NEW RESULTS IN PHOTOIONIZATION OF
LASER-EXCITED ATOMS
F. J. Wuilleumier1,2, D. Cubaynes1,2, M. Meyer2, S. Canton3,
E. Kennedy4, J. Bozek3, J.-M. Bizau1,2 and N. Berrah5
1LIXAM,
UMR CNRS 8624 Universite Paris Sud, B. 350, 91405 Orsay, France
2LURE, Campus d'Orsay, B. 209d, 91405 Orsay, France
3Advanced Light Source, Lawrence Berkeley Laboratory, Berkeley, Ca 97405, USA
4Dublin City University, School of Physics, Glasnevin, Dublin 9, Ireland
5Western Michigan University, Physics Department, Kalamazoo, MI 40849
Abstract. The use of high-spectral resolution VUV-photon beams in atomic photoionization experiments
has significantly increased over the past few years, owing to the number of operating third generation synchrotron
radiation sources and associated high-resolution beam lines. For laser-excited atoms, however, the number of
experiments is still low, because of the difficulties in combining the use of these two widely-different photon
sources. In the following, we will present the results of high-resolution photoionization experiments performed by
combining a cw dye laser and the photon beam available at the 10.0.1 station of the Advanced Light Source.
INTRODUCTION
In the experiments describe hereafter, we performed
high-resolution measurements in excited atoms by
exploiting simultaneously the high-spectral resolution
of the ALS photon beams, the high-electron resolution
of the Scienta electron spectrometer and the use of a
cw laser beam.
More than twenty years ago, we have pionnered this
field in performing the first experiments [1] combining
laser and synchrotron radiation to study resonant
photoionization processes in an excited atom, and to
determine oscillator strengths of transitions to evenparity autoionizing states [2]. Later, we succeeded in
measuring a strong enhancement of partial cross
sections for multiple inner-shell photoionization
(photoionization accompanied by excitation) into the
continuum for excited 2p63p 2P3/2 sodium atoms [3].
EXPERIMENT
The experimental set up consisted of three main
parts: the laser system, the synchrotron radiation
source, and the vacuum chamber including the
electron spectrometer and the oven for the production
of the sodium beam. The laser and synchrotron
radiation beams counterpropagated along the direction
of the incident photon beam. Relatively to this
direction, the sodium beam was emitted
perpendicularly, and the electrons were analyzed at an
angle of 54°44’ relative to the polarization axis of
synchrotron radiation. For photoionization of atoms in
the ground state (randomly oriented atoms), this angle
is the magic angle, at which the differential cross
section is proportional to the absolute cross section. A
cw ring dye laser, pumped by an argon ion laser, was
used for pumping the atoms into the excited 3p states,
with an output power of routinely several hundreds of
milliwatts in single mode operation. The linear
polarization of the laser could be rotated continuously
with respect to the linear polarization of the
At about the same time, were started the studies of
innershell excitation processes from aligned or
oriented atoms, by absorption of polarized laser
radiaton followed by inner-shell ionization with
synchrotron radiation [4,5]. More recently [6], we
have studied the formation of triply-excited states in
resonant photoexcitation of laser-excited lithium
atoms, using high–resolution monochromatic photon
beams available at the Advanced Light Source (ALS).
Two-photon experiments have also probed short lived
excited atomic states by using a pulsed laser
synchronized to the synchrotron radiation pulsed [7].
In most of the photoionization experiments, a
relatively low-resolution analyzer (a cylindrical mirror
electron spectrometer, with an ultimate resolution of
about 100 meV) was used to measure the
photoelectrons emitted in photoionization processes.
CP680, Application of Accelerators in Research and Industry: 17th Int'l. Conference, edited by J. L. Duggan and I. L. Morgan
© 2003 American Institute of Physics 0-7354-0149-7/03/$20.00
124
Fig. 1 - Photoelectron spectra of sodium atoms in the
ground- and 2P3/2 excited states as measured at Super ACO
(upper panel, ref. 3) and at the ALS synchrotron radiation
source (lower panel, this work).
synchrotron radiation beam. Approximately 1% of the
laser radiation was used to excite sodium atoms in a
reference chamber to lock the laser to the atomic
transition by observing the resonance fluorescence.
