X-ray fluorescence mapping makes images of spatially indexed characteristic X-ray emission. It can identify where elements contribute signals, but a brightly coloured pixel is not automatically an exact concentration, chemical compound or independently resolved object.
The gold standard is a calibrated mapping process with explicit beam geometry, spectra, counting statistics, reference standards and uncertainty. Follow the measurement chain before accepting a quantitative chemical claim.
eduKateSG · X-ray Fluorescence Mapping
Find your next learning step
Calculate scan dimensions, count rates and noise, then evaluate the evidence for a quantitative elemental map.
Full chapter index · Practice and transfer · Education root · How X Works
Full chapter index
Understand elemental maps · Chapters 1–2
Calculate pixel measurements · Chapters 3–4
Check spectral physics · Chapters 5–7
Interpret the sample · Chapters 8–10
Explore connected methods · Chapters 11–12
CHAPTER 1 OF 12 · Understand elemental maps
1. Elemental maps begin with an X-ray emission measurement
X-ray fluorescence (XRF) occurs when an incident X-ray excites a sample atom and a subsequent electronic relaxation produces characteristic emitted X-ray photons. Detecting their energies can reveal which elements contributed to a measured spectrum.
Scanning XRF microscopy records a spectrum or selected elemental-line counts from many spatial positions. Those spectra are associated with calibrated scan coordinates, producing a chemical image after suitable detector corrections and spectral interpretation.
The Diamond Light Source I18 microfocus beamline describes two- and three-dimensional XRF mapping of heterogeneous materials. Its examples span minerals, biological material and other scientifically complex specimens, but beamline capabilities are not universal instrument specifications.
A pixel is generally a measurement from a finite illuminated footprint and X-ray interaction volume, not the concentration at an infinitely small point. The observed intensity depends on element amount, excitation, attenuation, counting statistics and the geometry of the experiment.
The gold standard reports the sample, illuminating energy, beam width, scan step, pixel integration time, spectral processing, detector response, coordinate reference and the exact physical statement inferred from the map.
CHAPTER 2 OF 12 · Understand elemental maps
2. Elemental emission is not the same as chemical state or structure
Ordinary single-energy XRF maps can reveal spatial variation in the emitted lines associated with an element. They do not, on their own, uniquely identify its oxidation state, molecular compound, biological binding partner or crystalline phase.
X-ray absorption near-edge structure (XANES) varies incident energy across a core-electron absorption edge to investigate electronic environments. X-ray diffraction investigates structural order; X-ray computed tomography reconstructs attenuation geometry. These are related but distinct measurements.
An instrument may combine XRF with XANES, diffraction or coherent imaging. Their evidence still needs separate calibration, measurement boundaries and uncertainty reporting rather than treating one rich illustration as proof of everything visible.
Fluorescence photons can be absorbed before leaving the sample. A strong element concentration deep inside a thick object may produce less detector signal than a weaker region near its surface, even when excitation is otherwise comparable.
An XRF map therefore supports progressively stronger claims only with progressively stronger evidence: element presence, relative spatial patterns and finally carefully corrected quantitative composition under a validated calibration.
CHAPTER 3 OF 12 · Calculate pixel measurements
3. Worked example: scan positions, pitch and acquisition time
Imagine a hypothetical grid of 60 scan positions horizontally and 40 vertically. The experiment contains 60×40 = 2,400 nominal spatial acquisitions, assuming one recorded spectrum per grid position.
Let the commanded step be 2 micrometres along each direction. If each sample is represented as a square 2-µm-wide image cell, the nominal image extent is 120 µm by 80 µm. But the distance between the first and last scan-centre coordinates is 59×2 = 118 µm and 39×2 = 78 µm.
These two dimensions use different definitions. The scan-centre span and the plotted image-cell width are both legitimate if they are explicitly labelled; confusing them can create an apparent geometry or scale disagreement.
At 0.10 seconds nominal dwell per point, exposure adds to 2,400×0.10 = 240 seconds, or four minutes. Stage movement, settling, detector readout, dark-field checks and other activities may make wall-clock time longer.
This calculation does not establish radiation dose or spatial resolution. Real dose also depends on incident photon flux, energy, sample absorption and exposure history; spatial resolution is controlled by physical beam and detector responses as well as the scan step.
CHAPTER 4 OF 12 · Calculate pixel measurements
4. Worked example: compare counts after dwell-time normalization
Suppose a fitted elemental peak produces 12,000 counts at one pixel during 0.10 seconds. The arithmetic count rate is 12,000/0.10 = 120,000 counts/s, assuming the counts are appropriately defined.
At a second hypothetical pixel, a comparable peak produces 6,000 counts during only 0.05 seconds. This gives 6,000/0.05 = 120,000 counts/s, the same dwell-normalised rate despite half the raw counts.
That equality does not prove equal element concentration. Beam flux might differ, detector dead time may not be negligible, sample thickness could differ and spectral background or absorption could have changed.
