Chemical dynamics of the reaction K*(5p2P) + H2→KH(v=0; J)+H: Electronic orbital alignment effects

Article abstract of DOI:10.1063/1.478579

We report results from scattering state spectroscopic studies of the excited state reaction K*(5p ²P) + H₂→KH(ν″,J″)+H. The final state resolved action spectra allow a direct measurement of essential features of the excited state potential surfaces, including regions of local maxima and minima. We observe a pronounced blue-wing-red-wing asymmetry in the reactive to nonreactive branching ratio, peaking in the neighborhood of a strong blue wing satellite. These results show that the dominant reaction pathway passes over a small activation barrier (350 ± 100 cm^-1) in Σ⁺-like orbital alignment. This result is consistent with an electron jump mechanism through a K⁺H^-H ion-pair intermediate. In contrast, approach in Π-like alignment leads predominantly to nonreactive scattering. Our results suggest that a combination of steric and energetic effects determine the major quenching pathways for alkali metal atom-H₂ systems.

Full text of DOI:10.1063/1.478579

Chemical dynamics of the reaction K * (5p 2 P )+ H 2 →KH (v=0;J)+ H : Electronic orbital
alignment effects
T.-H. Wong, P. D. Kleiber, and K.-H. Yang
Citation: The Journal of Chemical Physics 110, 6743 (1999); doi: 10.1063/1.478579
View online: http://dx.doi.org/10.1063/1.478579
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/110/14?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Angular distributions for the F + H 2 → HF + H reaction: The role of the F spin-orbit excited state and comparison
with molecular beam experiments
J. Chem. Phys. 121, 5812 (2004); 10.1063/1.1784446
Quantum and classical studies of the O ( 3 P)+ H 2 (v=0–3,j=0)→ OH + H reaction using benchmark potential
surfaces
J. Chem. Phys. 120, 4316 (2004); 10.1063/1.1642580
State-to-state dynamics of H+HD→H 2 +D at 0.5 eV: A combined theoretical and experimental study
J. Chem. Phys. 116, 4769 (2002); 10.1063/1.1461818
Reaction dynamics of Mg (4 1 S 0 ,3 1 D 2 ) with H 2 : Harpoon-type mechanism for highly excited states
J. Chem. Phys. 113, 5302 (2000); 10.1063/1.1290125
Energy-dependent cross sections and nonadiabatic reaction dynamics in F ( 2 P 3/2 , 2 P 1/2 )+n– H 2 →HF
(v,J)+ H
J. Chem. Phys. 111, 8404 (1999); 10.1063/1.480182
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:
141.212.109.170 On: Mon, 24 Nov 2014 20:31:01

JOURNAL OF CHEMICAL PHYSICS
VOLUME 110, NUMBER 14
8 APRIL 1999
2
Chemical dynamics of the reaction K*„5p P…؉H ˜KH„v؍0;J…؉H:
2
Electronic orbital alignment effects
T.-H. Wonga) and P. D. Kleiberb)
Department of Physics and Astronomy, University of Iowa, Iowa City, Iowa 52242
K.-H. Yang
Department of Physics, St. Ambrose University, Davenport, Iowa 52803
͑
Received 30 October 1998; accepted 12 January 1999͒
We report results from scattering state spectroscopic studies of the excited state reaction
2
K*(5p P)ϩH →KH(vЉ,JЉ)ϩH. The final state resolved action spectra allow a direct
2
measurement of essential features of the excited state potential surfaces, including regions of local
maxima and minima. We observe a pronounced blue-wing–red-wing asymmetry in the reactive to
nonreactive branching ratio, peaking in the neighborhood of a strong blue wing satellite. These
results show that the dominant reaction pathway passes over a small activation barrier (350
Ϫ1
ϩ
Ϯ100 cm ) in ⌺ -like orbital alignment. This result is consistent with an electron jump
ϩ
Ϫ
mechanism through a K H H ion-pair intermediate. In contrast, approach in ⌸-like alignment leads
predominantly to nonreactive scattering. Our results suggest that a combination of steric and
energetic effects determine the major quenching pathways for alkali metal atom-H systems.
