Photodissociation Dynamics of Nitrobenzene Molecular Ion on a Nanosecond Time Scale

Article abstract of DOI:10.1021/jp9537872

Photodissociation dynamics of nitrobenzene molecular ion has been investigated on a nanosecond time scale by mass-analyzed ion kinetic energy spectrometry.Photodissociation has been found to occur through various reaction channels.Three dissociation channels to C6H5O(1+), C6H5(1+), and NO(1+) have been observed together with a consecutive dissociation to C5H5(1+) via C6H5O(1+).The photodissociation rate constant of the molecular ion and the kinetic energy release distributions in each channel have been determined.The rate constant of the second step of the consecutive reaction has been determined in real time also.It has been found that the direct bond cleavage to C6H5(1+) occurs in competition with the rearrangement processes to C6H5O(1+) and NO(1+) on a nanosecond time scale.Statistical theories have been used to gain insight into the dynamics of the reactions.

Full text of DOI:10.1021/jp9537872

J. Phys. Chem. 1996, 100, 9227-9234
9227
Photodissociation Dynamics of Nitrobenzene Molecular Ion on a Nanosecond Time Scale
Wan Goo Hwang,† Myung Soo Kim,*,† and Joong Chul Choe*,‡
Department of Chemistry and Center for Molecular Catalysis, Seoul National UniVersity, Seoul 151-742,
Korea, and Department of Chemistry, UniVersity of Suwon, Kyunggi 445-743, Korea
ReceiVed: December 20, 1995; In Final Form: March 6, 1996X
Photodissociation dynamics of nitrobenzene molecular ion has been investigated on a nanosecond time scale
by mass-analyzed ion kinetic energy spectrometry. Photodissociation has been found to occur through various
reaction channels. Three dissociation channels to C6H5O+, C6H5 , and NO+ have been observed together
+
with a consecutive dissociation to C5H5+ via C6H5O+. The photodissociation rate constant of the molecular
ion and the kinetic energy release distributions in each channel have been determined. The rate constant of
the second step of the consecutive reaction has been determined in real time also. It has been found that the
direct bond cleavage to C6H5+ occurs in competition with the rearrangement processes to C6H5O+ and NO+
on a nanosecond time scale. Statistical theories have been used to gain insight into the dynamics of the
reactions.
I. Introduction
Baer.1 In the continuous-wave IRMPD study of nitrobenzene
ion by Moini and Eyler,9 C6H5 product was not observed at
+
Nitrobenzene has been among the most attractive systems
studied in the fields of both ionic1-18 and neutral19-24 dissocia-
tion dynamics. A number of possible dissociation pathways of
the nitrobenzene or its molecular ion have been the main
motivation for such research efforts. To understand the dynam-
ics and mechanism of the nitrobenzene ion dissociation, various
experimental techniques have been employed such as photo-
electron-photoion-coincidence (PEPICO) spectrometry,1-3 laser
photodissociation,3-7 infrared multiphoton dissociation (IRM-
PD),8,9 and conventional mass spectrometry.10-18 Also, the
experimental rate constant and kinetic energy release data have
been compared with theoretical predictions based on statistical
theories such as Rice-Ramsperger-Kassel-Marcus (RRKM)
theory,25 which is equivalent to quasi-equilibrium theory (QET
or RRKM-QET)26 in ionic field and phase space theory.27
Despite such efforts, its dissociation dynamics and mechanism
have not been fully understood yet, and some controversy still
remains.
all even though its appearance energy was comparable to those
for other product ions. This was taken as evidence for the
excited state hypothesis mentioned above. In a recent study
by Osterheld et al.,8 however, production of C6H5+ was observed
in pulsed IRMPD, indicating that it occurs in the ground
electronic state. To accommodate this result with the previous
rate measurement with PEPICO by Panczel and Baer, Osterheld
et al. proposed an ion-dipole intermediate (IDI) mechanism.
In this mechanism, nitrobenzene ion first forms IDI, which either
+
dissociates to C6H5 directly or reassociates to an isomeric
structure (nitrite form) followed by dissociation to C6H5O+ or
NO+.
Another interesting feature in the nitrobenzene ion dissocia-
+
tion is the production of C5H5+. It is well known that C5H5
is produced by a consecutive reaction via C6H5O+. In particular,
in the previous visible laser photodissociation studies carried
out with a double-focusing mass spectrometer,4-6 PEPICO,3 and
an ion cyclotron resonance mass spectrometer,7 the major
The major fragment ions in the dissociation of nitrobenzene
fragment ions observed were C6H5 and C5H5+. Absence of
+
ions are C6H5O+, C6H5+, and NO+. In addition, consecutive
+
C6H5O+ was explained by its further dissociation to C5H5 at
+
dissociation via C6H5O+ results in abundant C5H5
. Rate
a nitrobenzene ion internal energy higher than 2.5 eV.3,7
constants for the main channels, namely, the productions of
C6H5O+, C6H5+, and NO+, in the dissociation of the energy-
selected nitrobenzene ions with an internal energy of 1.3-1.9
eV were measured with PEPICO by two research groups.1,2 In
the earlier study by Panczel and Baer,1 it was reported that the
In the present study, the dissociation dynamics of nitroben-
zene ions at the internal energy 2.48-2.98 eV has been
investigated with the recently developed photodissociation
method using mass-analyzed ion kinetic energy spectrometry
(PD-MIKES).28-33 Nanosecond rate data can be obtained with
the present technique. We observed that the production of
C6H5+ is competitive with the dissociation channels leading to
NO+ and C6H5O+. In this paper, the dynamics in the individual
dissociation channels is to be discussed on the basis of the
measured and theoretically calculated rate constants and kinetic
energy release distributions.
+
rate for the appearance of C6H5 was much faster than those
for C6H5O+ and NO+, indicating that these two groups of
channels are not in direct competition. This result combined
with the earlier report4-6 that C6H5+, but not C6H5O+ and NO+,
was observed in the photodissociation of nitrobenzene ion at
514.5 nm led the above investigators to suggest that the
+
production of C6H5 occurs from an excited electronic state.
On the other hand, in the later study by Nishimura et al.,2 all
the above three channels were found to occur competitively and
could be explained by RRKM-QET.
