Introduction High aspect ratio materials (HARM) like carbon nanotubes (CNT) exhibit material properties that enable innovative applications but also raised concerns about potentially harmful effects to humans due to their asbestos-like morphology. Control banding by grouping of HARM by hazard- and exposure-related properties is a promising approach to risk assessment and risk mitigation for a large family of materials. Limit values enabling a differentiation of HARM with low, moderate or high dustiness as well as information about the dust morphology are necessary for control banding.

Methods We have developed a dustiness test for powdery HARM, in particular for CNTs. In the Shaker method, a laminar low volume air flow passes through a vertically vibrating powder column resulting in powder fluidization. Vibration is required to overcome adhesive forces between powder grains that would otherwise hinder fluidization. The Shaker method combines aerosol monitoring over the dust generation process to determine the emission intensities with simultaneous dust sampling. Subsequent sample analysis by means of scanning electron microscopy (SEM) obtains information about dust morphology.

Results and discussion Using a standard operation procedure, we performed dustiness tests on 20 CNTs. The results lead to material rankings based on the emission intensity and the definition of limit values for low, moderate and high dustiness. The SEM-aided morphological analysis allowed us to rank the materials for the propensity to emit individual fibres and their grade of agglomeration. We also measured the diameter and length distributions to identify potential rigid fibres and those matching the WHO-criteria for hazardous fibres. These results are the basis of our proposed new grouping strategy for control banding, which classifies HARM using a risk matrix that considers both intrinsic material and process-related properties such as bio-durability, toxicity as well as dustiness, grade of agglomeration and presence of hazardous WHO-fibres.