- Email: [email protected]
Narrow-bandwidth VUV laser measurements of fine-structure predissociation linewidths in the Schumann-Runge bands of O2
Journal of Electron Spectroscopy and Related Phenomena 80 (1996) 29-32
Narrow-bandwidth
VUV laser measurements
of fine-structure
p r e d i s s o c i a t i o n l i n e w i d t h s in t h e S c h u m a n n - R u n g e
bands of 02
P. M. Dooley, B. R. Lewis, S. T. Gibson, and K. G. H. Baldwin Research School of Physical Sciences and Engineering, The Australian National University, Canberra ACT 0200, Australia Fine-structure predissociation iinewidths have been measured for the Schumann-Runge bands o f O2 B 3 E ~ X '3x2] ). Using two excimer-pumped dye lasers, radiation was generated in the 1800 A region by four-wave mixing in Xe, achieving a bandwidth of-,~ 0.10 c m - 1 full-width at half-maximum, and resolving lines to the Doppler limit. The predissociation linewidths are found to vary significantly with vibration, rotation and fine-structure. A predissociation model which includes spin-orbit interactions between the B a ~ , state and the repulsive 1 1II,,, 2 3E+ and 1 '5IL, states, and spin-orbit and L-uncoupling interactions with the repulsive 1 alia, state, provides a good description of the measurements.
1. I N T R O D U C T I O N The
Schumann-Runge (SR) system of Oe, X 3 ~ - , plays an important role in the photochemistry of the terrestrial atmosphere. It controls the depth of penetration of solar vacuum ultraviolet (VUV) radiation, and predissociation of the discrete levels of the B aE~ state is an important source of energetic oxygen atoms in the atmosphere. Accurate knowledge of the predissociation linewidths of the SR bands (1750 - 2050 A) is essential for photochemical modelling of the upper atmosphere. B 3,,,[
~
"o\\ \ s°
' ,is . . . .
2'.°. . . . L . . . . 3; R (A)
Figure 1. Potential-energy curves of 02 relevant to the B 3 2~-state predissociation. Energies are given relative to the minimum in the X aE)- ground state potential-energy curve. 0368-2048/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S0368-2048 (96) 02915-5
The SR bands have been the subject of much study; the many spectroscopic investigations have been reviewed by Yoshino et al. [I], while Lewis et aL [2] have recently reviewed work on the B 3E~-_state predissociation. Figure 1 shows potential-energy curves lbr the B state and the other states which perturb the B state leading to its predissociation. The B-state curve is a Rydberg-Klein-Rees (RKR) potential determined by Lewis et al. [3], while the repulsive curves are the a b initio calculations of Partridge et al. [4] (1 3II~,, 1 51]u)and Partridge [5] ( 1 lIIu, 23E + ). The four repulsive states shown in Fig. 1 perturb the B state through spin-orbit interactions and the 1 3II~, state also perturbs the B state via spin-electronic and L-uncoupling mechanisms [6,7]. These interactions cause predissociation of every vibrational level and are expected to produce irregular perturbations in the spectroscopic and triplet-splitting constants of the B state. Although the SR predissociation rates are expected to vary significantly with the triplet fine-structure level [2], very few measurements of fine-structureresolved predissociation linewidths have been reported [2,8]. The linewidths for the broader B-state levels with v -- 3 - 8 are comparable with, or exceed, the fine-structure splittings, making it difficult to determine fine-structure widths, due to line overlap. However, most of the narrower levels, with v > 12, can be resolved with an experiment of sufficient resolving power. In this paper, following a suggestion by Lewis et al. [2], we present preliminary results of a
30 comprehensive study in which fine-structure-specific linewidths were measured for the B-state levels with '~' -- 9 - 18.
20.0 15.o 10.0
l ~ Xe*5ps2P~ 2560.1 2560.~ 3900-4800A~-- Xe5ps(2p~)6p[5/212
LU 5.0
0.o
by a second PMT. The output pulses from the PMTs were processed by a boxcar gated integrator and the signals were averaged over fifty laser pulses. Each wavelength scan was done three times, with the absorption cell evacuated during the second scan to enable background correction. The detector signal was divided by the monitor signal to account for shot to shot variations, and slow drifts were compensated for by averaging the first and third (full) scans. The full-cell ratios were divided by the empty-cell ratios to give absolute cell transmittances and photoabsorption cross sections were calculated from the Beer-Lambert law.
