Energetics of halogen amine reactions with triplet metastables: NFCl2 with Ar(3P0,2), He(23S), and N2(3Σ+u) and NCl3 with Ar(3P0,2)

Article abstract of DOI:10.1021/jp961790v

The collisions of NFCl₂ with the metastable species Ar(³P_0,2), He(2³S, 2¹S), and N₂(³Σ⁺u) and of NF₂Cl and NCl₃ with Ar(³P_0,2) were carried out using a discharge flow system, and diatomic product channels were identified on the basis of emission from the fragments. The mechanisms for these reactions were determined from the emission spectra of the products, the energy available in the reacting species and released in the fragments, and the conservation of spin angular momentum. To carry out the energy analysis, the ΔH_form's of NF₂Cl, NFCl₂, and NCl₃ were calculated using G1 and G2 methods with Gaussian 94, yielding average theoretical values of -5.3, 23.0, and 48.3 kcal/mol, respectively. These values are compared to estimated experimental ΔH_form's which have been reported for NF₂Cl and NCl₃.

Full text of DOI:10.1021/jp961790v

840
J. Phys. Chem. A 1997, 101, 840-848
Energetics of Halogen Amine Reactions with Triplet Metastables: NFCl2 with Ar(3P0,2),
He(23S), and N2(3Σ+ ) and NF2Cl and NCl3 with Ar(3P0,2)
u
Yevgeniya I. Gershanovich and Julanna V. Gilbert*
Department of Chemistry & Biochemistry, UniVersity of DenVer, DenVer, Colorado 80208
ReceiVed: June 18, 1996; In Final Form: October 31, 1996X
The collisions of NFCl2 with the metastable species Ar(3P0,2), He(23S, 21S), and N2(3Σ+ ) and of NF2Cl and
u
NCl3 with Ar(3P0,2) were carried out using a discharge flow system, and diatomic product channels were
identified on the basis of emission from the fragments. The mechanisms for these reactions were determined
from the emission spectra of the products, the energy available in the reacting species and released in the
fragments, and the conservation of spin angular momentum. To carry out the energy analysis, the ∆Hform’s
of NF2Cl, NFCl2, and NCl3 were calculated using G1 and G2 methods with Gaussian 94, yielding average
theoretical values of -5.3, 23.0, and 48.3 kcal/mol, respectively. These values are compared to estimated
experimental ∆Hform’s which have been reported for NF2Cl and NCl3.
Introduction
mechanisms can be written for the formation of ClF(D′) and
for the formation of the high-lying electronic states in the 193
nm experiments.
Halogen amines are a fascinating class of compounds that
have been studied in our laboratory for several years. In
particular, the series NCl3, NFCl2, and NF2Cl have been of
interest to us because of the changes observed in stability as
chlorine atoms are replaced with fluorine atoms and because
of the ease with which excited state fragments can be generated
both chemically and photolytically from the parent amines (for
example, the a1∆ and b1Σ+ states of NF and NCl and the D′
states of Cl2 and ClF).1-5 It has been observed that conservation
of spin angular momentum directs the product formation in the
photolytic and chemical processes of these compounds studied
to date.
NFCl2 + hν f NFCl2*
(1a)
NFCl2* f Cl2(A′3Π2u) + NF(X3Σ-)
Cl2(A′) + hν f Cl2(D′)
(1b)
(1c)
These data complement the results obtained from photolysis
studies of NFCl2 in low-temperature argon matrices, where the
formation of NF was observed.1
Pertinent to the work presented here are the results from the
gas phase and low-temperature matrix photolysis experiments
of NFCl2, NF2Cl, and NCl3.1,3,5,6 The UV absorption spectra
of these compounds consist of structureless features, charac-
teristic of dissociated electronic states. NFCl2 has two maxima
at 275 and 190 nm,3,7 which are accessible with KrF (249 nm,
4.98 eV) and ArF (193 nm, 6.42 eV) excimer laser wavelengths.
When gas mixtures of NFCl2 in argon were photolyzed at 249
nm, both UV and visible emissions were observed, with features
at 258, 284, and 440 nm.3 The UV emission peaks were
assigned to the D' f A' transitions in Cl2 and ClF, respectively.
The weak visible emission at 440 nm observed in the 249 nm
photolysis study was tentatively assigned to NFCl, an assign-
ment that was later confirmed by low-temperature matrix
isolation experiments.8 Photolysis of NFCl2 at 193 nm resulted
in UV emission bands at 201, 209, 216, 225, 234, and 284 nm,
with no emission in the visible region.3 The 284 nm band was
assigned to ClF(D′ f A′). When the pressure of argon was
increased from 2 to 26 to 54 Torr, the bands between 200 and
235 nm were quenched and Cl2(D′ f A′) emission at 258 nm
emerged, indicating that these were emission from high-lying
electronic states of Cl2 (electronic states above the D′ state).
For both photolysis wavelengths, the species responsible for
the emission were determined to be initially formed photofrag-
ments, and an energy balance analysis indicated that a two-
photon process was responsible for the formation of the ion
pair states of Cl2 and ClF. The proposed mechanism is shown
here for the formation of Cl2(D′) at 249 nm, but analogous
The UV absorption spectrum of NF2Cl contains two broad
structureless features with maxima at 185 and 235 nm.1,7
Photolysis at 193 nm was used to probe the 185 nm feature,
and emission was observed at 284 nm, the D′ f A′ emission
of ClF, and at 215 nm, emission from ClF originating from an
ion pair state above the D′ state.5 A two-photon mechanism,
analogous to the mechanism of reactions 1a-1c was proposed
for the 193 nm photolysis in which ClF(A′3Π) and NF(X3Σ-)
were formed via the absorption of the first photon, with the
second photon promoting ClF to the high-lying excited triplet
ion pair states.
The UV absorption spectrum of NCl3 displays an intense
maximum at 220 nm and a somewhat weaker feature at 260
nm with a red tail that extends weakly into the visible.2,9 When
NCl3 was photolyzed at 308 nm (4.03 eV) and 249 nm (4.98
eV) with XeCl and KrF excimer lasers, respectively,6 no
emission from diatomic fragments was observed, in contrast
with NFCl2 and NF2Cl.1,29 Visible emission was observed and
has since been assumed to be from NCl2 generated in the
photolysis pulse. The excited electronic states of NF2Cl, NFCl2,
and NCl3 generated in the photolysis experiments were assigned
as singlets since the ground state is a singlet, and in the case of
the first two, they are easily observed in absorption.
This paper describes a set of experiments in which NFCl2,
NF2Cl and NCl3 were reacted with triplet metastable species.
NFCl2 was reacted with metastables Ar(3P0,2), He(23S), and N2-
(3Σ+u) to examine the product formation as a function of
metastable energy. NF2Cl and NCl3 were reacted with Ar(3P0,2)
to compare the product channels with those of NFCl2. Spec-
troscopic studies have shown that reactions between singlet
X Abstract published in AdVance ACS Abstracts, January 1, 1997.
