Initial state resolved electronic spectroscopy of HNCO: Stimulated Raman preparation of initial states and laser induced fluorescence detection of photofragments

Article abstract of DOI:10.1063/1.475190

Stimulated Raman excitation (SRE) efficiently prepares excited vibrational levels in the ground electronic state of isocyanic acid, HNCO. Photofragment yield spectroscopy measures the electronic absorption spectrum out of initially selected states by monitoring laser induced fluorescence (LIF) of either NCO (X ²II) or NH (a ¹Δ) photofragments. Near threshold, the N-H bond fission is predissociative, and there is well-resolved rotational and vibrational structure in the NCO yield spectra that allows assignment of K_a, rotational quantum numbers to previously unidentified vibrational and rotational levels in the ν₁ N-H stretch and ν₃ N-C-O symmetric stretch fundamentals in the ground electronic state of HNCO. The widths of NCO yield resonances depend on the initial vibrational state, illustrating one way in which initial vibrational state selection influences dissociation dynamics. Initial excitation of unperturbed ν₁ (N-H stretch) states leads to diffuse NCO yield spectra compared to excitation of mixed vibrational levels. The higher energy dissociation channel that produces NH (a ¹Δ) has coarser structure near its threshold, consistent with a more rapid dissociation, but the resonance widths still depend on the initially selected vibrational state.

Full text of DOI:10.1063/1.475190

Initial state resolved electronic spectroscopy of HNCO: Stimulated Raman preparation
of initial states and laser induced fluorescence detection of photofragments
Steven S. Brown, H. Laine Berghout, and F. Fleming Crim
Citation: The Journal of Chemical Physics 107, 8985 (1997); doi: 10.1063/1.475190
View online: http://dx.doi.org/10.1063/1.475190
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/107/21?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Analysis and prediction of absorption band shapes, fluorescence band shapes, resonance Raman intensities,
and excitation profiles using the time-dependent theory of electronic spectroscopy
J. Chem. Phys. 127, 164319 (2007); 10.1063/1.2770706
Observation of a new phosphorus-containing reactive intermediate: Electronic spectroscopy and excited-state
dynamics of the HPBr free radical
J. Chem. Phys. 123, 144304 (2005); 10.1063/1.2055227
Twisted intramolecular charge transfer states: Rotationally resolved fluorescence excitation spectra of 4, 4 ′ -
dimethylaminobenzonitrile in a molecular beam
J. Chem. Phys. 122, 084309 (2005); 10.1063/1.1850092
Dissociation dynamics of thiolactic acid at 193 nm: Detection of the nascent OH product by laser-induced
fluorescence
J. Chem. Phys. 120, 6964 (2004); 10.1063/1.1667878
Rotational coherence spectroscopy of para-cyclohexylaniline by stimulated Raman-induced fluorescence
depletion and stimulated emission pumping
J. Chem. Phys. 113, 7830 (2000); 10.1063/1.1314375
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50

Initial state resolved electronic spectroscopy of HNCO: Stimulated Raman
preparation of initial states and laser induced fluorescence detection
of photofragments
Steven S. Brown, H. Laine Berghout, and F. Fleming Crim
Department of Chemistry, University of Wisconsin-Madison, Madison, Wisconsin 53706
͑Received 22 July 1997; accepted 19 August 1997͒
Stimulated Raman excitation ͑SRE͒ efficiently prepares excited vibrational levels in the ground
electronic state of isocyanic acid, HNCO. Photofragment yield spectroscopy measures the electronic
absorption spectrum out of initially selected states by monitoring laser induced fluorescence ͑LIF͒
2
1
of either NCO (X ⌸) or NH (a ⌬) photofragments. Near threshold, the N–H bond fission is
predissociative, and there is well-resolved rotational and vibrational structure in the NCO yield
spectra that allows assignment of Ka rotational quantum numbers to previously unidentified
vibrational and rotational levels in the ␯ N–H stretch and ␯ N–C–O symmetric stretch
1
3
fundamentals in the ground electronic state of HNCO. The widths of NCO yield resonances depend
on the initial vibrational state, illustrating one way in which initial vibrational state selection
influences dissociation dynamics. Initial excitation of unperturbed ␯ ͑N–H stretch͒ states leads to
1
diffuse NCO yield spectra compared to excitation of mixed vibrational levels. The higher energy
1
dissociation channel that produces NH (a ⌬) has coarser structure near its threshold, consistent
with a more rapid dissociation, but the resonance widths still depend on the initially selected
vibrational state. © 1997 American Institute of Physics. ͓S0021-9606͑97͒01444-X͔
angle from its nearly linear ground state geometry18,19 to a
I. INTRODUCTION
bent excited state configuration,20–23 as Fig. 1 illustrates. The
S1 state has minima for both cis and trans isomers,13,20 al-
though the excitation of the trans isomer appears to dominate
the electronic excitation spectrum out of the ground vibra-
tional state in S0 .24
Double resonance excitation to a dissociative electronic
state using an excited vibrational level as the intermediate
state is a means of studying fully quantum state resolved
photodissociation.1–5 Recently, we reported the double-
resonance photodissociation of isocyanic acid,6–8 a molecule
that has three dissociation pathways that follow ultraviolet
Isocyanic acid has complex vibrational spectroscopy in
addition to rich photochemistry. Using stimulated Raman ex-
citation ͑SRE͒ ͑Ref. 25͒ as a vibrational state preparation
excitation to the S (1A ) state,7,9–12
Љ
1
1
1
NH a ⌬͒ϩCO X ⌺ϩ͒ D0ϭ42 765 cmϪ1
,
͑1͒
͑
͑
technique, we have explored the spectroscopy of the ␯ N–H
1
stretch8 and the ␯ and ␯ N–C–O symmetric and antisym-
HNCOϩh␯→H X S͒ϩNCO X ⌸͒ D ϭ38 380 cmϪ1
,
2
2
͑
͑
3
2
0
metric stretch fundamentals.26,27 In this paper, we describe
the excitation from initially selected vibrational levels to the
S1 state that dissociates to produce selected quantum states
of the fragments in dissociation channels ͑1͒ and ͑2͒.
