H2OK?

Werner Brack is head of the Department of Effect-Directed Analysis at the Helmholtz Centre for Environmental Research (UFZ) in Leipzig, Germany. Brack’s department bridges chemicals and water by developing tools to identify (eco)toxicologically relevant chemicals in (mostly aquatic) environments including water, sediments and biota. Brack is also involved in the network on emerging pollutants (NORMAN), where he heads a working group on effect-directed analysis.

Annemieke Kolkman is teamleader of the Laboratory of Material Analysis and Chemical Analysis at the KWR Watercycle Research Institute in Nieuwegein, The Netherlands. The focus of Kolkman’s research is the development and implementation of new analytical techniques and methods for safeguarding chemical water quality with respect to human health.

Ron van der Oost is a toxicologist at Waternet Institute for the Urban Water Cycle, Amsterdam, Netherlands. Waternet is the first Dutch Water Cycle company and gives van der Oost the opportunity to perform applied research on new monitoring methods for water quality, thus bridging the gaps between science and practice. Together, the staff at Waternet investigates alternative ways to assess the chemical quality of drinking water, surface water and waste water.

Juliane Hollender is head of the Department of Environmental Chemistry at Eawag (the Swiss Federal Institute of Aquatic Science and Technology) in Dübendorf and an adjunct professor of environmental chemistry in the Department of Environmental Systems Science at the ETH Zürich, Switzerland. Hollender focuses on the occurrence and fate of organic micropollutants like pesticides, biocides, pharmaceuticals in natural and engineered aquatic environments.

What are the broad challenges in water management today?

Juliane Hollender: I can think of four main challenges. First, the direct and indirect reuse of water in densely populated areas, such as in central Europe (for example, the river Rhine) and in slums in developing countries. Second, water scarcity and the compounding effect of climate change. Third, the pollution of water by agriculture, industry (especially in countries where there is rapid industrialization), and households (through the increasing consumption of chemicals). And fourth, developing appropriate technologies for sanitation and drinking water treatment; this should focus both on end-of-pipe solutions and on alternatives, such as separation at source and green chemistry.

Ron van der Oost: As Julienne indicates, the main issues are managing water quantity (to prevent of flooding or drought), increasing water quality (to minimize environmental and human health risks) and the creation of a sustainable water cycle (that is carbon neutral, reuses water, reduces energy usage, and contributes to a circular economy). As a toxicologist I am, of course, most interested in the absolute chemical quality of water.

Werner Brack: One major challenge is to ensure that European surface waters are ecologically sound. The presence of certain chemicals in these waters is at odds with this goal. Multi-target monitoring of complex mixtures of chemicals in surface waters is hard to implement at a European scale, and the issue cannot be solved by a “priority pollutants” approach as chemical production and use is far too dynamic. We must therefore approach the goal of (eco)toxicologically safe water by integrating batteries of sensitive bioassays that cover a broad range of toxicological endpoints with chemical screening approaches. This should address ecosystems and human health at the same time, but to achieve it screening technology needs to be developed to cover a much bigger fraction of components.

The second major issue is mitigation of the problems that we identify. Water managers often do not have the capacity to act beyond end-of-pipe solutions in water treatment. A more holistic approach that includes land-use, production and use of chemicals, energy production and so on, is required.

The third, and most specific challenge, is that we must be vigilant about micro-pollutants, including pharmaceuticals and their transformation products, dissolved organic compounds (DOCs) in reservoirs, and disinfection by-products, particularly in countries with extensive chlorination of the drinking water supply.
Annemieke Kolkman: Julienne, Ron and Werner have covered many of the key issues there. Perhaps I should summarize in a single sentence: we must focus on the provision of enough safe drinking water for everybody and a constant drive towards a pollution-free environment.

How has water management developed over the past 20 years?

RO: There have been many developments in water cycle sustainability, often focusing on the use of wastewater as a resource. From an analytical point of view, a recent development is the realization that we should focus on the risks of the entire chemical mixture rather than the levels of individual substances.

WB: We’ve improved water quality by using the best available treatment technologies,  and by phasing out many persistent organic pollutants (POPs) and other problematic compounds. The EU’s Water Framework Directive (WFD) moves us to a more holistic approach that considers overall quality rather than the regulation of individual chemicals. To further develop WFD, we need to implement more diagnostic monitoring tools to identify causation, which is essential for effective and efficient management.

