Font Size: a A A

Preparation Of SnO2-based Thin Films By Atomization-Assisted CVD Method And Study On Optical-Electrical-Thermal Properties

Posted on:2022-12-21Degree:MasterType:Thesis
Country:ChinaCandidate:Y H ChenFull Text:PDF
GTID:2481306755999389Subject:Master of Engineering
Abstract/Summary:PDF Full Text Request
With the continuous development of our country's economy,the role and position of the construction of ecological civilization is increasingly prominent.People pay more and more attention to energy conservation and emission reduction.The field of building energy conservation is an important part of implementing energy conservation and emission reduction measures and realizing ecological civilization society.The low-emissivity performance of SnO2-based transparent conductive film has important theoretical significance and practical value in the field of building energy conservation.Traditional film preparation methods are complicated,difficult to operate and maintain,and high cost.In view of the low-emissivity glass on-line coating process,the atomization-assisted CVD film deposition system structure is optimized by using the three-dimensional structure design software and the finite element analysis software.A simple,easy to operate and low-cost film deposition device is developed.SnO2 and SnO2:Sb thin films were prepared on quartz substrates using this device.The effects of deposition parameters on the structure,morphology and photoelectric properties of the films were studied by X-ray diffraction,atomic force microscopy,scanning electron microscopy,ultraviolet-visible photometer,four-probe tester,etc.A parameter optimization method based on machine learning was proposed to optimize the deposition parameters of thin film.High quality SnO2:Sb low-emissivity thin film was prepared,and its optical,electrical and thermal properties were tested,which accumulated experience for the application of multifunctional low-emissivity glass.The main results are as follows:(1)Equipment and process development acquired atomization-assisted CVD thin film deposition system.The structure of the film deposition system was optimized by using the three-dimensional structure design software and the finite element analysis software.The results show that the gas carrier module,the atomization module,the liquid level constant module and the reaction chamber gap are important to the film deposition system.Finite element analysis of temperature field distribution in reaction chamber shows that reaction chamber with fine channel structure can deposit film more uniformly and stably.(2)SnO2 and SnO2:Sb thin films were prepared on quartz substrate.The effects of different deposition temperatures on the growth of SnO2 films were studied.The results of structure and surface morphology showed that the optimal deposition temperature was around 425?.In order to obtain high-performance low-emissivity coating materials,the growth of SnO2:Sb films with different deposition time and different precursor solution composition was further studied.The results show that the growth of films under different deposition time has three stages:amorphous transition,preferred growth and competitive growth,and the more appropriate deposition time is between 15 min and 45 min;HCl content regulates the growth behavior of the film to a certain extent.The optimal HCl content of high-performance SnO2:Sb low-emissivity film is about 4 m L.(3)The process parameters of atomization-assisted CVD deposition of SnO2:Sb thin films were optimized by machine learning method,and high-quality SnO2:Sb low-emissivity thin films were prepared.According to the statistics of electricity,optics and corresponding FOM numerical box diagram,the film deposition system based on self-developed is stable and reliable,and the experimental repeatability is high;The variance analysis of statistical data shows that the factors affecting the change of electrical properties are contradictory to the factors affecting the change of thin film.HCl content and deposition time are the key factors affecting the FOM of thin film.Using SVR and BOA to establish a stable SVR model,the model can stably predict the FOM value under any combination of parameters in the preparation process parameter space,and the relative error is kept within 25%.Its reliability is much higher than the FOM value calculated by predicting electrical and optical properties respectively.The two-dimensional cloud image is created by SVR model to explore the best preparation conditions of SnO2:Sb thin films in parameter space.The experimental results show that the best average FOM of 324.7 can be obtained when the Sb doping concentration is 2 at%,the deposition temperature is425?,the HCl content is 4.5 m L and the deposition time is 35 min.(4)Under the optimal process parameters,the SnO2:Sb thin film has a smooth surface and a dense structure,with an average grain size of about 120 nm,a relatively uniform grain size,an average thickness of about 380 nm,an average visible light transmittance of86.61%,and a sheet resistance of 21.1??S-1,the band gap width Eg of SnO2:Sb thin film calculated according to the Tauc equation is 4.26 e V.The thermal insulation performance test of SnO2:Sb film shows that the indoor temperature difference between SnO2:Sb film glass room and ordinary glass room can reach 2.4?after vertical irradiation of infrared lamp for 3 hours,which has good thermal insulation ability;It can reach 100?at a voltage of 10 V,and the high heating efficiency makes it have the application potential of radiant heating;the surface contact angle test of the film shows that compared with ordinary glass,the coated glass has better hydrophilicity,and its wetting angle is about15.26°.The above research results are of great significance for exploring and improving the online coating process of low-emissivity glass,improving the low-emissivity performance of coated glass,and expanding the application research of multi-functional low-emissivity glass.
Keywords/Search Tags:Atomization-assisted CVD, SnO2:Sb thin film, low-emissivity performance, machine learning, process parameter optimization
PDF Full Text Request
Related items