To facilitate the use of the rate file as a standard package, we have calculated
the pseudo-time-dependent evolution of the 395 species in a dark cloud model in
which 
, 
, and 
magnitudes.
We have adopted the same initial elemental abundances as in our calculation
with the 1990 version: H:He:C:N:O:Na:Mg:Si:P:S:Cl:Fe = 
, and give both
early-time (
) and steady-state (
) abundances in
Table 5. In addition to the gas-phase reactions of Table 4, we have included the
grain surface formation of 
with a rate coefficient of
.
Table 5: Fractional abundances with respect to 
at 
(early time) and 
(steady state) years
Table 5: continued
Table 5: continued
The calculated abundances show some significant differences with those published in Millar et al. (1991b). We discuss these noting those reactions which have caused the changes and indicating the most important of these reactions which can be be studied further in the laboratory. We concentrate on the differences in steady-state results since it is more apparent which differences are caused by the new rate coefficients; at early time, differences can be caused by small changes in the time scale for chemical evolution.
The abundance of C atoms is an order of magnitude larger in this model. The
major source of the increase in CI is due to the reaction C
+ NO
C + NO
. Although the rate coefficient for this reaction
is unchanged in this revision of the ratefile, the NO abundance is larger
by 
100 due to an increase in the rate coefficient of the N + OH
NO + H reaction. The increase is due to an adoption of a
different dependence on temperature, from 
to 
,
an increase of 
13 at a temperature of 10 K. This behaviour also is
an indication of some of the difficulties in tracing the origin of the
changes in abundances; an alteration to one rate coefficient can propagate
through to species other than those directly involved in the reaction
concerned. In our analysis here, we have made use of software which
tabulates, for all species, the relative weights of the various
formation and destruction reactions.
The OH abundance is also an order of magnitude larger due to a decrease in the
rate coefficient of the O + OH 
O
+ H reaction. The O
abundance is unchanged by this change to the rate coefficient because the
`throughput', which depends on the product kn(O)n(OH) is unchanged. This
can only occur when the reaction is the major route to formation. At steady
state the reaction contributes 83
to the loss rate of OH.
The increased C and NO abundances have a minor effect on the abundance of
CN since they form this molecule at the 30
level. The major route to
CN is the dissociative recombination of HCNH
with electrons, although
the branching ratio to CN (and those to HCN and HNC) are only estimated
theoretically. An experimental determination of these ratios is needed. The
rate of formation of CN is increased by about a factor of 1.5 but the CN
abundance falls by about 30 because the reaction CN + O
CO + NO is about 70 times faster in RATE95 than in RATE90. This
reaction
has, through altering the abundance of CN, an affect on the
abundance of other species. Note that the CN + O
reaction rate coefficient
has been measured down to 13 K.
The increased abundances of NO and OH also cause the N atom abundance to decrease by about 40 because the reactions of N with NO and OH are very efficient at converting atomic nitrogen into nitrogen-bearing molecules.
Other molecules, particularly oxides, are affected by the increased OH
abundance. They include CO
, which is a factor of 10 larger, and which
is formed by the CO + OH reaction. CO
is unobservable from the ground,
except indirectly via its protonated form, HOCO
, but should be
detectable at IR wavelengths by the ISO mission. The only phosphorus
molecule yet detected in interstellar clouds, PN, decreases by an order of
magnitude due to the indirect effects of OH.
The uncertainty attached to the calculated abundance of OCN in this
model calculation is difficult to quantify as it is uncertain how this
species forms in interstellar clouds. In the ratefile, it is formed in the
reaction of CN with O
, which is less efficient due to the decrease in the
CN abundance. The OCN abundance is a factor of 20 lower in this model.
Finally, the abundance of HPO decreases by about 30 due to the inclusion
of the reaction O + HPO 
PO + OH, which was not included in
RATE90.
The C
H abundance is a factor of around 100 less because of more rapid
loss with O and O
. These reactions,which were assumed to have
activation energy barrriers of 250 K and 3500 K, respectively, are now
taken to be activationless.
The C
H
molecule, which in RATE90 was produced primarily by the
C
H + C
H
C
H
+ H reaction, falls by an
order of magnitude as the C
H abundance is lower.
HC
N also falls by an order of magnitude. Although the formation
reaction CN + C
H
HC
N + H has a larger rate
coefficient in RATE95,
the decrease in the CN abundance by a factor of 30 more than offsets this.
The C
abundance falls by 
10
due to the inclusion of a
new rapid destruction channel, reaction with O atoms, which dominates
over the loss reactions of C
with ions. However, since the products
of this reaction are assumed to be C
and CO, the C
abundances
increases by 
200. Rapid destruction of C
with O atoms is
included in both RATE90 and RATE95.
The formation of CH
CHO increases by around 100 due to a large
increase in the rate coefficient of the radiative association
H
O
+ C
H
C
H
O
+ h
,
from which CH
CHO forms by dissociative recombination.
The heavy molecule CH
C
N decreases in abundance by 
1000 due to the inclusion of an additional channel in the products
of the dissociative recombination of H
C
N
, protonated
CH
C
N. In RATE90, the only products of the recombination
are CH
C
N + H. Thus, proton transfer, followed by
dissociative recombination, simply recycles the neutral molecule
and its effective destruction is small. In RATE95, we include,
with an equal branching ratio, a channel to CH
+ HC
N,
which breaks the recycling process and leads to a much larger
destruction rate for CH
C
N.
Because the results of the calculation shown in Table 5 are for physical
parameters similar to those in the dark dust cloud TMC-1, we show in Table 6
a comparison between the calculated abundances at early time
(
) and steady-state (
) and those observed
toward TMC-1. This calculation has not been optimised in the sense of
searching for the best-fit through looking for the best time,
varying elemental abundances, cosmic ray
ionisation rate, etc., but it does show that the chemistry is particularly
suited to this type of source. In general, around one-half of the
molecules agree to within a factor of 5
at early time, although there are some notable exceptions. SO
and 
both increase in abundance rapidly at late times and agree with the
observations around steady-state, although the 
abundance is too
large at this time. Some molecules, 
(n =
3-6) for example, are
much too large at early time. This may indicate that they have faster loss
reactions with O atoms than adopted in the rate file, where we have assumed
an activation energy barrier of 250 K. 
is too low at all times and
this might indicate that the radiative association rate coefficient for the
reaction has been underestimated (see
Millar & Herbst 1990
for a discussion). The abundances of HCS
and HCNH
are roughly
an order of magnitude below the observations. This difference can be
resolved by the adoption of ion-dipolar rate coefficients for the proton
transfer reactions of CS and HCN. Finally, we note that contrary to
many statements in the
literature, the 
abundance agrees with the observations at early times
and does not need a special chemistry to be invoked.
Table 6: Comparison of observations toward TMC-1 (Ohishi
et al. 1992)
with fractional abundances with respect to 
at 
(early
time)
and 
(steady state) years
