High Resolution Mass Spectrometry and Principal Component Analysis for an Exhaustive Understanding of Acidic Species Composition in Vacuum Gas Oil Samples

Abstract

This article illustrates the study of the Ox-class composition of acidic species in different vacuum gas oil (VGO) samples by using electrospray ionization high resolution mass spectrometry (ESI HRMS) and principle component analysis (PCA). The conventional total acid number (TAN) is determined by titration of a strong acid species with milligrams of KOH. The inability of TAN to explain the total acidity is verified by spiking studies of model acidic compounds in VGO samples by using autotitrator and UV–Vis method. Furthermore, HRMS investigations were carried out on 20 VGO samples with TANs ranging from 0.39 to 8.90 mg of KOH/g. The HRMS results could not explain the trends for the variations in O1, O2, O3, and O4 species with respect to TAN. The complexity of the trends in the above Ox classes and the intragroup variations of each Ox class prompted the use of unsupervised PCA models for the combined O2+O4 (i.e., mono- and di-carboxylic acids) and O1+O3 (i.e., phenolic and hydroxycarboxylic acids) species. The first principal components (PC1) of the PCAs of O1+O3 and O2+O4 cover 83.47 and 73.48% of the total variance, respectively. In PCA1, the projection of the data with the greatest variance of the O2 class is along the first axis, explaining the observed TAN. The PCA1 had limitations, explaining the low-TAN samples, where the O2 variance was merely fluctuating, especially for lower TANs that were close in value to each other. This led us to determine the intragroup compositional variances of O1 and O3 by conducting PCA2. PCA2 basically explained the remaining weak acid species that were neglected by TAN but served as a contributing factor to corrosion.

Cloud Computing Storage Data Access Control Method Based on Dynamic Re-Encryption

“…e original cloud computing storage data are standardized and input into the principal component expression to obtain the principal component scores of different data. According to the calculation results, the data with high scores can be retained to realize the dimensionality reduction processing of high-dimensional cloud computing storage data [15,16].…”