Photodissociation dynamics of C1N3 at 203 nm: The NC1 (a1Δ/X3∑-) product branching ratio

Article abstract of DOI:10.1016/S0009-2614(02)01886-9

Velocity map imaging was applied to study the photodissociation dynamics of ClN₃ near 203 nm under collision free conditions. Images of state-selected N₂(X¹∑_g⁺, v = 0. J = 68) characterize the interna

Full text of DOI:10.1016/S0009-2614(02)01886-9

Chemical Physics Letters 368 (2003) 568–573
www.elsevier.com/locate/cplett
Photodissociation dynamics of ClN at 203 nm: the
3
1
3
À
NCl ða D=X R Þproduct branching ratio
a
a,
b
N. Hansen , A.M. Wodtke , A.V. Komissarov , M.C. Heaven
b
*
a
Department of Chemistry and Biochemistry, University of California, Santa Barbara, CA 93106, USA
b
Department of Chemistry, Emory University, Atlanta, GA 30322, USA
Received 13 August 2002; in final form 14 November 2002
Abstract
3
Velocity map imaging was applied to study the photodissociation dynamics of ClN near 203 nm under collision free
1
þ
conditions. Images of state-selected N
reveal the energetics of reactions (*) and (**):
2
(X R , m ¼0, J ¼68) characterize the internal energy of the photofragments and
g
1
þ
1
ðÃÞ
ClN
3
3
!N
!N
2
2
ðX R ÞþNClða DÞ DE ¼þ0:21 Æ0:07 eV
g
1
þ
3 À
ðÃÃÞ
ClN
ðX R ÞþNClðX R Þ DE ¼À0:93 Æ0:08 eV
g
These experiments provide the most accurate thermodynamics for these reactions presently available and are in good
0
~
1
1 0
~
agreement with other recent experimental results. Photofragment angular distributions indicate that B A
excitation is the most important pathway to photoproducts. Branching measurements between the dominant spin-al-
X A
1
3 À
lowed channel (1) and the spin-forbidden channel (2) showed a D=X R ¼0:78=0:22.
Ó 2002 Elsevier Science B.V. All rights reserved.
1
1
. Introduction
ded measurements are NCl(a D) removal rate
constants. If the rate constants are measured under
pseudo-first order conditions, photolysis of ClN3
can be used without knowledge of the absolute
Recent interest in photochemical production of
1
NCl(a D) derives from demonstrations that this
metastable can be used as an effective energy car-
rier in chemical iodine lasers [1–3]. Furthermore,
initial concentration, ½NClðaފ. However, for
0
rate constants that can only be measured under
second-order conditions such as the important
3
photolysis of ClN has been used as a convenient
1
source of NCl(a D) for kinetic measurements [4–8]
that are required for further development of this
chemical laser system. One of the key sets of nee-
self-annihilation reaction, knowledge of ½NClðaފ
0
is essential. In this case, ½NClðaފ may be mea-
0
3
sured or deduced from the fraction of ClN re-
moved by photolysis as long as the quantum yield
1
for formation of NCl(a D) is known.
Unfortunately the present understanding of the
primary photochemical dynamics of chlorine azide
*
Corresponding author. Fax: 1-805-893-4120.
E-mail address: wodtke@chem.ucsb.edu (A.M. Wodtke).
0
009-2614/02/$ - see front matter Ó 2002 Elsevier Science B.V. All rights reserved.
PII: S0009-2614(02)01886-9

N. Hansen et al. / Chemical Physics Letters 368 (2003) 568–573
569
is incomplete. The photolysis of ClN
3
by near UV
quantum yields of highly reactive photoproducts
may be less accurate unless measured under colli-
sion free conditions. Velocity map imaging [14,15]
solves both of these problems. First, it is a collision
free molecular beam technique. Second, the mea-
surement of specific velocities of state-selected
photofragments may be used together with the
principles of conservation of linear momentum
and energy, to clearly identify the electronic state
of the undetected counter-fragment, allowing un-
ambiguous identification of the reaction from
which the photofragment is being produced.
