Photodissociation of Kr2F(4 2Γ) in the ultraviolet and near-infrared: Wavelength dependence of KrF (B 2Σ) yield

Article abstract of DOI:10.1063/1.473567

The photoabsorption spectrum of the 4 ²Γ excited state of Kr₂F has been measured in the 280-850 nm region by fluorescence suppression spectroscopy. Both the 9 ²Γ←4 ²Γ and 6 ²Γ←4 ²Γ bands, peaking at 320 and ～710 nm, respectively, have been observed - the latter for the first time. Although the position of the ultraviolet band is consistent with both theory and previous experiments, its spectral width is ～40% of that measured by Greene and McCown [Appl. Phys. Lett. 54, 1965 (1989)] in absorption experiments in a pulsed discharge. The relative yield of KrF(B) produced in the photodissociation of Kr₂F(4 ²Γ) was found to be independent of wavelength over the 280-360 nm and 590-840 nm spectral intervals and the mechanism responsible for photodissociation appears to be predissociation of the Kr₂F 9 ²Γ and 6 ²Γ states by potentials correlated with the KrF(D) + Kr and KrF(C) + Kr limits, respectively.

Full text of DOI:10.1063/1.473567

Photodissociation of Kr 2 F (4 2 Γ) in the ultraviolet and near-infrared: Wavelength
dependence of KrF (B 2 Σ) yield
J. H. Schloss, H. C. Tran, and J. G. Eden
Citation: The Journal of Chemical Physics 106, 5423 (1997); doi: 10.1063/1.473567
View online: http://dx.doi.org/10.1063/1.473567
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/106/13?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Insights on the CN B 2Σ+ + Ar potential from ultraviolet fluorescence excitation and infrared depletion studies of
the CN–Ar complex
J. Chem. Phys. 136, 234303 (2012); 10.1063/1.4723694
Fluorescence-dip infrared spectroscopy and predissociation dynamics of OH A Σ + 2 ( v = 4 ) radicals
J. Chem. Phys. 122, 244313 (2005); 10.1063/1.1937387
Quantum yields for product formation in the 120–133 nm photodissociation of O 2
J. Chem. Phys. 121, 10437 (2004); 10.1063/1.1809114
Intracluster stereochemistry in van der Waals complexes: Steric effects in ultraviolet photodissociation of state-
selected Ar–HOD/H 2 O
J. Chem. Phys. 120, 8443 (2004); 10.1063/1.1697394
Photofragment emission yield spectroscopy of acetylene in the D̃ 1 Π u , Ẽ 1 A, and F̃ 1 Σ u + states by vacuum
ultraviolet and infrared vacuum ultraviolet double-resonance laser excitations
J. Chem. Phys. 117, 1040 (2002); 10.1063/1.1485064
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:
141.100.74.200 On: Wed, 26 Nov 2014 08:40:22

Photodissociation of Kr2F(4 2⌫) in the ultraviolet and near-infrared:
2
Wavelength dependence of KrF (B ⌺) yield
J. H. Schloss,a) H. C. Tran,b) and J. G. Eden
Everitt Laboratory, Department of Electrical and Computer Engineering, University of Illinois, Urbana,
Illinois 61801
͑Received 27 June 1996; accepted 24 December 1996͒
The photoabsorption spectrum of the 4 2⌫ excited state of Kr2F has been measured in the 280–850
nm region by fluorescence suppression spectroscopy. Both the 9 2⌫←4 2⌫ and 6 2⌫←4 2⌫ bands,
peaking at 320 and ϳ710 nm, respectively, have been observed—the latter for the first time.
Although the position of the ultraviolet band is consistent with both theory and previous
experiments, its spectral width is ϳ40% of that measured by Greene and McCown ͓Appl. Phys.
