XRF sample preparation methods for powders, pressed pellets, and fused beads

Why XRF sample preparation matters

XRF sample preparation is the practical step that turns a raw material into a specimen the spectrometer can measure reliably. For powders, ores, cement, ceramics, catalysts, metals, slags, and many process materials, the preparation method can change measured intensity even when the chemical composition has not changed. A flat surface, consistent density, fine particle size, and homogeneous element distribution help the instrument measure a representative specimen rather than a local grain, void, or contaminated surface. Published XRF guidance, including ASTM E1621 for wavelength dispersive XRF methods, treats calibration, sample preparation, and sample homogeneity as core contributors to analytical accuracy. In routine work, the right preparation method is not necessarily the most elaborate one. It is the method that fits the material, the elements of interest, the required precision, and the calibration strategy.

The main preparation routes used in XRF analysis

Most laboratory XRF workflows use one of three routes: loose powder, pressed pellet, or fused bead. Each route has a different balance of speed, cost, accuracy, contamination risk, and suitability for trace or major element analysis. Solid metals may also be machined, polished, or ground directly, while liquids can be measured in cups with film when the instrument and method support it. This article focuses on powders and powder-derived specimens because preparation choices often have the largest effect on routine results in these materials.

Preparation route	Typical use	Main advantages	Main limitations
Loose powder	Screening, rapid process checks, limited sample amount	Fast, low preparation cost, minimal chemical change	Particle size, packing density, segregation, moisture, and cup film effects can reduce repeatability
Pressed pellet	Routine powder analysis where speed and repeatability are both important	Flatter surface and more consistent density than loose powder; compatible with many production labs	Still affected by mineralogy and particle size; may need binder; die contamination must be controlled
Fused bead	Major and minor elements in minerals, ores, cement, ceramics, glass, and similar matrices	Produces a highly homogeneous glass disc and reduces particle-size and mineralogical effects	More time, flux cost, dilution, possible volatile losses, and possible crucible damage if chemistry is not controlled

The practical rule is that unknown samples and calibration standards must be prepared in the same way. A calibration built with fused beads should not be applied to pressed pellets unless the method has been validated for that conversion. Different preparation routes change density, surface condition, matrix behavior, and sometimes the chemistry presented to the spectrometer.

What changes during XRF sample preparation

XRF does not simply read a bulk chemical label from a container. The measured signal depends on how X-rays enter the specimen, interact with the matrix, and leave the specimen toward the detector. Several physical effects are strongly preparation-dependent.

Particle size effects

Coarse grains can make the illuminated area unrepresentative. If a powder contains minerals or phases with different compositions, the beam may strike too much of one phase and too little of another. Light elements are especially sensitive to surface and near-surface conditions because their fluorescent X-rays are more easily absorbed. Grinding to a fine, consistent particle size reduces this effect, but the target particle size should come from the validated method rather than from a universal rule.

Mineralogical effects

Two samples can contain the same total concentration of an element but hold that element in different minerals or crystal structures. Pressed pellets retain those mineralogical differences. Fusion largely removes them by dissolving the sample into a glass matrix, which is why fused beads are widely used for major and minor oxide analysis in geological, cement, ceramic, and metallurgical laboratories.

Surface and density effects

A rough surface, broken pellet edge, air gap, or uneven powder bed changes the effective path through which X-rays travel. Pressing creates a more defined surface than loose powder, while fusion creates a glass surface when casting and cooling are well controlled. For all routes, the measuring surface should be flat, clean, and free from cracks, fingerprints, loose dust, or visible defects.

Contamination and loss effects

Sample preparation can add or remove elements. Hardened steel grinding tools or pellet dies may introduce iron, chromium, or other metals. Fluxes, binders, films, and grinding aids can contribute background or dilution. Heating can change moisture, carbonates, sulfides, sulfur species, alkalis, or other volatile components. These effects do not make XRF unreliable; they mean the preparation route must be controlled and documented.

