Classification Strategies for Regulatory Nanodefinitions - Safety of Nanomaterials along Their Lifecycle - page 50

rods was determined, finding a median (D50) of 82.7 nm in number metrics, with a
standard deviation of 30.5 nm, and min/max values of 13 nm/208 nm. In summary
58% of the primary particles are below 100 nm. For the smaller axis even a smaller
D50 of 20 nm and even higher percentages below 100 nm result as the decisive cri-
terion. Without doubt, this is a nanomaterial in the sense of the EC nanodefinition
recommendation.
But a drastically simpler method also performs well for powder materials: Volume
specific surface area (VSSA) was acknowledged as an agglomeration-tolerant ensem-
ble method (Figure 3.1) with low cost and wide availability to identify nanomateri-
als (Allen 1997b; Kreyling, Semmler-Behnke, and Chaudhry 2010). VSSA has the
important advantage over classifying and counting techniques (including TEM) that
it does not involve dispersion protocols and achieves few-percent precision (Hackley
and Stefaniak 2013). In the specific case of the iron oxide pigment, the mass-specific
surface area is 80 m²/g, as determined by the BET method; multiplied by the mass
density of 3.9 g/cm³, one obtains the VSSA of 312 m²/cm³ and a sphere-equivalent
diameter of 20 nm, in agreement with the TEM evaluation. In a pilot round robin,
the VSSA measurements from six labs were reproducible within a scatter of 10% on
this substance (Gilliland and Hempelmann 2013). There are concerns that surface
modifications on particles may dominate the BET and thus induce false positive clas-
sifications, but in such cases the number size distributions are to prevail (EC 2011).
Further, false negative classifications by VSSA occur only for platelet-shaped
particles (Gilliland and Hempelmann 2013), suggesting that VSSA could also be a
reliable method to screen for non-nanomaterials in a full testing strategy.
3.3 METHODS TO MEASURE THE SIZE DISTRIBUTION
IN NUMBER METRICS: FOR SUSPENSIONS
The vast majority of nonmicroscopy based techniques for particle size measurement
( Fig u re 3.1)
1. Provide size distributions in terms of an equivalent spherical diameter that
is an average and not a minimum dimension measurement.
2. Interpret agglomerates and aggregates as individual particles.
Round-robin exercises (Lamberty et al. 2011) have typically employed relatively
monodisperse materials that may not convey the complications associated with
many real industrial materials, and have not substantiated the interexchangeabil-
ity of number and volume distributions. Fractionating, counting, and microscopic
techniques can still quantify smaller particles in the presence of agglomerates:
Methodical comparisons on deliberately mixed multimodal distributions have evi-
denced that fractionating methods (Figure 3.1) successfully identified nanomaterials
whereas ensemble methods failed and the innovative counting methods were too
operator-dependent (Anderson et al. 2013; Wohlleben 2012). However, validation
on well-dispersed spherical particles neglects irregular shapes and incomplete dis-
persion. A pilot round robin has compared the ensemble techniques (Figure 3.1) of
Dynamic Light Scattering (DLS), Laser Diffraction (LD), and Centrifugal Liquid