Lett. 54, 1965 ͑1989͔͒ in absorption experiments in a pulsed discharge. The relative yield of
KrF(B) produced in the photodissociation of Kr2F(4 2⌫) was found to be independent of
wavelength over the 280–360 nm and 590–840 nm spectral intervals and the mechanism
responsible for photodissociation appears to be predissociation of the Kr2F 9 2⌫ and 6 2⌫ states by
potentials correlated with the KrF(D) ϩ Kr and KrF(C) ϩ Kr limits, respectively. © 1997
American Institute of Physics. ͓S0021-9606͑97͒00313-9͔
2
I. INTRODUCTION
near-UV ͑308–351 nm͒ yields KrF͑B ⌺͒ with a quantum
efficiency approaching unity.11 An analogous observation for
Ar2F͑4 2⌫͒ was later reported in Ref. 12 but early experi-
ments on Xe2Cl had searched unsuccessfully for this
process.4
The lowest electronic excited state of the rare gas-halide
triatomic molecules, denoted 4 2⌫, is ionic in character ͑i.e.,
Rgϩ2 XϪ, where Rg and X represent rare gas and halogen
atoms, respectively͒ and ab initio calculations1–3 suggest that
the effect of the halogen anion on the structural and optical
properties of the rare gas dimer cation is negligible. Conse-
quently, the absorption spectra for the 9 2⌫←4 2⌫ and
6 2⌫←4 2⌫ transitions of Kr2F, for example, are expected to
be similar to those for the corresponding transitions of Krϩ2 :
Experiments in which the Kr2F͑4 2⌫͒ absorption spec-
trum was recorded from 280 to 850 nm are described here.
The bands associated with both the 9 2⌫←4 2⌫ and
6 2⌫←4 2⌫ transitions and lying in the UV ͑280–390 nm͒
and near-infrared ͑550–850 nm͒, respectively, are observed
in their entirety. In particular, the 6 2⌫←4 2⌫ absorption
band of the molecule which peaks at ϳ710 nm is observed
for the first time and the position and breadth of the band
profile are consistent with theoretical predictions. In addition
to absorption spectra obtained by fluorescence suppression
spectroscopy, the production of KrF excited states by photo-
dissociating Kr2F͑4 2⌫͒ was studied by recording the peak
KrF(B→X) emission intensity as the excitation wavelength
was scanned. A comparison of the absorption ͑fluorescence
suppression͒ and excitation spectra reveal that the relative
yield of KrF(B) in the photodissociation of Kr2F͑4 2⌫͒ is
independent of wavelength from 280 nm to beyond 800 nm.
D ⌺ϩg ←A ⌺ϩu and B ⌸g←A ⌺ϩu , respectively.
2
2
2
2
Measurements of the photoabsorption cross section for
the 4 2⌫ state of Xe2Cl at several wavelengths in the ultra-
violet ͑UV͒ were described by McCown et al.4 in 1985 and,
with the fluorescence suppression approach, McCown5 and
Hakuta and co-workers6 subsequently determined the
Kr2F͑4 2⌫͒ absorption cross section at 248 nm. Similar mea-
surements were made by Geohegan and Eden7,8 for Kr2F at
35 wavelengths in the 248–572 nm region with a dye laser
and were supplemented by continuous scans from 335 to
ϳ600 nm. These results showed that the peak of Kr2F͑4 2⌫͒
absorption in the UV lay between 310 and 335 nm. By pro-
ducing Kr2F in a transverse discharge, Greene and McCown9
were able to directly observe Kr2F͑4 2⌫͒ in absorption be-
tween 225 and 455 nm and confirmed that the 9 2⌫←4 2⌫
band peaks at 315 nm.
II. EXPERIMENT
The experimental apparatus and data acquisition tech-
niques have been discussed in detail previously8 and will
only be reviewed here. A pump–probe approach to selec-
tively examining the optical properties of Kr2F was adopted
in which the initial laser pulse serves to produce the tri-
atomic species. The second, time-delayed ͑probe͒ pulse de-
pletes the Kr2F͑4 2⌫͒ population by absorption and the ob-
servables are the magnitude and wavelength dependence of
the Kr2F͑4 2⌫→1 2⌫͒ fluorescence suppression and the
concomitant appearance of KrF(B) emission. Briefly,
Kr2F͑4 2⌫͒ molecules were produced by the beam from an
ArF laser ͑193 nm͒ which was focused into a quartz cell
In 1988, McCown and co-workers10 were the first to
demonstrate the direct production of the rare gas-fluoride di-
atomic species when a UV photon is absorbed by the tri-
atomic in the 4 2⌫ state. Specifically, these experiments re-
vealed that the photodissociation of Kr2F͑4 2⌫͒ in the
a͒
Present address: SSI Technologies, Inc., P.O. Box 5011, Janesville, WI
53547.
b͒
Present address: Northrop Corp., 600 Hicks Rd., MS-L5300, Rolling
Meadows, IL 60008.
