 |  |
Astron. Astrophys. 333, 1007-1015 (1998)
4. Isomerization of H2 NC (1 A1) into HCNH ( )
Given the barrier calculated for formation of H2 NC
in its excited 3 B2 state
from C + NH3 reactants, the only
possibility for production of this metastable ion via reaction (1)
would appear to lie in the stabilization of ground state H2
NC , which is formed as a primary product by the
reaction occurring on the ground potential surface (Figs. 1 and 4).
Although there is formally enough total energy for isomerization of
H2 NC into HCNH
(Fig. 4) over the H2 NC transition
state, there are several reasons that isomerization need not be rapid
or even occur at all. First, the exothermicity of reaction may go
sufficiently into relative translation between H2 NC
and H that there is not enough internal energy
left in H2 NC for it to overcome the
transition state barrier. Secondly, there is a large amount of angular
momentum in ion-polar molecule collisions and this angular momentum,
if converted into rotational angular momentum of the H2 NC
product, rather than into relative angular
momentum of the two products, effectively slows the rate of
isomerization since only vibrational energy is useful in this process
(Herbst 1996). Thirdly, relaxation of the ground state H2
NC ion product via emission of infrared
radiation (and, in the laboratory, by collision) to
vibrational-rotational states below the barrier to isomerization may
be more rapid than isomerization. For these three reasons, it is
necessary to estimate the rate of isomerization and then determine the
fraction of product that can remain as H2 NC
.
Our first consideration is to estimate how much energy goes into
relative translation between the products. Although both statistical
theories and assorted experiments (Illies et al. 1983) show that the
energy of exothermic reactions tends to wind up as vibrational energy
in polyatomic products, experiments also show that in the case of
potential surfaces with barriers, a significant amount of
translational energy, with a distribution centered on the potential
energy difference between transition state and products, can be
released (Jarrold et al. 1986). Reference to Fig. 4 shows that the
transition state between the H3 NC
energy minimum and the H2 NC
(1 A1) + H products lies at an energy of
21.4 kcal/mol above the reactants. It is most likely that the
separating products carry away this much translational energy, leaving
the remainder of the reaction exothermicity of H2 NC
(1 A1) + H products
( kcal/mol) as vibrational-rotational energy of
H2 NC (1 A1).
Since the transition state energy for isomerization of this ion into
the lower energy HCNH ion is only 17.9 kcal/mol
(see Fig. 4), the H2 NC ion possesses
sufficient internal energy to isomerize as long as its vibrational
energy exceeds 17.9 kcal/mol. To have less vibrational energy than
this, the ion must relax rapidly or have large amounts of angular
momentum. Both possibilities are addressed below.
The rate of isomerization over a barrier can be treated by the RRKM
statistical method (Herbst 1996). In this formulation, the rate
k(s-1) as a function of vibrational-rotational energy E and
angular momentum quantum number J is given by the expression
![[EQUATION]](img45.gif)
where h is Planck's constant, N refers to the
total number of vibrational states of the transition state from its
minimum allowable energy through
, and refers to the
density of vibrational states of the H2 NC
isomer at energy . Note
that kcal/mol is the energy difference between
H2 NC and the transition state. The
rotational energy for either the ion or
transition state can be approximated as that of a spherical top with
one rotational constant B equal to the geometric mean (cube root) of
the product of three rotational constants. For both the number of
states of the transition state and the density of states of the
complex, the semi-classical expression of Whitten and Rabinovitch
(1964) can be used.
With our calculated energies as well as the vibrational frequencies
(Table 3) and rotational constants of the H2 NC
ion and transition state (see Fig. 3), we can
determine k(E,J). Using the assumption that 21.4 kcal/mol of reaction
energy goes into relative translational energy and is thus not
available for isomerization of H2 NC ,
we still determine that isomerization is very rapid for a wide range
of angular momenta and vibrational energies. Let us consider two
extreme cases: when the ion possesses no angular momentum (J = 0) and
when it possesses a large angular momentum, quantum number J = 78,
which is the maximum quantum number for which there is sufficient
vibrational energy for isomerization to occur. For the former case, we
calculate that the isomerization rate for H2 NC
into HCNH before any
vibrational relaxation can occur is while the
equivalent rate after sufficient vibrational relaxation such that
H2 NC can just overcome the
transition state barrier is . For the latter
case, we calculate the isomerization rate to be
(vibrational relaxation in this case will prevent isomerization).
