Zero kinetic energy proton and deuteron production from photoionization of H2 and D2

Article abstract of DOI:10.1063/1.470692

A zero ion kinetic energy spectrometer has been developed to study the production of near zero energy protons and deuterons from dissociative photoionization of H2 and D2.Both H⁺ and D⁺ spectra show four peaks on top of a continuum.The continuum was found to be in excellent agreement with the single center Coulomb calculation for the direct dissociation through the X ²Σ⁺_g state of H2⁺.The observed structures were shown to originate from autoionization of the doubly excited Q₁ ¹Σ⁺_g(1), Q₁ ¹Σ⁺_u(1), Q₁ ¹Σ⁺_u(2), and Q₂ ¹Σ⁺_u(1) states, of which the Q₁ ¹Σ⁺_g(1) state is dipole forbidden.

Full text of DOI:10.1063/1.470692

Zero kinetic energy proton and deuteron production from photoionization of H2 and D2
Z. X. He, J. N. Cutler, S. H. Southworth, L. R. Hughey, and J. A. R. Samson
Citation: The Journal of Chemical Physics 103, 3912 (1995); doi: 10.1063/1.470692
View online: http://dx.doi.org/10.1063/1.470692
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/103/10?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Multichannel quantum defect theory simulation of the zerokineticenergy photoelectron spectrum of H2
J. Chem. Phys. 101, 923 (1994); 10.1063/1.467746
The nonresonant twophoton zero kinetic energy photoelectron spectrum from the electronic ground state of H2S
J. Chem. Phys. 98, 3592 (1993); 10.1063/1.464036
Kinetic energy distribution of excited H atoms produced through dissociative photoionization excitation of H2
J. Chem. Phys. 92, 1713 (1990); 10.1063/1.458053
Proton production probability in the dissociative photoionization of H2
J. Chem. Phys. 81, 2652 (1984); 10.1063/1.447974
Neutron Production from Proton and Deuteron Bombardment of Nickel
Rev. Sci. Instrum. 28, 56 (1957); 10.1063/1.1715707
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:
130.113.86.233 On: Tue, 02 Dec 2014 15:11:01

Zero kinetic energy proton and deuteron production from photoionization
of H2 and D2
Z. X. He and J. N. Cutler
Department of Physics and Astronomy, University of Nebraska, Lincoln, Nebraska 68588-0111
S. H. Southworth and L. R. Hughey
National Institute of Standards and Technology, Gaithersburg, Maryland 20899
J. A. R. Samson
Department of Physics and Astronomy, University of Nebraska, Lincoln, Nebraska 68599-0111
͑Received 18 April 1995; accepted 1 June 1995͒
A zero ion kinetic energy spectrometer has been developed to study the production of near zero
energy protons and deuterons from dissociative photoionization of H2 and D2. Both Hϩ and Dϩ
spectra show four peaks on top of a continuum. The continuum was found to be in excellent
agreement with the single center Coulomb calculation for the direct dissociation through the X ⌺gϩ
2
state of Hϩ2 . The observed structures were shown to originate from autoionization of the doubly
excited Q1 1⌺gϩ(1), Q1 1⌺uϩ(1), Q1 1⌺ϩu (2), and Q2 1⌺ϩu (1) states, of which the Q1 1⌺ϩg (1)
state is dipole forbidden. © 1995 American Institute of Physics.
similar to that of Strathdee and Browning.4 More recently,
I. INTRODUCTION
with the use of synchrotron radiation Latimer et al.19 and Ito
Dissociative photoionization of simple molecules such
et al.20 have remeasured the energy distribution of the Hϩ
as H2 has been the subject of extensive studies for the past
fragments. Their results appear to be in better agreement
twenty years because of its importance in astrophysics and
with the calculated energy spectrum of Kirby et al.16,17 How-
ever, the oscillatory behavior predicted by Kirby et al. was
not observed.