S1089-5639(96)01790-2 CCC: $14.00 © 1997 American Chemical Society

NFCl2 Reactions with Ar, He, and N2
J. Phys. Chem. A, Vol. 101, No. 5, 1997 841
ground state molecules and triplet metastables usually excite
the molecule to the triplet state manifold.10 Consequently, triplet
states of the halogen amines should be accessed in these
reactions, and the dissociation products will be different from
those of the singlet states produced in the photolysis work.
Similar studies were carried out by Zhao and Setser with PF3,
and emission from PF and from the PF2 radical was observed,
indicating diatomic and triatomic dissociation channels.11 In a
separate set of experiments, Zhao and Setser generated NCl-
(b1Σ+) via the reaction of NFCl2 with He metastables and
measured its radiative lifetime and several quenching rate
constants.12 For the experiments reported here, it was antici-
pated that similar products would be formed, and while no
emission from triatomic radicals was detected, there was
identifiable emission from diatomic fragments NF, NCl, and
Cl2 in the NFCl2 and NF2Cl studies and from NCl and Cl2 in
the NCl3 studies. Mechanisms for the formation of these
fragments are proposed on the basis of the electronic states of
the fragments identified via their emission spectra, the energetics
of the systems for which the ∆Hform, using G1 and G2 methods
with Gaussian 94, were calculated,13-17 and the spin constraints
of the system.
submerged in a methanol/dry ice bath. The cold bath was
removed, and any condensed Cl2 was allowed to evaporate.
Then, the NCl3 vapor was carried by the flow of argon through
1/4 in. black Teflon tubing to the discharge flow cell. Black
tubing was used to minimize the photodecomposition of NCl3.
A Teflon metering valve, placed before the cell, allowed a small
fraction of the NCl3/Ar gas mixture to enter the gas discharge
system, and the rest was returned to the hood. It should be
noted that NCl3 can explosively decompose and should be used
with adequate safety precautions.
Halogen amine/Ar or halogen/Ar gas mixtures were admitted
to the stream of metastables through a perforated circular Pyrex
ring placed inside the dc discharge-flow cell and located 1 cm
below the lower electrode. The pressure in the system was
monitored with an MKS Baratron gauge (Model 122A), and
for the data presented here, the total pressure in the flow dc
discharge cell was 5.8 Torr for the Ar* and N2*experiments
and 15 Torr for the He* experiments. Reagent (with the
exception of NCl3, which decomposes upon passing through a
flow meter) and carrier gas mass flows were measured with
calibrated MKS mass flow meters, and the linear flow velocity
of the system was 200 cm s-1 at the pressures used in these
experiments. The reagent densities in the flow cell were
calculated from the reagent flows, the carrier gas flow, and the
total pressure. The NFCl2 or NF2Cl density varied between 0.56
× 1014 and 1.1 × 1014 molecules/cm3, and the density of
metastables in the dc discharge was assumed to be 0.01% of
the total density of Ar, He, or N2.10
A circular 2.54 cm o.d. fused quartz viewing port was located
4 cm below the Pyrex ring on the side of the dc discharge flow
tube (2.54 cm o.d.). For a second series of experiments, the
circular viewing port was removed, a rectangular opening
extending from 1 to 5 cm below the Pyrex ring was cut into
the side of the Pyrex tube, and a fused quartz window was glued
over the rectangular opening. Ultraviolet and visible emissions
were dispersed by a Jarrell Ash 0.275 m monochromator with
a 2 nm band-pass, and detected with a cooled GaAs photomul-
tiplier tube (Hamamatsu R943-02). A cooled indium detector
was used in place of the photomultiplier tube to look for infrared
emission. The signal from the detector was transformed into a
voltage signal with an electrometer (Keithley Instruments, Model
610C) and visualized with a strip-chart recorder (Alltech
Associates). In some experiments, a yellow glass filter was
placed between the viewing port of the discharge cell and the
entrance slit of the monochromator to block second-order UV
emission lines in the visible region.
Experimental Section
The discharge-flow method described by Zhao and Setser11
was used to study the reactions of gaseous NFCl2, NF2Cl, and
NCl3 with the various metastables. The flow of metastables
was produced by passing Ar, N2, or He gas through the high-
voltage zone of the discharge cell. A 180 kohm resistor was
placed in series with the discharge, and the voltage between
the two tantalum electrodes was 400, 450, and 890 V to generate
N2*, Ar*, and He*, respectively. Passing pure N2 gas through
a dc discharge generates what is known as “active nitrogen”,
which contains reactive species in addition to N2(A3Σu+),
including atomic nitrogen, N2(C3Πu), N2(B3Πg), and vibra-
tionally excited N2(X1Σg+). The effect of these other species
in the flow system will be addressed.
NFCl2 and NF2Cl were synthesized and purified as described
in ref 1 and 4. Briefly, a stainless steel reactor tube (30 cm
long and 1.8 cm o.d.) was filled with 14 g each of NH4Cl and
NaCl and clean, dry copper shot. The charged reactor was
heated to 110 °C for 4 h while passing dry N2 gas through it to
completely dry the contents. The reactor was cooled (to 55 °C
to optimize the NFCl2 production or to 35 °C to optimize the
NF2Cl production), and the synthesis was started by passing a
20% mixture of F2 in helium (Spectra Gases) through the reactor.
The net chemical reaction has been proposed to be18
Results
2NH4Cl + 3NaCl + 7F2 w
The energies of the excited metastables Ar(3P0,2), N2(A3Σu+),
and He(23S, 21S) are listed in Table 1. Also included in Table
1 is a summary of the emitting products detected when NFCl2
was admitted to a stream of each of these metastables. In the
Ar metastable experiments, both visible and UV emission were
observed. The visible emission appeared to the eye as a very
intense orange flame, and its spectrum is shown in Figure 1B.
To achieve the spectral resolution in Figure 1B, the bandwidth
of the monochromator was reduced to 1 nm. The single blue
degraded peak with a maximum at 530 nm is identified as the
∆V ) 0 sequence band of the b1Σ+ f X3Σ- transition in NF.
The emission bands between 710 and 535 nm are assigned to
the b1Σ+ f X3Σ- transition in NCl. The most intense peak at
664 nm is part of the ∆W ) 0 sequence band, and the other
sequence bands are assigned as indicated in the figure, with
the individual band maxima listed in Table 2. These data are
in good agreement with previously reported NCl emission data
NFCl2 + NF2Cl + 8HF + 3NaF + Cl2 (2)
The reaction products were collected in a Teflon U-trap
submerged in liquid N2. NFCl2 or NF2Cl was separated from
the other products by vacuum distillation techniques, and Pyrex
bulbs containing between 5 and 10% NFCl2 or NF2Cl in argon
were prepared. UV absorption spectra, taken with a 7.5 cm
Pyrex cell equipped with fused quartz windows and a Milton
Roy Spectronic 3000 array spectrometer, were used to verify
the presence of NFCl2 or NF2Cl in the prepared bulbs.
