The objective of this work was to compare reference methodologies to Fourier transform infrared spectroscopy for analysis of pH and freezing point (FP) under different storage conditions on samples of raw bovine milk. Milk samples were collected from bulk tank milk on 57 farms. Each sample was subdivided into 45 vials (40 mL) which were distributed according to storage temperature (-20ºC, 7ºC and 25ºC), age of the sample (0, 3, 6 and 9 days) and levels of water addition (0, 2, 4 and 6%). Bronopol was added into 44 flasks and one was maintained without addition of preservative for control. Frozen point and pH were determined by the standard methods (thermistor cryoscope and potentiometer, respectively) and by the alternative methodology by using MilkoScanTM FT+. The results from each method, at different conditions of conservation of the samples, were evaluated by analysis of variance and comparison of means. The linear regression analysis was performed to evaluate the methodologies (reference and alternative) for results of the frozen point in function of the age of the sample. Bronopol did not alter pH means, but it reduced frozen point. Correction factors were calculated to eliminate this effect on the results of the frozen point. Means of pH in three-day samples, kept at -20ºC and 25ºC were lower than the samples at 7ºC. The age of the samples affected the results of pH when they were submitted to 25ºC. The increase of storage temperature to 25ºC reduced means of frozen point in samples without addition of water and at nine days of storage. The increase in age of sample at -20ºC and 7ºC does not affect means of frozen point. There is a strong correlation among methodologies in function of age of the sample for frozen point.
bulk tank milk; freezing point; milk acidity