II. Experimental Section
Production of C6H5+ from nitrobenzene ion has been the focus
of some research efforts since the first report by Panczel and
The experimental setup has been described in detail else-
where28 and will be reviewed only briefly here. A double-
focusing mass spectrometer with reversed geometry (VG
Analytical Model ZAB-E) modified for photodissociation study
was used. Nitrobenzene ions generated by charge exchange in
the ion source and accelerated to 8 keV were mass-analyzed
* Authors to whom correspondence should be addressed.
† Seoul National University.
‡ University of Suwon.
X Abstract published in AdVance ACS Abstracts, May 1, 1996.
S0022-3654(95)03787-7 CCC: $12.00 © 1996 American Chemical Society

9228 J. Phys. Chem., Vol. 100, No. 22, 1996
Hwang et al.
TABLE 1: Relative Abundances (%) of Fragment Ions in
the Photodissociation of the Nitrobenzene Ion
fragment ion (m/z)
wavelength (nm) C6H5O+ (93) C6H5+ (77) C5H5+ (65) NO+(30)
photodissociation
607.5
514.5
488.0
11 ( 3
0
0
68 ( 5
78 ( 3
80 ( 3
19 ( 4
22 ( 3
20 ( 3
2 ( 1
0
0
region appear at different translational energies in the PD-MIKE
spectrum. Therefore, the fact that the peaks at m/z 30, 77, and
93 are asymmetrically broadened in Figure 1a (enlargements
are shown in Figure 2) indicates that the photodissociation rate
constant of the molecular ion can be determined for the
experiment with 607.5 nm excitation. At higher photon
+
energies, however, the C6H5 peak is symmetric, as shown in
Figure 1b,c. This means that the dissociation rate of the
molecular ion after the absorption of a 514.5 or 488.0 nm photon
is too fast to be measured with the present method. The
photofragment ion with m/z 65 appears as two peaks designated
A and B in Figure 1, which will be assigned later. Its spectral
pattern changes dramatically with wavelength as can be seen
more clearly in Figure 3. Relative abundance of each photof-
ragment signal was determined after various calibrations to
account for the instrumental effects as previously.31 The results
are listed in Table 1.
Figure 1. PD-MIKE spectra for dissociation of nitrobenzene molecular
ion at the photoexcitation wavelengths of (a) 607.5, (b) 514.5, and (c)
488.0 nm. The molecular ions with 8 keV translational energy entered
the electrode assembly floated at 1.85 kV. Chemical formula and m/z
of the photofragment ion corresponding to each peak are indicated.
B. Rate Constants. The photodissociation rate constants
have been determined by analyzing PD-MIKE peak shapes as
described in detail previously.28-30 Instead of a single value
of rate constant, its distribution was determined, which is mainly
due to internal energy distribution of the molecular ions. The
average value will be taken as the rate constant representative
of the distribution. In our previous photodissociation
studies,30,32-33 not only the thermal energy of the parent
molecules but also the collisional relaxation of the molecular
ions occurring in the ion source were found to affect the internal
energy and its distribution of the molecular ion. An experi-
mental technique to correct for the latter in the rate data was
developed.30 Such a correction was not needed here because
the effect of the collisional relaxation on the rate data was not
significant. This is thought to be due to the fact that the internal
energy of nitrobenzene ions generated by charge exchange with
by the magnetic sector. Then, the ion beam was crossed with
the laser beam in the field region of an electrode assembly
located near the intermediate focal point of the instrument. The
607.5 nm output of a dye laser (Spectra Physics Model 375B)
and the 514.5 and 488.0 nm lines of an argon ion laser (Spectra
Physics Model 164-09) were used. The translational kinetic
energy of the fragment ions was analyzed by the electric sector.
This is called mass-analyzed ion kinetic energy spectrometry
(MIKES). Since a MIKE spectrum contains contributions from
metastable ion decomposition (MID) and collision-induced
dissociation (CID) by residual gas, phase-sensitive detection was
adopted to record a MIKE spectrum originating from photo-
dissociation, namely, a PD-MIKE spectrum. To improve the
quality of a PD-MIKE spectrum, signal averaging was carried
out for repetitive scans. Errors quoted in this work were
estimated from several duplicate experiments at 95% confidence
limits.
CS2•+ is small (0.21 eV ignoring the thermal energy, Vide infra).
Figure 2 shows the PD-MIKE peak shapes for NO+, C6H5
,
+
and C6H5O+ at 607.5 nm excitation which have been analyzed
to obtain rate constants. The peak shapes recalculated with the
rate constant distributions thus obtained are also shown. The
average photodissociation rate constants obtained from the three
peaks are listed in Table 2. It is to be noted that the average
total rate constants determined from C6H5+ and NO+ peaks are
essentially the same. This means that these channels are in
direct competition. Since the production of C6H5O+ and NO+
are known to be in direct competition, one expects that the rate
constant determined from C6H5O+ would be the same as the
other two. However, the former rate constant tends to be a little
smaller than the latter two, even though still within experimental
error. This, we believe, is due to consecutive dissociation of
CS2 was used as a reagent gas for charge exchange ionization.
The ion source chamber was maintained at 140 °C.
III. Results
A. General Appearance of the PD-MIKE Spectrum. To
measure the photodissociation rate constant, a high voltage was
applied on the electrode assembly and the parent ions were
photoexcited in the field region of the assembly. The PD-MIKE
spectrum for the nitrobenzene ion obtained at 1.85 kV applied
voltage using a 607.5 nm photon is shown in Figure 1a. Four
photofragment ions are identified and assigned at m/z (mass-
to-charge ratio) 30, 65, 77, and 93 in Figure 1a, which
correspond to NO+, C5H5+, C6H5+, and C6H5O+, respectively.
