Challenges and Opportunities
A. Theoretical Challenges in EXAFS and XANES
Although EXAFS is now probably the most well understood of all
x-ray spectroscopies, the accuracy of the theory can and should be improved.
Certainly the one-electron, multiple scattering theory of EXAFS (which is
the basis of most EXAFS codes including FEFF), is largely well understood.
But remaining errors in the theory can still adversely affect the
determination of distances and coordination numbers from
experiment. Because of the huge interest of the synchrotron radiation
user community in these spectroscopies, we feel that one of the main goals
of a Theory Center should be to overcome these limitations.
- 1) XANES
- The theory of the near edge in XAS is still far from complete,
making a quantitative interpretation difficult. There are a number
of reasons for this. First, the edge itself is defined by the Fermi
energy which depends critically on local electronic structure,
local binding, etc. An accurate theory thus requires self-consistent
calculations, probably with scattering potentials that go
beyond the muffin-tin approximation. The benefits would include
improved scattering phase shifts for many spectroscopies, as well
as a definitive calculation of threshold energy (the Fermi level)
which is needed for accurate distance estimates from XAFS.
Similar advances in theory are also needed to interpret NEXAFS
experiments, i.e., the shape resonances observed in low Z materials.
In particular, technqiues that combine quantum-chemistry methods
and scattering theory are needed.
- 2) XAFS Amplitudes
- Perhaps the biggest error in XAS theory now is the overall
amplitude which is now still only accurate to 10% and which often
has to be adjusted semi-phenomenologically.
Two major causes of the discrepancy are the following: 1) Debye-Waller
factors: The effect of thermal vibrations is usually approximated
by a simple phenomenological model (e.g. the correlated Debye model).
This works well in homogeneous materials, but is likely inadequate in
complex, heterogeneous materials. Needed for an accurate treatment
are local (i.e. real space) total energy codes applicable to general
periodic or non-periodic systems. The advantage of a total energy code
is that it can be combined with XAS data to give a refined determination
of local structure of complex systems, one of the principle aims of
synchrotron experiments.
- 3) Many body amplitude factors S_0^2
- The theoretical EXAFS signals are usually somewhat high, (about 20%)
due to certain fundamental many-body effects in the x-ray absorption
process (i.e., intrinsic and extrinsic losses and interference processes).
Presently this factor has to be fit semi-empirically.
Much additional theoretical development is required to understand and
interpret XANES (x-ray absorption near edge structure) quantitatively. A fully
satisfactory theory will require better understanding of many-body effects,
such as the energy dependence of the electron self-energy and exchange
interactions and an improved treatment of the scattering potentials including
non-spherical corrections. At the same time, such advances will lead to
better treatments of inelastic losses which are important in the theory
of LEED and PD. These tasks alone will likely require a few
more PhD or Postdoc projects to reach adequate resolutions. This is a case
where many of the theoretical ideas needed are already available, but
lack of funding has limited progress. However the payoff for such
research will likely be large, essentially promising more accurate results for
thousands of researchers using XAS and other synchrotron techniques at
modern synchrotron facilities.
B. Need for "User-Friendly", Portable Codes
We stress that to be most useful to the field, it is important that
such theoretical tools be readily available to scientists using the
synchrotron facilities and relatively easy to use. This was one of the keys
to the success of the FEFF codes in the experimental community.
Advanced electronic structure and quantum chemistry codes
are of limited utility if they always require expert users to run them.
Similarly, to maintain user-friendliness and transportability, any theoretical
extensions of current codes should be done in a way that is largely compatible
with existing versions. One way to do this would be to insist
on a modular structure that independent scientists could add to.
Second, it is important that fast, preferably on-line, data analysis
tools be available, to interface to existing theory. Presently
available analysis codes are quite sophisticated,
but the degree of automation and user-friendliness could and should be improved
significantly. Needed for example are on-line tools, graphical
interfaces, etc. Both novice and expert modes are needed. A
highly automated and relatively foolproof novice mode is needed for
the relatively large fraction of users who cannot be expected to be
theoreticians or electronic structure experts, yet must be assured of
reliable results. In our view this is a crucial need if the modern
synchrotron facilities are going to be useful to the industrial scientific
community. Finally it is desirable
to be able to link XAS codes to other advanced electronic structure
and quantum chemistry codes when total-energy configurations and
other chemical and electronic information is needed.
C. Other Spectroscopies
The light sources have spawned a large number of spectroscopies for
studies of condensed matter, each with its own advantages for probing
a given experimental niche. Correspondingly each of these spectroscopies
requires special theoretical treatment, even though the fundamental
underlying structure of the theory may be similar.
Thus developments along the above lines are needed for many other
spectroscopies of interest at the modern synchrotron centers. For example,
in photoelectron diffraction (PD), an theoretical effort analogous to FEFF
has been carried out, with major coding developments by the Fadley and
Van Hove groups at LBNL. Like FEFF, these codes have been extended
recently to include effects of photon and spin polarization, which
are of particular significance for magnetic studies being planned at the
ALS. The extension of such codes to valence-band photoemission will
permit studies of both the electronic and magnetic structures
of surfaces and interfaces. Other spectroscopies of interest include XES,
(x-ray emission spectroscopy), resonant x-ray scattering, XPS,
DAFS, XRD, XMCD etc. Since all of these involve the same fundamental
process of x-ray absorption and scattering, developments in one
can often be carried over to others. For example, many subroutines in
the FEFF codes could with a relatively small effort, be adapted to most of these
spectroscopies, thus avoiding duplication of effort.
Another example is the rapidly developing field of X-ray
Magnetic Circular Dichroism (XMCD) and many related spectroscopies
designed to probe magnetic properties of matter. The corresponding
theories are similar to those of XAS, but usually require a more
careful treatment of spin and relativistic effects, and thus a
more detailed theory. Without such a theory, however, the synchrotron
data is not very useful. For example, lack of an effective theory led to
about a 10 year gap in the quantitative analysis of the first XMCD measurements.
Although the FEFF7 code can now treat XMCD (FEFF7 is a spin dependent
generalization of the FEFF7 code which gave the first
quantitative calculations of XMCD) it is not yet automated for that purpose
due to limited theoretical support.
Other spectroscopies include valence-band photoemission to study
the valence electronic and magnetic structure of
surfaces and interfaces. Calculating core-level energy relaxation
effects in XPS will also permit the study of electronic structure
and bonding properties of atoms and molecules at surfaces and interfaces.
Similarly X-ray emission spectroscopy, and vibrationally resolved XPS,
and resonant Auger Raman spectroscopy will explore the local nature
of bonding, especially for molecules bonded to solids.
Spectromicroscopy using diffraction and holography will
detect the local structure of small grains.
Soft x-ray fluorescence will also examine the solid-
liquid interface, including electrochemical properties,
corrosion, and crystal growth. In particular, x-ray
fluorescence holography will determine interfacial and
local bulk structures with elemental specificity.
Return to the front page