Quantitative Analysis of Cotton Lint Trash by Fluorescence Spectroscopy

Authors: Gamble, Gary; Foulk, Jonn

Submitted to: Journal of Agricultural and Food Chemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/20/2007
Publication Date: 6/27/2007
Citation: Gamble, G.R., Foulk, J.A. 2007. Quantitative analysis of cotton lint trash by fluorescence spectroscopy. Journal of Agricultural and Food Chemistry. 55(13):4940-4943.

Interpretive Summary: The presence of cotton plant botanical components, such as leaf and stem, in baled cotton is an important criterion for assessing the classing grade, which in turn impacts the value of the cotton. It is desirable to identify specific trash types present in the baled cotton in order to establish procedures during ginning the processing which will optimize the removal of the problematic trash. Individual trash particles may be difficult to classify since they may be decreased in size to the point that they cannot be identified based on geometrical features. There are currently no established methods available to adequately classify trash types based on botanical origin. The present work addresses this issue by using fluorescence spectroscopy and chemometric analysis. A calibration model is developed from 112 mixtures of six different trash types, and is demonstrated to be capable of predicting trash component concentrations with a high degree of confidence.

Technical Abstract: The presence of cotton plant botanical components, or trash, embedded in lint subsequent to harvesting and ginning is an important criterion in the classification of baled cotton by the USDA-Agricultural Marketing Service. The trash particles may be reduced in size to the point that specific trash types are not identifiable by image or gravimetric analysis, and it is desirable to classify these ground trash particles so that processing lines may be optimized for removal of the problematic trash in order to enhance processing performance and cotton lint quality. Currently, there are no methods available to adequately classify trash based on botanical origin. The present work attempts to address this issue through the analysis by fluorescence spectroscopy of Dimethylsulfoxide extracts of mixtures of six botanical trash types. The fluorescence data are subsequently subjected to chemometric analysis. The resulting Partial Least Squares calibration model obtained from 112 mixtures is demonstrated to be capable of predicting individual trash component concentrations with a high degree of confidence.