Theoretical Computation On Adsorption Capacity Of Simple Gas Molecule For Wooden Activated Carbon

In recent years,environmental problems induced by volatile organic components(VOCs)are abstracting more and more attention world widely.With large surface area and adsorption capacity,activated carbon is one of gas adsorbents playing a very important role in dealing with those problems mentioned above.As for the adsorption performance of activated carbon primarily determined by its porous character,the characterization and analysis on the relationship between porous structure in activated carbon and its adsorption performance becomes the main fundamental problem of activated carbon and other amorphous materials.With the experimental adsorption isotherm,the pore size distribution(PSD)model constructed by solving adsorption integral equation(AIE)numerically is significant for traditional specific activated carbon screening.This work primarily focus on the application of modified density functional theories to improve the efficience of numerical procedure for the effective pore size distribution of activated carbon.With above results as the basis,it is to make computation on working capacity with multi-model,which is meaningful for the theoretical background for reducing the screening expenditure.There are following research results in this work.For screening substantial activated carbon from differnet resources,it is efficient to organize and utilize the characterization with data mining procedure.The neural network model is regressed from machine learning on the characterization of activated carbon through the gas adsorption method.It is to calculate the n-butane working capacity(BWC)of activated carbon through such classification model with the average deviation from experiment data nearly 6.78 % in thirty repeats.It shows that such model is reliable to combine the structural feasure and BWC of activated carbon.The reliablity of gas adsorption capacity of slit pores in theoretical pore size analysis is compared in backward error analysis.With five methods(MFA,FMSA,WDA(Yu),WDA(Liu)and GCMC),it is to calculate the local excess adsorption of methane and carbon dioxide in substantial slit pores.Compared with traditional theoretical method MFA,the theoretical error bound scale(TEBS)of WDA(Yu)is remarkably reduced,which means WDA(Yu)is more suitable for porous analysis of activated carbon.For more representative characterization,the modified density functional theory WDA(Yu)is adopted for porous analysis of activated carbon.The accuracy and stability of WDA(Yu)are improved dramatically for fitting nitrogen isotherms at 77 K of 64 activated carbon samples within the relative pressure range from 0.1 ~ 0.9,which means WDA(Yu)is meaningful for the required pore size distribution of activated carbon.Moreover,there are six numerical algorithms(P1 ~ P6)to study PSD of activated carbon for fitting carbon dioxide isotherm.It is notable that P1 method with B-spline function for the form of PSD is more accurate,with the average fitting deviation of which from experiment data around the range of theoretical error,which means the required pore size distribution is more representative for the porous analysis of activated carbon.The robustness of theoretical porous analysis computation methods are contrast according to the corresponding predictive deviation through the characterization of carbon dioxide adsorption method.With the characterized PSD from fitting CO2 adsorption isotherm,it is to calculate the adsorption capacity of CH4 within the pressure range from 0.2 to 2.0 MPa at 298 K.It is obvious that the required PSD of samples are more representative when adopted theoretical methodology WDA(Yu)for kernel building,solving AIE with P1 after discretization with trapezoid method in N1 hypothesis form on the heterogeneity of solid surface.It is able to make predictive computation on methane adsorption capacity of corresponding activated carbon in various conditions through theoretical models with the characterized PSD.