Adsorption Simulation And Prediction Of Amorphous Porous Nanomaterials

Carbon tetrafluoride（CF₄）is the most potent and longest lasting greenhouse gas emitted by human activities,and the capture and separation of CF₄ and its nitrogen-containing mixture is a challenging process.To solve this problem,this paper selects amorphous carbon materials as membrane materials to adsorb and separate these gases to mitigate the greenhouse effect,and the main solutions of this paper are as follows.（i）Firstly,this paper investigates the adsorption properties of CF₄and its nitrogen-containing gas mixture on 108 porous rigid amorphous carbon materials at 298 K and 10 bar by using advanced techniques such as Monte Carlo molecular simulation with giant regular system synthesis,and this paper simulates and calculates the adsorption isotherms of single components,from which the optimal adsorption materials for CF₄ and its nitrogen-containing gas mixture are screened for adsorption isotherm analysis,and The isenthalpic heat of adsorption of the optimal adsorption material was calculated to explain the adsorption phenomena of the optimal material.（ii）Then,the separation characteristics of the optimal adsorbent material for CF₄ were simulated.First,the availability of the Mu Si C program was verified using IAST,and then the competing adsorption properties of CF₄ and N₂ at two different molar fractions（0.5 and 0.5）were calculated by using the Mu Si C program in this paper to obtain the selected adsorption and mixed adsorption capacities of the optimal adsorbent material for CF₄.To understand the adsorption selectivity of CF₄ and its nitrogen-containing gas mixture at different mixing ratios in more detail,the adsorption selectivity properties of CF₄ and N₂ at mixing ratios of 20:80 and 80:20,respectively,were also calculated in this paper.（iii）In this paper,the above CF₄ adsorption data on 108 materials were then randomly divided into a training set and a prediction set of the BP neural network model.The trained BP neural network model was used to predict the adsorption materials in the test set and compared with the adsorption data in the test set.Meanwhile,based on the structure of the original BP model,Genetic Algorithm（GA）is introduced in this paper to build the GA-BP neural network model.The GA-BP neural network model was trained,and the adsorption data were predicted using the same training and prediction sets,respectively,and the differences in the prediction results between the BP neural network model and the GA-BP neural network model were compared.