2
©
1999 American Institute of Physics. ͓S0021-9606͑99͒00814-4͔
I. INTRODUCTION
excited metal atom ‘‘p’’-states with H , we have used a
2
‘
‘half-collision’’ pump-probe technique to measure the con-
The reactions of Group I and II excited metal atoms in
tinuum or ‘‘scattering state’’ absorption profiles of the M–H2
their ‘‘p’’-states with H to form MH product have been
5,10–12
2
collision complex for a number of metal atom reagents.
studied for many years.1 These interactions are useful test-
ing grounds for dynamical models that include nonadiabatic
interactions. Depending on the metal atom reagent, two dis-
tinct dynamical mechanisms have been identified. One
–4
The experimental methods have been recently reviewed.11
Essentially the pump-probe experiment involves tuning a la-
2
ser into the collision broadened line wings of the M(nЈp P
←
2
ns S) atomic resonance transition to excite the transient
1
mechanism ͓as identified in the reaction Mg*(3p P)
collision complex directly in the reactive entrance channel; a
second tunable laser then state-selectively probes the nascent
branching into distinct product channels. In the Born–
Oppenheimer, Franck–Condon approximation, the far wing
absorption occurs in a vertical transition between well-
defined molecular electronic states of the collision complex.
Because the different molecular eigenstates correspond to a
particular alignment of the M atom ‘‘p’’-orbital in the colli-
sion complex, it is possible to selectively align the reagents,
and to study the effect of this alignment on the reaction prob-
ability and product energy partitioning. Measurement of the
final state resolved absorption profiles ͑the action spectra͒
gives information about the shape of the excited state poten-
tial energy surfaces ͑including local extrema͒, and about the
dynamical evolution in the excited state from the Condon
point of absorption to the final product state, including in-
sight into both the nuclear motion dynamics and the elec-
tronic nonadiabatic interactions which couple the Born–
Oppenheimer adiabatic potential surfaces.
ϩH →MgHϩH͔ involves a surface crossing mechanism in
2
near C geometry between the excited MH (1B ) surface
2
V
2
2
and the ground state 1A1 surface.1 The approach to the
–6
surface crossing region follows an insertive bond-stretch pro-
cess where the good molecular orbital overlap in the 1B2
state between the metal atom ‘‘p’’-orbital and the
␴
*-antibonding orbital of H allows for efficient transfer of
2
electron density, weakening and stretching the H–H bond. In
the resulting H–M–H triangular insertion configuration, the
ground state (1A ) surface crosses the excited (1B ) surface
1
2
in a C2V symmetry-allowed crossing seam, which opens a
path for efficient quenching to the ground state surface with
no significant activation barrier.
Another dynamical pathway for reaction ͓as identified in
2
the case Cs*(7p P)ϩH →CsHϩH͔ is thought to involve an
2
abstractive long-range electron-jump in end-on attack geom-
etry via the repulsive MH (np ⌺ ) surface. The reaction
proceeds through a crossing of the neutral M*(np P)ϩH
potential surface with the ionic M H H surface, which goes
smoothly to the MHϩH product valley since the ground state
of MH is essentially ionic.8
2
ϩ
7–9
2
2
2
ϩ
Ϫ
2
In the case of the reactive quenching of Na*(4p P) by
H , our scattering state spectroscopy experiments have
2
,9
In order to further explore these differing interactions of
shown the presence of two distinct dynamical pathways, one
through the attractive ‘‘⌸-like’’ surface and the other on the
ϩ
15
repulsive ‘‘⌺ -like’’ surface, in agreement with the ‘‘in-
a͒Present address: Department of Chemistry, Columbia University, New
York, NY 10027.
Electronic mail: paul-Kleiber@uiowa.edu
sertion’’ and ‘‘abstraction’’ mechanisms suggested
b͒
10,11,13
above.
Our experimental work showed that these dis-
0021-9606/99/110(14)/6743/6/$15.00
6743
© 1999 American Institute of Physics
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:
41.212.109.170 On: Mon, 24 Nov 2014 20:31:01
1

6744
J. Chem. Phys., Vol. 110, No. 14, 8 April 1999
Wong, Kleiber, and Yang
tinct reaction pathways could be selectively excited by con-
trolling the electronic orbital alignment of the Na*(4p) or-
bital in the transient reaction complex.10
In order to further clarify the energetic and dynamical
effects that control the competitive branching into different
reactive and nonreactive quenching pathways we report here
the results from spectroscopic studies of the excited state
reaction
getically accessible. However, the KH rotational quantum
state distribution is narrowly peaked at low JЉ(JЉϭ6), with
a ͑full width at half maximum͒ range from ϳJЉϭ3 to JЉ
ϭ12. As a result we are unable to probe much higher rota-
tional quantum states. We do have limited data for KH (vЉ
ϭ0, JЉϭ3) as well; these results are similar to those pre-
sented here for JЉϭ6 and 10, but with poorer signal-to-noise
owing to the lower initial state population.
