What Preparation Method is Better for Elemental Impurity Analysis, Exhaustive Extraction or Total Digestion?

The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) Guideline Q3D and United States Pharmacopeia (USP) Chapters <232>/<233> outline a risk-based approach when analyzing elemental impurities (1,2). In a recent study, a team of researchers delved into these guidelines, discussing the main challenges laboratories encounter when implementing the regulations in these documents.

Published in the Journal of Trace Elements and Minerals, the work also tested the reproducibility of mercury (Hg) and vanadium (V) using inductively coupled plasma–mass spectrometry (ICP-MS) and X-ray fluorescence (XRF) spectroscopy. In Part II of our conversation with James Harrington of RTI International in Research Triangle Park, North Carolina, who was the lead author of this study, as well as coauthor Donna Seibert of Kalamazoo, Michigan, they talk about the reproducibility for Hg and V, as well as the ICP-MS and XRF results compare to one another.

James Harrington (right) and Donna Seibert (left), authors of this study. | Image Credit: © James Harrington and Donna Seibert

Can you discuss the reproducibility for mercury (Hg) and vanadium (V)? What factors contributed to the variability in these elements?

The easy answer is for vanadium. This element is primarily found as the ⁵¹V isotope in nature (99.75% abundant), and this isotope suffers from a chloride-based polyatomic interference in standard resolution ICP-MS (³⁵Cl¹⁶O). At the time we performed the study, there were most often two approaches used to adjust for this interference and improve accuracy at trace concentrations: spectral resolution on a high-resolution ICP-MS and reaction cell technology with ammonia gas. However, none of the participating labs used a high-resolution ICP-MS for analysis, and only one laboratory used ammonia gas as the reaction cell gas, while others used helium or a hydrogen:helium mixture. This was not surprising because of the relatively specialized nature of ammonia reaction cell use and our stipulation that laboratories would not be required to purchase special consumables to participate in the study. He and He/H₂ gas approach works well at higher concentrations, but can’t fully break down the interference at ultra-low concentrations. Therefore, we found that in the raw materials where no V was expected, most laboratories reported measurable concentrations, which we considered false positives. On the one hand, this phenomenon often won’t have an impact because it is rare that V is included in a product risk assessment, but it is still important to be aware of this phenomenon in cases where it could play a role.

In the case of mercury, all the laboratories demonstrated significant loss of the element from all the tableted test materials compared to the reference values. This observation was also consistent with the 2014 study (3). We believe that this may be a result of volatilization of the element from the test material over time rather than an analytical or preparation effect, and that the physical properties of the tablet’s matrix may not be favorable for retaining the volatile element. Notably, the silicon dioxide material from which Hg was sourced did not demonstrate a similar drop in Hg concentration over time. Additional experiments may be necessary to verify the hypothesis and determine the implications for risk assessment and best laboratory practices.

The study compared exhaustive extraction and total digestion. Can you elaborate on why total digestion exhibited lower variability and what this means for laboratories choosing a preparation method?

Our working hypothesis is that total digestion may provide a more uniform and aggressive digestion approach that mitigates noise and matrix effects. However, that’s not to say that labs should necessarily resort directly to a total digestion method for this analysis. There’s still a practical tradeoff of requiring additional safety protocols because of the use of hydrofluoric acid or fluoroboric acid to ensure a complete digestion of all remaining sediment. This requires specialized equipment, facilities, and training for lab personnel and is overall not feasible for many laboratories. The approach we took during method development was to use the most aggressive digestion approach first to see how close to the expected concentrations we could get and then separately develop and optimize an exhaustive extraction method to compare the results. Our reference laboratory was able to obtain comparable results for both methods, which were also acceptably close to the expected concentrations (even mercury, which is consistent with our hypothesis above, because the MD work was done shortly after the tablets were prepared). Our analysis of the reported concentrations from our participants using both methods was also consistent with this observation.

These observations suggest that both exhaustive extraction and total digestion can provide comparable concentrations for the same material; however, it is important that wherever possible, this should be a goal of the method development effort that is backed up with data to justify the use of a less harsh digestion method.

The study also included X-ray fluorescence (XRF) spectroscopy. How did XRF results compare with ICP-MS, and what are the advantages and limitations of XRF for elemental impurity analysis?

In this study, we found that some reported elemental concentrations were comparable between labs performing ICP-MS and those performing XRF analysis. This was notably the case for As, Cd, and Ni, although other elements (including Co and Pb) were comparable to some extent. As discussed earlier, Hg and V were also problematic for XRF analysis. Although the reproducibility for almost all elements was better by XRF, this observation should be put in the appropriate context. Fully empirical XRF analysis requires a series of matrix-matched standards to mitigate any matrix effects, which requires a significant mass of tablet or raw materials. Since all the tablets produced here were made specifically for this study and in small batches, it was necessary to prepare only single replicates and reanalyze to give an idea of within-laboratory variability, an approach that would be expected to naturally result in less variability. Standards were produced using a fundamental parameters approach where levels of each material were intentionally varied to create calibrations for each element. However, sample limitations will not always be the case, and the comparable accuracy, limit of detection, and reproducibility will make XRF a desirable option.

References

1. Harrington, J. M.; Seibert, D. S.; Williams, G.; et al. The Product Quality Research Institute Elemental Impurity Interlaboratory Study: Results and Implications for Industry. J. Trace Elem. Miner. 2025, 12, 100227. DOI: 10.1016/j.jtemin.2025.100227
2. Thermo Fisher Scientific, Elemental Impurities Analysis Information. Thermo Fisher. Available at: https://www.thermofisher.com/us/en/home/industrial/pharma-biopharma/pharma-biopharma-learning-center/pharmaceutical-qa-qc-information/elemental-impurities-information.html (accessed 2025-05-09).
3. D.S. Seibert, Analytical Challenges/Round Robin Study, PQRI/USP Workshop on Elemental Impurity Requirements in a Global Environment – Next Steps?, Silver Springs, MD, 2015.