Vol.3, No.8, 661-682 (2011) Natural Science
http://dx.doi.org/10.4236/ns.2011.38090
The influence of the different elements of an organic
molecule structure on the main kinetic parameters of its
unimolecular reaction in the high-pressure region
David Krinkin
Yozmot HaEmek Technological Incubator, Migdal HaEmek, Israel;
Received 29 June 2011; revised 17 July 2011; accepted 22 July 2011.
ABSTRACT
The most general dynamic tendencies of the
energy redistribution in the high-pressure re-
gion are considered. Their influence on the
possible deviations from the kinetic concep-
tions, which is now generally accepted, is ex-
amined. In this way, the structural elements of
an organic molecule that promote internal en-
ergy mobilization in the high-pressure region
and, conversely, hamper it, are defined. The first
of these elements reduces both the Arrhenius
parameters of the unimolecular reactions while
the second leads to the opposite results. Some
well-known exceptions to existing kinetic theo-
ries, which find an explanation in the framework
of these proposed concepts, is considered. The
proposed concept is very general as distinct
from the existing dynamic studies, which inves-
tigate more particular details of the separate
bond behaviors. The proposed general concept
can broaden the study of chemical kinetics.
Keywords: Arrhenius Parameters; A-Factor;
High-Pressure Region
1. INTRODUCTION
The present paper is devoted to the dynamics of en-
ergy mobilization inside an energized molecule in order
that a reaction will proceed in the high-pressure region.
Presently, the dynamics picture of the energy flow in
chemical kinetics is studied practically exclusively by
the method of collision-induced relaxation of a vibra-
tionally excited molecule, which has been the subject of
continuing interest for the past several decades [1-6].
The problem of the internal energy mobilization of a
molecule is studied mainly in the fall-off region, where it
plays a decisive role and determines in particular the
transition in the high-pressure region. The existing dy-
namical method is concerned with the detailed treatment
of molecular vibrations and the behavior of particular
molecular coordinates as a function of time. Generally,
this relaxation is studied during and after one collision.
In the present paper, in contrast to the existing dy-
namics method, an attempt was made to formulate those
elements of the organic molecule structure that promote
the concentration of the energy surplus of an energized
molecule on the bond that is being broken or, conversely,
that detain part of the energy on some structural group of
bonds, disperse it between these bonds, and in this way
impede energy concentration on the bond that is being
broken. The first structural elements reduce the general
energy quantity necessary for molecular activation (E),
and also reduce the energy and the entropy of the transi-
tion state (A-factor) and the energy that remains on the
reaction fragments. Note that in this case the negative
transition state entropy, i.e., negative entropy of activa-
tion, will be even in the simple bond fission reaction,
which is absolutely impossible in the framework of the
modern statistical Rice-Ramsperger-Kassel-Marcus (RR
KM) theory. In so doing, energy activation also may be
lower than the strength of the bond that is being broken.
This is also impossible in the high-pressure region ac-
cording to statistical theories, whereas in the fall-off
region the usual well-known situation exists. Conversely,
the second type of structural element leads to an increase
in energy activation (E) and the pre-exponent (A-factor).
In this case, the high A-factor and “loose” transition
state may be even at its ringed form, when it has lost a
few degrees of freedom. This is also impossible in the
framework of the modern statistical theories.
This dynamic influence is contrary to the statistical
one in many respects: the “loose” transition state means
a lowering of the rate constant, and the rigid transition
state means a rate increasing. The resultant of the both is
more complicated than the statistical influence only.
This is the real cause of many exceptions to the statisti-

D. Krinkin / Natural Science 3 (2011) 661-682 662
cal RRKM theory (see below).
In the Introduction, both types of the above-mentioned
molecular structures and some semi-empirical rules that
result from them are outlined. In the subsequent parts of
the article are discussed: the agreement between existing
kinetic data and these semi-empirical rules; their limita-
tions and the extent of their applicability; the connection
between them and the existing dynamic method; and
their ability to explain some known exceptions to mod-
ern statistical theories.
The main source of the existing kinetic data is, of
course, the articles in the current kinetic literature. Other
than this, a very comprehensive collection of rate data on
thermally induced unimolecular, homogeneous gas phase
reactions by Sydney Benson and O’Neal [7] should be
noted, which unfortunately includes kinetic data only up
to February 1968, and a fundamental book on uni-
molecular reaction by P. Robinson and K. Holbrook [8],
which, in addition to kinetic theory, contains a signifi-
cant collection of the experimental data.
The proposed method is not an alternative approach to
the existing dynamical one, but rather a qualitative sup-
plement, and a very general approximation. Nevertheless,
it introduces a few new concepts that are very important
for any branch of science, in that it aids our understand-
ing and interpretation of experimental data, and confirms
the direction that further research should take, particu-
larly that using existing conventional methods. It is also
important in that it explains the exceptions to existing
kinetic theories. Probably existing dynamic methods and
some others (particularly quantum scattering theory) will
in the future incorporate this approach to study similar
generalizations, and correct some of them.
We tried to cite the kinetic data of the usual static and
flow method, which are intentionally obtained for the
high pressure region only, and the data of the VLPP
(shock tube technique) method only if they were ob-
tained under strongly comparative conditions, since, in
this case, the values of both the Arrhenius parameters are
slightly dependent on the experimental condition.
The early kinetic data [7,8] were utilized only if their
interpretation was not disproved in later works. We do
not refer to the works in the field of quantum scattering
theory, in particular the numerous works of the Miller
Group (mostly published in the J. Chem. Phys.) in spite
of many common ideas. These works, even for quantum
mechanical transition state theory by the semi-classical
method, represent a statistical but not a dynamical ap-
proach.
The general preconditions. When at its stretching a
bond that is being broken falls into the region where the
attraction between the atoms that it is forming is weak-
ened (the anharmonic part of the potential curve), the
neighbouring bonds are generally shortened as a result of
a decrease in their resistance to this shortening. This
liberates a part of their energy for the bond breaking.
The shortening of the neighbouring bonds and the stret-
ching of the broken bond mutually accelerate one an-
other. A shift of the neighbouring bond atoms as a result
of their shortening is further transferred along the mo-
lecular chains, which may participate in this manner in
the internal energy mobilization of a molecule for the
bond breaking.
A weak bond in the vicinity of stronger neighbouring
bonds falls into its anharmonic region more often than
they fall into theirs. It therefore uses part of their energy
for stretching and breaking more often than conversely.
This weak bond in such a surrounding is an acceptor of
their common vibrational energy. Therefore, the break-
ing of a weak bond in such a surrounding increases the
internal energy mobilization of a molecule. If the reac-
tion takes place in the part of a molecule with stronger
bonds, this reduces the rate of energy mobilization and
promotes the energy dispersion, because the weak bond
absorbs a part of their common energy.
The increase in the internal energy mobilization rate
reduces the energy necessary for a molecule's activation
and also the energy, and consequently the entropy, of the
transition state, i.e., it reduces both Arrhenius parameters:
energy activation (E) and pre-exponent (A factor). The
dispersion energy in a molecule leads to the opposite
result: an increase in both Arrhenius parameters. The
pre-exponent (A) is a more reliable indicator of these
two processes than energy activation, because the latter
is exposed to fewer influences of other factors. A surplus
or lack of energy in the molecule produced as a result of
these two contrary processes is taken away by reaction
fragments with further relaxation.
The mobilization of the internal vibrational energy of
a molecule in the high-pressure region is a purely kinetic
effect, equally accelerating the direct and the back reac-
tion. If the potential energy curve of the direct reaction
has a back slope after the top, i.e., the back reaction has
energy activation and consequently an anharmonic re-
gion, it is accelerated in the same manner and to the
same extent. The absence of the back slope decelerates
both reactions (direct and back). Actually, in the absence
of the back slope on the potential energy curve of the
direct reaction, there is a high probability that, after
achieving the barrier’s top, the separated fragments will
be captured back during the next vibration. The steeper
the back slope, the lower is this probability. In such a
manner, the internal energy mobilization of a molecule
for bond fission in the high-pressure region, reducing
both Arrhenius parameters, takes place in the most
prominent form only when the remarkable activation

D. Krinkin / Natural Science 3 (2011) 661-682
663663
energy of the back reaction is available. This condition is,
probably, not so critical for a ringed transition state.
Here, the energy redistribution (inner relaxation) takes
place inside a ring and the fragments moving off are, as
a rule, neutral molecules but not radicals.
A less incisive example of internal energy mobiliza-
tion is the case of a triatomic molecule whose bonds are
both identical, but whose peripheral atoms have different
masses. When the common vibration energy of two
bonds approaches the detachment energy of one atom
(E), the quicker light atom on average more often falls
into the anharmonic region than does the heavy one, and
therefore more often uses the energy of its bond than
conversely. As a result, in opposition to statistical theo-
ries, the light atom has an additional advantage over the
heavy one in the detachment energy, i.e., it has less en-
ergy activation in spite of the identical potential energy
threshold (E). The energy advantage can lead to a sig-
nificantly greater difference between the reaction rates
of the two isotopes (isotope effect) than only the differ-
ence in mass and velocity. Naturally, this advantage may
be completely realized only at infrequent collisions of
the triatomic molecule with surrounding particles (in the
extreme case when there are no collisions at all), because
these collisions destroy the dynamic interaction between
two bonds and shift the bond breaking reaction in the
region of the usual isotope effects.
The simplest example of the interaction of the two
identical bonds, which is convenient for computer cal-
culation, is represented by a hypothetical triatomic “mo-
lecule,” restricted to a linear configuration and only two
modes of linear vibration (no bending modes). The sta-
tistics of the light and heavy atoms’ detachments from
such a triatomic “molecule” has been studied at different
surplus energies over a potential energy threshold [9].
For each energy surplus (E), many vibrations are per-
formed at the random initial conditions, in order to cal-
culate the average parameters of the detachment sepa-
rately for light and for heavy atoms. The interaction
model and numerical procedure is usually accepted now
for computer calculation of collision-induced vibration
[1,4]. The “molecule” vibrates until one of two atoms,
the heavy or the light one, detaches from the diatomic
molecule remaining. The smaller the energy surplus (E)
is, the greater the prolongation of the vibration. Under
such conditions, the ratio of the frequencies of light and
heavy atom detachments significantly exceeds the
known values of isotope effects. The light atom takes
away on average a greater part of their surplus energy
than the heavy one, i.e., it makes better use of their
common vibrational energy for its detachment. It can be
said that light and heavy atoms compete with one an-
other for a common energy excess and the possibility of
their detachment. The light atom, since its speed is
greater (on average) than that of a heavy atom, benefits
more frequently from this competition. These calcula-
tions have been performed in connection with the ex-
perimental discovery [9] of an anomalously large kinetic
isotope effect for oxygen isotopes separation, which is
also a motive for the present paper.
It is also necessary to consider the structural elements
that impede the energy concentration on the bond that is
being broken. As mentioned above, the significant dif-
ference in the parameters of the neighbouring bonds
promotes the energy concentration on one of them and
impedes the free passing of energy fluctuation along a
carbon chain. The proximity of these parameters (in the
extreme case, their full identicalness) must, conversely,
promote the free circulation of the energy fluctuation in
the system of these identical bonds, retarding its exit
from this system to the bonds with other parameters.
This proximity includes also the identicalness of the
mutual locations of these bonds, i.e., the symmetry of
this bonds system. Such a system of similar or fully
identical bonds is, to some extent, the energy fluctuation
trap, detaining the energy fluctuation on its bonds and
dispersing it between them.
As the branching of the carbon chain increases the
number of identical bonds linked to one central atom, so
also a ring having similar or fully identical bonds is rep-
resentative of such a similar bonds fluctuation trap. In
the first case of the branched chain the effectiveness of
the energy dispersion (demobilization) has to increase in
the row: primary, secondary, tertiary, quaternary (neope-
ntane type) carbon atom. In addition, a carbon atom, in
particular the other elements of the IV group of the pe-
riodic table, Si, Sn, Pb, may be in the center of the
branching. In the second case of a ring that can include
not only a carbon but also a heteroatom, the effective-
ness of the demobilization process will be increased with
more light movement of energy fluctuation inside the
ring. This lightness depends on the proximity not only of
the bonds’ strengths but also of other parameters, first of
all frequencies, masses of atoms, etc.
The case when the bond that is being broken is a part
of the fluctuation trap corresponds to the most effective
demobilization. When the bond that is being broken is
not found in the fluctuation trap, the effectiveness of the
demobilization depends on the structure and dimension
of those parts of a molecule that are found between them.
The more free energy circulating between these two
parts of a molecule, the more the effectiveness of the
demobilization increases.
Structural elements of an organic molecule, those that
increase the energy mobilization and also those that re-
duce it, influence both the Arrhenius parameters: energy

