Collisional mixing among the z3DJ and z3FJ states of Fe atoms in He and Ar

Article abstract of DOI:10.1063/1.472575

Collisional mixing among the z³D_J and z³F_J states of Fe[3d⁶4s(a⁴D)4p] atoms was investigated in He and Ar by laser-induced fluorescence method. The z³D_J and z³F_J states of Fe atoms were generated directly by photodissociation of Fe(CO)₅ followed by single photon absorption within a laser pulse using an unfocussed laser beam with atomic transition frequencies of Fe. When the z³D₃ level was excited, the emissions from this level showed a double exponential decay. The fast and slow components of the decay constants from the z³D₃ level were 10.7×10^-10 and 0.3×10^-10 cm³ molecule^-1 s^-1 in He, and 8.8×10^-10 and 1.6×10^-10 cm³ molecule^-1 s^-1 in Ar, respectively. When the z³F₄ level was pumped, the emissions from this level showed a single exponential decay and the decay constants were the same as those of the slow components of z³D₃. The emissions from higher-lying levels were single exponential at low pressures and the decay constants were in the range of 0.7-3.6 ×10^-10 cm³ molecule^-1 s^-1. It is found that the collisional mixing between the Z³D₃ and z³F₄ levels is very fast in both buffer gases while the mixing among the higher-lying four levels is relatively slow. The radiative lifetimes of the z³D_J and z₃F_J levels were 280-370 and 770-1100 ns, respectively, depending on J. Kinetic simulations of time profiles from the laser excited and collisional product levels revealed that intermultiplet mixing appeared to be more efficient than intramultiplet mixing.

Full text of DOI:10.1063/1.472575

Collisional mixing among the z 3 D J and z 3 F J states of Fe atoms in He and Ar
J. S. Goo, K. Lee, S. C. Bae, and J. K. Ku
Citation: The Journal of Chemical Physics 105, 7485 (1996); doi: 10.1063/1.472575
View online: http://dx.doi.org/10.1063/1.472575
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/105/17?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Collisionenergy/electronenergy resolved twodimensional study of Penning ionization of Ar by He metastable
atoms 23 S and 21 S
J. Chem. Phys. 105, 7536 (1996); 10.1063/1.472579
Destruction cross sections for low energy collisions of H+ 3 and D+ 3 with rare gas atoms
J. Chem. Phys. 104, 6149 (1996); 10.1063/1.471280
Collisional energy spread in hotatom reactions caused by thermal motions of the reagents
AIP Conf. Proc. 216, 301 (1990); 10.1063/1.39925
Satellite of forbidden line Ar (4s3P2–1S0) induced by collisions with helium atoms
AIP Conf. Proc. 216, 287 (1990); 10.1063/1.39919
Noble gas induced collisional line broadening of atomic Li Rydberg states nS and nD (n=4 to 30) measured by
trilevel echoes
AIP Conf. Proc. 146, 439 (1986); 10.1063/1.35714
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

Collisional mixing among the z3DJ and z3FJ states of Fe atoms in He
and Ar
J. S. Goo, K. Lee, S. C. Bae, and J. K. Ku
Department of Chemistry, Pohang University of Science and Technology, San 31 Hyoja-Dong, Pohang,
Kyungbuk 790-784, Korea
͑Received 6 May 1996; accepted 24 July 1996͒
Collisional mixing among the z3DJ and z3FJ states of Fe[3d64s(a4D)4p] atoms was investigated
in He and Ar by laser-induced fluorescence method. The z3DJ and z3FJ states of Fe atoms were
generated directly by photodissociation of Fe͑CO͒ followed by single photon absorption within a
5
laser pulse using an unfocussed laser beam with atomic transition frequencies of Fe. When the z3D3
level was excited, the emissions from this level showed a double exponential decay. The fast and
slow components of the decay constants from the z3D3 level were 10.7ϫ10Ϫ10 and 0.3ϫ10Ϫ10
cm3 moleculeϪ1 sϪ1 in He, and 8.8ϫ10Ϫ10 and 1.6ϫ10Ϫ10 cm3 moleculeϪ1 sϪ1 in Ar, respectively.
When the z3F4 level was pumped, the emissions from this level showed a single exponential decay
and the decay constants were the same as those of the slow components of z3D3 . The emissions
from higher-lying levels were single exponential at low pressures and the decay constants were in
the range of 0.7–3.6ϫ10Ϫ10 cm3 moleculeϪ1 sϪ1. It is found that the collisional mixing between the
z3D3 and z3F4 levels is very fast in both buffer gases while the mixing among the higher-lying four
levels is relatively slow. The radiative lifetimes of the z3DJ and z3FJ levels were 280–370 and
770–1100 ns, respectively, depending on J. Kinetic simulations of time profiles from the laser
excited and collisional product levels revealed that intermultiplet mixing appeared to be more
efficient than intramultiplet mixing. © 1996 American Institute of Physics.
͓S0021-9606͑96͒01241-X͔
I. INTRODUCTION
cies of Fe in the 315–325 nm region at low laser pulse en-
ergies and the emissions from z3DJ→a3FJ
and
Ј
z3FJ→a3FJ transitions are easy to observe; ͑ii͒ the radia-
It is well known that the refractive metal atoms in the
gas phase can be generated by multiphoton dissociation
͑MPD͒ of metal carbonyls or organometallic compounds.1–6
However, most of the MPD studies were done using a fo-
cused laser beam, and various excited states of metal atoms
were found to be generated simultaneously under these ex-
perimental conditions.1–3 Since the product states are not
unique when a focused beam is used for MPD of metal car-
bonyls or organometallic compounds, it is difficult to study
kinetic behavior of a state-selected excited state of metal
atoms with a one-color laser system.
Ј
tive lifetimes of these states are expected to be relatively
long due to the small Einstein A coefficients8 so that they are
good candidates for studying kinetic behaviors; and ͑iii͒ both
the z3D and z3F terms arise from the same electronic con-
figuration [3d64s(a4D)4p] with similar energies but differ-
ent orbital angular momentum. It is found that the collisional
mixing between the z3D3 and z3F4 levels is very efficient,
and kinetic simulations are needed to determine the colli-
sional mixing rate constants and the radiative lifetimes of
these levels. However, the collisional mixing among the
other higher-lying four levels are not so fast as the lower two
levels and the radiative lifetimes of these levels are deter-
mined directly from the Stern–Volmer plots obtained from
analyzing low pressure time profiles.
