Comparison of laser diffractometry and pipetting methods for particle size determination: A pilot study on the implications of result discrepancies on soil classification

Abstract

In recent decades, the determination of particle size distribution (PSD) using the laser diffraction method (LDM) has become increasingly common, supplanting traditional sedimentation techniques. Advances in everything from sample preparation to software settings have been realized globally, whether through recommendations from laser diffraction (LD) manufacturers or through user experiences. These developments seek to enhance accuracy and diminish the uncertainties associated with new methodologies. Particularly in the determination of PSD using LDM on various LD instruments and in comparison with the sieve–pipette method (SPM), discrepancies in PSD frequently arise. This article aims to mitigate these discrepancies by predefining parameters, specifically through the adjustment of LD software settings and sample preparation (employing a uniform set of dispersed samples in potassium hydroxide) on two widely used LD instruments for soil measurements: Mastersizer 3000 and Analysette 22. Additionally, these samples were analyzed using the traditional SPM (ISO 11277, 1998), with the results from LDM and SPM subsequently compared. The paper also explores the impact, range of user options, and limitations of predefined software settings on both LD instruments. Eighty soil samples were analyzed for PSD, collected from arable land in the cadastral area of Hrušky, district of Břeclav (Czech Republic), in spring 2022, from depths of 0- to 10-cm and 10- to 20-cm. Significant differences in PSD were confirmed, although the trends of the grain size distribution curves were very similar to those of LDM. Results from the Mastersizer underestimated the clay fraction by an average of 17% compared to SPM, at the expense of the sand fraction, whereas the silt fraction was underestimated by a maximum of 4%. Conversely, Analysette 22 overestimated the silt fraction by an average of 37% at the expense of the sand fraction, confirming only a slight difference in the clay fraction: 3%. Moreover, the quantity of sample entering the dispersion unit was identified as a significant issue when comparing LD instruments, despite the obscuration value being nearly identical, that is, 20%–30%. Therefore, it was not possible to achieve the same or similar weight when introducing suspension into circulation. The robustness of the obtained results underscores the importance of understanding input parameters when interpreting results between different methods.