Considerable noise between the peaks is due to strong back-
grounds originating from MID and CID by residual gas. At
photon wavelengths of 514.5 and 488.0 nm, C6H5O+ and NO+
peaks were not observed in the PD-MIKE spectra as shown in
Figure 1b,c. The results at 514.5 and 488.0 nm are consistent
with the previous laser photodissociation results,3-7 as men-
tioned in a previous section. Photofragment ions with a
particular m/z value formed at different positions in the field
C6H5NO2 to C5H5 via C6H5O+. In the second step of this
consecutive reaction, C6H5O+ ions with higher internal energy
will dissociate more readily. Then, the signal detected as the
PD-MIKE peak of C6H5O+ is due to those ions with lifetimes
longer than the flight time to the detector (∼8 µs). Among the
molecular ions with some internal energy distribution, those ions
with lower internal energy, on the average, will contribute to
this signal. Then, it is reasonable to observe that the average
total rate constant determined from C6H5O+ channel is smaller
than those from NO+ and C6H5+. To summarize, the average
•+
+

Photodissociation of Nitrobenzene Molecular Ion
J. Phys. Chem., Vol. 100, No. 22, 1996 9229
+
Figure 3. PD-MIKE spectra for production of C5H5 ion from
nitrobenzene ion obtained with 1.85 kV applied on the electrode
assembly. (a) 607.5, (b) 514.5, and (c) 488.0 nm excitations. Experi-
mental and calculated results are shown as filled circles and solid curves,
respectively. The arrow denotes the translational energy of C5H5+ if it
were produced directly from the molecular ion outside the field region
of the electrode assembly.
Figure 2. PD-MIKE profiles for (a) NO+, (b) C6H5+, and (c) C6H5O+
at 607.5 nm excitation. Experimental conditions are the same as in
Figure 1. Experimental and calculated results are shown as filled circles
and solid curves, respectively.
TABLE 2: Dissociation Rate Constants Derived from
PD-MIKE Peaks
assembly and peak B is that occurring outside the field region.
If C5H5 were produced directly from the molecular ion, the
+
rate constant (108 s-1
)
photodissociation
peak corresponding to dissociation outside the field region would
appear at 4233 eV, as marked with an arrow in Figure 3. Figure
3 shows that the relative abundance of peak A to peak B
increases with decreasing wavelength or increasing photon
energy. This means that the second step of the consecutive
reaction occurs faster as the internal energy increases, as will
be shown later.
wavelength (nm) C6H5O+ a
C6H5
NO+ a
C5H5
+ a
+ b
607.5
514.5
488.0
5.7 ( 2.9 8.6 ( 0.4 7.9 ( 1.5 0.52 ( 0.21
c
c
2.6 ( 1.0
3.5 ( 1.1
a Total dissociation rate constant (kT). b Second step rate constant
(k65) of the consecutive reaction 123•+ f 93+ f 65+. c Higher than
109 s-1
.
Full analysis of a PD-MIKE peak shape arising from a
consecutive reaction is rather complicated because two rate
constants and their distributions are involved. The rate constant
for the first step, which is kT, is not available for photodisso-
ciation at 514.5 and 488.0 nm. Even for photodissociation at
607.5 nm, it is likely that the effective rate constant for the
total rate constants determined from the three channels show
that all these channels are in direct competition, on a nanosecond
time scale.
The total dissociation rate constant, kT, is the sum of the
individual rate constants for competitive channels. The indi-
vidual rate constants for the above channels, k30, k77, and k93
can be evaluated from kT and branching ratios. Here, kT has
been estimated by averaging the total rate constants determined
+
first step is larger than kT determined from NO+ and C6H5
channels for the reason explained above. Fortunately, however,
the consecutive reaction can be treated easily by considering
the second step only since the second step rate constant is much
smaller than kT. Then, a simple first-order kinetics applies and
the rate constant or its distribution can be determined just as
for a direct reaction.25b The PD-MIKE profiles recalculated
from the rate constant distributions thus obtained are shown as
solid curves in Figure 3. A successful analysis as is evident in
from NO+ and C6H5 channels. In estimating the relative
+
abundances and hence the branching ratios, the intensity of
C5H5 has been added to that of C6H5O+ because the former
+
+
ion originates from the latter (consecutive production of C5H5
via C6H5O+ is evident, as will be shown below). The resultant
values for k93, k77, and k30 are (2.4 ( 0.6) × 108, (5.7 ( 1.1) ×
108, and (1.7 ( 0.3) × 107 s-1, respectively, for the experiment
with 607.5 nm excitation. For the experiments with 514.5 and
488.0 nm excitations, the fact that the C6H5+ peak is symmetric
means that kT is larger than the maximum (∼1 × 109 s-1) that
can be determined with the present method.
+
Figure 3 further supports that all the C5H5 ions are produced
consecutively via C6H5O+. The average values of the second
step rate constant, k65, at three photon wavelengths are listed in
Table 2.
C. Kinetic Energy Release Distributions. In our previous
investigations on the photodissociation dynamics of molecular
ion beam, a kinetic energy release distribution (KERD) was
evaluated from a PD-MIKE profile recorded without applied
voltage on the electrode assembly.28,30,32-33 This is to avoid
+
The PD-MIKE peak shape for C5H5 is different from the
others, as shown in Figure 1 and also in Figure 3. The methods
to identify a consecutive channel and analyze the PD-MIKE
peak to evaluate the rate constant have been described in detail
in our previous study on n-heptane molecular ion.29 Following
the same method, it can be readily shown that peak A in Figure
3 is due to C5H5+ generated by the consecutive reaction 123•+
f 93+ f 65+ occurring in the field region of the electrode
the peak tailing arising from kinetic reasons. The same method
+
has been used here to obtain KERD in the production of C6H5
.
KERD thus obtained at 607.5 nm excitation is shown in Figure
4b. The average kinetic energy release evaluated from the

9230 J. Phys. Chem., Vol. 100, No. 22, 1996
Hwang et al.
TABLE 3: Thermochemical Data
species
(structure)
∆Hf°0Ka (kJ/mol)
•+
C6H5NO2
C6H5ONO•+
C6H5O+
1045b
(phenylnitrite ion)
(phenoxy ion)
(phenyl ion)
900c
888b
1143.8b
1064d
987.7d
66d
+
C6H5
+
C5H5
(cyclopentadien-1,3-yl-5 ion)
NO+
C6H5O•
(phenoxy)
•
NO2
36.0b
NO•
CO
90.78b
-113.80b
a Heat of formation at 0 K. b Reference 34. c AM1 calculation in ref
1. d Values converted from 298 K data in ref 34.
