# XRF analyzers

Using X-ray fluorescence for material inspection with controlled calibration, surface preparation, detection limits and radiation handling.

Reviewed: 2026-11-01

Tags: inspection, elemental-analysis, xrf, positive-material-identification

X-ray fluorescence (XRF) identifies elements from characteristic radiation. Quantitative results depend on calibration, matrix and the element being measured.

**Illustrative decision problem:** a requirement limits an impurity to **0.05 mass %**, equivalent to **500 mg/kg**. A method whose detection/reporting capability is only **0.10 mass %** cannot demonstrate compliance from a nondetect. Even a quantified result below the limit needs the applicable uncertainty and decision rule. These values illustrate the logic; they are not specifications for a named analyzer.

## Measurement principle and instrument selection

Characteristic X-ray energies help identify elements; measured intensities are converted into concentrations using calibration and corrections. Absorption, enhancement and spectral interferences mean intensity is not simply a universal concentration scale. Elvatech describes these matrix effects and the need for an analytical method matched to the elements, concentration range, geometry and sample material. [Elvatech, *XRF Analyzers: How XRF Works, Accuracy, Elements, Limitations, and Practical Questions*, undated](https://elvatech.com/everything-you-need-to-know-about-xrf-and-analyzers-qa/)

| Inspection need | Capability to verify |
| --- | --- |
| Alloy PMI | Relevant elements, grade library, matrix calibration and acceptance criteria |
| Coating thickness/composition | Layer/substrate model, reference samples and supported layer combinations |
| Trace contamination | Element-specific method detection limit in the actual matrix |
| Small or curved component | Measurement spot, support, shielding and geometry qualification |
| Powder, soil or heterogeneous material | Preparation, sample presentation and representative sampling |

Handheld, benchtop and laboratory instruments differ in excitation, detection and sample environment. A capability advertised for one model or analytical mode does not transfer to all XRF instruments.

## What a grade match cannot verify

For the handheld PMI instruments discussed in Evident's **December 22, 2016** guidance, elements lighter than magnesium—including **carbon, lithium and beryllium**—are not directly measured. The article notes that other alloying elements can still identify some grades, but a requirement to quantify those light elements needs another analytical method. Do not accept a carbon-dependent steel requirement solely because the library displays a familiar grade name. [Evident, *3 Important Considerations when Using XRF for PMI*](https://ims.evidentscientific.com/en/insights/considerations-using-xrf-pmi)

Evident's **April 4, 2019** FAQ also distinguishes elemental amounts from chemical structure. Thus an elemental match should be treated as evidence for the measured chemistry, not proof of phase, compound identity or mechanical performance. Retain the element results and decision criteria alongside the library label. [Evident, XRF Frequently Asked Questions](https://ims.evidentscientific.com/en/insights/x-ray-fluorescence-xrf-frequently-asked-questions-faqs)

## Surface, thickness and representativeness

XRF does not necessarily sample the entire part thickness. Evident describes the effective sampling region as dependent on material and warns that paint, galvanizing, sealers, dirt and preparation residues can skew PMI. A clean-looking coated surface may represent the coating/substrate combination rather than the bulk alloy. Establish suitable preparation and record it; do not use one universal penetration depth for all elements and materials. [Evident, sample-condition guidance](https://ims.evidentscientific.com/en/insights/considerations-using-xrf-pmi)

Plan measurement locations around the inspection question. A weld, deposited layer, corroded surface and parent metal may need separate readings. A result at one accessible point does not demonstrate homogeneity across a casting or lot. Thin samples, edges, curvature and incomplete coverage of the measurement window require a qualified presentation method and appropriate support, not an assumption that longer exposure fixes geometry.

For environmental materials, EPA CLU-IN distinguishes rapid in-situ screening from intrusive sampling/preparation. Its soil examples show why drying, homogenization and other preparation can change data quality. Those procedures are specific to that matrix and purpose; they are not metal-PMI preparation recipes. [U.S. EPA CLU-IN, X-Ray Fluorescence technology overview](https://www.clu-in.org/characterization/technologies/xrf.cfm)

## Bounded manufacturer examples

Evident's **Vanta specifications sheet, copyright 2024**, differentiates models within the family: it lists a **large-area silicon drift detector** for **Vanta Max**, a silicon drift detector for **Core/Element-S**, and a silicon PIN detector for **Element**. The sheet offers optional **3 mm collimation** for Max, Core and Element-S, while Element has no internal collimation. These are configuration choices, not universal accuracy or detection-limit specifications. The sheet's filename contains 2023, but its printed copyright is 2024. [Evident, Vanta Specifications](https://adobeassets.evidentscientific.com/content/dam/downloads/276827051/Vanta_Max_Core_Element_Specifications_EN_2023_12_Web.pdf)

A coating application demonstrates the importance of calibration scope. Thermo Fisher's **MCS-AN1685-EN, 12/25** describes **Niton XL5 Plus** measurement of Zn-Ni coating thickness and composition on steel using **24 reference samples** for an empirical calibration. Validation used laboratory-characterized samples excluded from that calibration set. The reported success supports the demonstrated coating system and method; it does not establish arbitrary multilayer performance or bulk-alloy accuracy. [Thermo Fisher, *Real-time zinc-nickel coating measurements*, December 2025](https://documents.thermofisher.com/TFS-Assets/CAD/Application-Notes/an1685-en-12-2025-2911000.pdf)

## Precision, detection and acceptance

Separate repeatability/counting statistics, systematic error and sampling variation. Elvatech explains that four times the measurement time improves the counting-statistical component of precision by approximately two times. Longer counting does not remove an incorrect calibration or unrepresentative surface. [Elvatech, counting statistics and method limitations](https://elvatech.com/everything-you-need-to-know-about-xrf-and-analyzers-qa/)

EPA distinguishes instrument detection limits from method detection limits and notes that matrix effects can worsen the latter. “Not detected” therefore does not mean zero, and a catalogue detection limit does not automatically support a particular acceptance threshold. [EPA CLU-IN, detection limits and limitations](https://www.clu-in.org/characterization/technologies/xrf.cfm)

Use suitable certified reference materials and documented checks for the method. The **Vanta Max/Core manual, 10-040355-01EN Rev. 2, December 2023**, directs users investigating unexpected results to test certified reference material, inspect window/sample cleanliness and homogeneity, and examine the relevant spectral peaks. Do not relabel an unexplained discrepancy as a new grade. [Evident, Vanta Family User's Manual](https://adobeassets.evidentscientific.com/content/dam/downloads/276827047/10-040355-01EN-Vanta_Monarch--User.pdf)

## Handling and reporting

XRF generates ionizing radiation. Use trained operators, the actual instrument's safety procedures and applicable local requirements. The cited Vanta manual prohibits pointing toward people, firing into the air and holding samples in fingers or hands during analysis. Use the specified support/shielding arrangement for small samples; interlock configuration can vary and does not replace the operating procedure. [Evident, Vanta radiation-safety instructions, Rev. 2](https://adobeassets.evidentscientific.com/content/dam/downloads/276827047/10-040355-01EN-Vanta_Monarch--User.pdf)

Record part/lot and location, instrument/configuration, analytical mode, reference checks, preparation, acquisition conditions, element concentrations and units, detection qualifiers, uncertainty and acceptance rule. State which requirements XRF could verify and which require complementary testing.

## Related reference

[Inspection and quality documents](../../sourcing/documents/index.md)