Compare total counts, time-normalized counts, incident-flux-normalized counts and fitted concentration maps as different products. A chart title that merely says ‘XRF intensity’ does not tell the reader which one it shows.
A credible cross-session analysis records dwell time per pixel, photon flux monitors, detector corrections, beam position and energy calibration. Any conversion to concentration needs a further physical and reference-material model.
CHAPTER 5 OF 12 · Check spectral physics
5. Worked example: photon counting uncertainty
An idealized independent photon-counting process may be approximated by a Poisson distribution, whose variance equals its mean. For a count N in that approximation, standard deviation is about √N.
For a hypothetical pixel with 100 relevant photons, √100 = 10, giving 10% relative counting standard deviation. At 10,000 photons the standard deviation is 100, or about 1% of the count.
More photons improve this particular statistical precision but do not automatically remove systematic errors. A mistaken elemental peak assignment or faulty detector calibration can remain consistently wrong at very high counts.
A weak line fitted above background has uncertainty from both the line and the estimated background, and potentially from overlapping elemental lines. The precision of its area can therefore differ materially from the naive √N result.
Longer dwell can raise photon counts but also increase sample modification or detector saturation risk. The scientifically useful acquisition balances uncertainty, spatial coverage, specimen preservation and instrument capabilities.
CHAPTER 6 OF 12 · Check spectral physics
6. Overlapping XRF peaks and detector artifacts require spectral fitting
Real XRF spectra contain multiple lines, elastic and inelastic scatter and instrument contributions. An energy-dispersive detector has finite energy resolution, so some elemental peaks overlap or lie close enough to require model-based separation.
A simple energy-window count can include photons from multiple components. For element-specific inference, a suitable fitted spectrum may be needed, including a justified continuum and the expected line energies and relative intensities.
Detector dead time, pulse pile-up and escape peaks can complicate high-count-rate interpretations. A saturated bright region may look extremely rich in a chosen element even when the response is no longer linear.
Retain representative raw spectra from apparently rich, poor and anomalous map locations. A spectral residual and calibration record can reveal a false assignment that a smoothed colour image hides.
NIST’s micro-XRF chemical-imaging work emphasizes limitations from the source spectrum, sample geometry and matrix effects. Calibration and spectral interpretation should be demonstrated rather than assumed from a visually plausible elemental label.
CHAPTER 7 OF 12 · Check spectral physics
7. Spatial resolution is not identical to grid pitch
Suppose a hypothetical XRF beam has approximately 3 µm effective width and the scan grid advances by 1 µm. Many adjacent samples overlap. The map can have one-micrometre pixels without independently resolving two objects separated by exactly one micrometre.
Effective resolving performance also depends on the illumination point-spread function, mechanical drift, stage motion, sample thickness, detector collection and any interpolation or deconvolution.
Argonne’s Advanced Photon Source describes dedicated micro- and nanoprobe facilities, with capabilities that vary by beamline. A quoted small beam size belongs to the particular configuration and does not remove specimen-dependent limitations.
Use calibrated coordinate references and show scale bars. Repeated images should be registered using actual sample or fiducial features, especially when comparing a treatment or reaction before and after exposure.
A proper resolution claim uses relevant measured evidence such as reference structures or a suitable instrument response characterization, not just the number of image pixels.
CHAPTER 8 OF 12 · Interpret the sample
8. Thickness, geometry and matrix effects change map brightness
Characteristic fluorescence photons can be reabsorbed in a specimen before reaching the detector. The amount depends on their energy, the surrounding material and the path length through the sample.
Two regions with equal elemental mass can therefore have different observed intensity if one lies deeper or beneath a more attenuating matrix. Sample density, geometry and orientation may also change incoming excitation and emitted-photon collection.
Quantitative XRF methods can require reference materials, measured sample thickness, fundamental-parameter modelling and appropriate absorption and enhancement corrections. Which approach is valid depends on the specimen and instrument.
An ideal thin uniform film cannot automatically calibrate a rough heterogeneous rock, hydrated tissue or multilayer structure. The latter systems can introduce multiple depth and matrix effects that must be explicitly handled.
Report whether the display is qualitative element contrast, comparable normalized counts, estimated areal mass or an independently validated concentration. NIST notes that micro-XRF can be excellent for qualitative chemical imaging while remaining highly sensitive to matrix and geometry.
CHAPTER 9 OF 12 · Interpret the sample
9. Specimen preservation and radiation exposure influence the record
Biological cells, polymers and other sensitive specimens can change through irradiation, dehydration, movement or temperature variation during a long scan. A map of an evolving sample is not equivalent to one measurement of a static original state.
Argonne’s Bionanoprobe includes cryogenic sample-handling capabilities and specialized trace-element mapping. The aim is to support suitable sample preservation and high-resolution analysis, not to guarantee all damage mechanisms disappear.
