Multisignal calibration

It has been shown (Danzer and Wagner [1993] Danzer and Venth [1994] Venth et al. [1996]) that the reliability of quantitative analyses can be increased when several signals are used for calibration and evaluation. From theoretical considerations it is expected that a multivariate evaluation increases both the sensitivity and the precision and, therefore, the detection power, too. 

Multisignal evaluation is carried out by means of Principal Component Analysis (PCA) or (Partial Least Squares (PLS) regression. The fundamentals 

Because the first principal component is always situated in direction of the largest variation of the data, the intensity of ypc should be increased compared with the original intensities from which it was calculated, namely theoretically according to the -dimensional Pythagoras principle 

According to the fundamentals of PCA, simultaneously the precision should be improved because the noise is preferably to find in the higher pcs as shown by simulations and experimentally in Danzer and Wagner [1993], OES determination of Mn, Ni, and Cr in steel (Danzer and Venth [1994]) 

Mass Abundance (%) Molecule ion interferences Abundance (%) Isobaric Abundance interferences (%) 

---

Danzer K, Venth K (1994) Multisignal calibration in spark- and ICP-OES. Fresenius J Anal Chem 350 339... 

Danzer K, Wagner M (1993) Multisignal calibration in optical emission spectroscopy. Fresenius J Anal Chem 346 520... 

For evaluation of multisignal measurements, e.g. in OES and MS, more than one signal per analyte can be evaluated simultaneously by means of multiple and multivariate calibration. The fundamentals of experimental calibration and the relating models are given in Chap. 6. 

---