Preliminary Discrimination of Cheese Adulteration by FT-IR Spectroscopy Lucian CUIBUS 1 , Ruben MAGGIO 2 , Vlad MUREȘAN 3 , Zoriţa DIACONEASA 1 , Florinela FETEA 1 , Carmen SOCACIU 1 1) Department of Food Science, Faculty of Food Science and Technology, University of Agricultural Sci- ence and Veterinary Medicine Cluj-Napoca, Romania; 2) Department of Organic Chemistry, School of Pharmaceutical and Biochemical Sciences, National Uni- versity of Rosario and Institute of Chemistry of Rosario (IQUIR, CONICET-UNR), Suipacha 531, Rosario S2002LRK, Argentina; 3) Department of Food Engineering, Faculty of Food Science and Technology, University of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, Romania; * Corresponding author e-mail: lucian.cuibus@usamvcluj.ro Bulletin UASVM Food Science and Technology 71(2) / 2014 ISSN-L 2344-2344; Print ISSN 2344-2344; Electronic ISSN 2344-5300 DOI: 10.15835/buasvmcn-fst:10795 Abstract The present work describes a preliminary study to compare some traditional Romanian cheeses and adulterated cheeses using Attenuated Total Reflectance-Fourier transform infrared spectroscopy (ATR-FTIR). For PLS model calibration (6 concentration levels) and validation (5 concentration levels) sets were prepared from commercial Dalia Cheese from different manufacturers by spiking it with palm oil at concentrations ranging 2-50 % and 5-40 %, respectively. Fifteen Dalia Cheese were evaluated as external set. The spectra of each sample, after homogenization, were acquired in triplicate using a FTIR Shimatsu Prestige 21 Spectrophotometer, with a horizontal diamond ATR accessory in the MIR region 4000-600 cm -1 . Statistical methods as PLS were applied using MVC1 routines written for Matlab R2010a. As first step the optimal condition for PLS model were obtained using cross-validation on the Calibration set. Spectral region in 3873-652 cm -1 , and 3 PLS-factors were stated as the best conditions and showed an R 2 value of 0.9338 and a relative error in the calibration of 17.2%. Then validation set was evaluated, obtaining good recovery rates (108%) and acceptable dispersion of the data (20%). The curve of actual vs. predicted values shows slope near to 1 and origin close to 0, with an R 2 of 0.9695. When the external sample set was evaluated, samples F19, F21, F22 and F24, showed detectable levels of palm fats. The results proved that FTIR-PLS is a reliable non-destructive technique for a rapid quantification the level of adulteration in cheese. The spectroscopic methods could assist the quality control authority, traders and the producers to discriminate the adulterated cheeses with palm oil. Keywords: ATR FT-IR spectroscopy, cheese adulteration, cheese quality, food authenticity, PLS, cheese, MIR INTRODUCTION A major concern for researchers, food indus- try and consumers is the authenticity of cheese products. The determination of food authenticity and geographic origin becomes now a crucial issue in food quality control and safety (Barile, 2005; Karoui, 2007). Development of new sophisticated techniques for the authentication of food products grows rapidly due to increased consumer aware- ness of food safety and quality issues (Reid et al. 2006; Koca et al. 2010). One of the most important type of cheese for romanian consumers is Dalia. Dalia yellow cheese is an traditional romanian product obtained from whole or normalized cow pasteurized milk, adding selected microorganisms cultures and curd clotting.