We have used the high-resolution SCIENTA electron
spectrometer and the undulator photon beam available
at the 10.0.1 beamline of ALS over the 30-110 eV
photon energy range. Ultimate resolution (FWHM
including spectral width, Doppler effect, and
spectrometer resolution) was 13 meV. A resistively
heated oven has been built to produce a beam of Na
vapor. To populate significantly some specific excited
states of sodium atoms, the dye laser was successively
tuned to the wavelength of the 2p63s 2S1/2 → 2p63p
2P
2
1/2 or P3/2 transitions, respectively. Synchrotron
radiation
was
used
to
probe
inner-shell
photoionization in these laser-excited atomic states.
atoms in the ground state appear first (2p53s 3P2,1,0,
and 1P1 ionic states), then the main lines for
photoionization of atoms in the excited 2P3/2 state
(2p53p 3S1, 3D3,2,1, 1D2, 1P1, 3P2,01, 1S0). At higher
binding energies, one sees the interchannel coupling
(IC) 2p53p satellites produced in photoionization of
atoms in the ground state which correspond to the
same final states as the main lines in the excited state
(although relative intensities within the two manifolds
are strongly different as different pathways are
followed in either case). Then, the IC satellites from
the excited atom result from partly resolved transitions
to both 2p54s (∆l = - 1) and 2p53d (∆l = + 1) final
ionic states, offering the possibility to simultaneously
study transitions with a gain or a loss of one unit of
angular momentum. They are followed by the mostly
resolved shake up satellites from the excited state
(2p54p 3S, 3D, 1D, 1,3P, 1S, ∆l = 0), and from the
RESULTS
In the upper panel of Fig.1, we show, as a reference,
the low-resolution photoelectron spectrum resulting
from photoionization of sodium atoms in the ground
state and in the 3p 2P3/2 state, as measured by using
the synchrotron radiation emitted by the Super ACO
storage ring [3]. Since only 10 to 15 % of the atoms
are transferred into the excited state, most of the atoms
in the vapor are still in the ground state. Thus, the few
lines measured in this spectrum result from
photoionization of atoms in both ground and excited
states, and some treatment had to be applied to extract
the photoelectron spectrum originating merely from
the atoms in the excited state. In the lower panel, we
show the full photoelectron spectrum measured at
ALS when both synchrotron radiation and laser beams
illuminate the vapor. Most of the photoelectron lines
corresponding to the various final ionic states are fully
resolved. In order of increasing binding energies (from
right to left), the main lines for photoionization of
ground state (2p54s), and by higher-order (2p53d, ∆l =
2) satellites from the ground state. The resolution is
such that every group of lines is completely separated
one from each other. No further operation is needed to
determine the true photoelectron spectrum from the
excited state.
In Fig. 2, we show, on an magnified scale, the
photoelectron spectrum, recorded at 48 eV photon
energy, over the binding energy range corresponding
to the resolved 2p-main lines from the 2P3/2 excited
state (middle panel) and 2P1/2 (lower panel), and to the
IC satellites (upper panel) measured in photoionization
of sodium atoms in the ground state. The electronic
configuration of the residual ionic states we are
looking at is 2p53p in all cases, but several pathways
are followed to reach the same final states. The upper
pannel of Fig. 2 presents the photoelectron spectrum
corresponding only to these final ionic states detected
after photoionization of sodium atoms in the ground
state. From this initial state, the 2p53p final ionic states
are reached by correlation effects according to : 2p63s
2S
5
1,3L + εl. The total relative
1/2 + hν → 2p 3p
intensity of these unresolved interchannel coupling
satellites relatively to the main lines has already been
measured [8], but here, we are able to determine the
individual behavior of most of the components. One
can note that the 3S satellite dominates the spectrum,
and that the relative intensities of the lines within the
3D
manifold is distributed according to the
3,2,1
statistical ratio 7: 5: 3. In the middle panel of Fig. 2,
125
one can see the photoelectron spectrum (shifted along
the kinetic energy axis by the amount of energy (2.109
eV) brought by the laser) measured after single
photoionization of sodium atoms excited in the 2P3/2
state according to the pathway : 2p63s 2S1/2 + hνL
excited in the 2p63p 2P1/2 state. Again, there are strong
changes in the relative intensities. The 3D3 line has
completely vanished, as expected from dipole
selection rules. The 1D2 line can still be measured but
its relative intensity is strongly reduced. The oscillator
strength within the 3D manifold is redistributed among
the 3D1 and 3D2 final states which dominate now the
spectrum, together with the 1P1 state. The intensity of
the 1,3S components is severely reduced like in the
case of 2P3/2 optical excitation.
