A novel pure component contribution algorithm (PCCA) for extracting components' contribution from severely overlapped signals; an application to UV-spectrophotometric data.

A novel, simple and accurate algorithm capable of extracting the contribution of each component from a mixture signal where the components are completely overlapped was developed. It is based on the development of a coded function which eliminates the signal of interfering components using mean centering as a processing tool; finally the pure contribution of each component is extracted. The algorithm allows the determination of each component as a single one. It was validated by the use of simulated data set of three overlapped signals and tested against simulated random noise. Two fit values were developed and calculated for optimization, one to test that that the absorptivity values of the extracted spectra are within the confidence limits of the slope and the other is the correlation between the pure and extracted spectra. It has been successfully applied to real UV data of binary mixture of Ibuprofen and Paracetamol and ternary mixture of Amiloride hydrochloride, Atenolol and Hydrochlorothiazide in tablets and capsules, respectively. The results were compared to previously reported separation method and no significant difference was found regarding both accuracy and precision.