Repeated short checks or comparison of early and late acquisitions can help reveal drift or sample modification. However, those checks also contribute exposure and must be planned under the instrument’s qualified safety procedures.
Sample preparation and coating can introduce contamination or redistribute analytes. The location of a bright fluorescence patch should be interpreted alongside specimen history, mounting and, where appropriate, correlative microscopy.
Keep raw spectra, scan order, beam conditions, preparation notes and exposure record in the scientific archive. Without them, apparent changes between maps may be artefacts of acquisition rather than genuine chemical redistribution.
CHAPTER 10 OF 12 · Interpret the sample
10. Quantitative maps need calibration and independent cross-checks
A fluorescence line’s count rate depends on fluorescence yield, incident spectrum, detector sensitivity, specimen geometry, attenuation and other factors. Raw signal magnitude is therefore not a direct universal abundance scale.
Imagine one pixel displays 200 fitted counts per second and another 400 under apparently comparable acquisition. The second reports double the count rate, but its elemental concentration need not be double when matrix, thickness or excitation differ.
A validated quantitative method reports the reference materials, calibration function, detection limit, correction model, measurement uncertainty and applicable concentration and geometry range. Values outside that range need clear qualification.
Independent checks may use suitable chemical analysis, a reference specimen, repeated scans or another imaging modality whose observation matches the claim. XANES can sometimes add oxidation-state context; ptychography may provide structural imaging; neither automatically validates absolute XRF concentration.
The strongest conclusion distinguishes directly recorded photon counts from inferred composition. If two alternative physical explanations remain consistent with the map, identify what additional measurement would separate them.
CHAPTER 11 OF 12 · Explore connected methods
11. Practice, FAQs and connected X-ray learning routes
Practice A: A grid of 60×40 positions contains 2,400 spectra. At 0.10 seconds each, ideal integration time is 240 seconds. Why does a real scan usually take longer?
Practice B: Compare 12,000 counts in 0.10 seconds with 6,000 counts in 0.05 seconds; both correspond to 120,000 counts/s. Why might their concentration estimates still differ?
Practice C: Estimate relative Poisson counting uncertainty for 100 and 10,000 photons: approximately 10% and 1%. Name two non-Poisson uncertainty sources.
FAQ — Is an XRF colour map an elemental concentration map? Only with a suitable calibration and interpretation model; it may instead show intensity.
FAQ — Does 1 µm pixel spacing mean 1 µm achieved resolution? No. Illumination and sample-response physics can limit resolvable detail.
FAQ — Does bright XRF alone identify a chemical oxidation state? Not ordinarily; appropriate spectroscopic evidence is needed.
Follow the connected Gold Standard reading routes to distinguish XRF from X-ray absorption, phase imaging, tomography and the statistical basis of element-mapping uncertainty.
These verified eduKateSG routes connect the different measured quantities and keep their evidence boundaries visible:
- The Gold Standard of X-ray Fluorescence — the photon-emission measurement
- The Gold Standard of X-ray Absorption Near-Edge Structure — chemical-state interpretation
- The Gold Standard of X-ray Ptychography — coherent imaging and phase retrieval
- The Gold Standard of X-ray Optics — X-ray beam formation and detectors
- The Gold Standard of X-ray Computed Tomography — spatial attenuation rather than elemental mapping
- The Gold Standard of Spectroscopy — peak interpretation and spectral models
- The Gold Standard of Chemistry — element and compound distinctions
- The Gold Standard of Statistics — uncertainty and calibration
CHAPTER 12 OF 12 · Explore connected methods
12. Institutional sources, method limits and the next step
Diamond’s I18 microfocus beamline, Argonne’s Advanced Photon Source and NIST’s chemical-imaging programme describe real XRF instruments, their complementary methods and the fundamental problems posed by sample geometry and matrix effects.
Every scan size, dwell interval, photon count and brightness ratio above is an invented teaching case. These numbers are not a specimen analysis, instrument specification or safe dose recommendation.
A useful learning project creates a transparent measurement ledger for one map: excitation energy, beam profile, scan positions, raw spectra, dwell and incident-flux normalization, spectral fit, matrix model, detection limit, reference materials and uncertainty.
Primary institutional and scientific sources:
- NIST — X-ray Chemical Imaging Methods — matrix, source-spectrum and geometry limitations
- Diamond Light Source — I18 Microfocus Beamline — XRF mapping, micro-XAS and microdiffraction
- Diamond — I18 Beamline Information — instrument energy and spatial response
- Advanced Photon Source — 2-ID-D X-ray Microscopy — fluorescence microscopy and complementary modes
- Argonne — Bionanoprobe — cryo trace-element mapping and structural methods
- Advanced Photon Source — PtychoProbe — complementary ptychography and fluorescence
The immutable Clementi Secondary 1 Mathematics reference illustrates teaching progression; it is not an X-ray mapping operating manual.