radiation has been examined by Coombe and co-
workers [4,9–12], Henshaw et al. [5], and Komis-
sarov et al. [6,7]. In these studies photodissociation
was accomplished using pulsed excimer lasers op-
erating at 248 and 193 nm. At these wavelengths,
1
00
~
1
0
~
excitation within both the A A
0
X A and
X A systems is possible, yet the relative
1
0
~
1
~
B A
importance of these absorption systems is not
known. At both wavelengths the dominant disso-
ciation channel is believed to be the spin-allowed
channel (1)
1
1
þ
In this work, we report velocity map imaging
ClN
3
þhm !NClða DÞþN
2
ðX R Þ
ð1Þ
g
3
data for single photon dissociation of ClN at
3
À
Production of NCl(X R ) is possible via spin-for-
bidden channel (2)
k ꢀ203 nm, providing the first look at the pho-
todissociation of ClN under collision free condi-
3
3
À
1
þ
tions. Velocity map images obtained using
resonance enhanced multiphoton ionization (RE-
ClN
3
þhm !NClðX R ÞþN
as well as the spin-allowed channel (3)
2
ðX R Þ
ð2Þ
g
1
þ
MPI) of selected states of recoiling N (X R ) were
2
g
3
À
3
þ
u
resolved and unambiguously assigned to dissocia-
tion channels (1) and (2). The relative importance
of these two channels at 203 nm was obtained and
shows that ground state NCl(X R ) is produced
mainly by the spin-forbidden channel (2). Analysis
of the kinetic energy release data yielded an ac-
curate determination for the exoergicity of
ClN þhm !NClðX R ÞþN ðA R Þ
ð3Þ
3
2
3
þ
N
2
ðA R Þpresumably from channel (3) has been
u
1
þ
3
À
detected [9,10], along with NCl(b R ) which can
be produced by a second spin-allowed channel
[
9,13].
1
þ
1
þ
ClN
3
þhm !NClðb R ÞþN
2
ðX R Þ
ð4Þ
g
1
1
þ
ClN
tribution of the photofragments observed at this
wavelength is consistent with initial excitation of
3
!NClða DÞþN
2
ðX R Þ. The angular dis-
g
Up to now, a quantitative knowledge of the relative
importance of these four channels is not in hand.
Coombe et al. [10] reported that formation of
1
0
1
0
~
X A absorption system.
~
the B A
1
þ
3
þ
NCl(b R ) and N
% of the 193 nm photolysis products. Komissarov
et al. [7] examined the prompt and delayed pro-
2
(A R ) amount to only about
u
1
2. Experiment
3
À
duction of NCl(X R ) using absorption spectros-
copy after ClN photolysis at 248 nm. Assuming
3
The technique of velocity map imaging has been
described fully elsewhere [14]. Chlorine azide
(ClN ) was formed by passing a mixture of 5% Cl
that the delayed signal came from collisional re-
1
laxation of NCl(a D), they were able to estimate
that 80% of the primary NCl was formed in the a
3
2
in He over the surface of moist sodium azide
(NaN ). Astandard drying agent was used to re-
1
þ
state and, apart from the 1% of NCl(b R ), the
remaining NCl was produced in the electronic
ground state. Similar measurements for 193 nm
photolysis indicated production of approximately
0% NCl(a D) with 30% NCl(X R ) [6].
Measurements of photofragment quantum
yields are not equivalent to knowledge of the
quantum yield for a specific reaction, since
photofragments may be produced by more than
one reaction. Furthermore, measurements of
3
move water from the ClN =Cl =He mixture at the
2
3
exit of the generator. The mixture was then ex-
panded through a pulsed nozzle (General Valve)
with a backing pressure of ꢀ0.5 bar.
