Document Type

Thesis - University Access Only

Award Date

1998

Degree Name

Master of Science (MS)

Department / School

Nutrition, Food Science, and Hospitality

Abstract

High Performance Liquid Chromatography (HPLC) has been a traditional method for isoflavone analysis in soybeans. However, this method is costly and time consuming. Ultra-violet visible spectrophotometry (UV-Vis) and near infrared reflectance spectroscopy (NIRS) were used for the first time in developing alternate methods to HPLC. Isoflavones were extracted with 80% methanol from 210 soybean varieties. The extract was analyzed by HPLC and then scanned by an UV-Visible spectrophotometer. The absorbance values were regressed against the corresponding HPLC analyzed isoflavone values. Mathematical models were developed for predicting individual and total isoflavones based on the scanned data. Predicted values had a high correlation with HPLC analyzed values (R2 for total isoflavones was 0.86). UV-Vis was found to be an effective method for analysis of most isoflavones

NIRS calibrations were developed using HPLC as the reference method. Calibrations were effective for daidzin (R2=0.88), genistin (R2=0.81), malonyl daidzin (R2=0.90), malonyl genistin (R2=0.85), and total isoflavones (R2=0.93). The calibration models for the analysis of glycitin, daidzein, malonyl glycitin, acetyl genistin and genistein had poor correlation because the reference method itself was not accurate enough to measure these isoflavones that are found in negligible amounts. NIRS calibration models showed potential as a rapid, cost-effective method for analysis of isoflavones in large number of samples. UV-Vis and NIRS were compared for their effectiveness. Sixty-one soybean samples, selected by NIRS, were used to develop the calibration equations for both the UV-Vis and NIRS. An independent set of 42 soybean samples with a wide range of isoflavone concentrations (1098 to 2406 μg/rnl) was used to validate the calibration equations. The coefficient of determination (R2) between the predicted values and those from the HPLC were used to evaluate the effectiveness of the two methods. It was found that UV-Vis was a much more effective method than the NIRS. The former has a great potential in rapid analysis of isoflavones in soybeans.

Library of Congress Subject Headings

Flavonoids -- Spectra Flavonoids -- Analysis Soybean -- Composition

Format

application/pdf

Number of Pages

92

Publisher

South Dakota State University

Download

Share

COinS