Physical signals, chemometrics

Some of the classical applications of chemometrics are to evolutionary data. Such a type of information is increasingly common, and normally involves simultaneously recording spectra whilst a physical parameter such as time or pH is changed, and signals evolve during the change of this parameter. 

Equations (74) and (75) are valid only for a limited concentration range, known as the linear dynamic range. This range is limited at the upper end by physical phenomena, such as detector saturation, and at the lower end by the limit of determination. This limit is typical of a given analytical procedure and is the lowest concentration at which a determination can be performed with a certain precision [32] (- Chemometrics ->Trace Analysis, ->Trace Analysis). For 99.86% certainty, and provided the fluctuations of the limiting noi.se source can be described by a normal distribution, the lowest detectable net signal is three times the relevant standard deviation ... 

Despite the fact that direct analysis methods exclude a cost-intensive separation step overall analysis cost may still be high, namely by the need for more sophisticated instrumentation (allowing for a physical rather than chemical separation of components) or extensive application of chemometric techniques. The wide variety of additives that are commercially available and employed complicate spectroscopic data analysis. For multicomponent analysis some kind of physical separation of additive signals is often quite helpful, e.g. based on mobility (as in LR-NMR or NMRI), diffusion coefficient (as in DOSY NMR), thermal behaviour (as in a thermal analysis and pyrolysis techniques) or mass (as in tandem mass spectrometry). The power of signal processing techniques (such as multi-wavelength techniques, derivative spectrophotometry) is also used to the fullest extent. 

Fischer et al. [101] investigated the simultaneous quantification of the content of several additives in PVC with an in-line diffuse reflectance probe. The signal from diffuse reflectance can be affected by a number of physical properties of the sample, rather than just its chemical make up. This makes obtaining quantitative data very difficult. Chemometric analysis showed the possibility of detecting even small amounts of additives (3%) with an absolute prediction error of 0.3%. Step-scan PA-FTIR spectroscopic studies were used to study surfactant exudation and film formation in PS-nBA latex films [102]. 

---