Evaluation of the Use of Polyvinyl Alcohol in Preparation of Pressed Samples for X-ray Fluorescence Analysis

The reliability of results of X-ray fluorescence (XRF) analysis, a method widely used to determine the elemental composition of various materials, depends significantly on sample preparation quality. In preparing pressed pellets for XRF determination of trace elements, use is often made of a binder in powder form, which leads to considerable time expenditures for thorough homogenization. Here, we report our results on the use of a polyvinyl alcohol solution as a binder in the preparation of substrate-free pressed samples for XRF determination of trace elements. We present parameters that ensure preparation of a durable, homogeneous pellet that meets the condition of minimizing the effect on the intensity of analytical lines of the elements to be determined. The optimal sample weight in the proposed pressed pellet fabrication process is 7 g, polyvinyl alcohol is added in the ratio 7 : 1, and the optimal pellet drying time after pressing is 1 h. Calculated coefficients of variation of the analytical signal (<2%) suggest that the pellets we prepared are homogeneous. The maximum standard deviations for the ranges of elements chosen in the pressed pellets are far smaller than the acceptable standard deviations. This suggests that the proposed sample preparation procedure with the use of polyvinyl alcohol is appropriate. We demonstrate that, during prolonged storage and repeated measurements, pressed pellets prepared using a polyvinyl alcohol solution retain their initial structure, without changes in the intensity of the analytical lines of their constituent elements. Measurement accuracy in subsequent determination of rock-forming elements is consistent with their certified contents. The present results can be helpful for reducing the time needed for the preparation of a representative pressed sample in determination of elements without additional measures (coating films) and also in trace element and silicate analysis with the use of one sample.