XRF sample preparation guide for pressed pellets, fused beads, and liquids

XRF sample preparation covers the practical steps used to make a specimen representative, stable, and suitable for elemental analysis by X-ray fluorescence. The right route depends on the sample form, the elements of interest, the required detection limits, and the purpose of the work, whether it is screening, process control, or validated quantitative testing. Loose powders can be useful for quick checks. Pressed pellets often fit routine solid analysis and many trace element programs. Fused beads are preferred when mineralogy and particle-size effects would bias major and minor element results. Liquids need controlled cups, films, fill height, and mixing. For related laboratory workflows, see the sample preparation section.

Why sample preparation controls XRF results

XRF is often described as a low-preparation technique, but that phrase can be misleading. The instrument measures fluorescent X-rays emitted from the sample after excitation. If the exposed surface is uneven, wet, segregated, contaminated, or not thick enough for the chosen method, the measured intensity can shift even when the bulk chemistry is unchanged.

The main preparation risks fall into four groups. Physical effects, including particle size, surface roughness, pellet density, and sample thickness, change the path of X-rays through the specimen. Matrix effects occur when other elements absorb or enhance the signal from the analyte element. Mineralogical effects can appear when the same element is locked in different phases that respond differently during measurement. Contamination or loss can also be introduced during crushing, grinding, binding, drying, fusion, or transfer.

Authoritative methods and guides, including ISO 18227 for soil-like materials, EN ISO 12677 for fused cast beads in refractory products, ASTM E1621 for WDXRF procedure development, ASTM D7343 for petroleum products and lubricants, and EPA SW-846 Method 6200 for field-portable XRF in soil and sediment, treat preparation and calibration as connected. A calibration is valid only when standards, reference materials, and unknown samples are prepared in a comparable way.

Choose the preparation method by sample type and data objective

The best starting point is not the instrument brand or the fastest accessory. It is the analytical question. A mine site grade-control check, a cement oxide report, a petroleum sulfur test, and a soil screening survey may all use XRF, but they do not need the same specimen form or the same level of preparation control.

As a practical rule, preparation effort should increase when decisions become more quantitative, when light elements are important, when the material is heterogeneous, or when small differences affect compliance or payment. Screening can tolerate more uncertainty if results are verified by confirmatory testing. Reporting-grade work needs a written method, preparation records, reference materials, and defined acceptance limits.

Pressed pellets for routine solid XRF analysis

A pressed pellet converts powder into a compact, flat specimen. The route is widely used because it is faster than fusion, uses a larger test portion than many wet-chemical digestions, and avoids the high-temperature losses that can affect volatile components. In soil-like materials, ISO 18227 describes pressed pellets and fused beads as quantitative preparation options, with pressed pellets especially relevant for trace elements and volatile elements.

A typical pressed-pellet workflow starts with representative sampling and size reduction. The laboratory dries the material when moisture would affect mass, packing, or matrix behavior, then crushes and mills the sample under controlled conditions. Grinding should be consistent from sample to sample because particle-size distribution affects XRF intensity. Geological workflows commonly aim for fine powders that pass a defined sieve, while some ultra-fine methods go further when trace-element precision is the priority.

Binder selection is not a cosmetic choice. Waxes and cellulose-based binders can improve pellet strength, but they dilute analyte concentration and may add background for light elements. The binder-to-sample ratio should be fixed, documented, and used for calibration materials as well as unknowns. If no binder is used, pressure and particle-size control become even more important because fragile pellets can crack, shed powder, or present uneven density.

Pressing pressure, dwell time, die cleanliness, and pellet thickness also matter. A pressure that gives a strong pellet for one mineral powder may not work for a clay-rich or carbon-rich material. Published geological procedures have used high compaction pressures, but laboratories should not copy a number without validating it for their matrix, die diameter, and safety limits. The aim is repeatable density, a smooth surface, sufficient thickness for the analytical condition, and no visible cracks or laminations.

• Use the same milling equipment, time, sample mass, and cleaning procedure for calibration standards and unknowns.
• Check whether grinding media can contaminate target elements, such as tungsten from tungsten carbide or zirconium from zirconia.
• Keep binder type, binder ratio, pellet diameter, pressure, and dwell time constant within a method.
• Reject or remake pellets with cracks, raised rims, poor cohesion, fingerprints, or loose powder on the measurement face.
• Store pellets in clean containers and avoid humidity changes when hygroscopic materials are involved.

Fused beads when mineralogy and particle size are the problem

Fused beads are prepared by mixing a powdered specimen with a flux, heating the mixture until it dissolves, and casting the melt into a glass disc. EN ISO 12677 describes the principle for refractory products: fusion with a suitable flux breaks down mineralogical and particulate structure, and the resulting bead is measured by XRF against calibrations prepared in the same way.