NCl3 was synthesized following the method described in ref
6 in which a gaseous mixture of Cl2 and argon was passed
through an aqueous solution of (NH4)2SO4 in 1M H2SO4. Due
to its high vapor pressure and low solubility in water, NCl3 was
swept out of the solution with the argon and the excess chlorine
and was condensed as an amber-colored liquid in a cold trap

842 J. Phys. Chem. A, Vol. 101, No. 5, 1997
Gershanovich and Gilbert
TABLE 1: States and Energies (E) of He, Ar, and N2 Metastable Species and the Emission Observed upon Reaction of These
Metastables with NFCl2
species
Hea
state
21S
23S
E, cm-1
E, kcal/mol
E, eV
emission observed
166 272
159 850
475.3
457.0
20.6
19.8
NF(b1ΣfX3Σ); ∆V ) 0
NCl(b1ΣfX3Σ); ∆V ) 0
Cl2 emission at 258, 246, 236, 226, 216 nm
Ara
3P0
3P2
94 554
93 144
270.3
266.3
11.7
11.5
NF(b1ΣfX3Σ); ∆V ) 0
NCl (b1ΣfX3Σ); ∆V ) -1, 0, +1, +2, +3
Cl2 (D′fA′) at 258 nm, secondary reaction
Cl2 emission at 198 nm, secondary reaction
NF(b1ΣfX3Σ); ∆V ) 0
b
+
N2
3Σu
50 934
145.7
6.31
a See Table 1 in ref 10 for state assignments and energies. b N2(3Σu+) energy calculated from [Te + ωe/2] from ref 30.
systems.26 However, at the pressures used in these experiments,
this emission should be effectively quenched according to the
data provided in ref 26. Consequently, this emission is assigned
to Cl2. The D′ f A′ transition in ClF is a blue degraded band
which is known to lie at 284 nm.27 A band at 284 nm is
observed in Figure 2A; however, it is red degraded and is also
observed in Figure 2B and in Figure 2C (the spectrum of the
emission from the argon discharge with no added reagents), so
is associated with emission from the discharge. As is apparent
in Figure 2, this emission is greatly attenuated when either
NFCl2 or Cl2 is added to the system.
The spectra of Figures 2A,B appear to be nearly identical,
and the question arose as to whether or not Cl2 contamination
in the NFCl2 sample was responsible for the UV emission. Cl2
is a product of the synthesis, as shown in reaction 2, and the
separation of NFCl2 from Cl2 requires that the last step in the
vacuum distillation procedure be repeated many times. It is
judged to be complete when the Cl2 absorption at 330 nm does
not appear in the UV spectrum of the sample. The small
extinction coefficient of Cl2 (60 L mol-1 cm-1 at 330 nm)28
and the fact that the tail of the NFCl2 absorption peak at 270
nm extends weakly into the 330 nm region makes it possible,
however, that some Cl2 was present in all of the samples used
in this study. To address this problem, experiments were run
in which Cl2/Ar samples were admitted to the Ar metastables,
and the emission intensity at 258 and 198 nm as a function of
Cl2 density was monitored. Calibration plots of the intensity
vs Cl2 density were prepared from these data, and the plot
obtained for the 258 nm emission is shown in Figure 3A. Figure
3B is a plot of the Cl2 emission observed when an NFCl2/Ar
sample was added to the stream of argon metastables. (The
Cl2 + Ar(3P0,2) were performed on the same day as the NFCl2
+ Ar(3P2) experiments to reduce the possibility that the
metastable density had changed for the Cl2 and the NFCl2 runs.)
An upper limit on the Cl2 contamination in the NFCl2/Ar sample
was calculated from the UV absorption spectrum of the NFCl2
sample, and the Cl2 emission intensity expected for this density
of Cl2 was read from the calibration plots. It was assumed that
this intensity could be subtracted from the total Cl2 emission
intensity observed in the NFCl2/Ar experiments. However, the
decays observed in Figure 3 indicate that quenching of the Cl2
emission becomes important when the reagent density gets larger
than about 1.5 × 1013 molecules cm-3, and this complicates
the correction of the Cl2 emission intensity in the NFCl2
experiments for Cl2 contamination. Consequently, only NFCl2
data in the low-density regime, where the intensity increases
with density, could be corrected for Cl2 contamination. Curve
A of Figure 4 shows the total Cl2 emission intensity in the NFCl2
experiment, and curve B shows the Cl2 emission intensity
corrected for the Cl2 contamination. Clearly, some of the Cl2
emission observed in the NFCl2 experiment can be ascribed to
the presence of Cl2 contamination in the sample, but not all of
it. For example, at NFCl2 densities of 6 × 1012 molecules/
Figure 1. Visible emission spectrum observed when (A) NF2Cl, (B)
NFCl2 and (C) NCl3 were added to a stream of Ar(3P0,2) metastables.
The bandwidth of the emission monochromator was 2 nm for (A) and
(C), and 1 nm for (B).
by Colin et al.19 and by Pritt and Coombe and co-workers.20,21
When Cl2 was reacted with the argon metastables, no such
emission was observed.
Since the NFCl2 + Ar(3P0,2) reaction generates the b1Σ+ states
of NF and NCl, the possibility that the a1∆ states of NF and
NCl were also produced was explored. When the PMT was
replaced with the cooled indium detector to probe the infrared
region, the only emission that was observed was from the B3Πg
+
f A3Σu transition in N2,22 with no evidence of the a1∆ f
X3Σ- transitions in NCl (at 1080 nm) and NF (at 874 nm).20,23
The nitrogen source was determined to be a small air leak.
The spectrum of the UV emission, shown in Figure 2A,
consists of two features with maxima at 258 and 198 nm. Both
bands were also observed when a 5% mixture of Cl2 in argon
was admitted to the stream of argon metastables as shown in
Figure 2B. The 258 nm band is the well-known transition
between the D'(3Π 2g) and A′(3Π 2u) states of Cl2.24 Cl2
emission at 198 nm has been reported, and although the specific
electronic states involved are debated, it is likely that this
transition originates from an ion pair state at a higher energy
than the D′ state.25 The C f A transition of ArCl has been
observed in the 198 nm spectral region, and there is certainly a
possibility of direct abstraction of Cl by an Ar(3P0,2) in these

NFCl2 Reactions with Ar, He, and N2
J. Phys. Chem. A, Vol. 101, No. 5, 1997 843
TABLE 2: Individual Band Maxima (in nm) for the NCl (b1Σ+fX3Σ-) Emission Shown in Figure 2
sequence
assignment
sequence
assignment
sequence
assignment
sequence
assignment
∆V ) -1
V′, V′′
∆V ) 0
V′, V′′
∆V ) +1
V′, V′′
∆V ) +2
V′, V′′
703
697
691
686
680
674
0, 1
1, 2
2, 3
3, 4
4, 5
5, 6
664
660
665
0, 0
1, 1
2, 2
626
619
619
615
611
608
604
600
1, 0
2, 1
3, 2
4, 3
5, 4
6, 5
7, 6
8, 7
589
587
584
581
578
575
572
569
566
563
3, 1
4, 2
5, 3
6, 4
7, 5
8, 6
9, 7
10, 8
11, 9
12, 10
Figure 2. UV emission spectra observed upon the addition of (A)
NFCl2 and (B) Cl2 to a stream of Ar(3P0,2) metastables. The bands at
258 and 198 nm are assigned to Cl2. (C) UV spectrum of the Ar
discharge.