J. Chem. Phys. 106 (13), 1 April 1997
0021-9606/97/106(13)/5423/6/$10.00
© 1997 American Institute of Physics
5423
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:
141.100.74.200 On: Wed, 26 Nov 2014 08:40:22

5424
Schloss, Tran, and Eden: Photodissociation of Kr2F
FIG. 2. Five superimposed Kr2F fluorescence wave forms illustrating the
dependence of the suppression profile on ⌬t for a fixed probe laser intensity
and ␭ϭ320 nm. The dashed lines reflect the slowing of the recovery of the
Kr2F population that results from the rapidly declining KrF(B,C) popula-
tions.
FIG. 1. Typical Kr2F͑4 2⌫→1 2⌫͒ and KrF(B→X) emission wave forms:
͑top͒ Kr2F fluorescence wave forms produced by ArF laser irradiation of a
1000 Torr Kr/1 Torr F2 gas mixture and the suppression of the fluorescence
by a dye laser pulse ͑␭ϭ340 nm͒; ͑bottom͒ KrF(B→X) emission generated
by the ArF laser pulse ͑early fluorescence peaking at ϳ40 ns͒ and the
trailing dye laser ͑probe͒ pulse. Notice that, for illustrative purposes, ⌬t is
quite different for the two sets of wave forms ͑⌬tϳ35 and 90 ns for top and
bottom traces, respectively͒.
ns following the onset of the ArF laser pulse and exhibits a
decay time which is consistent with the known spontaneous
emission lifetime for the 4 2⌫ state and the rate constants for
quenching of Kr2F͑4 2⌫͒ by Kr and F2 ͑Refs. 8 and 13͒.
When the time-delayed probe pulse ͑␭ϭ340 nm͒ is intro-
duced, strong suppression ͑ϳ46%͒ of the Kr2F fluorescence
is observed but note that, following the termination of the
probe pulse, the Kr2F population partially recovers ͑to Ͼ85%
of its pre-probe value͒. In previous experimental studies of
Kr2F absorption by fluorescence suppression, recovery of the
Kr2F population was slight5 or not detectable,6–8 an interest-
ing effect that is closely tied to the temporal width of the
probe pulse and three body rate constant for Kr2F formation
and which will be discussed in more detail later.
The trace in the lower portion of Fig. 1 shows the
KrF(B→X) spontaneous emission that is produced by the
initial ͑ArF pump͒ pulse and the probe radiation ͑␭ϭ340 nm,
⌬tӍ90 ns͒. For the remainder of this paper, we will focus on
the latter. The temporal history of the KrF fluorescence pro-
duced by the probe ͑dye͒ laser pulse replicates the dye laser
wave form itself; one concludes, therefore, that Kr2F͑4 2⌫͒ is
converted to KrF(B) on a time scale of Շ1–2 ns. A super-
position of five Kr2F fluorescence suppression wave forms,
recorded for different values of ⌬t and a probe laser wave-
length of 320 nm, is presented in Fig. 2. Not only do these
traces illustrate the reproducibility of the experiment but they
also demonstrate that the rate and degree of Kr2F͑4 2⌫͒
population recovery following the probe pulse ͑which is re-
flected by the magnitude of the slope of the wave form dur-
ing recovery—cf. Fig. 2͒ falls rapidly as ⌬t increases. This is
precisely what one would expect since the re-formation of
Kr2F͑4 2⌫͒ depends upon the KrF(B,C) populations, both of
which have considerably shorter effective lifetimes than that
for Kr2F͑4 2⌫͒.
containing a room temperature mixture of Kr and F2. Pro-
duction of Kr excited states and Krϩ ions by nonresonant,
two photon excitation and 2ϩ1 REMPI, respectively, trig-
gers a series of collisional processes that culminates in the
formation of KrF(B,C) and Kr2F͑4 2⌫͒. An adjustable time
delay (⌬t) later, a pulse from an Nd:YAG-pumped dye laser
͑ϳ10 ns, ⌬␯Ӎ0.2 cmϪ1, ϳ1 mm2 cross-sectional beam
˜
area͒, directed along the axis of the cell and coincident with
the focus of the excimer, excites a fraction of the Kr2F mol-
ecules. The absorption of dye laser radiation by Kr2F͑4 2⌫͒ is
monitored by observing the suppression of Kr2F͑4 2⌫→1 2⌫͒
fluorescence in the violet with a photomultiplier and appro-
priate optical filters. As noted earlier, the ultimate fate of the
Kr2F species is also of interest; consequently, the relative
peak intensity of the KrF(B→X) emission produced by the
second ͑dye͒ laser pulse was also recorded as the laser wave-
length was varied. Unless noted otherwise, the gas mixture
can be assumed to be 1000 Torr Kr/1 Torr F2 and ⌬t is
typically set at 60 ns.