JH: Positive developments include improved wastewater treatment with nutrient (nitrogen and phosphorous) elimination, industrial and household water saving, and, of course, WFD implementation. Both the Stockholm Convention on Persistent Organic Pollutants, in effect since 2004, and the WFD’s priority compound list have helped to reduce contamination of the environment by very persistent compounds.

How would you define the role of analytical science within water management?

Annemieke Kolkman : Analytical science is used to determine chemical water quality, both for legal monitoring parameters and for research purposes, for example, to investigate new challenges facing the water sector, such as emerging compounds.

JH: Providing data and thereby supporting decisions. Also, as Annemieke indicated, raising risk awareness of new pollutants and thereby triggering mitigation measures.

RO: Agreed. Put simply, the role of analytical science within water management is to assess the potential risks of chemical and microbial pollution of the aquatic environment (surface and waste water) and human health (drinking water).

WB: Analytical science has done an excellent job in sensitively detecting and reliably quantifying most target chemicals in trace concentrations; however, we must not rest on our laurels. We are still far from understanding the chemical space in our waters –the ultimate goal. As noted above, in vitro bioanalytical tools will become increasingly important for detecting chemicals in sum and grouped according to effect patterns.

What trends have hit water analysis over the last two decades?

WB: The availability of liquid chromatography-mass spectrometry (LC-MS) methods shifted the focus of chemical analysis to polar water contaminants from metals and non-polar POPs. Many of these so-called emerging pollutants are included in today’s monitoring programs. More recently, transformation products and metabolites have shifted the focus again. In this field, non-target approaches have been quite successful, supported by information on parent compounds. Presently, analytical chemists are trying to achieve real non-targeted analysis of the numerous unknowns in water samples, which is still very challenging and not always successful. The development of new computer tools, spectra libraries and databases will doubtless enhance the success rate.

AK: As Werner noted, there has been a clear shift towards the analysis of more polar compounds, with a consequent shift from gas chromatography to LC analysis. Moreover, the sensitivity of instrumentation has increased tremendously, lowering detection limits. In my opinion, one of the major breakthroughs is the introduction of high-resolution MS (HR-MS) (Orbitrap, QToF) analysis in the environmental field. Analyses are no longer limited by a predetermined set of chemicals (targeted analysis), but are able to provide a much broader view of water quality. Furthermore, it is possible to retrospectively search MS data.

RO: Chemical analyses are more sensitive, so lower levels of pollutants can be detected. The complex mixtures of enormous numbers of (emerging) substances that are now observed in the water cycle means that it is virtually impossible to predict risks and effects. Therefore, in recent years increasing focus has been directed towards the monitoring of mixture effects with bioanalytical toxicity tests (bioassays). In terms of microbial analyses, molecular biological methods have been introduced to enable speciation of bacteria and viruses more rapidly.

JH: From a separation/detection view point, there has been a broad shift from non-polar organic compound analysis using GC coupled with flame ionization detection, electron capture detection or MS and HPLC coupled with UV or fluorescence detection to the analysis of polar compounds using HPLC-MS/MS after electrospray ionization (ESI) and atmospheric chemical ionization.

What have been the major milestones in analysis development ?

AK: HRMS is perhaps the biggest milestone. And, as Werner and Ron said, there is a move away from just looking at which chemicals are present towards questioning the effect of the total mixture of chemicals present in a sample by using effect-directed tools, in other words, in vitro bioassays.

JH: The introduction of LC-MS enabled analysis of polar compounds without derivatization, while the use of hybrid instruments  means that we can elucidate structure without reference standards. Besides targeted screening over the last 10 years, the coupling of LC to HRMS has enabled screening of compounds that we expect to see in the environment, as well as non-targeted screening of compounds that we were previously unaware of, such as transformation products formed in the environment or during technical treatment.
Inductively coupled plasma (ICP)-MS has enabled simultaneous analysis of many metals – and coupled to chromatography it can also determine speciation.

WB: To provide some historical perspective, in the late 1960s we saw successful coupling of GC and MS, which led to a major focus on its use for volatile and semi-volatile chemicals until the 1990s. As Julienne notes, in the late 1980s, new ionization techniques, such as ESI, were linked to LC, triggering a shift to polar and emerging pollutants.

RO: In addition to the above, molecular methods and ‘omics’ technologies have become much more affordable in recent years, which is a big milestone in my opinion.

How can we meet the big challenges in water analysis?