Recently, we have reported that a state-selected excited
state of Fe atoms can easily be generated by MPD of
Fe͑CO͒ using an unfocused weak UV laser pulse whose
5
frequency matches exactly to the corresponding Fe atomic
7
*
transition line. The efficiency of Fe formation depends on
the frequency of the atomic transition and pulse energy of
the laser. In general, the fluorescence from a state-selected
excited state of Fe atoms is easily observed when the corre-
sponding two-photon energies of a specified atomic transi-
tion line are about 3000 cmϪ1 in excess compared to the sum
of the five Fe–CO bond energies. The power dependence of
the fluorescence intensities revealed that the overall process
II. EXPERIMENTAL METHODS
A detailed experimental set up has been reported in
elsewhere.9 Briefly, the MPD/LIF cell was made of a 2l
Pyrex bulb and two pairs of 1 in. Pyrex O-ring joints were
attached to allow laser beam path and to connect to the gas
handling vacuum rack. The ports for the fluorescence detec-
tion were made by attaching 1.5 in. Pyrex tubings to the bulb
and cutting the arms with a glass saw as close as possible to
*
for the formation of the Fe atoms is a three-photon process.
In this work, we report the results of kinetic studies on
the behaviors of the z3DJ and z3FJ states of Fe atoms in He
and Ar. These states are chosen for the following reasons; ͑i͒
the bulb. The sample gas ͓0.5% Fe͑CO͒ in He or Ar͔ was
5
premixed and stored in a storage bulb in the vacuum rack.
The gas mixture was slowly flowed through the cell and the
flow rate ͑ϳ0.5 mmol/min͒ was controlled by adjusting the
they are easily formed by the MPD of Fe͑CO͒ using
5
a5DJ→z3DJ and a5DJ→z3FJ atomic transition frequen-
Ј
Ј
J. Chem. Phys. 105 (17), 1 November 1996
0021-9606/96/105(17)/7485/10/$10.00
© 1996 American Institute of Physics
7485
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

7486
Goo et al.: Mixing among the states of Fe
openings of the inlet needle valve and exit Teflon valve.
Then, the unfocused UV laser pulses ͑1.5–2.5 mJ/cm2͒ ob-
tained from the frequency doubling of the dye laser output
were passed through the cell. The linewidth of our laser͑Qu-
antel YG681-TDL 60 with NBP and DGO͒ was narrow
͑Ͻ0.1 cmϪ1͒ enough to excite the single spin–orbit state
selectively. The fluorescence from the excited Fe atoms was
detected at 90° from the laser beam direction through a 0.5 m
monochromator ͑Spex 1870C͒ equipped with a holographic
grating͑2400 grooves/mm͒ and a Hammamatsu R928 photo-
multiplier͑PM͒ tube. It was possible to monitor the near-
lying fluorescences from the z3D3 and z3F4 levels separately
by employing 200 ␮m of the slitwidth. Atomic emission
wavelengths from other levels were not so close that 300 ␮m
of the slitwidth was good enough to monitor the emissions
from the different spin–orbit levels. The signal from the PM
tube was fed into a transient digitizer ͑Tektronix 7912HB͒
and transferred to a laboratory computer for signal averaging
and storage. Small portions of the dye laser beam was di-
rected to a fast response photodiode and the signal from the
photodiode was used to trigger the digitizer. The Fe͑CO͒
5
was purchased from Aldrich, and transferred to a loading
vessel in a glovebox filled with dry N2. Then the sample was
further purified by freeze and thaw method on the vacuum
rack before making a mixture.
FIG. 1. Schematic energy level diagram for Fe͑a5DJ , a3FJ , z3FJ , and
z3DJ͒.
III. RESULTS
The decay curves for the z3D3→a3F4 emission from the
z3D3←a5D4 excitation in He and Ar were analyzed by a
double exponential fitting. The pressure dependence of the
decay rates of the fast component is plotted in Fig. 3͑a͒ and
that of the slow component is shown in Fig. 3͑b͒. Also
shown in Fig. 3͑b͒ is the pressure dependence of the decay
rates for the z3F4→a3F4 emission from the z3F4←a5D4
excitation analyzed by a single exponential fitting. It is inter-
esting that the pressure dependence of the decay rates of the
z3F4 level is virtually the same as that of the slow compo-
nent of the laser excited z3D3 level in both buffer gases.
Furthermore, the zero pressure intercepts for the fast and
slow components are the same within the experimental er-
rors. All the rate constants and the zero pressure intercepts
are shown in Table I.
A. Kinetics between the z3D3 and z3F4 levels
Schematic energy levels of the relevant states in this
work are shown in Fig. 1. When the laser frequency was
tuned to the z3D3←a5D4 transition line of the Fe atom and
directed to the cell containing a mixture of 0.5% Fe͑CO͒ in
5
He/Ar, the fluorescences from the z3D3 and z3F4 levels were
observed. Typical time profiles for the z3D3→a3F4 and
z3F4→a3F4 emissions observed from the z3D3←a5D4 and
z3F4←a5D4 excitation at 2.0 Torr of a He mixture are plot-
ted in Fig. 2. Although the difference in the fluorescence
wavelengths for the z3D3→a3F4 and z3F4→a3F4 transi-
tions is small, the emissions can be effectively separated by
using 200 ␮m slitwidth of the monochromator. The separa-
tion of the fluorescences from these levels was confirmed
from the different shapes in the rise times of the time profiles
for the laser excited and product levels. When the z3D3 level
was populated, the fluorescence from the laser excited level
was strong and clearly showed a double exponential decay.
The emission from the z3F4 level was weak and appeared to
be a single exponential decay as shown in Fig. 2͑a͒. When
the z3F4 level was populated, the fluorescence from the laser
excited level was weak compared to that from the z3D3 level
as shown in Fig. 2͑b͒. The same trend was observed in an Ar
mixture. Since the energy difference between the z3D3 and
z3F4 levels is about 15 cmϪ1, collisional mixing between the
two levels appears to be very efficient. Also, the strong emis-
sions from the z3D3 level when the z3F4 level was excited
suggest that the radiative lifetime of the z3D3 level is much
shorter than that of the z3F4 level.
We have attempted to find out the physical meaning of
the experimental rate constants as well as the zero pressure
intercepts by fitting this system to the well-known three-level
kinetic scheme,10 which is described by the following Eq.
