| Thursday, October 31, 2019 @ 11:00 am EDT — Thursday, October 31, 2019 @ 12:00 pm EDT | ||||
| Read across in nanosafety research: Dissolution behaviour and classification of a library of 37 nanomaterials in simplified physiological media Anastasios G. Papadiamantis,1,2 Emily Guggenheim, 1 Antreas Afantitis,2 Sophie Briffa, 1 Georgia Melagraki,2 Iseult Lynch1 and Eugenia Valsami-Jones1 1. School of Geography, Earth and Environmental Sciences, University of Birmingham, UK 2. Nanoinformatics Department, NovaMechanics Ltd., Cyprus The dissolution of engineered nanomaterials (ENMs) is a complex process, which remains poorly understood despite being one of the key prerequisites for the risk assessment of ENMs, and having a direct influence on toxicological and eco-toxicological effects. It thus requires a fundamental review in order to be described and parameterised, both for the sake of a better (eco)toxicological assessment of ENM behaviour, as well as their regulation. As there are currently no standardised ENM dissolution protocols, it is hard to identify the physicochemical, morphological and/or structural ENM parameters driving the process. The dissolution of a range of fully characterised (chemical composition, morphology, coating, coating charge, size, polydispersity index, hydrodynamic diameter, ζ-potential, energy band gap, specific and geometric surface area, core metal electronegativity, valency and atomic/ionic radii) ENM from the EU-FP7 project NanoMILE library was studied, using a protocol which monitored their dissolution up to 48 hours, in simplified physiological media (pH: 1.5 and 7.0). In general, ENM and bulk analogues demonstrated the highest ionic release at low pH, although this was not always the case suggesting a potential process of dissolution/reprecipitation. It was observed that certain descriptors (chemical composition, size, morphology, surface area, core metal valency) were common for the whole duration of the experimental procedure at both pH values. At the short-term (2-hour) time point, electronegativity was also found to significantly affect dissolution for both groups, while pH-specific significant descriptors were also identified: ζ-potential (pH: 1.5) and atomic/ionic radii (pH: 7.0). A classification model was also developed, based on the categories defined during the OECD case study on Ag ENMs. The model was in good agreement with the statistical analysis of the acquired experimental results and the most significant descriptors driving dissolution were identified. By applying the EnaloskNN algorithm (Enalos Chem/Nanoinformatics tools) we were able to study the selected training neighbours for each test ENM. In that way, it was possible to search for patterns and similarities in the neighbourhood space, do a grouping of the ENMs and qualitative illustrate the neighbouring relationships in a read across concept. The applicability domain, based on Euclidian distance of the used descriptors, was also defined to promote the model in practical applications and ensure predictions reliability. |
||||
| Export to My Calendar (ics) | ||||