Use of Chemometrics: Principal Component Analysis (PCA) and Principal Component Regression (PCR) for the Authentication of Orange Juice

ABSTRACT

Principal Component Analysis (PCA) was applied to a set of physico-chemical variables obtained from 41 samples of summer orange juice, in order to reduce the number of variables. Working with the covariance matrix, three components (which explained 98.27% of the variance) were taken. With the correlation matrix, four components which explained: 85.65% of the variance were taken. With the scores corresponding to both matrixes a principal component regression (PCR) was carried out against the dependent variable of Brix grades, so as to obtain two statistical models that would allow the detection of adulterations in pure orange juice, based on dilution and later masking by the addition of sugar. The models were tested with simulated dilutions of 41 samples of juice, to assess the effectiveness of each for the detection of adulterations. Both models turned out to be equally effective, detecting adulterations starting from about 15 % of dilution.

Keywords::

• Chemometrics
• Principal component analysis
• Principal component regression
• Authentication
• Orange juice