͑1͒:
K
2
Aϭz3D3 ꢀ z3F4ϭB
K
Ϫ2
K1↓
↓K3
a3F4
͑1͒
K1ϭ1/␶AϩkQA M ,
͓
͔
J. Chem. Phys., Vol. 105, No. 17, 1 November 1996
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

Goo et al.: Mixing among the states of Fe
7487
FIG. 2. Typical time profiles obtained from 2.0 Torr of 0.5% Fe͑CO͒ in
5
FIG. 3. The pressure dependence of the the decay constants in He͑᭝–᭝͒
and in Ar͑᭺–᭺͒ for ͑a͒ the fast and ͑b͒ the slow components when the z3D3
level was excited. Also shown in ͑b͒ are the pressure dependence of the
single exponential decay constants of the z3F4 emissions in Ar͑᭹–᭹͒ and
in He͑᭡–᭡͒ when the z3F4 level was excited.
He: ͑a͒ emissions from the z3D3 ͑᭹–᭹͒ and z3F4͑᭡–᭡͒ levels when the
z3D3 level was excited, and ͑b͒ those from the z3D3͑᭹–᭹͒ and z3F4͑᭡–᭡͒
levels when the z3F4 level was laser excited. Only one from five data points
are plotted for the decay part of the time profile.
K2ϭkAB M ,
͓
͔
The standard solution to the coupled equations is
K
Ϫ2ϭkBA M ,
͓ ͔
Ϫ1
t
␭ ϩK ϩK ͒eϪ␭
ϩ
A t͒ ϭ A 0͒ ␭ Ϫ␭ ͒
͓ ͑ ͓ ͑ ͔͑
͔
͓͑
ϩ
Ϫ
Ϫ
1
2
K3ϭ1/␶BϩkQB M ,
͓
͔
ϩ ␭ ϪK ϪK ͒eϪ␭
,
t
Ϫ
͑
͔
ϩ
1
2
␶Aϭradiative lifetime for the z3D3 ,
␶Bϭradiative lifetime for the z3F4 ,
kQAϭquenching rate constant for the z3D3 by M,
kQBϭquenching rate constant for the z3F4 by M,
kABϭmixing rate constant from the z3D3
to the z3F4 ,
Ϫ1
t
t
ϩ
B t͒ ϭ A 0͒ K ␭ Ϫ␭ ͒
͓ ͑ ͓ ͑
eϪ␭Ϫ ϪeϪ␭
,
͑3͒
͔
͔
͑
͓
͔
2
ϩ
Ϫ
where the two decay constants are given by Eq. ͑4͒,
1
2
␭Ϯϭ
K ϩK ϩK ϩK ͒Ϯ K ϩK ϪKϪ2ϪK3͒
͓͑
͑
͕
1
2
Ϫ2
3
1
2
2
1/2
ϩ4K2KϪ2
.
͑4͒
͔
͖
The square root term in Eq. ͑4͒ can be rewritten as Eq. ͑5͒,
kBAϭmixing rate constant from the z3F4
to the z3D3 ,
K ϩK ϪKϪ2ϪK3͒2ϩ4K2KϪ2
1/2
͓͑
͔
1
2
2
ϭ
K ϩK ͒2ϩ K ϪK ͒
Mϭbuffer gas.
͓͑
͑
2
Ϫ2
1
3
1/2
The coupled differential equations for the above scheme can
ϩ2 K ϪK ͒ K ϪK
͒
͔
.
Ϫ2
͑5͒
͑
͑
1
3
2
be written as Eq. ͑2͒,
If (K2ϩKϪ2)2ӷ(K1ϪK3)2ϩ2(K1ϪK3)(K2ϪKϪ2), the
second term of Eq. ͑4͒ can be approximated as Eq. ͑6͒,
d A
͓ ͔
ϭϪ K ϩK ͒ A ϩK B ,
͓ ͔ ͓ ͔
Ϫ2
͑
1
2
dt
K ϩK ϪKϪ2K3͒2ϩ4K2KϪ2 1/2ϷK2ϩKϪ2
.
͑6͒
͓͑
͔
1
2
d B
͓ ͔
ϭK2 A Ϫ K ϩK ͒ B .
͓ ͔ ͓ ͔
͑2͒
͑
Ϫ2
3
dt
Then, the decay constants can be reduced to Eq. ͑7͒,
J. Chem. Phys., Vol. 105, No. 17, 1 November 1996
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

7488
Goo et al.: Mixing among the states of Fe
TABLE I. Measured rate constants from the Stern–Volmer analysis for z3DJ and z3FJ levels of Fe.
Buffer
gas
Laser excited
level
Quenching rate
constantsa
Intercepts
Radiative
lifetimes ͑ns͒
͑ϫ106 sϪ1
͒
b
He
z3D3
z3F4
10.7Ϯ0.3
0.30Ϯ0.05
0.30Ϯ0.05
2.0Ϯ0.5
2.0Ϯ0.2
2.0Ϯ0.2
͑370Ϯ20͒
fast comp.
slow comp.
coupled
b
͑770Ϯ100͒
decay
z3D2
z3D1
z3F3
z3F2
z3D3
2.3Ϯ0.2
3.0Ϯ0.3
3.6Ϯ0.3
2.0Ϯ0.3
8.8Ϯ0.3
1.6Ϯ0.4
1.6Ϯ0.4
2.7Ϯ0.3
3.6Ϯ0.3
1.4Ϯ0.4
0.9Ϯ0.2
2.2Ϯ0.6
1.9Ϯ0.3
2.0Ϯ0.3
370Ϯ40
280Ϯ30
700Ϯ200
1100Ϯ200
b
Ar
͑370Ϯ20͒
fast comp.
slow comp.
coupled
z3F4
͑770Ϯ100͒
b
decay
z3D2
z3D1
z3F3
z3F2
1.2Ϯ0.2
1.8Ϯ0.2
2.3Ϯ0.5
0.7Ϯ0.2
3.1Ϯ0.3
3.6Ϯ0.2
1.1Ϯ0.3
0.9Ϯ0.2
320Ϯ40
280Ϯ30
900Ϯ200
1100Ϯ200
aUnits, 10Ϫ10 cm3 moleculeϪ1 sϪ1
.
bAssigned from kinetic simulations of time profiles.
1
kABϭ6.1ϫ10Ϫ9͑He͒; 4.0ϫ10Ϫ9͑Ar͒ cm3 moleculeϪ1sϪ1
,
␭ Х K ϩK ͒ϩ
K ϩK ͒
1 3
͑
͑
ϩ
2
Ϫ2
2
kBAϭ4.5ϫ10Ϫ9͑He͒; 2.8ϫ10Ϫ9͑Ar͒ cm3 moleculeϪ1 sϪ1
kQAϭkQBϭ0.3ϫ10Ϫ10͑He͒;
,
1
1
2
1
1
ϭ
kABϩkBA͒ϩ
kQAϩkQB
͒
M ϩ
ϩ
,
͑
͑
͑
͓
͔
ͫ
ͬ
ͩ
ͪ
2
␶
␶
B
A
1.6ϫ10Ϫ10 Ar͒cm3 moleculeϪ1 sϪ1
,
͑
1
1
2
␭ϪХ
K ϩK ͒ϭ
kQAϩkQB͒] M
͑
͓
͔
␶Aϭ370Ϯ20 ns,
␶Bϭ770Ϯ100 ns.
