Collisional stabilization and thermal dissociation of highly vibrationally excited C9H12+ ions from the reaction O 2+ + C9H12 → O2 + C9H12+

Article abstract of DOI:10.1021/jp048132s

Highly vibrationally excited n-propylbenzene cations, C₉H ₁₂⁺*, were prepared by the charge transfer reaction O₂⁺ + C₉H₁₂ → O₂ + C₉H₁₂⁺* in a turbulent ion flow tube. The subsequent competition between fragmentation of C₉H ₁₂⁺* into C₇H₇⁺ + C₂H₅ and stabilization in collisions with N₂ was studied at temperatures in the range 423-603 K and at pressures between 15 and 200 Torr. Most of the C₇H₇⁺ is the aromatic benzylium isomer, while the fraction of the minor species, seven-membered-ring tropylium, increases with T, from 5 to 20%. Minor fragments are C ₆H₆⁺, C₇H₈⁺, and C₈H₉⁺, Energy-transfer step sizes (ΔE) for collisional deactivation are obtained by combining the stabilization versus fragmentation ratios measured as a function of pressure in this study with fragmentation rates from the literature. The values are compared with related information for other excited molecular ions and are similar to those of their neutral analogues. At the highest temperatures, C ₉H₁₂⁺ was also observed to pyrolyze after collisional stabilization. Employing unimolecular rate theory, the derived rate constants for thermal dissociation of C₉H₁₂⁺ are related to values derived from the specific rate constants k(E,J) for fragmentation. Good agreement is found between measured and predicted pyrolysis rate constants. This allows us to confirm the dissociation energy of C ₉H₁₂⁺ into C₇H₇ ⁺ (benzylium) and C₂H₅ as 166.9 (±2.2) kJ mol^-1 (at 0 K).

Full text of DOI:10.1021/jp048132s

9652
J. Phys. Chem. A 2004, 108, 9652-9659
+
Collisional Stabilization and Thermal Dissociation of Highly Vibrationally Excited C9H12
Ions from the Reaction O2 + C9H12 f O2 + C9H12
+
+†
Abel I. Fernandez,‡,§ A. A. Viggiano,*,‡ Thomas M. Miller,‡ S. Williams,‡ I. Dotan,|
J. V. Seeley, and J. Troe#
Air Force Research Laboratory, Space Vehicles Directorate, 29 Randolph Road, Hanscom AFB,
Massachusetts 01731-3010, NRC Research Associateship Program, Keck Center of the National Academies,
500 Fifth Street, NW, GR 322A, Washington, D.C. 20001, Department of Natural Sciences, The Open
UniVersity of Israel, Ramat AViV, Tel AViV, Israel, Department of Chemistry, Oakland UniVersity,
Rochester, Michigan 48309-4401, and Institute for Physical Chemistry, UniVersity of Goettingen,
Tammannstrasse 6, D-37077 Goettingen, Germany
ReceiVed: April 29, 2004; In Final Form: July 8, 2004
Highly vibrationally excited n-propylbenzene cations, C9H12+*, were prepared by the charge transfer reaction
O2+ + C9H12 f O2 + C9H12+* in a turbulent ion flow tube. The subsequent competition between fragmentation
+
of C9H12+* into C7H7 + C2H5 and stabilization in collisions with N2 was studied at temperatures in the
+
range 423-603 K and at pressures between 15 and 200 Torr. Most of the C7H7 is the aromatic benzylium
isomer, while the fraction of the minor species, seven-membered-ring tropylium, increases with T, from 5 to
20%. Minor fragments are C6H6+, C7H8+, and C8H9+. Energy-transfer step sizes ∆E for collisional deactivation
are obtained by combining the stabilization versus fragmentation ratios measured as a function of pressure in
this study with fragmentation rates from the literature. The values are compared with related information for
other excited molecular ions and are similar to those of their neutral analogues. At the highest temperatures,
C9H12+ was also observed to pyrolyze after collisional stabilization. Employing unimolecular rate theory, the
+
derived rate constants for thermal dissociation of C9H12 are related to values derived from the specific rate
constants k(E,J) for fragmentation. Good agreement is found between measured and predicted pyrolysis rate
+
+
constants. This allows us to confirm the dissociation energy of C9H12 into C7H7 (benzylium) and C2H5 as
166.9 ((2.2) kJ mol-1 (at 0 K).
1. Introduction
us to better investigate pressure dependences. Recent improve-
ments to the TIFT allow measurements over a broader temper-
Charge transfer can generate chemically activated molecular
ions which subsequently undergo a variety of reactive and
nonreactive processes. Reactive processes include dissociation
at high excitation energies and secondary reactions. The
unreactive or energy-dissipative processes include collisional
stabilization and radiative cooling of vibrationally excited
molecular ions. At high temperatures, the stabilized molecules
subsequently undergo thermal dissociation.
In previous work, we have generated highly vibrationally
excited alkylbenzene cations by charge transfer from small ions
to alkylbenzenes in a high-temperature flowing afterglow
(HTFA) and in a selected ion flow tube (SIFT).1-3 Branching
ratios were determined as a function of the temperature, and
mechanisms were determined. The comparison of charge
transfer breakdown curves taken at about 1 Torr to data from
low-pressure photodissociation experiments suggested that the
buffer gas had an influence by either collisionally stabilizing
and/or thermally dissociating the ions. The small pressure
variation (at most 0.2-1 Torr), however, did not allow for a
systematic analysis of pressure effects. The subsequent con-
struction of a high-pressure turbulent ion flow tube4,5 (TIFT),
operating at buffer gas pressures from 15 up to 700 Torr, allows
ature range than in previous studies. Studies of this type, for
+
example, the reaction of O2 with ethylbenzene, allowed for
an investigation of the competition between unimolecular
fragmentation of ethylbenzene cations (forming methyl radicals
+
and C7H7 isomers: the aromatic benzylium and the seven-
membered-ring tropylium ions6) and collisional stabilization of
the excited ethylbenzene ions.7 Since the specific rate constants
k(E) for fragmentation are known from separate absolute rate
measurements,8,9 the effective rates for collisional stabilization
could be deduced.7 These results were analyzed in a simplified
master equation approach10 leading to the product of the total
collision frequency, Z, and average step size for deexcitation,
∆E , at each temperature. ∆E values were obtained by
identifying Z as the ion-induced dipole capture rate constant.
The ∆E values were found to be similar to those for the
corresponding neutral alkylbenzenes.
