Trace nickel analysis in chloride-containing pharmaceutical materials: How matrix-matching keeps Epsilon 4 XRF accurate

X-ray fluorescence (XRF) spectrometry is widely used in the pharmaceutical industry to both quantitatively and qualitatively analyze elemental impurities in starting materials, intermediates, and active pharmaceutical ingredients (APIs).

To quantify these impurities, XRF compares the API sample against external calibration standards. These standards contain known concentrations of the target element, dispersed in a bulk powder such as cellulose.

Why cellulose calibration breaks down in chloride-containing samples

Cellulose is used because of its elemental similarity to organic APIs, which are mainly made up of carbon, hydrogen, nitrogen, and oxygen. This means that cellulose absorbs X-rays weakly and predictably, much like an organic API.

When the standards and the sample absorb X-rays to the same degree, the concentrations XRF reports are accurate. When they don’t, the numbers drift – and when you’re testing elemental impurities against ICH Q3D limits, a calibration that drifts off target is a compliance risk.

However, cellulose only mimics organic samples that share its light-element profile. When a pharmaceutical material contains a heavier element – for example, chlorine – that element absorbs far more of the fluorescence signal than carbon, hydrogen, nitrogen, or oxygen do.

The sample will therefore absorb the fluorescence signal more strongly than a plain-cellulose standard. This means that the calibration no longer applies, and the analyte is recovered outside the 70-150% recovery window expected in pharmaceutical elemental analysis – the benchmark set by the established technique, inductively coupled plasma (ICP) analysis.

One fix is standard addition, which spikes known amounts of analyte into the sample. It’s an effective solution, but it needs at least three standard preparations per batch, sacrificing the speed and simplicity that make XRF advantageous in the first place.

What is the matrix-matching method?

Matrix-matching, a method developed and optimized by Barcelona-based pharmaceutical company Esteve CDMO, keeps calibration simple instead.

How much a material absorbs X-rays is captured by a single value known as its mass attenuation coefficient (MAC), which combines predictably when you blend materials. This lets you tune a standard’s absorption by mixing cellulose with a strong absorber like sodium chloride (NaCl), raising the standard’s MAC until it matches the sample’s.

One matched set of standards then quantifies the sample accurately, with no per-sample spiking.

Quantifying Ni in a chloride-containing API using Epsilon 4

In a recent application note, we describe how we analyzed nickel (Ni) in a chloride-containing starting material. We calculated the starting material’s MAC was more than double that of plain cellulose. Blending cellulose with NaCl in a 4.8:1.0 ratio matched the standards MAC to the sample’s exactly.

Measurements were performed on our Epsilon 4 EDXRF spectrometer in air, with no vacuum pump or helium required. Following ICH Q3D, where the limit of Ni for a 10 g daily dose is 20 ppm, we worked to an internal limit of 5 ppm, well within the 30% control threshold. Recoveries across spiked samples came in at 92-99%, comfortably inside the acceptable range.

Our Epsilon 4 EDXRF spectrometer is built for this kind of analysis:

• Non-destructive: Little to no sample preparation is required, and no sample is consumed.
• No vacuum or helium required: Epsilon 4 operates in air for most applications, cutting overhead time and running costs.
• Calibrated for years: A low-drift X-ray tube paired with automatic drift correction keeps results compliant for years without re-calibration.
• Compliance-ready: Epsilon 4 supports elemental impurity testing in line with ICH Q3D, USP<232>, and USP<233>, with optional Audit Trail Software available for GMP/GLP environments and FDA 21 CFR Part 11.

Keeping it simple: Why matrix-matching matters

Matrix-matching corrects matrix effects without standard addition. Because you calibrate once against matched standards instead of spiking every batch, quality control workload drops significantly. Plus, the method only needs ~200 mg of material, analyzed non-destructively, which is ideal for a pharmaceutical setting where sample is often scarce.

Download our application note for the full method and validation data behind our Epsilon 4 spectrometer trace analysis of Ni in pharmaceutical compounds.