XRF

XRF is a relatively old and established technique, with the first commercially available instruments being launched in 1945. Since that time significant advances have been made and today “hand held” XRF analysers deliver detection limits that are well below even SGVs for domestic applications for most heavy metals.

The most significant advance has been the change from using radioactive isotopes for the source of X-Rays to a simple X-Ray tube (XRT). The new XRTs deliver a much wider energy spectrum that will cause all the heavy elements (and many light elements) to fluoresce in the X-Ray spectrum. Using a radioactive source often meant having to have several different isotopes to be able to detect the full range of heavy metals.

The other advantage of this system is that as no radioactive sources are present, it is possible to carry the XRF unit around in complete safety. Airlines will happily accept the unit as hand luggage, even with the heightened security that is currently in place.

XRF works in a similar way to ICP or AAS. In each case the atoms in the sample are energised and as they relax, they emit a characteristic series of wavelengths. For ICP and AAS these wavelengths are in the visible spectrum. For XRF the wavelengths are in the X Ray spectrum. The advantage of XRF over these other two methods is that the samples can be solid, so unlike ICP and AAS, soil samples do not need dissolving in concentrated acid.

X Rays are very powerful, so care is needed when using the instrument, but numerous safety features are built in so it is very unlikely to cause X Ray exposure to the operator or any bystanders. QROS staff are trained to use the instruments safely and effectively.

Sample analysis is very rapid, with a typical scan taking just 60 seconds. This will generate an X Ray spectrum (simlar to an ICP or AAS spectrum) that the software interprets. An interesting feature of XRF is that it does not need to be calibrated. The system works by using Fundemental Parameters (FP) that relies on the fact that the intensity of the X-Ray produced from an element is directly proportional to the amount of the element present. This calibration is ideal for relatively high metal concentrations ranging from 200 mg/kg up to % concentrations and compensates for differing soil types such as clays, chalks, sand and high organic content samples.

More sophisticated systems can use Empirical Parameters (EP) to modify the FP method by constructing a calibration curve using real soil matrixes. This can compensate for potential matrix effects in samples with high concentrations of chalks, sand, clay etc . Empirical Parameters is the preferred method when looking for metals at low concentrations, typically below 200 mg/kg. When high concentrations of metals are present, the EP calibration can cause the actual concentration to be under reported. This is why it is important when using EP to have the facility to set a concentration flag for each metal where this can happen and have the instrument flag this up if the metal exceeds the EP range..

QROS uses the Oxford Instruments XMET 3000 TXS with both Fundemental and Empirical calibration systems fitted. For unusual matrixes, QROS can also construct a specific calibration curve and load this into the instrument. Our skilled operators will know which calibration to use and this instrument has the capability of switching between calibrations very easily on site so the appropriate calibration is used every time, regardless of the metal concentrations found.

Correlation with Laboratory Analysis

XRF can provide excellent correlation with conventional laboratory analysis, providing the appropriate calibrations and sample preparations are used. A point to bear in mind however is that XRF will “see” all of the target metal in the sample. The laboratory may not get 100% extraction efficiency. Unfortunately it is very difficult for a laboratory method to determine the extraction efficiency because for most samples, approximately 70% of the solids are not dissolved during the extraction process. This means there is the possibility of some of the target metals being trapped within the matrix. This is sometimes the case for Chromium, Nickel, Cobalt and Arsenic, all of which form compounds within silicates that typical laboratory methods will not be able to extract. This is usually the reason for the XRF reporting higher results than the laboratory. (see case study 2 and case study 5 for more data)

The typical volume of sample scanned by the XRF is matrix dependent, but the X Rays will penetrate between 5 - 20mm into the soil. The X Ray beam has an area of approximately 2 square centimetres, giving a total scanned volume of between 1 and 4 cubic centimetres. Assuming a density of around 1.8 for soil, this is equivalent to a sample weight of between 1.8 - 7.2g. A good technique will scan a 500g sample at three different points and the results are then averaged. This means that when using XRF, between 7.2g and 21.6g of soil are actually analysed. This is compared to the typical laboratory sample where only 1g of the original 500g of soil collected is actually being extracted and analysed.

Mercury and Cadmium are the two heavy metals that are not detected at the SGVs for domestic housing development sites. Typically these metals are detected above 20 ppm, which is more than adequate for industrial sites. For the domestic limits, QROS would use an ASV technique that can detect these metals at concentrations below 1ppm in soil.

Data Reporting

The data from the XRF can either be shown on the instrument or quickly downloaded to a PC. QROS uses its own in house software to compile the data into an easily readable report shown below.
(Click here for larger image).

Each sample is analysed three times and an average obtained. The average value is then given an homogeneity indicator by colouring the value green, yellow, orange or red. Green is for samples that show very good homogeneity. This means that the sample is very likely to be a good indicator of the soil concentration in that area. Samples that have a red indicator shows the three results vary by over 400%. This means that the metal in the sample is very unevenly dispersed, giving hot spots in one small patch and low levels in another. This is a clear indication that the sample may not be truly representative of the area sampled. It also shows that the laboratory results will also be highly variable and may not correlate with the XRF result or with a second laboratory result. Caution should be used when interpreting the data with a red indicator and it is advisable to analyse several samples from this area to obtain a better average value. Using the XRF technique means this is possible without incurring significantly higher analytical costs.

A yellow indicator shows the sample has some variability, but is not significant. Orange shows that there is significant variability, but the spread of results is not as extreme as a red indicator. More samples from the area should be analysed and averaged when red or orange shows in order to obtain a more representative average value.