fundamental physics.1–9 Under the bombardment of vacuum
uv radiation, hydrogen molecules dissociate into ionic and
͑or͒ neutral fragments with a certain amount of kinetic en-
ergy. Collisional reaction rates are often strongly dependent
upon the energy of these reactants. Thus, the local chemistry
of interstellar clouds and the atmospheres of planets can be
strongly influenced by the dissociative effects of vuv radia-
tion. The dissociation process is also of fundamental interest
in theoretical physics because H2 has the simplest molecular
structure and hence relatively simple models can be em-
ployed. Calculations of the dissociative photoionization cross
sections of H2 and D2 of the ratios Hϩ/H2ϩ and Dϩ/D2ϩ for
direct transitions into the vibrational continuum of the
In an effort to better understand the dissociative ioniza-
tion process in the Q1 and Q2 states more quantitatively it is
desirable to measure the production of fragment ions of spe-
cific energies as a function of the incident photon energy. In
the present work we have chosen a specific energy of near
zero ͑0 to 10 meV͒. There are several advantages in this
choice. One, of course, is that the collection efficiency of
such low energy ions can be ϳ102 times greater than that for
energetic ions. Second, it allows measurements of the rela-
tive cross section for dissociative ionization through transi-
X ⌺ϩg ground state ion have been made by O’Neil and
2
tions into the vibrational continuum of the X ⌺ϩg ionic state
2
Reinhardt5 and by Ford et al.6,10,11 Good agreement between
the experimental data and that of the theoretical results of
Ford et al. was found between the dissociative ionization
threshold ͑18 eV͒ and about 23 eV.10–14 However, contribu-
tions to the experimental data from dissociation of excited
ionic states and from autoionization of the Q1 and Q2 doubly
excited neutral states of H2 and D2 prevented a check on the
validity of the calculation at higher energies.
at all photon energies above the threshold. This is possible
because a direct transition into the vibrational continuum of
the Hϩ2 X ⌺ϩg state produces Hϩ fragments with a charac-
2
teristic energy distribution. This distribution starts with a
maximum at zero kinetic energy and decays exponentially to
a negligible value at about 1 eV.3–6 For incident photon en-
ergies above 19 eV the kinetic energy distribution of Hϩ ions
remains constant and independent of the incident photon en-
ergy. Thus by sampling a constant fraction of this energy
distribution we can obtain relative cross sections. By restrict-
ing the ion kinetic energy to nearly zero we can exclude the
Hϩ fragments generated from high lying ionic states because
these repulsive states are known to produce nonzero energy
species. Any zero energy ions produced by autoionization of
the Q1 and Q2 states will only produce discrete peaks super-
imposed on the continuum background. Our results of these
measurements obtained by use of a zero ion kinetic energy
͑ZIKE͒ spectrometer are reported below.
The mechanism of dissociative ionization from the dou-
bly excited Q1 and Q2 states appears to be less well under-
stood. Results from both theoretical and experimental studies
are often in contradiction with each other. Both classical and
quantum mechanical resonant models have been proposed by
Hanzi15 and Kirby et al.,16,17 respectively, which predict that
a significant amount of low energy protons are present dur-
ing dissociation, whereas the early experimental data of
Strathdee and Browning4 clearly showed otherwise. Later,
Kanfer and Shapiro18 have used a more accurate continuum
coupled-channel method, yielding a kinetic energy spectrum
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:
3912 J. Chem. Phys. 103 (10), 8 September 1995 0021-9606/95/103(10)/3912/5/$6.00 © 1995 American Institute of Physics
130.113.86.233 On: Tue, 02 Dec 2014 15:11:01

He et al.: Photoionization of H2 and D2
3913
FIG. 2. Ratios of Hϩ/H2ϩ and Dϩ/Dϩ2 as a function of photon energy. ᭡,
theoretical results of Ford et al., Ref. 10, 11;—, present data of Hϩ/H2ϩ , ••• ,
present data of Dϩ/D2ϩ . Notice that the Q1
⌺
ϩg (1) state is abbreviated as
1
FIG. 1. Schematic diagram of the zero ion kinetic energy ͑ZIKE͒ spectrom-
eter.
