VUV (photoemission) spectroscopy

1
VUV (photoemission) spectroscopy K.C.
Prince, Sincrotrone Trieste, Trieste, Italy
Gas phase. 1. Doubly excited states of helium. 2.
Biomolecules. 3. Dichroism 4. Two
colours Surfaces and solids. 1. Classical
application band mapping 2. Resonant
photoemission from thin film catalysts,
CeO2/Cu 3. Adenine/Cu(110) The future the Fermi
free electron laser light source.
2
Experimental
Our beamline
Gas Phase photoemission beamline, Elettra. VG
220i electron energy analyser.
Our undulator
3
Doubly excited states of helium are two electron
states in which neither electron is in a 1s
orbital Naively nlnl n, ngt1 Lowest energy
series 2snl, 2pnl, i.e. the states below the
second IP, N2. For Helium, they are in the
energy range 60-80 eV.

• Why study them?
• - The simplest three body problem in atomic
  physics,
• - the simplest system where correlation is
  important,
• a benchmark system, etc.

4
The doubly excited states of helium a brief
history
Doubly excited states were sought (and one state
found 1,2) as early as 1930. After this, there
was a pause.
Madden and Codling in the early 60s then used
synchrotron radiation to measure two of the 3
predicted series 2. Fano and Cooper provided
the theory, and there was another pause (for
experimentalists) while theoreticians thought
hard.
1990s third generation synchrotron light sources
threw some light on the matter. Domke et al
observed all three 1Po series below N2, plus
many more below higher N.
1 P. G. Kruger, Phys. Rev. 36, 855 (1930). 2
R. P. Madden and K. Codling, Phys. Rev. Lett. 10,
516 (1963) Astrophys. J. 141, 364 (1965). 3
M. Domke et al, Phys. Rev. A 52, 1424 (1996).
5
The doubly excited states of helium recent
history

• Recent results
• partial VUV fluorescence yield spectra gives a
  different view of the doubly
• excited states compared with ion yield 4, 5
• the ion yield does not give the true
  cross-section for these states 5, 6
• spin-orbit coupling is important, even for He
  7
• with photons, you can excite not only 1P states,
  but also triplets 3P, 3D 8
• He offers a window on quantum chaology 9
• - Lifetime measurements of the fluorescence tell
  us about correlation 10
• - Stark effects are also interesting (more later)
  11, 12

4 M. K. Odling-Smee et al, Phys. Rev. Lett. 84,
2598 (2000). 5 Jan-Erik Rubensson et al, Phys.
Rev. Lett. 83, 947 (1999). 6 K.C. Prince et al,
Phys. Rev. A. 68, 044701 (2003). 7 Thomas Ward
Gorczyca et al, Phys. Rev. Lett. 85, 1202 (2000).
8 F. Penent et al, Phys. Rev. Lett. 86, 2758
(2001). 9 R. Püttner et al, Phys. Rev. Lett.
86, 3747 (2001). 10 J. Lambourne et al, Phys.
Rev. Lett. 90, 153004 (2003) 11 J. R. Harries
et al, Phys. Rev. Lett. 90, 133002 (2003). 12
X.M. Tong and C. D. Lin, Phys. Rev. Lett. 92,
223003 (2004)
6
The difference between VUV fluorescence and ion
yield.
Ion yield (upper curve, left scale) and
fluorescent UV photon yield (lower curve, right
scale) at the (2,-13), (2,14) resonances.
The long lived state decays by fluorescence,
while the shorter lived state decays by
autoionization (ion yield). So that state is
stronger in partial fluorescence yield.
7
Some recent work the Stark effect for doubly
excited states. The Stark effect the response
of a quantum system to an external electric
field energy shifts and splitting of magnetic
sub-levels. An early triumph of quantum
mechanics.
Fields up to 84 kV/cm new propensity rule.
8
3 ways of looking at the Stark effect

-
Classical electrostatics charge moves from one
side of atom to other gives an approximate
value of Stark shift.

Mixing of atomic orbitals picture explains
asymmetric charge distribution.