1
3
À
7
Laser light at k ꢀ203 nm was generated by up-
converting the output from a Nd:YAG/dye laser
combination (Continuum) operating on a rhoda-
mine 610/640 mixture. The output from the dye
laser was frequency doubled in a KDP crystal and

570
N. Hansen et al. / Chemical Physics Letters 368 (2003) 568–573
then the fundamental and second harmonic was
summed in a BBO crystal. About 1 mJ/pulse of
this light was focused with a 10 cm focal length
lens onto the molecular beam. One photon of the
203 nm light was used to dissociate the ClN and
3
three more photons were used to carry out 2+1
REMPI of the N photoproducts [16].
2
Velocity map images were accumulated while
the laser frequency was scanned repetitively over
the Doppler profile of a single rotational transi-
tion. The images were taken with the laser verti-
cally polarized, such that the electric vector was
parallel to the image plane. Background subtrac-
tion of ions produced by the ion decomposition
channel [17] was required and image reconstruc-
tion was accomplished with the Basis-Set-Expan-
sion-Method [18].
The velocity space was calibrated by recording
D images produced by three-photon dissociation
þ
of D
þ3hm !D ðn ¼1ÞþD ðn ¼2Þ
The electronically excited atomic fragment, D
2
[19]
Fig. 1. Inverse Abel transform of the velocity map image of the
product recoiling from ClN photolysis. The excitation
and probes
N
2
3
D
2
ð5Þ
wavelength used, k ¼203:205 nm, dissociates ClN
3
1
þ
the N
2
ðX R
, m ¼0; J ¼68Þdissociation products. The veloc-
g
ities marked with arrows reflect a center-of-mass frame energy
difference of 1.14 eV. When this is compared to the singlet–
triplet splitting in NCl (1.15 eV), compelling evidence is found
for assignment of the inner ring to reaction (*) and the outer
ring to reaction (**).
(
n ¼2), subsequently absorbs an additional pho-
þ
ton and is ionized, producing the D ions with
kinetic energies of about 1.87 eV. The high velocity
þ
D ions were needed to calibrate the system for the
detection of the fast moving fragments that were
expected from one-photon dissociation of ClN3.
N ðm ¼0; J ¼68Þmolecules. At ꢁ4.75 eV and
2
lower, one sees an abrupt rise in intensity, corre-
1
sponding to the onset of formation of NClða DÞin
3
. Results
coincidence with the N ðm ¼0; J ¼68Þ. Exploit-
2
ing the thermodynamic cycle shown in Fig. 3, we
have used this threshold to extract the energetics of
þ
Fig. 1 shows the velocity map image of the N2
produced by laser excitation at k ¼203:205 nm,
which probes the N
1
reaction (*), assuming that NClða D; m ¼0; J ¼0Þ
2
ðm ¼0; J ¼68Þphotoprod-
can be formed at threshold. To accomplish this
one must use the rotational and centrifugal dis-
tortion constants [20] for N to find the rotational
uct. This image was recorded using a high accel-
eration voltage (repeller: 7000 V and extractor:
2
4
900 V) to ensure that fragments produced with
energy of N ðJ ¼68Þ: E ðN Þ¼1:14 eV. The
2
rot
2
the highest possible velocities would land on the
active area of the detector.
derived results (þ0:21 Æ0:08 eV) represents a rig-
orous upper limit to the exoergicity of reaction (*).
Also indicated in Fig. 2 is the singlet–triplet
splitting in NCl, DEða ÀXÞ. From this one can see
that there is an additional threshold (albeit less
abrupt) for formation of ground electronic state
NCl observed in the data. Using the literature
value for DEða ÀXÞ¼1:15 eV [21], we arrive at
the following thermodynamic results:
The center-of-mass frame kinetic energy distri-
bution derived from the inverse Abel transforma-
tion is shown in Fig. 2. Two peaks separated by
about the NCl singlet–triplet splitting are assigned
to channels (1) and (2). The high kinetic energy
peak corresponds to formation of ground elec-
tronic state NCl in coincidence with the detected

N. Hansen et al. / Chemical Physics Letters 368 (2003) 568–573
571
ð6Þ
ð7Þ
1
þ
1
ClN
3
!N
2
ðX R ÞþNClða DÞ
g
DE ¼þ0:21 Æ0:08 eV
1
þ
3
À
ClN
3
!N
2
ðX R ÞþNClðX R Þ
g
DE ¼À0:93 Æ0:08 eV
Also from Fig. 2, we see that the most probable
internal energy in the NCl counter-fragment is
0.51 and 0.42 eV for channels (1) and (2), respec-
tively. It is interesting to note that the exoergic
channel tends to produce NCl with less ro-vibra-
tional excitation. This suggests that there are sig-
nificant differences between the singlet and triplet
potential surfaces governing the dynamics of each
reaction.
