Comprehensive quality evaluation of anti-wear properties and oxidation stability of different vegetable oils based on QSPR-factor analysis

• Published in: Fuel (Guildford) 2024-02, Vol.358, p.130120, Article 130120
• Main Authors: Zhang, Rongrong, Yang, Sicheng, Liu, Ting, Zhang, Yaoyun, Yi, Chenglingzi, Jia, Dan, Liu, Jianfang
• Format: Article
• Language: English
• Subjects: anti-wear property; factor analysis; oxidation stability; QSPR; quality evaluation
• ISSN: 0016-2361; 1873-7153
• DOI: 10.1016/j.fuel.2023.130120

•(1) Based on 119 groups of fatty acid structure information and principal component regression method, a model containing 9 descriptors was construct to predict the anti-wear properties of fatty acids.•(2) According to the model analysis, the main factors affecting the anti-wear properties of fatty acids were the carbon chain length and double bonds number.•(3) The WSDs of common fatty acids in vegetable oils and different vegetable oils were estimated.•(4) Unsaturated fatty acid content (UFA%), polyunsaturated fatty acid content (PUFA%), saturated fatty acid content (SFA%) and U/S (UFA%/ SFA%) values were selected as evaluation indexes for oxidation stability. WSD was a reflection index of anti-wear properties. Finally, factor score and cluster analysis were used to evaluate the quality of different vegetable oils. With the increasing awareness of human environmental protection, it is urgent to find potential bio-lubricants. Vegetable oil has attracted the interest of scientists because of its good lubricity and biodegradability. The external properties of vegetable oils are closely related to their internal chemical composition. Fatty acids are the main chemical components in vegetable oils. The industrial capacity of fatty acids is accurately predicted and evaluated by theory, which is very helpful for the auxiliary design of green lubricants. Combining 9 meaningful molecular descriptors and principal component regression, a calculation scheme for predicting the anti-wear properties of fatty acids and vegetable oils was proposed. The training set and validation set of the model proved that the model had excellent predictive performance. Then, on this basis, through literature retrieval and factor score study, the research group successfully screened out a class of vegetable oils with the broadest industrial application potential, including yuzu, sesame, olive and so on. This project aims to provide new ideas about the design of green bio-based lubricants and to promote the development of vegetable oils in the industry.