-

Stark operator mixes states of opposite
parity. Nearby states of even parity are
mixed. Photoabsorption sees only 1Po states
there are dark S, Pe, D, F, etc.
states. (Triplet states allowed by spin-orbit
coupling, recently observed.)
9
Experimental

Purpose built apparatus parallel plate
condenser, gap 5 mm. Cost approx. 50 cents.
10
Total VUV photon yield.
Electric field F P, photon polarization.
(2,-1n) and (2,1n) states lose intensity. (2,0n)
states gain intensity (maybe an artefact) a broad
shoulder develops at higher energy for higher
n. Fields are very low!
11
Static field F perpendicular to polarization P
Quantum defect (1) -0.4350.005. Matches 1Pe
series.
A new series is observed, labelled 1. Another new
broad series labelled 2 is observed (as in other
geometry.) Big increases in intensity at high n.
12
How do we explain all this?
First, there are selection rules. Parallel
geometry, ?M0 perpendicular geometry, ?M1. M
zero for ground and S states, so in the
perpendicular geometry there is no mixing of S
states. Then, theory confirms the series 1 is
1Pe. The broad features 2 are due to a pair of
1De states. In the parallel geometry, mixing
occurs with 1Se states.
13
Quantitative modelling of energies, 5 kV/cm, F
perpendicular to P. First order perturbation
theory A. Mihelic and M. Žitnik, Ljubljana.
14
Conclusions
VUV spectroscopy is a powerful method for
investigating the fluorescence decay dynamics of
He doubly excited states. The wave functions in
the excited state can be probed in detail and
agreement with experiment is satisfactory. Stark
effects on these states can be seen at moderate
fields (lt 1 kV/cm). A new series of the He
doubly excited states observed, the 1Pe
series. Indications of other series observed, 1Se
and 1De.
V. Feyer, M. Coreno, CNR-IMIP, Montelibretti
(Rome), Italy, and INSTM, Trieste, Italy, R.
Richter, Sincrotrone Trieste, Trieste, Italy, M.
de Simone, A. Kivimäki, INFM-TASC,Trieste,
Italy,and INSTM, Trieste, Italy, A. Mihelic and
M. Žitnik, J. Stefan Institute, 1000 Ljubljana,
Slovenia
K.C. Prince et al, Phys. Rev. Lett. 96 (2006)
093001.
15
Biomolecules interaction with UV radiation
Interest damage by ionizing radiation astrobiolo
gy synthesis and destruction of pre-biotic
molecules in space (particularly Lyman a and
He I) conformation and dynamics of this class of
molecules mass spectrometry There exists a
large body of electron impact ionization why
use UV radiation? Because it is more specific
dipole selection rules apply.
16
Photofragmentation with laboratory sources of UV
glycine
proline
Aliphatic amino acids and Pro show strong
fragmentation, even close to threshold. Mostly
loss of HCOO. (Our work and Lago et al, Chem.
Phys. 307(2004) 9).
leucine
17

• Methionine is different
• - loss of COOH is not the main channel
• at low energies, the parent ion is the dominant
  peak
• Why?

