A Collaborative Approach to Water Analysis

February 8, 2016
Alasdair Matheson

The Column

The Column, The Column-02-08-2016, Volume 12, Issue 2
Pages: 2–6

Contaminants in surface water and drinking water supplies arising from pharmaceutical and personal care product use as well as other compound sources pose a difficult challenge for analytical chemists. Thomas Letzel from the Technical University of Munich in Germany, spoke to The Column to discuss collaborative research taking place in Europe to address contaminants of emerging concern in water analysis.

Contaminants in surface water and drinking water supplies arising from pharmaceutical and personal care product use as well as other compound sources pose a difficult challenge for analytical chemists. Thomas Letzel from the Technical University of Munich in Germany, spoke to The Column to discuss collaborative research taking place in Europe to address contaminants of emerging concern in water analysis.

Q. In your view, what are the biggest challenges faced by analysts in the field of water analysis?

A: Currently, the main challenge in water analysis is to establish comprehensive analytical strategies and workflows for complex questions. Therefore, the analytical community has to address the demands of analyzing various water sources in a variety of political and regulatory frameworks. This leads to various hurdles for environmental analysts. On one hand, there are various bodies of water containing different types and concentrations of contaminants of emerging concern (CECs). For example, wastewater sewage treatment plant (WSTP) influent, WSTP effluent, surface waters, and drinking waters are totally different aqueous matrices, even though they have the same basis. Furthermore, internationally the public consumption of chemical compounds differs in use, number, and amount. On the other hand, different political bodies have different approaches to determining organic molecules in water. The precautionary system and risk-based systems are two examples of different approaches to addressing CECs. Our department chair, Prof. Joerg Drewes, is involved in research project No. 44941 financed by the Water Research Foundation (based in Denver, Colorado, USA) studying (and currently publishing the results of) CEC management strategies in Australia, the European Union, Germany, Switzerland, and the United States.

Although political views on this topic are very diverse, the analytical strategies tend to have the same focus. Generally, analytical chemists are looking for molecular identification of CECs and their quantification. One can detect many compounds in water (from both biogenic and anthropogenic sources), sometimes up to several thousand, with strategies using reversed-phase liquid chromatography–mass spectrometry (LC–MS) and gas chromatography–mass spectrometry (GC–MS). Typically LC–MS analysts use reversed-phase high performance liquid chromatography (HPLC), electrospray ionization (ESI), and accurate mass spectrometry (mostly tandem mass spectrometry). However, there are many compounds that one cannot observe using those techniques. Currently, extending the polarity range of molecule separation is a fascinating research field. In this research, the focus is on hydrophilic interaction chromatography (HILIC), ion-exchange chromatography, capillary electrophoresis (CE), and supercritical fluid chromatography (SFC). Both HILIC and ion-exchange chromatography can easily be coupled with reversed-phase LC, using a 2D or on-line separation. Polarity range extended LC using these combinations has clearly shown that a further hundred to a thousand very polar molecules can be monitored. Also, the renaissance of ionization techniques like atmospheric pressure chemical ionization (APCI) and atmospheric pressure photoionization (APPI) as well as recent developments like ion mobility make it possible to obtain more information about the compounds in a sample. Furthermore, after using atmospheric pressure ionization techniques in GC–MS, one can search for molecular ions in chemical databases like Chemspider and STOFF-IDENT. With current ionization techniques like electron ionization, only fragmentation patterns can be compared with analytical databases.

In addition to following the precautionary strategy of identifying as many compounds as possible in a water sample by name, another focus should be learning about the properties of contaminants as they relate to human toxicology or ecotoxicology. This approach requires interdisciplinary collaboration among chemists, biologists, and toxicologists. Also, bioinformatics has to be involved, because computational modelling has become increasingly relevant in these studies (such as in using chemical structures to make chromatographic or toxicological property predictions).

Q. There are three main approaches commonly applied in water analysis - non-target screening, suspected-target screening, and target screening. What are the key differences between each one? Are particular techniques used in each approach?

