Why Do an XAFS Experiment?

Some common "counter-arguments":

1. XAFS is just a structural tool -- one can just do diffraction/Mossbauer/Raman/NMR using laboratory instrumentation.
2. Techniques like ICP and AAS can identify the presence of elements with very high sensitivity.

3. Laboratory techniques like kinetic phosphorescence analysis can identify particular valence states (such as U⁶⁺).

4. EDAX & EELS can be performed on many electron microscopes, which is much easier than synchrotron x-ray fluorescence & absorption.

1: Complementarity of XAFS as a structural tool

Length scale

• XAFS is a local probe, thus is only sensitive to atomic correlations on the scale of 1nm or less. C
  • Complements Bragg/Reitvelt diffraction, which is only sensitive to correlations over the entire length scale of the crystal.
  • Complements diffuse scattering, which is sensitive to correlations down to the meso-scale.
  • Complements PDF (or high energy WAXS, or whatever you want to call it), which should have similar sensitivity to the sub-nm correlations, but which measures total pair correlations. XAS measures partial pair correlations, thus provides key information to making the PDF signal interpretable

Time scales

• XAS is super-fast -- dominated by the core-hole lifetime, around 1-2 femtoseconds. This is much faster than the excitations in Mossbauer, Raman, or NMR, thus XAS may be sensitive to different dynamical behavior.

Entire periodic table

• Mossbauer is also a local probe, but it is restricted to the Mossbauer active isotopes. Some elements do not have active isotopes (Cu and Co for example). Other elements are prohibitively expensive to measure. While it is true that osmium has three active isotopes -- they are all small minority isotopes. A sample of ¹⁸⁹-enriched Os would be enormously expensive. On the other hand, all elements cost the same, in a real sense, at the synchrotron and virtually all elements can be measured by XAS. Note, though, that all beamlines cover specific energy ranges, so coverage of the entire periodic table requires a suite of properly equipped beamlines.

Symmetry and the state of the sample

• Unlike Raman and Bragg diffraction, XAS does not make any assumption of symmetry or periodicity. You can measure XAS on any state of condensed matter, including crystals, liquids, amorphous solids, surface sorbed species, quasicrystals, solutions, organic matter, etc.

Sample preparation

• Sample preparation is relatively simple and samples can stay very close to their original states. As one example, a moist soil sample need not be dried. The XAS measurement can (and should) be made on the wet sample. This leaves the sample as close as possible to its original chemical state. This is in contrast to many other analytical methods (such as anything in the electron microscope) which require extensive preparation which may alter the chemical state of the sample.

Sample containment

• Because X_rays are deeply penetrating into matter, it is easy to measure XAS while the sample is contained inside of some relevant instrumentation. For example, an electrochemistry cell can be made with kapton or beryllium windows, allowing the X-rays to penetrate into the active area of the cell and exit to the detector. Other examples of sample containment might include furnace, cryostat, high pressure cell, high field magnet, peristaltic flow cell, and so on.

Identification of elements

It is true that ICP and AAS are vastly more sensitive when quantifying the concentration of a particular element. Indeed, it is common to sacrifice a part of a sample to be measured by XAS for the purpose of quantification by one of these techniques.

XAS, however, is non-destructive. After the XAS measurement you still have your sample and in many cases the sample is completely unchanged for having been through an XAS experiment.

In the case of precious sample (cost, uniqueness, need to perform other measurements), ICP and AAS are not options. Using x-ray fluorescence to quantify concentrations is your best option.

Identification of valence states

KPA is used as a laboratory monitor of reduction rates in a redox reaction. In the case of U⁶⁺, one can monitor the rate at which U⁶⁺ is reduced to U⁴⁺ in some reaction. This is undoubtedly more convenient than trying to time a laboratory experiment to coincide with an XAS experiment.

KPA, however, does not identify which U⁶⁺ species is present, only that the U is in the 6+ state. XAS is, in that sense, a more complete valence speciation tool. You can quantify concentrations of valence states with (in many cases) similar accuracy and precision as KPA, but with the added benefit of being able to identify the precise species of the element in the different valence states.

KPA and XAS should be used together to understand redox. KPA is the laboratory monitoring tool which is used to identify the interesting samples for measurement at the synchrotron.

Assumptions about valence in a sample or supernatant on the basis of solubility can be misleading. In the presence of chelators or siderophores -- i.e. soluble compounds with very strong affinities for metal ions -- nominally insoluble valence states can be held indefinitely in solution. In that case, an argument about valence based on solubility can be grossly incorrect. In biotic systems, in particular, solubility arguments are particularly suspect as many microbes generate siderophores as part of their normal metabolism.

Multiple element analysis and imaging

EDAX and XRF are extremely similar -- down to the level of measuring the identical physical interaction (promotion of a deep core electron and measurement of a fluorescence photon). EDAX has several advantages -- it can be done in an electron microscope, thus is probably easier to schedule than an XRF experiment. EDAX has superior spatial resolution compared to all but a small handful of beamlines around the world.

On the other hand, EDAX requires a sample that can be put into a UHV environment. This makes it, in general, more appropriate for materials science problems than for life ro environmental science problems. XRF can be performed on a wet sample or on a sample contained in a special environment or in special containment.

Also, x-rays are much mere deeply penetrating into matter than electrons. EDAX is thus strictly surface sensitive while XRF is a better probe of the bulk. Both, or course, are important.

Faisal's comments

1. Only technique that provides species-specific atomic and electronic structural information from nearly every element of the Periodic Table. (BR: Expanded "entire periodic table" paragraph.)

2. The penetrative power of hard X-rays allow for in-situ studies in reaction cells. (BR: added sample containment paragraph)

3. Local structure information even if long-range structure is not periodic; useful in glasses and organics. (BR: Discussed in "symmetry and state of the sample" paragraph.)