O. Plekan et al, Chem. Phys. 334 (2007) 5363.
18
UV photoionization removes an electron from a
valence orbital.
The Highest Occupied Molecular Orbital of
aliphatic amino acids has nitrogen lone pair
character. The HOMO of methionine has S lone
pair character. The S atom accommodates the
charge without fragmenting.
19
Amino acids with aromatic groups.
Phenylalanine
Tryptophan
Tyrosine
Tryptophan mass spectrum
For all three amino acids, there is a
significant parent ion signal at low photon
energy. Fragmentation pattern differs aromatic
ring breaks off from rest of molecule. HOMOs have
p character.
20
Next step coincidences.
E.g. Methanol, CD3OH.
For amino acids we measured all ions produced by
ionization of several orbitals. If we measure
ions and photoelectrons in coincidence, we
observe the fragments due to the ionization of a
specific orbital.
Work done at Spring-8, in collaboration with R.
Richter, K. Ueda, G. Pruemper.
21
What is the conformational (folding) structure of
a free bio molecule?
Dilute species, biomolecules
Circular dichroism (CD) in the near UV is a
standard tool for secondary structure
determination/control of large molecules.
I(left)-I(right)/I0.001-0.0001. Sample in
solvent, structure is true structure. - Sample
in solvent, wavelength range limited to 180-250
nm circa. - low info content (few peaks).
Conventional CD is best for monitoring
conformational changes due to a perturbation,
quality control etc. Less good for absolute
structure. Being extended to 120 nm (Daresbury-gt
Australian synchrotron, Brookhaven etc.)
CD spectrum and secondary structure of proteins.
22
Dichroism
Recently natural CD of small chiral molecules
investigated. Optimal conditions CD
signalI(left)-I(right)/I0.04.
Can we use VUV spectroscopy (and later Fermi) to
obtain structural (folding) information? Higher
energy -gt no windows -gt molecules in
vacuum. Free molecules -gt range extended above
the IP -gt potentially larger info content
S. Turchini, N. Zema, G. Contini, G. Alberti, M.
Alagia, S. Stranges, G. Fronzoni, M. Stener, P.
Decleva, and T. Prosperi, Phys. Rev. A 70, 014502
(2004)
23
Proline, valence band spectrum
Dichroism from 0 to 4.
24
Valence band photoemission. He I. The two highest
MOs are due to two pairs of conformers.
By measuring spectra as a function of
temperature, can extract free energy difference,
6-9 kJ/mol.
25
Surfaces and solids
Photoemission from O/Ag(110).
One of the most important applications of VUV
photoemission spectroscopy has been band
mapping. It will continue to be a standard
method.
26
What science will be done? The valence band
gaps in high Tc superconductors, transport
properties, bonding, etc. Oxides and related
materials have large unit cells -gt high momentum
resolution required -gt high angular resolution at
low electron energy.
Z.-X. Shen and co-workers are prolific users of
low energy photons and Angle Resolved UPS. Huge
output of results on oxide materials. ARUPS-gt
direct access to valence band structure.
K.M. Shen et al, PRL
27
Microscopy Pb/Au/Si(111)
SPELEEM photoemission microscope, (now closed,
upgraded to Nanospectroscopy). Sample
Au/Si(111) 5 ML Pb. The Au induces layer by
layer growth. Area about 1 micron
diameter. Time one photon energy, one kinetic
energy, all angles about 60 s. One photon
energy, all valence kinetic energies, all angles
tens of min.
Valence band.
28
Inverse model catalyst CeO2/Cu
Many catalysts consist of metal particles
supported on oxides. Oxide maintains dispersion,
is a reservoir for oxygen, participant in SMSI
(strong metal support interaction), catalyst,
etc. Difficult to prepare single crystal oxides,
easy to prepare metal crystals. -gtPrepare oxides
on metals.
HR-TEM image of (a) Cu loaded ceria powder (b)
elemental mapping of the area white - cerium,
red copper.
F. Šutara, V. Matolín et al, Thin solid Films, in
press.
29
We can grow defect-free, well-ordered epitaxial
CeO2 layers on Cu(111)