1
3
2
͑9͒
1
2
1
1
ϩ
ϩ
.
͑7͒
ͩ
ͪ
Figure 4 shows the comparison of the experimental and cal-
␶
␶
B
A
Now the physical meanings of the slopes and intercepts ob-
tained from Fig. 3 can be interpreted from Eq. ͑7͒. It is clear
that the zero pressure intercepts for the fast and slow com-
ponents should be the same and they are the average value of
the lifetimes of the coupled levels. Also, the slope for the fast
component of the z3D3 level corresponds to the sum of the
forward and backward mixing rate constants plus the average
value of the purely quenching rate constants of the two lev-
els. The latter is also obtained from the slope of the slow
component. It is evident that kinetic simulations of the time
profiles are needed to determine the radiative lifetimes of the
z3D3 and z3F4 levels and collisional mixing rate constants,
because they cannot be determined directly from the Stern–
Volmer analysis. Since the experimental rate constants from
the Stern–Volmer analysis and the equilibrium constant ex-
pression given by Eq. ͑8͒, provide good boundary conditions,
it was possible to extract the radiative lifetimes of these lev-
els and mixing rate constants by kinetic simulations of the
time profiles,
KeqϭkBA/kAB
ϭ gA/gB͒exp Ϫ⌬E/kT͒ϭ0.724 at 300 K.
͑8͒
͑
͑
FIG. 4. Comparison of the experimental ͑•••͒ and simulated ͑—͒ time pro-
files for ͑a͒ z3D3→a3F4 and ͑b͒ z3F4→a3F4 emissions at three different
pressures when the z3D3 level was excited. ͑a͒, ͑d͒ 1.0 Torr; ͑b͒, ͑e͒ 2.0
To fit the shapes and relative intensities of time profiles from
both z3D3 and z3F4 levels, the following sets of rate con-
stants were assigned in He and Ar, respectively
Torr; and ͑c͒, ͑f͒ 4.0 Torr of 0.5% Fe͑CO͒ in He. Only one from five data
5
points were plotted for the decay part of the experimental time profiles.
J. Chem. Phys., Vol. 105, No. 17, 1 November 1996
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

Goo et al.: Mixing among the states of Fe
7489
culated time profiles at three different pressures using the
above rate constants. Because of the small energy difference
between the z3D3 and z3F4 levels, the quenching rate con-
stants for both levels were assumed to be the same. Note that
the radiative lifetime of the z3D3 level has to be assigned
much shorter than that of the z3F4 level to fit the relative
emission intensities of the time profiles shown in Fig. 2. It is
interesting that the collisional mixing between these two lev-
els is more efficient in He than in Ar. While less than 10% of
the experimentally measured apparent quenching rate con-
stants are assigned to other states in He, about 30%–40% of
the apparent quenching rate constants are assigned to other
states in Ar to fit the shapes of experimental time profiles.
B. Kinetics of the higher-lying z3D2,1 and z3F3,2 levels
1. Radiative lifetimes and apparent quenching
rate constants
The higher-lying four levels were excited at
319.70(z3D2←a5D3), 320.08(z3D1←a5D2), 318.49(z3F3
←a5D3), and 318.08 nm(z3F2←a5D2), and the emis-
sions
were
monitored
at
522.72(z3D2→a3F3),
527.04(z3D1→a3F2), 519.49(z3F3→a3F3), and 521.63
nm(z3F2→a3F2), respectively, as shown in Fig. 1. The ex-
citation wavelengths in the 318–320 nm region were chosen
for convenience since the fluorescence from the z3D2,1 and
z3D3,2 were strong enough to investigate kinetic behavior
when these levels were generated from the a5D3,2 levels. The
time profiles from these higher-lying levels showed apparent
single exponential decay at low pressures but they revealed
double or even triple exponential decay behavior at higher
pressures. Since the reliable analysis of the time profiles ob-
tained at higher pressures was difficult, we concentrated our
efforts to obtain good time profiles at low pressures ͑Ͻ2
Torr͒ and analyzed them by a single exponential fitting. The
pressure dependence of the decay constants below 2 Torr of
total pressures is plotted in Fig. 5. The fluorescence from the
z3F3 and z3F2 levels was very weak at low pressures, and a
large number of signal averaging were needed to obtain ana-
lyzable time profiles.
FIG. 5. The pressure dependence of the the decay constants in He͑᭝–᭝͒
and in Ar͑᭺–᭺͒ for ͑a͒ z3D2 , ͑b͒ z3F3 , ͑c͒ z3D1 , and ͑d͒ z3F2 levels.
cm3 moleculeϪ1 sϪ1, and collisional mixing among these lev-
els is expected. Indeed, collisional mixing among these lev-
els is identified by taking time-resolved fluorescence spectra.
Because of the relatively long radiative lifetimes of these
levels, the time-resolved fluorescence spectra were taken at
low pressures to get informations on the primary collisional
product levels. A typical time-resolved fluorescence spectra
obtained from the z3F3 excitation at 1.0 Torr of 0.5%
Fe͑CO͒5 in Ar are shown in Fig. 6. During the 0–16 ns time
period, only the z3F3→a3F3 emission peak appears at 519.5
nm, but emissions from other levels appear at later time pe-
riods. Since the radiative lifetimes of the z3DJ levels are
much shorter than that of the initially excited z3F3 level, the
emissions from the z3DJ levels are ascribed to the conse-
quence of the collisional mixing among these levels. Also,
all the z3DJ levels appear to be primary collisional products
with different branching fractions when the z3F3 level was
excited.
The apparent quenching rate constants as well as the
radiative lifetimes for these levels are shown in Table I. The
magnitudes of the apparent quenching rate constants are sub-
stantially larger in He͑2.0–3.6ϫ10Ϫ10 cm3 moleculeϪ1 sϪ1
͒
than in Ar͑0.7–1.8ϫ10Ϫ10 cm3 moleculeϪ1 sϪ1͒. The
quenching rate constants for the z3F3 and z3D1 levels are
substantially larger than those for the z3F2 and z3D2 levels
in both buffer gases probably due to the near-lying energy
levels. The radiative lifetimes obtained from the two buffer
gases agree well within the experimental errors, however, the
radiative lifetimes from the Ar buffer gas appear to be more
reliable than those from the He medium due to the small
quenching rate constants.
