【作者】 胡贵华；

【导师】 俞涛； 朱文华；

【作者基本信息】 上海大学 ， 机械制造及其自动化， 2009， 博士

【摘要】 能源与环境问题是21世纪人类面临的两大主要问题。世界能源将以石油化工能源为主的能源逐步转变为太阳能为主的可再生能源。太阳能是永不枯竭、无污染的清洁能源。高效率多结叠层太阳电池(GaAs基系多结太阳电池)是目前世界上最具竞争力的新一代太阳电池,它是航天飞行器急需的高性能长寿命空间主电源。在研制GaAs叠层太阳电池的方法中,金属有机化学气相沉积(Metal Organic Chemical Vapor Phase Deposition, MOCVD)技术是最优的。当前国内外关于GaInP薄膜电池生长的研究都是从实验方面研究GaInP/GaAs/Ge叠层电池的工艺特性,从而优化工艺和结构参数,但是缺乏从计算机模拟方面研究GaInP薄膜生长的动力学机理。然而,计算机模拟往往在模型建立和计算分析方面具有较大优势,但缺乏直观和交互特性。虚拟现实(Virtual Reality, VR)可视化技术弥补了计算机模拟的不足,可以帮助设计师和工艺师发现设计中的问题,为改进设计提供理论依据。因此,将计算机模拟与虚拟现实可视化技术有机结合可以创造一个更人性化的集成仿真系统,为优化结构设计和工艺参数提供指导依据。本论文以MOCVD设备研制高效多结太阳电池为背景,用计算机模拟和可视化研究的方法,从宏观和微观两个角度分别研究GaInP外延材料的生长动力学机理。其主要研究内容和成果如下:1.运用计算流体力学(computational fluid dynamics, CFD)方法对生长高效三结GaInP/GaAs/Ge叠层太阳电池的MOCVD反应室内气体的热流场进行了数值模拟研究。通过对不同工艺参数下反应室内气体热流场的数值模拟,得出结论:其他条件一定时,在气体总流量5L/min和系统压力0.1atm时,衬底上的热流场分布比较均匀。同时,要求衬底的长度不能超过两端排气口,为优化MOCVD反应室衬底的尺寸设计提出了合理化建议。2.利用虚拟现实可视化技术对MOCVD反应室内气体的热流场进行了可视化仿真。可视化仿真结果准确直观地显示反应室内温度场和速度场的分布情况,研究人员可以透过反应室表面看清室内的温度场和速度场的分布情况,从而及时调整加热器的温度和进口气体的总流量等工艺参数,在Ge衬底表面获得均匀的热流场分布。3.运用正交实验法对MOCVD生长GaInP薄膜电池的工艺参数进行了数值模拟试验,对试验结果进行了极差分析和方差分析,得出了影响恒温区厚度H的工艺参数的主次顺序,并得出了最佳参数组合;用回归分析的方法建立了基于恒温区厚度的MOCVD工艺参数数学模型;经过试验验证了MOCVD工艺参数影响作用的普遍性,为MOCVD反应室热流场的恒温区厚度提供了经验公式,为合理确定MOCVD工艺参数提供了理论基础。4.以MOCVD技术生长高效GaAs叠层太阳电池为对象,提出了一种基于动力学蒙特卡罗方法的模拟多元化合物GaInP薄膜生长模型及模拟算法。对GaInP薄膜生长的过程进行了动力学蒙特卡罗模拟。5.将动力学蒙特卡罗模拟的结果与虚拟现实可视化技术结合,实现了MOCVD反应室内GaInP薄膜生长过程的可视化仿真。模拟仿真结果准确直观地展示了MOCVD反应室内GaInP薄膜生长的过程,揭示了模拟时间、衬底温度及沉积速率对GaInP薄膜形貌的影响规律;模拟仿真结果与实验结果具有一致性,为优化MOCVD生长GaInP薄膜的工艺参数提供理论依据。6.在Windows环境下,基于Visual Studio.net2003和Open Inventor开发了一个高效太阳电池生长机理仿真系统的原型系统,实现了总体设计中的各个功能模块。通过本文的研究,获得在一定条件下高质量GaInP薄膜材料的工艺参数,对提高太阳电池的效率和质量具有重要的意义。更多还原

