Risks of using membrane filtration for trace metal analysis and assessing the dissolved metal fraction of aqueous media - A study on zinc, copper and nickel

Membrane filtration is commonly performed for solid-liquid separation of aqueous solutions prior to trace metal analysis and when assessing 'dissolved' metal fractions. Potential artifacts induced by filtration such as contamination and/or adsorption of metals within the membrane have been investigated for different membrane materials, metals, applied pressures and pre-cleaning steps. Measurements have been conducted on aqueous solutions including well-defined metal standards, ultrapure water, and on runoff water from corroded samples. Filtration using both non-cleaned and pre-cleaned filters revealed contamination and adsorption effects, in particular pronounced for zinc, evident for copper but non-significant for nickel. The results clearly show these artifacts to be non-systematic both for non-cleaned and pre-cleaned membranes. The applied pressure was of minor importance. Measurements of the labile fraction by means of stripping voltammetry clearly elucidate that membrane filtration followed by total metal analysis cannot accurately assess the labile or the dissolved metal fraction. - Highlights: → Membrane filtration for trace metal analysis can introduce significant artifacts. → The dissolved metal fraction cannot be assessed by membrane filtration. → Non-specified filtration procedures are inadequate for scientific studies. → Artifacts caused by membrane filtration need to be addressed by regulators. - Membrane filtration cannot be used to assess the dissolved metal fraction of aqueous media and needs to be defined in detail in standard tests.