Previously,8,26 we recorded HNCO Raman spectra by two
methods, photoacoustic Raman spectroscopy ͑PARS͒ ͑Ref.
28͒ and action spectroscopy. The latter is a photodissociation
experiment in which the Stokes laser in the SRE step scans
through rovibrational transitions at fixed photolysis and
probe laser wavelengths. Figure 1 is an energy level diagram
that shows the vibrational and electronic excitation steps, the
dissociation channels and their energies, and the laser in-
͑2͒
3
1
NH X ⌺Ϫ͒ϩCO X ⌺ϩ͒ D0ϭ30 080 cmϪ1
.
͑3͒
͑
͑
The dissociation of the N–H bond, channel ͑2͒, occurs
indirectly by a crossing from S1 to S0 , at least in the region
within a few thousand wave numbers of its threshold.7,10
There is a substantial barrier to N–H bond fission on S1 ,13
and NCO photofragment yield spectra and product state dis-
tributions from photodissociation of HNCO in a molecular
beam are consistent with barrierless dissociation on S0 .10
Channel ͑1͒ is the spin-allowed dissociation of the C–N
bond, and its mechanism has received careful scrutiny. It
appears, based on NH (a ⌬) and CO (X ⌺ϩ) rotational
distributions,11 fragment recoil anisotropies,14–16 and the ob-
servation of vibrational state-specific effects in the
photodissociation,6,8 that channel ͑1͒ occurs mainly by direct
dissociation on the S1 surface, which has a small
(ϳ500 cmϪ1) barrier to C–N bond fission.7,11,13 Channel ͑3͒
is a spin-forbidden process on S1 and, therefore, must occur
via a crossing to the triplet surface (T1).15,17 We have not
explored the triplet channel in the work described here. The
electronic transition to S1 changes the equilibrium N–C–O
1
1
2
duced fluorescence ͑LIF͒ interrogation of NCO (X ⌸) and
1
NH (a ⌬).
Figure 2 presents the PAR spectra of the two most in-
tense Raman transitions in HNCO, the ␯ N–H stretch and
1
the ␯ N–C–O symmetric stretch fundamentals. The ground
3
electronic state is a near prolate symmetric top with a large A
rotational constant ͑more than 30 cmϪ1͒.18,29,30 The intense
bands in the PAR spectra are QQK bands with well resolved
K ͑Fig. 2 uses the asymmetric rotor notation, Ka͒ and unre-
J. Chem. Phys. 107 (21), 1 December 1997
0021-9606/97/107(21)/8985/9/$10.00
© 1997 American Institute of Physics
8985
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50

8986
Brown, Berghout, and Crim: Initial state resolved spectroscopy
FIG. 1. Energy level diagram showing the Raman and electronic excitation,
photodissociation pathways and LIF detection schemes.
FIG. 2. Photoacoustic Raman spectra ͑PARS͒ of the ␯ N–H stretch ͑a͒ and
1
␯
N–C–O symmetric stretch ͑b͒ fundamental transitions in HNCO. The
3
assignments for the Ka quantum number appear over each Q branch. In the
spectrum the assignments for perturbed vibrational states also appear.
solved J structure. The PAR spectra are highly irregular be-
cause of numerous perturbations of the vibrationally excited
states. The N–C–O symmetric stretch mixes with several
combination states containing two quanta of bend.26 There
␯
1
are a number of combination bands that perturb the ␯ spec-
tors in the electronic transition to the bent N–C–O geometry
in S1 . Here we assign rotational quantum numbers (Ka) to
the previously unobserved bands in the ␯ spectrum, ␯␣ , ␯
1
trum, each consisting of a quantum of either ␯ or ␯3 , the
2
N–C–O antisymmetric or symmetric stretch, and several
1
␤
quanta of bending excitation ͑␯4 , ␯ or ␯6 , the HNC bend,
and ␯␥ , and to the bands in the ␯ region of the Raman
5
3
the NCO bend and the out-of-plane bend, respectively͒.8 The
perturbations shift the energy origins of the Ka sublevels
spectrum by observation of the structure in the electronic
transitions out of states prepared by SRE. Scanning the pho-
tolysis laser at fixed Stokes and probe laser wavelengths ͑see
Fig. 1͒ produces a photofragment yield spectrum. In addition
to providing rotational assignments, photofragment yield
spectroscopy measures resonance widths in the transition to
the dissociative electronic state. Because the initial Raman
state selection prepares a well-defined rovibrational level, it
removes thermal congestion in the electronic excitation step
and provides greater resolution than the room temperature
ultraviolet absorption spectrum of HNCO reported by Dixon
and Kirby.20 Recent photofragment yield spectroscopy by
unpredictably and alter the effective b-axis rotational con-
¯
stant Bϭ(BϩC)/2 in the vibrationally excited states, mak-
͓
͔
ing the shading of the unresolved J lines in the Raman
Q-branch profiles irregular as well.
Many of the bands in the ␯ and ␯ PARS are previously
1
3
unobserved and, hence, unassigned. We have used a simple
rotational constant analysis based on the degree and the sign
of the Q-branch shading to suggest which vibrational states
are likely perturbers.8 For example, in the ␯ spectrum, the
1
states labeled ␯ and ␯ have the form (001
6) where
ϩ 6ϭ3. The state ␯ may have the same form or
v
v4v5v
␣
␤
Reisler and co-workers10,15,17 of NH (a ⌬), NCO (X ⌸),
1
2
ϩ
v4 v5
␥
3
may be made up of five quanta of bend, and the ␯ state,
and NH (X ⌺Ϫ) from supersonically cooled HNCO also
x
which splits the ␯ Kaϭ2 band into a pair of transitions
removes thermal congestion in the S1←S0 transition. The
molecular beam experiment further has the potential to pro-
vide vibrational assignments24 since the photolysis laser may
scan over an arbitrary frequency range without background
from photolysis of thermally populated states. Despite its
more limited frequency range, the SRE initial state prepara-
1
whose infrared absorptions are well characterized, is most
likely ͑010110͒.29,31 Although some of these previously un-
observed bands are quite weak in the photoacoustic spec-
trum, their intensities in the action spectra are comparable to
the other bands because they have good Franck–Condon fac-
J. Chem. Phys., Vol. 107, No. 21, 1 December 1997
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50

Brown, Berghout, and Crim: Initial state resolved spectroscopy
8987
FIG. 3. NCO (X 2⌸) photofragment yield spectra out of each of the Ka sublevels in the ␯ manifold. The vertically plotted ␯ PARS illustrates the initial state
1
1
selection.
