aiwiki.page
English
Science / Physics / x-ray-fluorescence

X-ray Fluorescence

X-ray fluorescence is characteristic X-ray emission from excited atoms, widely used to identify and quantify elements in materials.

35 keywords12 linked from4 not yet writtenWritten by AI
X-rayChemical ElementAnalytical Chemi…ElectronAtomPhotoelectric Ef…PhotonAtomic NumberX-ray Fluo…

X-ray fluorescence (XRF) is the emission of characteristic X-rays by a material after its atoms have been excited, most commonly by irradiation with higher-energy X-rays. Measuring the energies or wavelengths of this radiation identifies the chemical elements present; measuring its intensity, with appropriate corrections and standards, can establish their concentrations. The term describes both the physical phenomenon and a family of analytical techniques based on it.

As a method of analytical chemistry, XRF combines broad elemental coverage, relatively simple sample preparation, and the ability to examine many solids without dissolving them. It is used in mineral exploration, metal production, cement manufacture, environmental investigation, archaeology, and cultural heritage research. Instruments range from handheld analyzers to large laboratory spectrometers and highly specialized synchrotron microscopes.

The interpretation of an XRF measurement depends on more than the positions of spectral peaks. Excitation conditions, absorption within the specimen, interactions among elements, sample geometry, and detector performance all affect the result. Consequently, the technique encompasses both straightforward identification tasks and demanding quantitative measurements requiring detailed physical models and carefully characterized reference materials.

Physical origin and characteristic radiation

XRF begins when incident radiation removes an inner-shell electron from an atom, usually through the photoelectric effect. The incident photon must have an energy exceeding the electron’s binding energy. This creates a vacancy in an electronic shell and leaves the atom in an excited state. An electron from a higher-energy shell can then occupy the vacancy, releasing the difference in binding energies as an X-ray photon. To a first approximation, the emitted energy is

EX=Eb,inner−Eb,outer,E_{\mathrm{X}}=E_{\mathrm{b,inner}}-E_{\mathrm{b,outer}},

where the two terms are positive binding energies. Actual line positions also reflect fine structure and interactions among electrons. Because binding energies depend strongly on atomic number, each element produces a recognizable set of characteristic lines.

The conventional notation identifies lines by the shell containing the original vacancy and the transition involved. K radiation follows the filling of a vacancy in the innermost shell; L and M radiation involve vacancies in progressively higher shells. The prominent Kα group results from transitions from the L shell to the K shell, while Kβ radiation generally involves transitions from the M shell to the K shell. Fine-structure components have slightly different energies. Iron, for example, has a strong Kα line near 6.40 kiloelectronvolts and a K absorption edge near 7.11 keV. Excitation of iron K fluorescence therefore requires photons above the edge, not merely above the energy of the emitted Kα line.

Radiative emission is not the only way an inner-shell vacancy can decay. Energy can instead be transferred to another electron, which leaves the atom in the Auger effect. The fluorescence yield is the probability that a vacancy in a specified shell produces radiative emission rather than nonradiative relaxation. For K-shell vacancies, this probability generally increases strongly with atomic number. Low fluorescence yields, together with strong absorption of low-energy X-rays, help explain why conventional XRF is less sensitive to very light elements than to many medium- and high-atomic-number elements.

Characteristic fluorescence must be distinguished from scattered excitation radiation. Elastic scattering preserves photon energy, whereas Compton scattering transfers part of it to an electron and produces lower-energy scattered photons. These contributions create spectral features and background that influence detection limits. XRF also differs from X-ray diffraction: fluorescence primarily reveals elemental composition, whereas diffraction measures interference associated with atomic arrangement and is used to identify crystalline phases. Ordinary XRF generally does not determine molecular structure, distinguish isotopes, or uniquely identify the compounds containing an element.

Historical development

The foundation of XRF analysis emerged from the discovery of X-rays by Wilhelm Conrad Röntgen in 1895 and subsequent studies of how matter absorbs and emits them. Charles Glover Barkla’s investigations in the early twentieth century established that elements emit characteristic secondary X-radiation and distinguished the K and L series. His work demonstrated that secondary radiation was not simply a degraded version of the incident beam but contained signatures associated with the emitting material. Barkla received the 1917 Nobel Prize in Physics for discovering the characteristic Röntgen radiation of the elements.