The results from the photodissociation experiment described
in the previous section show that the following dissociation
pathways are acceptable.
k77
+
•
C6H5 (m/z 77) + NO2
(2a)
k30
•+
C6H5NO2
•
NO+ (m/z 30) + C6H5O
(2b)
(2c)
(m/z 123)
k93
•
C6H5O+ (m/z 93) + NO
Figure 4. Kinetic energy release distributions in the photodissociation
of nitrobenzene ion to (a) NO+, (b) C6H5+, and (c) C6H5O+ with 607.5
nm excitation. Experimental results are shown as circles. Theoretical
(PST) results are shown as solid curves. Bars represent error limits.
k65
C6H5+ (m/z 65) + CO
(2d)
Dissociation dynamics for the individual channels is discussed
below.
A. Production of C6H5+. To gain insight on the dynamics
of this channel, the experimental rate constant and KERD have
been compared with predictions from the statistical theories.
KERD for this reaction has been calculated by phase space
theory formalism.27
distribution increases with the photon energy. These are 0.089,
0.133, and 0.134 eV, respectively, in the photodissociation at
607.5, 514.5, and 488.0 nm. KERD in the production of
C6H5O+ at 607.5 nm has been determined similarly. The result
is shown in Figure 4c. The average kinetic energy release
evaluated from the distribution is 0.66 eV. In the case of NO+,
which appears as a very weak signal in the spectrum, determi-
nation of the KERD from the PD-MIKE profile recorded without
the field was rather difficult due to the presence of metastable
noise. In the PD-MIKE spectrum recorded with the field, the
peak does not suffer from the above noise, as shown in Figure
3. Since the peak tailing is not severe, the higher energy half
of the peak has been used to determine KERD. The result is
shown in Figure 4a. The average kinetic energy release
evaluated from the distribution is 0.29 eV.
0
n(T;J,E)
E-E -T F′(E - E0 - T - R)P′(T,J,R) dR (3)
∫
R*
Here, n(T;J,E) is the distribution, namely, KERD, of the kinetic
energy release (T) at a given angular momentum J and internal
energy E. The root-mean-square average J value evaluated at
the ion source temperature has been used.27g F′ and P′ are the
product vibrational and angular momentum state densities,
respectively. R is the product rotational energy, and R* is its
minimum. E0 is the reaction critical energy. Assuming that
the reverse barrier for this channel is negligible as suggested in
the PEPICO study by Nishimura et al.,2 E0 has been taken as
the endoergicity at 0 K of 1.37 eV estimated from the
thermochemical data listed in Table 3. The molecular param-
eters used in the calculation of KERD are listed in Table 4.
KERD calculated with eq 3 is shown in Figure 4b, which is in
excellent agreement with the experimental result. This indicates
that the production of C6H5+ occurs statistically on the ground
electronic state potential energy surface without significant
reverse barrier. The statistical nature of the kinetic energy
release in this channel was also reported by Nishimura et al.2
and by Bunn et al.,3 who compared average kinetic energy
release determined by PEPICO and PEPICO-PD, respectively,
with the phase space theory calculation.
IV. Discussion
The internal energy of the molecular ion generated by charge
exchange ionization and then photoexcited can be estimated by
the following relation.
E ) RE - IE + Eth + hν
(1)
Here, RE is the recombination energy of CS2 and IE is the
ionization energy for nitrobenzene. Their best literature values34
are 10.07 and 9.86 eV, respectively. Hence, charge exchange
alone results in the production of nitrobenzene molecular ion
with 0.21 eV internal energy. Eth is the thermal internal energy
of nitrobenzene ion at 140 °C, which can be estimated from
the molecular parameters. The average vibrational internal
energy was 0.23 eV at 140 °C. Thus, the average total internal
energies for the photoexcited molecular ions are 2.48, 2.85, and
2.98 eV with 607.5, 514.5, and 488.0 nm excitations, respec-
tively.
The rate constant for this reaction has been calculated by
RRKM-QET formalism.25
W*(E - E0)
k(E) ) σ
(4)
hF(E)
Here, F is the density of states of the reactant ion, W* is the
state sum at the transition state, and σ is the reaction path

Photodissociation of Nitrobenzene Molecular Ion
J. Phys. Chem., Vol. 100, No. 22, 1996 9231
TABLE 4: Molecular Parameters Used in the Calculations
of KERDs
TABLE 5: Vibrational Frequencies Used in the Calculations
of Rate Constants (cm-1) for the Production of C6H5+, NO+,
and C6H5O+
vibrational frequenciesa (cm-1
)
•+ b
C6H5NO2
3080(4), 3050, 1610, 1590, 1520, 1480, 1460,
1350, 1320, 1310, 1180, 1170, 1160, 1120,
1110, 1000(2), 970, 940, 850, 840, 790,
700, 680, 670, 610, 530, 420, 400(2), 260,
180, 140
+ b
C6H5
3080, 3060(2), 3050(2), 1600(2), 1490(2), 1330, 1310,
1180, 1150, 1040(2), 1010, 1000, 990, 970(2), 850,
700, 670, 610(2), 410(2)
• b
NO2
1620, 1320, 750
C6H5Oc,d
3070, 3060(2), 3050(2), 1720, 1600(2), 1490(2), 1330,
1310, 1180(2), 1150, 1040(2), 1010, 990(2), 970(2)
850(2), 670, 610(2), 560, 410(2)
C6H5NO2•+ (TS1)c,d 3080(4), 3050, 1610, 1590, 1520, 1480, 1460,
1350, 1320, 1310, 1180, 1170, 1160, 1120,
1110, 1000(2), 970, 940, 840, 790, 600, 500,
NOc,e
1900
400, 380, 350, 300, 270, 180, 160, 150, 120
C6H5NO2•+ (TS2)e,f 3080(4), 3050, 1610, 1590, 1520, 1480, 1460,
1350, 1320, 1310, 1180, 1170, 1160, 1120,
rotational
polarizability
(10-24 cm3)
constants (cm-1
)
1110, 1000(2), 970, 940, 850, 840, 790, 700,
+•
C6H5NO2
0.057b
0.086f
0.152b
2.00e
680, 670, 610, 530, 420, 400(3), 320
C6H5O+
C6H5NO2•+ (TS3)f,g 3080(4), 3050, 1610, 1590, 1520, 1480, 1460,
1350, 1320, 1310, 1180, 1170, 1160, 1120,
+
C6H5
NO+
1110, 1000(2), 970, 940, 850, 840, 790, 700,
600, 470, 420, 350, 330, 310, 300, 240, 180
C6H5O•
0.101f
1.121b
1.67e
10.5g
2.5b
•
NO2
a Numbers in parentheses denote the degeneracies of vibrational
NO•
0.0171e
modes. b Reference 1. c Transition state for the production of C6H5
.
