limited ranges of low temperatures, low con-
unknown value. The motivated neglect of longitudinal
centrations, and low flow rates. Below, we describe a
diffusion in the traditional approach requires checking
method, which we have developed, that makes it possi-
inequality (7) and, therefore, assumes the knowledge of
ble to reduce constraints and complications.
the rate constant, which is sought, and the value of D.
It is known that, in a stationary one-dimensional jet, We developed a method of kinetic studies that allows
the concentration of a substance consumed by a first- one to take into account longitudinal diffusion over a
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

FAST REACTIONS OF ATOM SUBSTITUTION FROM POLYATOMIC MOLECULES
He, H , D
153
wide range of temperatures and concentrations not
2
2
requiring the knowledge of any rate constants and dif-
fusion coefficients. Along with the determination of
rate constants, the method allows the independent
determination of the role of longitudinal diffusion and
the diffusion coefficients of atoms and radicals. Below,
we describe this method together with the procedure of
measurements.
φ18
1
EXPERIMENTAL
φ0.5
to pressure gauge
2
Experiments were performed in jet vacuum systems
combined with EPR-2M or EPR E-104A (Varian) spec-
trometers (depending on the reaction series). Hydrogen
and deuterium atoms were obtained with radio-fre-
quency and microwave discharges (13.6 MHz and
2'
RH
He
φ2
3
9
8
2
.4 GHz) in jets H and D , respectively, with helium.
2 2
An atomic reactant, strongly diluted with the inert gas,
7
passed from discharge section 1 of a quartz tube
(Fig. 1) through nozzle 2 into a reactor. Side nozzle 2'
4
6
was necessary for the admission of a molecular reactant
or the inert gas into the reaction zone. After the reaction
zone, the mixture passed through a cavity of spectrom-
eter (5). An electric furnace with bifilar winding (3)
provided the uniform distribution of temperature in the
reaction zone. The rates of gaseous flows were con-
trolled by valves and measured using flowmeters with
an accuracy of ±2%; the jet was laminar. The pressure
was measured with the same accuracy at the beginning
and at the end of the reaction zone.
φ10
4
5
to pressure gauge
to pump
The experimental procedure was as follows. A
molecular reactant diluted with an inert gas was added
in excess over an atomic reactant through a side nozzle
to the jet of atomic reactant, which entered from the dis-
charge zone. When the jet left the reaction zone, the rel-
ative concentration of atomic component was mea-
sured. Then, at the constant parameters of the jet, the
molecular reactant was replaced by the same amount of
inert gas, and the concentration of atomic reactant was
measured again. Because the molecular reactant
Fig. 1. Reactor with a discharge section: (1) discharge sec-
tion; (2, 2') nozzle; (3) electric furnace; (4) magnet poles;
(5) spectrometer cavity; (6) mobile bottom cover of a cav-
ity; (7) tube for water cooling; (8) reactor; and (9) thermo-
couple. Dimensions are given in millimeters.
The value of B takes into account the consumption of
atomic reactant after the jet left from the constant-tem-
perature zone and before it reached the registration
point (that is, the center of the spectrometer cavity).
This area is an order of magnitude shorter than the con-
stant-temperature reaction zone. Moreover, because of
a decrease in temperature in this area, the reaction dras-
tically slows down due to a decrease in the rate con-
(hydrocarbon, RH) was taken in excess over the atomic
reactant, we were able to take into account only the
consumption of atomic reactant. Then, the value of k in
Eqs. (1)–(5) is
k = khet + bk [RH],
(8) stant. Therefore, the consumption of atoms after the jet
RH
left the constant-temperature zone can be neglected
where k and k are the rate constants of the heteroge-
het
RH
without loss of accuracy; that is, B is much smaller than
the first term on the right-hand side of Eq. (9) in agree-
ment with the results of measurements (see below).
neous decay of atoms and their reaction with the molec-
ular reactant, respectively, and b is the number of atoms
consumed by the reaction of one RH molecule (that is,
the stoichiometric coefficient). The value of b was
When the molecular reactant is replaced by an equal
determined by mass spectrometric analysis of reaction amount of the inert gas, the consumption of the atomic
products. According to Eqs. (4) and (5), a decrease in n reactant is due to recombination, and a decrease in n is
in the presence of the molecular reactant corresponds to determined by the expression
the equation
0
t = 0
0
RH
t = 0
RH

k D

n
n
kD

v
het
0

ln-------- -- = k t 1 – ------------ --  + B ,
(9a)
ln-------- -- = kt 1 – ------------ + B.
(9)
0
het
2
RH

2

n

v

n
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

1
54
AZATYAN
0
where (similarly to B) B takes into account the con- to the free term. In this case, it is not necessary to keep
sumption of atomic reactant after the jet left from the the initial concentrations constant when varying the
constant-temperature zone in the absence of the molec- rate of a flow. The ratio of the slope and the free term
ular reactant.
can be used to determine the diffusion coefficient. At
the same time, the slope of the line in the coordinates of
Eq. (11) is an indicator of the role of longitudinal diffu-
sion in changes of the concentrations of atomic reactant
in the flow. When the inequality (7) is fulfilled (the role
of longitudinal diffusion is insignificant), the straight
lines are horizontal. Under these conditions, Eqs. (10)
and (11) are simplified and take the following form:
The superscripts RH and 0 correspond to the cases
when the molecular reactant or the inert gas, respec-
tively, is added to the jet leaving the discharge zone.
Since the molecular and atomic reactants are strongly
0
diluted with the same inert gas, the values of D and
RH
D
are very close. Taking into account Eq. (8), the dif-
ference of Eqs. (9) and (9a) can be presented in the fol-
lowing form, which is convenient for the processing of
experimental data:
0
n
0
ln------- = bk [RH]t + B – B .
(12)
RH
RH
n
0
RH
1
n
D
-
------------ ln------- = bk – ---------
RH
RH
2
Because the above equations for calculation involve
the concentrations of the atomic reactant only in the
form of a ratio, not only the accuracy of the method is
high, but also any method for the registration of an
atomic reactant can be used if the sensor signal inten-
sity is proportional to the concentration. Since the pri-
[
RH]
t
n
v
(10)
2
0
RH

k
het
D – D 
2
2
×
 k b [RH] + 2k bk – ---------------------------------  .
RH RH het
RH
[
RH]