Although the experiments were performed at high
sodium vapor density (at least 1013 atoms/cm3), we
observed significant dichroism effects in the
photoelectron spectra measured with the laser
polarization axis either parallel or perpendicular to the
synchrotron polarization axis, respectively. An
intensity variation occurs when the atoms are optically
excited into the 2P3/2 state, demonstrating that at least
partial alignment of the atoms occurs in the vapor.
This linear dichroism is particularly clear for the
photoelectron lines corresponding to the final ionic
states having the 2p53p and 2p54p configurations,
although the variation of the relative intensities is not
strong enough to change the qualitative conclusion
given above. This suggest that radiation trapping is
weak in the volume seen by the electron spectrometer,
as compared to direct photoexcitation. This is
understandable when considering the very small
diameter of the interaction zone where the laser and
synchrotron beam overlap. In recent experiments [911] studying aligment effects, the fluorescence
photons measured to check the depolarization of the
target with increasing vapor densities were likely
emitted from a region larger than the interaction
volume. In further experiments, we will study in
details the laser-induced aligment effects on the
photoelectron lines, including the satellite lines. The
data measured following inner-shell photoionization of
atoms laser excited to the 2P1/2 state are free from any
dichroism effect [12].
As a second example of the potentialities of twophoton experiments performed at high resolution, we
have been able to determine some lifetimes of evenand odd- parity strongly autoionizing states. Tuning
the laser to the wavelength of 3s 2S1/2 + hνL → 3p
2P
variation, as a
3/2 transition, we measured the
function of the photon energy, of the number of
electrons emitted in the decay of several of 2p53s3p
resonantly excited states. In Fig. 3, we show the result
of a scan obtained in measuring the decay of the
[2p5(3s3p 2P3/2)] 2D5/2 state excited near 31.40 eV
photon energy. Preliminary analysis of the decay curve
was made by fitting the experimental profile to a Voigt
Fig. 2 - Photoelectron spectra showing the inter-channel
coupling satellite lines for sodium atoms in the ground state
(upper panel), and the main lines emitted from sodium atoms
laser-excited to the 2p63p 2P3/2 (middle panel) and 2P1/2
(lower panel) (with a vertical offset), respectively.
→ 2p63p 2P3/2, and 2p63p 2P3/2 + hνRS → 2p53p 1,3L
+ εl’. The same photoelectron lines are observed, but
with relative intensities differing considerably. The
3D line fully dominates the spectrum, while the
3
intensity of the 1,3S lines is strongly reduced. The
relative intensity of the 3D1 and 3D2 lines is very
weak, far from the value of the statistical ratio. The
1,3P and 1D lines do not show significant change in
2
relative intensity as compared to the 3D3 line. In the
lower pannel of Fig. 2, one sees the photolectron lines
measured after single ionization of the sodium atoms
126
profile resulting from the convolution of a Gaussian
shape instrumental profile with a Lorentzian natural
profile. We obtained a value of the natural witdh of the
autoionizing state that is a little lower than 1 meV,
corresponding to a long lifetime of several hundreds of
femtosecondes. More accurate values will be obtained
in further experiments, but this long lifetime suggests
that time-resolved probe of transitions induced by a
Fig. 4 – Shake up photoelectron lines (2p54p final ionic
configuration) emitted in 2p-inner shell photoionization from
the 2p63p 2P1/2 (full line) and 2P3/2 (dashed line) states,
respectively.
the main lines (Fig.2, middle panel). In both cases, the
relative intensity of the 1,3S states is also quite
reduced. Thus, as it could be expected, we observed
that the satellite spectra are almost a faithful image of
the main lines, with similar transfer of the oscillator
stength when compared to the IC satellites from the
ground state atoms. This results shows that the
measured enhancement of the overall intensity of the
satellite states is also valid for each individually
resolved state and does not dependend on the laser
transition.