Figure 3. (A) Cl2 emission intensity at 258 nm as a function of Cl2
density upon the addition of a 5% mixture of Cl2 in argon to the stream
of Ar(3P0,2) metastables. (B) Cl2 emission intensity at 258 nm as a
function of NFCl2 density upon the addition of a 5% sample of NFCl2
in argon to the stream of Ar(3P0,2) metastables.
cm3, one-half to two-thirds of the Cl2 emission is from NFCl2
and the remainder is attributed to the Cl2 contamination. The
same analysis was performed for the 198 nm emission feature,
with a similar result. Since the Cl2 density calculations were
upper limits, these analyses give lower limits on the contribution
that NFCl2 gives to the Cl2 emission at both 198 and 258 nm.
Although the quenching process in these experiments has not
been identified, it is not direct quenching of the Cl2 ion pair
states because Cl2 ion pair state lifetimes are on the order of 10
ns in the system, and the gas kinetic collision rate is 8 µs-1
Torr-1. Therefore, under the conditions of these experiments
a Cl2 molecule in an ion pair state will see less than one collision
before it emits. The quenching process must, however, involve
a species that leads to the formation of Cl2 in its ion pair state
and could be quenching of Ar(3P0,2) atoms due to changes in
the walls as the reagents are introduced into the system.
No emission from ClF was observed in these experiments.
To ascertain whether this was due to an experimental problem
or due to the fact that no emitting states of ClF were generated
in the NFCl2 plus metastable reactions, a tank of ClF was
obtained (which unfortunately was contaminated with Cl2), and
a bulb of ClF in argon was prepared and added to the stream of
metastables. Still, no ClF emission was observed. With the
idea that perhaps the ClF emission was too weak to see 4 cm
(0.019 ms) downstream of the mixing zone, the rectangular
viewing port described in the Experimental Section was
installed, allowing the emission 1 cm below the mixing region
Figure 4. (A) Total Cl2 emission intensity at 258 nm as a function of
NFCl2 density from Figure 3A. (B) Cl2 emission intensity at 258 nm
corrected for Cl2 contamination plotted as a function of NFCl2 density.
An upper limit on the Cl2 contamination in the NFCl2 sample was
calculated as described in the text.
to be viewed. All of the ClF and NFCl2 reactions were repeated,
but no emission from ClF emission was observed in any of the
experiments, and the conclusion is that ClF emission cannot be
detected in this system.
The spectrum of the emission produced when N2 was passed
through the dc discharge with no added reagent consisted of
N2(B3ΠgfA1Σu+, first positive) emission in the visible, and

844 J. Phys. Chem. A, Vol. 101, No. 5, 1997
Gershanovich and Gilbert
TABLE 3: Emission from Fragments Generated in the
Reactions of Ar(3P0,2)a with NFCl2, NF2Cl, and NCl3
molecule
NFCl2
emission observed
NF(b1Σ+fX3Σ-); ∆V ) 0
NCl(b1Σ+fX3Σ-); ∆V ) -1, 0, +1, +2, +3
Cl2(D′fA′) at 258 nm, secondary reaction
Cl2 emission at 198 nm, secondary reaction
NF (b1Σ+fX3Σ-); ∆V ) 0
NF2Cl
NCl3
NCl(b1Σ+fX3Σ-); ∆V ) +1, +2, +3 (weak)
ArF(2BfX) at low pressures of argon
NCl(b1Σ+fX3Σ-); ∆V ) -1, 0, +1, +2, +3
Cl2(D′fA′) at 258 nm, secondary reaction
Cl2 emission at 198 nm, secondary reaction
UV emission must be due to residual chlorine present in the
sample as a contaminant. Cl2 is a product of both the synthesis
and NF2Cl thermal decomposition, and it is, consequently, nearly
impossible to completely eliminate Cl2 from the sample. Again,
as in the NFCl2 experiments, no ClF emission was detected in
this experiment. When the Ar pressure was reduced to 2.0 Torr,
a peak at 192 nm assigned to the B f X transition in ArF
appeared as a second maximum just to the blue of the 198 nm
Cl2 feature.26
Addition of the NCl3 sample below the argon discharge
generated an intense orange emission flame similar to that
observed for the NFCl2 + Ar(3P0,2) reaction, and the spectrum
(Figure 1C) indicates the ∆V ) 0, +1, +2, and +3 sequence
bands of the b1Σ+ f X3Σ- transition in NCl. The UV emission
was again essentially the same as that observed in the NFCl2
and NF2Cl reactions and was similarly assigned to ion pair
transitions in Cl2.
Figure 5. UV emission spectra observed upon the addition of (A)
NFCl2 and (B) Cl2 to a stream of He(23S, 21S) metastables. The band
at 284 nm is assigned to the He discharge since it was observed when
no reagent was added to the system and looked very similar to the
feature that appears in Figure 2C. The other bands are assigned to Cl2
(see text).
N2(C3Πuf B3Πg, second positive) and NO(A2Σ+fX2Π, γ
band) emission in the UV.22 When NFCl2 was added, the NF-
(b1Σ+fX3Σ-) emission at 528 nm appeared in the visible, and
there was a general decrease in the intensity of the N2 first
positive and of the NO γ band emission. No UV emission from
Cl2 or ClF was observed, however, from the addition of either
NFCl2 or Cl2. Visible and UV emission were observed when
NFCl2 was reacted with He metastables. Two visible bands at
528 and 664 nm were assigned to the ∆V ) 0 sequence bands
of the b1Σ+ f X3Σ- transitions in NCl and NF. The UV
emission spectrum is shown in Figure 5A and consists of a peak
with a maximum at 258 nm assigned to Cl2(D'fA') and a series
of bands at shorter wavelengths with individual band maxima
at 246, 236, 226, 216, and 206 nm. These bands correlate well
with the short wavelength bands observed in the 193 nm
photolysis of NFCl2 and are assigned to transitions originating
from high lying electronic states Cl2.3,24 The relative intensity
and the distinct nature of these UV bands were a function of
the He pressure, and to see them well required helium pressures
of 15 Torr. The emission observed when Cl2 was added to the
flow of He metastables is shown in Figure 5B. Zhao and Setser
reported NCl(bfX) emission from the reaction of NFCl2 with
He metastables; however, since they were only interested in
the NCl emission, they did not report any other emission that
they may have observed.12
Emitting fragments generated in the NF2Cl or NCl3 plus Ar
metastable rare gas atom reactions are summarized in Table 3,
with the NFCl2 reactions for comparison. When NF2Cl was
admitted to the stream of Ar metastables, the emission appeared
to the eye as an emerald green flame. Consistent with this
observation, the spectrum, shown in Figure 1A, reveals an
intense single blue degraded peak at 528 nm assigned to the b
f X transition in NF. The weak emission features in the 650-
570 nm region are the ∆V ) 1, 2, and 3 sequence bands of the
NCl(bfX) transition. Nitrogen first positive emission appears
in the 650-700 nm region and obscures the NCl(bfX) ∆V )
0 sequence band. The UV emission spectrum was virtually
identical to the NFCl2 + Ar(3P0,2) UV emission spectrum with
Cl2 features at 258 and 198 nm. Since Cl2 cannot be formed in
a direct reaction of NF2Cl with Ar metastables, the observed
Discussion
When molecular species react with metastable argon atoms,
Penning ionization, associative ionization, neutral excitation,
neutral dissociation, and neutral dissociative excitation may
occur.10 The last of these, neutral dissociative excitation, is
assumed to be responsible for the production of excited state
diatomic fragments in the Ar(3P0,2) reaction systems. On the
basis of the product emissions observed in the NFCl2 system,
there are two pathways leading to diatomic fragments by which
the reaction proceeds, as shown here.