3. Time profiles from the laser excited and product
levels
Since the emissions from the collisional product levels
were observed, we attempted to find out the magnitude of
intra- and intermultiplet mixing rate constants as well as the
collisional branching fractions. For this purpose, we have
collected fluorescence time profiles from the laser excited
and collisional product levels by pumping each of the spin–
orbit levels at low pressures. Typical time profiles from the
laser excited and product levels in He and Ar are plotted in
2. Identification of collisional primary product levels
The apparent quenching rate constants for these levels
are not as large as those for the lower-lying z3D3 and z3F4
levels. However, they are still quite large, 0.7–3.6ϫ10Ϫ10
J. Chem. Phys., Vol. 105, No. 17, 1 November 1996
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

7490
Goo et al.: Mixing among the states of Fe
FIG. 6. Time resolved fluorescence spectra from z3F3 excitations; ͑a͒ 0–16
͑b͒ 16–32, ͑c͒ 32–62, ͑d͒ 62–92, ͑e͒ 92–122, and ͑f͒ 122–152 ns periods.
Figs. 7 and 8, respectively, which were obtained at 2.0 Torr
of total pressure. The kinetic behaviors observed from each
level excitation are described below.
FIG. 7. Comparison of experimental z3D2͑᭿–᭿͒, z3D3͑छ–छ͒,
z3D1͑᭺–᭺͒, z3F3͑᭡–᭡͒, z3F2͑᭝–᭝͒, and calculated͑—͒ time profiles in
He. The experimental time profiles are plotted one from every five data
points to avoid congestion. The emission intensities of the laser excited
z3D2 and z3D1 levels in ͑a͒ and ͑c͒ were reduced to 40% of their experi-
mental intensities.
a. Kinetics of z3D2 : The z3D2 levels locates 118 cmϪ1
below the z3F3 level and fast mixing between these two
levels is expected. When the z3D2 level was excited in He
and Ar, emissions from the z3D3 , z3D1 , and z3F3 levels
were observed as shown in Figs. 7͑a͒ and 8͑a͒, respectively.
Note that the emission from the z3D3 located 364 cmϪ1 be-
low the laser excited level was much stronger than that from
z3F3 and z3D1 levels. The shape and relative intensity of the
time profile from the z3D3 level in He are much different
form those in Ar although the rise times are similar. About
four times stronger peak intensity and much slower decay of
the z3D3 emission in He is consistent with the strong cou-
pling between the z3D3 and z3F4 levels as described in Sec.
III A. Much faster decay and weaker intensity of the z3D3
emission in Ar is ascribed to the weaker coupling between
the z3D3 and z3F4 levels and larger quenching rate constants
by Ar. The weak emission intensities from the z3F3 level
might be due to the much smaller Einstein A coefficient for
the z3F3→a3F3 transition ͑Aikϭ2.9ϫ105 sϪ1͒ than for
z3D3,2→a3F3,2 transition ͑Aikϭ2.0–2.5ϫ106 sϪ1͒ at the ob-
served wavelength.8
7͑b͒. The rise time of these product emissions is more or less
the same, but the peak intensities are in the order of
z3D2Хz3D3Ͼz3D1 . Also, the emission form the z3D3 de-
cays slowly compared to those from other levels. The slower
decay of the z3D3 emission might be ascribed to the tight
coupling with the z3F4 level in He. When the z3F3 level was
excited in Ar using the same laser power, the relative emis-
sion intensities are in the order of z3D2Ͼz3D1уz3D3 as
shown in Fig. 8͑b͒.
c. Kinetics of z3D1 : The z3D1 level locates 133 and
251 cmϪ1 above the z3F3 and z3D2 , respectively, and 196
cmϪ1 below the z3F2 . When this level was populated in He
and Ar, collisional product emissions from the z3D2 and
z3F3 levels were easily observed, but the emission from the
z3F2 level was very weak and not plotted. The emission
from the z3D2 level was much stronger than that from the
z3F3 as shown in Figs. 7͑c͒ and 8͑c͒. The rise and decay of
the z3D2 emission appeared to be somewhat than those of
the z3F3 in both buffer gases. The delayed appearance of the
z3D2 emission suggests that this level might be affected by
secondary collisional processes.
b. Kinetics of z3F3 : The z3F3 level locates 118 cmϪ1
above the z3D2 and 133 cmϪ1 below the z3D1 level. Since
the energy differences from these near-lying levels are small,
fast collisional mixing among them is expected. When the
z3F3 level was excited at 2.0 Torr of total pressure in He,
collisional product emissions from z3D1,2,3 levels were even
stronger than the directly pumped level as plotted in Fig.
d. Kinetics of z3F2 : The z3F2 level locates 196 and 329
J. Chem. Phys., Vol. 105, No. 17, 1 November 1996
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

Goo et al.: Mixing among the states of Fe
7491
FIG. 9. Kinetic scheme for the calculations of time profiles. ͑k1 , arbitrary
absorption rate constant; k2 –k7 , radiative rate constants; k8 –k29, inter- and
intramultiplet mixing rate constants; and k30–k35, quenching rate constants
to other states.͒
Thus, we set up more complicated kinetic scheme shown in
Fig. 9, which includes 6 species and 35 rate constants. When
the initially excited level, primary and secondary collisional
product levels are denoted as i, n, and m, respectively, the
time dependence of the initially excited and primary level
populations can be written as Eq. ͑9͒,
d N /dtϭF t͒Ϫ N 1/␶iϩ
k ϩkq
M
͓
͓
͔
͑
͓
͔
͔
͔
͚
i
i
in
ͩ
ͪ
i
ͭ
ͮ
n
FIG. 8. Comparison of experimental z3D2͑᭿–᭿͒, z3D3͑छ–छ͒,
z3D1͑᭺–᭺͒, z3F3͑᭡–᭡͒, z3F2͑᭝–᭝͒, and calculated͑—͒ time profiles in
Ar. The experimental time profiles are plotted one from every five data
points to avoid conjestions. The emission intensities of the laser excited
z3D2 and z3D1 levels in ͑a͒ and ͑c͒ were reduced to 10% of their experi-
mental intensities. The intensity of the z3D3 emission was magnified by a
factor of 2 to avoid congestion.
