Fiber optic technique for determination of moisture content in honey

This thesis presents a critical study on the application of fiber optic technique as a fast and accurate method to determine moisture content in honey based on reflection, transmission and absorption properties in the wavelength between 300nm and 800nm. Measurements were conducted using an Ocean Optics USB4000 UV-VIS-NIR spectrometer. The honey samples used in this work were from various different honey producers. The actual moisture content was found from standard oven drying method. Moisture content of fresh honey was approximately 14.3%. All the fibre optic measurements suggest strong presence of NADH and Flavins at approximately 340 nm and 480 nm, respectively. However, reflection measurement was imprecise due to the translucency of diluted honey. In contrast absorption measurement results for all honey samples of various percentages of moisture content were distinctively unique. Maximum absorbance was found for samples with moisture content 47.5% at all wavelengths due to high concentration of sugar, crystallization. A finite element method (FEM) was also used to calculate the absorbance in fresh honey. The FEM modeling was implemented using COMSOL Multiphysics version 3.5 software. The electric field distribution trough out fiber and sample were visualized to study the electric field pattern at different wavelengths and a comparison of energy levels trough honey and air were studied. Good agreement between measured and calculated results was obtained for fresh honey. Hence, the accuracy of predicting moisture content using fiber optic technique of honey was obtained. Good linear relationship was found between absorbance and moisture content of honey within the range of 430nm until 495nm where the board band excitation of Flavin occurs. However, for all other wavelengths, only polynomial second order relationships were found to give good regression coefficients due to crystallization of honey. Calibration equations have been established to predict moisture content in honey based on absorption measurement. The most accurate calibration equation was found at 460nm where the mean relative error between the true and predicted moisture content was 0.08