Methods of numerical differentiation in the analysis of thermodynamic data for the mercury–solution interface

Abstract

Methods of numerical differentiation for the calculation of surface charge densities and relative surface excesses from electrocapillary data are considered. Estimates of the random error in calculated derivatives are made by testing eight different differentiation methods with artificial data containing known levels of random error. The relative merits of these techniques and the number and spacing of data points needed to perform numerical differentiation precisely are discussed.