1
þ
Fig. 2. Observed center-of-mass frame kinetic energy distribu-
À
Angular distributions of the N ðX R ) frag-
2
g
1
3
tion for reactions (*) and (**): ClN
þ
3
þhm !NClða D; X R Þþ
ments were analyzed assuming the center-of-mass
angular distributions, f h), exhibit the usual form
1
N
2
ðX R Þ. The bimodal distribution is assigned to the elemen-
g
tary chemical processes as shown in the figure. The abrupt
threshold at 4.75 eV can be used to derive the energetics of re-
action (*), DEreac:ð1Þ¼þ0:21 eV. Also shown is the singlet–triplet
energy splitting in NCl, DEða ÀXÞ.
[
22]
f ðhÞ/1 þb2P
2
ðcos hÞ;
ð8Þ
were P
nomial. Fig. 4 shows the observed angular distri-
2
ðcos hÞis the second-order Legendre poly-
bution of N photofragments from reaction (*),
2
integrating over velocity at each recoil angle. The
solid smooth line shows the fit to Eq. (6). This
Fig. 3. Energy diagram showing species involved in this ex-
periment. Photolysis of ClN
À
3
at k ¼203:205 nm yields
Fig. 4. Observed angular distribution of N
obtained in a one-photon dissociation of ClN
2
photofragments
1
3
1
þ
NClða DÞor NClðX R Þand N
2
ðX R Þthrough exothermic
3
at k ¼203:205
g
1
þ
1
reactions. The total amount of excess energy released was ꢁ5.89
and 7.03 eV, respectively. The results derived from this work are
shown in bold face.
nm into N
the fit to Eq. (6) yielding an anisotropy parameter b ¼1:95ð6Þ
2
ðX R ÞþNClða DÞ. The solid smooth line shows
g
2
indicating a parallel transition.

572
N. Hansen et al. / Chemical Physics Letters 368 (2003) 568–573
fitting procedure (which was also performed for
data at higher recoil velocities important for the
triplet channel) allows us to derive the anisotropy
parameters for both singlet and triplet products.
We find that within experimental error both
channels exhibit the same anisotropy parameter
recent high-level ab initio calculation. Using a
MRSDCI-CASSCF(12e/10o)/D95+d level calcu-
lation, Morokuma [23] obtained a value for DE of
1
+0.52 eV, close to the value (+0.21) reported
here.
1
3
À
The NClða DÞ=NClðX R ) branching ratio of
0.78/0.22 obtained from the velocity map image
for 203 nm photolysis is close to the 0.70/0.30 ratio
observed using 193 nm photolysis with detection
ðb ¼1:95 Æ0:06Þ. This result suggests that both
2
singlet and triplet channels result after initial ex-
1
0
1
0
~
X A parallel absorption
~
citation in the B A
system. Presumably, production of triplet products
3
À
of NCl(X R ) by transient absorption spectros-
copy [6]. This agreement has several implications.
1
0
~
requires a curve crossing between the B A surface
and a nearby triplet surface.
3
À
First it strongly suggests that the NCl(X R )
produced by the UV photochemistry of ClN3
comes nearly exclusively from the spin-forbidden
channel (2). If the spin-allowed channel (3) were an
2
Having assigned the two rings in the N velocity
1
map to the production of NCl(a D) and
3
À
NCl(X R ) we integrated the velocity map image
to obtain the branching ratio between the a and
X-states of NCl appearing in coincidence with
3
À
important source of NCl(X R ), agreement be-
tween these very different experiments would be
quite unlikely. This also lends strength to the prior
measurements that reported the formation of
N
2
ðm ¼0; J ¼68Þ. We obtained a ratio of 0.78/
1
0.22 in favor of NCl(a D). While this value is
strictly speaking only applicable to the specific
3
þ
N ðA R Þas a minor channel [10].
