Silica gel as a particulate carrier of poorly water-soluble substances in aquatic toxicity testing

Aquatic toxicity tests were originally developed for water-soluble substances. However, many substances are hydrophobic and thus poorly water-soluble, resulting in at least two major implications. Firstly, toxicity may not be reached within the range of water solubility of the tested compound(s), which may result in the formation of solids or droplets of the tested substance and consequently an uneven exposure. Secondly, because of multi-phase distribution of the tested substance it may be complicated to keep exposure concentrations constant. To overcome such problems, we have introduced silica gel as a particulate carrier in a toxicity test with the benthic copepod Nitocra spinipes. The main objective of the current study was to evaluate whether a controlled exposure could be achieved with the help of silica gel for testing single poorly water-soluble substances. A secondary objective was to evaluate whether an equilibrium mass balance model could predict internal concentrations that were consistent with the toxicity data and measured internal concentrations of two model hydrophobic substances, i.e., the polybrominated diphenyl ethers BDE-47 and BDE-99. Larval N. spinipes were exposed for 6 days to BDE-47 and BDE-99, respectively, in the silica gel test system and, for comparative reasons, in a similar and more traditional semi-static water test system. Via single initial amounts of the model substances administered on the silica gel, effects on both larval development and mortality resulted in higher and more concentration-related toxicity than in the water test system. We conclude that the silica gel test system enables a more controlled exposure of poorly water-soluble substances than the traditional water test system since the concentration-response relationship becomes distinct and there is no carrier solvent present during testing. Also, the single amount of added substance given in the silica gel test system limits the artefacts (e.g., increased chemical load in test system) that a semi-static renewal may introduce when testing substances that partition to non-water phases. However, measured and modelled internal concentrations did not match toxicity, which may indicate that chemical equilibrium was not reached during the test. Further experiments are thus needed to explain the processes behind the observed positive effects of silica gel and a kinetic model would likely also be more appropriate to describe the concentrations and distributions in the two test systems.