D

0
RH
Here, n and n are the concentrations of atomic mary goal of this study was to check the possibility of
reactant when it leaves the reaction zone in the experi- the direct substitution of atoms from polyatomic mole-
ments with the inert gas and molecular reactant, respec- cules and to determine the parameters of these reac-
tively, added through a side nozzle. The left-hand side tions, we chose the ESR method, which is highly selec-
of this equation contains only measurable values. The tive to gas-phase paramagnetic species, including
rate of the jet is also directly measurable. It is seen from hydrogen and deuterium. An important advantage of
this equation that, when the molecular reactant concen- this method is also that the reaction mixture need not be
tration and the gas pressure are kept constant and the jet taken out of the reaction zone to, for instance, an ion-
rate is varied, one can determine the bkRH value for each ization chamber to determine the concentrations of
chosen temperature. This value is equal to the free term atomic reactants. Neither is it necessary to introduce
2
in the linear dependence of the left-hand side on 1/v . any accessory or affect the reaction in some other way.
Thus, to determine the rate constant of the reaction
Note that the procedure of alternating the supply of
between atoms and the molecular reactant, there is no
the molecular reactant and an inert gas through the side
need to know any values other than those measured in
nozzle was described in [42], where the rate constant of
this experiment. Note also that the atomic reactant con-
.
the reaction O + H = OH + H was determined. The
centration enters the calculation formula in the form of
2
0
RH
the ratio n /n ; that is, the absolute values of concen- concentration of atomic oxygen was determined in [42]
trations of these active species need not be known.
using special reaction of chemical titration by nitrogen
oxide, which is accompanied by chemiluminescence.
However, Clyne and Trush [42] restricted themselves to
using Eq. (12), which does not take into account longi-
tudinal diffusion. Besides that, the reaction time was
varied by changing the length of a reaction zone.
Because the properties of the reactor surface that deter-
mine the efficiency of heterogeneous decay of atoms
and radicals depend on the concentration of species in
(11) the gas phase (see, for instance, [43–45]), which
changes along the jet, variations in the reaction zone
length lead to some uncertainties in the results of rate
By treating the reactor surface by, for instance,
hydrofluoric acid, conditions can be achieved when the
rate of the heterogeneous consumption of the atomic
reactant is much lower than the rate of its consumption
in the reaction with the molecular reactant. Therefore,
Eq. (10) can be simplified as follows:
2
2
0
1
RH]
n
b kRH RH
f
-
------------ ln------- = bk – ----------- -- D 760-----,
RH 0
RH
2
[
RT
t
n
v
where f is the molar fraction of RH molecules, R is
constant determination. The consideration of this fact is
very important if the heterogeneous reactions of atoms
the gas constant, T is the temperature, and the sub-
RH
script at D0 indicates that this value refers to P = occur at a high rate. Therefore, we varied the reaction
7
60 torr [37, 39].
Thus, by carrying out experiments under the condi-
time by changing the rate of a jet at a constant length of
the reaction zone. A very slight change of it due to tem-
perature variations from one series of measurements to
another was specially taken into account.
tions when the heterogeneous consumption of atoms is
much slower than their homogeneous consumption,
one may use the linear dependence of the left-hand side
The temperature was measured using a chromel–
of Eq. (11) on 1/v to determine a value of bk equal alumel thermocouple placed between the reactor and a
2
RH
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

FAST REACTIONS OF ATOM SUBSTITUTION FROM POLYATOMIC MOLECULES
155
furnace (Fig. 1). Electrolytic hydrogen and deuterium
0


1
RH]
[H]
–10
 ------------ log--------------  × 10 , mol–1 cm3 s–1
were used. To remove olefin traces in the hydrocarbons,
the latter passed through a trap with bromine water. For
further purification, gas passed through a trap with a
concentrated solution of KOH and condensed in a trap
cooled with liquid nitrogen. The middle fraction of the
evaporated gas was allowed to pass to the reservoir of
the vacuum setup.
RH
[

[H]

2
.4
.2
1
2
1
Due to the presence of a nozzle between the sections
of the reaction and discharge, the gas pressure in the
discharge section was about twice as high as the pres-
sure in the reaction zone. Such a pressure difference
prevented the gas from passing from the reactor to the
discharger and allowed a variation in the reactor pres-
sure without any considerable change in the discharger
pressure. In the reaction zone, the concentration of
atomic reactants and the mixture pressure were low
enough to neglect the homogeneous recombination of
atoms with high accuracy. To suppress the recombina-
tion of atoms on the reactor walls, the quartz tube in the
reaction zone was washed with hydrofluoric acid.
3
8
2
4
6
0
.12
4
(× 100)
0.06
0
5
(× 100)
f/v2
1
2
0
1
[H]
Fig. 2. Dependence of ------------log------------- -- on f/v2
RH]
=
[
RH
[H]
2
α[RH]/Pv (where RH is C H ) at temperatures, K: (1) 368
3
8
Excited, metastable hydrogen and deuterium atoms
in a jet after the discharge zone were absent because, in
collisions of these atoms with the species of the gaseous
mixtures, they transfer from the metastable 2s-state to
the energetically close 2p state, from which they further
rapidly transfer to their ground 1s state (allowed transi-
tion). The transition from the metastable 3s state to the
ground state occurs in a similar way. This pathway of
fast deactivation of metastable atoms in the s states was
considered in [46].After leaving the discharge zone and
before mixing with a molecular reactant, hydrogen and
1
4
14
(
5
1
α = 7.46 × 10 ), (2) 345 (α = 6.32 × 10 ), (3) 321 (α =
.06 × 10 ), (4) 443 (α = 13.06 × 10 ), and (5) 404 (α =
1.1 × 10 ). Ordinates of lines 4 and 5 are decreased by a
1
4
14
1
4
factor of 100.
racy. The ratio of the concentrations of atomic reactants
in the computational equations was replaced by the
ratio of signal intensities in the spectrum. The reaction
time was determined by the consumption of a gaseous
mixture taking into account the temperature, the diam-
eter of the reaction tube, and the length of the reaction
zone.
5
deuterium atoms participate in more than 10 colli-
sions. This is many thousands times greater than is nec-
essary for the deactivation of metastable states via
intermediate -states according to the above mecha-
nism. An indicator of the absence of excited atoms in
the reaction zone is, for instance, the fact that the rate
constants of abstraction reactions determined in gener-
RESULTS AND DISCUSSION
The results of studies of reactions of atomic hydro-
ating the atomic reactant using a discharge agree well gen and deuterium with ethane and propane are given
with those determined by radically different methods below. These results are used further to interpret data on
where no discharge was used and hydrogen atoms were the detection and study of homolytic reactions of
formed in a chemical chain reaction (see below).
class (A).
Atomic hydrogen was registered using the ESR
spectrum (g = 2.003) consisting of two identical com-
ponents at a distance of 509 Oe from each other. Spec-
trum splitting is due to the hyperfine interaction of elec-
tron and proton spins. Atomic deuterium was also reg-
Reactions of Atomic Hydrogen
with Hydrogen-Containing Molecules
As Figure 2 shows, in the case of propane, the
0
RH
(
1/[RH])log{ [H] /[H] } value linearly depends on
2
2
istered using the ESR spectrum. As it is known, the f/v = α[RH]/Pv with the negative slope in good
latter consists of three identical components stipulated agreement with Eq. (11). The values of intercepts on
by the hyperfine interaction of the electron and the ordinate axis were used to determine the bk values.
nucleon-deuteron spins. The distance between the com- With a decrease in the temperatures of runs (and with a
ponents is 83 and 90 Oe from the sides of lower and corresponding decrease in the rate constant), the slopes
higher fields, respectively. The strong difference of the straight lines also decreased, pointing to a weak-
between the ESR spectra of free hydrogen and deute- ening of the role of longitudinal diffusion. Below
rium atoms enabled easy identification of these species. 313 K, the straight lines in the coordinates of Eq. (11)
The sensitivity of spectrometers made it possible to are almost horizontal, showing that the role of diffusion
observe changes in their concentrations with high accu- is insignificant. At these temperatures, the experimental
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