JH: Quantification of the enormous number of possible pollutants is a problem. About 30,000 compounds that are in daily use in households may end up in the aquatic environment. The use of multi-target analysis, combination of exposure modelling and analysis, suspect screening of expected classes and non-target screening using HRMS will resolve a large proportion of these unknowns. However, the coupling of chemical analysis to effect analysis is another necessary focus. Faster and more efficient effect-directed analysis would help us to focus on the most toxic compounds for structure elucidation.

AK: In truth, a big challenge is developing cost-effective tools. Here, the implementation of bioassays in combination with chemical analysis looks like a very promising direction.

RO: Further increases in the sensitivity chemical analyses are not very relevant. Rather, as indicated above, the focus should be on broad-spectrum screening methods. For toxicological testing, we must develop high-throughput methods to increase the speed and reduce the cost of toxicity testing of complex mixtures. A good example that combines both these areas is the ‘High-Throughput Effect-Directed Analysis’ project at the Free University of Amsterdam (tas.txp.to/0314/water1), in which HPLC separation is followed by both MS detection of chemicals and micro-scale bioassay measurement of effects.

WB: We must unravel our chemosphere – something that may be comparable with unravelling the human and other genomes. It requires intelligent, integrated and automated approaches that combine cutting-edge analytical tools with new developments in chemoinformatics and powerful software tools for fragmentation, ionization efficiency, chromatographic retention… and prediction.
Direct and automated combinations of chromatography, high-throughput in vitro assays and MS or other spectrometric tools may help to unravel toxic mixtures. 
Another important issue is the need to move away from measuring the external burden in water and sediments to measuring the internal burden (“exposome” is the new buzzword). Unravelling the complex mixture of xenobiotic metabolites and endogenous chemicals that result from all kind of stress factors will be an enormous challenge for analytical chemistry.

Where must we drive analytical capability?

WB: Personally, I feel a lot of progress is needed on the software side rather than in analytical hardware, along with integration of bioanalytical tools. A big, but maybe unrealistic, step forward would be an enhancement in the sensitivity and throughput of other spectrometric tools, such as NMR, so that they could be routinely applied in environmental analysis, complementing MS methods.

JH: We need further improvement of the mass resolution, mass accuracy and isotope abundance to enable that quantification of isotope fine structure, for example,  to resolve 15N, 34S, 18O. This would improve assignment of molecular formulae to unknown peaks, subsequent assignment of molecular formulae to MS/MS fragments and, finally, structure elucidation.
Coupling of ion chromatography with MS would be useful for identifying and quantifying small ionic compounds, like disinfection-by-products.
And, as Werner stresses, we should focus on coupling of chemical analysis with biological tests to enable identification of toxic compounds.

AK: Better tools for the identification of unknown contaminants are essential, but their development is challenging and successes have been few and far between so far. I also agree with Werner that the development and maturation of software tools to extract relevant data from broad screening (HRMS) is important; data analysis is a bottleneck and often the most time-consuming part of the whole analysis. A general trend outlined in this article is the need to develop bioassays into a more mature tool – which is to say, high throughput – so that they can be cost-effective and implemented in regular monitoring. Moreover, we need to understand the link between bioassay status and
human health.

Which groups of pollutants need the most attention?

RO: Organic micropollutants need most attention. The water cycle can be polluted with more than 100,000 substances and it is virtually impossible to chemically analyse all of them. In addition, nothing is known about mixture effects and the effects of metabolites. Therefore, a paradigm shift in risk assessment is needed, from measuring levels of a limited number of substances – current monitoring – to the effects of the entire mixture – future monitoring. Currently, the latter is only applied in scientific research, but it should be implemented in regular monitoring. However, we cannot develop scientifically sound threshold levels for bioanalyses because we need to know which substances cause the bioassay effect in order to predict whole organism and ecosystem effects. Therefore, we need to develop trigger values to indicate potential risks and the need for further chemical and/or toxicological research.

AK: Transformation products and metabolites demand further research. Organic contaminants can undergo bio-transformation once they are in the environment; for example, a pharmaceutical rarely exits the human body in its original form but rather as a more polar metabolite that enters the environment through the sewage system. Such transformation products should get more attention: are they present and, if so, at what levels and risk?

JH: In general, I think water analysis should be as comprehensive as possible and not focus too much on a single field. However, emerging pollutants that are not yet included in any European regulation need to receive more attention in governmental and regional laboratories. Increased suspect and non-targeted analysis with HRMS would help here. Spectral libraries for LC-MS analysis, like the open-access library MassBank (tas.txp.to/0314/massbank), need to be filled with HRMS data to enable the exchange of mass spectra and help elucidation of unknowns without reference standards.