tion experiments described in this work have the advantage
of exciting a variety of different initial rovibrational states, in
particular the widely separated Ka energy levels in HNCO.
rated KOCN solution with concentrated phosphoric acid,35 as
described previously.36 The HNCO vapor flows from a
Ϫ78 °C reservoir of the liquid ͓vapor pressure on the order
of 1 Torr ͑Ref. 37͔͒ through the fluorescence cell, where its
temperature and pressure are approximately 25 °C and 0.1
Torr.
II. EXPERIMENTAL APPROACH
We have detailed the experimental approach
elsewhere7,8,27 and provide only a brief description here. The
experiment requires four laser light sources: two to carry out
the Raman excitation, one to photolyze vibrationally excited
states prepared by SRE, and one to probe fragments by LIF.
The fixed frequency pump light for the SRE is a fraction of
the second harmonic of an injection seeded Nd:YAG laser,
and the tunable Stokes light comes from a dye laser pumped
with the remaining light from the same Nd:YAG laser. The
photolysis light is the frequency doubled output of another
Nd:YAG laser pumped dye laser system. The LIF probe laser
is a XeCl excimer laser pumped dye laser. For probing NCO
III. RESULTS AND DISCUSSION
A. The H؉NCO „X 2⌸… channel
Figure 3 shows the NCO ͑000͒ photofragment yield
curves from several different Ka sublevels in the ␯1 , N–H
stretch state. On the right side, plotted vertically, is the pho-
toacoustic Raman spectrum, and on the left side, plotted
horizontally, are the photolysis yield spectra for the initial
states prepared by the Raman excitation. The spectral range
includes the approximately 3000 cmϪ1 between the appear-
ance threshold for the NCO fragment from photolysis of the
32,33
2
2
by LIF on the (A ⌺ϩ←X ⌸) transition
we use the
fundamental of the dye laser directly at wavelengths between
␯ states and the onset of background signal due to photoly-
1
1
1
420–480 nm, and for probing the NH (c ⌸←a ⌬)
transition34 we frequency double the dye laser output to pro-
duce the necessary 324–330 nm light. We align all four
beams on the axis of a cylindrical, Pyrex cell with quartz
windows mounted at Brewster’s angle. The cell is equipped
with a microphone for photoacoustic measurements and a
quartz viewing window with an f/1 optical system and a
photomultiplier tube ͑PMT͒ for collecting fluorescence sig-
nals. We synthesize isocyanic acid by the reaction of a satu-
sis of thermally populated states. Any vertical cut through
the photolysis yield curves in Fig. 3 gives an action spectrum
at a specific photolysis wavelength. The spectra of electronic
transitions out of the vibrational states in the ␯ manifold are
1
highly structured, indicating a long-lived electronically ex-
cited state, and the structure varies considerably because of
rotational resonances for different Ka states and different
Franck–Condon factors for the distinct vibrational states.
The presence of well-defined structure with resolved rota-
J. Chem. Phys., Vol. 107, No. 21, 1 December 1997
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50

8988
Brown, Berghout, and Crim: Initial state resolved spectroscopy
laser is fixed to interrogate a different NCO vibrational level,
in each case using the Q1 bandhead of the lower spin-orbit
state of the unique Kϭ␯2ϩ1 state.33,39 All four traces have
an identical set of resonances, with the most important dif-
ference among spectra probing higher NCO vibrational states
being the energy shift of the appearance threshold. The dis-
sociation from a bent NCO group in the S1 state of HNCO to
a linear NCO fragment preferentially populates excited bend-
ing states in NCO,40,41 ͑where ␯ is the degenerate bend͒ as
2
shown by the relatively poor signal to noise ratio obtained
probing NCO ͑100͒ ͑where ␯ is the symmetric stretch͒.
1
2. Vibrational progressions in the S1—S0 transition
Dixon and Kirby20 observed several vibrational progres-
sions in the ultraviolet absorption spectrum of HNCO vapor
at room temperature, but they did not assign the transitions to
specific vibrational modes in the upper state. Their progres-
sions, labeled A, B, C, and D, all had a characteristic spacing
of approximately 500–600 cmϪ1, consistent with excitation
of a bending vibration in S1 . The most likely candidate is the
NCO bend ͑␯ in S0͒, whose experimentally measured fre-
5
quency in S0 is 577.4 cmϪ1 ͑Ref. 42͒ and whose calculated
frequency in S1 is very nearly the same.23,27 Dixon and Kirby
suggested that the presence of several similarly spaced pro-
gressions could be the result of transitions to both cis and
trans isomers in S1 . Ab initio calculations of the S1 surface
do indeed show the presence of cis and trans minima and a
substantial barrier to isomerization. The slightly bent NCO
angle ͑172°͒ ͑Ref. 18͒ in S0 places the Franck–Condon point
closer to the trans isomer than to the cis.13 Recent electronic
absorption spectra of HNCO cooled in a supersonic jet show
that there is only a single important progression with a spac-
ing consistent with the NCO bending frequency.10,24 The
electronic absorption from vibrationally excited levels in S0 ,
particularly those that contain some NCO bend character,
may access either well on the S1 surface and may thus give
rise to additional bands.
