Surface Modification And Gas-Sensing Properties Of Hierarchical Tungsten Oxide

Gas sensors play a very important role in people’s production and life.They can be used to detect harmful gases in the environment,such as carbon dioxide,carbon monoxide and formaldehyde,to ensure our health and safety.In order to build high performance gas sensors,scientists have been exploring and researching gas-sensitive materials,aiming to find more sensitive and selective materials.At the same time,researchers also need to consider how to improve the stability and reliability of gas-sensitive materials to ensure the long-term stability and accuracy of the sensor.In short,the development of gas sensor is of great significance to people’s production and life,and the exploration and research of gas sensitive materials will continue to become the key research direction in this field.In hierarchical structured gas-sensitive materials,nanoparticles are usually used as the base unit and macrostructures are assembled by self-assembly or other methods.This structure improves the sensitivity and selectivity of the material by increasing its porosity and active sites,and enables the detection of lower concentrations of gases.In this paper,the three-dimensional structure of tungsten oxide as a gas sensitive material,mainly through surface modification means to control its performance,with the aid of machine learning algorithm to develop high performance gas sensor.The main research contents of this paper are as follows:1.A simple oxygen-defect-assisted photoreduction strategy is adopted,the excited electrons in the defective WO_3-xpartially restore theπnetwork within the GO nanosheets under the illumination conditions,and then the in situ tightly bound r GO/WO_3-xcomposites are designed.The introduction of r GO not only broadens the visible light absorption range of the defective WO_3-x,but also accelerates its carrier migration rate.The morphological structure and forbidden band width of the composites were analyzed by SEM,UV-vis,Raman and other characterization means,and the reduction of GO and oxidation of WO_3-xduring the light exposure were further analyzed.The effect of GO incorporation on the gas-sensitive performance of WO_3-xwas systematically investigated,and the composite showed a 198%increase in response value to 40 ppm triethylamine at220°C,a 24.6%reduction in response time,and a decrease in the minimum detection limit from 138 ppb to 92 ppb compared to pure WO_3-x.The improved performance was attributed to the in situ interface formation of the r GO/WO_3-xcomposite The improved performance was attributed to the in situ interface formation of the r GO/WO_3-xcomplex structure,which suppressed the electron-hole complexation and improved the receptor function.2.The BiVO₄-modified WO₃composites were prepared by ultrasonic and hydrothermal methods,and the effect of the introduction of BiVO₄on the gas-sensitive properties of WO₃was systematically investigated.The results showed that the BiVO₄/WO₃composites exhibited a 400%higher response value to 40 ppm triethylamine at 190°C compared to pure WO₃,and had a lower lower limit of detection（57 ppb）and better selectivity and stability.The superior performance is attributed to the successful modification of monoclinic BiVO₄on WO₃,and the close contact between BiVO₄and WO₃promotes carrier transport.For this composite we developed an intelligent framework with good visibility using a machine learning classifier to identify ppm-level triethylamine and predict its exact concentration.The classifier（KNN）is able to clearly distinguish the decision boundary map with an accuracy of 92.3%.In addition,our proposed regression model can correctly predict the untrained triethylamine concentration with an accuracy of 91.1%.This work not only explores the basic gas-sensitive properties of BiVO₄/WO₃composites,but also provides an intelligent strategy for accurately identifying and predicting trace amounts of triethylamine.3.Pt was modified on the WO₃surface by ascorbic acid（AA）reduction method and Pt/WO₃composites were prepared.The gas-sensitive test results showed that the Pt modified WO₃composite exhibited excellent gas-sensitive performance with a response value of 8.2 for 40 ppm triethylamine at 190°C.In addition,the composite exhibited excellent humidity and long-term stability to triethylamine,ammonia and isopropanol.The improved performance is mainly attributed to the spillover and catalytic effect of Pt.We also developed an analytical method to achieve selective detection of triethylamine,ammonia and isopropanol gases with a single Pt/WO₃gas sensor.S（response value）,T（response time）,Y（response phase offset）and CV（highest value under 4 levels of discrete wavelet transform decomposition）were used as four eigenvalues to effectively distinguish between gas species.Machine learning classifiers such as ANN,KNN,SVM,NB,RL and DT are applied to the dataset degraded by PCA.Among them,the ANN classifier predicts the target gas with 100%accuracy.This work lays the foundation for intelligent application of individual metal oxide gas sensors.