The 5 most common ways to prepare samples for XRF analysis

Sample Preparation for XRF Analysis

XRF (X-ray Fluorescence Spectrometry) is a comparative chemical analysis technique that is capable of analysing a wide range of materials in different forms for a large part of the periodic table. This versatility makes it applicable to a wide range of applications: from quality control for steel, to the analysis of sulphur in gasoline to heavy metals in plastics and electronics.  XRF can analyse almost any material you can present to the spectrometer, but the better you prepare a sample the more accurate your analytical results.  

How do you choose what sample preparation method is best for your application ?  Here we review the 5 most common ways to prepare samples for XRF analysis and what you need to consider with each method. 


Solid Samples

Solid samples can be anything from unprepared pieces of metal or electronics or plastics to cut and polished metal samples.  The ideal sample for XRF analysis will have a perfectly flat surface.  Irregular sample surfaces change the distance from the sample to the x-ray source and introduce error.  All XRF systems are calibrated based on a fixed sample to source distance.  Changing the distance can increase or decrease the intensity coming from any element contained in the sample. 

Solid samples can be analyzed with no sample preparation or they can be cut and polished for a more qunatitative anaysis as is commonly done for the quality control of metals.

Even for largely flat samples, surface finish can effect your analysis particularly for lighter elements.  Rough surfaces can cause scattering and reabsorption of longer wavelength elements .  This effect is energy dependent so while the Ni signal may not be effected, the signal from C or S could be dramatically reduced. Quantitative analysis of solid samples often requires finishing the surface with a lathe or grinding paper.  The finer the finish the better the results will be for the lightest elements

Loose Powders

The analysis of loose powdered material usually required that the sample be placed into a plastic sample holder with a plastic support film.  This insures a flat surface to the X-ray analyzer and the sample to be supported over the X-ray beam.  The more finely ground the sample the more likely it is to be homogenous and have limited void spaces providing for a better analysis.  Sufficient powder should be used to insure infinite thickness is obtained for all of the elements of interest.  Usually this requirement can be met by using 15g of most samples.

Pressed Pellets

Pressing powder into pellets is a more rigorous sample preparation than pouring loose powders into a sample cup.  The process includes grinding a sample into a fine powder, ideally to a grain size of <75um, mixing  it with a binding /grinding aid and then pressing the mixture in a die at between 20 and 30T to produce a homogenous sample pellet. The binding /grinding aid is usually a cellulose wax mixture and combines with the sample in a proportion of 20%-30% binder to sample.

This sample preparation approach provides better analytical results than loose powders because the grinding and compression creates a more homogenous representation of the sample with no void spaces and little sample dilution.  This leads to higher intensities for most elements than loose powders.  Pressed pellets are still susceptible to particle size effects if not ground fine enough, but the biggest limitation to this approach are the mineralogical effects which most predominantly effect the analysis of major elements.  Pressed pellets are excellent for the analysis of elements in the ppm range. Pressed pellets are also relatively simple and inexpensive to prepare only requiring a pulverising mill and sample press.



Fused Beads

Sample prepared as fused beads provide a near perfectly homogeneous representation of the sample to the XRF and is considered by many to be the ideal sample preparation method for solids.  Fused beads are created by mixing a finely powdered (<75um) sample with a flux  in a flux/sample ratio of 5:1 to 10:1 and then heated to 900C-1000C in a platinum crucible.   The sample is dissolved in the flux (usually a lithium tetraborate or tetraborate/metaborate mixture) and cast into a mold  with a flat bottom.  The resultant glass disc or fused bead is a homogenous representation of the sample free of mineral structures.  The benefits of this approach are the reduction of mineralogical or matrix effects leading to more accurate analyses and the ability to combine several different sample type into the same calibration program.  The downside of fused beads include the relatively high sample dilution which has a negative effect on the analysis of trace elements.  Typical fused beads are only approximately 3mm thick and thus are susceptible to infinite thickness issues for heavier elements.  Fused beads usually require higher initial costs between platinumware and a fusion device, but then have similar cost/sample to prepare as pressed pellets.



Liquids are prepared by pouring them into a plastic sample cup in the same way as loose powdered samples.  There are limited options for analysing liquid samples and the main trick is to choose the correct support film that provides a balance of strength and transmission capabilities and contamination.  Mylar is a good general purpose film often used for the analysis of sulphur in fuels or lubricating oils. Polypropylene has better transmission than Mylar but has a lower tensile strength.  Kapton is the "bomb proof " film but dramatically attenuates your signal for lighter elements and is susceptible to strongly basic solutions.  If you are going to analyse liquids you will need to do a little research into selecting the best support film for your analysis goals.

There are many ways to prepare samples for XRF analysis and the method you choose will be a balance of the sample type, the amount of effort you are willing to expend and the quality of results you require.  


Download White Paper

Download the White Paper "Common sources of error in sample preparation for XRF Analysis and the Capabilities of Standalone Automation"


Download Application NoteDownload the Application Note "FLSmidth Centaurus V2 application note reproducibility of major element oxides in the analysis of raw meal"