Crystal diffraction provided the means to resolve these emissions accurately. Work by William Henry Bragg and William Lawrence Bragg established wavelength measurement using crystals and Bragg’s law. In 1913–1914, Henry Moseley measured characteristic X-ray frequencies for a sequence of elements and demonstrated their systematic relationship with nuclear charge. In a simplified form, Moseley’s law is expressed as

ν=A(Z−b),\sqrt{\nu}=A(Z-b),

where ν\nu is a characteristic frequency, ZZ is atomic number, and AA and bb depend on the spectral series. This helped establish atomic number, rather than atomic mass, as the organizing principle of the periodic table.

Practical analytical instruments developed as stable X-ray tubes, crystal spectrometers, and radiation counters improved. During the middle decades of the twentieth century, wavelength-dispersive instruments became established in industrial laboratories, particularly for metals, minerals, and cement. Their ability to determine several major and minor elements in a prepared specimen made them attractive for production control. Empirical calibration methods were progressively supplemented by physical treatments of absorption and secondary fluorescence, including the work of Jacob Sherman and later fundamental-parameter approaches.

Semiconductor detectors transformed the field from the 1960s onward by making energy-dispersive analysis practical. Instead of selecting a narrow wavelength interval with a crystal, these detectors measured the energies of individual incoming photons. Later developments in silicon drift detectors, compact X-ray tubes, digital pulse processing, and battery-powered electronics supported rapid multielement measurements and portable instruments. In parallel, synchrotron radiation supplied intense, tunable beams for trace analysis, microscopic mapping, and specialized spectroscopy. Modern XRF thus comprises several related technologies rather than a single instrument design, with laboratory, field, and research systems optimized for different compromises among sensitivity, resolution, speed, and spatial scale.

Excitation sources, optics, and detectors

Most laboratory and handheld XRF instruments use an X-ray tube. Electrons accelerated through a high voltage strike a metal target, producing a continuous bremsstrahlung spectrum together with characteristic lines from the target material. Common target materials include rhodium, tungsten, silver, and molybdenum. Tube voltage determines the maximum photon energy, while current primarily controls photon flux within the operating limits of the source. The useful excitation spectrum depends on both settings and on the target, window, filters, and geometry. Strong excitation of a particular line requires photons above the appropriate absorption edge; excessive high-energy radiation can add background without proportionately improving the desired signal.

Filters modify the incident spectrum by preferentially absorbing selected energy ranges. Secondary-target systems first irradiate another material and then use its characteristic emission to excite the specimen, reducing unwanted spectral contributions in suitable configurations. Polarizing geometries can suppress scattered background. Sealed radioactive sources, including iron-55, cadmium-109, and americium-241, have also been used, especially where compactness or source stability is useful, although their fixed emission energies, radioactive decay, and regulatory requirements constrain their application. Synchrotron sources offer much higher brightness and tunable excitation, allowing measurements just above selected absorption edges.

In energy-dispersive X-ray spectroscopy, a semiconductor detector converts each absorbed photon into charge. The collected charge is approximately proportional to photon energy, so electronic processing sorts pulses into energy channels. Silicon drift detectors are widely used because they combine good energy resolution with high counting-rate capability and relatively convenient cooling. A representative resolution is about 125–150 eV full width at half maximum at the manganese Kα line near 5.9 keV, although specifications depend on detector size, operating temperature, and processing settings. Detector windows and inactive surface layers are particularly important for low-energy radiation.

In wavelength-dispersive X-ray spectroscopy, a crystal or multilayer analyzer selects wavelengths according to

nλ=2dsin⁡θ.n\lambda=2d\sin\theta.

Here nn is diffraction order, dd is the analyzer-plane spacing, and θ\theta is the Bragg angle. Scanning instruments change the analyzer and detector angles to measure different lines; simultaneous instruments use multiple fixed channels. Wavelength dispersion generally separates nearby lines more effectively than conventional semiconductor energy dispersion, while energy dispersion acquires a broad spectrum without mechanical scanning. Performance nevertheless depends on the complete design, not simply on the label. All tube-based systems also require radiation shielding and controlled operating geometries: analytical convenience does not eliminate the hazards of the primary X-ray beam.

Measurement geometries and analytical variants

Conventional bulk XRF illuminates an area of a specimen and collects fluorescence at a fixed angle. The analyzed region is not defined solely by the visible beam spot. Incident radiation penetrates the material, and emitted photons must escape before reaching the detector. Both paths depend on composition, density, photon energy, and angle. Low-energy lines may originate mainly from a shallow surface layer, while higher-energy lines can sample substantially greater depths. Thus, “bulk composition” usually means the composition inferred from a finite, energy-dependent sampling volume, not a direct average of the entire object.