The LIF was collected at right angles to the laser axis
and focused onto the entrance slit of monochromator. The
photomultiplier output signal was collected with a gated in-
tegrator interfaced to a personal computer for data acquisi-
tion and analysis. A series of experimental linearity tests
were performed to verify that the observed LIF signals were
2
K*͑5p P͒ϩH →KH͑vЉ,JЉ͒ϩH.
͑1͒
2
The kinetics for this reaction have been previously studied.9
Our half-collision approach to the study of reaction ͑1͒ can
be schematically represented
2
2
K͑4s S͒ϩH ϩh␯→͓K͑5p͒H ͔*→K*͑5p P͒ϩH
2
2
2
͑
linear in incident laser power, H reagent gas pressure, and K
2
2a͒
atom density, over the range of operating conditions for
these experiments.
or →K͑nl͒ϩH†2
͑2b͒
͑2c͒
We have used two separate methods for observing the
nonreactive collisional channels in this work. First we mea-
sured the direct atomic fluorescence from the laser-excited
or →KH͑vЉ,JЉ͒ϩH.
Channel ͑2a͒ corresponds to the usual quasielastic collisional
absorption process. Channels ͑2b͒ and ͑2c͒ correspond to the
half-collision analog of the nonreactive (E→T,V,R) and re-
active collisional quenching processes, respectively.
2
K*(5p P) state to the ground state. Narrow spectrometer
slits were used to spectrally resolve the collision-induced
atomic fluorescence signal from Rayleigh-scattered light due
to the off-resonant pump laser. The resulting signals corre-
spond to a measurement of the quasielastic scattering chan-
nel ͑2a͒. To gain insight into the nonreactive quenching pro-
cess ͑2b͒ we also used the probe laser to detect population in
Our experimental results show that the dominant reac-
tion pathway involves passing over a small barrier ͑ϳ350
Ϫ1
ϩ
cm ͒ in ⌺ -like alignment, indicative of an abstractive
ϩ
Ϫ
electron jump mechanism through a K H H ion pair
intermediate. In contrast, approach in ⌸-like alignment
leads predominantly to nonreactive scattering channels.
These results are consistent with a suggestion that a combi-
nation of steric and energetic factors determine the major
9
2
the K*(4p P) state by laser-induced fluorescence on the
2
2
K(9d D←4p P) atomic transition. Since the two-step cas-
cade fluorescence process 5p→3d/5s→4p is unlikely to be
important for the short pump-probe delay time conditions
used in these experiments ͑because of the small 5p
8
quenching pathways for alkali metal atom–H systems.
2
→
3d/5s radiative decay rates͒, we believe that 4p state is
predominantly populated through collisional relaxation via
II. EXPERIMENTAL ARRANGEMENT
the 3d intermediate states ͑vide infra͒.
The experimental methods have been previously de-
scribed and only a brief review will be given here.5,10,11
A
III. EXPERIMENTAL RESULTS
frequency doubled and tripled 30 Hz Nd:YAG laser was
used to pump two tunable pulsed dye lasers simultaneously.
The pump dye laser was operated with DCM laser dye and
sum frequency mixed in a nonlinear mixing crystal to the
The main experimental results from this work are pre-
sented in Fig. 1, which shows the final state resolved far
wing absorption spectra ͑i.e., the action spectra͒, correspond-
ing to branching into the different product channels ͑2͒ as a
function of laser detuning (⌬ϭ␻ Ϫ␻ ) in the red and blue
2
2
spectral region near the K(5p P←4s S) second resonance
transition at 404.4 nm. The probe dye laser was operated
with Coumarin 480 laser dye and tuned into the region of the
L
0
2
2
wings of the K(5p P←4s S) second resonance transition.