ϩ
k
M
N
ϩ
k
M
N
k
,
͑9a͒
͓
͔͓
͔
͓
͔͓
͚
͚
ni
n
mi
m
n
m
d N /dtϭk
N
M Ϫ N 1/␶ ϩ k ϩ
nmϩkq
n
͓
͔
͓
͔͓
͔
͓
͔
͓
͚
n
in
i
n
n
ni
ͩ
ͪ
ͭ
m
ϫ M
ϩ
k
M
N
͓
͔
͔͓
͔
͚
mn
m
cmϪ1 above the z3D1 and z3F3 levels, respectively. When
this level was excited in He and Ar, emissions from the
z3D1 , z3F3 , and z3D2 levels were observed. The emission
from z3D2 level lying 463 cmϪ1 below was stronger than
other emissions but appeared with time delay as shown in
Figs. 7͑d͒ and 8͑d͒. The emission from the z3D2 level might
also be affected by secondary collisional processes as for the
z3D1 excitation.
ͮ
m
ϩ
k
M
N
.
͔
n
Ј
͑9b͒
͓
͔͓
͚
n n
Ј
n
Ј
The F(t) is the formation rate of the species by the laser
pulse, ␶ is the radiative lifetime, kjp represents collisional
mixing rate constant from j to p level belonging to the z3D
and z3F term, kq is the quenching rate constant to other
states and subscript nЈ denotes other near-lying primary lev-
els. Since the intensities of the experimentally observed time
profiles depend on the magnitude of Einstein A coefficient at
the observed wavelength, and since the Einstein A coeffi-
cient for each level at the observed wavelength shows large
differences, conversion of I(t) to N(t) was needed to com-
pare not only the shapes but also the relative intensities of
the experimental and calculated time profiles. Thus, those
experimentally obtained time profiles [I(t)] were divided by
the Einstein coefficients Aik to convert populations, where
Aik is the Einstein A coefficient for the observed line.8 The
assignment of state-to-state collisional mixing/quenching
rate constants was based on the apparent total quenching rate
4. Kinetic simulations and assignment of collisional
branching fractions
Although the collisional energy transfer processes
among the z3D2,1 and z3F3,2 levels in He and Ar seemed to
be very complicated, those kinetic behaviors of each level
described above provided some clue for assigning collisional
branching fractions among these levels by kinetic simula-
tions of the time profiles. When one of these levels was
populated by a laser pulse, emissions from all the near-lying
levels appeared. Also, preliminary calculations showed that
those time profiles from the collisional product levels could
not fit satisfactorily using two-level coupling scheme.10
J. Chem. Phys., Vol. 105, No. 17, 1 November 1996
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

7492
Goo et al.: Mixing among the states of Fe
TABLE II. Rate constants used for kinetic simulation and branching fractions ͑⌫͒ for z3DJ and z3FJ levels of Fe in He and in Ar.
c
Rate
⌬Eb
␴
Process
constantsa
⌫
͑cmϪ1
͒
͉
⌬L
͉
͉
⌬J
͉
͑Å2͒
z3D3ϩHe→z3F4
61
0.93
0.01
Ͻ0.01
0.05
0.36
0.52
0.03
0.02
0.06
0.70
0.12
0.12
0.06
0.91
0.01
Ͻ0.01
0.07
0.39
0.19
0.36
0.03
0.42
0.37
0.29
0.02
15
Ϫ364
Ϫ482
1
0
1
1
1
0
46.9
0.6
0.5
2.5
6.9
10.0
0.6
0.5
1.2
18.5
3.1
3.1
1.5
34.6
0.5
0.3
2.5
9.2
4.6
8.5
0.8
6.2
5.4
4.2
0.4
→z3D2
→z3F3
0.8
0.6
3.3
9.0
13.0
0.8
0.6
1.5
24
4.0
4.0
2.0
45
0.6
0.4
3.3
12
6.0
11
→others
z3D2ϩHe→z3F3
Ϫ118
364,379
Ϫ251
1
0,1
0
1
1,2
1
→z3D3 ,z3F4
→z3D1
→z3F2
Ϫ447
1
0
→others
z3D1ϩHe→z3F3
→z3F2
133
Ϫ196
251
1
1
0
2
1
1
→z3D2
→others
z3F4ϩHe→z3D3
→z3D2
Ϫ15
Ϫ379
Ϫ497
1
1
0
1
2
1
→z3F3
→others
z3F3ϩHe→z3D2
→z3D1
118
Ϫ133
482,497
1
1
1,0
1
2
0,1
→z3D3 ,z3F4
→others
z3F2ϩHe→z3D1
→z3F3
1.0
8.0
7.0
5.5
0.5
196
329
447
1
0
1
1
1
0
→z3D2
→others
z3D3ϩAr→z3F4
40
0.68
Ͻ0.01
Ͻ0.01
0.31
0.26
0.32
0.03
0.02
0.37
0.54
0.10
0.11
0.25
0.60
Ͻ0.01
Ͻ0.01
0.39
0.16
0.09
0.24
0.50
0.28
0.18
0.28
0.26
15
Ϫ364
Ϫ482
1
0
1
1
1
0
76.9
0.6
0.4
34.6
6.3
7.7
0.8
0.4
9.0
16.3
2.9
3.5
7.5
53.8
0.4
0.4
34.6
7.7
4.4
11.5
23.0
3.5
→z3D2
0.3
0.2
18
z3F3
→others
z3D2ϩAr→z3F3
3.3
4.0
0.4
0.2
4.7
8.5
1.5
1.8
3.9
28
0.2
0.2
18
4.0
2.3
6.0
12
Ϫ118
364,379
Ϫ251
1
0,1
0
1
1,2
1
→z3D3 ,z3F4
3
→ D1
→z3F2
→others
Ϫ447
1
0
z3D1ϩAr→z3F3
→z3F2
133
Ϫ196
251
1
1
0
2
1
1
→z3D2
→others
z3F4ϩAr→z3D3
→z3D2
Ϫ15
Ϫ379
Ϫ497
1
1
0
1
2
1
→z3F3
→others
z3F3ϩAr→z3D2
→z3D1
118
Ϫ133
482,497
1
1
1,0
1
2
0,1
→z3D3 ,z3F4
→others
z3F2ϩAr→z3D1
→z3F3
1.8
1.2
1.8
1.7
196
329
447
1
0
1
1
1
0
2.3
3.5
3.3
→z3D2
→others
aUnits, 10Ϫ11 cm3 moleculeϪ1 sϪ1
.
b⌬EϭE͑initial level͒ϪE͑product level͒.
c
␴ϭk/
;
ϭ͑8kT/␲␮͒1/2ϭ1300 m/s ͑in He͒ and 520 m/s ͑in Ar͒ at 298 K.
v
͗ ͘ ͗ ͘
v
constant shown in Table I, and adjusted to match the shapes
and relative intensities of the time profiles from the laser
excited and collisional product levels. The solid lines in Figs.
