Pyrolysis of biological wastes for bioenergy production: Thermo-kinetic studies with machine-learning method and Py-GC/MS analysis

This present study aimed to evaluate the bioenergy potential of abundant biological waste generated in the process of producing bio-product via pyrolysis for the first time. These bio-wastes were selected from the effluents in the ethanol, butanol, acetic acid and lactic acid production plants which mostly included bacterial biomass (BB). We used FTIR spectroscopy for the simultaneous, qualitative analysis of bio-macromolecular components and specific functional groups in BB such as proteins (1700-1500 cm(-1)), lipids (3000-2800 cm(-1)), and carbohydrates (1200-900 cm(-1)). The pyrolytic behavior of the BB has been studied using thermogravimetric analysis (TGA). The thermo-kinetic studies were conducted using iso-conversional methods such as Kissinger-Akahira-Sunose (KAS) and Ozawa-Flynn-Wall (OFW). The activation energy (average: 85.22-175.93 kJ/mol), Gibb's free energy (average: 171.76-175.56 kJ/mol), and the calculated high heating value (HHV) ranged from 16.98 +/- 0.02 to 17.37 +/- 0.02 MJ/kg, showed the remarkable bioenergy potential of the selected feedstock. Lower difference between the activation energy and enthalpy change of reaction (similar to 5 kJ/mol) declared that the product formation is favorable during pyrolysis. Principal component analysis (PCA) was applied as a statistical method to reduce the dimensionality of differential thermogravimetric (DTG) data and to identify the principal reactions. Pyrolysis-Gas Chromatography/Mass Spectroscopy (Py-GC/MS) at 500 degrees C, was further employed to characterize the potential chemical products, indicating the presence of a range of aromatic, aliphatic, and nitrogen-containing compounds. The comparative study of the results, predicted from support vector regression (SVR-based) model with the experimental data, showed satisfactory agreements (R-2 > 0.9999) and accurate results.