III. RESULTS AND DISCUSSION
A. Wave forms, typical data
Fluorescence wave forms representative of those ob-
served throughout these experiments are displayed in Fig. 1.
In the upper portion of the figure are shown two oscillograms
illustrating Kr2F͑4 2⌫→1 2⌫͒ fluorescence suppression aris-
ing from absorption by Kr2F͑4 2⌫͒ at 340 nm. In the absence
of a probe pulse, the Kr2F spontaneous emission peaks ϳ50
J. Chem. Phys., Vol. 106, No. 13, 1 April 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:
141.100.74.200 On: Wed, 26 Nov 2014 08:40:22

5425
Schloss, Tran, and Eden: Photodissociation of Kr2F
FIG. 3. Comparison of the temporal histories of the: ͑a͒ fractional suppres-
sion of Kr2F fluorescence in the violet, and ͑b͒ relative peak KrF(B→X)
emission, as functions of the time delay, ⌬t. The decay of both sets of data
is well described by a single exponential having a time constant of 76Ϯ2 ns
for ͑a͒ and 74Ϯ2 ns for ͑b͒. All data were obtained for a probe laser inten-
FIG. 4. Excitation spectrum for the production of KrF from Kr2F͑4 2⌫͒ in
the 280–900 nm region. In acquiring these data, the relative KrF(B→X)
emission was monitored and the probe laser intensity was fixed at ϳ20
MW-cmϪ2. For the sake of clarity, the 400–550 nm region has been mag-
nified and the data for ␭Ͻ390 nm have been smoothed. A few of the Kr2F
͑4 2⌫͒ absorption cross sections in the 337–385 nm interval from Ref. 8 are
also shown.
sity of ϳ20 MW-cmϪ2
.
Insight as to the mechanism responsible for the produc-
tion of KrF(B) is provided by the data of Fig. 3 which com-
pares the magnitude of the Kr2F violet fluorescence suppres-
sion with the peak KrF(B→X) emission intensity as a
function of ⌬t. The fluorescence suppression data were ac-
quired with a probe laser intensity lower than that of Fig. 2 in
order to avoid saturating the Kr2F absorptive transition. Note
that the decline in the Kr2F fluorescence suppression data is
well-described by a single exponential having a time con-
stant of 76Ϯ2 ns. Similarly, the KrF(B) emission decays
exponentially with ⌬t and, as shown by the solid curve in
Fig. 3͑b͒, the time constant that best fits the data is ͑74Ϯ2
ns͒Ϫ1. The fact that the decay constants for both sets of data
in Fig. 3 match ͑to within experimental uncertainty͒ each
other and the known effective lifetime of Kr2F͑4 2⌫͒ for this
gas mixture provides strong evidence that, as suggested by
McCown et al.,10,11 the KrF(B→X) fluorescence generated
by the probe pulse is, indeed, the result of the photodissocia-
tion of Kr2F͑4 2⌫͒. Also, results virtually identical to those of
Fig. 3 were obtained when the probe laser wavelength was
tuned to the expected position1 of the Kr2F͑6 2⌫←4 2⌫͒ tran-
sition in the near-infrared ͑␭ϳ710 nm͒—namely, the sup-
pression of Kr2F͑4 2⌫→1 2⌫͒ spontaneous emission corre-
sponded to the production of KrF(B). Consequently, not
only is the 9 2⌫ state of Kr2F unstable with respect to the
*
spectrum for the production of KrF in the photodissociation
of Kr2F͑4 2⌫͒ over the 280–900 nm region was obtained and
is shown in Fig. 4. Throughout the data acquisition process,
the peak intensity of the probe pulse was maintained at 20
MWϪcmϪ2 ͑2 mJ, ϳ10 ns͒. Two bands are clearly present
in this spectral region, the strongest of which lies in the UV
and peaks at ϳ320 nm. As was discussed earlier, virtually all
previous studies of Kr2F absorption have focused on this
band that arises from 9 2⌫←4 2⌫ transitions of the molecule.