A: Target screening is applied in water analysis the same as it is in fields such as food analysis and forensic analysis. In target screening, LC–ESI-MS–MS is used for the quantification of known molecules with isotopically labelled reference substances. Generally, quantification is performed using multiple reaction monitoring (MRM) with specific quantifier and qualifier fragments in tandem MS to perform quantitative analysis. This approach is typically applied for compounds both of general interest and of special interest (such as molecules on the “Watch List” under the EU Water Framework Directive or molecules listed in risk-based systems).

Suspected-target screening takes advantage of knowledge about the human usage of compounds and their degradation by metabolism or transformation during technological or chemical treatment. This technique is not yet clearly defined. Some laboratories include their measured molecules without isotopically labelled standards (thus without quantitative information). This process and its analytical systems are similar to those used in the target screening technique. Other laboratories include expected molecules in their analysis; examples are metabolic change prediction approaches such as the EAWAG (the Swiss Federal Institute of Aquatic Science and Technology) pathway prediction system and oxidative processes such as ozonation or OH radical treatment. Compound databases such as Chemicalize, Chemspider, and STOFF-IDENT can help in this approach.

That leads us directly to non-target screening, especially the subclass of hidden targets2 (formerly referred to as “known unknowns”3). For non-target screening, analytical laboratories typically use an accurate mass approach, mostly including accurate tandem MS fragments. The observed data can be used more efficiently in association with MS databases such as MassBank, with the MS prediction software MetFrag or MSforID, and with the vendor software.

The “real” non-target screening - that is, for compounds referred to as “unknown unknowns” - uses more rudimentary analytical information such as chromatographic and MS behaviour in comparison with known molecules. This type of water analysis is very time consuming and expensive, and therefore the approach is justified only for unknown compounds that have a toxicological relevance or another important function.


Q. With recent advances in analytical technologies, analysts are discovering unexpected molecules in addition to known and expected molecules. The terms “known unknowns” and “unknown unknowns” are increasingly used to describe such molecules. Can you explain the difference between the two?

A: The class of “known unknowns” has recently been described very nicely by J.L. Little and colleagues3 as comprising species that are known in the chemical literature or MS reference databases, but are unknown to the investigator. This definition conceals a clear dilemma of non-target screening: “Not targeted” does not always mean “unknown”. Non-target screening thus often uses the same path as suspected-target screening, which implements the identification of expected compounds using analytical and chemical databases. The future will show which definition covers the most applications; maybe there will be different definitions for different applications. A current definition for the identification of transformation products is given in our recently published water study.

Q. Your recently published study described the use of a number of LC–MS methods from three collaborating laboratories to detect trace levels of anti-hypertensive drugs in water samples.2 What led you to begin this study? Why did you choose to focus on anti-hypertension drugs from the sartan group?

A: In our opinion, the most powerful analysis is obtained if the information from various laboratories is combined and used for individual data evaluation. High-end analyses are very efficient, but not every
lab is capable of financing and handling that approach. Many laboratories are equipped with triple-quadrupole systems or single TOF-MS systems, which are significantly cheaper. These laboratories depend on interaction with other laboratories or the use of resources such as databases of potential anthropogenic molecules in the aqueous environment (such as STOFF- IDENT) or complementary data such as interlaboratory normalized retention times.

The sartans are a very interesting and varied group of compounds. They are widely used (for example, valsartan was on the list of 200 top selling pharmaceuticals in 2011). They look chemically similar, but they have different analytical properties and vary in their sewage treatment stability.4 Although sartans have no expected ecotoxicological relevance,4 they became one of our prioritized groups for demonstrating our holistic analytical strategies and risk assessment processes.

Q. Which screening strategy did you use? What factors did you take into consideration?

A: The sartan degradation products from a lab-scale sewage treatment plant (at the Bavarian Environment Agency) were identified by our colleagues in Langenau (Zweckverband Landeswasserversorgung) using non-target screening with an LC–hybrid quadrupole TOF-MS system, with a validation step performed using an LC–TOF-MS system in our lab. The identified molecules were then added to the environmental monitoring set of the Bavarian Environment Agency for target screening. The agency was able to synthesize one of the new identified transformation compounds, and it is now part of the agency’s quantitative monitoring list.

The study impressively reflects how effectively complicated analytical questions can be answered if laboratories with different analytical screening strategies can work together.