Xe5p~'SO |
Figure 2. Xe energy level diagram, showing the four-wave mixing scheme.
~t Gated
Detector
I
OxygenCell
-~tT~o-19SOA
~ /Microcomputerl
1 E,0e"meot
Radiation was generated using two-photonresonant four-wave difference-frequency mixing (FWDFM) in Xe [9]. The corresponding energylevel diagram is shown in Fig. 2 and the experimental set up is shown in Fig. 3. The two lasers used were Lambda Physik FL3002 pulsed dye lasers pumped by an EMG201 Lambda Physik excimer laser. The output of one dye laser was frequency doubled in a BBO I crystal to reach the 6p[5/212 two-photon resonance in Xe, while the other laser was scanned between 3900 ,~ and 4800/~, producing tunable VUV radiation between 1750 & and 1950/~. The two beams were focussed into the Xe cell using an off-axis lens of nominal focal length 25 cm. FWDFM can be phasematched for either positive or negative dispersive media, and so a Xe pressure could always be chosen (usually in the range 70 - 90 Torr) to optimise phase-matching for any VUV wavelength. After the Xe cell the light passed through a 0.2 m monochromator which discriminated against non-VUV light. The VUV radiation was then divided by a beamsplitter. The reflected radiation was directed into a solar-blind photomultiplier tube (PMT) to monitor the incident intensity, while the transmitted beam passed through the absorption cell (which was filled with 02 to pressures between 0.4 and 700 Torr) and was detected
d
Monitor I
l Controer 2. E X P E R I M E N T A L M E T H O D
i ~
DyeLaser
] X:ClmerLaser~
3900-4800A I
]3o,o'
°'aLaser
non
It
~
I
0.2mMonochromator
.
,,
I °e"
c: ~ f =25cm ~
Figure 3. Experimental apparatus
Use of intracavity 6talons reduced the nominal dyelaser bandwidth to 0.04 cm -1, but it was found that the bandwidth of the resultant four-wave mixing signal (incorporating five fundamental dye-laser photons) was ~ 0.10 cm -1 full-width at half-maximum (FWHM). This value varied significantly (±30%) from day to day, so it was necessary to monitor the bandwidth regularly using narrow reference lines. The instrumental bandwidth was deduced daily from a measurement of one of the reference lines, using a Voigt line-profile fitting program with the Doppler and predissociation linewidth components fixed at the known values. Two reference lines were used during this study. Their predissociation linewidths were determined using curve-of-growth analysis [10], which is independent of the instrumental bandwidth, and were found to be 0.051±0.007 cm -1 FWHM for the (14,0) /~1 (21 ) line and 0.069+0.010 c m - 1 FWHM for
31 the (16,0) Rt(23) line. Absolute wavenumber calibration was achieved by comparison with the measured wavenumbers of Yoshino et al. [ 1] for selected sharp, unblended lines of the SR system. 3. P R E D I S S O C I A T I O N
MODEL
The predissociation model used in this work is based on that developed by Julienne and Krauss [6] and Julienne [7] and optimised by Lewis et al. [2]. Briefly, the model employs a RKR potential for the/3 state and represents the four repulsive states in Fig. 1 as exponential potentials. Predissociation linewidths are calculated using the Golden rule. This results in a thirteen parameter model: four matrix elements for the spin-orbit interaction of the/3 3 ~ - state with each of the repulsive i 111~, 2 3x~+~,, 1 3II~, and 1 sII,, states; one for the L-uncoupling interaction with the 1 zII,, state, and eight parameters for the crossing points and slopes of the four repulsive states. The matrix element for the L-uncoupling interaction is J-dependent and is expected to produce significant effects at high rotation.