(3a)
NF + Cl2
NFCl2 + Ar(3P0,2
)
(3b)
NCl + ClF
Analogous diatomic pathways can be written for the NF2Cl/
Ar(3P0,2) system, and for the NCl3/Ar(3P0,2) system, a single
channel leading to the formation of NCl and Cl2 can be written.
Two aspects of this process will be analyzed here, the energetics
and the spin constraints.
To examine the energetics of the reactions requires that the
∆Hform values for the products and reactants and the energies
of the excited states of the products be known. For NFCl2, the
only ∆Hform available is the experimental lower limit of -10
kcal/mol determined from the NFCl2 gas-phase photolysis
experiments.3 Therefore, G1, G2-MP2, and G2-MP4 methods
were used with Gaussian 94 to compute the energies of NFCl2,
N2, F2, and Cl2.13-17 These energies were used together with
the formation reaction to compute the ∆Eform for NFCl2. The
∆Hform was calculated from ∆Eform + ∆nRT where ∆n is the
change in the number of moles in the formation reaction. The
results of the three calculations are included in Table 4 and
show about a 5% variation. To estimate the reliability of the
calculated ∆Hform of NFCl2, ∆Hform’s were calculated for NF3,
NCl3, and NF2Cl, species for which experimental values have
been reported. These values are also included in Table 3. For

NFCl2 Reactions with Ar, He, and N2
J. Phys. Chem. A, Vol. 101, No. 5, 1997 845
with Ar 3P2 atoms (11.7 eV) or with Ar 3P0 (11.5 eV) to generate
NF(b1Σ+) and Cl2(D′). To generate the Cl2 state that is
responsible for the emission at 198 nm, however, would require
an additional 4.15 kcal/mol in the reaction involving the more
energetic Ar(3P0) or 8.75 kcal/mol in the reaction with Ar(3P2).
As noted above, the spectrum of the Cl2 emission observed
in the NFCl2 + Ar(3P0,2) experiments is virtually identical to
that observed in the Cl2 + Ar(3P0,2) experiments. It is very
unlikely that these reactions would generate identical Cl2
emission spectra in view of the energetics of the two cases,
particularly since according to our calculations, there is insuf-
ficient energy to produce the 198 nm Cl2 emission in the NFCl2
reaction. It is, however, possible that the Cl2 ion pair states
are not direct products of the NFCl2 + Ar(3P0,2) reaction but
are formed in a secondary step as shown here.
TABLE 4: Theoretical and Experimental ∆Hform’s in kcal
mol-1 for NF3, NF2Cl, NFCl2, and NCl3
molecule
G1
G2-MP2 G2-MP4 av theor
exptl
NF3
-35.36
-3.94
+24.12
+49.54
-38.38
-5.90
+22.61
+47.67
-38.29
-6.16
+22.28
+47.59
-37.55
-5.33
+23.00
+48.32
-31.43a
NF2Cl
NFCl2
NCl3
+5b
+29.1c
+63d
a Reference 37. Reference 29. Predicted experimental value as-
b
c
d
suming that the average theoretical ∆Hform is 21% too low. Reference
9.
NFCl2 + Ar(3P2) f NF + Cl2
(4a)
(4b)
Cl2 + Ar(3P2) f Cl2(ion pair states)
The electronic states of the diatomic species in this mechanism
can be analyzed on the basis of the observed emission and spin
constraints. Since reactions between singlet ground state
molecules and triplet metastable atoms usually excite the
molecule to the triplet state manifold, then for our system an
intermediate triplet state of NFCl2 is produced in step 4a. The
emission data indicate that a channel exists in which the NF
fragment is produced in an excited singlet state, so spin
conservation rules predict that the Cl2 fragment will be a triplet.
There are three low lying 3Π states of Cl2, the B(0u+), the A(1u),
and the A′(2u) states.30 The B state generates well-known red/
orange emission which was not observed in these experiments.
No emission corresponding to the A and A′ states was observed,
although none would be expected due to their long lifetimes
and the high probability of collisional quenching in the system.31
A “dark” dissociation channel in which the ground state
fragments NF(X3Σ-) and Cl2(X1Σg+) are generated also cor-
relates with the triplet NFCl2 intermediate state, and this channel
is the likely source of the ground state Cl2 for step 4b, since it
would generate identical emission to what is observed when
Cl2 is added to the stream of Ar metastables.
The NCl + ClF channel can also be analyzed, and the left
side of the energy diagram (Figure 6) shows the energetics of
this channel using the known energies in electronvolts for NCl-
(b1Σ+) and ClF(D′).19,20,24 Since the energy of the ClF(A′) state
is not established, it was assumed to have the same energy as
the B3Π state because the analogous states are very close in
energy in Cl2 and F2.30 Spin constraints predict that the ClF
will be generated in a triplet state with NCl(b1Σ+), similar to
the NF(b1Σ+) + Cl2(triplet state) case above, and the energy
diagram indicates that there is sufficient energy in the NFCl2
+ Ar(3P2,0) reaction for the production of ClF(D′) with NCl-
(b1Σ+). There is also a spin and energetically allowed “dark”
dissociation channel possible leading to the formation of ground
state NCl and ClF.
The reactions of NFCl2 with N2 metastables (6.31 eV) and
He (19.8 eV) probes dissociating triplet states at energies lower
and higher than the state produced in the Ar metastable (11.7
eV) studies. For the reactions with N2 metastables, neutral
dissociative excitation is assumed to be responsible for the
production of excited state diatomic fragments as in the Ar
metastable reactions. The NF(bfX) emission at 528 nm shows
that the NF channel is open; however, no emission from Cl2
was observed. There are several reactive species in active N2,
and these must all be considered regarding their possible
Figure 6. Energy level diagram for NFCl2 and the diatomic channels
NF + Cl2 and NCl + ClF. The energies listed under “Total Energy”
were calculated from the energies of the diatomic fragments for each
channel indicated.
