Font Size: a A A

Simulation And Experimental Study On The Adsorption Performance Of Porous Materials For Airborne Molecular Contaminants Ammonia

Posted on:2023-05-27Degree:DoctorType:Dissertation
Country:ChinaCandidate:R Q ChenFull Text:PDF
GTID:1521307319993439Subject:Civil engineering
Abstract/Summary:PDF Full Text Request
With the continuous expansion of semiconductor scale,the product yield has aroused people’s attention.Airborne Molecular Contaminants(AMCs)can reduce the yield of products.AMCs such as ammonia and sulfur dioxide(SO2)will enter the semiconductor cleanroom along with the air conditioning system.Chemical filters containing porous adsorbents is used to purify AMCs to meet the requirements.In addition,ammonia pollution is becoming more and more serious,so it is very important to study ammonia adsorption performance and mechanism on porous adsorbents as well as the competitive adsorption mechanism with SO2 and water.In this study,four kinds of common activated carbons and three kinds of metal-organic frameworks(MOFs)were selected.The physicochemical parameter of the adsorbents were characterized.Ammonia,SO2 and water adsorption performance on adsorbents were studied by static volume method.The purification mechanism of ammonia was analyzed by thermogravimetric-mass spectrometry.The adsorption heat of activated carbon and MOFs was calculated by Clausius-Clapperron equation and density functional theory,respectively.The mechanism of ammonia purification and the competitive adsorption were revealed.The attention-based recurrent neural network(RNN)model is used to predict the adsorption isotherms using the characterization parameters of the absorbents as input.M-2 has the best ammonia adsorption performance at low pressure((10-6-5×10-4bar:1-500 ppm),which is about 6 orders of magnitude lower than the existing data.The maximum adsorption heat(low partial pressure)of ammonia and SO2 on M-2 is 89.80 k J/mol(chemisorption)and 35.20 k J/mol(physical absorption),respectively.Water adsorption heat is always close to 45 k J/mol(condensation heat of water),indicating that it is mainly water-water interaction.It shows that SO2 may not be able to replace the adsorbed ammonia,but water can cause ammonia/SO2 to be unable to contact the adsorption site in the pore,thus reducing the adsorption capacity.Interaction energies of three gases on MOFs:(1)HKUST-1:ammonia(-147.890 k J/mol)>SO2(O)(-78.890 k J/mol)>SO2(S)(-74.790 k J/mol)>water(-64.070 k J/mol);(2)MIL-100(Fe):ammonia(-80.280 k J/mol)>SO2(O)(-64.910 k J/mol)>water(-46.940 k J/mol)>SO2(S)(-40.920 k J/mol);(3)UIO-66:SO2(O)(-71.670 k J/mol)>SO2(S)(-65.720k J/mol)>ammonia(-47.080 k J/mol)>water(-30.000 k J/mol).Therefore,ammonia on HKUST-1/MIL-100(Fe)is likely to replace pre-adsorbed SO2/water,while ammonia on UIO-66 is difficult to be replaced by water but may be replaced by SO2.The adsorption mechanism and competitive adsorption mechanism at low pressure are revealed from the perspective of thermodynamics,which provides reference in the future.The ammonia and water RMSE of the attention-based RNN model were 0.006-0.089 and 0.145-0.686,respectively.The low RMSE value proves the applicability of the attention-based RNN model.It contributes to the rapid evaluation of adsorbent adsorption performance.
Keywords/Search Tags:Ammonia, Adsorption mechanism, Competitive adsorption, Molecular simulation, Machine learning
PDF Full Text Request
Related items