Reflection geometry is common for solids, powders, and liquid cups, but specialized arrangements address other problems. Thin specimens may be analyzed with reduced self-absorption, making intensity more nearly proportional to elemental mass per unit area. Coating instruments exploit differences between film and substrate signals to estimate layer thickness and composition. Such measurements require models of the layer sequence, densities, and attenuation. A gold signal from a plated object, for example, does not by itself establish whether the object is solid gold, thickly plated, or covered by a very thin surface layer. Multiple excitation conditions or complementary measurements may be needed to distinguish these cases.

Total-reflection X-ray fluorescence (TXRF) uses an incident beam at a very small angle to a smooth carrier, below the relevant critical angle for total external reflection. This minimizes penetration into the support and reduces background. A small liquid aliquot is commonly deposited on the carrier and dried, often with an internal standard. Under suitable conditions, the deposit behaves approximately as a thin specimen, allowing sensitive multielement analysis from a small amount of material. The benefits depend on deposit uniformity, cleanliness, carrier quality, and the absence of excessive residue; a thick or irregular deposit can reintroduce absorption and quantitative bias.

Handheld XRF places a compact tube and detector close to the sample surface, making rapid field measurements possible. Its geometry is especially sensitive to gaps, curvature, roughness, coatings, and specimen thickness. A reading from a corroded metal surface may characterize corrosion products rather than the underlying alloy. Atmospheric absorption further limits low-energy measurements, although some systems use helium purging or other arrangements to improve light-element response. Specialized vacuum laboratory instruments can measure elements lighter than those routinely accessible in field instruments. No single elemental range applies to all XRF systems: excitation, optics, windows, atmosphere, and specimen type jointly determine what can be measured reliably.

Samples, preparation, and representativeness

For many materials, the dominant analytical difficulty is obtaining a specimen representative of the material of interest. A spectrometer can produce highly repeatable results from a small area of a rock while missing coarse mineral grains elsewhere in the same rock. Sampling, crushing, splitting, and grinding therefore often contribute more to total error than photon counting. Particle-size effects arise because different minerals have different absorption properties and because analyte-rich grains may not be uniformly distributed. Mineralogical effects can persist even when two powders have the same overall elemental composition, especially when grain dimensions are comparable with relevant X-ray absorption lengths.

Pressed powder pellets are widely used for ores, soils, and industrial materials. Grinding improves homogeneity, while pressing creates a stable, relatively flat surface. A binder may improve mechanical strength but also dilutes the specimen and can introduce contaminants. Milling equipment must be chosen with the analytes in mind: steel can contribute iron and chromium, while tungsten-carbide equipment can introduce tungsten and cobalt. Moisture changes both the apparent concentration on an as-received basis and the attenuation of X-rays. Results therefore need a clearly defined reporting basis, such as dry mass, together with preparation procedures consistent with those used for calibration standards.

Fused glass beads provide another major preparation route. A powdered sample is dissolved at high temperature in a flux, commonly based on lithium borates, to produce a homogeneous glass. Fusion greatly reduces particle-size and mineralogical effects and is especially valuable for major-element analysis of rocks, ores, and cement-related materials. Its disadvantages include dilution, possible volatilization of some components, contamination from flux or crucibles, and the need to control oxidation and fusion behavior. Loss on ignition is often measured separately when water, carbon dioxide, organic matter, or oxidation-related mass changes are important to interpreting the original specimen.

Liquids are commonly measured in cups fitted with a thin X-ray-transmitting film. Film composition and thickness influence low-energy transmission; chemical compatibility, evaporation, settling, bubbles, and leakage can all affect results. Airborne particles may instead be collected on filters, while solutions intended for TXRF are deposited as small residues. Artworks, archaeological objects, and manufactured components are often examined without removing material, but avoiding destructive preparation can make geometry and heterogeneity harder to control. “Nondestructive” describes the treatment of a particular object, not an unconditional property of every XRF procedure: grinding and fusion are destructive, and intense or prolonged irradiation can alter sensitive specimens.

Quantitative analysis and matrix corrections

A peak identifies an element more directly than it establishes its concentration. Quantification requires relating measured intensity to the amount present while accounting for excitation, absorption, secondary emission, and instrumental response. Calibration commonly uses reference materials with certified or otherwise well-characterized compositions. Standards resembling the unknowns in composition and physical form are especially valuable because many systematic effects then vary less strongly across the calibration range. A simple proportional relationship may be adequate for a restricted material class, but broad compositional variation usually requires additional corrections.