1
56
AZATYAN
0
3
–1 –1
[
H]
logk1H, [cm mol s ]
log--------------
RH
[
H]
0
0
0
.75
1
2
3
4
1
1
1
9
7
2
.50
.25
0
3
0
.5
1.0
1.5
2.0
2.5
γ[C H ]t
3
8
1
2
1
3
–1
3
/T × 10 , K
Fig. 3. Dependence of log( [H] /[H]RH) on γ[C H ]t at
0
3
8
Fig. 4. Temperature dependence of the rate constant of
hydrogen atom reaction with C H (k ): (1) obtained in
9
temperatures, K: (1) 313 (γ = 12.82 × 10 ), (2) 295 (γ =
.39 × 10 ), and (3) 284 (γ = 8.97 × 10 ); γ is a scaling
factor.
2
6
1Ì
9
9
9
this work, (2) using the method ignition limits [37, 39, 50],
and (3, 4) obtained in [48, 49].
data are also described well by Eq. (12), which is evi- amount of consumed ethane. Owing to the similar set of
0
RH
reactions for the reaction of hydrogen atoms with pro-
dent from the proportionality of log([H] /[H] ) to
γ[RH]t (Fig. 3). At these temperatures, bk was deter-
mined both from the intercepts on the ordinate axis in
pane, the coefficient b equals 6.
The results of the determination of the rate constant
the coordinates of Eq. (11) (Fig. 2) and from the slopes of reaction (IH) between atomic hydrogen and ethane
of lines in Fig. 3. The straight lines pass through the ori- obtained by the method described above are presented
gin of coordinates in this figure attesting that the value in Fig. 4 in the Arrhenius coordinates. The rate constant
0
13
of B–B is equal to zero. That is, after the jet leaves the thus obtained is (8.13 ± 2.1) × 10 exp[–(4650 ±
3
–1 –1
constant-temperature zone, the consumption of atomic 340)/T] cm mol s . This value agrees well with data
reactant indeed becomes insignificant. Note that, at T > obtained in [37, 39, 50] by a variant of the method of
3
20 K, Eq. (12), which does not take into account lon- ignition limits developed by us earlier, which does not
gitudinal diffusion, does not describe the reaction require a knowledge of the rate constants of any other
0
RH
kinetics: the experimental values ln([H] /[H] ) deviate reactions. The above value is also confirmed by the
more and more drastically downwards from the linear results obtained more recently using radically different
dependence with an increase in [RH]t, whereas Eq. (11) methods and coincides with the value recommended in
describes well the results of measurements over the up-to-date reference literature in the interval of temper-
whole temperature interval (Fig. 2).
atures considered here [51, 52]. All these results pro-
vide further evidence for the accuracy of the method
described above. Specifically, this is evidence for the
absence of excited species in the reaction zone. Data
reported in [48, 49] and shown in Fig. 4 illustrate the
scatter of the rate constant values, which took place
before our publications [36, 37, 50].
Equations (11) and (12) describe the kinetics of
other reactions as well, including the reaction of H with
C H [37].
The obtained data agree well with [3, 47] and show
that, at pressures lower than several hundreds of kilo-
pascals, a reaction such as
2
6
Thus, the method enables a high-accuracy study of
the reaction kinetics and the determination of rate con-
stants. Note that the accuracy of the method also makes
.
H + C H = H + C H
(IH)
2
6
2
2
5
is rapidly followed by the steps of cracking and recom- it possible to determine a change in the ratio of the rate
bination
constants of the reactions of hydrogen atoms in methyl
and methylene groups of propane by varying the tem-
perature. This change is due to small variations in the
activation energies of hydrogen atoms at the primary
.
3
H + C H = 2CH ,
(IIH)
2
5
.
H + CH (+M) = CH (+M),
(IIIH) and secondary carbon atoms in a hydrocarbon mole-
3
4
cule.
where M is the third species that accepts excess energy.
Reactions of Atomic Deuterium
with Hydrogen-Containing Molecules
It follows from this mechanism that, in the reaction
of atomic hydrogen with ethane, the stoichiometric
coefficient b equals 4. This is evident from the fact that
It was found in the experiment with atomic deute-
the amount of methane formed is twice as high as the rium that, with the replacement of the inert gas intro-
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