WB: Actually, I think a lot of attention is already given to the analysis of emerging pollutants, such as pharmaceuticals, illicit drugs, personal care products, and biocides. The awareness of these chemicals as possible hazardous contaminants is increasing and appropriate LC-MS/MS techniques are becoming available. I also feel that the analysis and identification of metabolites and transformation products is sufficiently on the radar.
However, I think that screening and identification of unknowns using LC-MS/MS needs to become a real focus. The reason is that targeted analysis in water samples typically addresses only a very tiny portion of the peaks present. We have no idea about the vast majority – neither their structures nor their possible effects or risks. The screening of these unknowns is not getting the attention it needs, due in part to a lack of computer tools, good software for structure elucidation, prediction tools as well as spectral libraries. For this to happen, a joint effort by analytical instrument producers, computational chemists and the analytical chemists that apply the tools is crucial.

What’s your take on the marriage between bioanalytical tools and chemical analysis?

RO: This marriage is the future of water quality assessment, period. An initial screening with bioassays should provide a first impression of micro-chemical water quality, involving the impact of all potential (known and unknown) toxic chemicals, mixture effects and effects of break-down products. If no indications for significant effects are found, then it would be a waste of time and money to perform advanced chemical analyses. If significant effects are found, in other words, if the trigger values are exceeded, an effect-directed analysis (EDA) should be performed to identify the most relevant toxic substances that may cause adverse effects on environmental and human health. The obstacles to this are the lack of scientifically-accepted trigger values for bioassay responses and the high costs of EDA research. To address these obstacles, we have set up the ‘smart monitoring’ project, which is an alternative monitoring strategy to the WFD assessment of the chemical status of water bodies. We have designed a bioassay battery that covers the potential impact of a broad group of chemical pollutants at lower cost than a chemical analysis of the WFD priority substances. Trigger values for all bioassay responses will be proposed, to distinguish levels of risks caused by chemical pollution. The bioassay battery responses will be incorporated into a ‘toxicity traffic light’ model that clearly informs regulators on low (green), potential (orange) or high (red) micro-chemical risks for the ecosystem. For ‘orange’ sites, more efficient and cost-effective EDA methods should be developed to differentiate real risks (red) and artefacts (green).

AK: Certainly, such a marriage is a very promising and cost-effective combination of monitoring tools. However, in vitro bioassays need to mature and be performed in an automated high-throughput fashion. In addition, innovative chemical tools (fractionation, computational tools for structure elucidation, separation) are also needed, and must be integrated with the bioassays. We are involved in the EDA Emerge Network (tas.txp.to/0314/water2), which focuses on producing young scientists with the interdisciplinary skills required to meet the major challenges in the monitoring, assessment and management of toxic pollution in EU
river basins.

JH: Unfortunately, vendors have not yet developed instruments that combine the chemical and biological worlds, in terms of new detectors for existing systems. Developments in this area have come predominantly from academic research groups and have not yet been implemented in practice; an example is the coupling of MS with an acetylcholine esterase assay or luminescence test. The dilemma here is that many such detectors are only able to target one specific mode of action despite the fact there are many potential effects.

WB: I agree with Annemieke and Ron: the marriage of bioanalytical tools and chemical analysis is of great importance. Assuming that we will be unable to analyze and assess complex mixtures in water samples in the near future, we need, at the very least, good filters to focus our analysis on relevant chemicals rather than a fixed set of chemical targets. These filters will depend on the objective of the study but typically will focus on a biological effect, from receptor-binding to manifested toxicity.
We need to develop automated and miniaturized high-throughput approaches that combine high performance chromatography with multi-well-format bioassays and MS analysis. This requires an interdisciplinary approach that combines skills in analytical chemistry, in vitro bioassays, engineering and automation, which are often hard to bring together. The inclusion of more affinity-based separation approaches using toxicologically-relevant receptors would also be helpful. However, the commercial availability of such receptor-based chromatographic columns or solid phases for extraction is extremely limited at present.
The field has been combining bioanalytical tools and chemical analysis for a long time; a lot of effort has been devoted to individual tools and to the concept as a whole, which is being promoted through networks and collaborative projects. The focus moving forward must be on throughput and automation; the development of  biodiagnostic endpoints; miniaturization; separation techniques, and efficient compound identification and structure elucidation.