Reading an XRD pattern: what the peaks can and cannot tell you
Powder X-ray diffraction is often the first tool used to examine a synthesized ceramic or functional oxide. It is powerful, but a diffraction pattern is not a direct photograph of a crystal structure. It is a measured intensity distribution that must be interpreted using crystallographic models and experimental context.
Start with the measurement
A diffraction peak occurs when the scattering condition satisfies Bragg’s law:
[ n= 2d ]
Here, () is the X-ray wavelength, (d) is the interplanar spacing and () is the Bragg angle. A shift in peak position can therefore indicate a change in lattice spacing, but it can also arise from specimen displacement, zero error, residual stress or differences in measurement setup.
A defensible interpretation workflow
1. Phase identification
Compare the observed peak positions and relative intensities with reference patterns for chemically plausible phases. Do not assign a phase from one isolated peak if several reflections should be present. Consider overlapping peaks, preferred orientation and possible minor phases.
2. Lattice-parameter trends
If a series of samples shows systematic peak shifts, calculate lattice parameters using a consistent calibration and refinement procedure. A trend may be consistent with substitution or defect-induced lattice changes, but it does not independently prove which species occupies which crystallographic site.
3. Peak broadening
Peak width can contain contributions from instrumental resolution, finite coherent domain size, microstrain and overlapping reflections. Before interpreting broadening as crystallite-size reduction, account for instrumental broadening and state the model used. A Scherrer estimate is not automatically the same as a particle size measured by microscopy.
4. Rietveld refinement
Rietveld refinement fits a calculated diffraction profile to the entire observed pattern. The result depends on the structural model, background, peak-shape function, instrumental parameters and refinement strategy.
A lower residual is not, by itself, proof that the structural model is correct. Inspect the difference curve, parameter uncertainties, correlations and physical plausibility. Refine occupancies or complex defect models only when the data contain enough information to support them.
Common overinterpretations to avoid
- “No impurity peaks means 100% phase purity.” XRD has finite detection limits; amorphous or low-concentration phases may not be obvious.
- “A peak shift proves successful doping.” Instrumental offsets, strain and composition changes can produce similar observations.
- “Smaller peaks mean smaller particles.” Peak width reflects coherent diffraction domains and strain as well as size.
- “A good fit proves the proposed structure.” Model validity requires chemical, crystallographic and physical consistency, not just a numerical fit statistic.
What should be reported?
At minimum, state the radiation source, scan range, step size, geometry, sample preparation and reference database or structural model. For Rietveld work, describe the refined parameters, constraints, background and peak-profile treatment, and report fit indicators alongside the difference curve.
Further reading
- R. A. Young (ed.), The Rietveld Method, Oxford University Press (1993).
- H. M. Rietveld, “A profile refinement method for nuclear and magnetic structures,” Journal of Applied Crystallography 2, 65–71 (1969). https://doi.org/10.1107/S0021889869006558