Little information exists on collisional energy transfer from
molecular ions which are highly vibrationally excited. While
relative collision efficiency measurements are available,11-13 it
is difficult to put them on an absolute scale of energy transfer
rates. Since our O2 + ethylbenzene experiments7 have led to
+
values of the product Z ∆E , it appears desirable to derive
similar information also for other excited polyatomic ions.
† Part of the special issue “Tomas Baer Festschrift”.
‡ Air Force Research Laboratory.
§ NRC Research Associateship Program.
| The Open University of Israel.
Oakland University.
This is the aim of the present series of articles in which
experiments for the reactions of O2+ with n-propylbenzene (this
work) and with n-butylbenzene14 are described. A second goal
of the our work is to study the pyrolysis of the stabilized
# University of Goettingen.
10.1021/jp048132s CCC: $27.50 © 2004 American Chemical Society
Published on Web 10/05/2004

+
Stabilization and Dissociation of C9H12 Ions
J. Phys. Chem. A, Vol. 108, No. 45, 2004 9653
alkylbenzene ions and to attempt to relate specific rate constants
k(E) for fragmentation with the derived thermal dissociation rate
constants. The present work, involving n-propylbenzene ions,
presents this approach for the first time.
The chemistry of the O2+ + n-propylbenzene (C9H12) system
involves numerous processes:15
+
O2+ + C9H12 f O2 + C9H12
*
k1
k2
k3
k4
(1.1)
(1.2)
(1.3)
(1.4)
(1.5)
(1.6)
(1.7)
C9H12+* f C7H7+ (Bz+) + C2H5
C9H12+* f C7H7+ (Tr+) + C2H5
C9H12+* + M f C9H12+ + M
C9H12+* f C9H11+ + H
+
Figure 1. Typical low-resolution mass spectrum of the O2 + C9H12
system. The dashed peak at mass 32 corresponds to the undepleted
+
O2 signal before the addition of propylbenzene. The peak at mass
133 corresponds to a secondary reaction of benzylium (C7H7+) with
C9H12; see eq 2.1.
k5
k6
k7
C9H12+* f C8H9+ + CH3
is also heated. Last, it was found that preheating of the buffer
gas is needed. During the course of this work, the preheater
design was changed to allow for higher temperatures. The new
preheater consists of two Mellen split type (clam shell) tubular
heating elements that surround a tubular piece of copper; the
gas flows through narrow channels (which traverse the copper
tube along the axis) that have a radial-spoke cross section. The
pattern was chosen to promote efficient heating.
C9H12+* f C6H6+ + C3H6
(Bz+ denotes benzylium, Tr+ is tropylium). Channels 1.3 and
1.5-1.7 are minor under the present conditions and can be
mostly ignored. The bulk of this article focuses on reactions
1.1, 1.2, and 1.4. In addition, at temperatures approaching 600
A typical mass spectrum is shown in Figure 1, with and
without the propylbenzene reactant gas. In the absence of
+
K, collisionally stabilized C9H12 ions can pyrolyze,
+
propylbenzene, the O2 signal is dominant (>99%).
C9H12+(+M) f C7H7+ + C2H5(+M)
k8 (1.8)
Common impurities are H3O+(H2O)0,1 and O2+(H2O). Since
the relative concentrations of the impurities were usually less
than 2% (sometimes much less), no corrections to the recorded
branching ratios were made. When propylbenzene is added, ions
at 78, 91, 92, 105, 120, and 133 amu are observed. In Figure 1,
the 92 amu peak is hidden in the large 91 amu peak. Higher
resolution spectra are used to separate the two peaks, and an
isotope correction is made. The mass 78 and 105 amu peaks
are present at the same levels as impurities, ca. 0.5-2% at low
temperatures. At higher temperatures, their yields increase and
can be investigated. The dissociative and nondissociative charge
transfer products at 91 and 120 amu, respectively, dominate.
The peak at mass 133 is due to a reaction between benzylium
and propylbenzene3, that is,
Secondary reactions occur but are accounted for by extrapolating
to zero [C9H12].
There is also a practical aspect to the measurements. The
bottleneck for combustion of hydrocarbon fuels often is the
ignition delay caused by the primary fragmentation of the fuel
molecules. Injection of radicals or ions can considerably speed
up the ignition and may be required for advanced hydrocarbon-
fuelled airbreathing propulsion systems.1 The present work
contributes to this practical aspect of combustion technology
under conditions more appropriate to real life engines.
2. Experimental Technique
Our experiments were conducted in the TIFT which has been
described in detail before;4,16 therefore the technique is only
briefly summarized here. The TIFT is similar to low-pressure
flow tubes except that larger flow rates are used to achieve
higher pressures and Reynolds numbers. A liquid nitrogen
storage vessel supplies the N2 carrier gas, which is preheated
in a sidearm. The ions are created upstream in an off-axis corona
discharge source and are entrained into the N2 flow in the
sidearm. The gas enters the flow tube where the neutral reagent
is injected through a moveable on-axis tube. At the end of the
flow tube, most of the gas is removed by a large mechanical
pump while a small fraction is sampled through a 150 µm orifice
in a truncated nose cone. The core of the ensuing supersonic
expansion is sampled through a skimmer into a mass spectrom-
eter. The ions are analyzed by a quadrupole mass filter and
counted by a discrete dynode electron multiplier.
C7H7 (Bz+) + C9H12 f C10H13+ + C6H6
(2.1)
+
Branching ratios of the ion products were recorded as a function
of the neutral reagent concentration. Secondary chemistry was
taken into account by extrapolating to zero concentration. The
accessible temperature and pressure ranges were limited by the
allowable intensity of impurity ions (<2%) at low T and high
P, thermal decomposition of C9H12+ at high T, and small signals
at low P.
Up to 523 K, the pyrolysis reaction 1.8 could be neglected.
In this case, the mass 120 versus 91 ratio, which we call S/D,
can be symbolically written as
+
[C9H12
]
k4[M]
S
)
)
(2.2)
+
D
k2 + k3
[C7H7 ]
To elevate the temperature, six zones of heating are used and
the thermocouple that monitors the gas temperature is main-
tained to (2 K. The main flow tube is heated in four zones,
two short zones at either end, a long middle section, and a zone
inside the vacuum chamber just upstream of the nose cone. The
connection between the corona discharge tube and the sidearm
+
Here it is assumed that at zero pressure, all of the C9H12
dissociates, but in section 3, it is shown that the S/D values
exhibit a nonzero intercept at low pressures, indicating that some
of the C9H12+ transfer product is stable at zero pressure. Minor

9654 J. Phys. Chem. A, Vol. 108, No. 45, 2004
Fernandez et al.
channels are excluded in the dissociation signal. At T g 573
K, S/D values vary with the injector position and therefore time,
+
indicating that thermal decomposition of C9H12 is occurring.
