Determination of Accurate Specific Heat Capacities of Liquids in a Reaction Calorimeter, by Statistical Design

A method using statistical design is proposed to determine the measurement error of specific heat capacities (Cp) of liquids in any reaction calorimeter. This method only takes into account the experimental specific heat of the liquid, its volume and the reactors stirring rate. With the RC1e calorimeter used in this study, the Cp measured by its QuickCal mode was overestimated by up to 0.4 J·g–1·K–1, whereas with its RTCal mode, more accurate values of Cp (± 0.2 J·g–1·K–1) were obtained. The statistical model obtained based on the QuickCal mode predicts experimental conditions which improve the accuracy of Cp, and could also be used to correct the value of Cp in the rest of the experimental domain. The corrected values of Cp for test solvents using this model were close to those cited in literature (≤5%). As an application, the specific heat capacities at 25 °C of less well documented solvents were determined: anisole, ethyl-tertiobutylether, methyl-tetrahydrofurane, diethoxymethane, dimethyl-ethyleneurea and dimethylpropyleneurea.