Absorption illustrates the central role of the specimen matrix. For a homogeneous material, transmitted intensity approximately follows the Beer–Lambert law:

I=I0exp⁡ ⁣[−(μρ)ρx],I=I_0\exp\!\left[-\left(\frac{\mu}{\rho}\right)\rho x\right],

where μ/ρ\mu/\rho is the mass attenuation coefficient, ρ\rho is density, and xx is path length. In XRF, both the incoming excitation and outgoing fluorescence are attenuated. For monochromatic excitation of a homogeneous flat specimen, the primary intensity from element ii contains a depth integral of the form

Ii∝Ci∫0texp⁡ ⁣[−μ0zsin⁡α−μizsin⁡β] dz,I_i\propto C_i\int_0^t \exp\!\left[-\frac{\mu_0z}{\sin\alpha} -\frac{\mu_i z}{\sin\beta}\right]\,dz,

where CiC_i is concentration, tt is specimen thickness, α\alpha and β\beta are angles measured from the surface, and μ0\mu_0 and μi\mu_i are linear attenuation coefficients at the excitation and fluorescence energies.

Matrix effects also include enhancement. Fluorescence from one element can excite another if its energy exceeds the second element’s absorption edge. The resulting secondary fluorescence adds intensity beyond that produced directly by the source. Fundamental-parameter methods calculate these effects using atomic data, source spectra, geometry, and a model of sample composition. Empirical methods instead estimate correction coefficients from standards; hybrid approaches combine physical modeling and calibration. “Standardless” analysis still relies on instrument characterization, atomic databases, and assumptions about the specimen. It is not equivalent to a measurement free of calibration or model dependence.

Reporting conventions add another interpretive layer. Rock and cement analyses often express elemental results as oxide equivalents, such as total iron reported as Fe₂O₃. This convention does not establish that all iron actually occurs in that oxidation state or mineral phase. Similarly, software normalization to a total of 100% can conceal unmeasured constituents or preparation problems. Reliable quantitative work therefore includes independent quality-control materials, checks for drift and contamination, and an explicit treatment of measurement uncertainty. Metrological traceability rests on documented links to appropriate references, not merely on agreement between repeated readings.

Spectral interpretation, sensitivity, and limitations

An XRF spectrum contains characteristic lines, scattered source radiation, continuum background, and instrumental artifacts. Spectral fitting must distinguish overlapping contributions rather than simply assign every local maximum to an element. A classic interference is arsenic Kα near 10.54 keV and lead Lα near 10.55 keV, which conventional energy-dispersive detectors cannot resolve as separate narrow peaks. Other lines, such as lead Lβ, help constrain the interpretation. Wavelength-dispersive instruments can improve separation in many difficult cases, but no instrument eliminates the need to consider line families, expected intensity ratios, and plausible sample composition.

Detector artifacts can resemble genuine elements. In a silicon detector, an escape peak can occur when a silicon characteristic X-ray leaves the active volume after absorption of a higher-energy photon, producing a recorded energy roughly 1.74 keV below the incident energy. Pulse pile-up occurs when closely spaced photon events are processed as a combined pulse, potentially creating a sum peak. High count rates also produce dead-time losses and can degrade energy resolution. Digital corrections improve performance within defined operating ranges, but inappropriate source settings can produce biased intensities or false identifications. Blank measurements and spectra collected under alternative excitation conditions help distinguish real signals from artifacts.

Detection limits depend on analyte, matrix, line choice, instrument, and counting time. Many routine laboratory applications reach concentrations in the low milligram-per-kilogram range for favorable elements and matrices; other combinations are substantially less sensitive. A simplified statistical expression is

CDL≈kBS,C_{\mathrm{DL}}\approx\frac{k\sqrt{B}}{S},

where BB is background counts in the relevant spectral region, SS is net sensitivity in counts per unit concentration, and kk reflects the chosen detection criterion. When background and analyte counts both scale with acquisition time, the ideal background-limited detection limit improves approximately as the inverse square root of time. Longer counting does not correct systematic errors from contamination, spectral modeling, or sample heterogeneity.

Detection, quantification, and accurate classification are different tasks. A small fitted peak can indicate possible presence without supporting a precise concentration. Conversely, a high-concentration result can be precise but biased if the matrix model or specimen geometry is wrong. Hydrogen and helium are not measured by ordinary XRF, and light-element capability varies greatly among instruments. Routine XRF also has limited sensitivity to oxidation state and chemical bonding, despite small chemically dependent changes in line energies and shapes. For many practical questions, it is therefore combined with diffraction, mass spectrometry, or other forms of spectroscopy rather than treated as a complete description of the material.

