Signal Contribution Effects on Error Propagation in Self-Modeling Curve Resolution Methods

Multivariate curve resolution (MCR) methods are powerful tools to investigate complex chemical systems whenever there is little or no knowledge about the system [1]. Despite of its high resolution power in the analysis of overlapping chromatographic and spectroscopic signals, MCR-ALS solutions usually suffer from noise propagation uncertainties and rotational ambiguities. It is important to know the extent of uncertainties in the resolved solutions. Without quantifying these uncertainties, one might mistakenly consider that the obtained solution is the true one [2]. When ambiguity is present, every resolved profile can be represented by a band of feasible solutions instead of by a unique profile. In the presence of experimental noise, the estimation of this band of feasible solutions is even more difficult and uncertain [3]. If the intensity of a component in a data set changes, it does not change the overlap of the profiles, but the rotational ambiguity will change. The aim of this work is to investigate how signal contribution and reduction of the data set dimensions affect the reliability of feasible solutions. To study the effect of the signal contribution, several identical two-component systems with different intensities of concentration profiles were simulated and the uncertainties of feasible solutions were investigated. To study the effect of the reduction of the data set dimensions, several simulated data sets and their reduced sizes have been systematically investigated. The obtained results by simulated data sets showed that increasing the intensities of concentration profiles decrease the uncertainties of the feasible solutions, and reduction the data sets dimensions increase the uncertainties of the feasible solutions.