+
a Numbers in parentheses denote the degeneracies of vibrational
modes. b Reference 2. c The same values are used for the ion and radical.
d Reference 1. e Reference 35. f Estimated value. g Reference 36.
d The C-N stretching mode at 850 cm-1 was taken as the reaction
coordinate. e Transition state for the production of NO+. f The NO2
torsional mode at 140 cm-1 was taken as the reaction coordinate for
production of NO+ and C6H5O+. g Transition state for the production
of C6H5O+ ion.
109 s-1) that can be measured with the present method. This
+
explains why the experimental PD-MIKE peak for C6H5
appears symmetric at 514.5 and 488.0 nm.
Nishimura et al. measured the rate constant for this channel
on a microsecond time scale using PEPICO.2 RRKM-QET
fitting of their experimental data required a tight transition state
with E0 less than the endoergicity. On the other hand, the
experimental kinetic energy release could be well explained by
phase space theory, indicating involvement of a loose transition
state. The transition state switching model27f was invoked to
explain the results. In their study, however, the thermal internal
energy of the nitrobenzene ion (∼0.1 eV) was not considered
and the thermochemical data at 298 K instead of those at 0 K
were used in the calculation of the endoergicity. In addition,
improved thermochemical data have appeared since their study.
Hence, the rate-energy data of Nishimura et al. have been
corrected here, and the results are shown in Figure 5. The
corrected results are in excellent agreement with the present
RRKM-QET calculation.
Figure 5. Rate-energy dependence for the dissociation of nitrobenzene
ion to C6H5+. The present photodissociation result is shown as a filled
circle (b). The experimental rate-energy data of Nishimura et al. (0)
are also shown. The theoretical result (RRKM-QET) is shown as a
solid curve. Rate parameters used in the calculation are E0 ) 1.37 eV
and ∆Sq ) 6.4 eu. The bar represents error limit.
In the other PEPICO study by Panczel and Baer,1 the rate
constant for this channel was found to be larger than could be
determined by the technique. The fact that the productions of
C6H5O+ and NO+ occurred on a much longer (microsecond)
time scale than this channel was attributed to the lack of
competition between these two groups of channels. In particu-
degeneracy. In a usual RRKM-QET fitting of rate-energy data
for a particular channel, the critical energy and the activation
entropy at 1000 K (∆Sq) are treated as two parameters to be
adjusted.37 In the present case, E0 of 1.37 eV has been adopted,
as in the calculation of KERD, and ∆Sq has been adjusted to fit
the present experimental data. Also, an attempt has been made
such that the RRKM fitting for this channel and the fittings for
other channels (to be described below) result in branching ratios
at 514.5 and 488.0 nm excitations, which are in agreement with
the experimental data. The best fit has been obtained when
∆Sq is 6.4 eu, as shown in Figure 5. The vibrational frequencies
used in the calculation are listed in Table 5. The large positive
entropy of activation means that the reaction occurs statistically
via a loose transition state. This accords well with the above
analysis of KERD. According to the rate-energy curve in Figure
5, the rate constant at the internal energy of 2.85 eV corre-
sponding to the experiment with 514.5 nm excitation is 3 ×
109 s-1, which is larger than the maximum rate constant (1 ×
+
lar, the production of C6H5 from an excited electronic state
was suggested. This is in sharp contrast with the present result
on a nanosecond time scale that all three channels compete and
that the production of C6H5+ occurs statistically from the ground
electronic state of the molecular ion.
B. Production of NO+. The rate constants for reaction 2b
were determined on a microsecond time scale in the previous
two PEPICO studies.1,2 In the study of Panczel and Baer,1 a
faster channel in addition to the regular statistical one was
reported even though its origin was not clear. Hence, its time-
of-flight was analyzed using two-component rate constants. The
fast component was not observed in the PEPICO study of
Nishmura et al.2 carried out under a similar condition. In the
present study, a one-component rate constant with distribution
has been found to be adequate for data analysis. The RRKM-
QET calculation has been carried out to fit the present and the

9232 J. Phys. Chem., Vol. 100, No. 22, 1996
Hwang et al.
Figure 7. Rate-energy dependence for the dissociation of nitrobenzene
ion to C6H5O+. The present photodissociation result is shown as a filled
circle (b). The experimental rate-energy data of Nishimura et al. (0)
and those of Panczel and Baer (O) are also shown. The theoretical
result (RRKM-QET) is shown as a solid curve. Rate parameters used
in the calculation are E0 ) 1.20 eV and ∆Sq ) -1.1 eu. The bar
represents error limit.
Figure 6. Rate-energy dependence for the dissociation of nitrobenzene
ion to NO+. The present photodissociation result is shown as a filled
circle (b). The experimental rate-energy data of Nishimura et al. (0)
and those of Panczel and Baer (O) are also shown. The theoretical
result (RRKM-QET) is shown as a solid curve. Rate parameters used
in the calculation are E0 ) 1.21 eV and ∆Sq ) -7.1 eu. The bar
represents error limit.
two PEPICO rate data. The slow component rate constants of
Panczel and Baer have been used, and the data of Nishimura et
al. have been corrected for the thermal energy, as mentioned
above. The best fit was obtained using the parameters E0 )
1.21 eV and ∆Sq ) -7.1 eu, as shown in Figure 6. The
vibrational frequencies used in the calculation are listed in Table
5. E0 adopted here is similar to the one (1.27 eV) used in the
RRKM-QET analysis by Panczel and Baer, but higher than that
(1.02 eV) by Nishimura et al. The latter discrepancy is thought
to be due to the fact that thermal energy was ignored in the
estimation of internal energy. The large negative activation
entropy means that this channel occurs via a tight transition
state in accordance with the previous RRKM-QET analyses.1,2
Assuming the phenoxy structure of the neutral fragment
C6H5O•, the reaction endoergicity of ∼0.09 eV is estimated from
the thermochemical data in Table 3. The fact that this is well
below the critical energy of 1.21 eV obtained above means that
a substantial reverse barrier is present. Then, the kinetic energy
release in this reaction will be larger than statistically expected.