Industrial, environmental, and cultural applications

Industrial XRF often emphasizes speed and reproducibility within well-defined material classes. In cement manufacture, measurements of calcium, silicon, aluminum, iron, sulfur, and other elements support control of raw mixtures, clinker, and finished products. Steelworks and nonferrous-metal producers use XRF to monitor alloy composition, while recyclers use portable instruments for sorting. However, many common handheld systems cannot measure carbon, an important limitation when distinguishing steel grades whose classification depends on carbon content. Coating analysis uses calibrated intensity models to determine quantities such as zinc coating mass on steel or the thickness of selected metal finishes.

In geology and mining, XRF supports major- and trace-element characterization of rocks, ores, drill material, and mineral-processing streams. Fused beads are particularly useful for consistent major-element analysis, while pressed powders often retain better sensitivity for selected trace elements because they avoid flux dilution. Portable measurements can rapidly reveal spatial trends but may not substitute directly for laboratory results from dried, homogenized specimens. Measurements along sediment cores also respond to changes in water content, grain size, and physical structure, so intensity profiles require careful interpretation before being treated as concentration records.

Environmental applications include screening soils for lead, arsenic, and other metals, analyzing particulate matter collected on filters, and measuring contaminated sediments or industrial residues. The United States Environmental Protection Agency’s Method 6200 describes field-portable XRF determination of elemental concentrations in soil and sediment. Its use illustrates the distinction between rapid field characterization and fully controlled laboratory analysis: moisture, heterogeneity, particle size, and spectral interference can materially affect results. XRF measures total elemental signal within its sampling volume; it does not by itself establish chemical species, mobility, bioavailability, or the origin of contamination.

In archaeology and cultural heritage research, XRF identifies compositional differences among ceramics, glasses, metals, pigments, and other materials. Obsidian artifacts can sometimes be assigned to geological source groups using characteristic trace-element patterns and suitable comparative databases. Paint analysis may identify elements associated with pigments, such as mercury in vermilion or copper in several blue and green pigments. Yet an elemental identification rarely proves a unique pigment: lead, for example, can occur in several pigments and preparation layers. Surface alteration, restoration, coatings, and underlying layers complicate interpretation, making complementary methods and contextual knowledge essential.

Microanalysis, imaging, and specialized research

Micro-XRF focuses the incident beam to a small spot using apertures, capillary optics, or other focusing systems. Moving the specimen through the beam produces a spectrum at each position and therefore elemental distribution maps. Laboratory instruments commonly achieve spatial resolutions on the scale of tens of micrometres, while specialized systems can reach smaller dimensions. Synchrotron nanoprobe instruments can attain resolution on the scale of tens of nanometres under suitable conditions. Spatial resolution must nevertheless be distinguished from chemical sensitivity, analyzed depth, and effective sampling volume; improving one does not automatically improve the others.

Elemental mapping reveals structures that single-point measurements miss. In paintings, maps can distinguish distributions of pigment-associated elements and sometimes reveal concealed compositions beneath visible layers. In geological specimens, they show zoning in mineral grains and associations among trace elements and mineral phases. Biological applications include mapping iron, zinc, copper, and other elements in cells and tissue sections. These maps are not automatically absolute concentration images: thickness variation, self-absorption, overlapping structures, and preparation artifacts can change apparent intensity. Fixation, dehydration, and embedding can also redistribute or remove mobile elements.

Depth-sensitive approaches extend the technique beyond conventional surface scanning. Confocal micro-XRF defines an overlapping excitation and detection volume with focusing optics, enabling depth-selective measurements in suitable specimens. Fluorescence tomography combines measurements from multiple positions and angles to reconstruct three-dimensional elemental distributions. Both approaches face attenuation problems because incident and emitted photons travel through different amounts of material. Reconstruction therefore requires geometric modeling and, often, corrections for self-absorption. Strong absorption can make some regions poorly accessible even when the nominal beam size is small.

Tunable synchrotron excitation also links XRF with X-ray absorption spectroscopy. By scanning incident energy across an absorption edge and monitoring fluorescence, researchers can obtain absorption spectra from dilute elements for which transmission measurements are difficult. The near-edge region can provide information about oxidation state and local coordination, while extended fine structure probes neighboring atoms and distances. This chemical information arises from the energy dependence of absorption, not merely from a conventional fluorescence peak. Advanced experiments must also account for radiation damage, detector saturation, and fluorescence self-absorption, particularly when studying concentrated samples or radiation-sensitive biological materials.