|
Our rivers and oceans are contaminated with a wide range of chemicals, despite the best efforts of the municipal wastewater treatment plants and industrial sites. Trace amounts slip through the net on a regular basis, being discharged into receiving waters such as rivers and coastal water where they accumulate over time. All sorts of chemicals have been detected, complicating the life of the analyst whose job it is to find out what is out there and how much. For many of the "newer" pollutants originating from consumer products such as detergents, cosmetics and pharmaceuticals, their environmental fate and eco impacts are unclear, if not entirely unknown. Are they simply passive, drifting intact or degrading to harmless breakdown products, or are they toxic or acting as endocrine disruptors to aquatic life and beyond? The detection and measurement of these pollutants is complicated by their dilution in large bodies of water and the presence of other compounds naturally present, such as humic matter leached from soils. Special steps must be taken by the analyst to concentrate the compounds and remove any influences from other compounds. Although LC/MS methods appear to dominate the recent scientific literature, GC/MS remains one of the most popular techniques in the environmental lab. When linked with large volume injectors, they can cope with the need to process the increased volumes needed to transfer larger amounts of analytes to the GC column. This approach has been examined by a team of researchers who have optimised the extraction and analysis of a broad set of pollutants. Carlos Guitart and James Readman from the Plymouth Marine Laboratory, UK and the Institute of Marine Sciences, Barcelona, were seeking to measure pharmaceuticals, chemicals from personal care products, endocrine disrupting phenols and faecal steroids in one analysis. The drugs included diclofenac, ibuprofen and naproxen. The antibacterial compound triclosan was added because it is present in many products such as toothpastes, soaps, deodorants, cosmetics and the like. The four industrial phenols included bisphenol A and 2-phenylphenol. Five faecal steroids, including coprostanone and coprostanol, were analysed because they are useful for tracking sewage effluents. Guitart and Readman began by optimising the extraction step. Based on previous work and literature reports, they used an SPE HLB cartridge. The best eluate was found to be a mixture of dichloromethane, ethyl acetate and methanol, which was suitable for the wide spread of compound polarities. The extracts were reacted with silylating agents to prepare the tert-butyldimethylsilyl derivatives for GC/MS analysis with a programmed temperature-vaporising (PTV) inlet. After injection, the solvent was vented through a valve while the analytes became concentrated on the PTV liner. Subsequent ramping of the PTV temperature drove the analytes to the GC column, with a 5% phenyl dimethylsiloxane coating, for separation. The syringe injector speed, initial inlet temperature, solvent vent time, ramp rate and column temperature were all optimised to ensure minimal losses of analytes from the PTV unit during solvent venting and maximum transfer to the GC column. This afforded good recoveries and reproducibilities for all compounds. The electron ionisation mass spectra were recorded for identification of the silyl derivatives and the target analytes were measured by selected ion monitoring. The detection limits were almost all in the low ng/L range. Although the PTV inlet copes with large volumes, the improvement over conventional GC/MS was less than expected because the enhancement in signal-to-noise ratios did not vary linearly. For instance, s/n increased from 64 to 328 for coprostanol for a 50-fold increase in injection volume. Nevertheless, method performance was improved. In real samples of water, there is often so-called humic matter originating from the breakdown of plant and animal material and this is known to interfere in mass spectrometric analysis. In this study, the researchers found that humic acids reduced the recoveries of phenols and steroids but had no effect on the drugs. The removal of humic acids by eight different sorbents was tested, with a diol material being preferred. It trapped the humic compounds and eluted the target analytes within the first fraction. The optimised method was applied to sewage treatment plant effluent, river water near a sewage treatment plant, estuarine and coastal water. The sewage effluents had high levels of drugs, such as 2203 and 856 ng/L for ibuprofen and naproxen, respectively. The industrial phenols and faecal sterols were also found. Levels were lower downstream of a treatment plant, whereas the river and estuarine samples contained only trace amounts, presumably due to dilution effects.
The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.
|
The antibacterial triclosan, one of the target compounds, is found in many personal care products, including toothpaste |