The value of fusion is that it turns many heterogeneous inorganic powders into a smoother and more chemically uniform glass. This can greatly reduce the particle-size and mineralogical effects that remain in pressed powders. For major and minor oxides in cement, ceramics, rocks, ores, slags, and refractories, that reduction in physical bias can be more important than the extra preparation time.

Fusion is not a universal upgrade. The flux dilutes the specimen, which can raise detection limits for some trace elements. High temperature can affect volatile analytes. Reduced materials, sulfides, carbides, or samples with high metal content may require oxidation, loss-on-ignition treatment, or special precautions to avoid incomplete dissolution and damage to platinum alloy crucibles. Laboratories should validate flux type, sample-to-flux ratio, oxidizers, release agents, fusion temperature, fusion time, cooling, and bead inspection criteria.

Calibration is also preparation-specific. A fused-bead calibration should be built with standards or reference materials that go through the same fusion route as unknown samples. Pressed-pellet standards cannot automatically support fused-bead unknowns, and fused-bead standards cannot automatically support pressed-pellet unknowns. The preparation method is part of the measurement method, not a separate housekeeping step. See also: buying guides.

Liquids, oils, filters, and field samples need different controls

Liquid XRF samples are usually presented in cups sealed with thin film. The key controls are homogeneity, film compatibility, cup fill height, headspace, absence of bubbles, and stable measurement geometry. Petroleum products and lubricants are treated in ASTM D7343 as samples where sampling, storage vessels, subsampling, cups, and holders are part of method performance. If a liquid separates, settles, evaporates, or attacks the film, the measurement no longer represents the original sample.

For oils and fuels, mixing before subsampling is often essential. For aqueous or aggressive solutions, the film material must be chemically compatible and sufficiently transparent for the analyte lines of interest. For low-energy lines, film absorption can be significant, so laboratories should select film type based on the method rather than convenience.

Filters, thin films, and deposits require a different mindset. They may be intentionally thin, so the result depends on deposit uniformity and blank correction rather than infinite-thickness assumptions. Handling tools, filter media, backing materials, and storage containers can all contribute blank values. A method blank and a loaded reference or spike can show whether the preparation is suitable.

Field-portable XRF adds speed but increases exposure to uncontrolled surfaces. EPA SW-846 Method 6200 distinguishes in situ analysis from intrusive analysis of collected soil or sediment. In situ work requires removal of large nonrepresentative debris, a smooth contact surface, and avoidance of saturated conditions. Intrusive preparation may include homogenization, drying, grinding, sieving, and use of sample cups. Those steps take time, but they can reduce uncertainty caused by moisture, stones, roots, and uneven packing.

Quality control turns preparation into a defensible method

Preparation quality cannot be judged by appearance alone. A pellet may look perfect and still carry contamination from a mill. A bead may be clear but chemically biased by volatile loss. A liquid cup may be full but stratified. A defensible XRF workflow uses quality controls that test the likely failure modes.

Documentation should include sample receipt condition, drying temperature and time when used, milling equipment, sieve size, binder and ratio, pellet dimensions and pressure, fusion recipe, cup film type, lot numbers for consumables, and any deviations. The most important habit is consistency: standards, blanks, duplicates, and unknowns should follow the same preparation route unless the method explicitly justifies a difference.

Frequently asked questions

Is fused bead preparation always more accurate than pressed pellets?

No. Fused beads often improve major and minor element work by reducing particle-size and mineralogical effects, but they can dilute the sample and may not suit volatile analytes. Pressed pellets can be the better choice for many trace-element programs, volatile-sensitive materials, and high-throughput routine checks when the method is properly calibrated.

How fine should an XRF powder be before pressing?

There is no single particle size that applies to every matrix. The powder should be fine and homogeneous enough to control particle-size effects for the elements and precision required. Many published geological workflows use defined sieves and repeatable milling conditions; the laboratory should validate its own target using reference materials and duplicate preparations.

Can XRF measure a sample with no preparation?

Sometimes, especially for sorting, screening, metals verification, or field checks. However, no-preparation results are more exposed to surface condition, coating, corrosion, moisture, roughness, and heterogeneity. If the result will support a specification, payment, remediation decision, or regulatory report, controlled preparation and a confirmation plan are usually needed.

What is the most common reason XRF preparation fails?

The most common failure is treating preparation as a shortcut instead of a controlled part of the method. Inconsistent grinding, changing binder ratios, contaminated tools, cracked pellets, poor bead dissolution, incompatible cup films, and unrepresentative subsampling can all create errors that instrument software cannot fully correct.