Pyrolysis characteristics of metal-organic frameworks: TG-FTIR, TG-GC-MS, kinetic, thermodynamic and artificial neural network modelling

Abstract

Pyrolysis is one of the most promising methods for transforming metal-organic frameworks (MOFs) as precursors into metal-based products with tailored properties by decomposing their organic linkers into smaller molecules and carbonaceous. This research aims to explore the fundamental pyrolysis behavior of MOFs for potential upscaling. The experiments were performed on the laboratory-prepared nickel MOFs doped on magnetic graphene oxide composite (NiMOF@MGO) and its elemental and proximate analysis were investigated. The main active pyrolysis zones of MOFs were identified by thermogravimetric (TG) anlyzer, while TG-Fourier Transform Infrared (FTIR) spectroscopy and TG-gas chromatography-mass spectrometry (TG-GC-MS) were used to characterize the composition of the vapors emitted from each zone. The kinetic and thermodynamic parameters of the pyrolyzed MOFs were determined using different linear methods (Kissinger-Akihira-Sonoze, Flynn-Wall-Ozawa, and Friedman) and nonlinear Vyazovkin methods to determine the activation energy (Ea) required to terminate the reaction. In addition, an artificial neural network (ANN) algorithm was constructed and trained to predict the thermal degradation of MOFs under ambiguous heating rate parameters. Physical analysis showed that MOFs contain high content of volatile matter (72.04 wt%) and carbon element (21.30 wt%). The TG results showed three active pyrolysis zones (at 160–178 °C, 290–306 °C, 410–433 °C) with an overall weight loss of 73 wt%. Based on TG-FTIR analysis, CO bond (carbonyl) was the main group of the vapor emitted from the first active zone, while asymmetric stretching vibration of CO2was the main group of the other zones. While GC-MS showed that ethyl cyanoacetate (reactive core of linkers) was the dominant compound in all active regions with a abundance ranging from 60.94 % to 70.31 % at 5 °C/min. The kinetic analysis showed that MOFs linkers consume low Ea for decomposition in the range of 164–202 kJ/mol (Ea), while the thermodynamic parameters were estimated in the ranges of 158755–196,288 J/mol.K (enthalpy), −11,059 to −18,184 J/mol.K (Gibbs free energy), and 232–287 J/mol.K (entropy). The constructed ANN algorithm also showed high performance in predicting the degradation of MOFs with R ≥ 0.999 and MSE in the range of 0.0427–0.0683. Accordingly, pyrolysis can be considered a promising technology not only for producing metal-based products from MOFs, but also for valorisation of their organic linkers into value-added chemicals.