7 and 8 are the calculated time profiles at 2.0 Torr of total
pressure, and state-to-state rate constants are given in Table
II. In fact, the state-to-state rate constants shown in Table II
were obtained from such a comparison of experimental and
simulated time profiles at various pressures. The branching
fraction ⌫ and cross section ␴ are calculated from the simple
relation,
⌫i, jϭkij /͚kijϭkij /kQ,
␴ ϭk / ϭk / 8kT/␲␮͒
͑10͒
͑11͒
1/2
,
v
͗ ͘
͑
ij
ij
ij
J. Chem. Phys., Vol. 105, No. 17, 1 November 1996
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

Goo et al.: Mixing among the states of Fe
7493
where kQ is the apparent total quenching rate constant for the
laser excited level, and ␮ is the reduced mass.
electron configuration, 3d64s(a4D)4p, so that the different
radiative lifetimes of the spin–orbit levels belonging to the
same term could be attributed to the same reason.
In general, experimentally observed emission intensities
from the z3F3,2 levels were much weaker than those from the
z3D1,2,3 due to the small Einstein A coefficients at the ob-
served wavelengths. Since the Fuhr and co-workers’ Einstein
A coefficients6 were used to correct emission intensities of
the time profiles, and since the branching fractions in Table
II were obtained from the shapes and the time integrated
emission intensities of the time profiles, some difference in
the Einstein A coefficients of the z3F3,2 levels may propagate
substantial difference in the branching fractions. However,
the branching fractions in Table II clearly show the follow-
ing features for the collisional mixing among the z3DJ and
z3FJ levels: ͑i͒ Collisional mixing among the z3DJ and z3FJ
levels is much more extensive in He than in Ar, although the
absolute quenching cross sections for the laser excited level
appear to be opposite. Collisional branching fractions to
other states are less than 10% in He, but they become 25%–
50% in Ar. ͑ii͒ Energy loss ͑⌬EϾ0͒ collisions are more
efficient than energy gain ͑⌬EϽ0͒ collisions in both buffer
gases. When the z3D2 level was excited, formation of the
z3D3 and z3F4 levels lying 364 and 379 cmϪ1 below, respec-
tively, appeared to be more favorable than that of z3F3 level
lying 118 cmϪ1 above. When the z3F3 level was excited,
formation of z3D3 and z3F4 levels ͑⌬EϷ490 cmϪ1͒ also
seemed to be more efficient than that of z3D1 lying 139
cmϪ1 above. ͑iii͒ Intermultiplet mixing ͑⌬LϭϮ1͒ appears to
be more efficient in both buffer gases due to the smaller
energy differences, although the observed product emissions
from the z3DJ were much stronger than those from the z3FJ .
It was found that the experimentally measured radiative
lifetimes of the z3DJ and z3FJ levels were shorter than the
radiative lifetimes calculated from the sum of Einstein A
coefficients tabulated in Ref. 8. To investigate these discrep-
ancies, we scanned the monochromator to look for any miss-
ing radiative transitions from the laser excited level. It was
very difficult to find out missing transitions for the z3FJ
levels due to the small Einstein A coefficients as well as the
fast collisional transfer to the near-lying z3DJ levels, how-
ever, we were able to identify some of the missing transitions
from the z3DJ levels. For instance, the z3D2→a3F3 transi-
tion at 522.7 nm gave the strongest fluorescence from the
z3D2 excitation but was missing in Ref. 8. When the emis-
sion intensity at 522.7 nm was converted to the Einstein A
coefficient comparing with those of other tabulated transi-
tions, the sum of A coefficients for the z3D2 level gave good
agreement with the experimental value. Thus, the individual
A coefficient tabulated in Ref. 8 seemed to be reliable even
though the table does not include every plausible transitions.
Since the branching fractions shown in Table II were ob-
tained from the kinetic simulations of time profiles, and since
the Einstein A coefficients in Ref. 8 were used to correct
emission intensities of the time profiles at the observed
wavelengths, the reliability of the branching fractions in
Table II should not be so bad.
The apparent quenching rate constants for the z3DJ and
z3FJ levels of the Fe atom by He and Ar in this work are
much larger than those previously reported values for the
low-lying excited states of group 1, 2, and 12 metal atoms
tabulated by Breckenridge and Umemoto.15 One of the rea-
sons for those small quenching cross sections tabulated in
Ref. 15 is that they do not include intramultiplet mixing pro-
cesses. In fact, large intramultiplet mixing cross sections in
collision with rare gases have been reported for the low-lying
excited state of Li͑22PJ͒, Na͑32PJ͒, K͑42PJ and 52PJ͒,
Rb͑92DJ͒, and Ca(4s5p3PJ), where the energy gap between
the intramultiplet levels is small ͑⌬EϽ60 cmϪ1͒.16–22 It has
also been reported that the intramultiplet mixing cross sec-
tions are small for low-lying excited states of Sr, Zn, Cd, and
Hg where the energy gap between the intramultiplet levels is
large ͑⌬EϾ180 cmϪ1͒.23–29 Parson and Ishikawa30 also stud-
ied intramultiplet mixing collisions of Cr(3d54p z7PJ) with
various gases. The intramultiplet splitting of Cr͑3d54p z7PJ)
are 81.3 and 112.5 cmϪ1, respectively. They obtained 11–18
Å2 for the intramultiplet mixing cross sections by He and Ar
at the effective temperature of 405 and 951 K, respectively.
The results of these collisional intramultiplet mixing studies
reveal that the apparent collisional quenching rate constants
depend upon the existence of near-lying energy states and
the shapes of interaction potentials. Although the intramul-
tiplet splitting energies for the z3DJ and z3FJ levels of Fe
atoms are not small as shown in Fig. 1, the relatively small
energy differences between the nearest intermultiplet levels
IV. DISCUSSION
The kinetics between the z3D3 and z3F4 levels are very
interesting. The energy difference between the two levels is
only 15 cmϪ1 and fast collisional mixing between them is
expected. Because of the large collisional mixing rate con-
stants between the two levels and the relatively long radia-
tive lifetimes as well as the small average quenching rate
constants of the two levels, the experimental rate constant for
the fast component of the z3D3 level is dominated by the
sum of the mixing rate constants between the two levels.