【Abstract】 Energy and enviromment are two principal problems people confront in the 21st century. The world energy which used to be mainly petrochemical energy will gradually convert into renewable resources - solar energy. Solar energy is inexhausted and pollution-free. At present, high-efficiency multijunction tandem solar cells (GaAs basic series multijunction solar cells) are the most competitive in the new generation solar cells in the world, and are high-performance and long-life space main electrical sources which aerospace craft urgently requires. Among the development methods of GaAs tandem solar cells, metal organic chemical vapor deposition (MOCVD) is optimum. Nowadays all the studies at home and abroad have been paid attention to technical characteristics of GaInP/GaAs/Ge tandem solar cells from the experimental aspect, consequently optimizing processing and structural parameters, but the dynamic mechanism of GaInP film growth has little attention from the point of view of computer simulation.However, computer simulation has usually more advantageous at aspects of model building and computing analysis, being short of intuitive and interactive characteristics. Virtual reality (VR) visualization technology has made up the disadvantages of computer simulation, and can help designer and technician find the problems of design, consequently providing theoretical basis for improving design. Therefore, the organic combination of computer simulation and VR visualization technology can create a more humanized integrated simulation system, providing guidance basis for the optimization of structural design and processing parameters.With the background of high-efficiency mulitjunction solar cell made by MOCVD equipment, this thesis uses the study method of combining computer simulation with visualization, studying growth dynamic mechanism of GaInP epitaxial material from the macroscopic level and microscopic level respectively. The thesis’s main study contents and results are as follows:1. Method of Computational Fluid Dynamics (CFD) was applied to numerical simulation of gas’s thermal and flow fields of MOCVD reactor which grows high-efficiency three-junction GaInP/GaAs/Ge tandem solar cells. Through CFD numerical simulation of gas’s thermal and flow fields in reactor for different processing parameters, the results are acquired that, on a certain conditions, the distributions of thermal and flow fields over the substrate are relatively uniform at the total gas flux of 5L /min and system pressure of 0.1atm. Meanwhile, the substrate length must be not more than the distance between two ends of exhaust gas, thus providing rational suggestion for optimization design in size of the substrate in the reactor. 2. VR visualization technology was applied to the visualization of gas’s thermal and flow fields of MOCVD reactor. The visualization results truly and intuitively display distributing situation of gas’s temperature field and velocity field in MOCVD reactor. Researches can visually see the distributions of temperature and velocity fields inside reactor through wall, thus adjusting the processing parameters such as the heater’s temperature and the whole flow of gas’s inlet in time so as to acquire uniform thermal and flow fields over Ge substrate.3. Orthogonal experimental method was applied to numerical simulation experiment for processing parameters of GaInP thin film cell grown by MOCVD. With range analysis and variance analysis on experiments, the importance order of processing parameters affecting thickness of the flat-temperature zone is found, and the best level combinations of parameters are confirmed. Meanwhile, non-linear regression analysis is used to build mathematical models of processing parameters based on thickness of the flat-temperature zone. By means of the experimental verification, the universality of processing parameters’effect on thickness of the flat-temperature zone is verified, and experiential formula is offered for thickness of the flat-temperature zone of thermal and flow fields of MOCVD. All those provide theoretical basis for reasonably confirming processing parameters of MOCVD.4. With the object of growing high-efficiency multijunction tandem solar cells by MOCVD technology, this paper proposed a model and algorithm of multielement compound GaInP thin film growth based on method of Kinetic Monte Carlo (KMC). KMC simulation was carried out for GaInP thin film growth process.5. The result of KMC simulation was combined with VR visualization technology. The visualization emulation of the process of GaInP thin film growth in MOCVD reactor was realized. The results of simulation and visualization truly and intuitively display process of GaInP thin film growth in MOCVD reactor, and they reveal the rule of influence of simulation time, substrate temperature and deposition rate on GaInP thin film morphology. The simulation results well coincide with experimental results, providing the optimizations of processing parameters of GaInP thin film grown by MOCVD with theoretical basis.6. Based on Visual Studio.net2003 and Open Inventor, a prototype system of simulation of growth mechanism of high-efficiency solar cell was implemented in the environments of Microsoft Windows. All modules in the system design are realized in the prototype system.By the study of this paper, the processing parameters of high-quality GaInP thin film material are acquired under a certain conditions. The study is of important meaning for improving the efficiency and quality of solar cell.更多还原