Search

• Functional groups are effective variables of machine learning for FTIR spectra. • OH is a key factor discriminating the samples because of deoxidation reactions. • Pre-integration of initial samples increases the classification effects of PCA. Three biomass (rice straw, rice husk and wheat straw) and three coal samples (lignite, subbituminous coal and bituminous coal) were extracted via a thermal dissolution (TD) process, and the TD extracts were treated by catalytic hydrogenation to obtain reaction products. To reveal the similarity in molecular information among the samples and elucidate their chemical reactivity, four machine learning algorithms were applied to analyze the Fourier transform infrared spectra of both TD extracts and catalytic hydrogenation products. Functional groups were used as variables and the difference in peak area can be treated as the basis for sample classification. Aromatic C H, C O C, aliphatic CH 2 or CH 3 and aromatic C O or C C bonds were the main characteristic variables in principal component analysis algorithm to classify biomass- and coal-derivated samples. These samples were also grouped into four clusters by hierarchical clustering analysis algorithm according to the similarity and difference in the distribution of functional groups. For artificial neural network algorithm, aliphatic C H and OH bonds are the most important variables to classify these samples into four groups, and aromatic C H, OH, and C O C groups are the main variables contributed to the classification trees in random forest algorithm. Machine learning algorithms will provide methodological guidance for the data mining of the spectra of complex organic systems. [ABSTRACT FROM AUTHOR]

Directly converting lignite into arenes as commodity chemicals and drop-in fuels is a highly desirable target for researchers. However, this is severely limited by the presence of stable inter unit C-O linkages in lignite. Herein, a highly active catalyst, Co-MoS 2 , has been designed and prepared to be used for the selective catalytic cleavage of inter unit C-O bridged bonds (BBs) in the conversion of Naomaohu (NMH), a lignite, to yield arenes. Both catalytic hydrogenation conversion (CHC) and thermal dissolution (a non-catalytic process, named NCHC) of NMH were conducted using n-hexane as the solvent. The soluble fractions from the first stage of NCHC (SF NCHC-1) and CHC (SF CHC-1), as well as the second stage of NCHC (SF NCHC-2) and CHC (SF CHC-2) were obtained using the same solvent. After the first-stage catalytic treatment, the relative contents of arenes and alkanes in SF CHC-1 increased by 16.7 % and 9.3 %, respectively, compared to those in SF NCHC-1 , and the relative content of oxygen-containing compounds (OCCs), such as alcohols, phenols, esters, ketones, acids, and furans, decreased by 13.2 %. For the second-stage conversion, the relative contents of arenes and alkanes in SF CHC-2 were 53.5 % and 15.5 % higher than those in SF NCHC-2 , and the relative contents of OCCs was only 6.6 % in SF CHC-2. The catalytic mechanisms of cleavage of C-O BBs and deoxygenation were proposed based on the CHC of model compounds and density functional theory (DFT) calculation. The adsorption of OCCs and the activation of released hydrogen radical (H·) by Co-MoS 2 , followed by H· addition to the carbon atom connected to oxygen in C-O BBs and oxygen-containing functional groups, are crucial for the cleavage of C-O BBs and formation of deoxygenated compounds, particularly arenes. • Co@MoS 2 selectively cleave C-O bridged bonds in lignite. • Co@MoS 2 is effective for hydrodeoxygenation and generation of arenes. • C ar -O cleavage mechanism was revealed by catalytic model compounds and DFT calculation. [ABSTRACT FROM AUTHOR]