ACEnano – Proficiency testing Single particle Inductively Coupled Plasma Mass Spectrometry
Nanoparticles are increasingly used and a vast number of nanotechnological products including consumer products, are entering the market. To determine whether a product contains a nanomaterial, particle size and number concentration need to be measured. The EU project ACEnano innovates and optimises analytical techniques to detect and characterise nanoparticles. Proficiency testing is an integral part of the ACEnano project and the development of a proficiency testing scheme for nanomaterial analysis to assure comparable performance of laboratories is foreseen.
Participation in proficiency tests is essential to improve or maintain the quality of a laboratory. In 2018 the first of these proficiency tests was conducted for the detection and characterisation of nanoparticles using single particle inductively coupled plasma mass spectrometry. Not only ACEnano partners but also laboratories outside the ACEnano consortium were encouraged to participate which resulted in a total number of twenty-six participants. Participants were encouraged to use their in house method and practice experience to select the most appropriate conditions for the spICP-MS analysis. For guidance a reference was made to ISO/TS 19590:2017.
For particle size the lowest value reported is 30.2 nm and the highest value is 80 nm. The consensus value is 61 nm with a robust standard deviation of 5.3 nm. This is comparable to the fit-for-purpose value of 10% of the consensus value (=6.1 nm). With regard to the accuracy one result was questionable and two were unsatisfactory. The particle size results and performance of the participants are shown in Figure 1 and 2.
Figure 1. Particle sizes as determined by the participants
Figure 2. Graphical representation of the reported results as z-scores for the particle diameter
The lowest value reported for particle number concentration is 1·107 parts/l and the highest value is 9.776·1013 parts/l. The consensus value is 1.44·1013 parts/l with a robust standard deviation of 6.5·1012 parts/l. This is more than 2 times higher than the fit-for-purpose value of 20% of the consensus value (=2.9·1012 parts/l). With regard to the accuracy three results were questionable (PT025, PT026 and PT9994) and five were unsatisfactory (PT021, PT023, PT030, PT9955, and PT9995). Results and performance of all participants are shown in Figure 3 and 4.
Figure 3. Particle number concentrations as determined by the participants
Figure 4. Graphical representation of the reported results as z-scores for the particle number concentration.
Seventeen labs showed optimal performance by reporting a correct diameter and a correct particle number concentration. A total number of eleven questionable/unsatisfactory z-scores was reported, four for the particle size and seven for the particle number concentration. For the latter, some of the questionable/unsatisfactory z-scores are possible reported because the dilution factor of 1000000 was not taken into account. The full and detailed results of this study can be found in RIKILT report 2018.519, dd. November 2018.