2
u
channel observed in this work where N is formed
2
As mentioned earlier, 203 nm light excites the
1
1
0
~
0
~
in m ¼0 and J ¼68, we do not expect that the
B A ÀX A transition and the observed frag-
1
0
~
singlet triplet branching will be strongly dependent
on N rotational state. Furthermore, these results
are similar to previous cell experiments which are
ments arise from the dissociation of ClN ðB A ).
3
2
Consequently, the singlet–triplet branching ratio
depends on the probability of crossing from the
excited singlet to the ground-state triplet surface.
The agreement between the branching ratio ob-
tained here and that obtained in the cell experi-
ment [7] appears to substantiate our statements
above that the singlet–triplet branching ratio is at
2
integrated over all N internal states [6,7].
4
. Discussion
From a study of the dissociative photo-ioniza-
tion of ClN , Hansen et al. [17] obtained upper
least not strongly dependent on the N rotational
2
3
state. This fits well with an exit-curve-crossing
mechanism where there is not a large NANBN
bond angle change upon curve crossing. Further-
more, our experiments show that about 70–80% of
the available energy appears as translation. In light
of the large translational energies observed here
compared to the amount of rotational excitation,
we expect that the velocity at the curve crossing
will not be greatly affected by switching rotational
channels.
limits to the exoergicities of reactions (*) and (**).
3
In that work two-photon ionization of ClN gave
rise to ionic fragments and the exoergicity of the
þ
ion decomposition channel (forming NCl þN
2
Þ
could be obtained from velocity map images. Us-
ing previously determined values for the ionization
energies of ClN and NCl, the following results
3
were obtained : DE < 0:3 and DE < À0:85 eV,
1
2
which are in excellent agreement with the present
work. Both experiments yield upper limits for the
thermochemistry. However, now that two inde-
pendent means arrive at similar results, we con-
clude that these Ôupper limitsÕ represent the best
experimental values of the energetics of reaction
Alternative dissociation pathways for ClN ex-
3
cited at wavelengths below 320 nm include
2
ClN þhm !ClðPÞþN
ð9Þ
3
3
and
(*) and (**) and we recommend these be used in
the future. These results may be compared to
2
2
ClN
þhm !ClðPÞþNðDÞþN
:
ð10Þ
3
2

N. Hansen et al. / Chemical Physics Letters 368 (2003) 568–573
573
2
2
Both Cl( P) and N( D) atoms have been detected
in imaging experiments that will be described in a
subsequent publication. The branching fraction
for production of atomic products has not been
characterized, but indirect evidence indicates that
production of diatomic fragments dominates for
Acknowledgements
This work was supported by the AirForce Of-
fice of Scientific Research under Grant Numbers
F49620-95-1-0234 (UCSB) and F49620-01-1-0070
(Emory). We thank H. Reisler (University of
Southern California) for providing us with her
software program for analysis of the imaging data.
NH acknowledges the support of the Alexander
von Humboldt Foundation under a Fyodor Lynen
Stipend.
1
93 and 248 nm photolysis. For example, Ray and
1
Coombe [1] demonstrated an NCl(a D) driven
iodine laser using 193 nm photolysis of a mixture
3 2 2
of ClN and CH I . Analysis of the pumping ki-
netics shows that the quantum yield for generation
1
of NCl(a D) had to be greater than 0.5 to reach
lasing threshold. The results of this work are
consistent with that of Ray and Coombe [1].
References
[
[
1] A.J. Ray, R.D. Coombe, J. Phys. Chem. 99 (1995) 7849.
2] T.L. Henshaw, G.C. Manke, T.J. Madden, M.R. Berman,
G.D. Hager, Chem. Phys. Lett. 325 (2000) 537.
3] G.C. Manke, G.D. Hager, J. Mod. Opt. 49 (2002) 465.