FAST REACTIONS OF ATOM SUBSTITUTION FROM POLYATOMIC MOLECULES
157
duced through a side nozzle by the equivalent amount amount of deuterium atoms consumed in the reaction
of molecular reactant, the rate of deuterium atom con- with the hydrocarbon (ethane) and alkyl radicals
0
RH
sumption increases to a greater extent than the rate of increases; i.e, the ratio [H]/([D] – [D] ) increases.
atomic hydrogen consumption in analogous experi- Under certain conditions, this ratio reaches 0.85 (that
ments carried out with atomic hydrogen under the same is, 85% of consumed deuterium is replaced by atomic
conditions and with the same molecular reactant. This hydrogen).
result itself differs substantially from data reported in
19], where it was claimed that deuterium atoms are
The rates of hydrogen atom substitution are much
higher than the rates of abstraction reaction (I) from all
organic compounds studied by us. This is further evi-
dence for the formation of atomic hydrogen in the reac-
tions with these compounds not via the stepwise mech-
anism
[
consumed more slowly than hydrogen atoms in the
reaction with a hydrocarbon. The fast consumption of
deuterium atoms in the presence of a hydrocarbon is
accompanied by the appearance of atomic hydrogen,
which is registered by its ESR spectrum. Free hydrogen
atoms are formed only when a hydrogen-containing
molecular reactant is added through a side nozzle to the
jet containing atomic deuterium. Besides that, when the
.
.
D + CH = (CH D)* = CH D + H,
(VIIID)
3
3
2
but via direct substitution. Indeed, the rate of consecu-
.
supply of D to the discharge is discontinued, atomic tive reaction (VIIID) is limited by the rate of CH for-
2
3
mation, which is much slower than the apparent reac-
tions of deuterium with the molecular reactant and
atomic hydrogen formation.
hydrogen is not observed despite the supply of RH into
the reactor through a side nozzle. Thus, the formation
of atomic hydrogen is only due to the reaction of atomic
deuterium with the molecular reactant. The amount of
hydrogen atoms formed in the reaction of deuterium
atoms with the molecular reactant reaches several tens
of percent of the amount of consumed deuterium.
Thus, our experiment shows that, along with reac-
tions (ID)–(VID), the following fast reaction of hydro-
gen atom substitution occurs:
D + RH = RD + H.
(A1)
The appearance of atomic hydrogen in substantial
amounts is observed in the reactions of atomic deute-
rium even with compounds from which hydrogen
abstraction is virtually excluded under given condi-
tions. An example is the reaction with water vapor at
room temperature. The reaction
In the derivation of the calculation formula for the
determination of the rate constants, reaction (A) was
taken into account together with reactions (ID)–(IVD).
Obviously, by analogy with atomic deuterium, atomic
hydrogen can also substitute for a deuterium atom in a
polyatomic molecule. However, because in this experi-
ment the initial concentration of atomic deuterium and,
.
D + H O = HD + OH
(VIID)
2
is endothermic and the enthalpy increases more than by consequently, the concentration of deuterated products
6 kJ/mol [18]. The activation energy of the reverse is much lower than the concentration of C , the sub-
5
H
2 6
reaction (VIID) is 22 kJ/mol [51] (accurate to the value stitution of deuterium atoms for atomic hydrogen can
of the kinetic isotope effect). Therefore, the height of be neglected with a high degree of accuracy. Taking this
the activation barrier of reaction (VIID) is 78 kJ/mol. into account for the reaction with ethane, we obtain the
This means that, for the time required for the reaction following expression based on the reaction scheme for
mixture to pass through a reactor at room temperature, deuterium, which is analogous to Eq. (11) under the
the consumption of deuterium atoms by reaction conditions when the role of longitudinal diffusion is
(
VIID) is several orders of magnitude lower than the substantial:
threshold of the registration of a decrease in the concen-
tration of theses atoms. Meanwhile, the experiment
shows not only significant consumption of deuterium
atoms but the appearance of atomic hydrogen in large
amounts as well. Because abstraction reaction (VIID)
and the subsequent reaction
0
1
[D]
-
-------------ln------------ -- = 4k + kA
1
D
RH
[RH]t
[
D]
(13)
2
RH
0
(4k1D + k
) 760D
f
D
–
----------------------------------------------- -- -----
2
RT
v
.
.
D + OH = OD + H
and analogous to Eq. (12) under the conditions when
the role of this factor is insignificant:
do not occur under these conditions, then the formation
of atomic hydrogen results from its substitution from
0
[
D]
ln------------ -- = (4k + k )[RH]t.
(14)
H O molecules in their reaction with atomic deuterium.
1D
A
2
RH
[
D]
Obviously, the shorter the time during which the
mixture is in the reaction zone and the lower the tem-
The proposed method and the results of measure-
perature, the smaller the portion of hydrogen atoms is ments allow the separate determination of the rate con-
consumed after their formation. Therefore, with an stants of hydrogen substitution and abstraction reac-
increase in the flow rates and a decrease in temperature tions of the hydrogen-containing molecule with atomic
the ratio of the amount of hydrogen atoms to the deuterium. Taking into account the amount of substi-
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

1
58
AZATYAN
3
–1 –1
By subtracting Eq. (16) from Eq. (14), we obtain
logk , [cm mol s ]
i
RH
[
H]

D]

ln 1 + ------------ -- = k [RH]t.
(17)
A

RH

[
1
0
9
It is seen from this equation that the value k can be
A
determined from the slope of the linear dependence of
the left-hand side on [RH]t. In the processing of the
data, the ratio of the concentrations of atomic hydrogen
and deuterium was replaced by the ratio of signal inten-
sities in the spectra taking into account the fact that the
signals of atomic deuterium are somewhat broader and
that its spectrum contains three components, unlike the
spectrum of hydrogen atoms consisting of two compo-
nents.
2
1
8
2
.0
2.5
3.0
/T × 10 , K
3.5
3
–1
1
In the experiments with atomic deuterium, the tem-
perature was varied from 343 to 489 K and the pressure
Fig. 5. Temperature dependence of (1) k1D and (2) kD for
2
2
was varied from 1.3 × 10 to 4 × 10 Pa. Table 1 shows
the values of rate constants for reactions (ID) and (A),
their ratios, and the overall rate constant at different
deuterium atoms with ethane.
tuted atomic hydrogen, we obtain the following expres- temperatures obtained using Eqs. (13)–(16).
sion for the reaction of hydrogen atom abstraction from
the initial molecule with the formation of HD:
Figure 5 shows the results in the Arrhenius coordi-
nates. The following temperature dependences were
obtained for the rate constants of reactions (ID) and (A)
2
3
–1 –1
0
of deuterium atoms with ethane (cm mol s ):
1
[D]
16k1D
f
-
-----------ln----------------------------- = 4k – -------------D760-----. (15)
1D
RH
2
[
RH]
RT
13
[
D] + [H]
v
k1D = (34.9 ± 1.2) × 10 exp[–(4200 ± 400)/T],
1
3
Equations (13)–(15) were obtained taking into
k = (3.6 ± 0.4) × 10 exp[–(3300 ± 350)/T].
A
account that hydrogen atoms substituted in reaction (A)
are consumed in reactions (IH)–(IIIH) and in the reac-
tions of heterogeneous recombination. Unlike Eq. (12),
formula (15) describes the consumption of atomic deu-
terium only in abstraction reaction (ID) and subsequent
steps (IVD), (VD), and (VID). Therefore, the linear
It is seen from the ratio of the overall rate constant
kD = 4k1D + k to 4k that the apparent fast consump-
A
1D
tion of deuterium atoms cannot be explained by the
abstraction reaction and consequent cracking and
recombination reactions. Indeed, even if all hydrogen
atoms in the ethyl and methyl radicals were exchanged
with deuterium in reaction (VIIID), the rate constant of
2
dependence of the left-hand side of Eq. (15) on f/v
enables direct determination of the k value. Under the
1D
the overall reaction of atomic deuterium (k ) with
D
conditions when the role of longitudinal diffusion is ethane and radicals would be only 9/4 times higher than
insignificant (as is evident from the absence of the 4k . Meanwhile, as can be seen from the above data,
1
D
slopes of the straight lines), this equation takes a sim- the k value is greater than 4k from 2.6 to 4.6 times
D
1D
(
i.e. from 10.4/4 to 18.4/4 times) over the whole tem-
pler form:
perature interval considered here. The lower the tem-
perature, the greater the difference between these val-
(16) ues. Thus, the substitution of a hydrogen atom in the
molecular component has the most pronounced contri-
0
[
D]
ln----------------------------- = 4k [RH]t.
1
D
RH
[
D] + [H]
Table 1. Kinetic parameters of deuterium atom reactions with ethane
Temperature, K
Parameter
3
00
343
364
415
439
489
k1D × 10 , cm mol s–1
–
8
3
–1
0.28
6.2
2.2
25.1
11.1
33.9
3.8
3.6
30.7
8.5
15.1
118.6
7.8
29.0
182.0
6.3
63.2
453.0
7.2
k × 10 , cm mol s–1
–
8
3
–1
A
k /k
22.2
A
1D
k , cm mol s–1
3
–1
7.2
4.6
45.1
3.1
179.0
2.9
298.0
2.6
705.0
2.8
D
kD/4k1D
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