Consideration of tunneling under the transition state barrier (Herbst
1996) will enlarge these already large values. We conclude that
isomerization is a rapid process for all allowable J; in particular,
it is much more rapid than radiative relaxation, which is known to
occur at a rate of only 10-103 s-1 for
vibrational states (Herbst 1982; Herbst & Dunbar 1991). Thus we
are led to the following picture: after the H2 NC
ion is formed in its ground electronic state, it
will rapidly isomerize to the HCNH ion for
angular momenta from J = 0 to a maximum value, and, as can be shown by
calculations similar to the above, this ion will rapidly isomerize in
the backwards direction so that an equilibrium will quickly be
achieved between the two ions. As the ions slowly relax vibrationally,
the equilibrium will reflect the change in energy. The relevant
equilibrium for our purposes will be that corresponding to a
vibrational energy just equal to the transition state barrier. A
similar analysis was performed previously by De Frees et al. (1985)
for the CH3 CNH - CH3 NCH
isomerization.
To determine the equilibrium populations of ground state
H2 NC and HCNH ,
it is necessary to calculate the so-called equilibrium coefficient K,
which is related to the ratio of the forward and backward rates of
isomerization. The equilibrium coefficient in the microcanonical
formulation is given by the equation
![[EQUATION]](img54.gif)
where once again refers to the vibrational
density of states, the subscripts L and NL refer to the linear and
non-linear isomers, respectively, and refers to
the (positive) difference in energy between the two isomers. Part of
this expression can be obtained by dividing the RRKM rate equations
for the forward process by that for the backward process since the two
processes possess the same transition state. But, one must also
consider the rotational degeneracies of the two isomers, which are 2J
+ 1 for the linear ion and (2J + 1)2 for the non-linear ion
(assumed to be a spherical top). In addition, there is a symmetry
factor for the two-fold symmetry of the
non-linear ion. Expression (II) should be evaluated for vibrational
energies corresponding to the energy of the transition state.
Restricting attention to the energy at the barrier height, and using
the Whitten-Rabinovitch (1964) density-of-states formulation, we
obtain that K(J) = 3168/(2J + 1) for all values of J for which a
sufficient amount of vibrational energy can be obtained. For values of
, where is determined by
the amount of relative translational energy, a sufficient amount of
vibrational energy cannot be obtained, and K(J) = 0. For our standard
choice of translational energy, .
The rotationally-averaged equilibrium coefficient K is obtained
from the formula
![[EQUATION]](img60.gif)
where p(J) is the probability that the reaction produces
H2 NC in a specific rotational level
of angular momentum quantum number J. This probability can be
estimated by the statistical relation (Herbst 1996)
![[EQUATION]](img61.gif)
where N is a normalization coefficient. Eq. (IV) holds for values
of J from 0 to a maximum defined by the
stronger of several constraints posed by the details of the reaction
and by conservation of energy. If is
sufficiently large, then the averaged equilibrium constant can be
small, so that isomerization need not destroy a large amount of
H2 NC . After doing the rotational
averaging, however, we find that K = 44 if the relative product
translational energy is set at 21.4 kcal/mol (our preferrred value),
and K = 39 if the relative translational energy is set at 0. The weak
dependence on translational energy means that this parameter is not of
the utmost importance. Since the equilibrium coefficient tells us the
ratio of the population of HCNH product to
H2 NC product, we conclude that the
amount of the latter is only about 2-3% of the total and that HCNH
is the overwhelmingly dominant product for
reaction (1) when it occurs on the lowest energy potential
surface.
© European Southern Observatory (ESO) 1998
Online publication: April 28, 1998
helpdesk@link.springer.de  |