+
Our first attempt to determine the pyrolysis rate of C9H12
+
involved direct observation of the depletion of C9H12 as a
function of distance by using NO+ or C6H6 as reagent ions.
+
Because the ionization potentials of C6H6, NO, and C9H12 are
similar,17 this avoided the unwanted presence of highly vibra-
tionally excited C9H12+. Unfortunately, this approach yielded
uninterpretable results and was abandoned. Explanations for the
failure include interference from the isomerization of the
benzene cation to the five-membered-ring species, fulvene,18
and hydrogen abstraction reactions when NO+ was used as the
charge transfer agent.
The method that was successful involves the deconvolution
of the effects of time and pressure. An advantage over the
previous method is that (S/D) in the absence of thermal
decomposition (t ) 0), called (S/D)0, can also be obtained. S/D
was measured in the usual way at several distances (i.e., reaction
times) at the same T and P. This was repeated at numerous
pressures, P ) 30-250 Torr, at 573 and 603 K.
Figure 2. Thermal decomposition of C9H12+. [C9H12+]/[C7H7+] is
plotted vs reaction time at pressures of 100 Torr of N2. Best fits to eq
2.4, shown as curves, yield values of (S/D)0, the ratio of stabilization
to dissociation in the absence of thermal decomposition, and k8, the
+
first-order thermal decomposition rate coefficient of C9H12
.
A steady-state solution of the kinetics, modeled by eqs 1.1-
1.4 and 1.8 which include stabilization, dissociation, and thermal
dissociation, yields
k4[M]
[1 - exp(-k8t)]
(k2 + k3)
k4[M]
S
D
≈
(2.3)
k4[M]
k8t 1 +
-
[1 - exp(-k8t)]
(
)
(
)
(k2 + k3)
(k2 + k3)
Minor channels are neglected. Extrapolating to zero time leads
back to eq 2.2, as expected; this value is (S/D)0. At high
temperatures and nonzero time, substitution of eq 2.2 into eq
2.3 yields the following:
+
Figure 3. Rate constants k1 for the reaction of O2 with C9H12 as a
function of pressure at various temperatures.
yield to be measured by alternately using low and high
propylbenzene flows. Low concentrations of propylbenzene are
(S/D)0[1 - exp(-k8t)]
+
S
D
used to measure the total relative concentration of C7H7
,
≈
(2.4)
whereas at high concentrations Bz+ is completely removed.
Subtraction of the two limits yields the amount of Bz+. At 423
K, the Tr+ yield was found to be about 1%, and at 523 K the
value was roughly 7%. At 573 K, a yield around 20% was
found. The uncertainty is considerable since the propylbenzene
cation starts to decompose at this temperature and the kinetics
of reaction 2.1 has not been measured at high temperature.
k8t(1 + (S/D)0) - (S/D)0[1 - exp(-k8t)]
We assume that the decrease of the product of concentrations
[C9H12][O2+] during the course of the reaction is insignificant
since we extrapolate to zero extent of reaction. Nonlinear least-
squares fits of the time dependence of S/D at each pressure to
eq 2.4 yield both k8 and (S/D)0. Note that this involves both an
extrapolation to [C9H12] ) 0 and the fit to the time dependence,
thereby introducing noticeable scatter in the data. The fitted k8
values at 573 K varied by a factor of 2. The scatter at 603 K is
less since the pyrolysis rate is larger. The factor of 2 scatter is
an estimate of the accuracy. Figure 2 shows a typical time
dependence of S/D at T ) 573 and 603 K and at a nitrogen
pressure of 100 Torr. The pyrolysis decay is easily detected.
We, therefore, have the interesting situation of a chemical
activation process which is initiated by a charge transfer reaction
and which, after collisional stabilization of the highly excited
products, is followed by thermal dissociation of the stabilized
species. The intrinsic kinetics of chemical activation and thermal
dissociation is connected and must be interpreted in a consistent
way, and this is done here.
3. Results
The experimental results are of three types: (1) the total rate
constant k1 of the charge transfer process 1.1, (2) branching
ratios as a function of temperature and of pressure (mostly the
+
+
ratio of C9H12 to C7H7 but the other channels are also
reported), and (3) the thermal decomposition rate constants at
high temperature as a function of pressure. Rate constants k1
are plotted in Figure 3, and S/D values are listed in Table 1.
Table 2 lists the thermal decomposition rates, and Table 3 lists
the ratio of mass 91 to the minor products and to the fraction
of mass 91 that is tropylium.
The rate constants shown in Figure 3 show no dependence
of k1 on temperature or pressure over the ranges 423-523 K
and 30-200 Torr. The uncertainty of individual rate constants
was about (25%. Part of the scatter is because most of the
effort was focused on the more interesting branching information
where low depletion is used. Rate constants are better obtained
under high depletion. The average value of all measurements
The fragmentation of C9H12+* is known to lead to two
isomers of C7H7+, Bz+ and Tr+.15 The former contains the six-
membered aromatic ring and is reactive with propylbenzene to
produce C9H11+ (mass 119) and C10H13+ (mass 133) via reaction
2.1. By contrast, the latter is comprised of an unreactive seven-
membered ring. This difference in reactivity allows for the Tr+

+
Stabilization and Dissociation of C9H12 Ions
J. Phys. Chem. A, Vol. 108, No. 45, 2004 9655
TABLE 1: Stabilization vs Dissociation Yields S/D )
TABLE 3: Ratio of Concentrations of Major- to Minor-Ion
a
+
+
[C9H12+]/[C7H7+] in N2
Fragments from the Reaction O2 + C9H12 f C7H7
+
Other (Minor) Ion Fragments
T/K
P/Torr
[N2]
S/D
c1
c2
[C7H7+]/
[C7H7+]/
[C7H7+]/
[Tr+]/
423
15
30
0.31
0.65
1.15
1.69
0.29
0.62
1.05
1.58
2.11
3.19
0.28
0.60
0.99
1.51
2.03
3.04
0.26
0.54
0.92
1.37
1.81
0.505
0.835
1.26
1.68
2.52
3.35
4.19
0.474
0.798
1.19
1.59
2.39
3.19
4.00
1.18
1.61
2.28
3.18
1.07
1.07
1.43
1.69
1.94
2.22
0.87
0.96
1.13
1.29
1.48
1.74
0.75
0.66
0.83
0.86
1.00
0.63
0.67
0.82
0.90
0.95
1.08
1.36
0.46
0.54
0.65
0.72
0.66
0.92
1.32
0.59
1.44
+
+
+
+
a,b
T/K
P/Torr
[C7H8
]
[C6H6
]
[C8H9
]
[C7H7
]
tot
50
423
15
30
17.8
16.3
18.2
23.0
25.9
28.4
39.8
51.0
64.8
78.0
60.0
0.010
75
448
473
15
0.86
0.79
0.516
0.323
50
30
75
50
448
473
15
16.1
15.2
16.7
17.3
19.6
18.3
75
30
20.3
23.1
26.9
26.5
28.8
25.0
22.0
24.8
25.8
25.4
100
150
15
50
83.0
75.4
69.0
79.3
40.4
63.3
67.3
87.8
67.8
93.5
75
100
150
15
30
50
0.010
75
30
18.5
14.8
18.8
21.3
16.4
16.1
16.4
16.3
17.8
18.3
18.9
25.8
21.1
100
150
15
50
75
0.041
0.030
523
573
0.68
0.54
0.134
0.147
100
150
200
15
30
50
75
523
19.1
21.9
20.8
25.5
23.7
20.0
18.4
25.2
53.6
60.1
76.4
62.3
53.6
65.9
0.032
0.091
100
30
30
50
50
75
100
150
200
250
30
50
75
100
150
200
250
75
100
150
200
0.070
a Fraction of Tr+, the seven-membered-ring tropylium isomer of
C7H7+; the aromatic benzylium isomer is the major species. b Data of
603
0.48
0.0599
+
[Tr+]/[C7H7
]
tot
at 573 K, about 18% and invariant from 30 to 200
Torr, are not included here due to interference from thermal decomposi-
tion.
