Author
 |
Himmelsbach, David |
|
Submitted to: Diffuse Reflectance International Conference Proceedings
Publication Type: Abstract Only Publication Acceptance Date: 7/23/2004 Publication Date: 8/8/2004 Citation: Himmelsbach, D.S. 2004. Chemometric predictions of grain flour attributed using nir-ft/raman spectroscopy [abstract]. 12th International Diffuse Reflectance Conference. p.8. Interpretive Summary: Technical Abstract: NIR-FT/Raman spectroscopy can give comparable to improved chemometric results versus infrared absorbance spectroscopy for prediction of grain flour attributes, provided proper attention is given to data collection and pre-processing. An NIR laser is generally required for Raman excitation when investigating vegetative materials, such as grains, in order to reduce the effects of the normally stronger and competing fluorescence effect. Use of an NIR laser often does not completely eliminate fluorescence and residual fluorescence manifests itself as a sloping baseline superimposed on the Raman spectrum. Thus, baseline removal or derivative pre-processing is required to overcome this interference. Normalization of the resulting data is also typically required to overcome variations in laser response. Raman data collected over multiple time periods or instruments also necessitates it be processed with background correction and/or with standardization against reference samples. Attention to this kind of detail can yield results that take advantage of the fact that the Raman spectra are less convoluted than those of mid- or near-infrared. This permits the processing of data that is based on fundamental identifiable events with few overlapping bands. Examples of NIR-FT/Raman chemometric results will be presented on the prediction of attributes of rice and wheat flours. |