XRF-analys av förorenad mark
undersökning av felkällor och lämplig provbearbetning
X-ray fluorescence (XRF) detection of heavy metals is a cost- and time-effective method for investigation of polluted areas. Compared to laboratory analysis, XRF analysis is limited by high detection limits and uncertainties in some situations. Preparation of samples is known to affect the results of measurements.The purpose of this thesis is to bring a deeper understanding of how different factors affect the results of XRF-analysis. A large number of measurements have been made with the instruments Niton XLt 700 and Niton Xli 700. Results from measurements of lead, zinc and copper have been analysed.This study has shown that a greater moisture content will give a lower measured concentration for the same sample. If the moisture content is known, it is possible to make good approximations of concentrations in dry samples. No obvious differences were found between different metals.Proper homogenisation of samples was found to be one of the most critical issues. It should be noted that laboratories do not usually homogenise samples prior to analysis, so samples sent for laboratory verification should therefore be homogenised before sending.Sample preparation, including drying, sieving to finer grain-sizes and analysing in a special sample cup was found to increase the measured concentrations and decrease the variation between measurements. The concentrations differed with different sample matrices and with sample preparation. Because of the differences, measurements should always be correlated with laboratory analysis before any conclusions can be made. In situations when metals are well within detection range of the XRF instrument, the only preparation needed is in most cases homogenisation. If metal levels are just at the limit, or just below the detection limit, further preparations of samples should be made. When drying samples makes detection possible, more preparation is usually not worth the effort. If detection limits still are a problem after drying, the samples should be sieved and analysed in sample cups. In situations where concentrations still are below thedetection limit when analysing in sample cups, or if the measurements are disturbed for other reasons, another method than XRF-analysis must be used. When XRF analysis is possible, the method is preferable compared to only laboratory analysis. This cost- and time effective method makes a larger number of measurements possible, and gives a more complete overview of a polluted area.