Amorphous silica analysis in terrestrial runoff samples

The correct analysis of amorphous silica concentration (CASi) in natural waters is crucial if one wants to correctly quantify terrestrial and/or riverine ASi fluxes. Soil ASi measurements are conducted with a constant solid to solution ratio (). As the suspended particulate matter concentration (CSPM), and therefore, cannot be exactly known a priori in river samples. It is important to understand how variations in effect analysed CASi. The objectives of this paper are (i) to investigate whether and how variations in values affect measured CASi in river runoff samples and (ii) to investigate whether or not it is possible to define a range of within which CASi in runoff and/or soil samples can be accurately measured. For the laboratory experiment 30 runoff samples with a wide range of CSPM, typical for the Belgian Loam Belt, were prepared and analysed using the alkaline digestion method (0.1M Na2CO3). Our study confirmed that the alkaline digestion method proposed by DeMaster can be used for runoff samples provided that is within certain limits: at very low (<0.1kgm-3), subsample heterogeneity results in high variability of measured CASi while at higher values (>0.8kgm-3) incomplete dissolution of ASi as well as the reduction of mineral dissolution rates results in underestimated CASi. As both errors compensate one another, the range of applicable -values can be extended above the theoretically correct limit (1.6kgm-3). The finding that reliable measurements can be made within a relatively wide range of values (0.1≤≤1.6kgm-3) is important. It is now possible to propose a method for the measurement of ASi in runoff samples. We make recommendations for ASi analysis distinguishing samples with a low and high CSPM. For samples with a low CSPM (≤1.6kgm-3) the standard procedure is proposed while for samples with a high CSPM (>1.6kgm-3) an adapted procedure is proposed, analogue to that for soil samples. However, one should be aware that the range and limits for proposed here may depend on the type of sediment to be analysed: it is therefore recommended to evaluate the performance of the method again before it is used in other environments. © 2011