NF3 and NCl3, the average calculated ∆Hform’s are 19 and 23%
lower, respectively, than the experimental values. (It is noted,
however, that there is significant uncertainty in the experimental
∆Hform of NCl3.) Assuming that the theoretical value for NFCl2
is 21% too low predicts an “experimental” value of 29.1 kcal/
mol. The theoretical and experimental values of ∆Hform for NF2-
Cl were not used in this analysis because the validity of the
experimental value is debatable.29 The results of the G1, G2-
MP2, and G2-MP4 methods gave very similar ∆Hform’s for NF2-
Cl and yield an average value of -5.33 kcal/mol.
Figure 6 is an energy level diagram for NFCl2 and the
diatomic dissociation channels. The right side of the diagram
shows the energetics of the NF plus Cl2 pathway using the
known energies for NF(b1Σ+), Cl2(D′), and Cl2(A′), converted
to electronvolts.24,30 Using the adjusted ∆Hform for NFCl2, the
diagram indicates that there is sufficient energy in the reaction

846 J. Phys. Chem. A, Vol. 101, No. 5, 1997
Gershanovich and Gilbert
participation in the production of NF(b1Σ+) from NFCl2. For
nitrogen atoms to be involved would require that the N atom
extract an F atom from the NFCl2 directly or break apart the
NFCl2 molecule to produce electronically excited NF. This
seems very unlikely, particularly since the density of nitrogen
atoms below the discharge is quite small in typical systems.
The visible and UV spectra of the active nitrogen indicate the
presence of several excited electronic states of N2, A3Σu+, B3Πg,
and C3Πu.22 Of these, only the A state has a sufficiently long
lifetime to suffer enough collisions with NFCl2 to be an
important contributor to product formation. Figure 6 shows that
+
there is sufficient energy in the reaction with a 6.3 eV N2(A3Σu
)
metastable to generate NF(b1Σ+) plus Cl2 in a low-lying triplet
state and NCl(b1Σ+) plus ClF in a low-lying triplet state. The
details of the potential surface of the dissociating NFCl2 state
must be responsible for the fact that the NCl(b1Σ+) channel is
not observed, even though it is not prohibited by the energetics
of the reaction. Dark channels are possible both energetically
and by spin considerations as in the argon metastable reactions.
The fact that no D′ f A′ emission was observed when either
NFCl2 or Cl2 was added is consistent with the fact that there is
insufficient energy in the N2 metastable to promote ground state
Cl2 to the D′ state and does not preclude the dark channels.
The He metastables are the most energetic of the metastables
used in this study, and are the only ones with both triplet and
singlet components at 19.8 and 20.6 eV, respectively. Since
the ∆V ) 0 sequence band of the NCl(b1Σ+fX3Σ-) transition
and emission from several high-lying electronic triplet states
of Cl2 were observed, spin constraints predict that the dissociat-
ing state of NFCl2 is generated in the reaction with He
metastables is a triplet. The high-lying ion pair states of Cl2
are a product of the dissociating excited triplet state of NFCl2
since only the D′ f A′ emission at 258 nm was observed when
the Cl2/Ar sample was reacted with He metastables. This is in
contrast to reactions of NFCl2 with Ar or N2 metastables where
the Cl2 emission at 258 and 198 nm was assigned to secondary
reactions with ground state Cl2. The ∆V ) 0 sequence band of
the NF(bfX) emission was also observed in the He metastable
reactions, so the NF + ClF channel is open; however, as in the
other metastable reactions, no information regarding the elec-
tronic state of the product ClF is available from the data. In
the NFCl2 + He(1,3S) studies done by Zhao and Setser,12 it was
proposed that NFCl2 is ionized in the collision, and then NCl-
(b) is formed following the recombination of the parent ion with
an electron. It is not possible to verify whether this mechanism
or direct excitation of a triplet state of NFCl2 is responsible for
the excited state product formation from the data available. In
either case, however, the dissociating state of NFCl2 must be a
triplet state because of the spin multiplicity of the products.
Furthermore, if the excited triplet state of NFCl2 is generated
in a direct reaction with He, then it can be assigned to a reaction
involving He(3S) atoms rather than He(1S). This is reasonable
in view of the fact that there is a much larger density of He(3S)
generated in the discharge than of He(1S).10
Figure 7. Energy level diagram for NF2Cl and the diatomic channels
NF + Cl2 and NCl + ClF. The energies listed under “Total Energy”
were calculated from the energies of the diatomic fragments for each
channel indicated.
is sufficient energy to generate ClF(D′), as shown in the energy
level diagram, although no ClF(D′ f A′) emission was
observed. There are also other low-lying triplet states that are
energetically accessible, the B(3Π0+) and A′(3Π2) states. A
“dark” channel leading to the production of the ground state
fragments, NF(X3Σ-) and ClF(X1Σ+), also correlates with a
triplet intermediate state of NF2Cl. The left side of the energy
level diagram represents the energetics of the NCl + F2 channel.
Spin rules dictate that production of NCl(b1Σ+) must be
accompanied by production of the F2 fragment in a triplet state.
However, generation of the D′ state of F2 is not possible when
the energetics of the NCl + F2 channel are examined. This
channel would yield a lower limit for the ∆Hform of NF2Cl of
135 kcal/mol, an unreasonably high value and certainly out of
the question compared to the ∆Hform’s of Table 4. The
formation of one of the low-lying triplet states of F2(3Π+ or
0u
3Π2u) plus the b1Σ+ state of NCl and the “dark” channel leading
to NCl(X3Σ+) + F2(X1Σ+g) are allowed product channels based
on spin and energy considerations.
The NF2Cl/Ar(3P0,2) and NCl3/Ar(3P0,2) were carried out to
compare the processes in these chemically related systems to
those observed in the NFCl2/Ar(3P0,2) system. An energy
diagram for the NF2Cl system is shown in Figure 7. Since, as
mentioned above, the experimental ∆Hform value for NF2Cl is
not well established, the average theoretical value from Table
4 is included in the diagram. The experiment shows that the
NF(b1Σ+) fragment is formed in the reaction. Thus, assuming
that an excited electronic triplet state of NF2Cl is formed in the
NF2Cl + Ar(3P0,2) reaction, the spin constraints predict that the
second fragment, ClF, must be produced in a triplet state. There
The formation of ArF(2B) was observed in the Ar(3P0,2) +
NF2Cl reaction when the argon pressure was reduced to 2 Torr.
