6.1 - Multi-Modal-Spectroscopy and Multivariate Data Analysis as a Tool for Non-Invasive Process Analysis
- Event
- 11. Dresdner Sensor-Symposium 2013
2013-12-09 - 2013-12-11
Dreikönigskirche - Haus der Kirche Dresden - Chapter
- Neuartige Sensor-Systeme
- Author(s)
- R. Kessler, K. Rebner - Hochschule Reutlingen/D, W. Kessler - Steinbeis Transferzentrum Prozesskontrolle und Datenanalyse, Reutlingen/D
- Pages
- 102 - 107
- DOI
- 10.5162/11dss2013/6.1
- ISBN
- 978-3-9813484-5-3
- Price
- free
Abstract
The aim of Process Analytical Technology (PAT) is essentially to understand and control a process. Online and inline process spectroscopy is a well-established and widely used technique to noninvasively control complex reactions in industry on a molecular basis. In combination with standard multivariate data analysis (MVA), the spectral information is correlated to response variables. However, often these MVA results into empirical calibration models which are based on pure statistical correlations and therefore are lacking causality and first principle interpretation. This paper describes the procedure how to use multi-modal spectroscopy as a tool to separate multi-dimensional information. This allows e.g. to separate scatter from absorption when opaque systems are investigated. Furthermore, by means of a pushbroom imaging system it is possible to identify the distribution of the chemical components on a surface. Pushbroom imaging with multi point detection ability can reduce the costs for online or inline PAT in the near future.