LEED of CeO2/Cu(111), E 98 eV, (a)
discontinuous, (b) 2.5 ML, (c) 5 ML. Arrows mark
substrate spots.
Can we check for point defects (oxygen vacancies)
with resonant photoemission?
30
Resonant photoemission
Configurations CeO2, Ce4, (4d105p6) 4f0,
resonates at hv124.5 eV Ce2O3, Ce3, (4d105p6)
4f1, resonates at hv122 eV Ce metal, (4d105p6)
4f15d16s2, resonates at hv122 eV Resonant
process Ce 4d104fn ? 4d94fn1 ?
4d10fn interferes with direct valence
photoemission.
Resonant spectra show primarily
Ce4. Conclusion film is epitaxial (LEED) and
has low point defect density.
F. Šutara et al, Thin Solid Films, 516 (2008)
6120.
31
DNA bases II adsorption of adenine, C5H5N5, on
Cu(110)
A DNA base on a metal surface a prototypical
bio/metal interface. Has been studied by STM,
vibrational spectroscopy, theory.
Adenine, C5N5H5
Q. Chen and N. V. Richardson, Nature Mat. 2
(2003) 324 Q. Chen, D. J. Frankel, and N. V.
Richardson, Langmuir 18 (2002) 3219 D. J.
Frankel, Q. Chen and N.V. Richardson, J. Chem.
Phys. 124, 204704 (2006).
32
Experimental conclusions from the literature
the molecule is lying flat (Yamada et al) the
molecule is tilted slightly (Chen et al) or the
molecule is strongly tilted (McNutt et
al) Theory agrees bonding through amino
group. Molecule tilted up from surface, 18-26º.
Vibrational studies Q. Chen, D. J. Frankel, and
N. V. Richardson, Langmuir 18 (2002) 3219 A.
McNutt et al, Surf. Sci, 531 (2003) 131. T.
Yamada et al, Surf. Sci. 561 (2004) 233247.
Q. Chen, D. J. Frankel, and N. V. Richardson,
Langmuir 18 (2002) 3219
Preuss et al, PRL 94, 236102 (2005)
33
N 1s photoemission
0.6 ML
gt 1 ML
Ordering
Desorb excess
Deposited
Gas phase
Bonding is very different at high and low
coverage. N is involved in bonding. In
particular, the two amino N atoms have
unsaturated (imino) At low coverage N 1s (398.7
eV) characteristic of p bonded N.
34
NEXAFS of adenine Near Edge X-ray Absorption
Fine Structure Spectroscopy
0.3 ML, annealed 430 K
1 ML, annealed 430 K
Fractional monolayer molecule lying almost
parallel to the surface. Saturated monolayer
molecule(s) tilted
35
Collaborators.
Oksana
The Gas Phase team at Elettra O. Plekan, V.
Feyer, R. Richter, M. Coreno, M. de
Simone, Materials Science (Czech) Beamline T.
Skala, V. Chab, F. Sutara, V. Matolin
Vitaliy
Tomas
Robert
Marcello
Monica
36
Free Electron Lasers.
Fermi specs. A seeded free electron laser -gt
a conventional laser bunches the electrons -gt
the bunches pass through an undulator -gt the
electrons in each bunch emit coherently -gt
intensity is proportional to square of number of
electrons, and square of number of periods in
undulator. Fermi FEL 1 12-30 eV FEL 2 30-126
eV. Pulses of 50 fs-1 ps 50 Hz GW power levels,
1014 photons/pulse Start operation at the
beginning of 2009
37
If you can control the timing of light, you can
learn new things
Eardweard Muybridge, 1878 do all 4 horses
hooves leave the ground at once? A bet by Leland
Stanford, wealthy ex-governor of California
early research at Palo Alto. Sub-second
resolution.
Harold E. Edgerton, 1964. Microsecond strobe.
Or applied research, H. E. Edgerton, 1934.
38
Two photon double ionization
Physics is different for two photon (hngt39.5
eV) and many photon (hn few eV) double
ionization. -gtcontrol of relative field strengths.
39
FEL light is so intense that it ionizes all
molecules in its path -gt ultradilute
samples. Clusters and flying proteins.
Nanospray set-up for spectroscopy, with unfree
lasers, T.R. Rizzo et al, PPCM, EPFL.
Our schematic setup.
40
Setting up with Synchrotron Radiation two
photon spectroscopy of neon
Synchrotron light 30-65 ps bunch length, 2 ns
interval. Neon atoms excited with synchrotron
light. Then excited by a laser in a second step
to resonant ionizing states. Two photon
transitions ?l0,2. Proof of principle of
lifetime measurements on nanosecond time scale.
A. Moise et al, Nucl. Instrum. Methods A 588
(2008) 502.
41
Where is Fermi being built?
42
Brightness
ELETTRA