Activated carbon is extensively utilized in water and gas treatment due to its adsorption ability,low-cost and environmental-friendly.Construction of large-scale atomistic representations of activated carbon that reflect the real system allow exploration of structure-property relationships and adsorption phenomena from dynamic molecular simulations,which provide new insight into the mechnism of adsorption.However,due to limited data qualifying the distribution of structural features and limitations in construction techniques,the majority of these models were of small-scales(~100’s to 5000’s of atoms)while the construction protocols are either very simplistic or require high computational expense,which led to limited applications to sorption simulations.Herein,an automated construction protocol was applied to generate largescale presentations for tailored lignite-based activated carbon,based on experimentalderived characteristics,with highly controlling over the full range of structural features such as the stack distribution,orientation domains,and pore size distribution.Based on these models,the authors performed molecular simulations on benzene and water sorption,attempting to elucidate the influence of pore size distribution and surface oxidation on sorption behaviors,as well as competed sorption of benzene under different humidity.This approach represented activated carbon in a manner that captures experimentally-determined characteristics.A new semi-automated approach for HRTEM image analysis and a new technique for quantifying distributions were developed in Photoshop and Matlab respectively,which enabled quantitative characterization of the crystalline structure of lignite coal,including distributions of fringe size,interlayer spacing,stacking number,and fringe orientation.The fringe information can then be used,along with Fringe3 D to generate 3D representation of the stacks.While Fringe3 D attempted to duplicate the field of view from the micrograph,the models produced have considerable height and width but are of limited depth(all stacks being centered in the same z coordinate space).To overcome this limitation,a volumetric construction approach(Vol3D)was created with a similar philosophy of enabling control over the distribution of structural features,which populated the specific cubic volume from a collection of PAH stacks and pores.The constructed lignite-based activated carbon models are large-scale(100 x 100 x 100? in volume with ~33,000 atoms),which are the largest model far more reported.This modeling protocol resulted in greater experimental conformity,improved accuracy,and structural diversity.Inclusion of five-sided and seven-sided rings provided better representation of the curvature observed in HRTEM micrographs.This easier,more accurate,and less expensive approach may assist as the starting point for more computationally intense construction strategies that would be more suited to extensive undulating structures,as well as allow more frequent and improved explorations.As a starting point,molecular simulation was performed on the effect of pore size distribution and carbon surface oxidation on water sorption behaviors.It is concluded that sorption capacity of water onto activated carbon model was determined by porosity and surface oxidation.When carbon surface was slightly oxidized(6% carboxyl),there captured double-layer sorption.When carbon surface was largely oxidized(12~18% carboxyl),adsorption of water is initiated on the oxygen-related functional groups followed by growth and coalescence of the clusters.The average distance of water centers of mass was 3.07?~3.19?.Larger pores facilites the formation and growth of clusters,and sorption behavior was regulated by repulsion,dispersion,and hydrogen bonding via water-water and water-site interactions.In microporous models,the minimum pore size that water molecules could fit in was 3.9?.Water molecules were first occupied in pores of 13.7~17.2?,then in pore of 11.7~12.9?.While in micro/mesoporous model,water molecules adsorbed in micropores of 10.4~14.4?,then in mesopores,and micropore filling process dominated.At relatively low partial pressure of benzene(corresponding to 10,331 ppmv)and 303.15 K temperature,per micro/mesoporous model,the most favored sorption energy was-37.45 kJ/mol,with a preferred rotation angle of 20~30°,and a second favored angle of 30~40° relative to the graphene surface.These benzene molecules were aggregated in T-shaped and parallel-replaced configurations,with a separation distance of 5.75? from benzene centers of mass to carbon surface in a monolayer state.The behavior of benzene molecules sorption into the atomistic structure of activated carbon could be summarized in four steps.Relative to subsequent tailoring strategies,we note that if more arch-like structures could be created,their presence would improve sorption capacity for VOCs with configurations like benzene molecules.In comparison,at 318 K,the model simulated that benzene molecules preferred to be adsorbed parallel to—and between—two parallel carbon surfaces.Thus,sites such as these would also be beneficial when crafting tailored activated carbons for VOC removal.Based on previous research,this work further explored the sorption behavior of benzene at several different humidity.When a 10,331 ppmv concentration of benzene was applied,at relative humidity(RH)of 10~30%,the inhibitory effect of water vapor on benzene sorption capacity was slightly.When RH incrased to 50%,sorption capacity of benzene largely decreased,showing that at this state,the inbitory effect was enhanced.At RH=10%,water vapor and benzene created double-layer sorption onto the surface of the oxygen atoms.The first layer is mainly water vapor,which formed water cluster with the average distance of 3.07 ? to oxygen atoms.The second layer is mainly benzene,with the average distance of 5.09? to oxygen atoms and aggregated T-shape.When RH increased to 70%,water vapor sorption mainly occurred in microporous,which resulted in decreased effective pore volume for benzene to occupy,thus a small amount of benzene will move to 7.1~7.9? pores and mesopores.Water vapor exhibited a ≥2.99? separation from the oxygen atom surface and formed a multilayer sorption. |