90
M. Hodny, J.F. Hershberger / Chemical Physics Letters 645 (2016) 88–91
Figure 5. HCN yield as a function of CH3Br (solid circles) and C2H6 (solid squares)
reagent pressure. All data obtained at 298 K. P(ICN) = 0.1 Torr, P(SF6) = 0.5 Torr,
P(CH3Br) = variable (lower trace only), P(C2H6) = variable (upper trace only).
Figure 3. Arrhenius Plot of ln k (with k in cm3 molecule−1 s−1) vs. 1/T, for the
CN + CH3Br reaction. Error bar represents one standard deviation.
abstraction reactions, and is in contrast to the positive activation
energy previously reported for the CN + CH3Cl reaction [32], sug-
gesting that a different mechanism may dominate. An Arrhenius fit
to the data provides the following expression:
produced CN radicals have reacted with the reagent. At low reagent
pressures, competing CN removal processes such as self reaction,
diffusional decay, and reaction with trace oxygen result in lower
HCN product yields. The key result of this experiment is that it
demonstrates that the HCN yield of the CN + CH3Br is much lower
than that of the CN + C2H6 reference reaction (which is assumed
to produce HCN in unity yield). By comparing yields of HCN at
pressures of CH3Br or C2H6 of 1.0 Torr or greater, we estimate a
branching fraction of only 0.12 0.02 for channel (1a). This is in
qualitative agreement with ab initio calculations [33] that sug-
gested the possibility of alternate product channels such as (1b),
(1c), or (1d). Most HCN yield experiments were performed at 298 K,
but we did perform one experiment at an elevated temperature
(473 K), obtaining identical results as those at 298 K.
Note that in the above discussion, the assignment of unity
yield for HCN production in the CN + C2H6 reference reaction is an
assumption, and has not been experimentally proven to our knowl-
edge. We can, however, compare the HCN yields in this reaction to
a rough estimate of the initial number of CN radicals created in our
experiment, based on a previously measured 266-nm absorption
coefficient of the ICN precursor (∼0.009 cm−1 Torr−1, base e) [34],
assuming a unity quantum yield for CN production from ICN pho-
tolysis, and the measured pulse energies in this experiment. We
typically find that the HCN yield is ∼70% of the estimated [CN]0.
Although the uncertainties in this estimate are substantial, it does
demonstrate that HCN is at least the dominant channel of the ref-
erence reaction.
k1(T) = (2.20 0.6) × 10−12 exp(453 98/T) cm3 molecule−1 s−1
where the error bars represent one standard deviation. This leads
to an activation energy of Ea = −3.76 0.8 kJ mol−1. At 298 K, the
rate constant is k1 = (1.10 0.1) × 10−11 cm3 molecule−1 s−1
.
3.2. Product yields
Infrared diode laser absorption spectroscopy was used to detect
nals. The peak amplitudes of these signals were converted into
absolute number densities using linestrengths from the HITRAN
database [35]. For the P(19) line at 3251.822 cm−1, the linestrength
is Svj = 8.31 × 10−20 cm molecule−1. Figure 5 shows the resulting
HCN yield as a function of CH3Br and C2H6 pressures. As shown,
the HCN yield increases with increasing reagent (CH3Br or C2H6)
pressure up to a limiting value at which all of the photolytically
One issue that could affect the product yield results is the pos-
sibility that the reaction produces HCN with nascent vibrational
or rotational excitation. The linestrengths in the HITRAN database
that we use to convert transient signal amplitudes assumes a room
temperature Boltzmann distribution of quantum states. Any rota-
tional relaxation is expected to be quickly relaxed by collisions,
and the use of SF6 buffer gas probably causes vibrational excita-
tion to be also relaxed, but we are not aware of literature data
that verifies this. Therefore, in order to test whether this is an
issue, we performed some experiments in which we measured the
static HCN absorption after firing multiple photolysis laser shots
(typically 50), using our standard reaction mixtures. This approach
is generally less accurate than using transient signals, because of
possible secondary chemistry due to buildup of products, any sec-
ondary chemistry slower than the timescale of the transient signals,
and possible wall reactions. Nevertheless, this approach allows any
vibrational excitation to relax to a Boltzmann distribution. We find
that the ratio of static HCN produced by 50 photolysis shots upon
Figure 4. Transient infrared absorption signals for HCN product detection from the
CN + CH3Br and CN + C2H6 reactions at 298 K. P(ICN) = 0.10 Torr, P(CH3Br) = 1.6 Torr
(lower trace only), P(C2H6) = 1.6 Torr (upper trace only), P(SF6) = 0.5 Torr.