FIG. 4. NCO (X 2⌸) photofragment yield spectra originating in the
(␯1 /␯x)l Kaϭ2 level in HNCO and probing different vibrational levels in
the NCO fragment. The abscissa is the total photolysis energy, or the sum of
the photolysis photon energy and the Raman shift for the initial vibrational
excitation. The ␯ PARS plotted beneath the graph shows the initial state
1
selection.
tional transitions is the key to the assignment of Ka values to
the previously unobserved vibrationally excited states, ␯␣ ,
␯␤ , and ␯␥ .
Our initial state-selected NCO photofragment yield spec-
tra resolve rotational and vibrational progressions more
cleanly than the room temperature absorption spectrum but
still show numerous anomalies. The same type of vibrational
state mixing that is prevalent in the ground electronic state of
HNCO ͑Refs. 29, 43͒ probably contributes to the vibrational
energy level structure in S1 , giving rise to the irregular vi-
brational progressions that we observe. In addition, although
our Raman excitation prepares a single initial vibrational
state and Ka rotational level, it populates a range of J states
because we use Q-branch excitation. The electronic excita-
tion spectra therefore have some rotational congestion. As
Table I illustrates, we find two dominant progressions that
are consistent with the C and D progressions of Dixon and
Kirby. ͑The A and B progressions lie primarily below the
NCO production threshold.͒ We are also able to resolve
many other transitions that do not fall into a regular progres-
sion. Table II lists rotational assignments ͑described below͒
for the observed bands that correspond most nearly to the C
and D bands since the vibrational spacing in these progres-
1. Final state of NCO fragment
The NCO photofragment yield spectra in Fig. 3 have the
LIF probe laser fixed on the Q1 bandhead corresponding to
the lower (2⌸1/2) spin–orbit level of the vibrationless elec-
tronic ground state in the A ⌺ϩ(000)←X ⌸(000)
transition.32,33 The experiment cleanly resolves the final vi-
bronic NCO state but not the final rotational state because it
probes a bandhead that includes approximately the lowest
fifteen J states under the bandwidth of the probe laser.38
Because of its complex vibronic energy level structure,38,39
the NCO fragment presents many states to interrogate. Com-
bined with the choice of the initial HNCO vibrational state,
the choice of the final NCO state creates a matrix of possible
state-to-state photofragment yield spectra. As Fig. 4 shows,
however, the final NCO state has little influence on the reso-
nances in the NCO yield spectrum. The initial state in all
traces of Fig. 4 is the (␯1 /␯x)l Kaϭ2 level ͑as indicated by
the PAR spectrum at the bottom of the figure͒, but the probe
2
2
J. Chem. Phys., Vol. 107, No. 21, 1 December 1997
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50

Brown, Berghout, and Crim: Initial state resolved spectroscopy
TABLE I. Vibrational progression in NCO photolysis yield spectra.
8989
C Progression
Dixon and Kirby This work
␯ (cmϪ1
D Progression
Dixon and Kirby This work
␯ (cmϪ1
E Progression
This work
␯ (cmϪ1
)
⌬␯
)
⌬␯
␯ (cmϪ1
)
⌬␯
)
⌬␯
␯ (cmϪ1
)
⌬␯
38 174a
37 938
38 312.8
556
549
546
541
539
543
535
535
542
531
538.1
536.1
523.2
524
38 730
39 279
39 825
40 366
40 905
38 699.7
39 238.5
39 784.8
40 334
38 481
39 016
39 551
40 093
40 629
38 515.4
39 059.0
39 589.3
40 124.3
38 850.9
39 387.0
39 910.2
40 434
538.8
546.3
549
543.6
530.3
535.0
aMeasurement from RQ0 in the UV spectrum ͑Ref. 20͒. All other measurements in Dixon and Kirby’s work are
band maxima. All measurements from this work are RQ0 .
sions is relatively regular and approximately matches that of
Dixon and Kirby.20 There is one additional set of bands with
a regular spacing consistent with NCO bend excitation that
we arbitrarily label as E in Tables I and II. The energy range
with prolate top selection rules of ⌬KϭϮ1, ⌬Jϭ0, Ϯ1, and
the bent N–C–O angle in S1 leads to a large decrease in the
A rotational constant, ⌬AϭϪ26 cmϪ1 20 The prolate top ro-
.
44
2
¯
tational energy levels are E(J,K)ϭAeff K ϩBJ(Jϩ1),
over which we may scan the NCO photofragment yield spec- where A ϭAϪ¯B and ¯Bϭ(BϩC)/2, and the large ⌬A sepa-
eff
tra is limited to the approximately 3000 cmϪ1 between the
rates rotational transitions for different initial values of Ka
NCO appearance threshold and the onset of significant back-
ground signal due to photolysis of thermally populated
states. This limited range precludes the assignment of bands
to excitation of specific vibrational levels in the upper state.
by up to several hundred wave numbers. Table III lists the
anticipated line positions and splittings for P,RQK ͑prolate
top notation ⌬K⌬JK͒ branch transitions relative to RQ0 based
on the Ka rotational energy levels for S0 ͑Refs. 18, 29, 30͒
and the best fit values of Aeff in S1 from this work. In NCO
3. Initial rotations
yield spectra originating in Kaϭ0, there is a single Q branch
(RQ0) for each vibrational transition, while for KaϾ0 the Q
The rotational structure within each vibrational band re-
sults from the initial Ka quantum number selected in the SRE
branches occur in pairs (RQK ,PQK) that are progressively
step. The electronic absorption is a perpendicular transition
more red-shifted and have larger splitting with increasing
TABLE II. Line positions (cmϪ1) for KaЉϭ0–4 in the NCO photolysis yield spectrum.