D. Krinkin / Natural Science 3 (2011) 661-682 664
activation (E) and pre-exponent (A) uniformly, i.e., ei-
ther increase or reduce both of them. In the framework
of the effect considered in the present paper, both Ar-
rhenius parameters are closely bound one to another
quantitatively. Indeed, the mobilization of energy quan-
tity E from the internal energy of a molecule reduces
simultaneously, for the same value (E), both the energy
activation (E) and the energy of the transition state. A
decrease in the transition state energy for this value (E)
means also a decrease in its entropy for the value of
E/T, pre-exponent (A) by a factor eE/T and logA for
value E/2.3 RT. In the case of demobilization, all these
values will be increased for the same quantities. In such
a manner, the considered effect is bound to positive lin-
ear correlation between the logarithm of pre-exponent
(A) and energy activation (E) with a correlation coeffi-
cient (b) equal to 1/2.3.RT. It is, of course, the approxi-
mate evaluation of this coefficient, based on the statisti-
cal regularities, which are not completely applicable for
the dynamic picture that is being considered here. It can
be shown that this is the top evaluation. At the real lower
values of this coefficient (b), the low values of both the
Arrhenius parameters correspond to the high values of
the reaction rate (mobilization) and, conversely, the high
values of both of them correspond to the low rates (de-
mobilization).
It is interesting in this connection to consider the in-
fluence of different substituents in a molecule on both
the Arrhenius parameters of the same reaction, that is,
the correlation between both Arrhenius parameters of the
one group of derivatives. The absolute majority of de-
rivatives really shows a positive correlation between the
energy activation (E) and the pre-exponent (A), which in
the coordinates logA ~ E can be expressed as logA = a +
bE. Such graphs are quite informative. The derivatives
with structural elements, which promote internal energy
mobilization, in many cases are actually located in the
region of the low Arrhenius parameters at the left low
end of the correlation line, where the reaction rate is
higher, and those which promote demobilization are lo-
cated in the right high end. Unfortunately, the signifi-
cance of this information is limited by the influence of
other effects, as much on the energy activation (± I ef-
fect, conjugation) as on the pre-exponent (change in the
transition state degree of freedom). In spite of these
complications, the consideration of such graphs is useful,
because the significant deviation from the correlation
line is practically always very informative, as for in-
stance, at the change of the reaction mechanism.
In spite of the generally qualitative character of the
proposed method, as applied to derivatives it also allows
some quantitative evaluations. This is particularly true of
a qualitative evaluation of the energy quantity remaining
in the case of demobilization on the one degree of free-
dom for different types of structural elements (see be-
low). These structural elements preserve a more or less
similar value of internal energy surplus in the different
chemical substantives, a fact that indicates the objective
character of these evaluations.
The new concepts introduced by the proposed method
are internal energy mobilization and demobilization and
the regularity of the fluctuation energy passing in a
molecule through different bond types; its reflection, dis-
persion and concentration on a certain type of bond seem
somewhat unusual, although they represent the funda-
mental properties of the chemical kinetics. The appear-
ance of a new relevant concept is always very important
for any branch of the science. This paper is devoted to
the elucidation of relevance of the new concepts.
The main items of the proposed concepts may be
formulated briefly in the following way.
1) An organic molecule has structural elements that
promote the internal energy mobilization for its uni-
molecular decay in the high-pressure region, reduce both
Arrhenius parameters, and increase the reaction rate, and
elements that, conversely, detain part of internal energy,
disperse it, and impede its concentrations on the bond
that is being broken. This second type of structural ele-
ments increases both the Arrhenius parameters and re-
duces reaction rate.
2) The mutual arrangement of these two types of
structural element in the particular organic molecule
produces the row regularities that are discussed in this
paper.
3) A weak bond in the vicinity of stronger neighbour-
ing bonds is an acceptor of their common vibrational
energy in order to break. This bond represents the first
type of structural elements that promote an internal en-
ergy mobilization.
4) The number of the linked identical bonds: the
branching of the carbon chain or the bonds’ ring repre-
sent the second type of structural element, which dis-
perses internal energy between identical bonds and im-
pede its concentration on the bond that is being broken.
In the row: the primary, secondary, and tertiary carbon
atom increases the number of linked identical bonds and
increases also the energy dispersion.
5) The two types of the structural element, those that
increase the energy mobilization and also those that re-
duce it, influence both the Arrhenius parameters uni-
formly, i.e., either increase or reduce both of them si-
multaneously. This binds quantitatively energy activa-
tion (E) and A-factor (A) for the system of the deriva-
tives. Each derivative either increases or reduces both of
them. This correlation between both Arrhenius parame-
ters for the derivatives system produces the row regu-

D. Krinkin / Natural Science 3 (2011) 661-682
665665
larities discussed in the present paper.
The discussion begins with examples of the mobiliza-
tion of the internal energy of a molecule and continues
with demobilization examples. Between the presenta-
tions of these two groups of examples, the kinetic data,
which represent both tendencies, are presented. The in-
clusion of examples that are representative of both ten-
dencies somewhat complicates the construction of the
text.
2. DISCUSSION
2.1. The Thermal Decomposition of the
Hydrazines and the Amines)
These are simple bond fission reactions. The compre-
hensive collection of kinetic data on the decomposition
of the hydrazines and amines and their analysis is con-
tained in the volume by Benson and O’Neal [7] men-
tioned above. First, they called attention to the fact that
both Arrhenius parameters of these reactions are ex-
tremely low, being very far beyond the framework of the
modern statistical theories. The authors do not find an
explanation for this. They write: “The A-factors of the
hydrazine and the amine decompositions are all less than
the normal ekT/h values; some are significantly lower.
Such a situation implies negative entropies of activation,
which, in the case of simple bond fission reactions is just
not reasonable.” (page 34) The negative entropy of acti-
vation for bond fission reaction is typical of the in-
creased energy mobilization in the high-pressure region,
according to the effect being considered (see Introduc-
tion). High A-factors are presently more consistent with
the modern theory of simple bond fission reactions, than
are low A-factors. Such reactions as a rule have logA-
factors higher than “normal” values (13.5 - 14), being
about 15 - 16. At the same time the experimental values
logA-factors of the hydrazines and the amines are as a
rule lower than this “normal value, being 11.7 - 13.2 for
hydrazines and a little bit higher for amines – 12.8 - 13.4,
which means about 3 orders of magnitude lower than the
majority of other reactions with a single bond rupture.
The simple bond fission does not demand a particular
transition state conformation, reducing its degrees of
freedom and accordingly reducing its entropy and
A-factor. Therefore the low values of A-factor mean also
a significant reduction in the rotational and vibrational
energy remaining on the molecule and on the transition
state, i.e., in the significant internal energy mobilization
for bond breaking in the high-pressure region.
Presently, the activation energy (E) of the bond fission
reactions is universally accepted as the reaction enthalpy,
i.e., the kinetic factor is accepted as a thermodynamic
one, which leads to many complications. This also
means that the most common assumption has been that
the activation energy for radical-radial recombination
reactions is accepted as zero. This is more or less true
with some precision for many bond fission reactions. At
the same time, experimental values of the energy activa-
tion for the hydrazines decompositions are less than the
reaction enthalpy by 10 - 14 kcal/mol and for the amines
by 7.5 - 11. These values would probably represent the
highest discrepancy of their kind known. This means
that the activation energy of a molecule of a hydrazine or
an amine is significantly lower than the strength of the
bond that is being broken, i.e., a part of the energy for
bond breaking arrives from the internal energy of a
molecule. This also confirms that the internal energy
mobilization in the high-pressure region is in agreement
with the considered effect.
Both the conditions for bond fission reaction that are
accepted now, i.e., the high value of the A-factor and the
energy activation values exceeding bond strength in the
high-pressure region, are the consequence of the statisti-
cal character of the modern kinetic theories. This char-
acter excludes any tendency for energy redistribution
inside a molecule. This means that the transfer of an
energized molecule to a transition state is governed by
purely statistical consideration and its bond interaction,
which is studied by dynamic methods, is practically ig-
nored. Such indeed is the main concept of the modern
statistical RRKM theory that defines the general ap-
proach of the collection’s authors. Since the reported
experimental data for hydrazines and the amines contra-
dict significantly the statistical transition state theory, the
collection’s [7] authors presented also the preferred val-
ues of both the Arrhenius parameters (E and A-factor)
adapted to this modern theory in such a way as to pre-
serve the rate constant. They write: “Preferred values
will generally be those judged to be most consistent with
the experimental reaction rate and transition state the-
ory” (page 5). The usual procedure has been to equate
the activation energy to the reaction enthalpy and to in-
crease also the A-factor to preserve the reaction rate. In
this case, of course, the agreement between experimental
and adapted reaction rates was preserved only at the
mean reaction temperature, with a discrepancy for higher
and lower temperatures.
The thermal decomposition of the hydrazines and
amines, whose kinetic data are in the collection [7], were
performed in the gas phase at the temperature 750 -
950˚K. The toluene carrier and aniline carrier techniques
were employed. The thermal decomposition of hydra-
zine derivatives takes place by N-N bond breaking, in
accordance with the scheme:

D. Krinkin / Natural Science 3 (2011) 661-682 666
The rates were based on the production of ammonia.
The reaction of 1,1-dimethylhydrazine. The experimen-
tal data: energy activation (E) is 49.6 kcal/mol; loga-
rithm of pre-exponent logA = 13.2; logarithm of con-
stant rate kT (sec–1) logkT = –0.88 (770˚K). The heat of
the formation of the produced radicals (heat of formation
or broken bond strength) is calculated by the authors as
H˚ = 63.0 kcal/mol. The authors notice: “Rate con-
stants are probably reliable. Arrhenius parameters are
low” (page 442). In such a manner, in this case, energy
activation (E) is lower than bond breaking energy by
13.4 kcal/mol (a quarter of its value). The authors in-
creased the preferred value of energy activation (E) up to
about the heat of formation (H˚), i.e., up to 62.7
kcal/mol. They were forced to increase logA-factor up to
16.9, i.e., by 3.7 units, in order to preserve the rate con-
stant at the mean temperature (805˚K). This means the
A-factor increases by five thousand times. For these
preferred Arrhenius parameters, the value of the rate
constant at 805˚K is preserved quite precisely.
Similar situations with other hydrazine derivatives.
The experimental data of the methyl hydrazine decom-
position are: energy activation is 51.9 kcal/mol; logA is
13.19 units; the calculated heat of formation H˚ is 65.6
kcal/mol. The authors increased the preferred value of
energy activation up to about the heat of formation, 64.8,
i.e., by 12.9 kcal/mol, and logA by 3.51 units up to 16.7
to preserve the reaction rate at the mean of temperature
region.
Phenylhydrasine decomposition. The experimental
data: E = 40.0 kcal/mol; logA = 11.8. The calculated
heat of formation H˚ = 50.7 kcal/mol is 10.7 kcal/mol
more than experimental values of energy activations.
The authors increased again the preferred energy activa-
tion up to about the heat of formation –51.1 kcal/mol
and logA up to 15.5 unit, in order to preserve the reac-
tion rate.
The main conclusion drawn from the kinetic data for
hydrazine decomposition is that energy activations are
significantly less than the strength of the bond that is
being broken and A-factors are significantly less than
“normal” values for bond fission reactions. The thermal
decomposition of amines takes place with C-N bond
breaking.
Here also, the experimental values of energies activa-
tion are lower than the breaking bond strengths and the
A-factor is lower than “normal” values for the bond fis-
sion reaction, with a slightly smaller difference than for
hydrazines. The experimental data for n-methylbenzi-
lamine are: energy activation E = 57.7 kcal/mol; loga-
rithm A-factor 12.86 units; calculated heat of formation
H˚ = 65.2 kcal/mol. The authors increased the preferred
value of the energy activation a bit more than heat of
formation up to 69.0 kcal/mol, i.e., by 11.3 kcal/mol and
logA up to 15.7 i.e., by 2.84 units, to save the reaction
rate.
The experimental Arrhenius parameters of n-methy-
laniline: E = 60 kcal/mol and logA = 13.4 units. The
authors increased the preferred values to E = 67.7 and
logA = 15.3. The preferred E value is close to heat of
formation H˚ = 68.3 kcal/mol. In both cases, the au-
thors note that: “The rate constants are probably reliable.
Arrhenius parameters are suspect”.
The strength of the N-N bond in the hydrazine deriva-
tives is about half of the surrounding C-C bonds. Ac-
cording to the considered effect, such a difference must
lead to significant internal energy mobilization of a mo-
lecule for the weak bond that is being broken. In the case
of hydrazines, the weak bond (N-N) is an energy accep-
tor and the strong bonds (C-N, C-C) are the energy do-
nors. In the case of amines, the weak bond is C-N and
the strong bonds are C-C. The internal energy mobiliza-
tion must, in its turn, lead to a significant reduction in
the energy that is necessary for molecular activation (E)
and the entropy of activation.
The experimental data of the hydrazines and the
amine decomposition give a few independent reasons for
such a mobilization in the high-pressure region:
1) The energy activation of a molecule is lower than
the strength of the bond that is being broken.
2) The A-factors for the hydrazines and amines de-
composition are actually lower than A-factors for most
bond fission reactions by about three orders of magni-
tude.
3) As mentioned in the Introduction, in the framework
of the considered effect there must be a certain depend-
ency between the lowering of the energy activation (E)
and that of the A-factor. The average energy activation
of hydrazines decomposition is lower than the breaking
bond strength by about 13 kcal/mol. According to coef-
ficient 1/2.3RT, the lowering of logA by 3 - 3.5 units
can be expected. The average experimental value of
logA for hydrazines is about 12.5, which is actually less
than “normal” values (15 - 16) for bond fission reactions
by this value (3 - 3.5).
The situation for the amines is similar. Here the aver-
age experimental value of E is about 9 kcal/mol, which