between the fine-structure components is clearly visible. Lorentzian linewidths were deduced from such scans by fitting instrumentally-degraded Voigt profiles to each line, enabling specific account to be taken of the Doppler component (~ 0.012 c m - l FWHM). It was found that the Lorentzian component of the experimental linewidth had a pressure dependence due to collisional broadening. This was quantified by measuring absorption lines from the t = 1 vibrational level of the X 3 B ] ground state of O~ ("hot bands") which were then compared with the corresponding lines from the v = 0 level of the ground state. As the population of the t, = 1 level of the ground state is much lower than that of the v = 0 level, measurements of a given upper-state level could be made in very different pressure regimes at similar absorbances. This work gave a value for the collisionbroadening coefficient of--, 0.2 cm-~/atm, confirming the result of Lewis et al. [ I 1], which was determined at much higher pressures. It was necessary to correct the Lorentzian linewidths determined at high pressures for this effect.
04
8.0
I
'
I
R,(17) E 0
6.0
2
4.0
'~ ' ' l
''''1''''1'
'''1''''1''
I
(13,0) B 3Eu<--X sZ9
o.3
_ I~ k_--~__ ~*
''1
....
3
-B .~,,(v=15)
I'
//F3
R2(17) • Measured/~
v cO
R3(17)
2.0
"~. £
~ ,,"
0.0
H
0
Instrumentalbandwidth
,,,I 0"00
~s821.0 52823.0 , ss82s0 Wavenumber (cm-') Figure 4. A representative experimental scan showing the measured photoabsorption cross section for the ( 1:3, 0) R(17) fine-structure triplet.
4. RESULTS Figure 4 shows a representative experimental scan of the (13, 0)/~(17) fine-structure triplet. The finestructure is well resolved, and the difference in width
5
................. 110 115
,,,I 210
;5
.... 30
I,, 35
N Figure 5. Measured and calculated fine-structurespecific predissociation linewidths as a function of rotation for the v = 15 level of the/3' 3E~S state of 02.
Figure 5 shows preliminary results for the finestructure-specific linewidths of the (15,0) band as a function of rotation. These are the first comprehensive fine-structure-resolved measurements for this band and they exhibit strong dependence on N and finestructure. It is particularly notable that the difference
32 between the F1 and Fa linewidths increases with rotation. This effect is specifically dependent on the interference term involving the vibronic L-uncoupling (7/) and spin-orbit (~) interactions with the 13II~, state [2]. The results in Fig. 5 indicate that r/~ > 0. Figure 5 also shows good agreement between our experimental results and the predictions of the semiempirical predissociation model. The model has been optimised to our preliminary results for v = 9 - 18 and the results of Lewis et al. [2] for v = 0 - 8, which were based on the measurements of Cosby et al. [8] and Yoshino et al. [12]. The model parameters were varied only slightly from previous values [2] to obtain this agreement, the most significant difference being an improved value for the ratio q/~ of 0.025 compared with the previous value of 0.019 [2].
1.5
\.
I
'
\
'
I
/~
/
\ tI
~1.0 "1-
I
• F3-F1 measured - F2 calculated
\
~ / /
I
,~
....
REFERENCES
F3_F, calculated -
l.
,,
~
Preliminary results are presented from the first detailed study of fine-structure-specific SR predissociation linewidths with'v ~ > 2, measured using a narrowbandwidth VUV laser source. We also present calculations based on the predissociation model of Lewis et al. [2] with slightly modified parameters, which are in good agreement with the experimental measurements presented. Our new value for 71/~ is 0.025, significantly higher than the previous value of 0.019 [2]. The measurements presented here, and further measurements in progress, are being undertaken as part of a collaborative project involving groups from SRI International, California, the HarvardSmithsonian Centre tbr Astrophysics, Massachusetts and The Photon Factory, Tsukuba, in which predissociation linewidths obtained by three distinct narrowbandwidth techniques (laser-based photoabsorption spectroscopy, laser-induced fluorescence spectroscopy, and VUV Fourier-transform spectroscopy) will be critically compared.
• F2 measured
\
/
'
5. C O N C L U S I O N S
0.5
0.0
"'"
8
,
""
1
10
,
I 12
,
I 14
,
I 16
V
Figure 6. Vibrational dependence of the N = 20 F2 and F3 - F1 predissociation linewidths.