Band
RQ0
PQ1
RQ1
PQ2
RQ2
PQ3
RQ3
PQ4
RQ4
E
38 312.8 38 281.9 38 296.8
D
C
C
C
E
D
C
C
C
E
D
C
E
D
C
E
515.4
699.7
723.7
763.3
850.9
481.8
495.6
682.5
704.8
745.1
832.5
397.7
427.0
616.1
635.0
678.6
761.4
970.1
267.1
451.7 obscured
474.2 obscured
515.2
602.0
808.9
311.1
38 092.2
281.5
300.3
339.3 obscured
425.3 obscured
635.9
151.2
339.3
353.9
665.9
691.3
730.2
818.0
583.5a
604.9
645.5
732.9
940.4
567.3
646.5
854.8
39 059.0 39 026.0 39 040.2
696.6
199.8
238.5
262.3
387.0
589.3
784.8
910.2
40 124.3
334
39 082.2 obscured
121.8 39 152.8
145.2 obscured
270.6
466.3
670.5
794.5
40 008.5 40 039.1
208
316
818.1
881.1
302.6
498.8
701.6
824.8
242
354
434
KaЉϭ1
KaЉϭ2
KaЉϭ3
KaЉϭ4
Avg P–R splitting
Average AeЈff
14.2Ϯ1.1
3.6Ϯ0.3
30.4Ϯ1.7
3.8Ϯ0.2
45.1Ϯ1.1
3.8Ϯ0.1
58.5Ϯ2.9
3.7Ϯ0.2
aIndicates weak transition.
J. Chem. Phys., Vol. 107, No. 21, 1 December 1997
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50

8990
Brown, Berghout, and Crim: Initial state resolved spectroscopy
TABLE III. Calculated line positions and splittings (cmϪ1) for K Ͼ0.
Љ
number to several previously unidentified bands in the Ra-
man spectra of both the N–H stretch (␯1) ͑Ref. 8͒ and the
N–C–O symmetric stretch (␯3) ͑Ref. 26͒ regions. The char-
acteristic splitting of bands in the electronic transition results
directly from the initially selected Ka state, as shown in
Table III.
Position relative to RQ0
PQK-RQK
K
E(K )a
E(K )b
PQK
RQK
Splittingc
Љ
Љ
Ј
1
30.1
118.4
260.6
452.3
3.7
14.8
33.3
59.2
Ϫ33.8
Ϫ118.4
Ϫ249.5
Ϫ422.7
Ϫ19.0
Ϫ88.8
Ϫ205.1
Ϫ363.5
14.8
29.6
44.4
59.2
2
3
4
The comb at the top of Fig. 5 marks our identification of
the RQ0 line positions belonging to Dixon and Kirby’s C and
D progressions. Our line positions differ from theirs because
they identified overall vibrational band maxima without ro-
tational resolution in this region of the spectrum, whereas the
aE(K ) from infrared measurements, Ref. 30.
Љ
bE(K )ϭAeffK ; Aeffϭ(AϪB); Bϭ(BϩC)/2.
2
¯ ¯
Ј
c
␯ Ϫ␯ ϭ4K AЈ .
Љ
R
P
eff
R
double resonance experiment isolates Q0 more clearly ͑see
Table I͒. There are numerous transitions in Fig. 5 to vibra-
tional states not belonging to the C or D progressions, and
even the bands in the C progression itself appear to be split
into a series of three distinct transitions, as Table II illus-
trates. The most complete set of data in Table II is for lines
that originate in Kaϭ0 and 2. As described further below,
Ka . Because SRE excites a range of J states in the initial
state preparation, spectral overlap in the electronic transition
makes the Q branch transitions the most intense.
Figure 5 presents NCO ͑000͒ photolysis yield spectra
from three separate initial states in the ␯ manifold, all of
1
which have the same set of resonances when plotted against
a scale of total photolysis energy ͑Raman shift plus photoly-
sis photon energy͒. The lower trace originates in the ␯1 ,
Kaϭ0 pure stretch state, and the similarity of the electronic
because the structure in the ␯ Kaϭ1 NCO photolysis yield
1
spectrum is too diffuse to identify any lines, the Kaϭ1 data
in Table II come from initial preparation of ␯ Kaϭ1, which
3
has a lower vibrational frequency and thus a smaller range
over which we may scan the NCO yield spectrum without
significant single-photon dissociation of thermally populated
states. The Kaϭ3 and 4 data come from initial excitation of
both ␯ and ␯ states, but the rotational populations are small
spectra going through the previously unobserved states, ␯
␤
and ␯␥ , in the upper two traces allows a confident assign-
ment of rotational quantum number Kaϭ0 to these states as
well. In a similar fashion, we have assigned the Ka quantum
1
3
for these levels and the lines are more difficult to discern,
especially in the more diffuse, higher energy region of the
spectrum.
Table II also presents the average splitting between RQK
P
and QK bands and the corresponding value of AЈeff , which
directly determines the P–R splittings for the KЉaϾ0 transi-
tions ͑see Table III͒. There is considerable variation in the
P–R splittings in Table II that reflects both the difficulty in
locating the exact position of the peaks of the Q-branch tran-
sitions in our spectra and possibly also the different rota-
tional constants for different vibrational states in S1 . The
average excited state rotational constant from our data is
AeЈffϭ3.7Ϯ0.2 cmϪ1, close to Dixon and Kirby’s value of
20
AeЈffϭ4 cmϪ1
.
4. Initial vibrations
The character of the initial vibrational state affects the
structure in the transition to S1 by influencing the dissocia-
tion mechanism or dynamics and, therefore, the lifetime of
the excited electronic state. A clear example appears in the
NCO photofragment yield spectra of the ␯ Kaϭ1 and
1
Kaϭ3 states. Figure 3 shows that there is clearly resolved
rotational structure in the NCO yield spectrum of every ini-
tial rovibrational state in the ␯ manifold except for Kaϭ1.
