Font Size: a A A

Study On Surface Oxidation Properties Of Palladium Metal Based On Neural Network Potential Function

Posted on:2023-12-06Degree:MasterType:Thesis
Country:ChinaCandidate:M T YangFull Text:PDF
GTID:2531307172979519Subject:Materials and Chemical Engineering (Professional Degree)
Abstract/Summary:PDF Full Text Request
Due to its good stability and high selectivity,palladium is widely used in many important processes such as automobile exhaust neutralizer,electrical contact and coating material.As an important problem in science and industry,metal oxidation interface is not well understood at the atomic level.Palladium will be oxidized when exposed to air,and the oxidation of palladium surface will further affect the activity of catalytic reaction.Therefore,we hope to study the oxidation process of palladium metal and its effect on the catalytic reactivity from the atomic scale,and understand the stable structure of the oxidation process,the reconstruction phenomenon of surface atoms and the reaction mechanism.The microscopic image of the metal oxidation process and the corresponding reaction activation center is very important for the design of metal or metal oxide catalytic materials with high activity and selectivity.In this paper,the machine learning method(SSW-NN)is used to explore the oxidation kinetics of Pd surfaces with different structures(including Pd(110),Pd(100)and Pd(111)).This method combines global neural network(G-NN)potential energy with random Surface walking(SSW)global optimization method to achieve fast global structure search.Based on the SSW-NN method,10~6 structures can be searched globally for O_θ/Pd atomic system,so as to cover all possible structural ranges.On this basis,we further studied the reconstruction process of surface oxygen atoms and the effect of surface oxidation on the catalytic reactivity.The main conclusions are as follows:(1)We simulated and calculated the surface structure of Pd(100)and the corresponding oxidation remodeling process under different oxygen coverage ranges.When oxygen coverage is relatively low,oxygen atoms on the crystal surface of Pd(100)are preferentially adsorbed at the Hollow site.When oxygen coverage is relatively high,oxygen atoms gradually penetrate into the secondary surface layer.A Pd atom on the surface will be lifted by two oxygens.When the coverage is high(θ≥0.5 ML),the surface oxygen atoms begin to penetrate into the subsurface layer,and Pd atoms are tethered to the surface oxygen atoms and the subsurface oxygen atoms.(2)We constructed the oxidation and phase transition system of Pd(110)and Pd(111)crystal surface.The results showed that the surface structure of Pd(110)remained unchanged in the low oxygen coverage range,and the adsorbed oxygen atoms preferentially adsorbed fcc sites near the ridge.With the increase of the coverage of adsorbed oxygen atoms on the surface,when the ridge site on the surface is filled,the O atom will occupy three times the hollow site on the platform site in the valley position.The oxygen atoms on the surface of Pd(111)are preferentially adsorbed at the fcc site,which makes the structure the most stable and the crystal structure remains unchanged.When the oxygen coverage of Pd(111)surface is 0.42 ML,one Pd atom on the surface will be lifted by two oxygen atoms.When the oxygen coverage is relatively high(0.50 ML),the oxygen atom gradually penetrates into the sub-surface layer.(3)We investigated the oxidation process and reaction mechanism of formaldehyde(HCHO)on Pd(100)surface.When oxygen coverage increased,the main intermediates were HCOOH and HCOO.
Keywords/Search Tags:machine learning, Primary oxidation of palladium, Computational simulation
PDF Full Text Request
Related items