This product is allowed by the spin constraints of the system,
and there are several possible mechanisms by which it could
be generated. First, direct abstraction of a fluorine atom from
NF2Cl could lead to the ArF(2B) formation, leaving a doublet
NFCl as the other reaction product, although NFCl was not
detected in these studies. Ground state F2(X1Σ+g) generated in
the “dark” channel is also a candidate for a secondary reaction

NFCl2 Reactions with Ar, He, and N2
J. Phys. Chem. A, Vol. 101, No. 5, 1997 847
spectra were run at different amine densities. Since the
vibrational distributions were observed to not depend on the
amine densities, this assumption is valid. For the NFCl2/
Ar(3P0,2) and the NF2Cl/Ar(3P0) systems, only the ∆V ) 0
sequence band is observed in the NF(b f X) emission spectrum,
implying that the majority of the NF fragments are emitting
from the V ) 0 level of the b1Σ+ state. Quenching of NF-
(b1Σ+) is generally inefficient,32 and quenching of v′ > 0 levels
would require a polyatomic or diatomic species with a near-
resonant vibrational mode. Potential collision partners in the
flow system are argon atoms, the diatomic species NF, NCl,
ClF, or Cl2, and parent molecules. Of these, only argon is
present at high enough densities for a sufficient number of
collisions, but it would not be an efficient vibrational quencher
of NF(b1Σ+,v′>0). It is concluded, therefore, that the dissociat-
ing states of NFCl2 and NF2Cl produce NF(b1Σ+,v′) 0) directly
and that the NF bond of the dissociating state is approximately
the same length as re of NF(b1Σ-), 1.2998 Å.30 The length of
the NF bond in the ground electronic state of both NFCl2 and
NF2Cl is reported to be 1.382 Å, so the dissociating state has a
shorter NF bond than does the ground state in both amines.33,34
Vibrational excitation up to v′ ) 12 was observed in the NCl-
(bfX) emission spectra observed in all three amine systems
(NF2Cl, NFCl2, and NCl3). Therefore, using the same reasoning,
the NCl bond of the dissociating parent amine must be longer
than the bond length of the b state of NCl (re ) 1.5714 Å).30
For comparison, the NCl bond length of the ground state of
both NF2Cl and NFCl2 is 1.730 Å and of the ground state of
NCl3 is 1.759 Å.33,34 Applying this analysis to the NFCl2 +
N2(3Σ+) reaction, the fact that only ∆V ) 0 sequence bands
were observed in the b f X emission of both NF and NCl
indicates that the N-F and N-Cl bonds in the dissociating state
are very similar to the re values of the b states of NF and NCl.
For the metastable reactions of NFCl2 in which both NF and
NCl b f X emission was observed, the [NCl(b)]/[NF(b)] ratio
was evaluated with the following expression.
Figure 8. Energy level diagram for NCl3 and the diatomic channel
NCl + Cl2. The energies listed under “Total Energy” were calculated
from the energies of the diatomic fragments for each channel indicated.
with Ar(3P0,2) leading to the ArF(2B) + F(2P) production.
Finally, the mechanism of excimer formation via an excited
molecular halogen channel has been reported for the Cl2* +
Xe(Kr), Br2* + Xe, and IBr* + Xe reactions.27 By analogy
with these reactions, it is possible that ArF(2B) could also be
formed via an excited molecular fluorine channel where an
excited triplet state of fluorine acts as a precursor.
[
(SNCl(bfX)qNCl)]ANF
∑
[NCl(b)]
[NF(b)]
λ
)
[
(SNF(bfX)qNF)]ANCl
∑
λ
where the sum is over the wavelength range of the NCl or NF
emission, S is the signal intensity of the NCl or NF b f X
emission at wavelength λ, q is the system response at each
wavelength, and A is the Einstein A coefficient. A halogen
quartz lamp with known output was used to measure the system
response as a function of wavelength. Two values for the
lifetime of the NCl(b1Σ+) have been reported, 0.63 and 2.0 ms,
An energy level diagram constructed for the NCl3 + Ar(3P0,2)
reaction system is given in Figure 8. As can be seen from the
diagram, a sufficient amount of energy is present in the reaction
with either 11.7 or 11.5 eV Ar metastables to generate Cl2
emission at 258 nm. To generate the 198 nm emission,
however, an additional 6.92 or 11.5 kcal/mol is needed in a
reaction with a Ar(3P0) or Ar(3P2) metastable, respectively.
Therefore, analogous to the NFCl2 case, a two-step mechanism
is proposed to account for the generation of chlorine emission.
The dissociating state of NCl3 formed as an intermediate in the
NCl3(1X) + Ar(3P0,2) reaction correlates with two pathways both
leading to the production of diatomic fragments. The NCl(b1Σ+)
and Cl2(low-lying triplet state) fragments are generated in
channel one. Presumably, the ground states NCl(X3Σ-) and
Cl2(X1Σ+g) are formed in the second, “dark”, dissociation
channel. The ion-pair states of chlorine are assumed to be
formed via a secondary reaction between the ground state of
Cl2 formed in the “dark” channel and argon metastables.
The vibrational excitation of the nitrene products is a
reflection of the geometry of the dissociating halogen amine,
assuming no vibrational or electronic quenching of the b1Σ+
states of NF and NCl. To check this assumption, the visible
which yield Einstein A values of 1.59 × 103 and 5.0 × 102 s-1
,
respectively,21,35 and the Einstein A coefficient for NF(b1Σ+) is
44.3 s-1 36
In the NFCl2 plus Ar metastable reactions, the [NCl-
.
(b)]/[NF(b)] ratio is 5/4 using 630 ms for the NCl(b) lifetime
(or 4/1 using 2 ms for the NCl(b) lifetime). For the NFCl2 +
He metastable reaction, the choice of the NCl(b1Σ+) lifetime
changes the [NCl(b)]/[NF(b)] ratio from unity for the shorter τ
to 3/1 for the longer τ, but for either lifetime, the NF(b) channel
is not the preferred pathway. Comparison of the visible
emission spectrum observed in the NF2Cl/Ar(3P0,2) system
(Figure 6A) with that observed in the NFCl2/Ar(3P0,2) system
(Figure 1) indicates that, in the NF2Cl case, the NCl to NF ratio
is lower. Although the ratio was not calculated because of the
difficulty in resolving the NCl(bfX) emission in the NF2Cl
experiment, it appears that there is a preference for NF(b) over
NCl(b) production. Statistically, one would expect this to be

848 J. Phys. Chem. A, Vol. 101, No. 5, 1997
Gershanovich and Gilbert
the case, and apparently the potential curves do not select against
the statistical expectation.
our understanding of the excited electronic states of these
intriguing molecules.
Conclusion
References and Notes
(1) Conklin, R. A.; Gilbert, J. V. J. Phys. Chem. 1990, 94, 3027.
(2) Gilbert, J. V.; Smith, L. J. J. Phys. Chem. 1991, 95, 7278.
(3) Exton, D. B.; Williams, S. A.; Gilbert, J. V. J. Phys. Chem. 1993,
97, 4326.
(4) Exton, D. B.; Gilbert, J. V.; Coombe, R. D. J. Phys. Chem. 1991,
95, 7758.
(5) Hubbard, L.; Gilbert, J. V. Manuscript in preparation.
(6) Gilbert, J. V.; Wu, X. L.; Stedman, D. H.; Coombe, R. D. J. Phys.
Chem. 1987, 91, 4265.
