Transcript KLi_Ohio2007.ppt

An analytical potential for the for the a 3S+ state of KLi,
(derived from observations of the upper vibrational levels only)
Houssam Salami, Amanda Ross, Patrick Crozet
Université Lyon 1 & CNRS .
Robert J. Le Roy
University of Waterloo
Pawel Kowalczyk
Warsaw University
Wlodzimierz Jastrzebski
Polish Academy of Sciences, Warsaw
Motivation Reliable description of molecular levels close to atomic asymptotes
Looking for Born-Oppenheimer breakdown, K6Li / K7Li
K + Li
S+P
P+S
Cold atom
photoassociation
occurs R > 20 Å
S+S
Cold molecules ?
X 1S+ limit accessible via BX or AX fluorescence. But not a 3S+ .
How to reach the triplet manifold in KLi ?
1) Obvious Experiment
Pump A/b mixed levels
produced K2
2) Reliable Experiment
2 known (Warsaw) B/c
mixed levels
Gave B-X KLi, NaK
3) Others? All PE curves
are close at large R.
A priori, via B-X, high v’
Ab initio PE curves, from M. Frécon, associated with S. Rousseau, et al Chem. Phys. 247, (1999) 193-199.
Why did we have to resort to random searches ?
laser
Molecules formed in
the KLi heatpipe
IR beam completely absorbed by K2 vapour, doesn’t excite KLi
KLi spectrum is quite congested ~ 18300 cm-1
LIF does not discriminate transitions as well as the 2 photon polarisation labelling technique
Method : detect ~ 6000 cm-1 below laser (optical filter), then record spectrum
Pierced mirror to
reflect fluorescence
Gas phase
sample
Laser tuned to
chosen transition
~ Doppler
resolution
l
meter
cw
laser
Laser : SP 380D.
FTS
FTS : Bomem DA3. Source : heatpipe oven
First observation, laser = 18263.57 cm-1 R(16) 10-1, 39K6Li
Collisionally induced spectrum, no link with any known state!
To a 3S+
To X 1S+
Mixed B1P/triplet levels found in 39K6Li, v’ 20, 5 J  12
Combination differences match in B – X and B-a systems …
To v = 41, X state
?
Spectra establish an energy scale. vmax is 7 cm-1 below Li(2s) + K(4s) limit
Q(J) to X 1S+,
NOT enough information,
Found ONLY for K6Li.
The largest DGv in LIF
matches
the smallest DGv in CIF
Largest DGv ~ 20 cm-1,
Ab initio (Rousseau) we 47.3
So vmin(obs)  0
OP, SR,
Q (N’=J) lines
to 4 levels of a 3S+
Assigning vibrational quantum numbers
No isotopic data available.
No obvious match with ab initio predictions. DGv  vmin 5
Bv  vmin 4
Franck-Condon Factors ?
Requires V(R) for upper and lower state.
B1P – a 3S+ ?
Attempted RKR curves for possible vmin
Pointwise potential curve
Analytical potential, MLJ from DPotFit program
(results not suitable for this audience)
Modified Morse/Long-range potential form
Dispersion coefficients have been calculated.
C6 =1.119 x 107 cm-1Å6
Derevianko, Babb, Dalgarno, Physical Review A 63, (2001) 052704.
C8 =2.63 x 108 cm-1Å8
Porsev, Derevianko J. Chem. Phys. 119 (2003) 844
Ground state dissociation energy is known
De (X, K7Li) is known = 6217.37(5) cm-1
Martin et al J. Chem. Phys. 115 (2003) 4118
  u LR ( R )  MLR ( R ) y p ( R ) 
VMLR ( R )  De 1   u ( R )  e

  LR e 

with imposed long-range form :
u LR 
C6
C8
 8
R
R
6
* Model published : R. Le Roy, D. Henderson, Mol. Phys. 105, 663 2007
2
*
The other variables ?
yp (R) 
( R p  R ep )
p > 2 imposed by Cn constraints
( R p  R ep )

NS( or L )
 
MLR R   1  y p R 
i 0
i y p R i  y p R 
OPTIMIZE Re, NS, NL, p, i for possible values of Vmin
Larger p  larger number of  coefficients for given rms deviation
Data and Results
J
149 data, now includes
some isolated bands
Fit to experimental
uncertainty, 0.01 cm-1
RMS error is not an
adequate diagnostic
Options for Vibrational Numbering : clusters of similar results.
vmin=4 absurd
vmin = 5 closest to
Ab initio results :
De = 273 cm-1
Re = 4.97 Å
vmin = 6 few solutions
Concentrate on vmin = 5
Spectroscopic parameters from V(R) eliminate some options immediately
Find average Re from acceptable
solutions, fix, and re-fit data
Fitting parameters for V(R) of the a 3S+ state, KLi
149 data, rms error 0.009 cm-1
MLR takes
p=3
Ns = 1
NL = 6
Comparison with homonuclear alkalis
7Li
2
23Na
39K
2
23Na39K
De
Potential fit
(cm-1 ) ab initio
Harmonic mean
333.69a
173.65b
252.74g
207.86d
197.e
205.9
287. ± 4
273.f
287.6
we
65.4a
24.47h
21.71d
22.99i
22.63
23.09
44.2 ± 1.5
43.97f
43.56
2
39K7Li
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Re
Potential fit
4.173a
5.166b
5.773c
5.462d
4.99±0.09
(Å)
ab initio
5.45e
4.974f
Arithmetic mean
5.469
4.973
Potential fit
(cm-1 ) ab initio
Harmonic mean
Conclusion: Sensible solution with few observed levels
Binding Energy
(cm-1)
27.5 %
100
De = 287 cm-1
200
5
7
9
R (Ångström)
Possible to incorporate simultaneous fit to X and a.
Leads to De = 289 (2) cm-1, Re = 5.107 A.
v,J observed in X state of K6Li, in rectangle, also with transitions to a3S+