Aside from the data of Ref. 8 that spanned the 335–600 nm
region, the only other 4 2⌫ absorption cross sections in the
literature were measured at 248 nm ͑Refs. 5–8͒, 308 nm
͑Ref. 8͒, and 358 nm ͑Ref. 14͒. Although the location of the
9 2⌫←4 2⌫ band peak in Fig. 4 is in agreement with the
value reported by Greene and McCown9 ͑315 nm͒, our
present measurements ͑as well as the results of Ref. 8͒ sug-
gest that the bandwidth is only ϳ33 nm FWHM, or at least a
factor two smaller than that given in Ref. 9 ͑85 nm͒. The
difference may be attributable to the considerably different
environments in which the two spectra were recorded. In
Ref. 9, Kr2F was produced in a commercial, transverse dis-
charge device and absorption measurements were made in a
20 ns window during the active discharge or in the early
afterglow. It is likely, therefore, that excited vibronic levels
of the Kr2F͑4 2⌫͒ state populated by electron impact have the
effect of significantly broadening the 9 2⌫←4 2⌫ spectrum.
A much weaker band, extending from below 650 nm to
beyond 850 nm, is also observed in Fig. 4 and is assigned to
the 6 2⌫←4 2⌫ transition of Kr2F. The existence of this band,
*
KrF ϩKr limit ͑where the asterisk denotes an electronic ex-
cited state of the diatomic͒ but the 6 2⌫ state is as well.
B. KrF(B) excitation and Kr2F(4 2⌫) absorption
spectra
By monitoring the peak KrF (B→X) emission intensity
while scanning the probe laser wavelength, the excitation
which corresponds to the 1͑1/2͒g←1͑1/2͒ transition
u
J. Chem. Phys., Vol. 106, No. 13, 1 April 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:
141.100.74.200 On: Wed, 26 Nov 2014 08:40:22

5426
Schloss, Tran, and Eden: Photodissociation of Kr2F
fluorescence suppression spectra in the 280–360 nm region
are essentially identical, indicating that the yield of KrF(B)
in this wavelength range is also constant.
The uniform efficiency with which Kr2F is photodisso-
ciated over such a broad spectral range is interesting, particu-
larly in view of the fact that spontaneous emission from Kr2F
excited states higher than 4 2⌫ has not been reported in the
literature. Diatomics-in-molecules ͑DIM͒ calculations car-
ried out by Huestis and Schlotter2 for the lowest nine elec-
tronic excited states of Kr2F show that all are strongly bound
͑ϳ2 eV͒ with respect to the Krϩ2 ϩFϪ limits. However, only
one—4 2⌫—is stable with respect to the KrF(B,C,D)ϩKr
limits. Specifically, unpublished calculations by Huestis pre-
dict that the potentials correlated with KrϩFϪ(C,D)ϩKr are
dissociative. Since: 1͒ the symmetry species1 for the
KrF͑C ⌸3/2͒ϩKr and KrF͑D ⌺ϩ1/2͒ϩKr limits are B2 and
2A1, respectively, and 2͒ once spin–orbit effects are consid-
2
2
2
ered, the Kr2F͑6 2⌫͒ state has both A2 and B2 character,
then it is likely that the potential associated with KrF(C)
ϩKr is responsible for predissociation of the Kr2F͑6 2⌫͒
state. Similar considerations suggest that Kr2F͑9 2⌫͒ ͑which
is almost entirely A1 in structure͒ is predissociated by the
repulsive potential arising from KrF(D)ϩKr.
2
2
FIG. 5. KrF (B) excitation spectrum ͑top͒ and Kr2F͑4 2⌫͒ fluorescence sup-
pression ͑absorption͒ spectrum in the red and near-IR showing maxima at
655 and 710 nm attributed to the 6 2⌫←42⌫ transition of Kr2F.
2
1
It is, therefore, understandable that photoexciting the
9 2⌫ state of Kr2F would rapidly culminate in the production
of KrF(B→X) emission. The D and B states of KrF are
separated in energy by approximately the spin–orbit splitting
(B ⌸g←A ⌺ϩu ) of the parent Krϩ2 ion,1,8 was predicted by
Wadt and Hay1 but, to our knowledge, has not been reported
previously. Also, the peak intensity for this band relative to
that for the UV band is consistent with the ratio of oscillator
strengths calculated in Ref. 1 for the corresponding transi-
tions.