D. Krinkin / Natural Science 3 (2011) 661-682
667667
gives a correction for logA of 2 - 2.5 units. The experi-
mental values of logA are 13 - 13.5, which is approxi-
mately less than the “normal” values (15 - 16) by this
amount (2 - 2.5).
It is pertinent to note that the authors of the cited vol-
ume [7] adapted the experimental values of the
logA-factors for the hydrazines and amines, practically
up to these “normal” values for bond fission reactions
(15 - 16).
4) So sharp an internal energy mobilization (~13 kcal/
mol or about a quarter of the total activation energy)
probably demands a remarkable energy activation for the
back reaction of radicals’ recombination, as is discussed
in the Introduction. Benson and O’Neal in their volume
[7] collect also recombination rate constants for nitrogen
radicals. All nitrogen radicals whose recombination
produces hydrazines and amines have values of loga-
rithm rate constant (l/mol.sec) ranging from 5.8 to a
maximum of 7.5 (lower values for hydrazines); the
structurally similar hydrocarbon radicals have values
ranging from 8.7 up to 9.8. Other nitrogen radicals also
have these values ranging from 8.6 to 9.8. The recombi-
nation rate of radicals, which produce hydrazines by a
3.0 - 3.5 and amines by a 2.0 - 2.5 order of magnitude, is
less than the absolute majority of other radicals, whose
recombination energy activation are now accepted as
being zero. This means that the radicals that produce the
hydrazines and amines have a remarkable obstacle to
their recombination in the form of significant energy
activation. These data are not fully independent evidence
of internal energy mobilization because some of them
are calculated from the results of the hydrazines and
amines’ decomposition.
5) The recombination rates data of the radicals that
produce the hydrazines and amines make it possible to
estimate the energy activation (E) of these reactions. The
factor e–E/RT reduces the values of recombination rate by
a 3 - 3.5 order of magnitude for hydrazines and by 2 -
2.5 for amines. The energy activations (E) of the radical
recombination from this equation are very close to the
internal energy mobilizations of the decomposition, i.e.,
about 13 kcal/mol for hydrazines and about 9 - 10 for
amines. Such a relationship between the kinetics of the
direct and back reactions must be available in the
framework of the considered effect. This means that
direct and back reactions are accelerated similarly (a
purely kinetic effect).
The true cause of the remarkable energy activation for
nitrogen radicals’ recombination lies far beyond the
scope of the present article. Only one of a few possibili-
ties can be mentioned. A hydrazine molecule is consti-
tuted of two dipoles. Their positively charged ends (ni-
trogen atoms) are bound one to the other by the N-N
bond. The repulsion of these dipoles after N-N bond
breaking forms a back slope on the potential energy
curve of the N-N bond. It is also the reason for the acti-
vation energy arising in the back recombination reaction.
The dipole is formed by a positively charged nitrogen
atom that is found in the vertex of a trigonal pyramidal
structure and a negatively charged single non-bonding
pair of electrons in its base (sp3 hybridization). The two
orbitals of the nitrogen covalent bonds and one orbital of
nitrogen’s lone electron pair form the three arms of this
trigonal pyramidal structure.
The existence of a few independent pieces of evidence
of internal energy mobilization for the bond that is being
broken in the high-pressure region for the hydrazine and
amine makes this mobilization mechanism quite reason-
able. It contradicts statistical theories and reflects the
general dynamic picture of the bonds interaction ac-
cording to the proposed concept.
To the best of our knowledge, the very serious dis-
crepancy mentioned above between theory and experi-
ment for hydrazine and amine [7] has not been satisfac-
torily explained until now.
It is quite possible that the consideration of the dis-
cussed effect applied to other substances and reactions
will bring about a revision of the same assumptions,
which are accepted now a priori:
1) The high values of both Arrhenius parameters for
bond fission reactions;
2) The equality of the reaction enthalpy and the en-
ergy activation for the bond fission reaction and the
consequent absence of activation energy for the back
reaction of radicals recombination;
3) Other conclusions in the kinetics and thermodynam-
ics, that are bound to the above mentioned assumption.
2.2. Decomposition of Nitroalkanes
The thermal decompositions of the nitroalkane pro-
ceed through a five-center transition state, producing
olefines and nitrous acid:
The temperature region for this investigation is, as a
rule, 600 - 700˚K. The reaction rates are surface-sensi-
tive, and therefore the reaction chamber walls were con-
ditioned and, if necessary, radical chain reactions were
inhibited, to be sure that the residual reaction is molecu-
lar. The reduction in activation entropy in these reactions

D. Krinkin / Natural Science 3 (2011) 661-682 668
is a natural result of the restriction in the internal rotation
of a transition state. At the same time, this reduction was
significantly greater than can be theoretically estimated
on the basis of the statistical transition state theory,
(RRKM). Benson and O’Neal write in their collection
[7]: “By the transition state methods used to estimate
A-factor for the four- and six-center reactions, the
A-factor estimates of the five-center reactions are be-
tween one and two orders of magnitude higher than
those reported” (page 12).
The reported value of logarithm A-factors for the
thermal decompositions of nitroethane are 11.35; 11.53;
10. 83; 13.0; 12.4. Benson and O'Neal write: “The ex-
perimental A-factors require an extremely tight activated
complex (i.e., almost total loss of both internal rotations).
By analogy with six-center transitions state reactions,
one would expect an A-factor in excess of 1013 sec–1”
(page 130), i.e., logA is more than 13. The experimental
A-factors for 2-nitropropane decomposition are 11.34;
11.05; 11.05; 11.30. Benson and O’Neal write: “Ar-
rhenius parameters appear unreasonably low” (page 132).
In the reaction of nitroalkane decomposition, the weak
C-N bond (73 kcal/mol) is broken in the surrounding of
stronger bonds: the double N=O bonds (about 140 kcal/
mol) from one side and one or two C-C bonds (83
kcal/mol each) from the other. The non significantly
stronger C-C bond, however, in the process of this reac-
tion is significantly reinforced at the sacrifice of the
weakening C-N (scheme above) and is converted into a
double C=C bond (145 kcal/mol), in just the most criti-
cal position, defining the internal energy mobilization of
a molecule. So that, in this position, when the configura-
tion of a nitroalkane molecule approaches the transition
state conformation, the weakened C-N bond that is being
broken is surrounded by very strong bonds.
In addition, the limitation about the obligatory pres-
ence of the energy activation for a back reaction as for a
simple bond fission is absent here. The energy redistri-
bution here takes place inside a transition state ring, and
the forming molecules, as distinct from radicals, always
have remarkable energy activation for a back reaction.
This leads one to expect significant internal energy mo-
bilization and consequently a reduction in the A-factor,
as has been shown experimentally for nitroalkanes.
This is an example of the weak bond breaking in the
neighborhood of stronger ones in the ringed transition
state, which leads to additional internal energy mobiliza-
tion and A-factor reduction.
2.3. The Thermal Decomposition of
Tetra-Alkyl Orthosilicates (Silicon
Alkoides)
The tetraalkyl orthosilicates Si(OR)4 and other ele-
ments of the IV group (C, Ge, Sn, Pb) are also important
for kinetic investigation, because they make it possible
to vary widely both the nature of the central atom and
also the radical (R) structure.
The replacement in the Si(OR)4 of methyl radical (R)
instead of ethyl leads to very significant kinetic conse-
quences. For this reason Chu, Breslin, Wang, Lin, and co.
investigated [10,11] the relative stabilities of tetramethyl
(TMOS) and tetraethyl orthosilicate (TEOS). The pyro-
losis was carried out between 858 and 968 K for TMOS
and between 721 and 820 K for TEOS, with a very low
concentration of the reagents, using argon as diluent and
near atmospheric pressure to minimize the reactor wall
effect and to obtain a reliable first order rate constant in
the high-pressure region.
The rate constants for TMOS can be effectively rep-
resented by k = 1.4.1016 exp(–81200/RT) sec–1, conse-
quently logA = 16.15 and E = 81.2 kcal/mol. The rate
constant for TEOS is k = 7.4.1010 exp(–49500/RT) sec–1
steps in these two systems are very different. The
A-factor of the TMOS reflects a very loose transition
state, typical for a bond-breaking process producing
radical products: Si(OCH3)4   CH3 +  OSi(OCH3)3.
The low A-factor and activation energy measured for
TEOS suggest that the initial decomposition may occur
by a six-centered complex:
The possibility of such a mechanism is connected to the
presence of
 -hydrogen in the TEOS molecule. This
mechanism is similar to the known 6-centered molecular
decomposition process for diethyl carbonate:
whose unimolecular decomposition constant has been
measured to be K = 7 × 1013 exp(–4600/RT) sec–1. The
authors are confused that they obtained an A-factor of
the TEOS significantly lower (three orders of magnitude)
than the A-factor of diethyl carbonate and other six-
centered reactions. They refer also to the opinion of
Benson and O’Neal that the typical A-factors for the
6-centered molecular elimination process are 1013 - 1014

