Frank Michel1, Yong Chen2, Robert Shirey2
1Sigma-Aldrich (part of Merck KGaA, Darmstadt, Germany), Taufkirchen, Germany, 2MilliporeSigma, Bellefonte PA, USA
The analysis of water for volatile organic compounds is important due to their toxicity. The current methods for this determination lack of sensitivity, selectivity or capability for automation. This paper presents the new ISO 17943 Standard using Solid Phase Microextraction (SPME) and GC/MS. The sample preparation by SPME enables low limits of detection and easy automation of the whole method. GC/MS provides the required sensitivity and selectivity. This ISO Standard was validated by an interlaboratory trial, which results confirm the outstanding performance for this method.
Volatile Organic Compounds (VOCs) can occur from natural sources such as plant scents. However, a large amount of VOCs do have an anthropogenic origin, because they are released from products in daily use or emitted during the manufacturing of such products, as well as from polymers, adhesives, paints, petroleum products or pharmaceuticals. Typical applications for VOCs are use as additives for gasoline or as solvents and hydraulic fluids or for dry-cleaning. As many VOCs are toxic or are known or suspected human carcinogens, contamination of water resources is a serious human health concern worldwide.
Figure 1.SPME fiber holder with fiber immersed into aqueous sample
Because of this, many international regulations have been established to limit and control the amount of VOCs in drinking water, groundwater or surface water. Examples of such regulations are the Safe Drinking Water Act (SDWA)1 in the USA, and a corresponding law in Canada that established national standards for drinking water including VOC listings that are based on health considerations. Another example is the European Council Directive 98/83/EC on the quality of water intended for human consumption that regulates the values for individual volatile organic substances.2 In the EU Water Framework Directive (WFD) in article 16 of the Directive 200/60/EC3 a “strategy against pollution of water” is described.
According to Directive 2008/105/EC (EQS Directive)4 Environmental Quality Standards (EQS) values for single VOCs should be in the range of 0.4 to 20 μg/L. In annex V of WFD (standards for monitoring of quality elements) the use of ISO and CEN standards for the analysis of water is required, if available.
The existing ISO and CEN standards for the determination of VOCs in water are not state-of- the-art methods anymore. ISO 103015 uses Liquid/ Liquid Extraction (LLE) in combination with Gas Chromatography (GC) and detection using Flame Ionization Detection (FID) or Electron Capture Detection (ECD). ISO 114236 employs headspace (HS) sampling in combination with GC/FID or GC/ECD. For certain relevant VOCs, the required limits of detection cannot be achieved using these ISO standards because the detectors are not sensitive or selective enough.
ISO 156807 exhibits an alternative by using purge- and-trap enrichment and Gas Chromatography-Mass Spectrometry (GC-MS) analysis leading to better selectivity and limits of detection. The downside of purge-and-trap is the susceptibility of the trap to become contaminated and that automation is rather challenging to achieve.8
Solid Phase Microextraction (SPME) in combination with GC-MS is an attractive alternative for the determination of VOCs in water. SPME was developed by Janusz Pawliszyn in 19909 (Figure 1). Since then SPME has gained broader acceptance in environmental, pharmaceutical and food analysis as demonstrated by the growing number of publications on SPME developments and applications. The prevalence of this technique was additionally increased by the automation of SPME using regular GC autosamplers beginning in 1993. The use of SPME for the extraction of VOCs from water is described in several publications.10-12 In these publications, headspace SPME (HS-SPME) was proven to be a reliable and beneficial alternative to classical methods for VOC determination in water. Furthermore, SPME has been successfully used in many other official methods.13-15
Due to this, the new ISO standard 17943 was developed for VOCs in water. The scope of the standard is the determination of more than 60 VOCs from very different classes such as halogenated hydrocarbons, gasoline additives (like BTEX, MTBE and ETBE), volatile aromatic compounds and highly odorous substances like geosmin and 2-methylisoborneol in drinking water, groundwater, surface water and treated wastewater by HS-SPME and GC-MS. Of course the limit of detection depends on the matrix, on the specific compound and on the applied mass spectrometer, but for most compounds in ISO 17943, it is equal to or better than 0.01 μg/L. Additional validation data derived from standardization work show applicability of the method within a concentration range from 0.01 μg/L to 100 μg/L for individual substances.