1
ϩ
Q1
⌺
.
g
III. RESULTS AND DISCUSSION
II. EXPERIMENT
The measured ratios of Hϩ/H2ϩ and Dϩ/D2ϩ are shown in
Fig. 2. The data represents the relative measurement of the
continuous production of near zero energy ͑0 to 10 meV͒ Hϩ
and Dϩ ions from transitions into the vibrational continuum
A schematic diagram of our ZIKE spectrometer is illus-
trated in Fig. 1. Briefly, ionization products are produced by
the interaction of vuv radiation with a volume of hydrogen
gas inside the ion chamber. The ion chamber was designed
specifically to collect low energy ions ͑0–10 meV͒ very ef-
ficiently but to discriminate strongly against the energetic
of the X ⌺ϩg ionic state and from autoionization of the Q1
2
and Q2 doubly excited neutral states. In order to compare to
the calculated results of Hϩ/H2ϩ and Dϩ/D2ϩ, the present data
was normalized at 22 eV to the previous measured absolute
values of Chung et al.12 for hydrogen and to that of Brown-
ing and Fryar13 for deuterium. The accuracy of these previ-
ous results was quoted as Ϯ͑4–5͒% and Ϯ10% for Hϩ/H2ϩ
and Dϩ/D2ϩ, respectively.12,13 Also shown in Fig. 2 are the
calculated ratios of Hϩ/H2ϩ by Ford et al.10,11 for the direct
21
species. A SIMION simulation shows that the transmission
efficiency is nearly 100% for zero energy ions but falls very
rapidly to about 1%–2% for ions with у10 meV kinetic
energy. Due to this arrangement, mostly ions with near zero
kinetic energy were extracted out of the ion chamber and
then accelerated to a kinetic energy of 5 eV into the entrance
slit of a 180° hemispherical electrostatic energy analyzer.
The pass energy of the analyzer was chosen to be approxi-
mately 5 eV such that only the zero energy ions were able to
pass through. At this pass energy, the energy resolution of the
analyzer was found to be about 75 meV. After exiting the
energy analyzer, the ions then enter a 180° magnetic mass
spectrometer with a resolving power of approximately 1 in
65 mass units. After being energy and mass selected, the ions
were detected by a channel electron multiplier.
Vacuum uv radiation from both the Aladdin storage ring
at the Synchrotron Radiation Center ͑SRC͒, University of
Wisconsin, and the SURF II storage ring at the National Insti-
tute of Standards and Technology were used during the
present experiment. Initial studies of H2 were made using a
two-meter normal incidence low resolution beamline at SURF
II. The usable photon flux of this beamline drops rapidly
above a photon energy of 25 eV. In subsequent studies, a
high resolution 4-m normal incident monochromator beam-
line at Aladdin was employed to cover a spectral region from
15 to 45 eV at a wavelength resolution of 0.35 Å. An alumi-
num photodiode was used to monitor the photon flux. We
found the effect of scattered and higher order radiation to be
less than 5% over the measured 18 to 42 eV photon energy
range.
dissociation through the Hϩ2 X ⌺ϩg state. The agreement is
2
very good except for the photon energies from 23 to 38 eV,
where our data includes the excitation of the doubly excited
Q1 and Q2 states. The apparent difference between our data
and the theory near the threshold ͑18–19 eV͒ is due to the
fact that the present experimental technique samples near
zero energy Hϩ fragments only. The actual kinetic energy
distribution of the Hϩ fragments is increasing above the
threshold and does not become constant until the photon en-
ergy exceeds 19 eV. Thus our sampling of a fixed energy
interval overestimates the ratio between 18 and 19 eV. This
overestimation can be corrected if the fragment energy dis-
tribution is known, as such is the case for Hϩ and Dϩ pro-
duced from direct dissociation of the X ⌺ϩg ionic state.3–6 It
2
should be mentioned that both Hϩ and H2ϩ cross sections for
the direct dissociation have also been calculated by O’Neil
and Reinhardt.5 Their ratios of Hϩ/Hϩ2 have been omitted in
Fig. 2 because they differ from that of Ford et al.10,11 by as
much as 50% at a photon energy of 35 eV. Nevertheless,
there is a good agreement with our results from the dissocia-
tive ionization threshold up to about 22 eV.
