Determining the Age of Spoiled Milk from Dried Films Using Attenuated Reflection Fourier Transform Infrared (ATR FT-IR) Spectroscopy. Issue 9 (September 2019)

• Record Type: Journal Article
• Authors: Richardson, Zack; Perez-Guaita, David; Kochan, Kamila; Wood, Bayden R.
• Is Part Of: Applied spectroscopy. Volume 73:Issue 9(2019)
• Page Start: 1041
• Page End: 1050
• Publication Date: 2019-09
• Subjects: Milk -- Fourier transform infrared spectroscopy -- FT-IR -- attenuated total reflection -- ATR -- partial least squares regression -- PLSR -- bacteria -- milk spoilage; Spectrum analysis -- Periodicals; 543.505
• DOI: 10.1177/0003702819842548
• Languages: English
• ISSNs: 0003-7028

Abstract

Milk spoilage is an inevitable occurrence, which generates waste and can result in food poisoning. When milk spoils, the off-flavor and curdling are due to excessive proliferation of various bacteria which causes pH changes. Time, temperature, environment, and previous handling practice all affect the spoilage rate. There is a need for a fast reliable and accurate method that can identify in situ early spoilage of milk. Here we show the ability of attenuated total reflection Fourier transformed infrared spectroscopy (ATR FT-IR) in conjunction with multivariate data analysis to predict the age of milk. We found that dried films vastly increased the absorbance of important biomolecules within milk such as lipids, proteins, and sugars, compared to an unchanged milk sample. This allowed us to note the minor discrepancies that happened in spoilage. Spoilt milk was characterized by bands associated with increased lipids, proteins, and lactic acid and a decrease in carbohydrates. A semi-quantitative prediction model for milk spoilage at room temperature demonstrated ATR FT-IR spectroscopy can predict milk age with a root mean square error of prediction of approximately 14 h. The model showed poor performance in the first 40 h but the predictions improved significantly after this time. The experimental procedure proposed for detecting biomolecules within milk has the potential to improve common practice. Furthermore, the model would be a starting point for newer and improved methods to predict the spoilage date of milk, with potential commercial uses to reduce food waste and costs to the milk industry.