1
The Kaϭ3 state has some structure in its NCO yield spec-
trum, but also a large component of diffuse background. As
noted above, the Kaϭ1 spectrum is so diffuse that it cannot
be used to assign a Ka quantum number to the initial state
FIG. 5. NCO (X 2⌸) photofragment yield spectra originating in three dif-
ferent vibrational states within the ␯ manifold and probing NCO in the
1
͑the rotational assignment for ␯ Kaϭ1 comes from infrared
1
͑000͒ vibrational state. The ␯ PARS plotted beneath the graph shows the
1
spectroscopy29͒. The vibrationally excited states of HNCO
three different initial states, which are the combination states, ␯ and ␯
,
␥
␤
and the N–H stretch state, ␯ Kaϭ0.
tend to be perturbed by interactions that depend quite
1
J. Chem. Phys., Vol. 107, No. 21, 1 December 1997
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50

Brown, Berghout, and Crim: Initial state resolved spectroscopy
8991
strongly on the value of Ka , but the Kaϭ1 and 3 levels are
unmixed, pure N–H stretch states.8,29 The other Ka sublevels
in the ␯ manifold are mixed by varying amounts with back-
1
ground states that contain several quanta of bending excita-
tion. The mixed N–H stretch–bending states have favorable
Franck–Condon factors to reach S1 compared to the unper-
turbed N–H stretch states, as evidenced by the poor signal to
noise ratio in the ␯ Kaϭ1 NCO photofragment yield spec-
1
trum of Fig. 3.
The overlap between vibrational states in S1 and differ-
ent initially prepared vibrational states in S0 may be the key
to understanding the influence of initial vibrational states on
the resonance widths. There appear to be two sets of reso-
nances in all of the NCO yield spectra in Fig. 3, one of which
is structured and tends to be quite strong, and a second, much
weaker, broad continuum. The perturbed ␯ states have
1
strong transitions to the structured bands but also show weak
transitions to the continuum. The unperturbed ␯ levels only
1
have good overlap with the latter. We suggest that the struc-
tured bands are transitions to vibrational levels in S1 that
have excitation in the NCO bend or other coordinates that
improve the Franck–Condon factor with the ground state,
and that the continuous absorption excites states that have
N–H stretch excitation in S1 . The latter have shorter life-
times and broad resonances. We have shown elsewhere23
that initial vibrational excitation of the N–H bond leads to
preferential Franck–Condon overlap with states in S1 that
have N–H stretch excitation. The N–H bond length and fre-
quency do not change significantly between S0 and S1 , and,
FIG. 6. NH (a 1⌬) photofragment yield spectra originating in two different
vibrational and rotational states in the ␯ manifold and probing NH on the
1
Q(2) line of the (c 1⌸←a 1⌬) transition. The ␯ PARS in the insets of
1
each graph indicate the initial state selection of either the unperturbed ␯
1
Kaϭ1 state or the perturbing ␯ Kaϭ0 state.
␤
N–H stretch excitation in S1 . The S0 and S1 surfaces do not
intersect,13 but in the N–H bond coordinate their closest ap-
proach must lie along the inner repulsive walls since the S1
surface goes through a barrier at long N–H bond lengths
while S0 does not. Vibrational states with one or more
consequently, the diagonal transitions in the ␯ vibration are
1
more than ten times stronger than the off-diagonal transi-
tions. The states that perturb the ␯ fundamental have no
quanta of ␯ may have better overlap with high vibrational
1
1
N–H stretch excitation but significant NCO stretch and bend
excitation, and they contribute more strongly to the overall
Franck–Condon factors of the mixed states. Thus, the back-
ground state character dominates the Franck–Condon factors
of the mixed states, and those states overlap primarily with
states in S1 that have no N–H stretch excitation.
levels on S0 at the inner turning point of the N–H oscillator.
Zyrianov et al.17 used similar reasoning to explain a lack of
3
structure in the NH (X ⌺Ϫ) photofragment yield spectrum
1
above the NH (a ⌬) production threshold, postulating that
the crossing from S1 to T1 that leads to dissociation in the
triplet channel is most favorable at the Franck–Condon point
on S1 . Regardless of the mechanism, the important point is
that direct vibrational excitation of the N–H dissociation co-
ordinate on S0 leads to a shorter lifetime in S1 and, thus, a
broader set of resonances in the electronic transition.
There are two likely explanations for the broad reso-
nances to S1 . Depending on the size and width of the barrier
to N–H dissociation on the excited surface, N–H stretch
excitation may increase the tunneling rate and lead to a direct
dissociation on S1 and a consequent reduction in the lifetime.
A recent calculation of the barrier height suggests that, at
least for energies near the dissociation threshold, this mecha-
nism is unlikely. The two isomers in S1 ͑cis and trans͒ have
comparable minimum energies, but the barrier to dissocia-
tion from the trans well is about 8800 cmϪ1 above the
HϩNCO asymptote, while the barrier from the cis well is
B. The NH „a 1⌬…؉CO channel
1
The photofragment yield spectra probing the NH (a ⌬)
fragment are qualitatively different from the NCO yield
spectra in that they are more diffuse and do not display well-
defined resonances that depend on the initially selected rota-
tional state. The threshold for spin-allowed cleavage of the
C–N bond lies significantly higher in energy,7,10 and above
4500 cmϪ1 13 Even for a hydrogen atom, the tunneling prob-
.
1
ability across either barrier is too small at threshold to com-
pete with internal conversion to S0 .45 However, at energies
closer to the top of the barriers, the tunneling rate may be-
come comparable to the internal conversion rate, and the
initial vibrational level may influence the competition be-
tween direct and indirect dissociation.
the NH (a ⌬)ϩCO threshold direct dissociation on the S1
surface competes with crossing to S0 .11,14,16,46 Figure 6 dis-
1
plays NH (a ⌬) photolysis yield spectra for two different
initially selected states in the ␯ manifold, ␯␤ , which has
1
mostly background vibrational state character, and the ␯
1
Kaϭ1 unperturbed N–H stretch state. The former shows
The most likely explanation for the broad resonances is
an increase in the crossing rate to S0 from states that have
coarse vibrational structure, consistent with previous obser-
vations of the features in the NH (a ⌬) photofragment yield
1
J. Chem. Phys., Vol. 107, No. 21, 1 December 1997
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50

8992
Brown, Berghout, and Crim: Initial state resolved spectroscopy
spectrum from HNCO cooled in a molecular beam.10 The
latter, however, is relatively unstructured, indicating that an
absorption to a continuous background dominates the NH
ACKNOWLEDGMENTS
We thank the Division of Chemical Sciences of the Of-
fice of Basic Energy Sciences of the Department of Energy
for supporting this work. We also thank Professors Hanna
Reisler and Reinhard Schinke for useful discussions and for
sharing their results in advance of publication.
