Statistical Analysis of Adsorption Experimental Data – the Influence of the Selection of Error Function on Optimized Isotherm Parameters

Dawid Myśliwiec, Stanisław Chibowsk
2014 Annales UMCS Chemistry  
Experimental adsorption data were analysed by fitting them to nonlinear forms of Langmuir and Freundlich isotherms. Optimization of the parameters was performed by nonlinear least square regression with different forms of error function, namely: vertical, horizontal, orthogonal, normal and squared normal. The results showed, that isotherm parameters may be affected by the selection of error function and that they are more sensitive to its' form in case of Langmuir equation. We did not find any
more » ... e did not find any correlation between a type of the function and performance of the regressionprocedure requires optimization for every experimental dataset and every model being fitted.
doi:10.2478/umcschem-2013-0006 fatcat:doododa6avhs7bvshgu3hvyrey