In this regard, KERD has been calculated with phase space
theory to compare with the experimental one shown in Figure
4a. The molecular parameters used in the calculation are listed
in Table 4. The calculated KERD which is also shown in Figure
4a is significantly narrower than the experimental one. Namely,
the analysis of KERD suggests that the reaction occurs via a
tight transition state with a substantial reverse barrier, in
agreement with the result from the above rate constant analysis.
Nishmura et al. suggested that the production of NO+ occurs
via two different channels;2 it occurs consecutively via phe-
nylnitrite ion intermediate near the threshold energy with a small
average kinetic energy release, and it occurs directly at higher
internal energy with a large average kinetic energy release. Since
the average internal energy of photoexcited nitrobenzene ions
is well above the threshold, the latter of the two mechanisms
would apply in the present case, in agreement with the
observation of large kinetic energy release. The present result
does not allow us, however, to tell whether switching of the
pathway would occur near the threshold.
reaction 2c. The best fit has been obtained using E0 of 1.20
eV and ∆Sq of -1.1 eu as shown in Figure 7. The vibrational
frequencies used in the calculation are listed in Table 5. This
E0 value is the same as that reported by Panczel and Baer.1
The negative activation entropy indicates that the reaction occurs
via a tight transition state. In Figure 4c, the experimental KERD
is compared to the theoretical one calculated by phase space
theory using the parameters listed in Table 4. Phenoxy structure
of C6H5O+ was assumed in the calculation. The experimental
KERD deviates significantly from the statistical prediction. One
of the reasons for such a deviation is due to the fact that C6H5O+
further dissociates to C5H5+. Namely, C6H5O+ ions generated
with a smaller kinetic energy release would possess a larger
amount of internal energy than those with a larger kinetic energy
release. Thus, the former ions will dissociate more readily,
which results in smaller probability at the smaller kinetic energy
release in the experimental data. Even though such an effect
cannot be ignored, the substantial probability for large kinetic
energy release in the experimental data indicates that a consider-
able reverse barrier is present in this reaction. This is in
accordance with the result of the above rate constant analysis.
D. Production of C5H5+. The present work has confirmed
in real time that C5H5+ is produced by the consecutive reaction
via C6H5O+(reactions 2c and 2d). As shown by the data in
Table 2, the rate constant for the second channel (reaction 2d)
increases with the photon energy, indicating that the internal
energy of the intermediate, C6H5O+, also increases. Detailed
analysis of the dynamics of the second step is not easy, however,
because the structure of C6H5O+ and its internal energy content
are uncertain. It is assumed here that C6H5O+ has the phenoxy
structure, whose thermochemical data are available.34 Its
average internal energy has been estimated in two opposite
extreme situations. In the first method, the total energy available
for the products in reaction 2c, namely, the sum of the excess
energy at the transition state and the reverse barrier, has been
assumed to be partitioned statistically. Such an assumption is
in contradiction with the experimental observation that the
average kinetic energy release is much larger than statistical
prediction. Hence, the internal energy of C6H5O+ estimated in
this way will be an upper limit to the true value. The total
C. Production of C6H5O+. RRKM-QET calculation has
been carried out to fit the present and PEPICO1,2 rate data for

Photodissociation of Nitrobenzene Molecular Ion
J. Phys. Chem., Vol. 100, No. 22, 1996 9233
also shown which has been achieved with the critical energy
of 1.45 eV and the activation entropy of -1.1 eu.
The RRKM-QET analyses of the rate-energy data for reaction
2d estimated with two different internal energy scales require
critical energies which are noticeably different. The critical
energy of 1.91 eV estimated from the data reported by
Nishimura et al.2 agrees better with the one obtained with the
former method of internal energy estimation. However, we are
not in a position to support this method at all because the
estimation of the critical energy for the second step of a
consecutive reaction from an appearance energy measurement
can be erroneous. Reliable critical energy information will be
very helpful in this regard. Fortunately, the activation entropies
obtained from the two analyses are not significantly different,
both having values near zero. Hence, one can tell that reaction
2d occurs via a slightly tight transition state, as is expected from
the reorganization of the reactant structure needed in the
reaction.
E. Overall Dissociation Mechanism. From the data
presented so far, it is clear that the direct bond cleavage to C6H5
+
+
Figure 8. Rate-energy dependence for production of C5H5 from
C6H5O+ The internal energy is that of C6H5O+. The same experimental
rate data are plotted with two different internal energy scales (see text
for details). The rate-energy data using the upper-limit internal energy
scale are shown as solid circles (b) together with the corresponding
RRKM-QET fit (s). E0 of 1.76 eV and ∆Sq of 1.8 eu have been used
in the calculation. The rate-energy data using the lower limit internal
energy scale are shown as open circles (O) together with the
corresponding RRKM-QET fit(- - -). E0 of 1.45 eV and ∆Sq of -1.1
eu have been used in the calculation. Bars represent error limits.
competes with the rearrangement processes to NO+ and C6H5O+
on a nanosecond time scale. Such an overall picture is in
general agreement with the PEPICO study by Nishimura et al.
carried out on a microsecond time scale. Namely, even though
the rearrangement channels to C6H5O+ and NO+ dominate near
the threshold energy, the rate constant for the direct bond
+
cleavage to C6H5 increases more rapidly with the internal
energy and becomes dominant in the photodissociation at 514.5
and 488.0 nm. Such a changeover is well-known in the
dissociations of molecular ions where direct bond cleavage and
rearrangement compete.
available energy has been estimated from the thermochemical
data of the related species in Table 3. Then, its statistical
partitioning has been calculated using the relation proposed by
Klots.38
Even though the conclusion drawn from the present nano-
second result does not agree with the microsecond result reported
by Panczel and Baer, it is worthwhile to examine whether the
ion-dipole intermediate (IDI) mechanism proposed by Oster-
held et al.8 is still applicable to the present result. As was
mentioned in the introductory section, the IDI mechanism shown
below was devised to explain the result of Panczel and Baer
n
r - 1
E* ) kT* +
kT* + hν [exp(hν /kT*) - 1]-1 (5)
∑
i i
(
)
2
i)1
Here, E* is the total energy available for partitioning, r is the
number of rotational degrees of freedom of the products, the
νi’s are the product vibrational frequencies, and kT* is the total
average translational energy in the dissociation. The average
internal energies of C6H5O+ thus estimated are 2.76, 3.09, and
3.21 eV, respectively, at 607.5, 514.5, and 488.0 nm excitations.