Figure 6 shows the vibration dependence of the F2 predissociation linewidths and the F3 - F1 differencelinewidths for N=20. For this rotation, the B 3E ustate coupling is close to Hund's case (b) and the F3 - F1 widths are dominated by the smoothly varying ~1~ cross term associated with the inner-crossing 1 3II~, state. Conversely, the F2 widths are dominated by spin-orbit contributions, including those from the outer-limb crossings of the B state which result in rapid oscillation of linewidth with v. Figure 6 also shows the predictions of the present model which are in excellent agreement with the measurements.
K. Yoshino, D. E. Freeman, and W. H. Parkinson, J. Phys. Chem. Ref. Data, 13 (1984) 207. 2. B . R . Lewis, S. T. Gibson, and E M. Dooley, J. Chem. Phys., 100 (1994) 7012. 3. B . R . Lewis, L. Berzins, J. H. Carver, and S. T. Gibson, J. Quant. Spectrosc. Radiat. Transfer, 36 (1986) 187. 4. H. Partridge, C. W. Bauschlicher, S. R. Langhoff, and E R. Taylor, J. Chem. Phys., 95 (1991) 8292. 5. H. Partridge, (private communication, 1993). 6. E S. Julienne and M. Krauss, J. Mol. Spectrosc., 56 (1975) 270. 7. E S. Julienne, J. Mol. Spectrosc., 63 (1976) 60. 8. E C. Cosby, H. Park, R. A. Copeland, and T. G. Slanger, J. Chem. Phys., 98 (1993)5117. 9. R. Hilbig and R. Wallenstein, IEEE J. Quantum Electron., QE- 19 (1983) 194. 10. I. M. Vardavas, J. Quant. Spectrosc. Radiat. Transfer, 49 (1993) 119. 11. B. R. Lewis, L. Berzins, C. J. Dedman, T. T. Scholz, and J. H. Carver, J. Quant. Spectrosc. Radiat. Transfer, 39 (1988) 271. 12. K. Yoshino, J. R. Esmond, A. S.-C. Cheung, D. E. Freeman, and W. H. Parkinson, Planet. Space Sci., 40 (1992) 185.
Narrow-bandwidth
VUV laser measurements
of fine-structure
p r e d i s s o c i a t i o n l i n e w i d t h s in t h e S c h u m a n n - R u n g e
bands of 02
P. M. Dooley, B. R. Lewis, S. T. Gibson, and K. G. H. Baldwin Research School of Physical Sciences and Engineering, The Australian National University, Canberra ACT 0200, Australia Fine-structure predissociation iinewidths have been measured for the Schumann-Runge bands o f O2 B 3 E ~ X '3x2] ). Using two excimer-pumped dye lasers, radiation was generated in the 1800 A region by four-wave mixing in Xe, achieving a bandwidth of-,~ 0.10 c m - 1 full-width at half-maximum, and resolving lines to the Doppler limit. The predissociation linewidths are found to vary significantly with vibration, rotation and fine-structure. A predissociation model which includes spin-orbit interactions between the B a ~ , state and the repulsive 1 1II,,, 2 3E+ and 1 '5IL, states, and spin-orbit and L-uncoupling interactions with the repulsive 1 alia, state, provides a good description of the measurements.
1. I N T R O D U C T I O N The
Schumann-Runge (SR) system of Oe, X 3 ~ - , plays an important role in the photochemistry of the terrestrial atmosphere. It controls the depth of penetration of solar vacuum ultraviolet (VUV) radiation, and predissociation of the discrete levels of the B aE~ state is an important source of energetic oxygen atoms in the atmosphere. Accurate knowledge of the predissociation linewidths of the SR bands (1750 - 2050 A) is essential for photochemical modelling of the upper atmosphere. B 3,,,[
~
"o\\ \ s°
' ,is . . . .