Figure 1͑a͒ shows relative population in the nonreactive
product channels ͓corresponding to processes ͑2a͒ and ͑2b͔͒,
measured by direct fluorescence from the laser excited
K*(5p) state, or by probe LIF from the lower lying K
atomic 4p state as noted above. Figure ͑1b͒ gives the relative
populations in the reactive KH (vЉϭ0, JЉϭ6 and 10͒ reac-
tive product channels ͑2c͒ measured by probe LIF as dis-
1
ϩ
1
ϩ
KH(A ⌺ ←X ⌺ ) band near 490 nm. The probe dye laser
was optically delayed with a delay time of 6 ns. Typical
pulse energies were 300 ␮J each for the pump and probe dye
lasers. The pump laser power varies appreciably over the
spectral range and the observed signals are normalized to
constant laser power. The two beams were softly focused to
a spatial overlap at the center of a heat pipe oven. The oven
was operated at a temperature of 500 K, corresponding to a
2
cussed. In each case the signals are multiplied by ⌬ in order
13
3
K atom vapor density of ϳ2ϫ10 /cm . The oven is filled
to expand the scale and visually enhance the weak far wing
signals which arise from the strongest ͑i.e., smallest impact
parameter͒ collisions.
with H gas a typical operating pressure of 8 Torr.
2
Branching into the reactive product channels KH (vЉ
ϭ0, JЉϭ6 and 10͒, corresponding to ͑2c͒ above, were mea-
Pronounced satellites are obvious in the far wing absorp-
tion spectra in both the red and blue wings of Fig. 1. Strong
satellites are apparent in both the reactive and nonreactive
1
ϩ
sured by laser-induced fluorescence ͑LIF͒ on the A ⌺
1
ϩ
←
X ⌺ ͑7,0͒ band and detected on the corresponding ͑7,1͒
Ϫ1
band. The reaction exoergicity from the K*(5p) state is
action spectra at ⌬ϳϩ250 cm in the blue wing and ⌬
Ϫ1
Ϫ1
ϳ2900 cm , so that rotational states up to JЉϳ30 are ener-
ϳϪ400 cm in the red wing. There is also a weak satellite
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:
141.212.109.170 On: Mon, 24 Nov 2014 20:31:01

J. Chem. Phys., Vol. 110, No. 14, 8 April 1999
Wong, Kleiber, and Yang
6745
FIG. 2. Reactive to nonreactive branching ratio. The ratio of the KH (vЉ
ϭ0, JЉϭ6 and 10͒ LIF signals to the K*(5p) direct collisional fluorescence
signal, as a function of laser detuning (⌬ϭ␻ Ϫ␻ ) in the red and blue
L
0
wings of the K(4s–5p) resonance line.
the reactive-to-nonreactive branching ratio of Fig. 2. The re-
active branching increases rapidly with blue wing detuning
and peaks in the region of the blue wing satellite near ⌬ϳ
Ϫ1
ϩ250 cm . Beyond the satellite the total absorption drops
sharply. In the red wing, the reactive branching peaks in the
Ϫ1
region of the near red wing satellite at ⌬ϳϪ100 cm , and
then drops rapidly. Despite a strong overall absorption prob-
ability in the far red wing, no reaction product is observed
Ϫ1
for ⌬ϽϪ150 cm . We discuss the dynamical implications
of these observations in the next section.
FIG. 1. ͑a͒ Nonreactive action spectrum: Log–log plot of the K*(5p) direct
Consideration of the relative branching into the different
nonreactive product channels ͓͑2a͒ and ͑2b͔͒ is also informa-
tive, giving insight into the physical quenching pathways for
2
fluorescence ͑᭹͒ and K*(4p) LIF ͑᭝͒, each multiplied by ⌬ . ͑b͒ Reactive
2
action spectrum: Log–log plot of the LIF signals ͑multiplied by ⌬ ) from
KH (vЉϭ0, JЉϭ6) ͑᭹͒ and KH (vЉϭ0, JЉϭ10) ͑᭝͒.
the K–H system. Figure 3 presents the relative ratio of the
2
probe LIF signal from the K*(4p) state to the direct colli-
sional fluorescence signal from the K*(5p) state, as a func-
tion of pump laser detuning in the red and blue wings. If the
K*(4p) state were populated primarily through a multistep
process involving quasielastic scattering to the 5p state
feature at ⌬ϳϪ100 cmϪ1 in the near red wing of the reac-
tive action spectrum of Fig. 1͑b͒ ͕although it is not present in
the corresponding nonreactive profile ͓Figure 1͑a͔͖͒. These
satellites are analogous to rainbow scattering maxima and
result from Franck–Condon excitation into regions of local
extrema on the excited state/ground state difference potential
͓
͑
channel ͑2a͔͒, followed by radiative or collisional relaxation
through subsequent collisions͒ via the intermediate 5s or 3d
states to the 4p state, then the different nonreactive absorp-
tion profiles of Fig. 1͑a͒ would be similar and the ratio in
1
1
energy surfaces. We discuss the spectral assignment for the
observed satellite features below.
