Design Frameworks for Safer Chemicals

Cleaning products are routinely used in commerce, but specific ingredients within a product often range in persistence (P), bioaccumulation (B) and toxicity (T) (e.g., endocrine, skin sensitization, ecotoxicity) characteristics, thus presenting diverse risks to human health, biodiversity and ecosystem services. Subsequently, various approaches have examined hazards and risks of individual chemicals used in cleaning products, aimed to identify safer chemical ingredients, and labeled specific products as part of broader efforts to prevent, reduce or eliminate pollution (e.g., Safer Choice), Because cleaning product ingredients represent diverse classes of chemicals, our research team has examined hazards of cleaning product ingredients by class, by usage, and among toxicology endpoints. Such information has supported derivation of data-driven thresholds of toxicological concern and uncertainty factors by chemcial classes and identification of less hazardous chemical classes for substitution. Because specific cleaning products are inherently composed of chemical mixtures, our team has recently advanced probability based approaches to examine specific product hazards (e.g., PBT, skin sensitization, mixture assessment or allocation factor) by usage categories and by marketing strategies to consumers, including incorporation of Indigenous values. This approach provides a framework scaleable to other product types and to support sustainable molecular design of less hazardous chemicals.

The Chemical Hazard Data Trust (“Data Trust”), managed by ChemFORWARD, is a public interest asset that provides standardized chemical hazard data to support the selection of safer functional chemical alternatives. Originally designed to inform decision-making by the consumer product supply chain, the Data Trust has demonstrated broader utility supporting systems change for regulators, educators, investors, and other parties interested in advancing the adoption and proliferation of safer chemistry.

The Data Trust houses comprehensive GHS-based chemical hazard assessments (CHAs) spanning human and environmental endpoints that are completed by credentialed toxicology firms and independently peer reviewed. Overall hazard scores, “Hazard Bands” are assigned to all chemicals following an “A” to “F” scale with “?” for unknowns, enabling rapid identification of chemicals as high hazard, potentially safer, or uncharacterized, supporting efficient screening and prioritization. This model scales equitable access to verified and trusted hazard data across all value chain members and stakeholder groups.

Data sharing and transparency reduces cost and time while avoiding duplicative, potentially inconsistent, hazard identification efforts. This shared infrastructure supports three critical activities: (1) identification of potentially safer functional chemical alternatives, (2) promotion of verified safer trade-name materials in commerce, and (3) development and tracking of hazard-based metrics over time. This presentation will highlight how value-chain collaboration and shared hazard data accelerate safer product and material design.

Presenters will illustrate real-world applications of Data Trust integration, including overview of the web interface, hazard endpoint classification methodology, and use as an educational tool in Beyond Benign’s Green Chemistry Commitment signatory classrooms. Case studies will demonstrate (1) supplier collaboration to address data gaps and identify SAFER™ verified on-market alternatives, and (2) use of aggregated hazard metrics to visualize trends to drive innovation, improve safer material selection, and highlight data-generation priorities across a dataset of more than 48,000 beauty and personal care products.

Effective innovation in designing safer chemicals must come from the realization that predicting a chemical’s hazard is insufficient; rather, a comprehensive design approach should consider not only the safety of a chemical but also the intended functionality. Physicochemical properties influence chemical behaviors and understanding the underlying mechanisms of these behaviors leads to safe and functional chemical design. Much of the information needed to elucidate the relationships between properties, function, and hazard exists in databases or is obtainable through quantum chemical calculations. However, integration of such information that leads to clear design rules translatable across multiple chemical classes is difficult due to the diversity of chemical space, lack of unified functionality definitions, data uncertainty, and other considerations. In this work, the steps necessary for designing safer chemicals considering both efficaciousness and toxicity are examined via a case study of substances that serve the function of preventing the harmful effects of UV radiation. Challenges and decisions that arise throughout the process are addressed and existing databases, information, and design criteria that are useful for informing chemical toxicity and functionality are highlighted. Approximately 250 substances that serve to prevent the harmful effects of UV radiation were identified from several databases and literature sources. This chemical set was used to determine commonalities in physicochemical properties, including electronic properties computed from density functional theory (DFT), that influence the UV radiation protective functionality. Boundaries of the functional chemical space were then extended to safe chemical space as defined by applying property-based design rules for reduced toxicity, specifically properties related to bioavailability such as molecular weight, octanol-water partition coefficient, and melting point. Integration of information about both function and safety within a defined chemical space led to the identification of low to no hazard chemical candidates capable of performing the desired function.

Baseline toxicity or narcosis is the minimum toxicity of all chemicals that is caused by partitioning of the chemicals into cellular membranes. If a chemical exceeds the critical membrane concentration the membrane integrity is compromised, which leads to cell death. Baseline toxicity is not dependent on the type of molecule partitioning into the membrane, but only depends on the molar concentration of the chemicals in the membrane. The baseline toxicity concept has originally been used in aquatic toxicity studies and has rarely been applied in human toxicology, but recently it has been expanded to human cell lines. For the development of safer chemicals, we propose to use the baseline toxicity concept in conjunction with specificity analysis as guiding principles. Applying these concepts for chemicals hazard assessment has four major advantages: (i) it makes toxicity predictable, not only for individual chemicals, but also for mixtures, (ii) it helps to choose the correct dosing concentration for toxicity experiments, avoiding artifacts and unnecessary repeats, (iii) it allows a better interpretation of existing toxicity data by comparing the measured effect concentrations with predicted baseline toxicity and (iv) it can help to fill data gaps for difficult-to-test chemicals, e.g., compounds with low aqueous solubility. For the development of safer chemicals, the baseline toxicity concept can help to identify by which mechanisms a given chemical is causing toxicity. Safe chemicals should cause cytotoxicity only via membrane disruption and not by other mechanisms like DNA damage or oxidative stress. Furthermore, measured effect concentrations for different modes of endocrine disruption can be compared to measured or predicted cytotoxicity to identify chemicals of concern with high specificity, i.e. for which effect concentrations are observed far below cytotoxic concentrations. The degree of specificity can also be predicted by structural alerts and similarity methods. By combining these principles, chemical structures with low hazard potential can be identified and applied in future chemical design.