In Figs. 3 and 4 we present the dissociative ionization
cross section of Hϩ and Dϩ from photoionization of H2 and
D2 molecules. These cross sections have been corrected for
the known kinetic energy distribution of Hϩ and Dϩ
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:
J. Chem. Phys., Vol. 103, No. 10, 8 September 1995
130.113.86.233 On: Tue, 02 Dec 2014 15:11:01

3914
He et al.: Photoionization of H2 and D2
FIG. 3. Hϩ cross section as a function of photon energy. Theory: ᭡, Ford et
al., Refs. 10, 11;-––-––-, O’Neil and Reinhardt, Ref. 5. Experimental:—,
present results.
fragments.3–6 In order to obtain the absolute dissociative ion-
ization cross sections, we have extracted the branching ratios
͓i.e., Hϩ/͑HϩϩH2ϩ͒ and Dϩ/͑DϩϩDϩ2 ͔͒ from Fig. 2 and mul-
tiplied by the previously known photoionization cross sec-
tions of H2 and D2.12,22 The D2 total ionization cross section
is identical to that of H2 at most photon energies except
where transitions into repulsive states take place. This is
caused by the different widths of the H2 and D2 Franck–
Condon regions as can be seen by the energy shifts between
the structures in Figs. 2–4. Also included in Figs. 3 and 4 are
the calculated Hϩ and Dϩ cross sections5,10,11 for the direct
FIG. 5. Potential energy diagram of H2 showing the production of Hϩ frag-
ments from autoionization of the doubly excited neutral Q1 and Q2 states.
The limits of the Franck–Condon region were taken at 1.15 a0 and 1.80 a0 .
This figure is a modification of Fig. 5 in Ref. 28.
falls off very rapidly at higher photon energies and differs
from our data by nearly a factor of two at 40 eV. A small
difference ͑ϳ7%͒ in the Dϩ cross section ͑see Fig. 4͒ near
the photon energy of 19 eV is consistent with the large error
͑Ϯ10%͒ in the absolute ratio13 of Dϩ/D2ϩ, which was used
for normalization.
dissociation through the X ⌺gϩ state. As illustrated in Fig. 3,
2
there is an almost perfect agreement between the present data
of Hϩ and the calculation of Ford et al.10,11 except where the
excitation of the doubly excited states takes place. Such
agreement suggests that the single center Coulomb wave-
We also emphasize that the intensities of the autoioniz-
ing features observed in Figs. 2–4 are not on an absolute
scale. However, their values are relative to the continuum
production of Hϩ and Dϩ ions with kinetic energies lying
between zero and 10 meV. The production of these ions can
be explained by the resonant model proposed by Kirby et
al.16,17 Following the initial excitation from the ground state
function treatment for the final state ͑used by Ford et al.10,11
͒
is adequate for describing the direct dissociation process
even for photon energies higher than 23 eV, which was not
clear before the present study was attempted. The agreement
between the present data ͑both Hϩ and Dϩ͒ and those of the
calculation of O’Neil and Reinhardt5 is also rather good near
the dissociation threshold but the calculated cross section
of neutral X ⌺ϩg state of H2 into one of the doubly excited
1
Q1 and Q2 states, as illustrated in Fig. 5, the atoms begin to
separate along the potential energy surface of the doubly
excited state and gain a certain amount of kinetic energy. At
some internuclear distance R, the molecule may undergo
spontaneous autoionization into a particular vibrational level
of the X ⌺ϩg state of Hϩ2 . If the energy gained ͑⌬E͒ is
2
exactly the amount necessary to produce dissociation of H2ϩ
then zero kinetic energy fragments are produced. The prob-
ability of this occurring depends upon ͑i͒ the Franck–
Condon transition probability between the H2 X ⌺ϩg ground
1
state and the repulsive neutral state involved, ͑ii͒ the lifetime
of that state against autoionization, and ͑iii͒ the transition
probability from the repulsive state into the Hϩ2 X ⌺gϩ state.