1
(a ⌬) yield spectrum of the unperturbed N–H stretch states
in a fashion similar to the NCO yield spectra. The broaden-
1
ing of the NH (a ⌬) yield resonances, which probe disso-
ciation of the C–N bond, probably reflect a decrease in the
S1 lifetime due to dissociation of the N–H bond or crossing
to S0 that is influenced by vibrational excitation of the N–H
1
1
͑a͒ F. F. Crim, Annu. Rev. Phys. Chem. 44, 397 ͑1993͒; ͑b͒ R. L. Vander
coordinate. The NH (a ⌬) production threshold lies at ap-
Wal, J. L. Scott, F. F. Crim, K. Weide, and R. Schinke, J. Chem. Phys. 94,
3548 ͑1991͒; ͑c͒ R. L. Vander Wal, J. L. Scott, and F. F. Crim, ibid. 94,
1859 ͑1991͒.
proximately the same energy as the top of the barrier for
N–H dissociation from the cis well on S1 and about halfway
to the top of the barrier for dissociation from the trans
well.7,10,13 Therefore, it is possible that a direct dissociation
of the N–H bond competes with the direct C–N bond fission
on S1 to give the continuous absorption.
2
͑a͒ I. Bar, Y. Cohen, D. David, T. Arusi-Parpar, S. Rosenwaks, and J. J.
Valentini, J. Chem. Phys. 95, 3341 ͑1991͒; ͑b͒ D. David, A. Strugano, I.
Bar, and S. Rosenwaks, ibid. 98, 409 ͑1993͒; ͑c͒ Y. Cohen, I. Bar, and S.
Rosenwaks, ibid. 102, 3612 ͑1995͒.
3 J. Zhang, C. W. Riehn, M. Dulligan, and C. Wittig, J. Chem. Phys. 103,
6815 ͑1995͒.
4 M. Brouard and S. R. Langford, J. Chem. Phys. 106, 6354 ͑1997͒.
5 H. T. Liou, K. L. Huang, and M. C. Chen, Chem. Phys. Lett. 266, 591
͑1997͒.
6 S. S. Brown, R. B. Metz, H. L. Berghout, and F. F. Crim, J. Chem. Phys.
105, 6293 ͑1996͒.
IV. SUMMARY
7 S. S. Brown, H. L. Berghout, and F. F. Crim, J. Chem. Phys. 105, 8103
͑1996͒.
Photofragment yield spectroscopy is a state-to-state elec-
tronic absorption measurement from a well-defined, initially
selected vibrational and rotational level in the ground elec-
tronic state into a particular dissociation channel and frag-
ment quantum state. We use SRE to prepare selected Ka
sublevels in either the ␯ N–H stretch or ␯ N–C–O sym-
8 S. S. Brown, H. L. Berghout, and F. F. Crim, J. Chem. Phys. 106, 5805
͑1997͒.
9 H. Okabe, J. Chem. Phys. 53, 3507 ͑1970͒.
10 M. Zyrianov, T. Droz-Georget, A. Sanov, and H. Reisler, J. Chem. Phys.
105, 8111 ͑1996͒.
11 T. Droz-Georget, M. Zyrianov, H. Reisler, and D. W. Chandler, Chem.
Phys. Lett. 276, 316 ͑1997͒.
1
3
metric stretch fundamentals of HNCO. Both vibrational lev-
els are strongly perturbed, and each Ka sublevel has a differ-
ent vibrational character as a result of state mixing. We
detect N–H or C–N bond fissions by LIF of the NCO
12 C. R. Brazier, R. S. Ram, and P. F. Bernath, J. Mol. Spectrosc. 120, 381
͑1986͒.
13
¨
J. Klossika, H. Flothmann, C. Beck, R. Schinke, and K. Yamashita, Chem.
Phys. Lett. 276, 325 ͑1997͒.
14 M. Kawasaki, Y. Sato, K. Suto, Y. Matsumi, and S. H. S. Wison, Chem.
Phys. Lett. 251, 67 ͑1996͒.
2
1
(X ⌸) or NH (a ⌬) fragments, respectively. Near the
N–H dissociation threshold, the NCO photofragment yield
spectra show rotational and vibrational resonances that cor-
respond roughly to those established previously from the
room temperature absorption spectrum,20 although the
double resonance experiment is better able to resolve band
origins. The form of the resonances is invariant to the final
state of NCO but depends strongly on the initial state of the
HNCO molecule. Selection of the initial Ka rotational state
gives rise to a set of resonances in the electronic transition
that allow assignment of Ka to previously unidentified, per-
turbed vibrational states in both the ␯ and ␯ manifolds. In
15 T. Droz-Georget, M. Zyrianov, A. Sanov, and H. Reisler, Ber. Bunsenges.
Phys. Chem. 101, 469 ͑1997͒.
16 A. Sanov, T. Droz-Georget, M. Zyrianov, and H. Reisler, J. Chem. Phys.
106, 7013 ͑1997͒.
17 M. Zyrianov, T. Droz-Georget, and H. Reisler, J. Chem. Phys. 106, 7454
͑1997͒.
18 K. Yamada, J. Mol. Spectrosc. 79, 323 ͑1980͒.
19 A. L. L. East, C. S. Johnson, and W. D. Allen, J. Chem. Phys. 98, 1299
͑1993͒.
20 R. N. Dixon and G. H. Kirby, Trans. Faraday Soc. 64, 2002 ͑1968͒.
21 J. W. Rabalais, J. M. McDonald, V. Scherr, and S. P. McGlynn, Chem.
Rev. 71, 73 ͑1971͒.
