A parameter selection for Raman spectroscopy-based detection of chemical contaminants in food powders

Authors: DHAKAL, SAGAR - Forest Service (FS); Chao, Kuanglin - Kevin Chao; Qin, Jianwei - Tony Qin; Kim, Moon; Schmidt, Walter; Chan, Diane

Submitted to: Transactions of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/17/2015
Publication Date: 4/20/2016
Citation: Dhakal, S., Chao, K., Qin, J., Kim, M.S., Schmidt, W.F., Chan, D.E. 2016. A parameter selection for Raman spectroscopy-based detection of chemical contaminants in food powders. Transactions of the ASABE. 59(2):751-763.

Interpretive Summary: Recent reports of human illnesses resulting from the economically motivated addition of chemical contaminants have focused attention on this emerging food safety issue. Notable instances include the deliberate adulteration of milk products with melamine and the use of excessive maleic anhydride in starch production. Raman spectroscopy has been demonstrated to be an effective technique for qualitative detection of contaminants in dry food powders, while Raman imaging-based detection methods have the potential to screen larger volumes of food powders for quantitative contaminant detection. The depth of laser penetration into tapioca starch and wheat flour, the power of the laser, and the spatial resolution needed for effective quantitative imaging-based Raman detection of contaminant particles were investigated in this study. Laser depth and power were evaluated by testing the detection of melamine through layers of the food powders, while spatial resolution was tested for detection of benzoyl peroxide in flour and maleic acid in starch. Using the selected experimental parameters, a linear correlation between the detected numbers of contaminant pixels and the actual contaminant concentrations in the mixed powder samples was demonstrated, showing that the method developed in this study can be used for quantitative detection of chemical contaminants present in food powders at low concentrations. This Raman imaging-based technique shows promise for quantitative analysis of food ingredients that will benefit food processors and regulatory agencies.

Technical Abstract: Raman spectroscopy technique has proven to be a reliable method for detection of chemical contaminants in food ingredients and products. To detect each contaminant particle in a food sample, it is important to determine the effective depth of penetration of laser through the food sample and the corresponding laser intensity required for the penetration. It is also important to determine effective spatial resolution needed to detect each contaminant particle in the mixture. This study examined the depth of penetration of a 785nm laser through tapioca starch and wheat flour when using three different laser intensities. Melamine, known to exhibit identifiable Raman spectral peaks, was selected as a subsurface reference material for determining the depth of laser penetration through the food powder samples. The food powders were layered in 5 depths between 1 and 5 mm overtop a Petri dish packed with melamine. It was observed that the 785nm laser could penetrate to 3mm depth in starch; however, the penetration depth was limited to 2mm in flour. These depths were achieved using laser intensity of 200mw and 100mw for starch and flour, respectively. The selected depth and laser intensity parameters were next used to examine the effective spatial resolution required for detection of maleic acid in starch and benzoyl peroxide in flour, which was selected to be 0.5 mm. Finally, an experiment was conducted to demonstrate the use of these parameters for quantitative Raman imaging-based detection of these contaminants prepared in mixtures at 0.1%, 0.3%, and 0.5% (w/w) concentrations.