a [N2] in 1018 molecule cm-3, c2 in 10-18 molecule-1 cm3.
TABLE 2: Thermal Dissociation Rate Constants k8 for
+
+
C9H12 f C7H7 + C2H5
T/K
P/Torr
k8/s-1
T/K
P/Torr
k8/s-1
573
30
50
75
37
44
32
22
27
34
603
30
50
101
109
109
98
75
75
100
150
200
250
100
150
200
250
75
109
138
was found to be k1 ) 2.1 × 10-9 cm3 molecule-1 s-1, in
agreement with our previous measurements,1,3 and confirms the
accuracy of the TIFT approach. This value is also in good
agreement with the Langevin rate constant for capture of C9H12
by O2+, which is calculated as kL ) 1.89 × 10-9 cm3 molecule-1
s-1 on the basis of an estimated19 polarizability of C9H12 of R
) 16.1 × 10-24 cm3 (contributions from polarizability anisotro-
pies and quadrupole and permanent dipole moments at this level
can be neglected).
Figure 4. Ratio of stabilization to decomposition vs [N2]. Temperatures
range from 423 to 603 K, while pressure was varied from about 15 to
250 Torr. At T g 573 K, thermal decomposition is considerable and is
taken into account.
nondissociative products. Therefore, the mechanism of reactions
1.1-1.8 is extended to (excluding minor channels)
+
O2+ + C9H12 f O2 + C9H12
Rkcap
(3.1)
The measurements of S/D (or (S/D)0 at higher temperatures)
are presented in Table 1 and plotted as a function of the
temperature and the buffer gas concentration in Figure 4. The
data show that lower temperatures and higher pressures favor
the formation of the nondissociative product, that is, give larger
S/D. The measurements exhibit a roughly linear dependence
on [N2] or equivalently pressure. One also notices an apparent
intercept of the S/D versus [N2] curves as [N2] f 0 which is in
contrast to the predictions from eq 2.2.
+
O2+ + C9H12 f O2 + C9H12
*
(1 - R)kcap (3.2)
C9H12+* f C7H7+ + C2H5
k(E)
γcZ
(3.3)
(3.4)
C9H12+* + M f C9H12+ + M
C9H12+ (+ M) f C7H7+ + C2H5 (+ M)
k8 (3.5)
C7H7 in reaction 3.3 stands for the sum of Bz+ and Tr+ but,
as shown in Table 3, is mostly the former. E denotes the internal
energy of the chemically activated propylbenzene cations, and
+
The present results are similar to our previous observations
for the O2 + ethylbenzene reaction. The intercept indicates
that there is a fraction, R, of reaction 1.1 that leads to
+

9656 J. Phys. Chem. A, Vol. 108, No. 45, 2004
Fernandez et al.
TABLE 4: Energy Transfer Properties for Highly Excited
Propylbenzene Ionsa
All minor channels decrease with increasing pressure (relative
to [C9H12+]). However, the data had too much scatter to ascertain
if individual channels had different dependences. No attempt
was made to correct these channels for impurities, since the
procedure was uncertain. The most abundant of the minor
T/K
k(E)/γc s-1
I(T)
γc (T ) 0)
- ∆E /hc cm-1
423
448
473
523
573
603
7.18 × 108
2.34 × 109
1.12 × 109
2.89 × 109
2.87 × 109
7.33 × 109
0.278
0.240
0.207
0.152
0.110
0.0905
0.195
0.069
0.168
0.089
0.123
0.059
336
119
288
153
212
101
+
channels is C7H8+, which is approximately 5-6% of the C7H7
channel. There is no temperature dependence within our
+
+
uncertainty, as is true for all minor species. C6H6 and C8H9
are about 4 and 2% of the C7H7+ signal, respectively. The Tr+
is small except in the temperature range where thermal dis-
sociation becomes important, where 20% of C7H7 was
a See the text; analysis with eqs 4.4 and 4.5; Ech/hc ) 27 030 cm-1
,
k(Ech) ) 3.90 × 107 s-1, s* ) 7.608, Z ) 6.5 × 10-10 cm3 molecule-1
s-1, c1 and c2 from Table 1.
