Appendix II F. X-Ray Fluorescence Spectrometry

PRINCIPLE OF THE TECHNIQUE

X-ray fluorescence (XRF) analysis is based on measurements of the X-rays emitted by the constituent atoms of a sample when it is excited by an external source of radiation. If sufficiently energetic radiation impinges on an atom of the sample material, it may eject 1 of the inner-shell electrons of that atom. The vacancy created is filled by 1 of the electrons from an outer, higher-energy shell. The energy difference between the 2 electron shells involved in the process is released in the form of fluorescent X-rays. These X-rays are characteristic since their energy is specific to each element (atom). By measuring their energy and intensity, qualitative and quantitative data about the elemental composition of the test material is obtained.

XRF spectrometry is suitable for liquid, solid and powdered materials and is widely used as a means of screening pharmaceutical ingredients and products for toxic elements or elemental impurities, for quality control and in-process testing. It is also used to identify inorganic foreign elements within falsified medicinal products. As XRF can be non-invasive, it lends itself to process analytical technology (PAT) applications, such as the analysis of an unwanted trace catalyst in an active pharmaceutical ingredient.

EQUIPMENT

An XRF spectrometer (or analyser) consists of 3 essential components:

Depending on the X-ray detection method employed, either a wavelength-dispersive (WD) or an energy-dispersive (ED) XRF spectrometer is used.

In a WD-XRF spectrometer the X-rays from the sample are directed at a crystal, which diffracts them at precise angles depending on their energy. The intensity of the diffracted X-rays is measured sequentially by a detector counter.

In an ED-XRF spectrometer the X-rays from the sample are directed at a solid-state detector, which generates an electric pulse of an amplitude proportional to the energy of each X-ray photon detected. During measurement, the spectrometer acquires an X-ray spectrum of the sample that contains complete information about its composition. An ED-XRF spectrometer can also be combined with a scanning electron microscope (SEM). Substantial advances in miniaturisation and automation have led to the development of hand-held ED-XRF spectrometers for field measurements.

Some instruments are supplied with an initial factory-set calibration that allows semi-quantitative analyses to be carried out.

MATRIX EFFECTS AND INTERFERENCE

The intensity of the characteristic X-rays of the analysed elements is not necessarily linear with concentration, owing to matrix effects. The intensity of the fluorescence measured for a given element depends not only on the concentration of that element in the sample but also on the absorption of the incident and fluorescent radiations by the matrix. The presence and concentration of other elements (analytes) in the sample, the composition of the sample matrix, and the particle size of the sample material are known to contribute to matrix effects. The presence of matrix effects must be taken into account in any calibration method utilised for quantitative determination.

At low concentrations the linearity is usually well preserved, which greatly facilitates calibration of the spectrometer. The intensity of the fluorescent radiation emitted by an element in a given matrix and at a given wavelength is then proportional to the concentration of that element and inversely proportional to the mass absorption coefficient of the matrix at that wavelength.

SAMPLE PREPARATION

It is essential that the sample is sufficiently thick that the intensities of characteristic X-rays measured are not affected by variations in sample thickness.

Liquid samples Samples are analysed ‘as is’, provided that the solution consists of a clear, single phase and has sufficiently low volatility. A special liquid-sample holder and a commercially available support window composed of a suitable polymer film (transparent to X-rays and solvent resistant) are required. Alternatively, liquid samples can be transferred onto the surface of a disc and dried before analysis.

Powdered samples Samples may be analysed ‘as is’ in special X-ray sample cups with bottoms made of a thin polymer film, transparent to X-rays. After transferring the powder to a sample cup, the cup is tapped gently until no further settling of the powder is observed. If necessary, more powder can be added to the cup after tapping. Another alternative for preparation of a powdered sample, which is better suited to WD-XRF analysis, is to press the sample powder into a pellet, with a binder (for example cellulose powder, wax or ethylcellulose) or without a binder. The mass of reference material and sample material must be about the same and the resulting pellets must be approximatively 5 mm thick or more.

Solid samples Samples are analysed by placing them directly on the spectrometer measuring window, making sure they cover it completely. Solid samples for WD-XRF may need to be cut into a uniform shape with a flat surface for reproducible analysis, whereas samples can be measured ‘as is’ when using ED-XRF, provided there is adequate sample depth.

Fusion The fusion bead method can be used to prepare solid and powdered samples (e.g. minerals or oxides) if the element of interest is not volatile. The sample is homogeneously mixed with a flux reagent (e.g. dilithium tetraborate) and transferred to a platinum vessel; a releasing agent and/or oxidising agent may be added if necessary, for example to prevent damage of the vessel. The mixture is heated at an appropriate temperature while swirling the vessel until a homogeneous melt is obtained. If necessary, the melt is then transferred to a flat-bottomed mould, kept in a horizontal position and cooled under conditions maintaining its property as a glass.

PROCEDURE

Measuring conditions

The instrument is set up and used in accordance with the manufacturer′s instructions. The measurements may be carried out under vacuum, nitrogen or helium to improve the sensitivity of the method for the quantitation of light elements.

Reference standards

Standards required for the calibration, system suitability or control of equipment performance are prepared from certified reference materials (CRMs). Standards with a high carbon load may be more representative for pharmaceutical applications.

Calibration

The calibration model selected must be fit for purpose. Various calibration methods are available, including the ‘fundamental parameters’ approach, empirical calibration, Compton/Rayleigh normalisation and multiple linear regression (MLR).

System suitability test

This test must be carried out before the analysis to ensure that the performance of the measurement system is satisfactory. It may also be performed to verify the calibration of the system.

Acceptance criteria The measurement system is suitable if the concentration obtained for a check material containing the element(s) of interest within the used concentration range does not differ from the actual concentration by more than 5 per cent for an assay, and 20 per cent for impurity tests; when using concentrations determined from reference methods such as atomic absorption spectrometry, the accuracy of the XRF calibration should be aligned to that of the reference method; in this case an acceptance criterion of 10 per cent for the assay can be more realistic.

Analysis

The samples are measured with the same parameters as used during calibration of the instrument.

CONTROL OF EQUIPMENT PERFORMANCE

These parameters are also applicable for equipment qualification.

Specific procedures, acceptance criteria and time intervals for characterising XRF performance depend on the instrument and its intended application. Demonstrating stable instrument performance over extended periods of time provides some assurance that reliable measurements can be obtained.

The following tests may be carried out at appropriate intervals defined according to the user’s quality system procedures to ensure the adequate performance of the XRF instrument.

x- and y-axes

It is recommended that the x-axis (energy or peak angle) and y-axis (intensity) are verified at least on installation and thereafter at appropriate intervals that are defined according to the user’s quality system procedures. Consideration is to be given to peak position in ED-XRF and to peak angle in WD-XRF when verifying the x-axis, and to the count rate when verifying the y-axis.

Detector resolution

Calculate the resolution (total width at half-height) at the energy used during calibration of the instrument.

Acceptance criteria The resolution value does not deviate by more than 20 per cent for an assay, and 25 per cent for identity and elemental impurities tests, from the value determined during calibration of the instrument.

VALIDATION REQUIREMENTS

The objective of an XRF method validation is to demonstrate that the measurement procedure is fit for purpose (assay, content uniformity, limit tests and identification tests). Where sample preparation is necessary, it is essential that test materials are spiked before any preparatory steps. For example, if a test material is to be digested, the material must be spiked at the beginning of the digestion procedure.

For the determination of impurities, the validation requirements are given in general chapter 2.4.20. For other purposes, validation is performed according to the relevant ICH guidelines.