2
2
of Krϩ ͑2P3/2Ϫ2P1/2Ӎ0.67 eV͒. Consequently, the D ⌺ϩ1/2
2
state is ‘‘embedded’’ within the B state potential and a num-
ber of low D ⌺ϩ1/2 vibrational levels are nearly degenerate
2
with ЈϽ30 levels of the B ⌺ϩ state. Furthermore, at the
2
v
Aside from these two obvious bands, other faint but re-
producible structure appears in the 420–490 nm region and
in the vicinity of 600 nm. For convenience, the 400–550 nm
interval in Fig. 4 has been magnified and, although the signal
is too weak to make definitive statements, it is quite possible
that these features are associated with the 8 2⌫←4 2⌫ and
7 2⌫←4 2⌫ transitions of Kr2F that were predicted by Wadt
and Hay1 to peak at 478 and 603 nm, respectively. The latter
band appears to be partially blended with the blue tail of the
stronger 6 2⌫←4 2⌫ band discussed earlier.
1/2
gas pressures typical of these experiments, the B and C
states of KrF are closely coupled by collisions—hence, pre-
dissociation of the 6 2⌫ state of Kr2F would also quickly
produce the KrF(B) species.
C. Modeling of Kr2F fluorescence recovery: Estimate
of KrF yield
Figure 5 gives a comparison of the KrF(B) excitation
spectrum in the 550–900 nm region with the Kr2F͑4 2⌫͒ ab-
sorption spectrum measured by fluorescence suppression
over approximately the same wavelength interval. Although
the S/N ratio for the fluorescence suppression data is roughly
one half of that for the excitation spectrum, it is nevertheless
clear that the two spectra exhibit the same structure. Local
maxima at ϳ655 and 710 nm appear in both spectra and their
overall profiles are virtually indistinguishable. Recalling that
the excitation spectrum is a measure of the production of
In the discussion of Sec. III A, the partial recovery of the
Kr2F͑4 2⌫͒ fluorescence suppression wave forms following
the termination of the probe pulse was noted. One would
expect the rate and degree of recovery to depend critically on
several kinetic constants; in fact, the experimental wave
forms serve as a sensitive and convenient test of the rate
2
*
constant for the formation of Kr2F͑4 ⌫͒ from KrF by three
body collisions and, moreover, allow for a lower limit on the
*
yield of KrF in the photodissociation of Kr2F to be deter-
*
KrF as a function of wavelength, then normalizing the ex-
mined.
citation spectrum to the fluorescence suppression ͑absorp-
tion͒ spectrum provides a measure of the wavelength depen-
In order to assess the impact of various experimental
parameters and rate constants on the fluorescence suppres-
sion wave forms, a simple kinetics model of the KrF–Kr2F
system was constructed in which, for convenience, the pho-
todissociation of Kr2F ͑4 2⌫͒ was assumed to produce the
KrF (C) species. Photoexcitation of Kr2F can be described
by the following set of three coupled rate equations:
*
dence of the relative yield of KrF in the photodissociation
of Kr2F͑4 2⌫͒. Consequently, it is reasonable to conclude that
the KrF(B) yield is independent of wavelength in the 590–
840 nm region. The same result was obtained for the
9 2⌫←4 2⌫ spectra—the profiles for both the excitation and
J. Chem. Phys., Vol. 106, No. 13, 1 April 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:
141.100.74.200 On: Wed, 26 Nov 2014 08:40:22

5427
Schloss, Tran, and Eden: Photodissociation of Kr2F
FIG. 7. Kr2F͑4 2⌫͒ fluorescence suppression wave forms calculated for five
values of the three body Kr2F formation rate constant k3 , ranging from
FIG. 6. Simulated Kr2F͑4 2⌫͒ fluorescence suppression wave forms illustrat-
ing the dependence of the recovery on the KrF(C) quantum yield, ␤. The
actual dye ͑probe͒ laser wave form was incorporated into the computations
and the peak probe laser intensity was assumed to be 20 MW-cmϪ2. The
initial Kr2F͑4 2⌫͒ and KrF(B,C) number densities were set at 3ϫ1013 cmϪ3
and zero, respectively, and the probe laser wavelength was 337 nm
͑9 2⌫←4 2⌫͒.
10Ϫ32 cm6 sϪ1 to 10Ϫ30 cm6 sϪ1
.
respectively. Also, the initial Kr2F ͑4 2⌫͒ number density is
taken to be 3ϫ1013 cmϪ3 while the initial KrF(B,C) popu-
lations are set to zero. When ␤ϭ0, re-feeding of the Kr2F
͑4 2⌫͒ state does not occur and, on the basis of a comparison
with experimental wave forms, it is clear that ␤ must be
у0.75. This result is consistent with the conclusions reached
by McCown11 who measured ␤ for the photodissociation of
both Kr2F and Xe2Cl in the 193–351 nm spectral interval.
