To the Editor

Parallel to your Editorial 'Join the dialogue' (Nature Nanotech. 7, 545; 2012), the Hazardous Substances Advisory Committee, which advises the UK government, also considered the subject of reporting standards for nanotoxicological research and has published its findings online (http://go.nature.com/wxNgtl). We highlight the key findings here.

Implementing consistent standards is broader than simply reporting appropriate physical and chemical characterization; it starts with problem formulation and hypothesis generation. At its broadest level, procurement of nanomaterials and their characterization can be differentiated for regulatory and research purposes. For regulatory purposes, the standards applied and data generation required must be more prescriptive, whereas for research these must be primarily based on the hypothesis to be tested.

However, minimum requirements in both areas include the reason for the choice of the nanomaterials, appropriate sample pre-treatment, physical and chemical characterization, and measurement of actual dose to which an organism is exposed. Sample pre-treatment must be considered because, for instance, the use of traditional high-vacuum microscopy and spectroscopy is inappropriate for hydrated and organic-rich samples; analysis can change the conformation, aggregation state and other nanomaterial properties. Characterization must also be performed using an appropriate multi-method approach to obtain unbiased measurements. For example, the indiscriminate use of dynamic light scattering to measure the size of polydisperse, aggregated and non-spherical samples should be discouraged because the data produced are inaccurate. Electron microscopy should be made quantitative and additional detectors should be used to extract useful parameters such as shape, elemental composition and chemistry from a sufficient number of randomly chosen images. Dose, which is the concentration of toxicant to which an organism or cell is exposed, is a fundamental parameter in toxicology. Because dynamic changes (such as sorptive losses to container walls, aggregation or dissolution) can occur during exposure of nanomaterials to cells, the use of nominal (added) dose is dubious. Dose and characterization, therefore, must be measured over the appropriate exposure time and conditions. The details of and data from metrology and characterization should be fully reported in any published paper. Without such information the comparability and utility of biological hazard data is highly questionable.