+
observed as tropylium. However, the monitor ion chemistry was
not measured at high temperature.
k(E) is the specific rate constant for their fragmentation. We
identify Z with the Langevin capture rate constant between
4. Analysis of Experimental Results
+
C9H12 and M: Z ) 6.5 × 10-10 cm3 molecule-1 s-1. S/D is
represented in this mechanism as
4.1. Energy Distributions from Charge Transfer. The
observation of S/D values between 0.1 and 0.2 under single-
collision conditions indicates that part of the charge transfer
reaction 1.1 generates C9H12+* with very low energy. Either
the energy of this fraction is below the dissociation threshold
or can be stabilized by radiative cooling in ∼10-2 s. We
speculate that this fraction is due to production of electronically
excited O2.7 The intercepts in Figure 4 are larger than the low-
pressure limiting value, providing evidence for a second channel
where only partial energy transfer occurs. We suggest that this
involves formation of the complex (O2-C9H12)+ which upon
dissociation distributes energy statistically between O2 and
C9H12+. Finally, near-resonant charge transfer deposits the
difference between the ionization energies of O2 and C9H12 into
vibrational energy of C9H12+; at the same time, the thermal
+
[C9H12
]
γcZ[N2]
S
D
R
1 - R
)
≈
+
(3.6)
+
(1 - R)k(E)
[C7H7 ]
Linear extrapolations of S/D versus [N2] in Figure 4 to [N2] f
0 yield intercepts on the order of 0.66 for all temperatures giving
+
R ) 0.40, similar to the value of R ) 0.44 found in our O2
+
C8H10 experiments.7
The apparent intercepts in Figure 4 are significantly larger
than S/D values observed in our previous measurements1 of the
O2+ + C9H12 reaction at lower pressures; for example, measure-
ments in the selected ion flow tube and the high-temperature
flowing afterglow at 0.5 Torr of N2 at 500 K gave S/D ) 0.15.
Furthermore, experiments from a single-collision guided-ion
beam apparatus led to values of S/D ) 0.19 and 0.11 at 300
and 650 K, respectively. For C8H10+, we attributed this
“nondissociative” fraction to incomplete energy transfer in some
of the charge transfer encounters, which is quenched at lower
pressures than studied here. Several types of charge transfer
processes were postulated and were analyzed in detail in ref 7.
The present work employs an analogous interpretation. The three
processes are (1) resonant charge transfer in which the difference
+
energies of O2 and C9H12 are assumed to be nearly retained
as internal energies of O2 and C9H12+, respectively. Mainly the
resonant channel is the one affected by pressure in the present
experiments. In summary, the three channels are
+
O2+ + C9H12 f O2 + C9H12
*
ga
(4.1a)
+
O2+ + C9H12 f (O2-C9H12)+ f O2 + C9H12
*
gb
+
(4.1b)
of the ionization energies of O2 and C9H12 into C9H12 is
transferred, (2) complex formation where the energy transfer is
statistical, and (3) resonant charge transfer where the O2 neutral
is produced in an excited state.
O2+ + C9H12 f O2* (1∆ or 1Σ) + C9H12
gc (4.1c)
+
with branching fractions ga for near-resonant charge transfer,
gb for complex-forming charge transfer, and gc for near-resonant
or complex-forming charge transfer with electronic excitation
of O2, such that ga + gb + gc ) 1.
The results from Table 1 were fit to a linear expression
according to eq 3.5, S/D ≈ c1 + c2[N2], and the fit parameters
c1 and c2 are summarized in Table 1. The values for k(E)/γc )
Z(1 + c1)/c2 are given in Table 4. k(E)/γc in this representation
corresponds to averages over suitable energy distributions of
C9H12+*. A more detailed analysis is given below.
We modeled the energy distributions g(E,T) of C9H12+ arising
from channels 4.1a-c by analogy to ref 7. For convenience,
the thermal vibrational distribution was calculated with the
density functional theory (DFT) frequencies of C9H12+ derived
in ref 15. Employing the energy parameters from the same
reference, the difference of the ionization energies of O2 and
C9H12 amounts to Ech/hc ) 27 030 cm-1 whereas the dissocia-
The rate constant for the thermal dissociation of C9H12+ has
been measured at 573 and 603 K. The average values are k8 )
33 s-1 at 573 K and k8 ) 106 s-1 at 603 K and do not depend
on pressure, indicative of the high-pressure limit. Our analysis
presented below is in accord with this conclusion. The corre-
sponding Arrhenius expression is k8 ≈ 5.7 × 1011 exp(-112
kJ mol-1/RT) s-1. The factor of 2 uncertainty in the rates and
the small temperature range lead to uncertainty on the order of
70 kJ mol-1 in the activation energy. If one assumes the
activation energy to be the dissociation energy, E0 ) 167 kJ
+
+
tion energy of C9H12 into C2H5 + C7H7 (benzylium) at 0 K
is estimated to be E0/hc ) 13 950 cm-1
.
This value will be of particular importance for the analysis
of k8. Figure 5 shows a representative modeling result for g(E,
473 K) using ga ) 0.60, gb ) 0.27, and gc ) 0.13 and assuming
that the (O2-C9H12)+ complex distributes its energy statistically
mol-1 15
the two values of k8 correspond to
,
+
over the vibrations of O2 and C9H12 and the translation and
k8 ≈ 2.1 ((0.9) × 1016 exp(-167 kJ mol-1/RT) s-1 (3.7)
rotations of the relative motion.
Unfortunately, the available data on S/D at low and inter-
mediate pressure are not sufficient to fix the branching ratios
ga and gb with certainty. This prevents inclusion of a temperature
While this expression appears more reasonable, an internally
consistent derivation of k8 is given in the discussion section.

+
Stabilization and Dissociation of C9H12 Ions
J. Phys. Chem. A, Vol. 108, No. 45, 2004 9657
+
Figure 6. Specific rate constants k(E) for the dissociation C9H12
f
Figure 5. Energy distributions g(E,T) (in 1/cm-1) of excited C9H12
*
+
+
C7H7 + C2H5 (experimental points from ref 15; however, see ref
15a): long-short dashed curve, PST; dashed curve, SACM/CT
modeling from ref 20; dotted curve, RRKM fit from ref 15; full line,
empirical representation by eq 4.3. All curves (except PST) are fine-
tuned to the experimental k(E) at E/hc ) 32 260 cm-1 from ref 15; see
the text.
+
from the charge transfer O2 + n-C9H12 (T ) 473 K, representative
calculations with assumed branching ratios ga ) 0.60 (curve at higher
energies) and gb ) 0.27 (curve at lower energies), see the text).
dependence. The value of gc ≈ 0.13 follows from the single
collision results. For this reason, we cannot make an analysis
which accounts for the full energy distribution g(E,T). Instead,
we have to limit ourselves to an analysis based on eq 3.6. It
electrostatic and short-range valence character. The Massey
parameter for the C9H12 dissociation indeed signals nonadia-
+
batic dynamics20 so that an RRKM treatment is good only for
fitting purposes. There is an interesting apparent contradiction:
on one hand, kinetic energy distributions of the fragments often
are well represented by loose activated complex transition state,
or phase space, theory (PST); on the other hand, specific rate
constants k(E) are far below PST predictions. This apparent
contradiction can be resolved by realizing the nonadiabatic
character of the dynamics: (1) the dissociation bottleneck
governing k(E) is located in the range of the potential where
dynamical constraints arise from the marked anisotropy of the
potential such that k(E) is far below PST values; (2) nonadiabatic
dynamics changes the kinetic energy distribution among the
fragments beyond the bottleneck far into the range where the
potential becomes increasingly isotropic such that PST applies.