For Kr2F, the branching ratio was determined to 0.9Ϯ0.1
between 308 and 351 nm but vanished at 248 and 193 nm.
Consequently, the result reported here—that ␤ is constant in
the 280–360 nm and 590–840 nm regions—is in agreement
with the data of Ref. 11.
ץ KrF B͒
KrF B͒
͓
͑
͔
͔
͓
͑
͔
2
ϭϪk3 Kr KrF B͒ Ϫ
͔ ͓
͓
͑
͔
ץt
␶
b
ϩkm KrF C͒ Ϫ KrF B͒ ,
͔
͖
͑1͒
͑2͒
͓
͑
͔
͓
͑
͕
ץ KrF C͒
͓
͑
ϭkm KrF B͒ Ϫ KrF C͒
͔
͖
͓
͑
͔
͓
͑
͕
ץt
␴I t͒
KrF C͒
͑
͓
͑
͔
ϩ
␤ Kr F Ϫ
,
͓
͔
2
ប␻
␶
c
The model-generated fluorescence suppression wave
forms are also sensitive to the temporal width of the probe
laser pulse and to the assumed value for the rate constant k3.
Most of the previous measurements of Kr2F͑4 2⌫͒ absorption
cross sections by fluorescence suppression have involved an
excimer laser or excimer laser-pumped dye laser having
pulsewidths in excess of 20 ns. Since this is large compared
to the effective KrF(B) lifetime for the gas mixtures of in-
terest here, the simulations confirm prior observations5–7 that
little recovery of the Kr2F fluorescence following the termi-
nation of the probe laser pulse is to be expected. Only as the
probe pulsewidth falls below 10 ns ͓i.e., becoming compa-
rable to the KrF(B) lifetime͔ does recovery of the wave form
become observable. The results of Fig. 7 illustrate the sensi-
tivity of the degree and rate of fluorescence recovery to k3.
Values in the literature for this rate constant differ by more
than an order of magnitude but it is clear from the simulated
wave forms that accounting for the experimental data re-
quires that k3 be у5ϫ10Ϫ31 cm6 sϪ1. Finally, the influence
of re-feeding of the Kr2F state is clearly illustrated by Fig. 8
in which the variation of the measured fluorescence suppres-
ץ Kr F
Kr F
͓
͔
͓
͔
2
2
2
ϭϪkq
F
Kr F Ϫ
ϩk3 Kr KrF B͒
͔ ͓
͓
͔͓
͔
͓
͑
͔
2
2
ץt
␶
3
␴I t͒␤
͑
Ϫ
Kr F ,
͑3͒
͓
͔
2
ប␻
where brackets denote number densities, ͓Kr2F͔ represents
the population of the 4 2⌫ state, k3 is the three body forma-
tion rate constant for Kr2F ͑6.7ϫ10Ϫ31 cm6 sϪ1͒, ␶ is the B
b
state spontaneous emission lifetime ͑6.8 ns͒, km is the rate
constant for the mixing of the B and C states by two body
collisions with Kr ͑5ϫ10Ϫ10 cm3 sϪ1͒, ␶ is the C state ra-
c
diative lifetime ͑75 ns͒, kq is the rate constant for the
quenching of Kr2F͑4 2⌫͒ by F2 ͑4.3ϫ10Ϫ10 cm3 sϪ1͒, ␶3 is the
Kr2F͑4 2⌫͒ spontaneous emission lifetime ͑178 ns͒, and ␤ is
the quantum yield of KrF(C) in the photodissociation of
Kr2F͑4 2⌫͒ ͑Refs. 11, 13, 15, and 16͒.
A series of five simulated Kr2F fluorescence suppression
wave forms are shown in Fig. 6 for ␤ varied from 0 to 1. In
these computations, it was assumed that the probe laser
wavelength and peak intensity are 337 nm and 20 MW cmϪ2
,
J. Chem. Phys., Vol. 106, No. 13, 1 April 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:
141.100.74.200 On: Wed, 26 Nov 2014 08:40:22

5428
Schloss, Tran, and Eden: Photodissociation of Kr2F
Photoexcitation of Kr2F͑4 2⌫͒ in the UV or near-IR re-
sults in dissociation of the molecule and the production of a
KrF fragment in the C or D excited states. Comparing the
Kr2F absorption ͑fluorescence suppression͒ and KrF(B) ex-
*
citation spectra demonstrate that the quantum yield for KrF
production is independent of wavelength in the 280–360 nm
and 590–840 nm intervals. Modeling of the experimental
fluorescence suppression wave forms indicates that
the quantum yield for the process: Kr2F͑4 2⌫͒ϩប␻ ͑3.4–4.4
eV͒→Kr2F͑9 2⌫͒→KrF(D)ϩKr is у0.75. These results
largely confirm the predictions of ab initio calculations con-
cerning the positions and relative strengths of Kr2F absorp-
tion bands as well as predissociation of the 6 2⌫ and 9 2⌫
excited states.