D. Krinkin / Natural Science 3 (2011) 661-682
669669
s–1 but not 7 × 1010. They suspect the secondary reac-
tion’s influence is an explanation for this discrepancy.
Benson and O’Neal write in section 4.1, which is de-
voted to six-centered elimination reactions: “Observed
entropies activation are about half this expected amount;
therefore cyclic transition states must be reasonably
loose” (pages 12-13).
The strength of the C-O bond that is being broken is
80 - 85 kcal/mol. The strength of the neighbouring Si-O
bond is not less than 100 kcal/mol [10] and it is signifi-
cantly reinforced in the process of this reaction, transfer-
ring into double bond Si=O. Therefore, the weak C-O
bond being broken is neighbouring on a very strong
bond, exactly at the critical moment when the former is
found in its anharmonic region. In such a manner, the
mobilization of the internal energy of a molecule in the
high-pressure region may quite easily be the cause of the
unusually low values of the A-factors for these reactions.
Both the Arrhenius parameters of the TEOS pyrolysis
are significantly lower than the TMOS one, and the con-
stant rate is significantly higher. The transition to a ki-
netically more advantageous mechanism leads to a re-
duction in both the Arrhenius parameters in the same
fashion as the influence of the considered effect. How-
ever, this transition is accompanied by a much more sig-
nificant reduction in energy activation in relation to the
reduction in the A-factor. This leads naturally to a very
remarkable rate constant increase. Therefore, this transi-
tion in the coordinates logA~E for some derivative sys-
tems will be characterized by a deviation from the cor-
relation line to the side of the higher A-factor at the
same energy activation, i.e., up and leftward. Such a
deviation is a clear indication of the transition to another
reaction mechanism.
Actually, in the transition from the TMOS pyrolysis to
the TEOS one, the energy activation is reduced by 31.7
kcal/mol. The correlation coefficient between E and
logA (1/2.3RT) equals  0.24 mol/kcal for an average
experimental temperature of 900 K. The reduction in the
logA may be expected, in the framework of the consid-
ered effect, to be by a value of 31.7 × 0.24 = 7.7 units,
i.e., from 16.15 down to 8.45 units. The real value of the
logA-factor is actually remarkably higher –10.87 units.
For derivatives decomposing as a TMOS according to
the simple bond fission mechanism, the represented
TEOS point must be significantly deflected from the
correlation line to the upper left side, which will be the
indication of reaction mechanism change.
This is also an example of the weak bond breaking in
the ring in the vicinity of the stronger one, which leads
to a significant reduction in the Arrhenius parameters.
It is also an example of the change in the reaction
mechanism drawing the derivate out of the correlation
line.
2.4. The Influence of the Continuous
Chain Length on the Internal
Energy Mobilization
It is of interest to consider the substances whose
molecules contain both structural elements, which, just
as they promote internal energy mobilization, also sup-
press it. Among these are, in particular, alkanes contain-
ing the t-butyl group, which is bound with a continuous
chain. The strongly branched t-butyl group
detains energy fluctuations on its bonds (demobilization)
while the continuous chain promotes the internal energy
concentration on its bonds (mobilization).
The availability of the back slope on the potential en-
ergy curve in this case of the two contrary influences is
probably less critical than for other cases of mobiliza-
tion.
The influence of the length of the continuous chain
radicals (R) on the main kinetic parameters of the bond
breaking between this radicals and the t-butyl in the mo-
lecules with a common formula, R-t.butyl, is studied by
W. Tsang and J. Kiefer [12] and [13]. The main parame-
ters of this reaction for different alkanes with this com-
mon formula were studied under strongly comparative
conditions. This increases the precision of the compari-
son and may reveal a tendency that is of interest to us,
even though the difference between the results is not big.
The comparative data for these alkanes’ decomposi-
tion were determined from a single pulse shock tube
study at a temperature of 1000 - 1200˚K. This very
low-pressure pyrolysis (VLPP) method is a very good
one for obtaining reliable first-order parameters in the
high-pressure region for simple bond fission reactions.
The bond fission, an initial reaction in decomposition, is,
as a rule, complicated by the numerous secondary reac-
tions of the radicals that are produced. The good solution
is a very short reaction time and accordingly a very low
rate of conversion and concentration of the radical that is
produced. The single pulse shock tube was designed and
constructed to achieve extremely short reaction times
and conversions.
The logarithm of A-factors for methyl, ethyl and
n-hexyl groups (R) were: 17.0; 16.778; 16.477. These
values have a tendency to reduce with the increase in the
continuous chain’s length. The situation of energy acti-
vation is similar: 81.4; 76.8; 74.1 kcal/mol. In spite of

D. Krinkin / Natural Science 3 (2011) 661-682 670
the moderate difference between A-factors, the common
tendency is clear. The continuous chain lengthening in-
creases the internal energy mobilization.
The chain increase for one carbon atom from methyl
to ethyl radicals reduces the logA by 0.22 units. The
subsequent chain length increase reduces more moder-
ately. This influence fades along the chain’s length.
In this case, the chain’s increase for one carbon atom
close to the breaking bond is equivalent to an internal
energy mobilization equal to 2.3logA RT = 2.30.22
RT  0.5 RT. This is, of course, a very rough estimation.
This is an example of the carbon continuous chain
lengthening promoting the internal energy mobilization.
2.5. The Thermal Decomposition of the
Cyclobutane and Its Derivatives
This is a homogeneous unimolecular reaction yielding
two molecules of olefins. The reaction is expressed by
the following scheme:
Presently, the biradical mechanisms are commonly ac-
cepted, in particular because simultaneous breaking of
two opposite ring bonds does not agree with the high
A-factor. The logarithm of A-factors as a function of
energy activation (E) for cyclobutane derivatives with
this common formula R-C4H7 is shown in Figure 1. The
reaction is studied in the temperature region 695 - 740˚K.
The connection between these two parameters is very
well expressed by the correlation line:
logA = 2.12 + 0.216E,
with a moderate deviation from it. The correlation coef-
ficient equal to 0.305 corresponds to the preservation of
the constant reaction rate along the whole correlation
line. The real, lower correlation coefficient, 0.216, cor-
responds to the reaction rate increase in the left lower
part of the correlation line, where the simultaneous de-
crease of both the Arrhenius parameters reflects the mo-
bilization of the internal molecule energy for bond
breaking. Indeed, at a temperature of 720K the rate con-
stant for methylcyclobutate (point 1) in the lowest part
of the correlation line is 5.5 × 10–4 sec–1. At the same
time, t-butylcyclobutate (6) in the highest part of the
correlation line has a lower rate value: 2.58 × 10–4 sec–1.
All the derivatives that appear in Figure 1 can be di-
vided into two groups, those lying below and those lying
above cyclobutane (4), i.e., more to the left and more to
Figure 1. The logarithm of A-factor as a function of energy
activation (E) for cyclobutane derivates: 1-methylcyclobutane;
2-propylcyclobutane; 3-ethyl cyclobutate; 4-cyclobutate; 5-
isopropylcyclobutane; 6-t-butylcyclobutate; 7-cyclobutane d8.
the right of the line AB. The derivatives with continu-
ous-chain substitutes methyl, ethyl and proplycyclobu-
tanes lie to the left of and below cyclobutate (4), in the
region corresponding to the internal energy mobilization
of a molecule for the ring bonds breaking. Both the Ar-
rhenius parameters in this region are lower and the reac-
tion rate is higher than for derivatives on the right side of
the correlation line. A substitute branching shifts a de-
rivative right and upward along the correlation line. The
most branched t-butylcyclobutate is found in the almost
highest position (6); the less branched isopropylcy-
clobutane is found lower (5).
Cyclobutane has a high rate of symmetry and com-
pletely identical C-C bonds, which, according to the
effect being considered, promotes the dispersion of en-
ergy fluctuation into the ring and hampers the energy
concentration on its bond. The cyclobutane ring is a fac-
tor of the demobilization. Any substitute reduces its
symmetry rate and must therefore reduce the A-factor. In
spite of this, t-butyl significantly increases them (6), that
reflects the very high-energy dispersion ability of the
strongly branched chain. These two opposing influences
on the less branched isopropylcyclobutane (5) only re-
ciprocally counterbalance one another, so that it is found
practically on the same part of the correlation line as
cyclobutane (4).
Some conclusions about the dynamical influence of
the C-H bonds on the C-C ones can probably be drawn
from these data. This influence, from the viewpoint of
mobilization as well as that of demobilization, has to be
less than the mutual influence of C-C bonds, because
their parameters, and primarily all masses and frequen-
cies, are significantly different. The substitution in a
hydrocarbon of a hydrogen atom for deuterium ap-
proaches a little these parameters. In this connection, a
deuterated compound has to increase its dynamic influ-

D. Krinkin / Natural Science 3 (2011) 661-682
671671
ence (mobilization or demobilization) in comparison
with an initial hydrocarbon. The deuterated cyclobutate
(7) has a remarkably higher rate of demobilization, and it
is found significantly higher on the correlation line.
The broadening of a derivative group included in the
correlation picture increases, as a rule, the results’ devia-
tion from the correlation line. The inclusion of the de-
rivatives with two substitutes, for example 1.2 dymethy-
lcyclobutanes, in the above correlation as well increases
the deviation from the correlation line, to a large extent
for trans-compound and to a lesser extent for cis.
The 1,1,3,3-tetramethylcyclobutane, with its high rates
of symmetry and a branching, is found practically on the
continuation of the correlation line in the demobilization
region (dotted line). This cyclobutane derivative, accord-
ing to the effect being considered, actually must show a
high rate of demobilization.
The influence of the continuous-chain substitutes
(1,2,3) in this case, except of a mobilization, is probably
above all expressed in the reduction in the demobiliza-
tion influence of the cyclobutate (4). So that the deriva-
tive for which both the influences mobilization and de-
mobilization are counterbalanced lies near to these three
derivatives (1,2,3) or even lower on the correlation line.
This frontier derivative can have both its Arrhenius pa-
rameters maximum close to their estimation based on
statistical theories (for example, RRKM), without the
dynamical influence.
Generally, the analysis of such a correlation picture
for a derivatives system may give some kinetic informa-
tion that is useful, and even more specific for a narrower
group of derivatives.
There are here the derivatives that, according to the
proposed concept, promote the internal energy mobiliza-
tion and those that impede it. The first group lies in the
lower part of the correlation line (with low E and A val-
ues) and the second one in the higher part.
2.6. The Pyrolyses of Esters to Give Olefins
plus Carboxylic Acids
The esters decompose as a rule by purely molecular
mechanisms with some exceptions for primary esters,
which demand that a concurrent chain reaction be sup-
pressed by inhibitors for this. For the unimolecular
process, the six-centered transition state is commonly
accepted:
Inside the transition state ring, a very significant vi-
brational energy redistribution takes place. The strength-
ening bonds C-O and C1-C2 and the forming O-H bond
transfer their released energy to the C1-H and C2-O
bonds that are being broken (the energy flow is indicated
in the scheme by arrows). In the acid part of the ring, the
simple C-O and the double C=O bonds exchange places.
This is related to the energy exchange between them.
This affects the other bonds of the transition state ring to
only a moderate extent. The breaking of the bonds (C1-H
and C2-O) takes place in the alcohol part of the ring. For
this reason alone it is possible to assume that the mo-
lecular structure of the alcohol part of an ester influences
the ester’s decomposition more significantly than the
structure of the acid part (see below).
All the energy released at the conversion of the simple
C1-C2 bond into the double C1=C2 bond (~60 kcal/mol)
and the O-H bond formation (~100 kcal/mol) passes
through the carbon atoms C1 and C2 on the bonds that
are being broken, C1-H and C2-O. In the framework of
the effect considered here, the branching at these carbon
atoms C1 and C2 must have a significant influence on
this passage, hampering it. For example, the split of the
sole energy flow through the C1 carbon atoms in the di-
rection of the side chains, R3 and R4 (see scheme) re-
duces the energy entering on the bonds that are being
broken in the ring, C1-H. A similar situation exists with
regard to the C2 carbon atom and the branching into the
R1 and R2 side chains. In other words, the reduction in
the alkyl substituent’s numbers around the double C1=C2
bond in the reaction product accelerates the production
of this product from the viewpoint of the effect being
considered here. In such a manner, the effect explains
the well-known Hofmann rule of which this is a special
case. Robinson and Holbrook [8] formulate this rule:
“For esters which are capable of decomposing in differ-
ent ways, the predominant direction of elimination is
given by the Hofmann rule, i.e., the favored olefinic
product is that bearing the fewest alkyl substituents
around the double bond.” The influence of the usual
steric effect is also possible in this case.
The correlation picture of the esters’ decomposition in
the coordinates logA ~ E is split into three almost paral-
lel separate lines depending on the molecular structure of
the alcohol part of the molecule where the bond breaking
is taking place. The deviation from the correlation lines
is moderate, whereas the difference between them is
very significant, and it shows the very significant influ-
ence of the structure of the alcohol part of a molecule,
Figure 2. In so doing, the derivate’s position is practi-
cally almost independent of the structure of the acid part
of its molecule. The lowest correlation line (AB) corre-
sponds to the continuous chain alcohols and has minimum

