[Qualitative and quantitative detection of beet syrup adulteration of honey by near-infrared spectroscopy: a feasibility study].

In order to further investigate the utility of near-infrared spectroscopys (NIRS) in rapidly detecting honey adulteration, near-infrared spectroscopy in combination with chemometric methods was investigated for qualitative and quantitative detection of beet syrup adulteration of honey. Total prediction accuracy of testing set was 90.2% by partial least squares-discriminant analysis (PLS-DA) for authentic and adulterated honey samples. Total prediction accuracy of testing sets was all below 33.3% by different discriminant methods for classes of adulteration level. The quantitative analysis of adulteration level by PLS regression gave satisfying results if adulterated honey samples were got from the same one authentic honey sample: correlation coefficient (r)of actual values versus predicted values was 0.9829 and root mean square error of prediction (RMSEP) was 1.394 2 in testing set, otherwise it gave dissatisfying results for the adulterated samples from different botanical origins or the different samples of the same botanical origins. The results showed that NIRS could be applied for rapid detection of authentic and adulterated honey samples, but not for detection of classes of adulteration level and quantification of adulteration level with beet syrup.

[Prediction analysis of soluble solids content and moisture in honey by near infrared spectroscopy].

A new method for the analysis of soluble solids content (SSC) in honey by near infrared spectroscopy (NIR) was developed, and moisture was also analyzed. The partial least square regression models of SSC and moisture were built for different pretreatments of the raw spectra in different spectral range. Good predictions were always obtained for all models. The best models of SSC and moisture were obtained by using Norris (3,2) smoothing + first derivative + multiplicative signal correction in total spectral range. The coefficient of determination (R(CV)2) and root mean square error of cross validation (RMSECV), the coefficient of determination (R(p)2) and root mean square error of validation sets (RMSEP) were 0.9986, 0.190, 0.9985 and 0.127 respectively for SSC, while for moisture they were 0.9984, 0.187, 0.9986 and 0.125 respectively. NIR could be used to analyze SSC and moisture in honey. The result of this article was better than that of related documents for moisture.