Submitted to: Near Infrared Spectroscopy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: December 22, 2008
Publication Date: January 23, 2009
Citation: Liu, Y., Delwiche, S.R., Graybosch, R.A. 2009. Two-dimensional correlation analysis of near-infrared spectral intensity variations of ground wheat. Near Infrared Spectroscopy Journal. 17(1):41-50.
Interpretive Summary: Generalized two-dimensional (2D) correlation spectroscopy has been established as a viable means to analyze and extract useful information from conventional one-dimensional infrared (IR), near infrared (NIR), Raman, and visible spectra. However, most of the applications are limited to a small number of spectral data affected by simple and univariate static perturbations, and one of specific challenges might be how to implement the 2D correlation analysis in much large spectral sets, which are common in chemometric model developments and actually include diverse fluctuations in chemical and physical components. Here, we present an innovative approach to explore the multi-variable variations within the large NIR spectra of ground wheat by averaging the spectra of neighboring reference values and developing cluster classification via principal component analysis. The result reveals significant differences in spectral features between low and high protein wheat as well as low and high SDSS wheat, even between high protein wheat and high SDSS wheat. Researchers working on NIR spectral and imaging screening of grain quality and subsequent model developments will benefit from the findings of this research.
Generalized two-dimensional (2D) correlation analysis was applied to characterize the NIR spectral intensity fluctuations among many spectra of ground wheat with multi-variable variations. Prior to 2D analysis, the spectra having neighboring protein / SDSS reference values were averaged and then new spectral sets were subjected to PCA for cluster classification. Synchronous and asynchronous 2D correlation spectra enhanced spectral resolution and provided information about protein content- and SDSS index- dependent intensity changes not readily accessible from one-dimensional NIR spectra of wheat. The results revealed remarkable differences between the protein content- and SDSS index- induced NIR spectral response. Of the interest were several unique protein bands around 1980, 2040, 2200, 2260, and 2350 nm in high protein and SDSS wheat.