To clarify the chemical dynamics in the K*(5p)ϩH2
collision system, it is useful to consider the relative branch-
ing into different quenching channels as a function of detun-
ing in the far line wings. Information about reactive-to-
nonreactive product branching is given in Fig. 2 where we
have graphed the relative ratio of the observed reactive prod-
2
uct KH (vЉϭ0, JЉϭ6 and 10͒ LIF signals to the K*(5p P)
direct fluorescence signal ͑corresponding to the quasielastic
scattering channel͒ in the far-red and blue wings. This is, of
course, not the total reactive-to-nonreactive branching since
we do not measure all exit channels. Furthermore, due to
radiative trapping effects in the direct fluorescence measure-
ment, it is difficult to give a meaningful comparison of the
absolute signal levels; thus, the ratio in Fig. 2 is given only
as a relative value. Nevertheless, these data do give a clear
indication of the reaction pathway.
FIG. 3. The ratio of the K*(4p) laser induced fluorescence signal to the
K*(5p) direct collisional fluorescence signal, as a function of laser detuning
(⌬ϭ␻ Ϫ␻ ) in the red and blue wings of the K(4s–5p) resonance line.
Note the very strong red-wing–blue-wing asymmetry in
L
0
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:
41.212.109.170 On: Mon, 24 Nov 2014 20:31:01
1

6746
J. Chem. Phys., Vol. 110, No. 14, 8 April 1999
Wong, Kleiber, and Yang
2
2
ϩ
The 5p( A ) ⌺ -like surface shows an obvious local
1
maximum in the potential surface near ϳ12–13a . The lo-
0
cation is determined in large part by the interaction between
ϩ
Ϫ
2
the K*(5p)ϩH covalent curve and the K –H ion pair
2
curve which passes through this energy region, and the po-
tential surface maximum represents an activation barrier
which must be overcome to access the ion-pair surface cross-
ing region in the proposed electron-jump reaction mecha-
nism. ͓A similar barrier in the analogous Na*(4p)ϩH sys-
2
tem was also found at smaller internuclear distances and
higher energies due to the larger ionization potential of
14
2
2
ϩ
Na. ͔ Inside this long range barrier the 5p( A ) ⌺ -like
1
surface shows a deep well, and evidence for an avoided
2
2
ϩ
crossing with the lower lying surface of 3d( A ) ⌺ -like
1
character correlating with the K*(3d)ϩH asymptote.
2
The attractive ⌸-like surfaces are anisotropic because
molecular orbital overlap determines the degree of bonding
interaction. The local minimum in the 5p attractive ⌸-like
surface is considerably deeper for C2V approach symmetry
than for Cϱv symmetry, as expected for an insertive attack.
͓
Similar results were found for the Na–H interaction, al-
2
though in that case the chemical interaction was stronger due
to the smaller size of the Na(4p) atom which allows for
more efficient molecular orbital overlap with the ␴* anti-
14
bonding orbital of H . ͔
FIG. 4. Relevant potential energy curves for KH . The potentials are for
2
2
fixed H–H bond length and side-on (C ) approach geometry: solid lines
To identify the observed spectral features, we have car-
2
V
2
2
2
(
A ); short dashed lines ( B ); long dashed lines ( B ). Adapted from
ried out a simple quasistatic lineshape calculation based on
1
1
2
Ref. 14.
11,14
the Rossi and Pascale potentials of Fig. 4.
The model
absorption spectrum predicts a blue wing satellite near ⌬ϳ
Ϫ1
ϩ550 cm , correlated with Franck–Condon excitation to
the region of the ⌺ -like local surface maximum at long
Fig. 3 would be essentially flat. However, our experimental
design discriminates against such slow multistep processes
due to the short pump-probe delay. In fact the clear red-
wing–blue-wing asymmetry and structure in the ratio of Fig.
