Study On The Mechanism Of Catalytic Oxidation Of 5-hydroxymethylfurfural Over Metal Oxide Based Composites

With the orderly advancement of the"pesk carbon emission in 2030 and carbon neutrality in 2060"goal,biomass resources will play an important role as"zero carbon energy".The prospect of bio-resource refining to replace petroleum refining has been proved by research work at domestic and overseas,and a relatively clear target product has been established.In the process of high-value treatment of biomass resources,the selective oxidation of 5-hydroxymethylfurfural（HMF）still has disadvantages such as high conversion cost,unstable system and unclear synergistic mechanism among process factors.Based on the bottleneck problem of HMF selective oxidation,this work carried out a series of studies on high value-added products such as 2,5-furandicarboxylic acid（FDCA）and 2,5-diformylfuran（DFF）,mainly including the following four aspects:First,preparation of spinel-based catalysts and study on the mechanism of catalytic oxidation of HMF to produce FDCA.In view of the problems of high reaction temperature and participation of precious metals or inorganic base in the existing catalytic systems,a series of spinel-based catalysts were prepared to realize the efficient oxidation of HMF to FDCA.The A-site and B-site metals in the chemical structure of spinel"AB₂O₄"were optimized.Among them,the NiCo₂O₄ structure has the most excellent catalytic performance.Further,NiCo₂O₄ with different morphologies,such as nanosheet-like,sea urchin-like,pom-pom-like and corolla-like were prepared.nanosheet-like NiCo₂O₄ exhibited excellent catalytic activity and stability that the yield of FDCA reached 60.1%within 12 hours.The excellent performance of nanoheet NiCo₂O₄ was attributed to the abundant oxygen vacancies and acid sites generated by the exposed（400）crystal planes.The oxygen vacancies and acidic sites on the catalyst surface can form precise adsorption of-CHO and-CH₂-OH to HMF,respectively,and this synergistic effect promoted the efficient production of FDCA.This work provided a new strategy for constructing multifunctional catalytic systems containing oxygen vacancies and acidic sites.Second,preparation of CN-WO₃@MnO₂ catalysts and study on the mechanism of catalytic oxidation of HMF to produce DFF.In view of the problems of harsh conditions and uncontrollable oxidation process in photocatalytic systems,CN-WO₃@MnO₂ catalysts with efficient electron transport channels and oxygen adsorption sites were prepared.The experimental results showed that CN-WO₃@MnO₂ achieved a DFF yield of 61.8%and a DFF selectivity of 79.6%within 24 hours under 420 nm LED light.Through radical trapping experiments,electron spin resonance（EPR）characterization and density functional theory（DFT）over the catalysts,it is confirmed that the efficient transport of photogenerated electrons between CN-WO₃ and MnO₂can effectively control the generation of·OH.Meanwhile,the oxygen adsorption sites on the surface of MnO₂ can improve the generation efficiency of·O₂^-and·¹O₂.The synergistic effect of·O₂^-and ¹O₂ promoted the efficient preparation of DFF.This work is of great significance for the development of multifunctional catalysts with efficient electron transport channels and oxygen adsorption sites to promote synergy between free radicals.Third,preparation of alumite-supported manganese dioxide catalysts and study on the mechanism of catalytic oxidation of HMF to produce DFF.Manganese dioxide frequently leads to the deep oxidation of the target product in the HMF selective oxidation reaction due to its abundant oxygen vacancies and chemisorption sites.In order to improve the selectivity of manganese dioxide in the thermal catalytic oxidation of HMF to prepare DFF,a series of alunite-supported manganese dioxide（Alunite@MnO₂）catalysts were synthesized by hydrothermal method using alumina,potassium permanganate and manganese sulfate as raw materials.The experimental results showed that Alunite@40%MnO₂ exhibited excellent catalytic activity,achieving 90.7%HMF conversion,82.4%DFF yield and 90.8%DFF selectivity within6 hours at 120°C.In the Alunite@40%MnO₂ catalytic system,the introduction of alumite can regulate the surface oxygen vacancies of manganese dioxide.At the same time,the specific morphology and contact mechanism formed between the Alunite and MnO₂ phases caused the steric hindrance effect over the catalyst active phase interface,which enhanced the active site escape probability of DFF and ensured the efficient production of DFF.The application of Alunite@40%MnO₂ catalyst in the selective oxidation of HMF provided a new reference for the regulation of oxygen vacancies on the surface of transition metal oxides.Fourth,In-depth optimization of alumite-supported manganese dioxide catalytic oxidation process of HMF based on machine learning model.Aiming at the problems of lack of high-content database and weak model universality for parameter regulation in the catalytic reaction process,a data set of catalytic oxidations of HMF to DFF by alunite-supported manganese dioxide was constructed.Six factors including MnO₂content,temperature,time,concentration,dosage and oxygen pressure were selected as x variables,and HMF conversion,DFF yield and DFF selectivity were selected as evaluation indicators y1,y2 and y3.The datasets were trained by random forest algorithm,gradient boosting algorithm and adaptive boosting algorithm respectively.After comparison,the fitting results of the random forest algorithm were more credible.The characteristic importance order of y1（HMF conversion）and y3（DFF selectivity）are:Mn（MnO₂ content）>d（catalyst dosage）>c（HMF concentration）>T（reaction temperature）>P（oxygen pressure）>time（reaction time）.The characteristic importance order of y2（DFF yield）is:Mn（MnO₂ content）>d（catalyst dosage）>c（HMF concentration）>P（oxygen pressure）>T（reaction temperature）>time（reaction time）.The Partial Dependency（RF-PDP）algorithm is introduced to explore the interaction between multiple features.For y1（HMF conversion）and y2（DFF yield）,there is a set of strongly interacting feature pairs:MnO₂ content and catalyst dosage.For y3（DFF selectivity）,there are two sets of strongly interacting feature pairs:MnO₂content and catalyst dosage,MnO₂ content and HMF concentration.The establishment of this model improved the interpretability of machine learning and revealed the interaction between various process parameters in the catalytic system.