The RRKM-QET calculation has been carried out to fit the rate-
energy data thus obtained for reaction 2d varying both the
critical energy and the activation entropy at 1000 K. A good
fit has been obtained with the critical energy of 1.76 eV and
the activation entropy of 1.8 eu. The result is shown in Figure
8.
+
that the production of C6H5 did not compete with the other
rearrangement channels.
k1
+
•
•
C6H5+•••NO2
•
C6H5+ + NO2
C6H5NO2
(IDI)
k2
k3
k4
+
•
NO+ + C6H5O
•
C6H5ONO
(6)
•
C6H5O+ + NO
In the second method, it has been assumed that KERD of
C6H5O+ which dissociates further to C5H5+ is the same as that
of C6H5O+ which remains undissociated, namely, Figure 4c.
At a particular value of the kinetic energy release, the difference
between the total available energy and the kinetic energy release
becomes the energy available for partitioning in the vibrational
and rotational degrees of freedom of C6H5O+ and NO. This
energy has been partitioned statistically over the pertinent
degrees of freedom using the Klots formula. The internal energy
of C6H5O+ thus estimated has been averaged over the experi-
mental KERD in Figure 4c. These are 2.28, 2.62, and 2.74
eV, respectively, at 607.5, 514.5, and 488 nm excitations. Since
it is very likely that C6H5O+ generated with less kinetic energy
release will dissociate more easily, the internal energy estimated
in this way will be only a lower limit to the true value. The
rate-energy data for the second step have been plotted in Figure
8 with this internal energy scale. A good RRKM-QET fit is
C6H5+ + CO
Here, the ion-dipole intermediate (IDI) may either dissociate
+
•
to C6H5 and NO2 or isomerize to the phenylnitrite structure.
If k2 . k3, k4, NO+ and C6H5O+ will be observed with a smaller
apparent total rate constant (k3 + k4) than that (k1 + k2) for
C6H5+. Then, it would look as if the direct and rearrangement
channels do not compete, as was pointed out by Osterheld et
al.
In the present result obtained on a nanosecond time scale,
however, all three channels have been found to compete. Within
the IDI mechanism, such a situation can arise only when the
total rate constant for IDI, kT ) k1 + k2, is much smaller than
that (k3 + k4) for the dissociation of the phenylnitrite ion.
Namely, appearance kinetics of all three ions will be governed
by kT in this case. Predominance of the C6H5+ channel would

9234 J. Phys. Chem., Vol. 100, No. 22, 1996
Hwang et al.
(2) Nishimura, T.; Das, P. R.; Meisels, G. G. J. Chem. Phys. 1986,
require k1 . k2. The above consideration means that k3 and/or
k4 should increase rapidly with the internal energy for the IDI
mechanism to be applicable from a microsecond to nanosecond
time scale. To check such a possibility, we have calculated
the rate constant k4 on the basis of the PEPICO rate-energy
data of Panczel and Baer. The heat of formation of phenylnitrite
ion has been estimated as ∼900 kJ mol-1 from an AM1
calculation by Osterheld et al.8 Our own semiempirical
calculations resulted in similar values. The best RRKM-QET
calculation for k4 to fit the PEPICO data has been obtained using
E0 of 2.35 eV and ∆Sq of 12 eu. Then, k4 at the internal energy
corresponding to photoexcitation at 607.5 nm has been evaluated
with the RRKM-QET calculation using the same parameters.
Finally, the branching ratio information in Table 2 has been
utilized, which has resulted in k3 + k4 of 2.0 × 108 s-1. This
is much smaller than kT of 8.3 × 108 s-1 observed in the present
work, invalidating the applicability of the IDI mechanism on a
nanosecond time scale.
84, 6190.
(3) Bunn, T. L.; Richard, A. M.; Baer, T. J. Chem. Phys. 1986, 84,
1424.
(4) Mukhtar, E. S.; Griffiths, I. W.; Harris, F. M.; Beynon, J. H. Org.
Mass Spectrom. 1980, 15, 51.
(5) Griffiths, I. W.; Harris, F. M.; Mukhtar, E. S.; Beynon, J. H. Int.
J. Mass Spectrom. Ion Phys. 1981, 38, 127.
(6) Kingston, E. E.; Morgan, T. G.; Harris, F. M.; Beynon, J. H. Int.
J. Mass Spectrom. Ion Phys. 1983, 47, 73.
(7) Cassady, C. J.; McElvany, S. W. Org. Mass Spectrom. 1993, 28,
1650.
(8) Osterheld, T. H.; Baer, T.; Brauman, J. I. J. Am. Chem. Soc. 1993,
115, 6284.
(9) Moini, M.; Eyler, J. R. Int. J. Mass Spectrom. Ion Processes 1987,
76, 47.
(10) Meyerson, S.; Puskas, I.; Fields, E. K. J. Am. Chem. Soc. 1966,
88, 4974.
(11) Brown, P. Org. Mass Spectrom. 1970, 3, 1175.
(12) Brown, P. Org. Mass Spectrom. 1970, 4, 533.
(13) McLafferty, F. W.; Bente, P. F., III; Kornfeld, R.; Tsai, S.-C.; Howe,
I. J. Am. Chem. Soc. 1973, 95, 2120.
(14) Jones, E. G.; Bauman, L. E.; Beynon, J. H.; Cooks, R. G. Org.