Similar kinetics were observed for the collisional coupling
and relaxation study of N2(B) and N2(W) vibrational
levels11 and the vibrational relaxation study of the a3B1 state
of SO2.12
It is also interesting that the radiative lifetimes of the
intramultiplet levels of the z3D term have a tendency of
getting shorter as the energy is higher, and those of the z3F
term show an opposite trend. A significantly different radia-
tive lifetimes of intramultiplet levels have also been reported
for the z3P term of an Fe atom by Figger et al.13 Marek and
co-workers14 reinvestigated the radiative lifetimes of the
z3PJ and obtained 108, 94, and 37 ns for Jϭ0, 1, and 2,
respectively. They attributed their observations on the z3PJ
levels to the inappropriate LS-coupling scheme for the z3P
state. The z3F, z3D, and z3P states arise from the same
*
seemed to provide complicated Fe –He/Ar interaction po-
J. Chem. Phys., Vol. 105, No. 17, 1 November 1996
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02

7494
Goo et al.: Mixing among the states of Fe
tentials leading to large collisional quenching of the z3DJ
and z3FJ levels.
1 Z. Karny, R. Naaman, and R. N. Zare, Chem. Phys. Lett. 59, 33 ͑1978͒.
2 D. P. Gerrity, L. J. Rothberg, and V. Vaida, Chem. Phys. Lett. 74, 1
͑1980͒; J. Phys. Chem. 87, 2222 ͑1983͒.
3 G. W. Tyndall and R. L. Jackson, J. Chem. Phys. 89, 1364 ͑1988͒.
4 P. C. Engelking, Chem. Phys. Lett. 74, 207 ͑1980͒.
5 S. Leutwyler, U. Even, and J. Jortner, Chem. Phys. Lett. 74, 11 ͑1980͒.
6 S. A. Mitchell and P. A. Hackett, J. Chem. Phys. 79, 4815 ͑1983͒; Chem.
Phys. Lett. 107, 508 ͑1984͒; J. Chem. Phys. 93, 7813 ͑1990͒.
7 K. Lee, J. S. Goo, and J. K. Ku, Chem. Phys. Lett. 244, 213 ͑1995͒.
8 J. R. Fuhr, G. A. Martin, and W. L. Wiese, J. Phys. Chem. Ref. Data 17,
Suppl. 4 ͑1988͒.
V. CONCLUSION
We have studied collisional mixing among the z3DJ and
z3FJ states of Fe atoms in He and Ar. These excited states of
Fe atoms can be easily formed by two-photon dissociation of
Fe͑CO͒ followed by one-photon absorption within a laser
5
9 K. Lee, J. S. Goo, and J. K. Ku, Chem. Phys. Lett. 216, 483 ͑1993͒.
10 J. T. Yardley, Introduction to the Molecular Energy Transfer ͑Academic,
New York, 1980͒.
pulse using an unfocused laser beam with atomic transition
frequencies of Fe. The radiative lifetimes of z3D3 and z3F4
levels were obtained from kinetic simulations of the time
profiles using three level kinetic scheme since the Stern–
Volmer analysis of the decay frequencies gave an arithmetic
mean value of the two levels due to strong collisional cou-
pling between the levels. The radiative lifetimes of the
higher-lying four levels were obtained directly from the pres-
sure dependence of the decay frequencies. Collisional mix-
ing among the higher-lying levels was also investigated by
monitoring the time profiles of the laser excited and colli-
sional product levels. State-to-state mixing rate constants
were assigned from kinetic simulations of the time profiles.
Although the state-to-state collisional branching fractions as-
signed in this work is based on the Fuhr and co-workers’
Einstein A coefficients, it clearly shows that intermultiplet
mixing is more efficient than intramultiplet mixing for this
system.
11 N. Sadeghi and D. W. Setser, J. Chem. Phys. 79, 2710 ͑1983͒.
12 S. C. Bae, K. Lee, G. H. Kim, and J. K. Ku, J. Chem. Phys. 102, 1665
͑1995͒.
13 H. Figger, J. Heldt, K. Siomos, and H. Walther, Astron. Astrophys. 43,
389 ͑1975͒.
14 J. Marek, J. Richter, and H. J. Stanhke, Phys. Scr. 19, 325 ͑1979͒.
15 W. H. Breckenridge and H. Umemoto, Adv. Chem. Phys. 50, 325 ͑1982͒.
16 J. Elward-Berry and M. J. Berry, J. Chem. Phys. 72, 4500 ͑1980͒.
17 L. Krause, Adv. Chem. Phys. 28, 267 ͑1975͒.
18 C. B. Ke, S. H. Chou, K. C. Lin, and W. T. Luh, J. Phys. Chem. 97, 604
͑1993͒.
19 K. C. Wang, K. C. Lin, and W. T. Luh, J. Chem. Phys. 96, 349 ͑1992͒.
20 A. Spielfiedel, D. Gilvert, E. Roueff, and F. Lostas, J. Phys. B 12, 3693
͑1979͒.
21 T. R. Mallory, W. Kedzierski, J. B. Atkins, and L. Krause, Phys. Rev. A
38, 5917 ͑1988͒.
22 M. O. Hale and S. R. Leone, J. Chem. Phys. 79, 3352 ͑1983͒.
23 J. F. Kelly, M. Harris, and A. Gallagher, Phys. Rev. A 37, 2354 ͑1988͒.
24 N. P. Penkin and T. P. Redko, Opt. Spectrosc. 57, 598 ͑1984͒.
25 H. Umemoto, T. Ohnuma, and A. Masaki, Chem. Phys. Lett. 171, 357
͑1990͒.
26 H. Umemoto, J. Kikuma, A. Masaki, T. Takayanagi, and S. Sato, Chem.
Phys. 134, 193 ͑1989͒.
ACKNOWLEDGMENTS
27 H. Umemoto, A. Masaki, T. Ohnuma, T. Takayanagi, S. Sato, F. Misaizu,
and K. Fuke, J. Chem. Phys. 93, 4112 ͑1990͒.
This work is financially supported in part by KOSEF
through the Center for Molecular Science at KAIST and in
part by the Ministry of Education through Basic Science Re-
search Institute Program ͑BSRI-95-3438͒.
28 X. L. Han and J. F. Kelly, J. Chem. Phys. 94, 5481 ͑1991͒.
29 N. Bras, A. El Afghani, J. Butaux, J. C. Jeannet, and D. Perrin, J. Chem.
Phys. 91, 4142 ͑1989͒.
30 J. M. Parson and T. Ishikawa, J. Chem. Phys. 80, 3137 ͑1984͒.
J. Chem. Phys., Vol. 105, No. 17, 1 November 1996
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:
131.193.242.46 On: Fri, 28 Nov 2014 19:57:02