2
ϩ
range (ϳ12–13a ) ͑Fig. 4͒. We, therefore, assign the ob-
0
Ϫ1
served blue wing satellite feature near ⌬ϳϩ250 cm to
excitation into this region. The discrepancy in satellite posi-
tion shows that the pseudopotential calculations significantly
3
shows the effect of direct nonreactive collisional quench-
ing ͓channel ͑2b͔͒ in the initial far wing excitation process:
overestimate the barrier height in the KH system.
2
2
K͑4s S͒ϩH ϩh␯→͓K͑5p͒H ͔*
Analysis of the quasistatic lineshapes show that both the
-like and ⌺ -like surfaces can contribute to the total
2
2
2⌸
2
ϩ
2
2
→
K*͑3d D or 5s S͒ϩH ,
2
absorption in the red wing. The model predicts a broad con-
tinuum wing resulting from absorption to the attractive 5p
⌸-like surfaces throughout the spectral region of interest.
followed by radiative decay ͑or a second quenching colli-
sion͒ to the 4p state. In particular, note the strong resonance
peak for nonreactive quenching following excitation in the
region of the far red wing satellite near ⌬ϳϪ400 cmϪ1, cor-
responding to a region of the excited state surface which
channels flux into the nonreactive quenching channel.
2
ϩ
However, red wing excitation to the 5p ⌺ -like surface is
also possible inside the long range barrier where the
2
ϩ
5p ⌺ -like surface shows a deep well near ϳ8–9a . The
0
Ϫ1
quasistatic model predicts a satellite near ⌬ϳϪ450 cm
,
superimposed on the broad absorption continuum. The satel-
lite correlates with Franck–Condon excitation to the inner
IV. DISCUSSION
2
ϩ
well region of the 5p ⌺ -like surface. The experimental red
Ϫ1
Franck–Condon excitation into the spectral region of a
far wing satellite corresponds to pumping the transient colli-
sion complex directly into the neighborhood of a local extre-
mum on the excited state potential energy surface ͑accessing
a local activation barrier, saddle point, or well͒. This allows
a direct measure of branching into distinct product channels
from the transition state region of the excited state surface.
Classical satellites are expected based on the pseudopotential
wing action spectral satellite observed near ⌬ϳϪ400 cm
can then be assigned to Frank–Condon excitation into region
ϩ
of this potential well on the ⌺ -like potential energy surface.
However, it is important to realize the real potential en-
ergy surfaces in the region from ϳ7–9a are much more
0
complicated than is readily apparent in the restricted geom-
etry curves of Fig. 4. Note that, in C geometry ͑as in C ),
2
v
ϱv
2
2
ϩ
2
the 5p( A ) ⌺ -like surface crosses the attractive 5p( B
and B ) ⌸-like surfaces in the region near the deep mini-
1
1
2
calculations of Rossi and Pascale for the KϩH collisional
2
2
1
4
molecule. The Rossi and Pascale potential curves for C2v
mum at ϳ8a . However, in the more general ͑and more
0
2
2
geometry are reproduced in Fig. 4 for convenience.
probable͒ C approach geometry, the B and A surfaces
s 2 1
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:
141.212.109.170 On: Mon, 24 Nov 2014 20:31:01

J. Chem. Phys., Vol. 110, No. 14, 8 April 1999
Wong, Kleiber, and Yang
6747
are both of AЈ character and will show an avoided crossing
Excitation in the region of the near red wing satellite at
Ϫ1
͑
conical intersection͒ leading to an appreciable distortion of
⌬ϳϪ100 cm probably corresponds to excitation of the lo-
2
2
2
ϩ
2
the potential energy surfaces in this region. The B ( AЉ)
state does not interact and will cross the AЈ surfaces. The
cal minimum on the upper 5p ⌺ -like ( AЈ) reactive sur-
face above the avoided crossing conical intersection in CS
geometry. Chemical reaction on this surface follows directly
by an electron jump mechanism as proposed. This explains
why the KH reactive product rotational state branching for
excitation in the red wing is identical to that found through-
out the blue wing, because the reaction pathway is the same
following excitation in both wings. In fact, it is difficult to
explain this observation any other way.