Mass Spectrom. 1973, 7, 185.
(15) Beynon, J. H.; Bertrand, M.; Cooks, R. G. J. Am. Chem. Soc. 1973,
95, 1739.
(16) Procter, C. J.; Kralj, B.; Brenton, A. G.; Beynon, J. H. Org. Mass
Spectrom. 1980, 15, 619.
(17) Porter, C. J.; Beynon, J. H.; Ast, T. Org. Mass Spectrom. 1981,
16, 101.
(18) Cameron, D.; Clark, J. E.; Kruger, T. L.; Cooks, R. G. Org. Mass
Spectrom. 1977, 12, 111.
(19) Apel, E. C.; Nogar, N. S. Int. J. Mass Spectrom. Ion Processes
1986, 70, 243.
(20) Marshall, A.; Clark, A.; Jennings, R.; Ledingham, K. W. D.;
Singhal, R. P. Int. J. Mass Spectrom. Ion Processes 1992, 112, 273.
(21) Marshall, A.; Clark, A.; Jennings, R.; Ledingham, K. W. D.;
Singhal, R. P. Int. J. Mass Spectrom. Ion Processes 1992, 116, 143.
(22) Lemire, G. W.; Simeonsson, J. B.; Sausa, R. C. Anal. Chem. 1993,
65, 529.
(23) Galloway, D. B., Bartz, J. A.; Huey, L. G.; Crim, F. F. J. Chem.
Phys. 1993, 98, 2107.
(24) Komidis, C.; Ledingham, K. W. D.; Clark, A.; Marshall, A.;
Jennings, R.; Sander, J.; Singhal, R. P. Int. J. Mass Spectrom. Ion Processes
1994, 135, 229.
(25) (a) Robinson, P. J.; Holbrook, K. A. Unimolecular Reactions;
Wiley-Interscience: New York, 1972. (b) Forst, W. Theory of Unimolecular
Reactions; Academic Press: New York, 1973.
(26) Rosenstock, H. M.; Wallenstein, M. B.; Wahrhaftig, A. L.; Eyring,
H. Proc. Natl. Acad. Sci. U.S.A. 1952, 38, 667.
(27) (a) Light, J. C. J. Chem. Phys. 1964, 40, 3221. (b) Pechukas, P.;
Light, J. C. J. Chem. Phys. 1965, 42, 3281. (c) Klots, C. E. J. Phys. Chem.
1971, 75, 1526. (d) Chesnavich, W. J.; Bowers, M. T. J. Am. Chem. Soc.
1976, 98, 8301. (e) Chesnavich, W. J.; Bowers, M. T. J. Chem. Phys. 1977,
66, 2306. (f) Chesnavich, W. J.; Bowers, M. T. J. Chem. Phys. 1981, 74,
2228. (g) Choe, J. C.; Kim, B. J.; Kim, M. S. Bull. Korean Chem. Soc.
1989, 10, 167.
(28) Choe, J. C.; Kim, M. S. J. Phys. Chem. 1991, 95, 50.
(29) Choe, J. C.; Kim, M. S. J. Phys. Chem. 1992, 96, 726.
(30) Yim, Y. H.; Kim, M. S. J. Phys. Chem. 1993, 97, 12122.
(31) Yim, Y. H.; Kim, M. S. Int. J. Mass Spectrom. Ion Processes 1993,
123, 133.
V. Conclusions
Photodissociation of nitrobenzene molecular ions with a
visible photon occurs via several reaction channels. The first
generation dissociation channels producing C6H5O+, C6H5+, and
NO+ have been found to occur competitively. Among the
C6H5O+ ions produced, those with high internal energy dis-
sociate further to C5H5+, which has been observed in real time
in this work. The photodissociation rate constant has been
determined at the molecular ion internal energy of 2.48 eV on
a nanosecond time scale together with KERDs in each channel.
+
It has been found that the production of C6H5 on this time
scale occurs statistically on the ground electronic state potential
energy surface in accordance with RRKM-QET. The rear-
rangement channels producing C6H5O+ and NO+ also occur
statistically. Namely, the results on a nanosecond time scale
are quite typical for cases when direct cleavages and rearrange-
ments compete in the dissociation of a molecular ion. In
particular, no evidence has been found on a nanosecond time
scale to support the ion-dipole intermediate mechanism pro-
posed by Osterheld et al.8 The possibility of this mechanism
on a microsecond time scale cannot be completely ruled out,
however, because some of the nitrobenzene ion dissociation
pathways which can be ignored at high internal energies may
become important at low internal energies. Further investiga-
tion, both experimental and theoretical, seems to be needed in
this regard. Analysis of the rate-energy data for the second step
of the consecutive reaction to C5H5+ indicates that the reaction
occurs statistically also, even though lack of reliable knowledge
on the structure and internal energy content of the intermediate
makes such a conclusion provisional.
(32) Yim, Y. H.; Kim, M. S. J. Phys. Chem. 1994, 98, 5201.
(33) Cho, Y. S.; Kim, M. S.; Choe, J. C. Int. J. Mass Spectrom. Ion
Processes 1995, 145, 187.
(34) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin,
R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1988, 17, Suppl. 1.
(35) Herzberg, G.; Huber, K. P. Molecular spectra and molecular
structure; Van Nostrand Reinhold: New York, 1979; Vol. 4.
(36) Hirshfelder, J. O.; Curtiss C. F.; Bird, R. B. Molecular theory of
gases and liquids; Wiley: New York, 1954.
Acknowledgment. We are grateful to Prof. T. Baer for his
critical reading of the manuscript and valuable discussion. This
work was supported by NON DIRECTED RESEARCH FUND,
Korea Research Foundation, by the Ministry of Education,
Republic of Korea, and by the Center for Molecular Catalysis
and the Korea Science and Engineering Foundation.
(37) Lifshitz, C. AdV. Mass Spectrom. 1989, 11, 713.
(38) Klots, C. E. J. Chem. Phys. 1973, 58, 5364.
References and Notes
(1) Panczel, M.; Baer, T. Int. J. Mass. Spectrom. Ion Processes 1984,
58, 43.
JP9537872