2'.°. . . . L . . . . 3; R (A)
Figure 1. Potential-energy curves of 02 relevant to the B 3 2~-state predissociation. Energies are given relative to the minimum in the X aE)- ground state potential-energy curve. 0368-2048/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S0368-2048 (96) 02915-5
The SR bands have been the subject of much study; the many spectroscopic investigations have been reviewed by Yoshino et al. [I], while Lewis et aL [2] have recently reviewed work on the B 3E~-_state predissociation. Figure 1 shows potential-energy curves lbr the B state and the other states which perturb the B state leading to its predissociation. The B-state curve is a Rydberg-Klein-Rees (RKR) potential determined by Lewis et al. [3], while the repulsive curves are the a b initio calculations of Partridge et al. [4] (1 3II~,, 1 51]u)and Partridge [5] ( 1 lIIu, 23E + ). The four repulsive states shown in Fig. 1 perturb the B state through spin-orbit interactions and the 1 3II~, state also perturbs the B state via spin-electronic and L-uncoupling mechanisms [6,7]. These interactions cause predissociation of every vibrational level and are expected to produce irregular perturbations in the spectroscopic and triplet-splitting constants of the B state. Although the SR predissociation rates are expected to vary significantly with the triplet fine-structure level [2], very few measurements of fine-structureresolved predissociation linewidths have been reported [2,8]. The linewidths for the broader B-state levels with v -- 3 - 8 are comparable with, or exceed, the fine-structure splittings, making it difficult to determine fine-structure widths, due to line overlap. However, most of the narrower levels, with v > 12, can be resolved with an experiment of sufficient resolving power. In this paper, following a suggestion by Lewis et al. [2], we present preliminary results of a
30 comprehensive study in which fine-structure-specific linewidths were measured for the B-state levels with '~' -- 9 - 18.
20.0 15.o 10.0
l ~ Xe*5ps2P~ 2560.1 2560.~ 3900-4800A~-- Xe5ps(2p~)6p[5/212
LU 5.0
0.o
by a second PMT. The output pulses from the PMTs were processed by a boxcar gated integrator and the signals were averaged over fifty laser pulses. Each wavelength scan was done three times, with the absorption cell evacuated during the second scan to enable background correction. The detector signal was divided by the monitor signal to account for shot to shot variations, and slow drifts were compensated for by averaging the first and third (full) scans. The full-cell ratios were divided by the empty-cell ratios to give absolute cell transmittances and photoabsorption cross sections were calculated from the Beer-Lambert law.
Xe5p~'SO |
Figure 2. Xe energy level diagram, showing the four-wave mixing scheme.
~t Gated
Detector
I
OxygenCell
-~tT~o-19SOA
~ /Microcomputerl
1 E,0e"meot
Radiation was generated using two-photonresonant four-wave difference-frequency mixing (FWDFM) in Xe [9]. The corresponding energylevel diagram is shown in Fig. 2 and the experimental set up is shown in Fig. 3. The two lasers used were Lambda Physik FL3002 pulsed dye lasers pumped by an EMG201 Lambda Physik excimer laser. The output of one dye laser was frequency doubled in a BBO I crystal to reach the 6p[5/212 two-photon resonance in Xe, while the other laser was scanned between 3900 ,~ and 4800/~, producing tunable VUV radiation between 1750 & and 1950/~. The two beams were focussed into the Xe cell using an off-axis lens of nominal focal length 25 cm. FWDFM can be phasematched for either positive or negative dispersive media, and so a Xe pressure could always be chosen (usually in the range 70 - 90 Torr) to optimise phase-matching for any VUV wavelength. After the Xe cell the light passed through a 0.2 m monochromator which discriminated against non-VUV light. The VUV radiation was then divided by a beamsplitter. The reflected radiation was directed into a solar-blind photomultiplier tube (PMT) to monitor the incident intensity, while the transmitted beam passed through the absorption cell (which was filled with 02 to pressures between 0.4 and 700 Torr) and was detected
d
Monitor I
l Controer 2. E X P E R I M E N T A L M E T H O D
i ~
DyeLaser
] X:ClmerLaser~
3900-4800A I
]3o,o'
°'aLaser
non
It
~
I
0.2mMonochromator
.
