A practical method to evaluate major statistical errors in isothermal titration calorimetry

Abstract

Continuous efforts have been devoted to improve the accuracy and reliability of isothermal titration calorimetry (ITC) in terms of instrumental design, modeling and data processing. Among various factors, background noise, proportional error and injection volume error lead to heteroscedasticity (non-constant variance) of the titration heat. The variance function method provides an efficient way to estimate the non-constant variance and weight of the titration heat. Following this approach, in this work, the heat residual recorded by calorimeter model of Nano ITC Standard Volume was examined in detail. Based on the CaCl2-EDTA saturated titration and water blank titration, 9 heat variances, each from 60 injections, have been analyzed, which provide reliable information about the variance function. The major statistical errors for general titrations were estimated and the non-constant variance distribution of typical titration (BaCl2 to 18-Crown-6) was successfully predicted.