4] A.J. Ray, R.D. Coombe, J. Phys. Chem. 98 (1994)
8940.
5
. Conclusions
[
[
We observed 2+1 REMPI signal and ion ve-
þ
1
locity maps of state-selected N
2
ðX R Þ, resulting
g
[5] T.L. Henshaw, S.D. Herrera, G.W. Haggquist, L.A.V.
Schlie, J. Phys. Chem. A101 (1997) 4048.
3
from the one-photon dissociation of ClN . Two
channels with a maximum 4.75 and 5.90 eV
translational energy were observed and assigned to
channels (1) and (2):
[6] A.V. Komissarov, G.C. Manke II, S.J. Davis, M.C.
Heaven, Proc. SPIE 3931 (2000) 138.
[7] A.V. Komissarov, G.C. Manke, S.J. Davis, M.C. Heaven,
J. Phys. Chem. A106 (2002) 8427.
[8] A.J. Ray, R.D. Coombe, J. Phys. Chem. 97 (1993) 3475.
1
1
þ
ClN
3
þhm !NClða DÞþN
2
ðX R Þ
ð1Þ
ð2Þ
g
[9] R.D. Coombe, D. Patel, A.T. Pritt, F.J. Wodarczyk, J.
3
À
1
þ
g
Chem. Phys. 75 (1981) 2177.
ClN þhm !NClðX R ÞþN ðX R Þ
3
2
[
[
[
10] R.D. Coombe, S.J. David, T.L. Henshaw, D.J. May,
Chem. Phys. Lett. 120 (1985) 433.
11] R.D. Coombe, M.H. Vanbenthem, J. Chem. Phys. 81
These results show that ClN
thermically to form N and NCl. We derived ex-
oergicities of DE < þ0:21 Æ0:08 eV and < À0:93 Æ
:09 eV for decomposition of ClN into N and
3
decomposes exo-
2
(
1984) 2984.
12] R.H. Jensen, A. Mann, R.D. Coombe, J. Phys. Chem. A
104 (2000) 6573.
0
3
2
1
3
À
[13] A.T. Pritt, D. Patel, R.D. Coombe, J. Chem. Phys. 75
1981) 5720.
14] A. Eppink, D.H. Parker, Rev. Sci. Instrum. 68 (1997)
477.
NCl(a D) or NCl(X R ), respectively. These re-
sults compare well with quantum chemical calcu-
lations and results obtained earlier from another
experiment in which the two-photon dissociative
(
[
[
3
15] D.H. Parker, A. Eppink, J. Chem. Phys. 107 (1997)
2357.
3
ionization of ClN was observed [17].
[
[
16] K.R. Lykke, B.D. Kay, J. Chem. Phys. 95 (1991) 2252.
17] N. Hansen, A.M. Wodtke, A.V. Komissarov, K. Moro-
kuma, M.C. Heaven, J. Chem. Phys. (submitted).
18] V. Dribinski, A. Ossadtchi, V.A. Mandelshtam, H. Reisler,
Rev. Sci. Instrum. 73 (2002) 2634.
We also obtained the singlet–triplet NCl
branching ratio (0. 78/0.22) in the UV photolysis
of ClN near 203 nm, that is, the relative impor-
3
[
tance of channels (1) and (2). These results also
À
3
show that NCl(X R ) is produced mainly by the
spin-forbidden channel (2) and confirm that the
[19] M.A. Buntine, D.P. Baldwin, D.W. Chandler, J. Chem.
Phys. 96 (1992) 5843.
[20] J. Bendtsen, F. Rasmussen, J. Raman Spectrosc. 31 (2000)
3
þ
formation of N
2
(A R ) is a minor channel. Fi-
u
4
33.
nally, angular distributions suggest that the
[
[
21] R. Colin, W.E. Jones, Can.J. Phys. 45 (1967) 301.
22] R.N. Zare, Mol. Photochem. 4 (1972) 1.
1
0
1
0
~
X A absorption system is responsible for
~
B A
production of photoproducts at this wavelength.
[23] K. Morokuma, private communication.