Q. What were your main findings in the overall RISK-IDENT project?

A: In addition to the results obtained by the research, we faced the challenge of bringing together chemists, biologists, engineers, programmers, ecotoxicologists, and public relations people. We found that workflows and strategies in risk assessment require deep interactions and clear communication inside and outside of the little box of research. Furthermore, the national consortium based on the funding measure “Risk Management of Emerging Compounds and Pathogens in the Water Cycle (RiSKWa)” (which includes
12 initiatives like our RISK-IDENT project) joined in interdisciplinary meetings. We were able to obtain sustainable results because a lot of experts put their experience together.

Another main topic was and still is the communication with the public. All partners tried to translate their findings and results continuously into understandable reports. These efforts were not always successful, but they moved the public discussion and understanding forward and provided useful suggestions such as “Do not wash pharmaceuticals into the sewer!”


Q. Did you face any difficulties or challenges? How did you overcome them?

A: It was not always easy to understand each other due to the different “languages” in the different disciplines. However, we mostly overcame this inconvenience by visiting the other partners and participating in their research.

As I noted in my answer to the previous question, it was not always easy for the public to understand us. We learned by ongoing interactions and by the help of professional institutions. In the end, we found that the public understanding of our research led to the acceptance of our conclusions and spurred the public to reevaluate environmental issues.

Q. You are the coordinator of the continuing project FOR-IDENT funded by the German Federal Ministry of Education and Research (BMBF). Can you tell us about the aims of this project and what you have achieved so far?

A: Our goal is to discover worldwide analytical strategies for studying CECs in water. This process will lead to combined knowledge and harmonized analysis. As part of the process, a so-called FOR-IDENT platform will be established in an open-access format that will include software tools to establish efficient analytical workflows. We have taken the first steps towards our goal to provide free data from an open platform. The international extension was started by
being part of the European Norman Network and by organizing international workshops (for example, at the 2015 ACS Fall Meeting in Boston), including and combining vendor ideas and software. Finally, we plan to edit a book titled Towards Harmonized Strategies and Workflows to Assess Transformation Products of Chemicals of Emerging Concern by Non-Target and Suspect Screening.

Q. What are the implications of your work for the future?

A: At this point, our interdisciplinary teams are working intensively together, and they will become more integrated in the future. Our analytical strategies and workflows for water analysis will find their counterpart in disciplines like metabolomics, forensic, process, food, and other analysis fields. On the long term, we expect that the analytical strategies can exceed the limits of single disciplines and they will end up in a holistic view.


  1. T. Rauch-Williams, J.E. Drewes, E. Dickenson, A. Fulmer, S. Snyder, S. Bieber, S. Deslauriers, and S. Dagnino, “Evaluation of Strategies to Manage Trace Organic Compounds in Water (WRF #4494)”, Water Research Foundation web seminar, 24 June 2015, available at http://www.waterrf.org/resources/webcasts/Lists/PublicWebcasts/Attachments/48/6-24-15-Final%20Presentation.pdf
  2. T. Letzel et al., Chemosphere137, 198–206 (2015).
  3. J.L. Little et al., LCGC Europe26(3), 163–168 (2013).
  4. A. Beyer et al., Environ. Sci. Pollut. Res.21(18), 10830–10839 (2014).

Prof. Thomas Letzel is an analytical chemist with almost 20 years of professional experience in the field of analytical screening techniques using liquid chromatography with mass spectrometric detection. Prof. Letzel is head of the Analytical Research Group at the Chair of Urban Water Systems Engineering at the Technical University of Munich (TUM). Currently, the key aspects in research cover technological, analytical-methological, and analytical-chemical properties and can be applied in water and wastewater analysis as well as in other relevant environmental matrices, food analysis, beverage and plant extract analysis, and others. A special focus is on the chemical analysis with simultaneous functionality analysis using mass spectrometric detection. He is author and co-author of about 100 publications and two books. He has experience with many national and international research projects and he actively participates in (inter)national environmental initiatives like RiskWa-BMBF, NORMAN Association and ESSEM COST Action ES1307. 

E-mail: t.letzel@tum.de