Fig. 3 - Excitation function of the [2p5(3s3p 2P3/2)] 2D5/2
resonant state in 2P3/2 excited sodium atoms as a function of
the excitation energy (lower scale, in eV).
sub-picosecond powerful laser between autoionizing
states may become feasible in the near future.
As a last example, we like to mention the strongly
different behavior of the shake up satellite lines
produced by photoionization of excited-atoms,
depending of the transition the laser was tuned to. In
Fig. 4, we show the shake up lines following
photoionization of sodium atoms optically excited
either in the 2P1/2 (full line) or 2P3/2 (dashed line)
excited states, respectively. In both cases, the
electronic configuration of the final ionic states is
2p54p. In the satellite spectrum measured from the
2P
1
3
1/2 state, the dominant L-states are the P1, D1, and
3D
ACKNOWLEDGEMENTS
The authors warmly thank Bruce Rude for helping in
taking the data, and Chantal Jucha for preparing the
manuscript in its final form.
REFERENCES
states, like in the main line spectrum, (Fig. 2,
lower panel), while in the spectrum from the 2P3/2
2
1. Wuilleumier, F. J., J. Physique 43, 347 (1982).
2. Bizau, J.-M., Wuilleumier, F. J., Ederer, D. L. Keller, J. L.,
LeGouët, J.-L., Picqué, J.-L., Carré, B., and Koch, P.,
Phys. Rev. Lett. 55, 1281 (1985).
3. Cubaynes, D., Bizau, J-.M., Wuilleumier, F. J., Carré, B.,
and Gounand, F., Phys. Rev. Lett. 63, 2460 (1989)
4. Meyer, M., Müller, B., Nunneman, A., Prescher, T., von
Raven, E., Richter, M., Schmidt, M., Sonntag, B., and
Zimmermann, P., Phys. Rev. Lett. 59, 2963 (1987).
5. Baier, S., Schulze, M., Staiger, H., Zimmermann, P.,
Lorenz, C., Pahler, M., Rüder, J., Sonntag, B., Costello, J.
T., and Kiernan, L., J. Phys. B 27, 1341 (1994).
state, the 3P, 1D2, and 3D3 satellite states dominate the
spectrum, again in similarity with the behavior of
127
6. Cubaynes, D., Diehl, S., Journel, L., Bouvellou, B., Al
Moussalami, S., Wuilleumier, F. J., Berrah, N., VoKy,
L. et al., Phys. Rev. Lett. 77, 2914 (1996)
7. Gisselbrecht, M., Marquette, A., and Meyer, M., J. Phys.
B 31, L977 (1998).
8. Cubaynes, D, Bizau, J. Richter, M., Wuilleumier, F.,
Eur. Phys. Lett. 14,747 (1991).
9. Dorhmann, T., von dem Borne, A., Verweyen, A.,
Sonntag, B., Wedowski, M., Godehusen, K., and
Zimmermann, J., J. Phys. B 29, 5699 (1996)
10. Wedowski, M., Godehusen, K., Weisbarth, F.,
Zimmermann, P., Martins, M., Dormann, Th., von dem
Borne, A., Sonntag, B., and Grum-Grzhmailo, A., Phys.
Rev. A 55, 1922 (1997).
11. Schulz, J., Wernet, P., Godehusen, K., Müller, R.,
Zimmermann, P., Martins, M., and Sonntag, B., J. Phys.
B 35, 907 (2002).
12.Verweyen, A., Grum-Grzhimailo, A., Kabachnik, N.,
Phys. Rev. A 60, 2076 (1999).
128