In the modeling study of ref 20, we have designed long-range/
short-range switching potentials and calculated k(E,J) using the
statistical adiabatic channel model/classical trajectory (SACM/
CT) approach. Our results for ethylbenzene cation dissociation
were found to be in fairly good agreement with experimental
results of k(E) over the range 103-108 s-1. This provides
confidence in the k(E) for the dissociation of n-propylbenzene
cations which was also modeled in ref 20. These modeling
results could well be represented by the empirical expression
+
was shown in our O2 + C8H10 study that this simplified
analysis leads to an overall uncertainty of about a factor of 2 in
∆E . The fraction 1 - R of channel 3.2 includes not only ga
from channel 4.1a but also the high-energy portion from channel
4.1b. Figure 5 illustrates the corresponding overlap of the
distributions ga and gb. Isomerization was considered as a source
of the intercepts but was dismissed because the temperature
dependence of the intercepts would be much larger for a process
with a large barrier.
4.2. Specific Rate Constants for the Dissociation of
C9H12+*. As the product Z ∆E is derived from a relative rate
measurement, its precision depends on that of the reference rate
constant, that is, the specific rate constant k(E) for dissociation
of C9H12+*, in addition to uncertainties of the energy distribution
g(E,T). Unfortunately, there exists experimental information on
k(E) only from a single study.15 These data have been fitted by
a rigid activated-complex RRKM model in ref 15, the results
of which can be represented by the expression
k(E) ) 3.6 × 108 s-1 [(E - E0)/hc 18310 cm-1]7.776 (4.2)
with E0/hc ) 13 950 cm-1. The experimental data in the range
(1-4) × 108 s-1, of course, are well reproduced by eq 4.2 since
they defined the fitted activated complex frequencies of the
RRKM representation.
k(E,J)0) )
3.6 × 108 s-1 [(E - E0(J)0))/hc 18310 cm-1]6.541 (4.3)
The k(E) data from ref 15 were measured at energies not much
above those reached in the present experiments. They are,
therefore, sufficient to evaluate our measured S/D values with
respect to energy transfer. However, without further consider-
ations they cannot be used to establish a relation to the thermal
dissociation rate constants. First, there is the lack of a proper
account for the J-dependence of the specific rate constants,
k(E,J), and in particular of the threshold energies E0(J), which
are essential for the calculation of the thermal rate constant,
k∞. Second, the RRKM fit corresponds to a vibrationally
adiabatic treatment of the dynamics. However, it was shown in
ref 20 that rigid activated complex RRKM theory does not
provide an adequate approach to k(E) for simple bond fission
processes when the Massey parameter for the transitional mode
dynamics corresponds to nonadiabatic behavior. This is par-
ticularly true for systems with transition state shifting, taking
place on a potential energy surface with both long-range
Equation 4.3 is fine-tuned to the experimental values k(E) )
3.6 ((0.4) × 108 s-1 at E/hc ) 32 260 cm-1. One finds that
k(E) is somewhat closer to PST results at lower energies and it
falls increasingly below PST at higher energies. This behavior
is illustrated in Figure 6 where k(E,J)0) from PST is compared
with the RRKM fit shown in eq 4.2, the empirical expression
of eq 4.3, the SACM/CT results,20 and the experimental points.15
+
Equation 4.3 has a similar energy dependence to that for C8H10
dissociation where an exponent of 4.824 was found. The larger
+
number of medium- to low-frequency oscillators in C9H12 is
responsible for the larger exponent in eq 4.3.
C8H10+ and C9H12+ have nearly equal dissociation energies.
Because of the larger number of oscillators, one would expect
+
that k(E), at the same E - E0, is smaller for C9H12 than for
C8H10+. However, the opposite is true: for the range of interest

9658 J. Phys. Chem. A, Vol. 108, No. 45, 2004
Fernandez et al.
in the present work, k(E) apparently is about a factor of 2 larger
+
for C9H12 than the corresponding value for C8H10+. Again,
the reason is that there are more medium- to low-frequency
oscillators in C9H12+. As a consequence, higher pressures for
+
collisional stabilization are required in C9H12+ than in C8H10
.
This is illustrated by comparing Figure 4 of the present work
with Figure 1 of ref 7.
4.3. Collisional Energy Parameters for Excited C9H12+*.
An evaluation of S/D versus [M] was made in two stages as
we did previously.7 First it was assumed that the nonresonant
charge transfer channels 4.1b and 4.1c only contribute to the
apparent intercepts, and second, the influence of the energy
distribution of channel 4.1b was taken into account. The former,
a simplified analysis, was preferred in the present work.
Nevertheless, we also show the results from a representative
simulation of the more precise analysis. However, we are unable
to fix the required parameters ga and gb with the same certainty
that was possible for C8H10+. The analysis7 from eq 3.6 leads
to the expression
Figure 7. Modeling of the dependence of S/D on [N2] (energy
distributions g(E,T) from section 4.1 with representative assumed values
ga ) 0.60, gb ) 0.27, and gc ) 0.13; experimental points as in
Figure 4).
using eqs 4.4-4.6 and accept an uncertainty of the derived
values of ∆E of about a factor of 2.
The mean value of ∆E from Table 4 is
Z[M](- ∆E )s*
(1 - R)k(Ech)(Ech - E0)
S
D
R
1 - R
≈
+
I(T)
(4.4)
with
∆E /hc ) -200 ((100) cm-1
(4.7)
I(T) )
at an average excitation energy of E/hc ≈ (30 000-33 000)
cm-1. This value of ∆E is in good agreement with the
corresponding result for stabilization of C8H10+* in collisions
with N2 which, at a similar excitation energy, gave ∆E /hc )
-285 ((150) cm-1. Although there remains an uncertainty of
about a factor of 2 in ∆E , we find again that ∆E values for
excited ions are similar to the corresponding ∆E values for
excited neutral molecules.21-25 In contrast, the collision frequen-
cies Z for ions and neutrals are markedly different.
s*
Ech - E0
Ech - E0 + Eth
Eth + Ech F(Eth)
Eth
kT
∞
exp -
dEth
∫
(
) (
)
(
)
0
Ech
Qvib
(4.5)
In this expression, we assume k(E) to be of the form k(E) (E
- E0)s*-1 used in eq 4.3, F(Eth) is the vibrational density of
+
states, and Qvib is the vibrational partition function of C9H12
.