,,
I °e"
c: ~ f =25cm ~
Figure 3. Experimental apparatus
Use of intracavity 6talons reduced the nominal dyelaser bandwidth to 0.04 cm -1, but it was found that the bandwidth of the resultant four-wave mixing signal (incorporating five fundamental dye-laser photons) was ~ 0.10 cm -1 full-width at half-maximum (FWHM). This value varied significantly (±30%) from day to day, so it was necessary to monitor the bandwidth regularly using narrow reference lines. The instrumental bandwidth was deduced daily from a measurement of one of the reference lines, using a Voigt line-profile fitting program with the Doppler and predissociation linewidth components fixed at the known values. Two reference lines were used during this study. Their predissociation linewidths were determined using curve-of-growth analysis [10], which is independent of the instrumental bandwidth, and were found to be 0.051±0.007 cm -1 FWHM for the (14,0) /~1 (21 ) line and 0.069+0.010 c m - 1 FWHM for
31 the (16,0) Rt(23) line. Absolute wavenumber calibration was achieved by comparison with the measured wavenumbers of Yoshino et al. [ 1] for selected sharp, unblended lines of the SR system. 3. P R E D I S S O C I A T I O N
MODEL
The predissociation model used in this work is based on that developed by Julienne and Krauss [6] and Julienne [7] and optimised by Lewis et al. [2]. Briefly, the model employs a RKR potential for the/3 state and represents the four repulsive states in Fig. 1 as exponential potentials. Predissociation linewidths are calculated using the Golden rule. This results in a thirteen parameter model: four matrix elements for the spin-orbit interaction of the/3 3 ~ - state with each of the repulsive i 111~, 2 3x~+~,, 1 3II~, and 1 sII,, states; one for the L-uncoupling interaction with the 1 zII,, state, and eight parameters for the crossing points and slopes of the four repulsive states. The matrix element for the L-uncoupling interaction is J-dependent and is expected to produce significant effects at high rotation.
between the fine-structure components is clearly visible. Lorentzian linewidths were deduced from such scans by fitting instrumentally-degraded Voigt profiles to each line, enabling specific account to be taken of the Doppler component (~ 0.012 c m - l FWHM). It was found that the Lorentzian component of the experimental linewidth had a pressure dependence due to collisional broadening. This was quantified by measuring absorption lines from the t = 1 vibrational level of the X 3 B ] ground state of O~ ("hot bands") which were then compared with the corresponding lines from the v = 0 level of the ground state. As the population of the t, = 1 level of the ground state is much lower than that of the v = 0 level, measurements of a given upper-state level could be made in very different pressure regimes at similar absorbances. This work gave a value for the collisionbroadening coefficient of--, 0.2 cm-~/atm, confirming the result of Lewis et al. [ I 1], which was determined at much higher pressures. It was necessary to correct the Lorentzian linewidths determined at high pressures for this effect.
04
8.0
I
'
I
R,(17) E 0
6.0
2
4.0
'~ ' ' l
''''1''''1'
'''1''''1''
I
(13,0) B 3Eu<--X sZ9
o.3
_ I~ k_--~__ ~*
''1
....
3
-B .~,,(v=15)
I'
//F3
R2(17) • Measured/~
v cO
R3(17)
2.0
"~. £
~ ,,"
0.0
H
0
Instrumentalbandwidth
,,,I 0"00
~s821.0 52823.0 , ss82s0 Wavenumber (cm-') Figure 4. A representative experimental scan showing the measured photoabsorption cross section for the ( 1:3, 0) R(17) fine-structure triplet.
4. RESULTS Figure 4 shows a representative experimental scan of the (13, 0)/~(17) fine-structure triplet. The finestructure is well resolved, and the difference in width
5
................. 110 115
,,,I 210
;5
.... 30
I,, 35
N Figure 5. Measured and calculated fine-structurespecific predissociation linewidths as a function of rotation for the v = 15 level of the/3' 3E~S state of 02.