Preparing powders and pressed pellets

Pressed pellets are often the practical middle ground for routine XRF sample preparation. They are faster and less chemically disruptive than fusion, while giving a more repeatable surface than loose powder. A robust pellet workflow usually includes drying, size reduction, splitting, grinding, mixing, pressing, and inspection.

1. Dry or condition the sample. Moisture changes mass, packing, and X-ray absorption. Drying temperature and time should match the material and target analytes, because aggressive heating can alter volatile components.
2. Crush and split the bulk material. A laboratory specimen is useful only if it represents the submitted sample. Use a suitable splitter or validated sub-sampling method rather than scooping from a segregated container.
3. Grind to a consistent fineness. The target particle size depends on the matrix and method, but the goal is to minimize grain-size and mineralogical segregation across the beam area.
4. Mix with binder if required. Binders can improve pellet strength, but they also dilute the sample and may add measurable elements. Use a fixed binder type and ratio in both standards and unknowns.
5. Press under controlled conditions. Pressure, dwell time, die diameter, backing cup, and release technique should be consistent. The exact conditions are method-specific and should be validated for mechanical stability and analytical repeatability.
6. Inspect before analysis. Reject pellets with cracks, laminated layers, chipped edges, loose dust, or visible segregation unless the method specifically allows them for screening only.

Loose powder can be useful when the goal is quick screening or when only a very small amount of material is available. The laboratory should treat loose-powder results cautiously if the sample contains coarse grains, light elements, moisture, or strong mineralogical variability. Cup film selection also matters because film composition and thickness can attenuate low-energy X-rays.

When fused beads are the better choice

Fused beads are prepared by mixing a weighed sample with a borate flux, heating until the mixture dissolves, and casting the melt into a glass disc. ASTM E1621 describes borate fusion as one of the preparation routes used for XRF methods, and ISO/TS 9516-4:2021 specifies a fused-bead performance-based method for iron ore analysis by WDXRF and EDXRF. ISO cement methods also describe bead preparation, storage, and temperature considerations for cement analysis by XRF.

Fusion is usually preferred when the laboratory needs better control of particle-size and mineralogical effects. The glass bead gives the spectrometer a more uniform matrix, which can improve calibration behavior for major and minor elements. This is why fusion is common in cement, glass, ceramics, rocks, ores, slags, and similar inorganic materials.

Fusion also introduces limits that must be considered during method development. The sample is diluted by flux, so trace-element sensitivity may suffer. Some volatile components can be lost or altered if temperature, oxidation, or fusion time are not suitable. Materials containing unoxidized metals, carbides, sulfides, carbonates, or high levels of reducing components may need pretreatment because they can affect the melt, damage platinum-gold ware, or change analyte recovery. For sulfur, chlorine, alkalis, and other more volatile analytes, a pressed pellet or a reduced-temperature fusion method may be more appropriate if validated.

Common fusion variables include flux chemistry, sample-to-flux ratio, oxidizers, anti-wetting agents, crucible material, fusion temperature, agitation, casting temperature, cooling rate, bead thickness, and surface quality. Published examples often use lithium tetraborate, lithium metaborate, or mixtures of the two. ASTM guidance gives common examples such as 1 g of sample with several grams of lithium borate flux, but the final ratio should be selected for the matrix, calibration range, and required detection limits.

Calibration and quality control must match the preparation method

Good preparation is inseparable from good calibration. A method should define how standards, blanks, drift monitors, reference materials, and unknowns are prepared. If certified reference materials are pressed as pellets but production samples are fused, the comparison may be biased even when the spectrometer is stable. The instrument responds to the prepared specimen, not just to the theoretical composition.

A practical QC plan for XRF sample preparation should include:

• Preparation blanks to check fluxes, binders, films, grinding media, and labware.
• Replicate preparations to separate preparation variability from instrument repeatability.
• Certified or well-characterized reference materials prepared by the same route as unknowns.
• Drift monitors to distinguish instrument drift from changes in preparation quality.
• Control charts for high-value analytes, light elements, and elements vulnerable to contamination.
• Written acceptance criteria for pellet integrity, bead appearance, recovery, repeatability, and recalibration triggers.

