|Barton Ii, Franklin|
Submitted to: UJNR Food & Agricultural Panel Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 11/6/2003
Publication Date: 12/3/2003
Citation: Himmelsbach, D.S., Meadows, F., Barton II, F.E. 2003. Determination of rheological properties from spectroscopy. UJNR Food & Agricultural Panel 32nd Annual Proceedings, Tsukuba, Japan. pp. 143-147. Interpretive Summary: The ability to rapidly predict rheological properties has eluded cereal scientists for many years. To accomplish this, Rapid Visco Analyzers (RVAs) are now widely used to assess the cooking and processing of starch and flours of grains. However, analysis time by RVA with sample preparation is still relatively long. This limits the usefulness of the method. On the other hand, an optical spectrum of a sample takes less than 2 minutes to acquire. Correlation of near-infrared and Raman spectra to RVA curves has permitted the development of predictive models that can be used in lieu of direct collection of viscograms. This shortens the time needed to determine cooking and processing qualities of flours.
Technical Abstract: The ability to predict viscoelastic properties from vibrational spectra has been further investigated. Both dispersive near-infrared (NIR) and Fourier-transform Raman (FT-Raman) spectra were used to generate two-dimensional matrix maps versus Rapid Visco Analyzer (RVA) viscograms. After calibration with partial least squares analysis type-2 (PLS2), models produced were used to predict entire viscograms and thus viscoelastic parameters. In this study 86 rice samples that varied in amylose content (0.41-24.90% w/w) and protein content (8.47-11.35% w/w) were used as subject materials. Since these spectroscopies are already employed to predict the chemical composition of samples, the additional capability of predicting rheological properties could eliminate the need to maintain both spectrometers and viscometers for grain analysis. In addition the time required to obtain all analyses is reduced to less than 2 minutes.