(7) Bouknight, S.; Gilbert, J. V. Unpublished vacuum-UV absorption
spectrum of NFCl2.
(8) Gilbert, J. V.; Okin, G. S. J. Phys. Chem. 1995, 99, 11365.
(9) Clark, T. C.; Clyne, M. A. A. Trans. Faraday Soc. 1970, 66, 372.
(10) Stedman, D. H.; Setser, D. W. Chemical Applications of Metastable
Rare Gas Atoms; Pergamon Press: Oxford, 1971.
(11) Zhao, Y.; Setser, D. W. Chem. Phys. Lett. 1993, 210, 362.
(12) Zhao, Y; Setser, D. W. J. Chem. Soc., Faraday Trans. 1995, 91,
2979.
(13) Gaussian 94 (Revision A.1): Frisch, M. J.; Trucks, G. W.; Schlegel,
H. B.; Gill, M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith,
T. A.; Petersson, G. A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham,
M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.;
Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala,
P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts,
R.; Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart,
J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian, Inc., Pittsburgh,
PA, 1995.
(14) Pople, J. A.; Head-Gordon, M.; Fox, D. J.; Raghavachari, K.;
Curtiss, L. A. J. Chem. Phys. 1989, 90, 5622.
(15) Curtiss, L. A.; Jones, C.; Trucks, G. W.; Pople, J. A. J. Chem.
Phys. 1990, 93, 2537.
(16) Curtiss, L. A.;Raghavachari, K.;Trucks, G. W.; Pople, J. A. J. Chem.
Phys. 1991, 94, 7221.
(17) Curtiss, L. A.; Raghavachari, K.; Pople, J. A. J. Chem. Phys. 1993,
98, 1293.
(18) Conklin, R. A. M.S. Thesis, University of Denver, 1989.
(19) Colin, R.; Jones, W. E. Can. J. Phys. 1967, 45, 301.
(20) Pritt, A. T.; Patel, D., Jr.; Coombe, R. D. J. Mol. Spectrosc. 1981,
87, 401.
(21) Coombe, R. D.; Patel, D.; Pritt, A. T., Jr.; Wodarczyk, F. F. J.
Chem. Phys. 1981, 75, 2177.
(22) Pearse, R. W. B.; Gaydon, A. G. The Identification of Molecular
Spectra, 4th ed.; Wiley: New York, 1976.
(23) Jones, W. E. Can. J. Phys. 1967, 45, 21.
(24) Tellinghuisen, J.; Chakraborty, D. K. Chem. Phys. Lett. 1987, 134,
565.
(25) Wilson, M. W.; Rothscild, M.; Rhodes, C. K. J. Chem. Phys. 1983,
78, 3779.
(26) Sadeghi, N.; Cheaib, M.; Setser, D. W. J. Chem. Phys. 1989, 90,
219.
(27) Diegelmann, M.; Hohla, K.; Rebentrost, F.; Kompa, K. L. J. Chem.
Phys. 1982, 76, 1233.
(28) Calvert, J. G.; Pitts, J. N., Jr. Photochemistry, Wiley: New York,
1967.
(29) Petry, R. C. J. Am. Chem. Soc. 1967, 89, 4600.
(30) Huber, K. P.; Herzberg, G. Molecular Spectra and Molecular
Structure. Constants of Diatomic Molecules; Van Nostrand Reinhold: New
York.
(31) Bondybey, V. E.; Fletcher, C. J. Chem. Phys. 1976, 64, 3615.
(32) Cha, H.; Setser, D. W. J. Phys. Chem. 1989, 93, 235.
(33) Ramasway, K.; Mohan, N. A. Naturforsch. 1970, 25B, 169.
(34) Vikov, L. V.; Nazarenko, I. I. Zh. Strukt. Khim. 1967, 8, 346.
(35) Zhao, Y.; Setser, D. W. J. Chem. Soc., Faraday Trans. 1995, 91,
2979.
(36) Tennyson, P. H.; Fontijn, A; Clyne, M. A. A. Chem. Phys. 1981,
62, 171,
(37) JANAF Thermochemical Tables; Dow Chemical: Midland, MI,
1971.
The studies reported here have yielded new information on
excited electronic states of NFCl2, NF2Cl, and NCl3 via reactions
of these amines with excited triplet metastable species. Excited
singlet states of these amines had been probed via photolysis
experiments carried out previously, and information on excited
triplet states is now available from the data presented here. For
NFCl2, two excited singlet states had been observed in the 249
nm (4.98 eV) and 193 nm (6.42 eV) photolysis experiments,
and evidence for three excited triplet states was observed
following the reaction of NFCl2 with triplet metastables of
energy 6.3, 11.7, and 19.8 eV. All five of these excited
electronic states have dissociation pathways leading to the
formation of diatomic fragments which can be understood on
the basis of conservation of spin angular momentum. Only in
the 249 nm photolysis was a nondiatomic dissociation pathway
producing NFCl observed. The direct production of ion pair
states of the diatomic halogens appears to be possible only in
the reactions of NFCl2 with He metastables; otherwise, low-
lying triplet states of the halogens are generated.
The singlet and triplet electronic states of NFCl2 observed to
date can be classified according to the preferred diatomic
dissociation channel. The NF + Cl2 channel is the favored
dissociation pathway for both the lowest excited singlet state
pumped at 4.98 eV and the triplet state accessed at 6.3 eV.
Dissociation of the second excited singlet state pumped at 6.42
eV and of the triplet states observed at 11.7 and 19.8 eV
generally favor the NCl + ClF over the NF + Cl2 channel.
The results of the NF2Cl/Ar(3P0,2) and the NCl3/Ar(3P0,2
reaction systems are similar to those of the NF2Cl/Ar(3P0,2
)
)
system, which is not surprising since these molecules are closely
related to each other. There are again diatomic dissociation
channels which are easily understood on the basis of spin
constraints and energetics. The variations observed in the NX
species generated are also not unexpected for the series NF2Cl,
NFCl2, and NCl3. In the NF2Cl system, both NF(b) and NCl-
(b) are generated with an apparent preference for NF(b). In
the NFCl2 system, again, both NF(b) and NCl(b) are generated,
but with a preference for NCl(b). Of course, only NCl(b) is
observed in the NCl3 system. As a part of this work, the
∆Hform’s for NF3, NF2Cl, NFCl2, and NCl3 were computed. The
experimental value for NF3 is fairly well established, and the
calculated value was 19% lower. For NCl3, an experimental
value had been reported, and the calculated value was similarly
too low (by 21%). This suggests that the experimental ∆Hform
for NFCl2 is ∼20% higher than the computed value. The trend
observed in the calculated ∆Hform’s for the series NF3, NF2Cl,
NFCl2, and NCl3 (-31.4, -5.3, 23.0, and 48.3 kcal/mol,
respectively) conforms to expectations regarding the stability
of these compounds. As the F atoms are replace by Cl atoms,
the stability decreases.
Computations of the electronic states are planned and, with
the information obtained from the metastable studies presented
here and the previously reported photolysis studies, will improve