2
Because of the finite lifetimes of the repulsive states and the
tendency for these lifetimes to be greater at smaller internu-
clear distances,17,23–27 we would expect, qualitatively, that
the probability for producing zero kinetic energy products
would be greater for transitions into the ␷Јϭ0 level of H2ϩ
FIG. 4. Dϩ cross section as a function of photon energy. Theory: -––-––-,
O’Neil and Reinhardt, Ref. 5. Experimental:—, present results.
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:
J. Chem. Phys., Vol. 103, No. 10, 8 September 1995
130.113.86.233 On: Tue, 02 Dec 2014 15:11:01

He et al.: Photoionization of H2 and D2
TABLE I. Starting and ending energy positions for peaks in Hϩ spectra. TABLE II. Starting and ending energy positions for peaks in Dϩ spectra.
Starting energy ͑eV͒ Ending energy ͑eV͒ Starting energy ͑eV͒ Ending energy ͑eV͒
3915
Theorya
Observed
Theoryb
Observed
Theorya
Observed
Theoryb
Theoryc
Observed
1
1
1
1
1
1
1
1
Q1
Q1
Q1
Q2
⌺
⌺
⌺
⌺
ϩg (1)
ϩu (1)
ϩu (2)
ϩu (1)
23.0
25.5
27.8
33.3
23.0Ϯ0.2
27.2Ϯ0.4
30.0Ϯ0.4
33.2Ϯ0.2
27.9
30.1
32.5
37.4
26.3Ϯ0.2
31.0Ϯ0.4
32.7Ϯ0.2
37.6Ϯ0.2
Q1
Q1
Q1
Q2
⌺
⌺
⌺
⌺
ϩg (1)
ϩu (1)
ϩu (2)
ϩu (1)
24.0
26.2
28.8
33.9
24.9Ϯ0.2
28.6Ϯ0.4
31.6Ϯ0.4
34.2Ϯ0.2
27.1
29.4
31.7
34.9
28.6
30.7
33.2
37.7
28.6Ϯ0.2
32.0Ϯ0.4
33.9Ϯ0.4
38.2Ϯ0.2
aReference 28.
aOuter semiclassical turning point for D2 ground stateϭ1.72 a0 .
bInternuclear distance for the outer semiclassical turning point of the ground
bInner semiclassical turning point for ␷ϭ0 ground state of D2ϩϭ1.69 a0 .
cInner semiclassical turning point for ␷ϭ1 vibrational state of D2ϩϭ1.57 a0 .
state of Hϩ2 ϭ1.63 a0 .
X ⌺ϩg . Thus in the analysis of our results we assume that
state could be observed as a very weak feature. The possibil-
ity of some other allowed states existing in this region is not
expected.30 Moreover, our assignment is also consistent with
the fact that the observed starting energies are in good agree-
ment with theory.
2
the major intensity of the structure observed is produced by
autoionization of the neutral repulsive states into the ␷Јϭ0
vibrational level of the Hϩ2 X ⌺ϩg state. Bearing this as-
2
sumption in mind, we are able to assign the observed four
peaks as excitation into the Q1 1⌺gϩ(1), Q1 1⌺uϩ(1),
Q1 1⌺ϩu (2), and Q2 1⌺ϩu (1) doubly excited states, respec-
tively.