FAST REACTIONS OF ATOM SUBSTITUTION FROM POLYATOMIC MOLECULES
159
3
–1 –1
bution to the overall consumption of atomic deuterium
in the reaction with the molecular component.
logk , [cm mol s ]
i
1
2
3
1
0.5
0.0
Analogous fast substitution of hydrogen atoms was
found in the studies of reactions of atomic deuterium
with other hydrogen-containing compounds, such as
methane, propane, ethanol, acetone, and water vapor.
The results of kinetic studies of reactions of hydrogen
and deuterium atoms with propane via directions (ID)
and (A) are presented in Fig. 6. It can be seen, that the
rate of the substitution reaction is much higher than the
rate of atom abstraction in this case as well.
1
9
.5
9.0
8.5
2
.0
2.5
3.0
3.5
3
–1
1
/T × 10 , K
When considering various molecular reactants, the
k /k ratio is higher for a reactant from which abstrac-
A
1D
Fig. 6. Dependence of the rate constants for reactions of
tion of a hydrogen atom is more difficult.
(
1) hydrogen and (2, 3) deuterium atoms with propane:
(
2) k and (3) k1D.
In the experiments with water vapor, helium was
bubbled through a saturator filled with distilled water
before passing to the side nozzle of the reactor. Satura-
tion was carried out at a temperature, which was higher
than the necessary value to achieve the desired concen-
tration of water vapor. Then, a flow of helium and water
vapor passed through a trap packed with glass balls at a
lower temperature to obtain the desired concentration
A
3
93 K showed that the values of the overall rate con-
stant k (and, hence, k ) are independent of the surface
D
A
material with an error of 10–15%, and the substitution
reaction thus occurs in the gas phase.
A great difference in the rates of reactions (A) and
of water vapor. Other details of the procedure were the (I) and the absence of correlations between the activa-
same as described above. Thus obtained, the rate con- tion energies and enthalpies show that these are differ-
stant of the reaction
D + H O = DOH + H
ent types of reactions with different transition states.
Most probably, the spontaneous substitution of a hydro-
gen atom by deuterium in hydrocarbon molecules
occurs as an attack of a deuterium atom on a carbon
(IXD)
2
6
3
–1 –1
was 5 × 10 cm mol s at room temperature, which
(rather than hydrogen) atom of a hydrocarbon. Our data
corresponds to an activation energy no higher than
8 kJ/mol at a normal value of the preexponential fac-
show that this attack only results in the C–H bond
cleavage rather than C–C bond cleavage, and deuterium
substitutes for a hydrogen atom in the initial molecule.
Indeed, it can be easily shown that, in the case of C–C
bond cleavage in the reaction of a hydrogen atom with
ethane, the stoichiometric coefficient would be equal
to 2, and it would be equal to 4 in the case of propane.
However, it was shown above that the experimental sto-
ichiometric coefficient in the reaction of hydrogen atom
with ethane is 4. This indicates that reactions (IH),
3
tor. This is more than 40 kJ/mol lower than the activa-
tion energy of hydrogen abstraction. Correspondingly,
in the cases considered here, the rate constant of substi-
tution is millions of times higher than the rate constant
of the abstraction reaction.
Note that the results of potential energy surface cal-
culations [31] and the activation energy obtained in that
work (38.8 kJ/mol) agree well with the experimental
value obtained by us earlier for the rate of hydrogen
(
IIH), and (IIIH) occur. In the reaction with propane,
substitution in H O. However, the authors of [31] were
2
the stoichiometric coefficient is 6 due to an analogous
mechanism, which also involves propyl radicals.
unfamiliar with our experimental data [39, 53] on the
fast substitution of H in H O at room temperature and
2
The above experimental data on hydrogen atom
abstraction by hydrogen and deuterium from the ethane
molecule correspond to the kinetic isotope effect
expressed as follows:
claimed that, contrary to their calculated data, the acti-
vation energy of hydrogen atom substitution H should
be much higher than their calculated value.
To find out whether heterogeneous factors contrib-
ute to the discovered substitution reactions of hydrogen
atom in polyatomic molecules, we studied reactions of
hydrogen and deuterium atoms with ethane in a special
k /k = 2.08exp(–450/T).
H
D
Consideration of reactions (A) also makes it possi-
series of experiments using a quartz reactor, whose sur- ble to explain data from [54], where it was found that
face properties were varied by chemical modification. the products of olefin hydrogenation in the mixture of
In one series of experiments, the surface was washed atomic hydrogen and molecular deuterium are partially
with hydrofluoric acid, and in the other series, the sur- deuterated. Some of the product molecules contain
face was covered with sodium tetraborate. Obviously, more than one deuterium atom. The authors of [54]
the surface of sodium tetraborate, an alkali metal salt explained this assuming that the free valence of a car-
with a weakly acidic anion, differs substantially from bon atom in the free radical simplifies the exchange of
quartz glass treated with hydrofluoric acid in its chem- deuterium with hydrogen atoms connected to this car-
ical properties and structure. Measurements at 295 and bon atom. This assumption was not supported by
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