Documentation should be detailed enough for another trained analyst to reproduce the specimen. Record the sample mass, flux or binder mass, lot numbers, grinding time, equipment used, pressing conditions, fusion temperature, fusion time, casting conditions, storage conditions, and any pretreatment such as drying, ignition, or oxidation.

Common mistakes that reduce XRF accuracy

Mistake	Why it matters	Better practice
Using one calibration for different preparation routes	Pressed pellets, loose powders, and fused beads produce different matrix and surface conditions	Prepare calibration standards and unknowns by the same validated method
Grinding too little or inconsistently	Coarse grains and phase segregation can distort measured intensities	Use a controlled grinding method and verify repeatability for the matrix
Ignoring moisture or loss on ignition	Water, carbonates, and volatile species change mass and sometimes chemistry	Define drying, ignition, or reporting basis in the method
Allowing die or mill contamination	Wear metals can create false positives or biased concentrations	Select grinding media and dies compatible with target elements; run blanks
Accepting cracked pellets or defective beads	Surface defects change the interaction geometry and measured signal	Inspect each specimen and prepare a replacement when defects affect the measuring area
Changing flux, binder, film, or cup type without validation	Consumables can change dilution, absorption, or background	Control consumable specifications and revalidate when they change

How to choose the right XRF sample preparation method

There is no single preparation route that is best for every XRF application. The decision should start with the analytical question.

• Use loose powder when speed matters more than the lowest uncertainty, the matrix is relatively uniform, and results are used for screening or internal process trends.
• Use pressed pellets when routine throughput is important but the laboratory needs better repeatability than loose powder can provide. This is often suitable for many industrial powders after method validation.
• Use fused beads when major and minor element accuracy is the priority and the matrix is affected by particle-size or mineralogical differences. Fusion is especially useful for many oxides, silicates, ores, cement materials, ceramics, and glass-related samples.
• Use machined or polished solids for metals and alloys when the material is sufficiently homogeneous and the method specifies surface finishing requirements.
• Use liquid cups cautiously when the spectrometer configuration, film, sample chemistry, and safety controls support liquid analysis.

The most defensible method is one that has been tested against reference materials and replicate preparations. If preparation contributes more uncertainty than the instrument, improving grinding, pressing, fusion, or sampling will usually deliver more value than changing counting time alone.

Frequently asked questions

Is fusion always more accurate than pressed pellets?

No. Fusion often improves accuracy for major and minor elements in mineral and oxide matrices because it reduces particle-size and mineralogical effects. However, it dilutes the sample and may not be ideal for some trace elements or volatile analytes. Pressed pellets can be the better choice when the method is validated for the required elements and concentration range.

Can XRF analyze a powder without preparation?

Yes, but the result may be suitable only for screening unless the method has been validated. Loose powders are vulnerable to segregation, uneven packing, surface roughness, particle-size effects, and film absorption. For routine reporting, pressed pellets or fused beads usually provide better control.

What particle size is needed for pressed pellets?

There is no universal particle-size limit that fits every matrix. Finer and more consistent grinding generally improves homogeneity, but the target should be set by the validated method, analytes of interest, and acceptable repeatability. Over-grinding can also increase contamination or change heat-sensitive materials.

Do standards and unknowns need the same preparation?

Yes. Calibration standards, quality-control samples, and unknowns should be prepared by the same route whenever quantitative accuracy is required. Changing from fusion to pressing, or from pressed pellets to loose powder, changes the specimen that the instrument measures.

What should be recorded in an XRF preparation method?

Record the sample mass, drying or ignition conditions, grinding equipment and time, binder or flux type and mass, pressing or fusion conditions, consumable lots, storage conditions, inspection criteria, and QC results. Good records make troubleshooting possible when results shift unexpectedly.