Mass spectrometric evaluation of the soluble species of Shengli lignite using cluster analysis methods. (15th January 2019)

• Record Type: Journal Article
• Authors: Yu, Ya-Ru; Fan, Xing; Chen, Lu; Dong, Xueming; Zhao, Yun-Peng; Li, Bei; Wei, Xian-Yong; Ma, Feng-Yun; Nulahong, Aisha
• Is Part Of: Fuel. Volume 236(2019)
• Page Start: 1037
• Page End: 1042
• Publication Date: 2019-01-15
• Subjects: Coal -- Expectation maximum algorithm -- Extraction -- Hierarchical cluster analysis -- R language -- Thermal dissolution; Fuel -- Periodicals; Coal -- Periodicals; Coal; Fuel; Periodicals; 662.6
• DOI: 10.1016/j.fuel.2018.09.063
• Languages: English
• ISSNs: 0016-2361

Abstract

Highlights: HCA and EMGM are introduced to reach in-depth statistic results for coal extracts. The similarities and differences of clusters are visually exhibited in a plot via HCA. Molecular structures in the same EMGM model have a certain similarity. Abstract: A coal was extracted and thermally dissolved with cyclohexane, acetone and methanol sequentially. The ultrasonic extracts and thermal dissolution (TD) products were analyzed using an Orbitrap mass spectrometer (MS) with an atmospheric pressure chemical ionization source in positive ion mode. Large amounts of MS data were obtained but there were challenges in obtaining meaningful information from the data. The purpose of cluster analysis is to reduce complex multivariate data into meaningful groups. Two cluster analysis methods, hierarchical cluster analysis and expectation maximum algorithm based on Gaussian mixture model (EMGM), run by R language were introduced to obtain in-depth statistical results for compounds in both extracts and TD products. Seven types of heteroatomic compounds (O1, O2, N1, S1, O1 N1, O1 S1 and N1 S1 ) in the extracts under the condition of ultrasonic extraction with cyclohexane were clustered with EMGM and possible structures of the related models can be inferred by analyzing the relationship between carbon number and double bond equivalent. The application of cluster analysis will provide methodological guidance in studying the structure of coal molecules.