Analyzing Expanding Clays by Thermoporometry Using a Stochastic Deconvolution of the DSC Signal

Abstract

AbstractA version of thermoporometry dedicated to analyzing the pore network of expanding clays is proposed here. The blurred, wide Differential Scanning Calorimetry (DSC) peak obtained upon the melting of a frozen clay sample is processed by means of a deconvolution analysis based on searching for such a temperature distribution of “pulse-like heat events” which, convolved with the apparatus function, gives a minimal deviation from the observed heat flux function, i.e. the calorimetric signal. As a result, a sharp thermogram was obtained which can be transformed easily into the pore-size distribution curve. Results obtained for samples of two Clay Minerals Society Source Clays (montmorillonites SWy-2 from Wyoming and STx-1b from Texas) at different water contents indicate a greater resolution and sensitivity than that achieved by classical thermoporometry using the unprocessed DSC signal. Phenomena corresponding to the evolution of the pore network as a function of the water content have been detected in samples with large water contents subjected to free drying prior to the experiments.