To illustrate how these assignments were made, let us
examine the feature peaked at ϳ35 eV, namely, the
Q2 1⌺ϩu (1) state. According to the Franck–Condon prin-
ciple, transitions occur within the Franck–Condon region de-
fined by the semiclassical turning points of the vibrational
state. The choice of the semiclassical turning points is some-
Notice in Table I that the observed ending point of the
Q1 1⌺ϩg (1) state is about 1.5 eV smaller than the predicted
value. This could be caused by a lifetime effect. The lifetime
of the resonant state against autoionization usually increases
as the internuclear distance becomes smaller.17,23–27 When
the lifetime becomes too long, the resonant state may un-
dergo decay with ⌬E larger than 2.648 eV, thus producing
energetic protons ͑deuterons͒ rather than the zero energy
fragments. The same effect could also account for the appar-
ent delays in the starting points of both the Q1 1⌺uϩ(1) and
Q1 1⌺ϩu (2) states. In this case, the lifetime of the resonant
state is probably too short. Therefore, the resonant state de-
cays into the ground state of Hϩ2 with ⌬E smaller than 2.648
eV and no atomic fragment is produced. It is clear that a
more detailed calculation needs to be made to account for
this difference quantitatively.
what arbitrary.
A
reasonable choice, according to
Guberman,28 would be the midpoints between intercepts of
the potential energy curve and two adjacent ␷ and ͑␷ϩ1͒
vibrational levels. As shown in Fig. 5, excitation into the
Q2 1⌺ϩu (1) state from the X ⌺ϩg ground state becomes pos-
1
sible at a photon energy of approximately 33.3 eV corre-
sponding to the internuclear distance of 1.8 a0 ,28 where a0 is
the Bohr radius. The observed onset value is about 33.2Ϯ0.2
eV. As the photon energy increases, the probability of the
transition into the Q2 1⌺uϩ(1) state becomes larger and
hence more Hϩ ions are expected. The production of Hϩ ions
is also dependent on the probability of the excited state au-
toionizing into the ground state of Hϩ2 . This transition prob-
ability becomes negligible when the internuclear distance is
smaller than the inner semiclassical turning point of H2ϩ
The observed features for the isotope D2 is also consis-
tent with the resonant model. As seen from Fig. 2 and Table
II, the starting energies of structures of Dϩ shift toward
higher values since the Franck–Condon region of the ground
state of D2 is about 16% smaller than that of H2. An inter-
nuclear distance of 1.72 a0 was chosen for the semiclassical
outer turning point of the D2 ground state. Similar to D2, the
Franck–Condon region is expected to be about 16% smaller
for ␷Јϭ0 vibrational level of Dϩ2 . The semiclassical inner
turning point was chosen to be 1.69 a0 , and the ending en-
ergies of D2 should be lower than H2 ͑see Tables I and II͒. It
is apparent that this is not the case. This discrepancy can be
resolved if the second lowest vibrational state of Dϩ2 ͑␷Јϭ1͒
is also allowed to participate in the decay mechanism. The
semiclassical inner vibrational turning point for ␷Јϭ1 state
lies at an internuclear distance of approximately 1.57 a0 ,
yielding predicted ending energies of 28.6, 30.7, 33.2, and
37.7 eV for Q1 1⌺gϩ(1), Q1 1⌺uϩ(1), Q1 1⌺ϩu (2), and
Q2 1⌺ϩu (1) states, respectively, which are in good agreement
with the observed values of 28.6Ϯ0.2, 32.0Ϯ0.4, 33.9Ϯ0.4,
and 38.2Ϯ0.4 eV, respectively. This reasonable agreement is
not fortuitous but probably correlated with the fact that the
␷Јϭ0 and ␷Јϭ1 states of Dϩ2 lie close to the lowest vibra-
tional state of Hϩ2 .
X ⌺ϩg ͑␷ϭ0͒, which is approximately 1.63 a0 . Since the de-
2
tected ions have nearly zero kinetic energy, ⌬E must be
equal to the dissociation energy, which is 2.648 eV.29 Based
on this argument, the Hϩ signal will drop to a minimum at a
photon energy of approximately 37.4 eV, which is in good
agreement with the experimental result of 37.6Ϯ0.2 eV.