1
2
avoided crossing in AЈ symmetry can result in a pair of local
extrema in the excited state surfaces ͑i.e., a local minimum in
2
2
the upper AЈ surface and local maximum in the lower AЈ
surface͒ which may be evidenced in a broad absorption sat-
ellite, or even a pair of absorption satellite features. Thus, the
Ϫ1
Ϫ1
observed satellites near ⌬ϳϪ100 cm and ⌬ϳϪ400 cm
in the red wing may be associated with excitation of a local
2
minimum on the upper AЈ surface and a local maximum on
Absorption in the region of the red wing satellite near
2
Ϫ1
the lower AЈ surface in the region of the avoided surface
⌬ϳϪ400 cm probably corresponds to direct excitation to
2
crossing in AЈ symmetry, respectively.
the neighborhood of a local maximum on the lower AЈ sur-
Despite these uncertainties in the details of the potential
energy surfaces, we can draw significant and unambiguous
conclusions about the quenching dynamics from the experi-
mental data. The strong red-wing–blue-wing asymmetry in
the reactive branching of Fig. 2 shows that the reaction oc-
face below the AЈ avoided crossing region in CS symmetry.
Excitation in this region shows a strong ‘‘resonance’’ for
scattering into the nonreactive quenching channel ͑Fig. 3͒.
Thus, the surface crossing region near the minimum of the
2
ϩ
deep 5p ⌺ -like potential well represents a ‘‘hole’’ in the
2
ϩ
ϩ
reactive 5p ⌺ -like surface that allows for escape or leak-
age of flux off the reactive surface to a direct nonadiabatic
collisional quenching channel to the lower lying (5s) or
curs exclusively in ⌺ -like alignment. In contrast, we note
that blue wing excitation to the outer wall of the repulsive
ϩ
⌺
-like state at long range results primarily in quasielastic
(
3d) states.
scattering and that the probability for direct nonreactive col-
lisional quenching from this region is small. Figures 2 and 3
may be understood in terms of reaction over the long-range
2
ϩ
V. SUMMARY
We have reported the results from scattering state spec-
barrier on the ⌺ -like potential energy surface. For smaller
detunings only the higher energy portion of the Maxwell–
Boltzmann velocity distribution has sufficient energy to clear
the top of the activation barrier and react, but as the laser is
tuned farther up the repulsive wall on the outer side of the
troscopic studies of the excited state reaction
2
K*͑5p P͒ϩH →KH͑vЉ,JЉ͒ϩH.
2
barrier, an increasing fraction of the Maxwell–Boltzmann
We observe clear satellite features in both the red and blue
spectral wings of the (5p–4s) absorption line of the KH2
collision complex. These spectral features are identified with
local extrema on the excited state potential energy surfaces;
the observed satellites are in qualitative agreement with the
distribution will surmount the barrier.1
0,11
Based on the ob-
served blue wing satellite position we estimate the activation
Ϫ1
barrier height at 350Ϯ100 cm . Note that these experimen-
tal techniques allow a direct spectroscopic observation and
measurement of the activation barrier. Reaction over an ac-
essential surface features predicted in the pseudopotential
ϩ
14
tivation barrier in ⌺ -like orbital alignment is fully consis-
calculations for KH by Rossi and Pascale. Specifically, a
2
Ϫ1
tent with the proposed electron-jump mechanism through a
strong blue wing satellite near ⌬ϳϩ250 cm is assigned to
ϩ
Ϫ
9
K H H ion pair intermediate. This reaction mechanism fa-
vors low rotational energy product states as observed. The
blue wing action spectra for branching to the different JЉ
ϭ6 and 10 ͑and to JЉϭ3, not shown͒ are identical to within
our experimental uncertainties, consistent with a single reac-
tion pathway.
excitation in the region near a local maximum at long range
ϩ
in the repulsive ⌺ -like potential energy surface. This po-
Ϫ1
tential maximum represents a small ͑350 cm ͒ activation
ϩ
barrier for reaction on the ⌺ -like surface. We have also
observed satellite features in the red wing of the profile and
assigned these to excitation into the region of an avoided
ϩ
The dramatic decrease in the reactive branching at larger
surface crossing ͑inside the long range barrier on the ⌺ -like
Ϫ1
ϩ
red wing detunings (⌬ϽϪ100 cm ) shows that approach in
surface͒ between the ⌺ -like and ⌸-like surfaces in AЈ sym-
⌸-like electronic orbital alignment does not lead to reaction
metry.
