A Research of CO2 Adsorption on MOFs by AI
A artificial Intelligence (AI) model combined a back-propagation neural network (BPNN) with a genetic algorithm (GA) based on experimental data as training samples was established to predict the CO2 adsorption capacity for metal organic frameworks (MOFs) of Ni/DOBDC.
The random function of the conventional BPNN model was modified by the GA−BPNN model for optimizing the initial weights and bias nodes. The amounts of adsorbed CO2 and corresponding isosteric heat of adsorption on Ni/DOBDC were synchronously studied within a wide temperature range (25−145 °C) and pressure range (0−3.5 Mpa).
Back-propagation Neural Network
This research used the thermal analyzer from SETARAM, which consists of four parts, namely, adsorption, calorimetric, gas supply, and data acquisition systems. A Sievert’s method based gas adsorption analyzer (SETARAM, PCTPRO) equipped with a 3D Calvet scanning calorimeter (SETARAM, C80) served as the adsorption and calorimetric systems.
The sample was pretreated to obtain high-purity Ni/DOBDC powder. The Ni/DOBDC powder was heated at 200 °C for 24 hours under vacuum using the Calvet calorimeter to remove water and CO2 absorbed by Ni/DOBDC from the air. The gas supply system, which contained three high-purity gas supply lines, namely, CO2, N2, and He, was connected to the adsorption system. Nitrogen was used to drive pneumatic valves in the Sievert PCTPro. Helium was flowed through the 3D Calvet calorimeter sample cell to calibrate the volume. CO2 (as the adsorbate gas) was fed to the sample cell to achieve the setting pressure.
The target temperature was controlled using the Calvet calorimeter. During the adsorption, the amount and heat of adsorption were synchronously determined using the Sievert PCTPro − Calvet calorimeter system. The test signals were recorded using data acquisition systems. The entire system configuration is explained in the below figure.
The Test System Configuration
For more details about this research, please refer to the original paper, “Prediction and Experimental Verification of CO2 Adsorption on Ni/DOBDC Using a Genetic Algorithm–Back-Propagation Neural Network Model”. https://pubs.acs.org/doi/abs/10.1021/ie404396p Please contact to ACTTR Technology, we will introduce the instruments as well as the consultant for building up your thermal analysis laboratory.