D. Krinkin / Natural Science 3 (2011) 661-682 672
Figure 2. The correlation between high-pressure Arrhenius
parameters for the pyrolyses of the esters, to give olefins and
carboxylic acids. Designation: -  - continuous-chain alcohols;
-  - isopropyl alcohols ; -  - branched-chain alcohols (at the
tertiary carbon atom); -  - cycloalkane alcohols; -- aromatic
alcohols; -  - t-butyl chloroacetates. 1-propyl formate (9.4;
39.7). 2. ethyltrimethylacetate (11.2; 44.0). 3. ethyl formate
(11.3; 44.1). 4. ethyl acetate (11.6; 44.0). 4. ethyl acetate (12.5;
47.8). 5. ethyl propionate (12.7; 48.5). 6. ethyl-d5 acetate (12.7;
49.5). 7. isopropul acetate (12.1; 42.9). 8. isopropyl formate
(12.4; 44.2). 8. isopropyl formate (12.6; 44.0). 9. isopropyl
trimethylacetate (12.9; 44.8). 7. isopropyl acetate (12.9; 46.1).
7. isopropyl acetate (13.0; 45.0). 10. sec-butylacetate (13.3;
46.6). 7. isopropylacetate (13.4; 46.3). 11. t-butyl formate (11.1;
34.6). 12. t-butyl propionate (12.8; 39.2). 12. t-butyl acetate
(13.3; 40.5). 14. t-amyl acetate (13.4; 40.3). 13. t-butyl acetate
(13.5; 42.1). 15. cyclopentyl acetate (12.9; 44.1). 16. cyclopen-
tyl acetate (12.6; 45.2). 17. 1.2 diphenylethyl acetate (13.0;
42.3). 18. t-butyl chloroacetate (13.1; 38.1). 19. t-butyl di-
chloroacetate (12.8; 36.1). The first figure in the brackets is
logA, the second one is E kcal/mol.
A-factors from all three groups. The alcohols with
minimum branching of the carbon chain (at the secon-
dary carbon atom) are found on a somewhat higher level
of A-factors, line CD.
The maximally branched alcohols (tertiary carbon
atom) are found in the highest position along the logA
axis, with average logA values for 3.2 units higher than
the continuous-chain alcohols - line EF (Figure 2).
So clear a split of the common correlation line into a
few separated, clearly distinguished groups, depending
on some parameters, appears quite seldom and testifies
to the distinct influence of these parameters. These data
are in the very good agreement with the effect being
considered here. As indicated in the Introduction, the
branching disperses the energy fluctuation into different
ways and hampers its concentration on the bond that is
being broken. This increases the energy remaining on a
molecule at the moment of breaking, i.e., increases the
A-factor and the energy, which is necessary for the mo-
lecule’s energization (E).
It is possible, very roughly, to estimate the energy
amount remaining on the branched groups, based on
these data. The tertiary branching increases logA for 3.2
units, i.e., 3.2 × 2.3RT = 7.36RT kcal/mol energy,
which remains on this branched group. If we assume the
main model of tertiary branching the t-butyl group (four
carbon atoms, 12 degrees of freedom), then each degree
of freedom of such very strong branching detains
7.36/12 = ~0.6RT additional energy. No account has
been taken in this calculation of the contribution of the
C-H bonds, which have very different parameters (fre-
quencies and strengths) from the C-C and C-O bonds.
The average energy amount remaining on the secondary
carbon atom branching can also be evaluated in a similar
way. The propyl group may be taken as a main model of
this branching (3 carbon atoms, 9 degrees of freedom).
This branching increases the average logA value for  1
units (Figure 2), i.e. , the remaining energy for 2.3RT or
~0.25 RT for each degree of freedom. This is signifi-
cantly less than the stronger tertiary branching.
It is quite clear that the hydrogen atom and C-H bonds
contribute somewhat to additional energy retention on a
molecule at the moment of breaking. For this reason, the
deutering, which to some extent brings together C-H and
C-C bonds’ properties, according to the considered ef-
fect, has to make some contribution to the A-factor’s
increase. The deutered ethyl-d5 acetate (6) really has
some more A-factor than the initial one (4) and lies
therefore on the highest part of the correlation line (AB)
for continuous chain alcohols. A similar situation per-
tains to deutered cyclobutane (see above).
A few additional esters are presented in the graph for
comparison with their three main types (esters of pri-
mary, secondary and tertiary alcohols). The esters of cyclo
alkanes, cyclopentyl and cyclohexyl acetates (15,16) are
found practically on the correlation line CD of the sec-
ondary alcohol esters. The ester of the aromatic alcohol-
1,2-diphenyl ethylacetate (17) - is found a bit nearer to
the more branched tertiary alcohol esters. This can serve
as a hint that the rings of cycloalkanes and aromatic
compounds, containing bonds that are similar or almost
similar to branched alkanes, refer also to some type of
fluctuation trap and detain on these bonds part of the
energy necessary for bond breaking. This of course mer-
its more detailed consideration here (see below).
Contrary to the branching in the alcohol part of an es-
ter, such branching has practically no influence when it
is found in the acid part. Ethyl trimethyl acetate (2) with
maximum branching in the acid part and ethyl formate
(3) with the simplest acid part have practically the same
A-factor values (their logarithms are 11.2 and 11.3) and
lie on the same correlation line AB, practically on the

D. Krinkin / Natural Science 3 (2011) 661-682
673673
same point. A similar situation exists for isopropyl for-
mate (8) and isopropropyl trimethylacetate (9), where
alcohol residua are the same (isopropyl), but acid ones
are very different by the branching rate. They also have
close A-factor values and lie on the same correlation line
CD.
Even the very significant change of the acid residuum
is connected to only a moderate deviation from the cor-
relation line, corresponding to the same alcohol part of
esters. The Arrhenius parameters of t-butyl chloroace-
tates (18, 19) lie not far from correlation line EF of the
t-butyl esters that are not chlorinated. These deviations
from the EF line are significantly less than the difference
between the lines depending on the structure of the al-
cohol parts (AB, CD and EF).
In such a manner, the presence of the carboxyl group
between two alkane chains hampers the fluctuation en-
ergy passing from the one carbon chain to the other, i.e.,
from the fluctuation trap (isopropyl, t-butyl groups) in
the acid part to the bond that is being broken in the al-
cohol part of a molecule. This reduces the efficiency of
the fluctuation trap, i.e., its ability to increase both Ar-
rhenius parameters. It is indicated in the Introduction
that the presence between two carbon chains of the bonds
and atoms whose properties (frequencies, strengths,
atomic masses) are significantly different from carbon-
carbon bonds must lead to such results.
The proportionality coefficient “b” between logA and
E, i.e., the slope of the correlation lines AB, CD and EF
in Figure 2 equal  0.37. This generally corresponds to
the temperature diapason for esters pyrolyses (500 -
800˚K) according to estimation b = 1/2.3 RT.
The reported data dispersion of energy activation (E)
and A-factor (A) by different authors is, as a rule, much
bigger than their rate constant data. Therefore, between
the results of different authors for A and E values of the
same reaction, there exists, as a rule, a more or less lin-
ear correlation, approximately preserving a similar rate
constant value. This correlation is quite near to those that
are stipulated by the effect being considered. For exam-
ple, all the four points of indicated data of the different
authors for pyrolyses isopropyl acetate (7) are near to the
correlation line CD for other derivatives with the iso-
propyl group (Figure 2).
The experimental data presented on Graph 2 are quite
consistent with the body of conclusions about the effect
being considered mentioned in the Introduction:
1) The influence of the carbon chain branching on the
increase of both Arrhenius parameters for unimolecular
reaction in the high-pressure region;
2) The increase in this influence with the increase in
the branching rate;
3) The known Hofmann rule finds its substantiation in
the considered effect;
4) The influence of a heteroatom in a carbon chain
between a branching and a bond that is being broken;
5) The linear character of the correlation line for a de-
rivatives system in the coordinates logA ~ E.
2.7. The Benzene Ring as a Fluctuation Trap
The dynamics of intramolecular energy flow in the
benzene ring is being studied very widely by the method
of collision-induced relaxation of vibrationally excited
molecules (at least many tens of articles). Therefore,
there are enough data to look at this problem from the
viewpoint that is interesting to us. The toluene molecule
is a particularly attractive one to study because of the
presence of both a benzene ring and a side chain. Most
authors study the collision-induced dynamics of highly
vibrationally excited toluene interacting with argon, us-
ing quasiclassical trajectory calculation [1,5] with the
reproduced model Figure 1. Either CHmethyl or CHring has
been set at its high-energy state (as a rule, a little bit be-
low its dissociation threshold), while all other vibrational
modes are in their ground state.
It is found [5] that the interaction near θ = 77˚ (Figure
3(b)) plays a dominant role in promoting energy loss
through intramolecular energy redistribution, because it
is connected with energy transfer by bend mode through
the most flexible C-C-Hmethyl region. The smaller and
greater angles lead to significantly less energy loss, be-
cause the attack of the incident atom in these cases is
directed to the much more rigid corners of the benzene
ring and not to the side chain. The reverse situation, i.e.,
the excitation of the bend mode of the side chain region
from the vibrational energy of the rigid benzene ring,
seems significantly less probable. This is consistent with
the well-known experimental fact that the highly excited
benzene derivatives, which have side chains, show a
much greater tendency to transfer their vibrational en-
ergy to translation than the highly excited benzene does
(up to ten times as much).
The main conclusion to be drawn from these studies
[1,5] is that the relaxation process is dominated by the
energy flow from the side chain (methyl group) to the
benzene ring, but not in the opposite direction. The au-
thors [5] write: “We do not find any evidence of energy
flowing out of C-Hring and returning to C-Hmeth yl. That is,
energy flows irreversibly out of the initially excited
C-Hmethyl vibration and deposits in the C-Hring after pass-
ing through two stretching and three bending modes”
(page 4864). These two notions: “irreversibility” and
“vibrational energy deposit” are exactly the key concepts
of the considered effect.
The benzene ring has many properties to make it a
practically perfect fluctuation trap: the almost identical

D. Krinkin / Natural Science 3 (2011) 661-682 674
(a)
(b)
bonds, and the angles between them and their surround-
ing; the flat hexedron right form; and the identical mass
distribution (six identical carbon atoms). All these prop-
erties are conditions for practically free fluctuation en-
ergy movement inside the ring, whereas their exit from
the ring is very hampered, because of the demands of the
side-chain bend mode excitation. At the same time, a
side chain of aromatic compounds has the bond (β),
which is very much weakened in comparison with the
neighbouring one (  ) as result of the I-effects of the
benzene ring (about 63 against 90 kcal/mol).
This bond (β) is broken in most bond fission reactions.
The nearness of a weak bond, which mobilizes sur-
rounding energy, must reduce the role of the benzene
ring as a contrary factor.
These investigations [1,5] also give valuable informa-
tion about the C-H bond influence on the processes of
concentration and dispersion of vibrational energy in the
system of C-C bonds. The highly excited C-Hmethyl bond
of toluene makes at least 15 vibrations before the colli-
sion with the Ar atom and ~30 after it, without any vis-
ual indications of this high energy transfer to neighbour-
ing C-C bonds, which have significantly less vibrational
energy, because they vibrate in their ground state [1,
Figure 2(a), page 1225]. This is so even in spite of the
fact that the excited C-H bond in this case is in its an-
harmonic region and therefore vibrates at a lower fre-
quency (484 cm–1) than the neighbouring C-C bonds
(1208 and 1494 cm–1), which are in their ground state. A
similar situation is described in [5]. 5, page 4863, Figure
5. The main role here is played, of course, by the mass
difference between carbon and hydrogen atom. It should
be noted that the highly excited C-C bond transfers its
energy to surrounding C-C bonds practically during a
few vibrations.
The vibrational energy from the excited stretch mode
of the C-Hmethyl bond transfers to the C-Hring through two
stretch modes of C-C bonds and deposits on this C-Hring
bond to a greater extent than on these two intermediate
C-C bonds [1, Figure 2(b), (c), page 1225].
These data, which are interesting for our consideration,
have been taken directly from the corresponding graphs
because the authors of these papers [1-5] discuss in the
text other aspects of these calculations in which they
were interested. All these data mean that the interaction
between C-H and C-C bonds is significantly lower than
between C-C themselves and C-H themselves. It can be
loosely said that the vibrational energy movement inside
a molecule takes place through the system of C-H bonds
and the system of C-C bonds to some extent independ-
ently. This statement is significantly more right for the
system of C-C bonds. The energy movement along C-H
bonds fades much more quickly, because it must pass
through C-C bonds with very different properties.
The main conclusion is that the identical bonds of the
benzene ring constitute a very effective fluctuation trap,
which disperses energy fluctuation and impedes its con-
centration on the individual bond of the ring or the side
chain.
The second conclusion is that the interaction between
C-C and C-H bonds systems is much weaker than the
interaction inside the system of the C-C bonds and be-
tween the C-C bonds themselves.
It is of interest to compare the kinetic parameters of α
and β bonds fission. The fission of an α-bond that
neighbors the benzene ring must, to a significantly
greater extent, reflect the true influence of the ring as a
fluctuation trap, i.e., its ability to increase A-factor and
energy activation. The β-bond fission is a subject to two

