APPLICATION OF PARTIAL LEAST-SQUARES REGRESSION TO REMOTE QUANTITATIVE-ANALYSIS BY THE RAMAN-LASER-FIBER-OPTICS (RLFO) METHOD

A quantitative analysis program using the PLS method has been implemented and coupled with a Raman-laser-fibre optics installation. A mean-centring pretreatment has been incorporated to reduce the fibre spectrum effect. The contribution of the fibre Raman signal remaining in each spectrum after this pretreatment was entered in the model calculation. This explicit modelling of the fibre Raman signal fitted the model better to the "real correlation" and thus improved precision. Differences of the spectra recorded at two different temperatures proved that temperature can also be estimated. As a partial example of the performance of this method the remote quantitative analysis of benzene/heptane mixtures was performed with fibre lengths up to 100 m. The speed and efficiency of this method and its low cost increase the possibilities of applying RLFO to remote and in situ real time chemical analyses and monitoring.