Smithers Chemistry experts presented the webinar “Practical Chemistry Approaches for “Difficult to Test” Substances in Environmental Risk Assessments.”  Read below for responses to the questions asked during the audience Q&A.

Q: Do you know if Smithers or your customers submitted any studies using SFC to any regulatory authorities, and were they accepted? 

A: Currently, Smithers does not use Supercritical Fluid Chromatography (SFC) system although there have been discussions regarding its deployment on to future studies. Whether regulators would accept studies performed with this type of system depends on the context. According to the “Guidance Document on Pesticide Analytical Methods for Risk Assessment and Post-approval Control and Monitoring Purposes” (SANTE/2020/12830, Rev.2), post-approval control and monitoring methods require instruments that are “commonly available” and SFC does not currently qualify. However, the document also states that methods developed for the purpose of risk assessment are not limited by this. Therefore, there would no reason for regulators not to accept data generated by SFC systems for risk assessment purposes. 

Q: When you say use "whole water analysis" for adsorptive compounds, do you mean GC Purge and Trap or something else?

A: In this context, “whole water analysis” means the entire sample is extracted and not a portion of the sample. Imagine we have 100 mL of an aqueous sample in a glass container with a given amount of an adsorptive compound. The compound may adhere to the glass vessels meaning it is not a homogenous mixture. If we then, for example, take a 10 mL portion for analysis, that portion (or aliquot) may not be representative. Instead, we would add our extraction solvent directly to the entire sample with the aim that the presence of the organic extraction solvent would prevent or reverse any such surface adsorption that may have taken place. This is why whole sample analysis can be important for substance that have a high potential adsorb to container surfaces.

Q: Do you work with protein? Would you consider it as a difficult test substance?

A: Protein-based test substances pose specific analytical challenges different from traditional chemical analysis. Our staff does have experience and knowledge with this type of analyte. However, there is little demand from our clients as we specialize in environmental risk assessment, and OECD-based regulators do not currently require us to test the environmental effects of artificially-derived proteins or similar large molecules. 

Most synthetic chemicals that have a potential to be released in the environment are small molecules so that remains our focus. Therapeutic protein-based medicines (such as biologics) are usually completely metabolized by biological systems and even if any were released in the environment, the perception is that these would quickly degrade and be rendered harmless, thus there is no need for a risk assessment. 

There is some disagreement on whether this is the correct stance for the regulators to take so Smithers will monitor the situation if anything changes. Finally, whether proteins would be considered “difficult” depends on the context. Stability, quality control and batch release assays often employ reverse phase and size exclusion liquid chromatography with ultra-violet detection, and this is usually quite straight forward. Quantification using liquid chromatography tandem mass spectrometry can present more challenges as collision induced dissociation (CID) or fragmentation patterns in the MS system can be extremely complex with multiple charged species. Structural elucidation increases the complexity even more requiring sophisticated deconvolution software. As with so many cases with chemistry, the real answer again is “it depends”.

Q: How can we choose the most relevant marker for a multi-constituent active substance like a UVCB? Should the ecotoxicity of each component be considered to select the most relevant marker?

A: This is a very good question, and one which is a source of some challenges. The answer is often that the choice is based on a series of value judgments advised by the relative abundance, relative toxicity and stability of the different moieties present in the test substance. A species that represents a high proportion of the total multiple components within the UVCB, but with a low (known or predicted) toxicity may be of less significance than a more toxic, but less prevalent one. Additionally, as more marker species are added, the cost of the analysis generally increases, and this can be a consideration. The best results in these situations have come following close discussions between sponsor, ecotoxicology experts and chemists. In general, regulatory bodies are more receptive to data where a larger proportion of the test substance is reported, although Smithers has not experienced any issues in this area when a well-argued case is made.  

Q: For UVCB substances with very complex composition, what is your advice to ensure that 100% of the substance has been identified and quantified, and you are not missing something? 

A: Continuing from the answer above, it is not always necessary to identify and quantify all components within a UVCB test substance. What is required is that a set of analytical markers that adequately capture the range of physical characteristics, toxicities and prevalences. This is best achieved via close discussions between the sponsor, ecotoxicology experts and chemists. As stated above, when a well-constructed case is presented to the receiving authority, it is unlikely that any issues will be encountered.

Q: What would you recommend for testing multicomponent biologicals like neem oil or other multi mixture plant extracts regarding the regulatory need to characterize individual components?

A: This question is also related to the two above. In the case of neem oil or (e.g.) pyrethrum extract, a multiresidue method to measure the limonoid/triterpinoids in neem oil or the pyrethrins in pyrethrum extract can be developed. The individual components can be measured against analytical standards of the actives. These can be obtained commercially, even as certified standards, but it should be noted that it can sometimes be time-consuming and expensive to do so. This approach permits characterization of the concentrations obtained and changes in concentrations during the tests. 

Q: Do you need to separate all 8 isomers of cypermethrin? If so, on what column? 

A: The answer is “it depends”. In the majority of risk-assessment type studies, separating and quantifying the four diastereoisomer pairs would be sufficient, and this can be accomplished on a number of different, readily available columns (both GC and HPLC/UHPLC). Separating all eight diastereoisomers is not routinely required, although Smithers has had requests for an assessment of whether the ratio of the diastereoisomers changes between the test substance as supplied and following exposure to organisms in an ecotoxicology test. These requests have significantly increased since the adoption of EFSA guidance regarding risk assessment of substances with stereoisomers. 

