XRF analysis is an excellent method for non-destructive elemental analysis. It can rapidly give us quantitative chemistry from Na to U with minimal sample preparation. However sometimes peak overlaps, background interference and low abundances can make it difficult to accurately quantify a specific element. Is this just a loss or is there something we can do to improve our results? Certain scan parameters can be altered to improve the response of certain elements. These are voltage, current, pressure, analysis time and filters.
An XRF spectrum is generated by firing X-rays at a sample (Fig. 1). This knocks an electron out of the inner orbital of an atom, leaving the atom in an unstable configuration. To stabilise, an electron falls from an outer orbital of the atom to fill the hole left in the inner orbital. As the electrons in the outer orbit have more energy, the electron releases fluorescence with a specific energy that can be used to identify the element. The resulting XRF spectrum, therefore, shows a series of peaks at energies characteristic of specific elements in the analysed area (Fig. 2). The height of these peaks can then be used to determine how much of each element is present in the sample to give a quantitative result.
Figure 1. The process of generating an XRF spectrum from the fluorescence of atoms within a sample.
Figure 2. A typical XRF spectrum showing a range of peaks at energies characteristic of specific elements.
Filters are thin discs composed of different metals that channel the X-ray energy towards different regions of the XRF spectrum to either cut out or focus on specific elements (Fig. 3). There are a few standard filters that are commonly used to manipulate what we can see in our spectra. These filters can be used by both pXRF and micro-XRF instruments to target certain elements during analysis.
Figure 3. The filters are positioned between the x-ray beam and the sample to channel the energy of the x-ray.
Blue filter – Ti
Yellow filter – Ti and Al
Red filter – Ti, Al and Cu
Green filter – Cu, Ti and Al
Portable XRF instruments have become a staple in ‘field’ analysis, for a range of industries such as mining and exploration, agriculture, food safety, and arts and conservation. Bruker’s portable XRF instruments (S1 TITAN and Tracer 5g) are leaders in the industry when it comes to customising the scan parameters to optimise analysis to the requirement of the sample. While the standard “off the shelf” calibrations for portable XRFs are generally good for the majority of elements, the ability to create custom calibrations for the Bruker instruments opens up the possibility of obtaining accurate and precise results for even the most challenging elements.
Both the S1 TITAN and Tracer 5g have the function to use multiple beams at varying voltages and currents that allows for optimisation across the full range of the XRF spectrum in a single analysis. In addition to this, both come with a 5 position filter wheel:
The Tracer 5g also has the option of a manual filter that allows for the user to make their own filter that is optimal for their requirement beyond the 5 options provided by the manufacturer.
The Bruker M4 Tornado micro XRF has several filters that can be applied when analysing a sample. The two figures below show the effect of an AlTi filter (Fig. 4) and an AlTiCu filter (Fig. 5) compared to the original spectrum where no filter was used. Both filters will improve the interpretation of heavier metal elements. As the AlTiCu filter is much heavier than the AlTi filter, a much larger proportion of the XRF spectrum is reduced and will particularly focus on the higher Z elements, while the AlTi filter will cover the majority of metal elements. You will notice that the peak heights have been reduced when using a filter. While the filters lower the peak heights for the elements, they also reduce the spectral background and spectral artefacts such as sum peaks and braggs peaks which leaves us with a more easily interpretable spectrum.
Figure 4. Comparison of an XRF spectrum collected without a filter (red) to a spectrum collected with an AlTi filter (green).
Figure 5. Comparison of an XRF spectrum collected without a filter (red) to a spectrum collected with an AlTiCu filter (blue).
The use of filters should not be overlooked when trying to optimise the detection of particular elements of interest. While they do prevent you from analysing the full elemental range in a single scan, their ability to improve the detection of problem elements can be invaluable by enabling critical information to be gained from a sample.
Portable Spectral Services’ Schedule of Hire, Support Services & Fees is now available!