A Performance Comparison of Two Portable NIRS Technologies for Olive Oil Adulteration

To date, the development of low-cost, easy to use real time and no-invasive sensors for safety and quality controls in agrifood is needed. One of the most promising techniques that complies with these requirements is Near Infrared spectroscopy (NIRS). In this work, two portable NIRS (Nanoscan Texas Instruments, and MicroNIR 1700, Viavi). have been tested as a feasible sensor to detect frauds in olive oils (olive oils adulterated with low-cost seed oils). Spectra of 99 oils were scanned in transmittance (Nanoscan) and reflectance (MicroNIR) modes. After transforming raw data with derivative and scattering math pretreatments, Partial Least Square regression models were developed. Results of Principal Component analysis (PCA) with Nanoscan raw data showed that exist differences between Olive oil and adulterated oils. In both portable NIRS devices, the coefficients of determination for calibration for all adulterations were at least 0.8.