Figure 5 shows preliminary results for the finestructure-specific linewidths of the (15,0) band as a function of rotation. These are the first comprehensive fine-structure-resolved measurements for this band and they exhibit strong dependence on N and finestructure. It is particularly notable that the difference
32 between the F1 and Fa linewidths increases with rotation. This effect is specifically dependent on the interference term involving the vibronic L-uncoupling (7/) and spin-orbit (~) interactions with the 13II~, state [2]. The results in Fig. 5 indicate that r/~ > 0. Figure 5 also shows good agreement between our experimental results and the predictions of the semiempirical predissociation model. The model has been optimised to our preliminary results for v = 9 - 18 and the results of Lewis et al. [2] for v = 0 - 8, which were based on the measurements of Cosby et al. [8] and Yoshino et al. [12]. The model parameters were varied only slightly from previous values [2] to obtain this agreement, the most significant difference being an improved value for the ratio q/~ of 0.025 compared with the previous value of 0.019 [2].
1.5
\.
I
'
\
'
I
/~
/
\ tI
~1.0 "1-
I
• F3-F1 measured - F2 calculated
\
~ / /
I
,~
....
REFERENCES
F3_F, calculated -
l.
,,
~
Preliminary results are presented from the first detailed study of fine-structure-specific SR predissociation linewidths with'v ~ > 2, measured using a narrowbandwidth VUV laser source. We also present calculations based on the predissociation model of Lewis et al. [2] with slightly modified parameters, which are in good agreement with the experimental measurements presented. Our new value for 71/~ is 0.025, significantly higher than the previous value of 0.019 [2]. The measurements presented here, and further measurements in progress, are being undertaken as part of a collaborative project involving groups from SRI International, California, the HarvardSmithsonian Centre tbr Astrophysics, Massachusetts and The Photon Factory, Tsukuba, in which predissociation linewidths obtained by three distinct narrowbandwidth techniques (laser-based photoabsorption spectroscopy, laser-induced fluorescence spectroscopy, and VUV Fourier-transform spectroscopy) will be critically compared.
• F2 measured
\
/
'
5. C O N C L U S I O N S
0.5
0.0
"'"
8
,
""
1
10
,
I 12
,
I 14
,
I 16
V
Figure 6. Vibrational dependence of the N = 20 F2 and F3 - F1 predissociation linewidths.
Figure 6 shows the vibration dependence of the F2 predissociation linewidths and the F3 - F1 differencelinewidths for N=20. For this rotation, the B 3E ustate coupling is close to Hund's case (b) and the F3 - F1 widths are dominated by the smoothly varying ~1~ cross term associated with the inner-crossing 1 3II~, state. Conversely, the F2 widths are dominated by spin-orbit contributions, including those from the outer-limb crossings of the B state which result in rapid oscillation of linewidth with v. Figure 6 also shows the predictions of the present model which are in excellent agreement with the measurements.
K. Yoshino, D. E. Freeman, and W. H. Parkinson, J. Phys. Chem. Ref. Data, 13 (1984) 207. 2. B . R . Lewis, S. T. Gibson, and E M. Dooley, J. Chem. Phys., 100 (1994) 7012. 3. B . R . Lewis, L. Berzins, J. H. Carver, and S. T. Gibson, J. Quant. Spectrosc. Radiat. Transfer, 36 (1986) 187. 4. H. Partridge, C. W. Bauschlicher, S. R. Langhoff, and E R. Taylor, J. Chem. Phys., 95 (1991) 8292. 5. H. Partridge, (private communication, 1993). 6. E S. Julienne and M. Krauss, J. Mol. Spectrosc., 56 (1975) 270. 7. E S. Julienne, J. Mol. Spectrosc., 63 (1976) 60. 8. E C. Cosby, H. Park, R. A. Copeland, and T. G. Slanger, J. Chem. Phys., 98 (1993)5117. 9. R. Hilbig and R. Wallenstein, IEEE J. Quantum Electron., QE- 19 (1983) 194. 10. I. M. Vardavas, J. Quant. Spectrosc. Radiat. Transfer, 49 (1993) 119. 11. B. R. Lewis, L. Berzins, C. J. Dedman, T. T. Scholz, and J. H. Carver, J. Quant. Spectrosc. Radiat. Transfer, 39 (1988) 271. 12. K. Yoshino, J. R. Esmond, A. S.-C. Cheung, D. E. Freeman, and W. H. Parkinson, Planet. Space Sci., 40 (1992) 185.