Fig. 3.

Correspondence between the VSSA-derived particle dimension d _VSSA and dmin_VSSA and the Feret_min from EM. In the left panel, the available data from the NanoDefine and the JRC/Eurocolour materials are shown, assuming the case of no pre-knowledge on the particle shape (i.e., D = 3 for all materials and no porosity correction). In the right panel, the same materials are shown, but this time dmin_VSSA was calculated using the best estimate for the number of small dimensions D (the shift which occurred in comparison to the case with D = 3 is indicated with black arrows), and the effect of an additional t-plot evaluation (always ISO method except for TiO₂ which is calculated according to Lecloux) is shown by red arrows. In both panels, a color code indicates the NM classification of the data points. In the yellow and green areas, the materials are consistently classified as nano and non-nano, repectively, by VSSA and EM. In the red and blue areas, both methods disagree on the classification. The error bars are the standard deviations resulting from size measurements (d ₅₀ for EM and dmin_VSSA, respectively) of different labs. Note that these are no combined measurement uncertainty budgets, which is particularly important in the case of platelet materials; as for them, EM cannot assess the smallest particle dimension