Effect of sample preparation procedure on standardless wavelength dispersive X-ray fluorescence analysis of plant samples

Abstract

XRF (X-ray fluorescence analysis) has been applied as the method for analysis of plant samples, but not in its full potential. This method would be highly suitable for monitoring programs and screening studies. A simple sample preparation procedure makes this technique time and cost-effective, although the absence of adequate calibration standards for some analyzed matrices can be a significant limiting factor. Nevertheless, this obstacle can be overcome by the development of the so-called “standardless method” approach. It is important to explore the possibilities and limitations of such method applied to an analysis of plant materials. The accuracy of XRF analysis is highly dependent on the physical characteristics of the sample, which can be significantly affected by the sample preparation procedure. Variations of 3 crucial sample preparation parameters: the mass of the samples (from 1 to 5 g), binder ratio (from 0 to 25% of wax), and pressing pressure (from 10 to 25 t) were tested on needles obtained from two widespread conifer species: Pinus nigra and Abies alba. For most elements, the correlation between the concentration and the change of the binder ratio was statistically significant, while there was no statistically significant correlation between the concentration and the change of the pellet mass and applied pressure. For both types of needles, an estimated systematic error was always higher than a random error. The difference between two types of errors has been higher among samples with different wax content than in samples with different pellet mass, while the variation of applied pressure did not introduce any significant systematic error. On average the change of wax content inflated the value of systematic error by 18 and 13% for the pine and fir needle samples (respectively), while the change of the sample mass has introduced systematic error in lesser extent, on average it was 11% for pine needles and 10% for fir needle samples. The change of pelletizing pressure affected the results by only 3.9% for both conifer needles.