Similar arguments can also be applied to the
Q1 1⌺ϩg (1), Q1 1⌺ϩu (1), and Q1 1⌺ϩu (2) doubly excited
states. The predicted starting and ending values for H2 are
tabulated in Table I along with the observed results. The
observation of a transition into the Q1 1⌺gϩ(1) state was not
expected since this excitation is dipole forbidden. However,
due to configuration interaction the transition may be pos-
sible when quadruple transitions are taken into account, even
though this probability is generally very small. On the other
hand, the Q1 1⌺ϩg (1) state is predicted to autoionize very
efficiently into HϩϩH.15 As a result, it is possible that this
It should be pointed out that in addition to Q1 1⌺gϩ(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:
J. Chem. Phys., Vol. 103, No. 10, 8 September 1995
130.113.86.233 On: Tue, 02 Dec 2014 15:11:01

3916
He et al.: Photoionization of H2 and D2
Q1 1⌺ϩu (1), Q1 1⌺uϩ(2), and Q2 1⌺ϩu (1) states other
known doubly excited states, which obey the dipole selection
rule, are Q1 1⌸ϩu (1), Q1 1⌸uϩ(2), and Q2 1⌸ϩu (1). Results
from previous fluorescence studies31,32 as well as the semi-
classical calculation33 suggest that both Q1 1⌸ϩu (1) and
Q2 1⌸ϩu (1) dissociate into neutral fragments with yields of
approximately 90%, therefore, their contribution to the pro-
duction of ionic fragments have been neglected. The behav-
ior of the Q1 1⌸ϩu (2) state is less well understood. However,
it is reasonable to assume its contribution to the production
of ions to be minimal based on the other 1⌸uϩ states behavior.
Finally, we also note that autoionization of the Q1 and
Q2 states will produce stable Hϩ2 and D2ϩ ions in various
vibrational levels. The kinetic energies of these ions are iden-
tical to those produced by direct ionization and hence these
Hϩ2 and Dϩ2 ions will also be collected in the present mea-
surements. This enhancement in the parent Hϩ2 and D2ϩ ion
signals will cause a slight lowering of the ratio in the vicinity
of the resonance features, but could only be observed when
Hϩ and Dϩ contribution from the resonance drops to zero.
We suggest this could be the reason why the minima between
the resonance features at 26.5 and 33 eV for Hϩ/H2ϩ and at
28.5 and 34 eV for Dϩ/D2ϩ in Fig. 2 both lie below the
expected extrapolation curves for the continuum state.
In conclusion, dissociative ionization of H2 and D2 have
been studied by the ZIKE technique. The single center Cou-
lomb treatment by Ford et al.10,11 for the final state appears
adequate for describing the direct dissociation through the
ATM-9410716. We are also grateful for the help of the per-
sonnel at NIST and at the Synchrotron Radiation Center
͑NSF͒ Grant No. OMR-9212658. Finally, we would like to
thank Professor A. L. Ford for providing the numerical val-
ues of his published and unpublished cross section data.
1 T. E. Cravens, G. A. Victor, and A. Dalgano, Planet. Space Sci. 23, 1059
͑1975͒.
2 W. T. Huntress, Jr., Adv. Atom. Mol. Phys. 10, 295 ͑1974͒.
3 J. L. Gardner and J. A. R. Samson, Phys. Rev. A 12, 1404 ͑1975͒.
4 S. Strathdee and R. Browning, J. Phys. B 12, 1789 ͑1979͒; 9, L505 ͑1976͒.
5 S. V. O’Neil and W. P. Reinhardt, J. Chem. Phys. 69, 2126 ͑1978͒.
6 A. L. Ford and K. K. Docken, J. Chem. Phys. 62, 4955 ͑1975͒.
7 C. J. Latimer, K. F. Dunn, F. P. O’Neil, M. A. MacDonald, and N. Kouchi,
J. Chem. Phys. 102, 722 ͑1995͒.
8 T. Masuoka, J. Chem. Phys. 81, 2652 ͑1984͒.
9 C. Y. R. Wu, T. S. Chien, and D. L. Judge, J. Chem. Phys. 92, 1713
͑1990͒.
