Choice of optimal conditions for X‐ray fluorescence analysis of milk products with varying fat content

The design of experiments was used to study sources of errors in the sample preparation and to choose optimal conditions for X‐ray fluorescence (XRF) analysis of milk products of varying fat content. The measurements were performed using the conventional wavelength‐dispersive XRF (WDXRF) and the total reflection XRF (TXRF) techniques. For WDXRF, the dried milk samples were pelletized, and the mathematical models were constructed, which described the dependence of XRF intensity on the pelletizing pressure, the pellet weight and the milk fat content. The effect of radiation time on the stability of pressed milk powder samples was also estimated. When applying TXRF, the sample preparation involved diluting milk samples with the ultrapure water, adding the internal standard (Ga) and drying of a sample aliquot on a quartz glass sample carrier. The mathematical models were designed, which described the dependence of the sample preparation error on the dilution ratio and sample aliquot pipetted on the carrier. A physical interpretation of the obtained mathematical models was proposed. Copyright © 2013 John Wiley & Sons, Ltd.