ABSTRACT

Ozone can react with fatty species such as vegetable oils to form oxidized compounds with interesting antibacterial activity. The iodine and peroxide indexes of these ozonated products are two crucial antibacterial characteristics. Currently, these two parameters are determined by chemical methods, which is disadvantageous because it requires long handling time, sample destruction and the use of toxic products. Spectroscopic analyses coupled with chemometric evaluations have already shown their usefulness for characterizing virgin oils via rapid, chemical-free methods. In addition, infrared spectroscopy is widely used to characterize ozonated compounds. The present study evaluates the possibility of coupling infrared spectroscopy with chemical analyses of ozonated compounds. We subjected 187 fat samples of varying compositions to different ozonation conditions to determine their chemical parameters. We analyzed the samples by infrared spectroscopy and modeled the results by combining chemical data with spectral analysis. The model results are quantified by the coefficient of determination R², the root-mean-square error, and the residual predictive deviation (RPD). The model produces RPDs greater than 3.5 for the mid-infrared spectral range, which means that they may be used to estimate indexes. In addition, with RPDs between 2 and 3, partial least squares near-infrared models demonstrate a capacity for rough screening of ozonated fats.

摘要

臭氧可以与植物油等脂肪物质发生反应，形成具有有趣抗菌活性的氧化化合物。这些臭氧化产品的碘和过氧化物指数是两个重要的抗菌特性。目前，这两个参数是通过化学方法确定的，这是不利的，因为它需要较长的处理时间、样品破坏和有毒产品的使用。光谱分析与化学计量学评估相结合，已经表明它们可用于通过快速、无化学物质的方法表征初榨油。此外，红外光谱广泛用于表征臭氧化的化合物。本研究评估了将红外光谱与臭氧化化合物的化学分析相结合的可能性。我们将 187 个不同成分的脂肪样品置于不同的臭氧化条件下，以确定它们的化学参数。我们通过红外光谱分析样品，并通过将化学数据与光谱分析相结合来模拟结果。模型结果由决定系数 R 量化²、均方根误差和残差预测偏差（RPD）。该模型产生的中红外光谱范围的 RPD 大于 3.5，这意味着它们可用于估计指数。此外，RPD 介于 2 和 3 之间，偏最小二乘法近红外模型展示了对臭氧化脂肪进行粗略筛选的能力。