(548d) Towards the Development of Safe and Sustainable By Design Chemicals, Materials and Products

The safe and sustainable by design (SSbD) concept is defined as a process to accelerate widespread market uptake of new and alternative chemical products and technologies that deliver greater consumer confidence in their safety, environmental and societal benefits and advance the transition towards a circular economy and climate-neutral society. Currently SSbD is at the core of institutional, process and technological systems innovation in the chemical industry in Europe. Given that the criteria for defining specifically the SSbD concept have been coined by the European Commission, the need to move towards the operationalization of SSbD is a key objective of the EU partnership on chemical risk assessment (PARC). Thus, an SSbD toolbox integrating tools for safety and sustainability assessment coming from different policy areas and strategies as well as new tools developed in PARC is being designed.

The overall PARC toolbox comprises interoperable models building in the EC SSbD framework, covering both the requirements for (a) newly developed chemicals and materials and (b) existing ones. The framework supports the redesign of their production processes by evaluating alternative processes to improve their safety and sustainability performance on the basis of SSbD criteria.

Towards this aim, a comprehensive framework that accounts for both the safety and sustainability criteria will be built, addressing the steps proposed by the EC SSbD framework, pertaining to:

• Step 1: Hazard assessment (intrinsic properties):
• Step 2: Human health and safety aspects of production and processes
• Step 3: Human health and environmental aspects in the final application phase
• Step 4: Environmental sustainability assessment
• Step 5: Social and economic assessment

Based on the above and aiming at the operationalisation of the EC SSbD framework, the PARC toolbox proposes a methodological framework that builds technically on innovative and beyond-the-state-of-the-art methods for hazard, exposure (both occupational and for the general population), human health risk assessment, and sustainability assessment. Starting from the user input where parameters related to chemical structure and a list of parameters, tools of different levels of complexity will be used to support the SSbD assessment.

Regarding the safe-by-design aspects, the source-to-impacts continuum is followed, aiming to provide the functional integration of environmental releases, environmental concentrations, multi-pathway and multi-route exposure, internal dose assessment, early biological effects, adverse outcome pathways, risk assessment and health impact assessment. With regard to impact sustainability, Life Cycle Assessment (LCA), protocols (ISO 14040, ISO 14044 and the International Reference Life Cycle Data System) is followed. Several methodological choices need to be made to assess the impacts associated to a certain product or process and to predict system performance upon upscaling, especially when prospective LCA is applied based on lab-scale data. The criteria to assess the quality of data and apply cut-off points, the methodology to be used to handle material recycling at the end of lifetime, and the way the environmental burdens and benefits are allocated among the different value chains, must be carefully selected, while different methodological options exist for the impact assessment, since these methodological choices can drastically influence the results of an LCA.

It has to be noted that:

• Tools of different level of complexity will be implemented for the assessment of chemicals, materials and processes across the different stages of innovation.
• An open architecture is followed in the computational toolbox allowing for the ready integration and use of new tools when appropriate; the overall methodology is built on the concept of concurrent engineering, able to integrate the methodological advances (e.g. omics, NAMS) that are of regulatory relevance.

A broad array of models is available within PARC, able to address both safety and sustainability issues. The aim of this work is the development of functional workflows (model pipelines) that involve the best suited tools for the specific user request. These pipelines are delivered automatically, either in the form of a new “pipeline tool”, (e.g. linking portal) or a dashboard. The automated pipeline requires that the tools have interfaces for automated running of the models. The pipelines will be tested and refined with the case studies through a continuous iterative process following a concurrent engineering approach. This is critical in order to continuously adapt the workflows based on the emerging needs of the SSbD community, as well as for the integration of novel tools.

Equally important is to ensure link with databases that relate to various aspects of chemical safety and sustainability such as data on chemical structures, tonnage, phys-chem properties, toxicity, ancillary exposure data, carbon footprint and other sustainability indexes. Databases considered include the ZZS similarity tool, ECHA CLP C&L Inventory, PubChem, eChemportal, ECOTOX, RISCTOX, NORMAN, CompTox Chemicals Dashboard, Registry of Toxic Effects of Chemical Substances (RTECS), CAMEO Chemicals, Chemical Entities of Biological Interest (CheBI), ChemSpider, ChemView, OpenFoodTox, ChEMBL, OSHA Occupational Chemical Database, openLCA nexus and Ecoinvent.

Another important feature of the developed workflows is that they follow a tiered approach regarding the complexity of tools and methods, relying on the stage gate innovation model. In this sense, models of different level of complexity are employed based on the innovation stage and data availability. Various models will be interconnected through APIs. Models for safety aspects include INTEGRA, OECD SAAToolbox, ConsExpo, OECD QSAR Toolbox, QSARs lab database, VEGAHUB, GreenScreen for Safer Chemicals AMBIT tool (LRI) and ECETOC TRA. With regard to sustainability, models in the pipelines include ProScale, EF-LCIA Method, SubSelect, USEtox and INTEGRA-LCA.

The study presented herein concerns the development of an integrated solution with a user-friendly interface as well as user requirements for minimum data needs. A central interface is created aiming at the description of the pipeline of the SSbD toolbox. Initially, this provides links to the individual model user interface included in the toolbox for early and late stages of innovation. Later on, user interfaces will be created for model pipelines that will be tailored for specific SSbD assessment requests based on the specification of potential scenarios. The user interface will be web-based; in some cases (for specific tools) they may be a client-side application, executed from the web browser window that will guide the user through the pipeline of the SSbD toolbox.

Compounds of interest in the case study presented herein include bisphenol A, F, S and bisphenol alternatives. All human exposure routes and pathways to the toxic compounds were integrated at the level of systemic internal dose, that was linked with adverse health outcomes on the endocrine system with cascade impacts on human reproduction, metabolic syndrome and, even, neurotoxicity after chronic exposure to the persistent organic chemicals found in plastic products and waste to quantitatively assess the related health risk. On top of that, the various dimensions of sustainability have been investigated as well. The toolbox is also empowered with quantitative structure-activity relationship models for property prediction supported by a machine learning-based search engine that explores the relevant chemical space for alternative molecules that would eventually minimize the environmental and human health hazards associated with their use in consumer products. Thus, an integrated framework for selection of chemicals for product redesign has been coined; the computer-aided workflow we present supports the redesigning of these chemicals to achieve the sought of environmental and human health objectives using in silico generated structure suggestions.