ACKNOWLEDGMENTS
Several valuable discussions with A. W. McCown and
D. P. Greene and the expert technical assistance of K. Kuehl,
K. Voyles, L. Cliff, D. Prest, and J. Sexton are gratefully
acknowledged. This work was supported by the National
Science Foundation and the Office of Naval Research.
FIG. 8. Calculated dependence of the percentage fluorescence suppression
on the probe laser intensity. The results are shown for five values of ␤. The
actual dye laser wave form was incorporated into the simulations.
sion on probe laser intensity is shown for five values of ␤.
Not surprisingly, as the efficiency for replenishing the
Kr2F͑4 2⌫͒ population is increased ͑␤→1͒, the apparent
͑measured͒ absorption cross section is smaller than the true
value and the effect is most pronounced at low probe laser
intensities. However, it should be mentioned that the magni-
tude of the correction for this effect is reduced dramatically
for longer probe laser pulses.
1 W. R. Wadt and P. J. Hay, J. Chem. Phys. 68, 3850 ͑1978͒.
2 D. L. Huestis and N. E. Schlotter, J. Chem. Phys. 69, 3100 ͑1978͒.
3 W. J. Stevens and M. Krauss, Appl. Phys. Lett. 41, 301 ͑1982͒.
4 A. W. McCown, M. N. Ediger, D. B. Geohegan, and J. G. Eden, J. Chem.
Phys. 82, 4862 ͑1985͒.
5 A. W. McCown, Appl. Phys. Lett. 50, 804 ͑1987͒.
6 K. Hakuta, H. Komori, N. Mukai, and H. Takuma, J. Appl. Phys. 61, 2113
͑1987͒.
7 D. B. Geohegan and J. G. Eden, Chem. Phys. Lett. 139, 519 ͑1987͒.
8 D. B. Geohegan and J. G. Eden, J. Chem. Phys. 89, 3410 ͑1988͒.
9 D. P. Greene and A. W. McCown, Appl. Phys. Lett. 54, 1965 ͑1989͒.
10 A. W. McCown, D. P. Greene, J. H. Schloss, and J. G. Eden, Paper ThA6,
Conference on Lasers and Electro-Optics ’88 ͑CLEO ’88͒, Anaheim, CA
͑April, 1988͒.
IV. SUMMARY AND CONCLUSIONS
Photodissociation of the 4 2⌫ state of Kr2F has been
studied in the 280–850 nm wavelength region by fluores-
cence suppression and laser excitation spectroscopy. Bands
attributed to the 9 2⌫←4 2⌫ ͑␭maxϳ320 nm͒ and 6 2⌫←4 2⌫
͑␭maxϳ710 nm͒ transitions of the molecule have been ob-
served, the latter for the first time. Weaker features lying in
the blue ͑420–490 nm͒ and orange ͑ϳ600 nm͒ regions of the
spectrum were also detected and tentatively assigned to tran-
sitions terminating on the 8 2⌫ and 7 2⌫ states, respectively,
of the triatomic.
11 A. W. McCown, J. Phys. Chem. 94, 8913 ͑1990͒.
12 K. Hakuta, S. Miki, and H. Takuma, J. Opt. Soc. Am. B5, 1261 ͑1988͒.
13 D. L. Huestis, G. Marowsky, and F. K. Tittel, in Excimer Lasers, 2nd ed.,
edited by C. K. Rhodes ͑Springer, Berlin, 1984͒, pp. 181–215.
14 J. G. Eden, R. S. F. Chang, and L. J. Palumbo, IEEE J. Quantum Electron.
QE-15, 1146 ͑1979͒.
15 J. G. Eden, R. W. Waynant, S. K. Searles, and R. Burnham, Appl. Phys.
Lett. 32, 733 ͑1978͒.
16 P. J. Hay and T. H. Dunning, J. Chem. Phys. 66, 1306 ͑1977͒.
J. Chem. Phys., Vol. 106, No. 13, 1 April 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:
141.100.74.200 On: Wed, 26 Nov 2014 08:40:22