D. Krinkin / Natural Science 3 (2011) 661-682
675675
contrary influences: the mobilization of the internal en-
ergy of a molecule by this weak β-bond for its breaking,
and, in contrast, the detention of a part of this energy by
the benzene ring. Unfortunately, the α-bond fission in
the hydrocarbon benzene derivatives with the availabil-
ity of the β-bond is almost impossible. In this case, the
weak β-bond is practically always broken. There are,
however, quite reliable data for an α-bond fission in the
benzene derivatives containing a heteroatom. It is also of
interest because the benzene ring properties as a fluctua-
tion trap have to be quite manifest in this case also.
2.8. The β-Bond Fission
The β-bond fission can be considered in the example
of the thermal unimolecular decomposition of ethylben-
zene, isopropylbenzene, and tert-butylbenzene [14]:
logA = 15.3; E = 72.7 kcal/mol; T = 1053 - 1234˚K
logA = 15.8; E = 71.3 kcal/mol; T = 971 - 1151˚K
logA = 15.9; E = 69.1 kal/mol; T = 929 - 1157˚K
These reactions have been studied using the very
low-pressure pyrolysis (VLPP) technique under com-
parative conditions. Each reactant is decomposed by
β-bond homolysis, producing methyl radicals and benzyl
or benzylic type radicals.
The tendency of the A-factor to increase in passing
from the ethyl to the t-butyl group is undoubtedly con-
nected with the growth of the branching rate in this row.
This regularity was already mentioned in particular for
the esters pyrolyses and will be discussed further. Here
the transfer from strongly branched t-butyl to the con-
tinuous chain of the ethyl group reduces the logarithm of
A-factor to 0.6 units only, whereas a similar transfer for
esters pyrolyses lowers this value to 3.2 units. This
might quite easily be a consequence of the fluctuation
trap influence produced by the benzene ring, which does
not allow the reduction of A-factor. That is, the influ-
ences of the strongly branched t-butyl group and the
benzene ring are quite commensurable and opposed.
The conclusion: The growth of the demobilization
factor (dispersion rate of the fluctuation) occurs together
with the increased number of identical bonds in the
branching (branching rate).
2.9. The α-Bond Fission
The α-bond fission can be considered in the example
of the unimolecular decomposition of nitrosobenzene
with phyenyl radical production [15].
The rate constant was measured at temperatures be-
tween 553 and 648˚K at atmospheric pressure and a
strong dilution with argon. This leads to minor secon-
dary processes and a reliable measurement of the high-
pressure first-order rate constant for the primary nitro-
sobenzene decay. The result of the experiments may be
presented as energy activation 55.06 ± 1.08 kcal/mol,
which is higher by as much as 6 kcal/mol than those
reported earlier, and a logarithm A-factor of 17.15 units.
Such a high A-factor value indicates essential vibrational
energy dispersion inside the molecule, i.e., the presence
of an effective fluctuation trap, a role which, in this case,
may be played only by the benzene ring.
The C-N bond that is being broken lies in the same
plain as the benzene ring, and forms identical angles
(~120˚) with both ring bonds. All this must promote the
energy exchange between the C-N bond and the benzene
ring and increase the efficiency of the benzene ring as a
fluctuation trap in relation to this bond. The absolutely
symmetrical phenyl radical is produced in the process of
the nitrosobenzene decay, which may be a main factor in
the vibrational energy dispersion inside its identical
bonds (see below). This may probably justify the high
A-factor value, in spite of the weak bond breaking, and
confirm the high level of efficiency of the benzene ring
as a fluctuation trap. This is consistent with the previous
example. At the same time, all these preliminary results
need more experimental confirmation.
2.10. The Fluorescence Spectra of Benzine
Derivatives
The lifetime of excited molecules and, in particular,
those containing a benzene ring, may be studied by
fluorescent spectroscopy. In this connection, very valu-
able information relevant to the effect being discussed
here may be obtained from the very interesting works
[16] of C. Parmenter and B. Stone and from the more
recent ones of D. Nesbit and R. Field. [17] We can
evaluate from these papers not only the energy fluctua-

D. Krinkin / Natural Science 3 (2011) 661-682 676
tion lifetime on the benzene ring but also symmetry’s
influence on it.
Light absorption, as it is known, produces an elec-
tronically excited molecule. It can return to its original
ground state by relaxation with the emission of radiation
(fluorescence). It can be converted also to the ground
state of the original molecule by vibrational relaxation
that is accompanied naturally by the disappearance of
the fluorescence. It enables us to measure the lifetime of
the excitation, i.e., the actual lifetime of a vibrational
energy fluctuation on a molecule.
The authors studied the transition from the lowest ex-
cited singlet (S1) configuration to the ground state (S0),
i.e., the S1 → S0 fluorescence spectra. They studied such
spectra for the symmetrical molecule of p-difluoroben-
zene (pDFB)
and the molecule of p-fluorotoluene (pFT),
i.e., they studied the result of the symmetry being de-
stroyed by the replacement of fluorine of pDFB with a
methyl group.
Thirty of the 39 vibrational degrees of freedom of pFT
are in common with pDFB, with almost no frequency
change. All of these 30 degrees of freedom relate of
course to benzene rings and do not include vibrations of
fluorine or the methyl group. The two rings are nearly
identical twins, and similar excitations can be produced
in each. Thus, the pair provides an opportunity to ob-
serve the specific effect of a methyl group on the in-
tramolecular vibrational redistribution (IVR) dynamics
of the same benzene rings.
The lifetimes of IVR for S1 levels of both molecules
are presented in [16]. Therefore, for an energy region of
1596 cm–1, the lifetime for S1 level of pDFB was 290 (a
few thousands vibrations) and for pFT only 10 p.s. There
is no doubt that the main role in this region is played by
the stretch vibrations of the six bonds of the benzene
ring whose frequencies are close to this value (presently
accepted as 1494 cm–1). It was mentioned above that all
these bonds, just as in the benzene ring of pDFB, so also
in the benzene ring of pFT, have almost identical fre-
quencies, strengths, and all other properties excepting
only one: their inside tensions are somewhat different as
a result of influences that are unsymmetrical outward
from the rings.
This result emphasizes the very significant role that
symmetry plays in the lifetime of the energy fluctuation
on a molecule. Really, the full equivalency of the dif-
ferent directions for its flow at a high level of symmetry
promotes a higher level of dispersion between these di-
rections. This conforms to the above-mentioned role of
the chain branching and other forms of identical bond
systems as a fluctuation trap. The very small change in
bond strength in the rings leads, in this case, to a quite
significant change in internal energy redistribution be-
tween the bonds of these rings (~30 times as much). This
is consistent with the original supposition on which the
effect being considered is based about internal energy
redistribution between weak and strong bonds. For, more
details of the symmetry's influence see [1].
For the energy region of 1988 cm–1, the lifetime for
the S1 level of pFB was 97 and for pFT ~10 p.s., i.e., the
difference between the two rings is lower than for the
lower energy region 1596 cm–1. For region 2382 cm–1,
the lifetime for S1 level of pDFB was 25 - 33 and for
pFT also 10 p.s. As the region of C-H bond frequencies
(2800 - 3100 cm–1) is approached, the differences be-
tween the two benzene rings are reduced. This means
that the C-H bonds interact with each other more weakly
than do the C-C ones and the dispersion of the fluctua-
tion energy through the system of C-H bonds takes place
more slowly than through the system of C-C bonds. The
C-H bonds can interact only through the C-C bonds,
while the C-C bonds interact directly. The energy fluc-
tuation passing through a bonds system with significant
frequency changes is hampered, as was also mentioned
above.
The general conclusion from these studies is that the
benzene ring must be a very effective fluctuation trap.
The excited degree of freedom, including the stretching
vibration of the C-C bonds of the benzene ring (~1600
cm–1), preserves a few thousand vibrations on it, before
it spreads (very slowly) onto other degrees of freedom.
This is consistent with the above-mentioned data [1,5],
where at least tens of vibrations occur on the ring with-
out any visual indications of an energy transfer to other
degree of freedom. This type of the C-C stretch vibration
is preserved for the longest time on the benzene ring
without mixing with other types. In such a manner, the
energy flow through the stretch C-C vibration of the six
ring bonds is probably the main factor in the formation
of a fluctuation trap. This is the main conclusion. This
confirms also the above-mentioned conclusion about
weak interaction between systems of C-C and C-H
bonds.
Such results [16,17] go beyond the usual symmetry
contribution to fluorescence spectra. The authors rec-

D. Krinkin / Natural Science 3 (2011) 661-682
677677
ommend further study of this phenomenon. They write:
“The source(s) of the methyl acceleration remains to be
established.” [16, page 4711].
2.11. The Branching of the Carbon Chain as
a Fluctuation Trap in the Bond Fission
Reaction
Bond fission reaction (i.e., the initiation reactions in
decomposition) has always been of prime interest to
kineticists and thermodynamicists. The kinetic study of
such a reaction is, as a rule, complicated by the numer-
ous secondary reactions of the produced radicals. One
solution is to reduce the reaction time and accordingly
the conversion rate, which diminishes the initial radicals'
concentration and their secondary reactions. Very
low-pressure pyrolysis (VLPP) makes it possible to re-
duce the conversion time to below 10–2 - 10–3 second.
The VLPP technique is well established as a method for
measuring the unimolecular rate constant. The Arrhenius
parameters obtained using the VLPP method usually
differ somewhat from those received using conventional
methods: static or flow techniques. The Arrhenius pa-
rameters for this reason must be compared in the
framework of the same experimental method, which is
important for our analysis. Below are presented Ar-
rhenius parameters of different reactions obtained by the
same authors for comparison, using the VLPP method.
The middle bond cleavage in the n-butane, 2,3 di-
methyl-butane and 2,2,3,3 tetramethylbutane has been
studied for comparison by W. Tsang and J. Kiefer [13]:
n-butane
logA s–1 = 17.8; E = 86.7 kcal/mol; [12, p. 79, Figure 9]
2,3 dimethylbutane
LogA = 18.9 s–1; E = 84.1 kcal/mol; [13, p. 80, Figure 10]
2,2,3,3 tetramethylbutate
LogA = 19. 7 s–1; E = 77 kcal/mol; [13, p. 84, Figure 11]
The experimental part of this work using the shock
tube technique (VLPP method) was reported in [12].
Transfer from continuous-chain n-butane to branched-
chain 2,3 dimethylbutane increases the logarithm of the
A-factor, i.e., increases the vibrational energy quantity
detained on the transition state and reaction fragments
(two identical radicals). The further increase of the
branching rate by transfer to 2,2,3,3 tetramethylbutane
also increases the logA value.
It is possible to evaluate very roughly the additional
energy quantity remaining on the transition state in these
reactions. The number of degrees of freedom for distri-
bution of this additional energy is, of course, quite con-
ditional. Transfer from n-butane to 2,3 dimethylbutane
increases the logA by 1.1 units, i.e., the additional en-
ergy increases to 2.3 × 1.1  2.5 RT. Setting the number
of degrees of freedom for 2,3 dimethylbutate as 15 this
gives ~ 0.17 RT per each degree of freedom. A similar
evaluation gives a value of ~0,2 RT for transfer from
n-butane to 2,2,3,3 tetramethylbutane.
For esters pyrolyses this tendency can be seen much
more distinctly. The general vibrational energy quantity
remaining on each degree of freedom is tenths of RT. As
known, the opposite process of intramolecular energy
mobilization for a bond that is being broken in the
fall-off region is accompanied by similar energy values,
removed by one degree of freedom (tenths RT).
2.12. The Bond Fission Reaction of
Neopentane and Its Derivatives
Neopentane is a maximum branched acyclic alkane,
which has a possible maximum of four identical bonds at
one carbon atom, with high central symmetry.
Therefore we can expect a very high rate of energy
dispersion inside its molecule at the bond cleavage (de-
mobilization process) and a significant reduction in this
dispersion for neopentane derivatives. The one methyl
group replacement in neopentane significantly reduces
the molecule symmetry as a result, and so also the num-
ber of identical bonds for energy dispersion.
It is interesting in this connection to compare the ki-
netics of the C-CH3 bond cleavage in neopentane and
4,4-dimethylpent-2-yne (DM2P), where one methyl in
neopentane is replaced by the acetylenic group [18,19].
logA = 17.3 (s–1); E= 80.8 kcal/mol.