Separation of each isomer of Cypermethrin is more challenging, but we have accomplished this using a Daicel Chiralpak AD-3 column with 1% ethanol in hexane as an isocratic mobile phase. Our method also employed mass spectrometry which is a challenge with this type of mobile phase in terms of achieving sufficient ionisation with conventional Electrospray Ionization (ESI). We had to introduce a flow of 0.01 M ammonium acetate in methanol (post column) in order to facilitate ionization. This setup proved challenging to achieve sufficient reproducibility but was ultimately successful.

European Food Safety Authority (EFSA), Bura, L., Friel, A., Magrans, J.O., Parra-Morte, J.M. and Szentes, C., 2019. Guidance of EFSA on risk assessments for active substances of plant protection products that have stereoisomers as components or impurities and for transformation products of active substances that may have stereoisomers. EFSA Journal, 17(7).

Q: What is your approach to bound residues in metabolism studies?  Do you start with solvents with different dielectric constants (mild approach) and then move to harsh acid/base, etc.?  If you cannot release bound residue that is a high % of the residue, do regulators accept your best effort globally typically, or do they still want you to try to identify the molecules?

A: For metabolism studies, the guidelines we follow give a flowchart for the characterization/identification of unextracted test substance. This lists; dilute acid/base, surfactants, enzymes and then harsh acid /base as the steps to be taken (generally taken in this order). If all steps are taken and the test substance is still shown to be bound at significant levels, then we can show that best efforts have been made to release the activity and this in part adds to the characterization. For some plant studies we have seen this go slightly further, with further extraction techniques after enzyme treatment to further characterise the residues into fractions e.g. lignin, hemicellulose and cellulose. We have also performed a similar approach on milk where we fractionated the sample into its different proteins.

Q: Suppose the test substance is non-polar with very low aqueous solubility; what is your recommendation to ensure the test substance is uniformly distributed and aquatic test organisms are exposed to the target concentrations, even using the maximum recommended solvent concentration?

A: In cases like this, it is a critical part of the test to ensure that the exposure concentration is at, but does not exceed, the solubility of the test substance in the designated medium. The former is critical to reduce the possibility of false negative results, and the latter is an absolute Guideline requirement. Centrifugation of test solutions is a useful technique for removal of suspended, undissolved test substance, but often the difference in density between test substance and water is small, and long centrifugation periods are required. Filtration can be used, but it is vital to ensure that dissolved test substance is not adsorbed onto the filter as this will deplete the solution and return a concentration below true solubility.

Techniques such as Tyndall effect or critical micelle concentration can be useful in ensuring that a true solution, rather than a microsuspension has been generated, but both of these techniques become ineffective at very low concentrations. The single best approach in our experience at Smithers has been to obtain accurate solubility data for the test substance in the actual medium. It should be noted that the solubilites in ecotox media are often markedly different to the solubility of the test substance in pure water.

Q: ECHA are indicating that biodegradability tests on UVCB are no longer sufficient to demonstrate substance as "not persistent".  Has Smithers applied MS techniques to ready biodeg studies to show degradation of specific UVCB components?

A: At the time of writing, Smithers has not directly done so for an OECD 301 study. However, in cases where mineralization should be confirmed, an OECD 310 study with chemical analysis maybe more appropriate. Given our experience of measuring fugitive materials, often in complex matrices, we are confident that it is entirely possible to do so in many cases, and to provide quantitative information on the rates of such degradation should they be required. Further, we have very significant experience in the identification of degradants and metabolites, should such information be required.

Q: Will the proposed changes to REACH and requirement for polymer testing pose any challenges?

A: Yes. An EU study in 2020 indicated that more than 30,000 different polymer and oligomers may require registration. Current thinking is that polymer substances will be grouped together to reduce the testing burden.

A polymer is defined as a substance consisting of molecules characterized by the sequence of one or more types of monomer units. These molecules must be distributed over a range of molecular weights, with differences in molecular weight due primarily to differences in the numbers of monomer units. 

A polymer includes: A simple majority by weight of molecules containing at least three monomer units covalently bonded to at least one other monomer unit or to another reactive substance;
An amount less than a simple weight majority of molecules of the same molecular weight. For the purposes of this definition, "monomer unit" means the reacted form of a monomer substance in a polymer.

Essentially, guidance ECHA shared in 2012 indicated that 50% of the weight of the substance must be polymeric molecules. In addition, the amount of polymer molecules with the same molecular weight must be less than 50% by weight of the substance.

These offer analytical challenges, as large molecules are often not best suited to LC/MSn techniques, and approaches like size-exclusion chromatography (SEC) to assess molecular weight distribution, and not always common in laboratories performing small molecule analysis.

Q: Is it sufficient for surfactants/surface-active substances to prepare a "traditional" WAF at the highest nominal test concentration, e.g. 100 mg/L? Will it result in a solution at the CMC?

A: There isn’t a simple answer to this question.  Assuming standard conditions of temperature, there are mechanical issue, which apply irrespective of the properties of the test substance.  These include how the test substance is introduced into the solution, and the material of the container: a smooth container type with minimal active sites is preferable. The type of mixing is important, with laminar flows being naturally preferred as there are no cross-flow collisions. Regarding the test substance itself, matters like the ionic strength of the test substance and the solvent (including possibly the dielectric constant) and obviously the pH can profoundly impact the result.

Finally, there are a series of minor factors unique, to each substance and strongly related to the factors above, for water-based matrices: Any structural affinity for secondary bonding with the solvent, for example hydrogen bonding; Secondary bonding with other molecules, essentially the propensity with which the test substance will create micelles or liposomes; The power of the solvent to separate and keep the molecules apart.