4.4. Relation between Specific Rate Constants k(E) and
We also assume that ∆E is proportional to E and does not
depend on T. The collision efficiency γc according to ref 10
can be related to ∆E by the expression
+
Thermal Rate Constants k(T) for C9H12 Dissociation.
Determining both specific rate constants k(E) and thermal rate
constants k8 for the dissociation C9H12+ f C7H7+(Bz) + C2H5
offers the rare possibility to check the internal consistency of
the derived data. This analysis requires the full details of
unimolecular rate theory. In particular, the contributions of
angular momentum (quantum number J) in k(E,J) need to be
accounted for, both in energy-specific and in thermal measure-
ments. This analysis has been done in our parallel SACM/CT
study20 where k(E,J) was derived from trajectory calculations
employing a model potential which was fine-tuned to reproduce
the experimental measurements of k(E).15 Based on these
calculations of k(E,J), thermally averaged rate constants
krec,∞(T) for the recombination C7H7+ + C2H5 f C9H12+ were
derived. An important result from this analysis was the
demonstration that the marked difference of the calculated k(E,J)
from the results given by PST also occurs in the thermal rate
constants: k(E,J) falls increasingly below the PST results with
increasing E, and krec,∞(T) analogously falls below the phase
space theory results with increasing T. In other words, krec,∞(T)
is expected to be much smaller than the Langevin rate constant
for capture of C2H5 by C7H7+. On the basis of the experimental
k(E) values, the recombination rate was predicted20 to be
γc
1 - γc2
- ∆E s*
E - E0
≈
(4.6)
If γc , 1, as found here, the left-hand side approaches γc. Table
4 includes the corresponding values for γc (0 K) ≈ - ∆E s*/
(Ech - E0) and I(T). The values for ∆E , such as derived from
the experimental values for k(E)/γc, are also given in this table.
The detailed analysis, accounting for near-resonant (eq 4.1a)
and nonresonant (eqs 4.1b,c) charge transfer, is more involved.
It has been described in ref 7 and will not be explained here
again. We only illustrate the dependence of S/D on [M] by
employing g(E,T) from section 4.1 with representative branching
ratios of ga ≈ 0.60, gb ≈ 0.27, and gc ≈ 0.13. Figure 7 compares
this modeling with the experimental S/D versus [N2].
One realizes that there are some deviations of the experi-
mental results from the modeled curves. We were unable to
trace the origin of these deviations, in particular for the
measurements at 423 K which are not included in Figure 7. In
addition, the modeled S/D versus [N2] plots are more curved at
low [N2] than the experimental data. The latter may indicate
inadequate branching ratios ga and gb or problems with the
statistical modeling of the charge transfer channel 4.1b. We
varied ga and gb within this model but could not obtain much
better agreement with the experiments. (Increasing ∆E or
decreasing ga would have produced steeper curves of S/D versus
[N2].) Therefore, we limited ourselves to the simplified analysis
krec,∞(T) ≈ 8.0 × 10-12(T/600 K)-1.16 cm3 molecule-1 s-1
(4.8)
over the temperature range 400-800 K. In contrast, the
Langevin rate constant is given by kL ) 1.0 × 10-9 cm3
molecule-1 s-1
.

+
Stabilization and Dissociation of C9H12 Ions
J. Phys. Chem. A, Vol. 108, No. 45, 2004 9659
To compare the predicted krec,∞(T) with the measured k8 from
section 3, the equilibrium constant was calculated as
are also of similar magnitude as found for the corresponding
excited neutral molecules.
The conversion of specific rate constants k(E) into high-
pressure thermal dissociation rate constants kdiss,∞(T) was per-
formed by unimolecular rate theory in the form of SACM/CT
calculations. As kdiss,∞(T) has a particularly large temperature
dependence, these quantities also allowed us to confirm the
+
Keq ) ([C7H7 ][C2H5]/[C9H12+])eq ) kdiss/krec (4.9)
and kdiss,∞(T) was obtained by combining Keq with krec,∞. It then
has to be decided whether the measured k8 is less than the high-
pressure limiting dissociation rate constant kdiss,∞ because of
falloff effects. For internal consistency, we calculated Keq with
the DFT frequencies from ref 15, keeping in mind that the
largest uncertainty stems from the bond energy E0 which, as
before, was taken to be E0/hc ) 13 950 cm-1. In this way,
between 500 and 600 K, we obtained
+
dissociation energy of C9H12 as 166.9 ((2.2) kJ mol-1 (at 0
K). The given analysis provided an intrinsically consistent
description of rate parameters such that it may serve as an
example for subsequent studies of other chemical/thermal
activation reactions of excited molecular ions.
Acknowledgment. This project was funded by the United
States Air Force Office of Scientific Research under Project
2303EP4 and Grant Award FA8655-03-1-3034. Financial sup-
port by the Deutsche Forschungsgemeinschaft (SFB 357 “Mole-
kulare Mechanismen unimolekularer Reaktionen”) is also
gratefully acknowledged.
Keq
)
1.67 × 1027(T/600 K)-0.33 exp(-E0/kT) molecule cm-3
(4.10)
References and Notes
To assess falloff effects, we calculated limiting low-pressure
rate constants kdiss,0 following the conventional method described
in ref 26. The relevant quantities entering kdiss,0 are Z ) 6.5 ×
10-10 cm3 molecule-1 s-1, âc ≈ 0.2, the vibrational density of
states Fvib(E0) ) 5.4 × 1014/cm-1, the vibrational partition
(1) Williams, S.; Midey, A. J.; Arnold, S. T.; Morris, R. A.; Viggiano,
A. A.; Chiu, Y.-H.; Levandier, D. J.; Dressler, R. A.; Berman, M. R. J.
Phys. Chem. A 2000, 104, 10336.
(2) Arnold, S. T.; Williams, S.; Dotan, I.; Midey, A. J.; Morris, R. A.;
Viggiano, A. A. J. Phys. Chem. A 1999, 103, 8421.
(3) Arnold, S. T.; Dotan, I.; Williams, S.; Viggiano, A. A.; Morris, R.
A. J. Phys. Chem. A 2000, 104, 928.
(4) Viggiano, A. A.; Miller, T. M.; William, S.; Arnold, S. T.; Seeley,
J. V.; Friedman, J. F. J. Phys. Chem. A 2002, 106, 11917.
(5) Midey, A. J.; Williams, S.; Arnold, S. T.; Viggiano, A. A. J. Phys.
