Abstract
The least squares fitting of experimental results with a non‐linear model can result in a serious loss of accuracy in the model parameters estimation if the statistical nature of the method is not correctly considered. This occurs when the experimental data is fitted to a set of functional parameters that depend in the model parameters to be estimated in the end. A realistic example can be found in the two state model of monomer‐excimer kinetics. The decay curves of the monomer and excimer are a sum and a difference of two exponentials, respectively. It is usual to fit the experimental decays in order to obtain the pre‐exponential factors and decay constants, thus using a reparametrization that is non‐linear with respect to the model parameters. This procedure is thoroughly discussed and a new method to analyse the decay curves that circumvents the problem of reparametrization is presented. The proposed method yields improved results with less than 7% bias in the recovered rate constants. Monte Carlo simulations have been performed in order to obtain confidence intervals for the fitting and model parameters.
Similar content being viewed by others
References
J.S. Alper and R.I. Gelb, J. Phys. Chem. 94 (1990) 4747; 95 (1991) 104.
P.R. Bevington, in: Data Reduction and Error Analysis for the Physical Sciences (McGraw-Hill, New York, 1969).
J.B. Birks, Rep. Prog. Phys. 38 (1975) 903.
P. Gans, in: Data Fitting in the Chemical Sciences (Wiley, New York, 1992).
D.W. Marquardt, J. Soc. Ind. Appl. Math. 11 (1963) 431.
W.H. Press, B.P. Flannery, S.A. Teukolsky and W.T. Vetterling, in: Numerical Recipies: The Art of Scientific Computing (Cambridge University Press, Cambridge, 1986).
S. Ross, in: A Course in Simulation (Macmillan Publishing Company, New York, 1991).
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Farinha, J., Martinho, J. & Pogliani, L. Non‐linear least‐squares and chemical kinetics. An improved method to analyse monomer‐excimer decay data. Journal of Mathematical Chemistry 21, 131–139 (1997). https://doi.org/10.1023/A:1019114217567
Issue Date:
DOI: https://doi.org/10.1023/A:1019114217567