D. Krinkin / Natural Science 3 (2011) 661-682 678
logA = 16.4 (s–1); E = 71.4 kcal/mol.
These results are obtained under comparable condi-
tions using VLPP techniques in the particular identical
temperature regions, 1000 - 1260˚K for neopentane and
903 - 1246˚K for DM2P,
Vibrational energy flows between similar bonds more
easily than between different bonds and also transfers
more easily from a strong bond to a weak one than vice
versa. Therefore, as a first approximation, it may be as-
sumed that the acetylenic group CH3-C  C-, which has
stronger bonds than C-CH3, does not participate in acti-
vation energy dispersion. In this case, only three C-CH3
bonds of DM2P participate in the forming of the fluctua-
tion trap, and DM2P has as fluctuation trap of only 9
degrees of freedom, i.e., three degrees of freedom fewer
than neopentane.
The energy surplus detained on the neopentane mole-
cule at the moment of bond breaking exceeds this value
for DM2P by logA × 2.3  RT = 0.9 × 2.3  RT = 2.07 
RT, that is, it exceeds it by ~ 0.7  RT per each addi-
tional degree of freedom.
The reduction in DM2P’s efficiency as a fluctuation
trap in comparison with neopentane is obviously con-
nected to the symmetry being destroyed, not only to the
fewer numbers of degrees of freedom.
It seems that maximally branched neopentane of the
hydrocarbon chain is really a very effective fluctuation
trap. If we take into account some contribution of the
acetylenic group to energy dispersion, which has been
ignored, the energy surplus in neopentane in comparison
with DM2P is even increased by values greater than 0.7 
RT per each degree of freedom. Because some energy
surplus is detained on the DM2P also, the absolute en-
ergy value which neopentane detains on its vibrational
degree of freedom may probably exceed a value of one
RT.
2.13. The Pyrolysis of n-Propyl Isopropyl
and Tert-Butyl Cyanides
This was investigated using the VLPP method under
the same conditions, in the temperature region 1050 -
1250˚K. [20] The kinetic data of bond fission with  CN
radical elimination are as follows:
n-propyl cyanide
logA = 15.4 (s–1); E = 79.0 ± 1.7 kcal/mol
isopropyl cyanide
logA = 15.7 (s–1); E = 79.0 ± 2.0 kcal/mol
tert-butyl cyanide
logA = 15.80 (s1); E = 74.9 ± 1.6 kcal/mol.
Here there is also a clear tendency for the A-factor to
increase along with the increase in the carbon chain
branching.
The vibrational energy surplus in comparison with
n-propyl cyanide, according to expression E=2.3  
logA  RT, for isopropyl cyanide equals 0.69  RT and
for t-butyl cyanide to 0.92 respectively, where  logA is
the difference between the logarithms of A-factors for
reactions 2 and 3 and reaction 1. Here the efficiency of
isopropyl and t-butyl radicals as a factor that disperses
vibrational energy along a molecule is significantly less
dominant than, for example, in esters pyrolysis, where
these values were 2.3 and 7.36  RT, respectively. In
principle, this may be a result of the reverse strong CN
bond influence, which leads to the mobilization of vibra-
tional energy for the neighbouring weak C-C bond.
2.14. The Pyrolysis of Tetramethyl and
Tetraethyltin
The pyrolysis of tetraalkanes of the IV group of chem-
ical elements, which have a high rate of central symme-
try, similar to neopentane’s, are accompanied also by
high A-factor values, which indicates that a remarkable
part of the activation energy is dispersed along the
molecule [21,22]. The pyrolysis of tetramethyltin has
been studied in a conventional toluene carrier flow sys-
tem at temperatures from 800 to 940˚K. No appreciable
heterogeneous reaction was detected and the first order
rate constant appears to have been determined at the
high pressure limit [21,22]. No remarkable secondary
reactions complicate the main reaction, and the Ar-
rhenius parameters obtained reflect Sn–CH3 bond cleav-
age
The decomposition of dimethyltindichloride has been

D. Krinkin / Natural Science 3 (2011) 661-682
679679
studied in a similar toluene carrier system under com-
parative conditions. The reaction of Sn-CH3 bond clea-
vage was first order and homogeneous:
logA = 14.3 (s–1); E = 58 kcal/mol
As may be seen, the destruction of the central symme-
try of tetramethyltin, with the reduction in the number of
identical bonds from 4 to 2, leads to a significant de-
crease in the logA value by 1.4 units. Here the part of the
activation energy dispersed inside the transition state
reduces by 2.31.4  3.2RT at the transfer from one of
the above reactions to another. Here it is very difficult to
evaluate this additional energy for tetramethyltin de-
composition per one additional degree of freedom. The
tetramethyltin has two additional Sn-CH3 bonds, six
additional degrees of freedom, in comparison with di-
methyltin dichloride, which means a loss of 3.2/6 ~0.5
RT per one degree of freedom. This evaluation must be
more if we take into account that here not loss but only
substitution of two symmetric bonds for another sym-
metric one takes place. As above mentioned, this figure
for neopentane probably exceeds the value of 0.7 RT.
Tetramethyltin, from this point of view, must not differ
from neopentane.
In the paper of Johnson and Price [21] there is some
rough evaluation of logA values for reactions of succes-
sive Sn–CH3 bond cleavage from the following species:
Sn(CH3)4; Sn(CH3)3; Sn(CH3)2; SnCH3 down to metallic
tin. This evaluation is based on previous studies in their
laboratory. They suppose it reasonable to assign the fol-
lowing approximate values to the logA factor: 15.7, 15.0;
14.0; 11.0 (s–1). Albeit these figures are very rough, the
general tendency is quite clear: the reduction in the iden-
tical bond number at one central atom leads to a signifi-
cant decrease in A-factor for the reaction of these bonds
cleavage. Such a conclusion agrees with our original
supposition that a reduction in the identical bonds num-
ber at one central atom, i.e., a reduction in branching rate,
leads to a decrease in the activation energy surplus re-
maining on a transition state at these bonds’ cleavage.
The replacement of all four methyl groups in the
tetramethyltin for ethyl groups leads to an increase of
logA-factor from 15.7 to 16.0 for the reaction of this
group releases:
Sn(C2H5)4 → Sn(C2H5)3 + C2H5
logA = 16.0 (s-1); E = 59.3 kcal/mol
The thermal decomposition of tetraethyltin has been
studied under comparative conditions with tetrame-
thyltin in a toluene carrier flow system over the tem-
perature range 725 to 833˚K. [22] This A-factor increase
is in good agreement with our supposition about the ac-
tivation energy dispersion between neighbouring identi-
cal bonds and groups that are bound with them. Natu-
rally, the complication of identical groups, i.e., their in-
ternal energy capacity, must increase the rate of disper-
sion and accordingly the A-factor. Unfortunately, this
A-factor increase is not too large, near to the experi-
mental precision.
The ethyl alkyls pyrolysis of the bivalent metals has
been studied under the same conditions [22]. These al-
kyls have, naturally, a lower degree of symmetry than
alkyls of tin with a valency of four. The reaction of bond
cleavage with the ethyl group have, according the effect
being considered, less lgA-factor than tetraethyltin,
which has an logA value of 16.0. The toluene carrier
study of diethylmercury gave logA ~ 15.4, and di-
ethylzinc ~14.3. It may be concluded that the substitu-
tion in the neopentane of the central carbon atom for
other elements of the IV group (Sn, Pb) changes not
very significantly the amount of energy remaining on
each degree of freedom.
2.15. The Thermal Decomposition of
Methane
It is interesting to consider the influence of the dis-
cussed effect on the C-H bond breaking in the example
of methane decomposition. The methane has C-H bonds
only, which interact one to the other directly, but not
through the C-C bonds, which weakens this interaction
very significantly. The methane molecule, like the neo-
pentane, has a possible maximum of four identical bonds
at one carbon atom, with high central symmetry. There-
fore, we can expect, as in the case of neopentane, a very
high rate of energy dispersion inside its molecule at the
cleavage of one of these four bonds.
The very detailed and careful work of J. Chen and H.
Back [23] is a good completion of many previous inves-
tigations where unfortunately not all the proper experi-
mental conditions were strictly kept (pressure-dependent
region, secondary reaction influence, influence of the
deposition of carbon on the surface vessel wall, etc.). It
was reasonably concluded in this work that the initial
stage of the decomposition (1) can be described by a
simple homogeneous, non chain radical mechanism:
CH4  CH3 + H (1)
H + CH4  CH3 + H2 (2)
2CH3  C
2H6 (3)

D. Krinkin / Natural Science 3 (2011) 661-682 680
In this work, the reaction products hydrogen (2), eth-
ane (3), ethylene, propylene and acetylene were ana-
lyzed as a function of time. In the initial stage of the
reaction, hydrogen (2) and ethane (3) are the only prod-
ucts that are in equal quantities. The results show clearly
that this initial rate of ethane and hydrogen formation is
a measure of the initial homogeneous rate of the de-
composition of methane, i.e., it is reflected at the pri-
mary stage that is of interest to us (1).
In this work, the effect of carbon deposit was negligi-
ble; after each experiment its traces were burnt. The in-
fluence of the vessel wall is also negligible; the rate of
ethane and hydrogen formation in the packed and un-
packed vessel was identical. The quartz tube was flushed
with hot concentrated nitric acid. The experimental
pressure was increased up to the clear entering into the
high-pressure region.
The pyrolysis of methane has been studied in a static
system at temperature 995 - 1103˚K. The Arrhenius pa-
rameters of interest to us at the primary stage (1) of
CH3-H bond cleavage were logA(s–1) = 16.45; E = 107.6
kcal/mol. For conventional methods (static and flow
systems), and undoubtedly for a simple bond cleavage of
the simplest organic molecule, such a value of logarithm
A-factor seems too high. The “normal” value in this case
probably does not exceed 14.5 - 15.0. The difference
between experimental and expected values, 1.5 - 2.00
units (logA), is equivalent to 2.3logART energy.
This additional thermal energy remaining on the methyl
radical that is formed is about 0.3 - 0.4 RT per each de-
gree of freedom. An excessive value of activation energy
(107.6 kcal/mol), in comparison with the usually ac-
cepted C-H bond strength ~98 - 102 kcal/mol, is also in
agreement with this conclusion about significant energy
remaining on the reaction fragments.
It seems that the energy dispersion rate in methane at
its bond cleavage is not significantly different from that
in neopentane. Probably, the energy dispersion in the
C-H bond system is near to that in the C-C bonds
2.16. The Alkyl Chloride Pyrolysis
The elimination of hydrogen chloride from alkyl chlo-
rides to yield the corresponding olefin proceeds through
a four-centered activation complex. This concept is now
widely accepted.
Since in the formation of the four-center transition
state a sizable internal rotation is lost and its structure
may be stiffening and strained, an A-factor greater than
normal kT/h is not reasonable. Benson and O’Neal [7]
estimate a logA-factor on the level of 12.9 - 13.8. They
indicate, however, that the experimental values in most
cases are significantly higher.
The strengths of all bonds that are being formed and
broken in this reaction are indicated on the figure
(kcal/mol). In the energized molecule, the energy con-
centration on the bonds that are being broken (C-Cl and
C-H) takes place mainly by energy redistribution inside
the transition state ring. These energy flows are indi-
cated on the figure by arrows. The H-Cl bond formation
releases 103 and the simple C-C bond conversion into
the double bond 62 more kcal/mol. The significant en-
ergy flow to the bonds that are being broken (C-H and
C-Cl) passes through both the carbon atoms in the ring.
Therefore the branching at these carbon atoms (R1, R2,
R3, R4) takes aside a part of energy from these bonds.
The lengthening of these radicals increases, naturally,
the energy quantity dispersed in such a way and the
A-factor value. This influence of the considered effect
can quite explain the above mentioned discrepancy be-
tween existing theory and experiment that bewildered
Benson and O’Neal [7]. They write: “The A-factor ap-
pears to increase with the increasing size of the alkyl
group in the n-alkylchloride series. Such an effect, if real,
is difficult to rationalize” (page 68).
Because of this, in the majority of cases, the authors
estimate the Arrhenius parameters as too high and mod-
ify them to lower values, in spite of the fact that they
state that the rate constants they estimated are reliable.
As a preferred value of the A-factor, they accept either
the lowest experimental value from among the many
results or a value which is lower than them. For example,
the experimental value of Arrhenius parameters for py-
rolysis of 1-chloro-2-methyl propane are logA = 14.0
and E = 56.85 kcal/mol. The authors modify them to the
values logA = 12.9 and E = 53.2. They write: “Rate con-
stant is reliable but the Arrhenius parameters are proba-
bly high” (ibid, p. 69).
The A-factor values estimated by the authors on the
base of RRKM are also lower than the experimental data.
The empirical Hofmann rule for the direction of esters
pyrolysis is similarly explained by the effect being con-
sidered here. The branching on the energy flow in the
six-centered ring defines this direction (see above).
At the decomposition of cyclobutane derivatives (para-
graph 2.5), the main internal energy for bonds breaking
arrives from a side chain, but not from the ring. In this
case, as distinct from the alkyl chlorides, the length of
the continuous side chain promotes energy mobilization.

