Feasibility study of a candidate reference material for ions in PM(2.5): does commutability matter also for inorganic matrices?

The existing Air Quality Directive 2008/50/EC establishes within the European Union (EU) member states limit values for fine air particulate matter (PM(2.5)) including the possibility to discount natural sources of pollution when assessing compliance with the legislation. In proving this, EU member states shall determine, amongst others, the rural background concentration of some anions (Cl(−), NO(3)(−) and SO(4)(2−)) and cations (Na(+), NH(4)(+), K(+), Ca(2+) and Mg(2+)). To deliver reliable data and to comply with the data quality objectives of the legislation, environmental control laboratories should use certified reference materials (CRMs) to validate or verify the performance of their analytical methods. Since no CRMs for anions and cations in PM(2.5) are presently available, we present the commutability issues encountered during the first attempt to develop such a material. We demonstrate that a dust, collected in a road tunnel and previously used for the production of two CRMs of a PM(10)-like material, does not behave as an authentic fine particulate matter collected according to EN12341:2014 when measured by an established method proposed by the European Committee for Standardization (CEN/TR 16269:2011). The water-soluble fractions of SO(4)(2−), NH(4)(+), K(+), Ca(2+) and Mg(2+) in a PM(2.5)-like candidate CRM produced from that road tunnel dust are only fully extracted after 3 h of sonication and not after 30 min, as stated in the method. Moreover, we found that the particle size of the test material influenced the extraction yield of K(+), Ca(2+) and Mg(2+), suggesting that these ionic species were incorporated in the core of the particles and inaccessible to the extraction procedure. These particular features make the material unsuitable for the measurements of ions with the CEN method. The difference in the extraction time can be seen as a commutability issue and the candidate CRM should be considered as not commutable with routine samples. This demonstrates that commutability studies should not only be considered for clinical CRMs, but also for inorganic CRMs when they are intended to be used to quantify operationally defined analytes.