Chem. A 2000, 106, 11726.
(6) Fridgen, T. D.; McMahon, T. B.; Troe, J.; Viggiano, A. A.; Midey,
A. J.; Williams, S. J. Phys. Chem. A 2004, 108, 5600.
(7) Troe, J.; Viggiano, A. A.; Williams, S. J. Phys. Chem. A 2004,
108, 1574.
function Qvib(603 K) ) 8.2 × 105, FE(600 K) ≈ 1.6, and Frot
≈
F
rot,max(603 K) ≈ 5.3. Combining kdiss,∞ ) krec,∞Keq from eqs
4.8-4.10 with the derived kdiss,0 indicates that the center of the
falloff curve kdiss,0([M]) ) kdiss,∞ at 600 K is located near
[N2]center ≈ 4 × 1013 molecule cm-3. Falloff curves connecting
kdiss,0 with kdiss,∞ have been illustrated in refs 26 and 27 for
systems of comparable size. On this basis, one can safely assume
that the measured k8 corresponds to the high-pressure rate
constant kdiss,∞ within a few percent. The absence of an
observable pressure dependence in Table 2 confirms this
conclusion.
(8) Kim, Y. H.; Choe, J. C.; Kim, M. J. Phys. Chem. A 2001, 105,
5751.
(9) Malow, M.; Penno, M.; Weitzel, K.-M. J. Phys. Chem. A 2003,
107, 10625.
(10) Troe, J. J. Phys. Chem. 1983, 87, 1800.
Combining the modeled krec,∞ with Keq from eqs 4.8-4.10
predicts
(11) Miasek, P. G.; Harrison, A. G. J. Am. Chem. Soc. 1975, 97, 714.
(12) Ahmed, M. S.; Dunbar, R. C. J. Am. Chem. Soc. 1987, 109, 3215.
(13) Barfknecht, A. T.; Brauman, J. I. J. Chem. Phys. 1986, 84, 3870.
Boering, K. A.; Brauman, J. I. J. Chem. Phys. 1992, 97, 5439.
(14) Fernandez, A.; Viggiano, A. A.; Miller, T. M.; Williams, S.; Troe,
J. J. Phys. Chem., to be submitted.
(15) Hwang, W. G.; Moon, J. H.; Choe, J. C.; Kim, M. S. J. Phys. Chem.
A 1998, 102, 7512. (a) One should note that the point with k(E) ) 3.4 ×
104 s-1 may not be a true experimental point; its energy E/hc ) 18 390
cm-1 was “estimated from the rate-energy curve”, i.e., presumably from
the RRKM calculation in ref 15. However, our own RRKM calculation of
k(E) using the same activated complex frequencies gives values of k(E) at
E/hc ) 18 390 cm-1 which are markedly lower than 3.4 × 104 s-1; see
Figure 6.
(16) Arnold, S. T.; Seeley, J. V.; Williamson, J. S.; Mundis, P. L.;
Viggiano, A. A. J. Phys. Chem. A 2000, 104, 5511.
(17) NIST Chemistry WebBook, http://webbook.nist.gov/chemistry.
(18) Rusyniak, M.; Ibrahim, Y.; Alsharaeh, E.; Meot-Ner (Mautner),
M.; El-Shall, M. S. J. Phys. Chem. A 2003, 107, 7656.
(19) Hirschfelder, J. O.; Curtiss, C. F.; Bird, R. B. Molecular Theory of
Gases and Liquids; J. Wiley: New York, 1964. Miller, Th. M. In Handbook
of Chemistry and Physics; Lide, D. R., Ed.; CRC Press: Boca Raton, FL,
2003; Section 10, p 163.
k8 ) 1.34 × 1016(T/600 K)1.49 exp(-E0/kT) s-1 (4.11)
The prediction is remarkably close to our direct measurement
of k8 given by eq 3.7 with E0 ) 166.9 kJ mol-1. The difference
between krec,∞ and kLangevin in the modeling has been calibrated
by the difference between the measured k(E) and k(E) from PST
such that realistic values of k8 should have been produced. The
good agreement between the measured k8 from eq 3.7 and the
predicted k8 from eq 4.11 confirms the validity and internal
consistency of the described approach. The agreement between
the measured and the modeled value also supports the E0 value
of E0/hc ) 13 950 cm-1 (or E0 ) 166.9 kJ mol-1) used in the
analysis. The difference of a factor of 1.6 in k8 under our condi-
tions would correspond to a difference of about 190 cm-1 in
+
+
E0/hc. The energetics of the dissociation C9H12 f C7H7
C2H5, therefore, should be accurate to better than about 2.2 kJ
mol-1
+
.
(20) Troe, J.; Ushakov, V. G.; Viggiano, A. A.; Williams, S. J. Chem.
Phys., to be submitted.
(21) Hold, U.; Lenzer, Th.; Luther, K.; Reihs, K.; Symonds, A. C. J.
Chem. Phys. 2000, 112, 4090.
(22) Lenzer, Th.; Luther, K.; Reihs, K.; Symonds, A. C. J. Chem. Phys.
2000, 112, 4090.
(23) Grigoleit, U.; Lenzer, Th.; Luther, K.; Mu¨tzel, M.; Takahara, A.
Phys. Chem. Chem. Phys. 2001, 3, 2191.
(24) Troe, J.; Wieters, W. J. Chem. Phys. 1979, 71, 3931.
(25) Hippler, H.; Luther, K.; Troe, J.; Wendelken, H. J. J. Chem. Phys.
1983, 79, 239.
5. Conclusions
This study described charge transfer generating excited
+
+
C9H12 ions which then dissociate mainly to C7H7 ions and
C2H5 radicals. Calibrating S/D ratios for chemical activation
using literature values of the specific rate constants for dis-
sociation, absolute values for collisional stabilization rates were
obtained. From these quantities, average energies transferred
per collision ∆E /hc ) -200 ((100) cm-1 were derived for a
(26) Troe, J. J. Phys. Chem. 1979, 83, 114.
bath gas N2 at excitation energies near E/hc ) 32 000 cm-1
.
(27) Troe, J.; Ushakov, V. G. Faraday Discuss. 2001, 119, 145.
(28) Hamon, S.; Speck, T.; Mitchell, J. B. A.; Rowe, B. R.; Troe, J. J.
Chem. Phys. 2002, 117, 2557.
+
These values are similar to those measured for excited C8H10
ions, where ∆E /hc ) -285 ((150) cm-1 was obtained. They