D. Krinkin / Natural Science 3 (2011) 661-682
681681
3. CONCLUSIONS
1) In this article, two types of organic molecule struc-
ture are considered, which very often deflect experi-
mental results from those predicted by statistical kinetic
theories for unimolecular reaction in the high-pressure
region.
a) The structural elements that promote internal en-
ergy mobilization accelerate the reaction rate, reduce
both the Arrhenius parameters - energy activation and
A-factor - and reduce the amount of energy that remains
on the reaction fragments.
b) The structural elements that, conversely, impede
internal energy mobilization, decrease the reaction rate,
and increase both the Arrhenius parameters and the
amount of energy remaining on the reaction fragments.
The final dynamic influence depends on the presence in
the organic molecule of these two structural elements as
much as on the molecular structure between them. The
influence of these two structural elements (a and b) on
both Arrhenius parameters is uniform, i.e., it either in-
creases or reduces both of them. Therefore, for the ma-
jority of derivatives there exists a positive correlation
between energy activation and A-factor. The derivatives
whose structure promotes internal energy mobilization
are located in the part of the correlation line where both
Arrhenius parameters are low, that is, at the lower, left
end of the line. The derivatives whose structure impedes
internal energy mobilization are located at the right
higher end.
2) The breaking of a weak bond in the vicinity of
stronger neighbouring ones is accompanied by the rein-
forcement of the internal energy mobilization of a mole-
cule (type a). The continuous carbon chain neighbouring
on this weak bond increases this effect. In this article,
this type of influence is considered using as an example
the simple bond fission reaction (hydrazine, amine), as is
also the ringed transition state reaction (nitroalkanes,
orthosilicates, cyclobutanes, and esters). In all these
cases, both Arrhenius parameters are significantly lower
than predicted by statistical theories.
3) The group of identical bonds that are linked to one
another represent the fluctuation energy trap. They pro-
mote easy energy circulation inside this group, disperse
it between identical bonds and impede its concentration
on the bond that is being broken (inside the group or
outside it). Such groups include a branched carbon chain
as well as a cyclic compound.
4) The influence of the branching is increased in the
row, primary, secondary, tertiary, and neopentane type
carbon atom; i.e., with the increase in the quantity of
identical bonds number in the branching from one to
four. This influence is considered using the examples of
thermal decomposition of esters, butanes, neopentane
and its derivatives, tetramethyl and tetraethyltin, alkyl
chlorides, and methane. In all these cases, the Arrhenius
parameters are significantly higher than those expected
based on statistical theories. The energy that remains on
the reaction fragments, by an order of magnitude, is
tenths of an RT per one degree of freedom; it grows with
the increase in the branching rate and at its maximum
(neopentane) probably approaches the value of one RT.
5) As a cyclic compound, the aromatic substances are
considered using toluene, nitrosobenzene, p-diflouro-
benzene, p-fluorotoluene as an example. The benzene ring
is an effective fluctuation trap, dispersing energy mainly
through the stretch vibrations of the six identical bonds
(type b). The excited degrees of freedom, including the
stretching vibration on the C-C bond of the benzene ring,
are preserved on these degrees of freedom up to thousands
of vibrations without a noticeable spread on the other vi-
brational mode.
6) The dynamic influence in the esters almost com-
pletely fades away during transfer though the carboxyl
group from the carbon chain of the alcohol part of a mo-
lecule to the acid part, or in the opposite direction. This is
a result of the significant difference between the parame-
ters of the bonds of the carboxyl group and carbon chain.
7) The correlation between energy activation and
A-factor for the thermal decomposition of cyclobutanes
and esters derivatives is considered. Except for the dy-
namic effect discussed in this paper, which is the basis for
the positive correlation for most of the derivatives, the
row of other factors influence both the Arrhenius parame-
ters. This leads to some deviation from the correlation line.
In spite of this deviation, the positive correlation between
two Arrhenius parameters for these two types of deriva-
tives is quite clearly observable; the slope of the correla-
tion line corresponds to the theoretical values for the ef-
fect being discussed, the derivatives with structural ele-
ments, which promote internal energy mobilization being
located in the region of the low Arrhenius parameters at
the left lower end of the line, and those which impede this
mobilization at the right higher end. The significant devia-
tion from the correlation line is even quite informative,
because, in particular, it corresponds to the change in the
reaction mechanism, which demonstrates the orthosilicate
decomposition.
8) The dynamic interaction between systems of C-H
and C-C bonds is considered using cyclobutanes, esters
and aromatic compounds as examples. This interaction is
much impeded as a result of the significant difference
between the parameters of C-C and C-H bonds. Therefore,
the C-H bonds’ influence on the dynamic picture of an
organic molecule, as a first approximation, may be ig-
nored. At the same time, the deutering of the C-H bonds
brings their properties a little closer to those of the C-C

D. Krinkin / Natural Science 3 (2011) 661-682 682
ones and other organic bonds, and the influence of C-D
bonds becomes more noticeable. The deutering of the
organic compounds shift them along the correlation line.
REFERENCES
[1] Ree, J., Chang, K.S., Kim, Y.H. and Shin, H.K. (2003)
Collision-induced intramolecular energy flow in highly
excited toluene, Bulletin Korean Chemical Sciences, 24,
1223-1226. doi:10.5012/bkcs.2003.24.8.1223
[2] Lenzer, T., Luther, K., Troe, J., Gilbert, R.G. and Lim,
K.F. (1995) Trajectory simulation of collisional energy
transfer in highly excited benzene and hexafluoroben-
zene, The Journal of Chemical Physics, 103, 626-641.
doi:10.1063/1.470096
[3] Clarke, D.L., Oref, I., and Gilbert, R.G. (1992) Colli-
sional energy transfer in highly excited molecules: Cal-
culations of the dependence on temperature and internal,
rotational, and translational energy, The Journal of
Chemical Physics, 96, 5983-5998.
doi:10.1063/1.462639
[4] Shin, H. K. (1975) Effects of the multiplicity of impacts
on vibration-translation energy transfer in collinear colli-
sions, Journal Chemical Physics, 62, 4130-4137.
doi:10.1063/1.430291
[5] Ree, J., Kim, Y.H. and Shin, H.K. (2002) Collision-indu-
ced intramolecular energy flow and C-H bond dissocia-
tion in excited toluene, Journal Chemical Physics, 11 6,
4858-4870. doi:10.1063/1.1452726
[6] Bernshtein, V. and Oref, I. (1996) Trajectory calculations
of relative center of mass velocities in collisions between
Ar and toluene, The Journal Chemical Physics, 104, 1958-
1965. doi:10.1063/1.470950
[7] Benson, S.W. and O’Neal H.E. (1970) Kinetic Data on
Gas Phase Unimolecular Reactions,United States De-
partment of Commerce, NSRDS-NBS, February.
[8] Robinson, P.J. and Holbrook, K.A. (1972) Unimolecular
reaction, Wiley-Interscience.
[9] Krinkin, D. (2007) Anomalously large kinetic isotope
effect, Central European Journal of Chemistry, 5, 1019-
1063. doi:10.2478/s11532-007-0048-2
[10] Jason, C.S., Chu, J.B., Wang, N.S. and Lin, M.C. (1991)
Relative stabilities of tetramethyl orthosilicate and tetra-
ethyl orthosilicate in the gas phase, Department of
Chemistry Emory University Atlanta, GA30322, 1-11.
[11] Chu, J.C.S., Soller, R., Lin, M.C. and Melius, C. F. (1995)
Thermal decomposition of tetramethyl orthosilicate in
the gas phase: An experimental and theoretical study of
the initiation process, The Journal Chemical Physics,
99, 663-672. doi:10.1021/j100002a034
[12] Tsang, W. (1981) Comparative rate single pulse shock
tTube in the thermal stability of polyatomic molecules, In:
Lifshitz, Shock Tubes in Chemistry, Marcel Dekker, New
York, 59.
[13] Tsang, W. and Kiefer, J.H. (1995) Unimolecular reactions
of large polyatomic molecules over wide ranges of tem-
perature, Advanced Series in Physical Chemistry, 6, Part
1, 87.
[14] Robaugh, D.A., Stein, S.E. (2004) Very-low-pressure
pyrolysis of ethylbenzene, isopropylbenzene, and tert-bu-
tylbenzene, Internal Journal of Chemical Kinetics, 13,
445-462. doi:10.1002/kin.550130503
[15] Park, J., Dyakov, I.V., Mebel, A.M. and Lin, M.C. (1997)
Experimental and theoretical studies of the unimolecular
decomposition of nitrosobensene: High-pressure rate
constant and C-N bond strength, Journal of Physics
Chemistry A, 101, 6043-6047.
doi:10.1021/jp9712258
[16] Parmenter, C.S. and Stone, B.M. (1986) The methyl rotor
as an accelerating functional group for IVR, The Journal
of Chemical Physics, 84, 4710-4711.
doi:10.1063/1.449999
[17] Nesbitt, D.J., Field, R.W. (1996) Vibrational energy flow
in highly excited molecules: role of intramolecular vibra-
tional redistribution, The Journal of Chemical Physics,
100, 12735-12756. doi:10.1021/jp960698w
[18] Baldwin, A.C., Lewis, K.E., Golden, D.M (2004) Very-
low-pressure pyrolysis (VLPP) of group IV (A) tetrame-
thyls: Neopentane and tetramethyltin, International
Journal of Chemical Kinetics, 11, 529-542.
doi:10.1002/kin.550110506
[19] King, K.D. and Nguyen, T.T. (2004) Very-low-pressure
pyrolysis (VLPP) of pentynes. II. 4-methylpent-2-yne
and 4,4-dimethylpent-2-yne. Heats of formation and
resonance stabilization energies of methyl substituted
propargyl radicals, International Journal of Chemical
Kinetics, 13, 255-272. doi:10.1002/kin.550130305
[20] King, K.D. and Goddard, R.D. (2004) Very-low-pressure
pyrolysis (VLPP) alkyl cyanides.II. n-propyl cyanide and
isobutyl cyanide. The heat of formation and stabilization
energy of the cyanomethyl radical, International Journal
of Chemical Kinetics, 7, 837-855.
doi:10.1002/kin.550070606
[21] Johnson, R.P. and Price, S.J. (1972) The pyrolysis of
tetramethyltin, Canadian Journal of Chemistry-NRC Re-
search Press, 41, 50-54.
[22] Daly, M. and Price, S.J. (1976) Determination of tetra-
ethyltin decomposition by the toluene carrier method and
studies of the reaction of ethyl radical with toluene, Ca-
nadian Journal of Chemistry-NRC Research Press, 54,
1814-1819.
[23] Chen, C.-J., Back, M.H., and Back, R. A. (1975) The
thermal decomposition of methane. I. Kinetics of the
primary decomposition to C2H6 + H2; rate constant for
the homogeneous unimolecular dissociation of methane
and its pressure dependence, Canadian Journal of Chem-
istry-NRC Research Press, 53, 3580-3590.