The production of Certified Reference Material for PFAS in fish and water
Within-laboratory validation characteristics such as precision, accuracy, and robustness need to be determined. One important tool for evaluating accuracy is the use of certified reference materials (CRMs), but the availability of CRMs for many organohalogen micropollutants is limited. Some CRMs are available for classical lipophilic compounds like polychlorinated biphenyls, organochlorine pesticides and dioxins in food items e.g. milk powder, fish and fish oil. Recent feasibility studies evaluated the production and certification of CRMs for brominated flame retardants in fish (EU-BROC) and sterilized milk, pork, fish, fish oil and feed (EU-DIFFERENCE). These studies showed that it is possible to successfully produce matrix-type CRMs while keeping the physical state of the sample matrix intact. This is very important as the CRM should closely approximate real sample matrices. In line with these past experiences, and the lack of CRMs for perfluoroalkylated substances (PFAS), the PERFOOD consortium decided to create two CRMs. The two matrices selected were water and fish which is in line with the aim of PERFOOD (i.e. food related exposure), and as well are two sources of exposure to humans. The responsible institute for the production of the CRM was IVM. The production of the material was performed in late 2011, and it is currently under validation, i.e. stability and homogeneity testing and certification of the concentrations of the selected PFAS.
In December 2010 the Joint Research Centre (JRC) Institute for Reference Material and Measurements (IRMM) expressed their interest to collaborate with IVM in the production line of the two CRMs. The role of IRMM is to become the responsible party for the storage and sales of the materials. Due to their expertise in the field they have been given valuable advice in the production. Representatives from IRMM, came to control and assist the fish and water CRM production performed at IVM and KWR (Nieuwegein) respectively.
Following is a demonstration of the work needed for the preparation of such a large sample volume and the importance to follow a well defined protocol to ensure homogeneity and stability of the material. All equipment used during the production was thoroughly cleaned and material used for storage was tested regarding its stability after freezing/unfreezing and absorption test to material.
Fish – A total of 80 kg of frozen pike-perch fillet from the Dutch rivers Nieuwe Merwede and Amer and was delivered to IVM the 15th of November 2011 (Figure 1). The samples were homogenized one week later with the aid of a Stephan cutter of 40 L (Figure 2). Due to the large volume of sample the homogenization was done over three batches and repeatedly (4 times) equally divided into the final containers. For an equal dividing the portions were weighted. The total homogenization in the cutter was more than 2 hours before the correct particle size was reached (non-fibrous particles <2mm, and fibers <5x1 mm, visually measured by isopropanol addition at IVM (Figure 3). In addition, two subsamples were taken out for particle size determination at IRMM.
After particle size confirmation a total of 1249 jars was filled manually and numbered (Figure 4). The jars were sterilized by autoclave (45 min, 120°C, 1.4 bar, Figure 5) and cooled to room temperature before the jars were labeled with labels provided by IRMM. The jars were transported to IRMM who will store the material at room temperature and will keep jars apart for stability tests at different temperatures (Figure 6).
Water – The water CRM was prepared at KWR the 23rd of November 2011. A total of 650 L of drinking water was collected from Nieuwegein (the tap) into a clean stainless steel barrel. The material was spiked with a PFAS mixture due to low natural levels. The spike solution (10 ml) was added to a volumetric flask of 10 L. This flask was used 65 times and hence well rinsed. Four times during the sampling was 0.4 liter of drinking water taken out as control blank (no added spike solution). After the filling of the barrel the total bulk of 650 L was stirred for three hours. A sample of 0.4 L was taken out for PFAS analysis.
The next day the material was stirred for 1 h before the portioning of 0.4 L into plastic bottles of 0.5 L. Bottles were labeled immediately after filling, strictly following the filling sequence (labels supplied by IRMM). After the PFAS concentration was confirmed by IVM the bottles were transported to Isotron Nederland B.B for sterilization by gamma irradiation (25kGy). After sterilization all bottles were transported to IRMM to be stored at -20 °C and some bottles will be kept apart at different temperatures for stability tests.
The whole process took around one week and was very labor intensive. The IRMM partner was satisfied with the procedure and quality of the produced material. Currently the homogeneity of the materials’ are tested, the PFAS concentrations will be investigated in the material and the short and long-term stability will be evaluated. The certification of the material is planned to be finished in early 2013 and after that the material will be available for purchase via IRMM.
Jana Weiss & Ike van der Veen