As part of the development of this new ISO standard, an international interlaboratory trial was conducted to validate the new method.16 Each of the labs had to determine the concentration of 61 compounds in the two water samples (one surface water, one wastewater). The surface water sample was taken from an urban and industrialized area (the Ruhr River in Muelheim, Germany). The municipal wastewater sample was taken from a plant effluent. Both samples had been pre-treated to stabilize them and had been spiked with concentrations unknown to the participating labs in the range of 0.02 – 0.80 μg/L (~ 50 % < 0.10 g/L) for the surface water and 0.05 – 3.0 μg/L (~ 50 % < 0.50 g/L) for the wastewater. The labs in the interlaboratory trial had to conduct four independent replicate analyses from each of the two samples, strictly following the procedure as prescribed in the draft standard method. All laboratories were provided with a set of calibration solutions placed in three ampoules each containing certified reference substances of the 61 VOCs dissolved in methanol. These stock solutions contained the individual substances in concentrations of 100 µg/mL each and were intended to be used for preparation of the corresponding aqueous multi-component reference solutions used for calibrating the total procedure. The results had to be delivered within 30 days after receipt of the samples.
The Supelco® Application Lab was one participant in the interlaboratory trial. The two water samples were analyzed according to the drafted ISO Standard 17943 (Table 1 & 2, Figure 2) using toluene-d8, benzene-d6 and fluorobenzene as internal standards. For the GC analysis a VOCOL® capillary GC column was used, which is an intermediate polarity column that is designed for analysis of VOCs and provides great retention and resolution of highly volatile compounds. For HS-SPME a DVB/CAR/PDMS fiber was used which was also used by the majority of the interlaboratory trial participants. A smaller share of the labs used a CAR/PDMS fiber.
According to ISO Standard 17943 both the Carboxen/PDMS (85 μm) and the DVB/Carboxen/PDMS (50/30 μm) fiber can be used.
Figure 2.Chromatogram of 61 VOCs in water after HS-SPME using a VOCOL® GC column on Agilent® GC/MS
More than 40 labs from all over the world registered for this interlaboratory trial. Out of these a total of 27 labs reported results to be included in the evaluation process according ISO 5725-2.17 Nine laboratories did not submit any results. Six labs had to be excluded from the valuation due to significant deviation from the prescribed procedure. Some single results had to be excluded due to outliers.
All 61 parameters had been analyzed by ten labs and nearly all parameters had been analyzed by nine labs. Expressed in a different way, this resulted in the fact that nearly each of the 61 VOCs had been analyzed by more than 20 labs, which provides a valid base for statistical evaluation. The data was analyzed for the overall mean of results (without outliers), the recovery rate (from assigned value), the reproducibility (variation between different labs) and the repeatability (variation within a lab).
One example of such an evaluation is shown in Figure 3 for 2-chlorotoluene. For this compound, results from 24 labs could be evaluated. The overall mean value (green line) is very close to the assigned value (purple line). The majority of the 24 labs, even those labs that were new to SPME, achieved results very close to the assigned value. The recovery rate for more than 90% of the compounds was between 84 and 116 % (surface water) and 81 and 118 % (wastewater). The reproducibility (variation between laboratories), for more than 90% of the compounds, was less than 31% (surface water) and less than 35% (wastewater), while the repeatability (variation within a lab) for more than 90% of the compounds was less than 10% (surface water) and less than 8% (wastewater).
Figure 3.Graphical presentation of the example of 2-chlorotoluene which shows the results of the interlaboratory trial for the validation of ISO 17943. The purple horizontal line is the assigned value; the green horizontal line is overall mean.
The outstanding results in the interlaboratory trial underscore the high performance, reliability and reproducibility of HS-SPME in combination with GC/MS for the determination of VOCs in water. The new ISO 17943 is an improvement on existing official methods for this determination in terms of sensitivity and selectivity. In addition, the capability for full automation of SPME is beneficial for running this analysis 24/7.
For more information on our complete environmental testing workflow solutions, please visit: SigmaAldrich.com/environmental-testing