1
60
AZATYAN
proper arguments and contradicts results obtained in the limits and the rate of nonthermal flame propagation
the studies of reactions of molecular deuterium with of CS and O .
2
2
photochemically generated alkyl radicals [55]. When
fast reactions (A) are considered, these data are easier
to explain. Indeed, because the activation energy of
reaction
Note that, in contrast to reaction (VIID) of atomic
deuterium with water vapor, which is the only reaction
of the observed formation of atomic hydrogen at low
temperatures, the formation of atomic sulfur in the
reaction mixture O + CS does not provide compelling
H + D = HD + D
2
2
evidence for substitution reaction (V). Indeed, fast
reactions
is not high (~30 kJ/mol), the reaction system studied in
[
54] also contains atomic deuterium, which undergoes
O + CS = SO + CS and O + SO = O + S (VI)
the addition to an olefin molecule and enters reaction (A)
with the hydrogenation product. The number of deute-
rium atoms in the deuterated product molecule is deter-
mined by the number of times it enters reaction (A).
2
2
occur one after another at high rates in such a mixture
and result in the formation of free sulfur atoms [56],
while the presence of self-ignition limits of the reaction
of CS with O unambiguously points to the fact that,
The discovery of fast gas-phase homolytic substitu-
2
2
2
N
tion reactions of the S type (A) suggests reconsider- along with chain branching, which has the second over-
all kinetic order with respect to the concentrations of
ing ideas that negate the occurrence of such reactions.
active intermediate species (with nonlinear branching)
Specifically, the high rate constants of reactions (A)
and stipulates nonthermal flame propagation, the reac-
obtained experimentally and the low activation energies
tion of linear branching also occurs. It is the substitu-
show that the calculated values reported in [1, 2, 5] can-
tion reaction (IV), which is characterized by the first
not be a reason to deny reactions of this type. The high
order with respect to the active intermediate species
rates of reactions (A) also point to the necessity of tak-
(
oxygen atom), and provides linear branching together
ing them into account when studying the mechanisms
of complex chemical processes accompanied by the
formation of atoms and radicals, using isotopically
with reaction (V).
Reaction (V) is an example illustrating that fast
labeled molecules, as well in the theory of elementary homolytic reactions of atom substitution from poly-
chemical reactions.
atomic molecules may involve atoms other than atomic
hydrogen or its isotopes, as well as other molecules.
These reactions may play an important role in the pro-
cess of chain combustion and pyrolysis.
Direct Substitution Reaction in the Process
of Carbon Disulfide Combustion
As expected, the direct homolytic substitution of
atoms from polyatomic molecules is not restricted to
substitution reactions of hydrogen atoms. To explain
the specific features of carbon disulfide oxidation, we
supplemented the chemical scheme of the process con-
taining the step of active species multiplication [56],
Heterogeneous Reactions of Atomic Substitution
in a Chain Process
Atomic substitution is also observed in the reactions
of gas-phase atomic reactants with solids. The substitu-
tion by atomic hydrogen, formed in a rarefied flame of
H with O , for sodium and potassium atoms to the gas
2
2
S + O = SO + O,
(IV)
2
phase from the respective solid chlorides was found
by the thermoneutral reaction of direct substitution of a in [58].
sulfur atom from the CS molecule:
2
A quartz reactor and the registration system are
described in Fig. 7. A quartz cylinder was placed into
the reactor. The internal surface of the cylinder was
covered with a salt to be studied. The reactor design
allowed us to vary the position of a cylinder with
respect to the gas inlet (positions A and B in Fig. 7).
This was necessary for the determination of the contri-
butions of reactions of the atomic reactant with the
solid salt and its vapors. The pressure of the mixture
and temperature were varied from 300 to 500 Pa and
from 543 to 783 K, respectively. Gas-phase alkali metal
atoms were directly registered by the method of emission
spectroscopy using an OSA-500 multichannel analyzer of
optical spectra (with a resolution of 1.4 Å/channel)
equipped with a system for signal accumulation. In spe-
O + CS = COS + S.
(V)
2
This made it possible to explain the presence of self-
ignition limits. Unlike the limit of nonthermal flame
propagation, these limits are known to be determined
by the competition of the reactions of active intermedi-
ate species that follow the first kinetic order with
respect to the concentrations of these species. Reaction
(
V) together with (IV) provides such chain branching.
The value of the rate constant of reaction (V), equal to
1
1
3
–1 –1
1
.84 × 10 cm mol s at 298 K [57], determined
later shows that this reaction is so fast that, together
with (IV), it can provide a rate of linear chain branching
corresponding to the observed values of self-ignition.
Simulations of a set of equations corresponding to cial runs where atomic hydrogen was obtained by H2
the reaction scheme that also takes into account nonlin- dissociation, sodium atoms were registered in the gas
ear chain branching [56] made it possible to describe phase by the method of resonance fluorescence. Fluo-
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

FAST REACTIONS OF ATOM SUBSTITUTION FROM POLYATOMIC MOLECULES
161
6
1
1
3
9
1
4
1
0
5
1
4
A
B
OSA-500
H (Ar)
2
8
7
a
1
3
9
2
1
2
6
7
O (H + Ar)
2
2
Fig. 7. Reactor for the registration of metal atom substitution from their salts: (1) reactor, (2) electric furnace, (3) resonance lamp,
(
4) spectral analyzer, (5) positions of a mobile tube covered with a chloride, (6) microwave discharger, (7) inert gas charge tube,
7a) H (and Ar) charge tube, (8) communication lines for pumping, (9) collimator, (10) quartz windows, (11) diaphragm, (12) light
(
2
trap, (13) nozzle, and (14) thermocouple.
rescence of atomic sodium escaped to the gas phase
In the runs where atomic hydrogen was obtained
was excited by high-frequency discharge irradiation in with a discharge, the radiation of excited sodium atoms
a quartz lamp containing NaCl (or metallic sodium in and the signal of resonance fluorescence of these atoms
some runs) and filled with a mixture of N (0.1 torr) and in the ground state were registered. Sodium atoms
2
Ar (2.0 torr). This source provided radiation at a wave- appear in the gas phase only if atomic hydrogen is
length of 589 nm.
available. When the flow of hydrogen atoms through
side inlet 7 is discontinued, the signals of sodium and
potassium atoms in the ground and excited states disap-
pear.
The reaction mixture was supplied from the flame
into the reactor through side inlet 7. It is known that, in
the combustion of the stoichiometric or hydrogen-reach
mixtures of H and O , hydrogen atoms are almost the
The pressures of argon supplied through inlet 7a and
a mixture coming from the discharge zone through a
side tube 7 were 0.55 and 1.10 torr, respectively, in the
first series of experiments and 1.87 and 3.73 torr in the
second series. Table 2 shows as an example the results
of measurements of radiation intensity (I*) and the sig-
2
2
only active intermediate species participating in heter-
ogeneous reactions, because their concentration is
much higher than the concentration of oxygen atoms
.
and OH radicals and because they are more mobile. In
the runs with a mobile cylinder with NaCl (or KCl) in
position A, hydrogen atoms only contacted salt vapors
carried by a flow of inert gas passing from inlet 7‡,
while in the runs with the cylinder in position B, hydro-
gen atoms also contacted the solid salt. Special mea-
surements were performed to ensure the constant flow
of salt vapors carried with a jet passing through inlet 7‡
along the whole length of the reactor: when the internal
cylinder was in position A, the value of the resonance
fluorescence signal and the intensity of Na atom radia-
tion were independent of the distance between the cyl-
inder and the place where atomic hydrogen was intro-
nal of resonance fluorescence (I ) of atomic sodium at
RF
a gas temperature of 718 K in two series of runs, where
atomic hydrogen was obtained in a discharge.
These data show that, if hydrogen atoms interact not
only with the solid salt but also with NaCl vapors, the
concentration of sodium atoms in the ground state reg-
istered by the method of resonance fluorescence is
1
2−13 times higher. In this case, the concentration of
excited sodium atoms is also much higher, although
deactivation makes this effect smoother.
In the study of NaCl interaction with atomic hydro-
duced, that is, nozzle 13. Because the temperature was gen formed in a flame, molecular hydrogen was sup-
constant along the reactor axis, a comparison of signal plied from the butt-end of the reactor, whereas oxygen
intensities made it possible to determine the contribu- was supplied through a side inlet. Ignition and combus-
tion of the heterogeneous reaction to the formation of tion occurred to the right of the nozzle. When the cylin-
Na and K atoms in the gas phase.
der covered with NaCl was in position B, atomic
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