10 A. L. Ford, K. K. Docken, and A. Dalgarno, Astrophys. J. 195, 819
͑1975͒.
11 A. L. Ford ͑private communication͒.
12 Y. M. Chung, E. M. Lee, T. Masuoka, and J. A. R. Samson, J. Chem. Phys.
99, 885 ͑1993͒.
13 B. Browning and J. Fryar, J. Phys. B 6, 364 ͑1973͒.
14 C. Backx, G. R. Wight, and M. J. Van der Weil, J. Phys. B 9, 315 ͑1976͒.
15 A. U. Hanzi, J. Chem. Phys. 60, 4358 ͑1974͒.
16 K. Kirby, T. Uzer, A. C. Allison, and A. Dalgarno, J. Chem. Phys. 75, 2820
͑1981͒.
17 K. Kirby, S. Guberman, and A. Dalgarno, J. Chem. Phys. 70, 4635 ͑1979͒.
18 S. Kanfer and M. Shapiro, J. Phys. B 16, L655 ͑1983͒.
19 C. J. Latimer, A. D. Irvine, M. A. McDonald, and O. G. Savage, J. Phys.
B 25, L211 ͑1992͒.
Hϩ2 X ⌺ϩg state even at higher energies. The structures ob-
2
20 K. Ito, P. Lablanquie, P. Guyon, and I. Nenner, Chem. Phys. Lett. 151, 121
͑1988͒.
served in both Hϩ and Dϩ spectra have been assigned as the
doubly excited Q1 1⌺gϩ(1), Q1 1⌺ϩu (1), Q1 1⌺ϩu (2), and
Q2 1⌺ϩu (1) states. It appears that zero energy Hϩ fragments
are most probably produced from autoionization of a reso-
nant state into the lowest vibrational state of Hϩ2 , whereas the
production of Dϩ fragments may be more probable through
autoionization into both ␷Јϭ0 and ␷Јϭ1 vibrational levels of
Dϩ2 . Further theoretical investigations are needed if a more
quantitative comparison is to be made between theory and
experiment.
21 A. D. Dahl and J. F. Delmore, 1988 Report Idaho Natl. Eng. Lab., P. O.
Box 1625, Idaho Falls, ID 83415, U.S.A.
22 J. A. R. Samson and G. N. Haddad, J. Opt. Soc. Am. B 11, 277 ͑1994͒.
23 C. Bottcker and K. Docken, J. Phys. B 7, L5 ͑1973͒.
24 C. J. Latimer, K. F. Dunn, N. Kouchi, M. A. McDonald, V. Srigengan, and
J. Geddes, J. Phys. B 26, L595 ͑1993͒.
25 I. Shimamura, C. J. Noble, and P. G. Burke, Phys. Rev. A 41, 3545 ͑1990͒.
26 L. A. Collins, B. I. Schneider, D. L. Lynch, and C. J. Noble, Phys. Rev. A
͑submitted͒.
27 L. A. Collins, B. I. Schneider, and C. J. Noble, Phys. Rev. A 45, 4610
͑1992͒.
28 S. L. Guberman, J. Chem. Phys. 78, 1404 ͑1983͒.
29 T. E. Sharp, At. Data 2, 119 ͑1971͒.
30 S. L. Guberman ͑private communication͒.
ACKNOWLEDGMENTS
31 M. Glass-Maujean, J. Chem. Phys. 85, 4830 ͑1986͒.
32 M. Glass-Maujean, J. Chem. Phys. 89, 2839 ͑1988͒.
33 A. U. Hanzi and K. Wiemers, J. Chem. Phys. 66, 5296 ͑1977͒.
This work was made possible with the financial support
from the National Science Foundation under Grant No.
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:
J. Chem. Phys., Vol. 103, No. 10, 8 September 1995
130.113.86.233 On: Tue, 02 Dec 2014 15:11:01