with any significant quantum yield, but rather results pre-
dominantly in nonreactive scattering. These observations are
in stark contrast with our previous work on the Na*(4p)
reaction wherein we found competitive branching into both
We observe a strong blue-wing–red-wing asymmetry in
the reactive to nonreactive branching ratio showing that the
dominant reaction pathway involves passing over the activa-
ϩ
tion barrier in ⌺ -like alignment. This result is consistent
ϩ
Ϫ
the reactive and nonreactive quenching channels following
with an electron jump mechanism through a K H H ion pair
intermediate that has been proposed for this system by Lin
far red wing absorption.1
0–12
As discussed above, the appar-
9
ent lack of any reactive quenching pathway through a bond
and co-workers. In contrast, approach in ⌸-like alignment
insertion process in the K–H case might be due to the larger
atomic size for K. Because of its larger size, there is less
efficient molecular orbital overlap with the small ␴* anti-
leads predominantly to nonreactive scattering. Experimental
evidence also suggests the surface crossing region near the
local minimum on the 5p ⌺ -like reactive surface allows a
path for nonreactive quenching to lower lying excited states
of K.
2
2
ϩ
bonding orbital of H , and a correspondingly weaker chemi-
2
cal interaction.
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:
141.212.109.170 On: Mon, 24 Nov 2014 20:31:01

6748
J. Chem. Phys., Vol. 110, No. 14, 8 April 1999
Wong, Kleiber, and Yang
6
7
These observations are compared with results from pre-
L. N. Ding, P. D. Kleiber, M. A. Young, W. C. Stwalley, and A. M.
Lyyra, J. Phys. Chem. 97, 2181 ͑1993͒.
J. M. L’Hermite, G. Rahmat, and R. Vetter, J. Chem. Phys. 93, 434
10,11
vious work on the Na*(4p)ϩH reaction.
Clear differ-
2
ences in the experimental observations for the KH and
2
͑
1990͒; F. X. Gadea, F. Spiegelmann, M. Pelissier, and J. P. Malrieu, J.
NaH systems suggest that a combination of steric and ener-
2
Chem. Phys. 84, 4872 ͑1986͒.
X. Huang, J. Zhao, G. Xing, X. Wang, and R. Bersohn, J. Chem. Phys.
104, 1338 ͑1996͒.
D. K. Liu and K. C. Lin, J. Chem. Phys. 105, 9121 ͑1996͒.
S. Bililign and P. D. Kleiber, J. Chem. Phys. 96, 213 ͑1992͒.
P. D. Kleiber, in Chemical Dynamics and Kinetics of Small Radicals,
edited by K. Liu and A. Wagner ͑World Scientific, Singapore, 1996͒, p.
getic effects determine the major quenching pathways for
8
8
alkali metal atom–H systems.
2
9
0
1
1
1
ACKNOWLEDGMENTS
We gratefully acknowledge the support of the National
Science Foundation for this work. This manuscript is dedi-
cated to the memory of our colleague and friend, Kenneth M.
Sando.
573.
12
P. D. Kleiber, T. H. Wong, and S. Bililign, J. Chem. Phys. 98, 1101
͑1993͒.
1
1
1
3
4
5
A. Sevin and P. Chaquin, Chem. Phys. 93, 49 ͑1985͒.
F. Rossi and J. Pascale, Phys. Rev. A 32, 2657 ͑1985͒.
At large M–H distances the internal structure of H may be ignored and
1
W. H. Breckenridge, J. Phys. Chem. 100, 14840 ͑1996͒.
W. H. Breckenridge and H. Umemoto, J. Chem. Phys. 80, 4168 ͑1984͒; N.
2
2
2
Adams, W. H. Breckenridge, and J. Simons, Chem. Phys. 56, 327 ͑1981͒.
P. Chaquin, A. Sevin, and H. Yu, J. Phys. Chem. 89, 2813 ͑1985͒.
K. C. Lin and C. T. Huang, J. Chem. Phys. 91, 5387 ͑1989͒; D. K. Liu, T.
L. Chin, and K. C. Lin, Phys. Rev. A 50, 4891 ͑1994͒.
P. D. Kleiber, A. M. Lyyra, K. M. Sando, V. Zafiropulos, and W. C.
Stwalley, J. Chem. Phys. 85, 5493 ͑1986͒.
the potential surfaces are roughly isotropic and similar to diatomic M–Ne
potential curves; for simplicity we use the diatomic state labeling,
‘‘⌺ -like’’ or ‘‘⌸-like’’ to describe the alignment of the M atomic
3
4
ϩ
p-orbital parallel or perpendicular to the M–H internuclear axis, respec-
5
2
tively.
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:
41.212.109.170 On: Mon, 24 Nov 2014 20:31:01
1