1
62
AZATYAN
Table 2. Dependence of radiation and resonance fluores- the color of the surface. It became dark purple. Fast
cence intensities for sodium atoms on the cylinder position
substitution of an alkali metal atom from its salt
described above is one of the pathways for the chemical
modification of the surface in the course of a chain pro-
cess. Obviously, surface changes under the action of a
chain process affect in their turn the characteristics of
combustion in general. In the general case, gas-phase
reactions of substituted metal atoms should also be
taken into account. Considerable changes in the fea-
tures of hydrogen combustion in oxygen in reactors
with surfaces covered with NaCl and KCl [60] is an
example of manifestation of modifications in the sur-
face state and its role.
(A and B in Fig. 7)
Intensity, rel. units
Signal
A
B
A
B
Series 1
Series 2
I*
172
27
444
362
27
10
124
120
IRF
Table 3. Dependence of luminescence intensity (rel. units)
Obviously, the transition of atoms from a solid
for atoms (Na and K) substituted from their solid salts NaCl matrix to the gas phase under the action of an atomic
and KCl on the position of the cylinder under conditions reactant can also take place in other process with the
of rarefied flame
participation of free atoms, including other chain pro-
cesses.
NaCl
KCl
Wavelength, nm
767
KCl
770
Cylinder position
CONCLUSION
5
89
Thus, when atomic reactants react with polyatomic
molecules, fast direct substitution reactions occur along
with atom abstraction reactions. The elucidation of
kinetics and mechanisms of these homolytic reactions,
including the configurations of transition states, is one
of the topical problems in the theory of elementary
reactions. This problem is also of interest for under-
standing the mechanisms and kinetics of gas-phase pro-
cesses occurring with the participation of free atoms
and radicals, such as pyrolysis and combustion. The
reactions of metal atom substitution from their salts by
atomic gas-phase reactants are of interest for both
topochemistry and gas-phase pyrolysis and combustion
processes.
A
B
1953
5407
345
273
815
1030
sodium was registered both in the gas phase in the
excited and ground states. In analogous experiments
with KCl, atomic potassium was registered. When the
surface contacted the fuel mixture in the absence of
flame, metal atoms were not registered in the gas phase.
To eliminate the effect of the internal cylinder on the
flame geometry, two identical cylinders were put into
the reactor in positions A and B in special runs. The
inner surface of one of them was covered with NaCl
and that of the other, with KCl. The role of the hetero-
geneous reaction in the appearance of an alkali metal in
the gas phase was judged on the basis of changes in the
luminescence signal intensities in positions A and B
ACKNOWLEDGMENTS
I thank I.I. Moiseev and Yu.A. Kolbanovskii for
when cylinders changed places. The ratio of lumines- helpful discussions of this work.
cence signal intensities in positions A and B correlates
with the corresponding ratio of resonance fluorescence
intensities. Table 3 exemplifies one experimental series.
REFERENCES
It is seen again from these data that, if hydrogen
atoms contact the solid salt, the concentrations of both
alkali metals are much higher than in the case of inter-
action with salt vapors only. Thus, the reaction with the
solid salt is the main channel for metal to escape into
the gas phase. The appearance of yellowish color in the
hydrogen flame in glass tubes can be explained by the
substitution of sodium atoms from the glass. These
atoms radiate the characteristic d-line on excitation in a
flame. Note that after our work [58], the escape of
sodium atoms from solid salts to the gas phase under
flame conditions was also observed in [59].
1. Glasstone, S., Laidler, K.J., and Eyring, H., The Theory
of Rate Processes, New York: Princeton Univ., 1941.
2. Waters, W., The Chemistry of Free Radicals, Oxford,
1946.
3
4
. Steacie, E.W.R., Atomic and Free Radical Reactions,
New York: Reinhold, 1954, vol. 2.
. Semenov, N.N., O nekotorykh problemakh khimicheskoi
kinetiki i reaktsionnoi sposobnosti (On Some Problems
of Chemical Kinetics and Reactivity), Moscow: Akad.
Nauk SSSR, 1958.
5
. Kondrat’ev, V.N., Kinetika khimicheskikh gazovykh
reaktsii (Kinetics of Chemical Gas-Phase Reactions),
Moscow: USSR Academy of Sciences, 1958.
The escape of Na and K atoms during the interac-
tions of their salts with atomic hydrogen obtained in a
flame or in a discharge is accompanied by a change in
6. Melander, L., Isotope Effects on Reaction Rates, New
York, 1961.
KINETICS AND CATALYSIS Vol. 43 No. 2 2002

