Prediction Methods for the Environmental Fate of Organic Chemicals

The ability to identify hazardous substances before they are, either purposefully or inadvertently, emitted into the environment is of high priority. Complete removal of xenobiotic chemicals from the environment entails transformation into innocuous products such as biological components (biomass) or complete mineralisation (CO2 and inorganic salts). Consequently, regulatory assessment is based on biodegradability experiments. The experiments are performed in a tiered manner with increasing complexity, where failure to meet a certain threshold in a simpler test triggers further testing. The more complex simulation tests are designed to emulate biodegradation in specific environmental compartments (for example, soil, sediment, and surface water). These tests commonly involve the application of chemicals labelled with stable or radioactive carbon isotopes (13C, 14C). In these tests, a significant portion of the carbon (up to 90%) remains unidentified as non-extractable residues (NER). For many years, NER have been considered a ‘black box’ as NER are commonly quantified from combustion (complete oxidation) of the sample. How NER are formed and the risks they pose have been topic of research for more than 60 years. Experimental insights of the past decade have unequivocally shown that microorganisms incorporate the carbon label into bio-molecules and use the chemical as a source of energy and nutrients . Thus, a considerable fraction of NER is in fact of biogenic origin (bioNER) posing no risk. The key parameter relating biodegradation to the growth of microorganisms is the growth yield, which is defined as the mass of microorganisms formed per mass of substrate consumed. The parameter can be determined experimentally, however, due to its widespread application in environmental biotechnology, fermentation and reactor design in wastewater treatment; many methods based on thermodynamic considerations have been developed and are routinely applied to estimate this parameter without resorting to experimentation. The scope of the PhD thesis was to develop predictive equations to estimate the formation of bioNER and develop a method to estimate the microbial growth yields using thermodynamic considerations and a minimum of information.
Predictive equations, able to estimate the formation of biogenic non-extractable residues, were derived from carbon mass balances. Only information of the microbial growth yield and measurements of the evolved CO2 are needed as input. In order to estimate the microbial growth yield, the Microbial Turnover to Biomass (MTB) method was developed and used to estimate the growth yield on common carbon substrates and on xenobiotics. The MTB method considers the nutritional value of the substrate – i.e. how much substrate is needed per mass of microorganism, and relies only on knowledge of the Gibbs energy of reaction and the carbon content of the chemical. The reduction in complexity compared to other methods reduces the number of confounding factors and amount of information needed while increasing the method flexibility. The MTB method outperformed two other yield estimation methods when estimating the growth yields on pesticides, but was outperformed when used to estimate the growth yield on common carbon substrates.
Coupling biodegradation to microbial growth and formation of bioNER using the growth yield provides a powerful tool in degradation models. The ‘unified model for biodegradation and sorption’ was successfully used as the fundamental framework to model the competing processes of biodegradation and NER formation. The model employs a two-compartment-sorption model calculating rapid (adsorption) and slow (sequestration) kinetics combined with Monod kinetics for microbial metabolism and growth. The developed modelling approach was shown able to capture complex phenomena such as competing transformation pathways and water-sediment mass transfer in a water-sediment simulation test (OECD TG 308) with incomplete transformation of glyphosate. The MTB method could provide additional insights into possible explanations for observed shifts of microbial degradation of glyphosate and accumulation of its main metabolite aminomethylphosphonic acid (AMPA), as these experiments employed multiple isotope labels in a single substrate.
Based on the findings presented in this PhD thesis, it is now possible to elucidate the nature of NER by combining novel analytics, first principles, and dynamic model simulations. The PhD thesis provides a theoretical foundation that can be used to predict bioNER formation prospectively but also retrospectively, as it can be used to interpret NER data obtained in already performed biodegradability experiments. Consequently, the methodology is recommended to be included in the coming update of the technical guidance document describing how to assess the persistent, bioaccumulative and toxic (PBT) properties of chemicals by the European Chemicals Agency (ECHA). However, the methodology is new and should be tested in further experiments (preferably using OECD tests relevant for persistence assessment).
As there is a need for standardised models to infer kinetic values needed to comply with the chemical regulation, it is envisioned to provide the MTB method and the dynamic model as a tool available to registrants and regulators.