Study on Raman Spectral Imaging Method for Simultaneous Estimation of Ingredients Concentration in Food Powder

This study investigated the potential of point scan Raman spectral imaging method for estimation of different ingredients and chemical contaminant concentration in food powder. Food powder sample was prepared by mixing sugar, vanillin, melamine and non-dairy cream at 5 different concentrations in a bench top acoustic mixer for 20 minutes. A 785 nm laser source was utilized to collect Raman spectral covering the surface area of 96 mm x 46 mm and sample depth of 3 mm for each concentration sample, with spatial resolution of 0.5 mm. Self modeling mixture analysis (SMA) was used by utilizing the reference spectra of each component to extract pure component spectra. This was used to identify each component in the sample based on their spectral information divergence (SID) values to the reference spectra at all the concentration levels. Raman spectral image of each concentration level was created using contribution images from SMA. The identified pixels of each component in Raman spectral image was used to develop a relationship with concentration of corresponding component in the sample. The method is useful for simultaneous estimation of ingredients concentration in food powder mixture.