FAST REACTIONS OF ATOM SUBSTITUTION FROM POLYATOMIC MOLECULES
163
7
. Kondrat’ev, V.N., Nikitin, E.E., Reznikov, A.I., and 34. Semiokhin, I.A., Strakhov, B.V., and Osipov, A.I., Kurs
Umanskii, S.Ya., Termicheskie bimolekulyarnye reaktsii
v gazakh (Thermal Bimolecular Reactions in Gases),
Moscow: Nauka, 1976.
khimicheskoi kinetiki (Chemical Kinetics), Moscow:
Mosk. Gos. Univ., 1995.
3
3
5. Kondrat’ev, V.N. and Nikitin, E.E., Khimicheskie pro-
tsessy v gazakh (Chemical Processes in Gases), Mos-
cow: Nauka, 1981.
8
. Emanuel’, N.M. and Knorre, D.G., Kurs khimicheskoi
kinetiki (Chemical Kinetics), Moscow: Vysshaya
Shkola, 1984.
. Eyring, H., Lin, S.H., and Lin, S.M., Basic Chemical
Kinetics, New York: Wiley, 1980.
6. Azatyan, V.V. and Filippov, S.B., Dokl. Akad. Nauk
SSSR, 1969, vol. 184, p. 625; Izv. Akad. Nauk SSSR, Ser.
Khim., 1969, no. 5, p. 1186.
9
1
1
1
1
1
1
1
1
1
1
2
2
0. Fujisaki, N., Gaumann, T., and Ruf,A., Ber. Bunsen-Ges.
3
3
3
4
4
7. Azatyan, V.V., Dokl. Akad. Nauk SSSR, 1972, vol. 203,
Phys. Chem., 1986, vol. 90, no. 2, p. 375.
p. 137.
1. Baulch, D.L., Cobos, C.J., Cox, R.A., et al., Combast.
Flame, 1994, vol. 98, no. 11, p. 59.
8. Azatyan, V.V., Filippov, S.B., and Khachatryan, M.S.,
Kinet. Katal., 1971, vol. 12, no. 1, p. 17.
2. Ogg, R.A. and Polanyi, M., Trans. Faraday Soc., 1935,
9. Azatyan, V.V., Doctoral (Chem.) Dissertation, Moscow:
vol. 31, no. 2, p. 482.
Inst. of Chem. Phys., 1978.
3. Bujuke, J.E., Pratt, M.W.T., and Noyes, R.M., J. Am.
Chem. Soc., 1961, vol. 83, no. 7, p. 1547.
0. Azatyan, V.V., Gazaryan, K.G., and Garibyan, T.A.,
Kinet. Katal., 1987, vol. 29, no. 1, p. 38.
4. Morikawa, K., Benedict, W.S., and Taylor, N.S.,
1. Frank-Kamenetskii, D.A., Diffuziya i teploperedacha v
khimicheskoi kinetike (Diffusion and Heat Transfer in
Chemical Kinetics), Moscow: Nauka, 1987.
J. Chem. Phys., 1937, vol. 5, no. 1, p. 212.
5. Farkas, A. and Melville, H.W., Proc. R. Soc., 1936,
vol. 157, no. 3, p. 625.
4
2. Clyne, M.A.A. and Thrush, B.A., Proc. R. Soc., 1963,
6. Geib, K.H. and Steacie, E.W.R., Trans. Roy. Soc. Can.,
vol. 275A, p. 599.
1
935, vol. 3, no. 1, p. 91.
43. Azatyan, V.V., Usp. Khim., 1985, vol. 54, no. 1, p. 33.
44. Azatyan, V.V., Zh. Fiz. Khim., 1998, vol. 72, no. 3, p. 391.
45. Azatyan, V.V., Kinet. Katal., 1999, vol. 40, no. 2, p. 180.
7. Geib, K.H. and Steacie, E.W.R., Z. Phys. Chem., 1935,
vol. 29, no. 1, p. 215.
8. Corchado, J.C. and Garscia, J.E., J. Chem. Phys., 1997,
vol. 106, no. 10, p. 4013.
46. Kampaneets, A.S., Teor. Eksp. Khim., 1972, vol. 8, no. 1,
9. Shiff, H.J. and Steacie, E.W.R., Can. J. Chem., 1951,
p. 147.
vol. 29, no. 1, p. 1.
4
4
4
5
5
5
5
5
7. Kazmi, H.A. and Le Roy, D.J., Can. J. Chem., 1964,
0. Steacie, E.W.R. and Phillips, N.W.F., J. Chem. Phys.,
vol. 42, no. 5, p. 1145.
1
936, vol. 4, no. 8, p. 461.
8. Berlie, M.K. and Le Roy, D.J., Disc. Faraday Soc.,
1. Kondrat’ev, V.N., Konstanty skorosti gazofaznykh
reaktsii (Rate Constants of Gas-Phase Reactions), Mos-
cow: Nauka, 1970.
1953, vol. 14, no. 1, p. 50.
9. Panfilov, V.N. and Voevodskii, V.V., Kinet. Katal., 1965,
vol. 6, no. 4, p. 577.
2
2
2
2
2
2
2
2
3
3
3
3
2. Trenner, N.R., Morikawa, K., and Taylor, H.S., Z. Chem.
0. Azatyan, V.V., Dokl. Akad. Nauk SSSR, 1971, vol. 196,
Phys., 1937, vol. 5, no. 1, p. 203.
no. 3, p. 617.
3. Chon, C.C. and Rowland, F.S., J. Am. Chem. Soc., 1966,
1. Tsang, W. and Hapson, R.F., J. Phys. Chem. Ref. Data,
vol. 88, no. 11, p. 2612.
1
986, vol. 15, no. 3, p. 1087.
4. Roor, J.W. and Rowland, F.S., J. Chem. Phys., 1967,
2. Baulch, D.L., Cobos, C.J., Cox, R.A., et al., Combust.
Flame, 1994, vol. 98, no. 1, p. 59.
vol. 46, no. 11, p. 4299.
5. Chon, C.C. and Rowland, F.S., J. Chem. Phys., 1969,
3. Azatyan, V.V., Grigoryan, F.A., and Filippov, S.B.,
vol. 50, no. 6, p. 2763.
Kinet. Katal., 1972, vol. 13, no. 6, p. 1389.
6. Menzinger, M. and Wolfgang, R., J. Am. Chem. Soc.,
4. Voevodskii, V.V., Lavrovskaya, G.K., and Mardaleish-
vili, R.E., in Voprosy khimicheskoi kinetiki, kataliza i
reaktsionnoi sposobnosti (Problems of Chemical Kinet-
ics, Catalysis, and Reactivity), Moscow: USSR Acad-
emy of Sciences, 1955, p. 40.
1
967, vol. 89, no. 24, p. 5992.
7. Estrup, P.J. and Wolfgang, R., J. Am. Chem. Soc., 1960,
vol. 82, no. 11, p. 2661.
8. Firestone, R.F., Lemr, C.F., and Trudel, C.F., J. Am.
Chem. Soc., 1962, vol. 84, no. 12, p. 2279.
5
5
5
5
5
5. White, E. and Steacie, J. Chem. Phys., 1953, vol. 21,
9. Lawrence, R.H. and Firestone, R.F., J. Am. Chem. Soc.,
p. 933.
1
966, vol. 88, no. 19, p. 4564.
6. Azatyan, V.V., Novozhilov, B.V., Posvyanskii, V.S., and
0. Martin, R.M. and Willard, J.N., J. Chem. Phys., 1969,
Semenov, N.N., Kinet. Katal., 1976, vol. 17, no. 6, p. 1386.
vol. 40, no. 7, p. 3007.
7. Atkinson, R., Baulch, D.L., Cox, R.A., et al., J. Phys.
Chem. Ref. Data, 1997, vol. 26, no. 6, p. 1329.
8. Rubtsov, N.M., Azatyan, V.V., Borodulin, R.R., et al.,
Khim. Fiz., 1984, vol. 3, p. 1789.
9. Baulch, D.L., Griffiths, J.F., and Richer, R., Combust.
Flame, 1991, vol. 85, nos. 1–2, p. 271.
1. Shatz, G.C., Colten, M.C., and Grant, J.L., J. Phys.
Chem., 1984, vol. 88, no. 14, p. 2971.
2. Ingold, K.U. and Roberts, B.P., Free-Radical Substitu-
tion Reactions, New York: Wiley, 1971.
3. Nikitin, E.E., Teoriya elementarnykh atomno-moleku-
lyarnykh protsessov v gazakh (The Theory of Elemen-
tary Atomic-Molecular Processes in Gases), Moscow: 60. Azatyan, V.V., Sklyarenko, V.I., and Shavard, A.A.,
Khimiya, 1970.
Khim. Fiz., 1982, vol. 1, no. 12, p. 1704.
KINETICS AND CATALYSIS Vol. 43 No. 2 2002