Engine oils are produced with a blend of almost 80% (w/w) base oils and 20% (w/w) of different additives. This study investigates, for the first time, the capabilities of NIR (Near Infrared) and EEM (Emission-Extraction Matrix) fluorescence spectroscopies coupled with chemometrics as low-cost, green and non-destructive methods in identifying the type of base stock into engine oil. In order to reach this goal, base stocks of different American Petroleum Institute (API) groups were analysed without any pre-treatment. PCA (Principal component analysis) performed on NIR and unfolded EEM spectra showed that samples form clusters according to their API groups and to chemical composition. PARAFAC (Parallel Factor Analysis) was also applied on 3-way fluorescence data and outcomes were consistent with PCA results. PLS-DA (Partial Least Squares Discriminant Analysis) was able to classify the base stock samples according to the API groups and satisfactory results were achieved: the correct prediction abilities on an external test set using NIR and EEM fluorescence spectroscopies were 87% and 85%, respectively. In addition, the determination of the base oil group at different gasoline engine oil performance levels was used as a method to evaluate the efficiency of the lubricants. Both spectroscopic methods appear to be fast and non-destructive to characterize the base stocks in analysing pure base stocks and engine oils with different performance levels.

中文翻译：

---

通过近红外和荧光光谱结合化学计量学鉴定发动机油中的基础油

发动机油由几乎 80% (w/w) 的基础油和 20% (w/w) 的不同添加剂混合而成。本研究首次调查了 NIR（近红外）和 EEM（发射-提取矩阵）荧光光谱结合化学计量学作为识别发动机基础油类型的低成本、绿色和非破坏性方法的能力油。为了达到这个目标，在没有任何预处理的情况下分析了不同美国石油协会 (API) 组的基础油。对 NIR 和展开 EEM 光谱进行的 PCA（主成分分析）表明，样品根据其 API 基团和化学成分形成簇。PARAFAC（平行因子分析）也应用于三向荧光数据，结果与 PCA 结果一致。PLS-DA（偏最小二乘判别分析）能够根据 API 组对基础油样品进行分类，并取得了令人满意的结果：使用 NIR 和 EEM 荧光光谱在外部测试集上的正确预测能力分别为 87% 和 85% ， 分别。此外，在不同的汽油发动机油性能水平下测定基础油组被用作评估润滑剂效率的方法。在分析具有不同性能水平的纯基础油和发动机油时，两种光谱方法似乎都可以快速且无损地表征基础油。使用 NIR 和 EEM 荧光光谱对外部测试集的正确预测能力分别为 87% 和 85%。此外，通过测定不同汽油发动机油性能水平下的基础油组，作为评估润滑油效率的方法。在分析具有不同性能水平的纯基础油和发动机油时，两种光谱方法似乎都可以快速且无损地表征基础油。使用 NIR 和 EEM 荧光光谱对外部测试集的正确预测能力分别为 87% 和 85%。此外，在不同的汽油发动机油性能水平下测定基础油组被用作评估润滑剂效率的方法。在分析具有不同性能水平的纯基础油和发动机油时，两种光谱方法似乎都可以快速且无损地表征基础油。