Extended X-ray Absorption Fine Structure (EXAFS)

What it tells you: the oxidation state and local structural properties of the sample.

EXAFS is a powerful technique for investigating the short-range bonding within a material, and bulk measurements can investigate the averaged local structure of the sample, within the extent of the beam. The EXAFS signal is a modulation of the measured absorption coefficient, determined by the inteference of backscattered photoelectrons. The signal is typically normalised by subtracting the pre-edge and post-edge lines from the dataset, and can be examined to determine the number, radial distance, and disorder of neighbouring atoms.

At I14, we can create a stacked map that has a EXAFS spectra for each pixel, allowing you to interrogate mixtures of samples at small scales. Using standard analysis techniques (PCA and cluster analysis), EXAFS signals can be averaged across regions significanltly improving signal quality.

To create an EXAFS map, we use our fluorescence detector to take an individual XRF map at a range of energies over and beyond the edge of your desired element. The default energy range is quite comprehensive (>250 photon energies), with 0.5 to 1 eV increases around the absorption edge, and in increasing steps for higher energies (proportional to the relative wave number). However, this energy list is totally customisable if you aim to acquire a longer range or finer energy steps at a particular region. For larger areas, in-situ experiments, or samples sensitive to x-ray dose, we can use our novel sparse EXAFS technique, where we collect a smaller portion of the total measurements and numerically recover the missing entries to speed up mapping time (for more information, see Sparse Scanning).

Things to consider:

Time – Similar to XANES, we collect XRF maps at over 250 different photon energies over the specific element edge, with additional maps measured at higher energies beyond the absorption edge. At higher energies, the dwell time (time per pixel) also increases with the wavenumber to ensure the signal has a sufficient signal to noise ratio. Both factors impact the total acquisition time, which can be very time consuming (roughly 3 x the time to perform a XANES experiment).

Signal – Insufficient amounts of the element of interest in your sample will lead to a noisy spectra. As rule of thumb, check that the intensity of the studied element has at least 500 counts, when measuring at 0.5 keV over the K- or L-edge electronic transition.

Standards – EXAFS relies on having adequate reference spectra to understand what is going on in your sample. This is because the spectrum of an unknown sample can generally be understood as a linear superposition of two or more spectra from known samples. Hence, you need to think carefully what compounds/minerals are able to chemically “explain” your sample, and bring these to the beamtime (as films, foils, or (diluted) pellets). Generally, an EXAFS standard for a reference material would take ~45-60 minutes acquisition time.

Post processing – Our in-house python scripts and automatic reconstructions methods will provide you a stacked, normalised and aligned EXAFS file, which you can interrogate using MANTiS or DAWN. Linear combination fitting analysis using Athena/Larch will help you to understand the extracted clusters or principal components by comparing them against the above-mentioned EXAFS standards. For more info, have a look at: I14 beamtime and access> After your beamtime> Data analysis and interpretation.