22 W.-H. Fang, X.-Z. You, and Z. Yin, Chem. Phys. Lett. 238, 236 ͑1995͒.
23 S. S. Brown, C. M. Cheatum, D. Fitzwater, and F. F. Crim, J. Chem. Phys.
105, 10 911 ͑1996͒.
1
3
addition, the character of the initial vibrational state influ-
ences the manifold of resonances to which there is favorable
Franck–Condon overlap in the electronic excitation step.
The unperturbed, excited N–H stretch states have transitions
to a weak, continuous absorption that probably corresponds
to states with N–H stretch excitation in S1 . These states
dissociate more rapidly, most likely because of an increase in
24 H. Reisler ͑private communication͒.
25
͑a͒ F. DeMartini and J. Ducuing, Phys. Rev. Lett. 17, 117 ͑1966͒; ͑b͒ G.
O. Sitz and R. L. Farrow, J. Chem. Phys. 93, 7883 ͑1990͒; ͑c͒ D. A. V.
Kliner, D. E. Adelman, and R. N. Zare, ibid. 95, 1649 ͑1991͒; ͑d͒ K.
Ravichandran, M. Yorgancioglu, and T. R. Fletcher, ibid. 101, 3406
͑1994͒.
26 S. S. Brown, H. L. Berghout, and F. F. Crim, J. Chem. Phys. ͑in press͒.
27 S. S. L. Brown, Ph.D. thesis, University of Wisconsin-Madison, 1996.
28 J. J. Barrett, in Chemical Applications of Nonlinear Raman Spectroscopy,
edited by A. B. Harvey ͑Academic, New York, 1981͒, pp. 89–169.
29 K. M. T. Yamada, M. Winnewisser, and J. W. C. Johns, J. Mol. Spectrosc.
140, 353 ͑1990͒.
1
the coupling to S0 . The NH (a ⌬) photofragment yield
spectra display more diffuse resonances with no resolvable
rotational structure because the threshold for this channel lies
higher in energy and the dissociation mechanism is direct.
30 M. Niedenhoff, K. M. T. Yamada, and G. Winnewisser, J. Mol. Spectrosc.
176, 342 ͑1996͒.
The unperturbed ␯ levels in S0 give rise to a diffuse, un-
1
1
structured NH (a ⌬) yield spectrum, probably as a result of
31 J. H. Teles, G. Maier, B. A. J. Hess, L. J. Schaad, M. Winnewisser, and B.
P. Winnewisser, Chem. Ber. 122, 753 ͑1989͒.
the decrease in the S1 lifetime due to excitation of the N–H
dissociation coordinate.
32
͑a͒ T. R. Charlton, T. Okamura, and B. A. Thrush, Chem. Phys. Lett. 89,
J. Chem. Phys., Vol. 107, No. 21, 1 December 1997
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50

Brown, Berghout, and Crim: Initial state resolved spectroscopy
8993
98 ͑1982͒; ͑b͒ R. A. Copeland and D. R. Crosley, Can. J. Phys. 62, 1488
40 J. Zhang, M. Dulligan, and C. Wittig, J. Phys. Chem. 99, 7446 ͑1995͒.
41 S. S. Brown, H. L. Berghout, and F. F. Crim, J. Phys. Chem. 100, 7948
͑1996͒.
͑1984͒.
33 D. Patel-Misra, D. G. Sauder, and P. J. Dagdigian, J. Chem. Phys. 93,
5448 ͑1990͒.
42 D. A. Steiner, K. A. Wishah, S. R. Polo, and T. K. McCubbin, J. Mol.
Spectrosc. 76, 341 ͑1979͒.
34
͑a͒ R. S. Ram and P. F. Bernath, J. Opt. Soc. Am. B 3, 1170 ͑1986͒; ͑b͒
W. Ubachs, G. Meyer, J. J. Ter Meullen, and A. Dymanus, J. Mol. Spec-
trosc. 115, 88 ͑1986͒; ͑c͒ B. Bohn, F. Stuhl, G. Parlant, P. J. Dagdigian,
and D. R. Yarkony, J. Chem. Phys. 96, 5059 ͑1992͒.
35 R. A. Ashby and R. L. Werner, J. Mol. Spectrosc. 18, 184 ͑1965͒.
36 S. S. Brown, H. L. Berghout, and F. F. Crim, J. Chem. Phys. 102, 8440
͑1995͒.
43 B. P. Winnewisser, in Molecular Spectroscopy; Modern Research, edited
by N. K. Rao ͑Academic, Orlando, Florida, 1985͒, pp. 321–419.
44 I. N. Levine, Molecular Spectroscopy ͑Wiley, New York, 1975͒.
45 T. Baer and W. L. Hase, in Unimolecular Reaction Dynamics ͑Oxford
University Press, New York, 1996͒, pp. 264–266.
46
37 M. Linhard, Z. Anorg. Allg. Chem. 236, 200 ͑1938͒.
͑a͒ W. S. Drozdoski, A. P. Baronovski, and J. R. McDonald, Chem. Phys.
38
Lett. 64, 421 ͑1979͒; ͑b͒ T. A. Spiglanin, R. A. Perry, and D. W. Chan-
dler, J. Chem. Phys. 87, 1568 ͑1987͒; ͑c͒ T. A. Spiglanin and D. W.
Chandler, ibid. 87, 1577 ͑1987͒; ͑d͒ B. Bohn and F. Stuhl, J. Phys. Chem.
97, 4891 ͑1993͒.
͑a͒ R. N. Dixon, Philos. Trans. R. Soc. London, Ser. A 252, 165 ͑1960͒;
͑b͒ P. S. H. Bolman, J. M. Browm, A. Carrington, I. Kopp, and D. A.
Ramsay, Proc. R. Soc. London, Ser. A 343, 17 ͑1975͒.
39 J. T. Hougen, J. Chem. Phys. 36, 519 ͑1961͒.
J. Chem. Phys., Vol. 107, No. 21, 1 December 1997
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
129.174.21.5 On: Wed, 24 Dec 2014 07:00:50