自适应基因算法在表面界面问题的应用

发布时间：2019-03-29 17:40

【摘要】：在第一章中,我在第一部分(1.1-1.4)对第一性原理计算的基础理论进行了简单介绍,重点是基于电子密度的密度泛函理论(DFT)。由于第一性原理计算的可靠性,计算机性能计算速度的大幅提升,以及并行计算机的普及应用,使得利用第一性原理理论计算,由体系的化学成分出发,预测体系稳定结构成为可能,在第二部分(1.5)节介绍了全局优化算法,其中重点介绍了由Kai-MingHo组开发的自适应基因算法。在第二章中,简要介绍了TiO_2的几何结构和电子结构,以及不同TiO_2相导带价带的带边高低顺序。讨论了光催化制氢的三种方法,光伏电催化制氢、光解水制氢和光电催化制氢。阐述了光催化反应的原理,以及提高光催化反应效率的三个关键因素:(1)光吸收效率,(2)光生载流子寿命,以及(3)表面反应活性。在第三章中,我们研究了表面拉伸应力与表面反应活性的关系,解释了长久以来理论预测与实验观测的矛盾。在接近20年的时间里,密度泛函理论一直预测锐钛矿相(Anatase)TiO_2(001)面与其(1 ×4)重构都是高度活泼的,由此诱发了大量实验工作合成高比例的Anatase(001)面,但是最近扫描隧道显微镜的实验结果恰恰与其相反,实验显示除缺陷结构以外,这个表面是非常惰性的。在这一章中,我们对锐钛矿相TiO_2(001)-(1X4)重构表面进行了系统的第一性原理计算研究。我们的计算表明,表面的拉伸应力(Tensilestress)对表面活性起着至关重要的作用,之前理论预测的表面高活性源于本征存在的表面应力,而表面应力很容易通过缺陷等方式得到释放,与此同时抑制表面高活性。因此,合成具有少量表面缺陷的锐钛矿相TiO_2(001)表面对提高其催化活性至关重要,我们的结果表明通过水分子的吸附可以做到这一点。我们的研究不仅解释了实验观测和理论预测在锐钛矿相TiO_2(001)表面活性上的矛盾,而且揭示了这个表面活性起源的本质,这样的结论可以推广到其他氧化物表面。在第四章中,我们对锐钛矿TiO_2(001)-(1×4)表面可能存在的各种缺陷结构做了系统研究。缺陷的存在对氧化物表面活性有至关重要,因此理解缺陷的几何结构与其电子结构也非常重要。锐钛矿TiO_2(001)-(l X4)的表面缺陷具有高活性,然而由于表面重构的出现,在不同的实验中缺陷有着复杂的特征,这使得确定缺陷的原子结构非常具有挑战性。在本章中,我们利用自适应基因算法(AGA)和密度泛函理论(DFT),对TiO_2(001)-(1X4)表面缺陷进行了系统的第一性原理研究。对于不同的Ti-O比例,我们给出了具有较低能量缺陷的原子结构,并利用杂化泛函方法研究了其电子结构,并得出各种缺陷在真实环境下的稳定条件。我们成功的找到了不同的氧空位(Ov)和钛填隙(Tiimi)结构,这些结构在电荷局域程度,磁性和扫描隧道显微镜明暗特征上都与传统缺陷有明显区别。我们的计算结果成功地解释了相关实验。在第五章中,我们研究了 Pt过渡金属纳米颗粒在MoS2单层上吸附的结构。在不同的衬底上合成含有少量Pt金属原子的纳米颗粒,对设计高效低价的催化剂具有重要意义。为了理解Pt金属原子在MoS2单层上成核和外延生长的行为机理。我们使用自适应基因算法预测了 Ptn(n55)金属纳米颗粒的原子结构,并结合密度泛函理论,分析了结构的生长行为和电子结构。我们给出了 Pt金属颗粒在MoS2单层上的幻数,为7, 9和11。我们得到的结构几乎在单层上都是六角晶格,不同层间为HCP堆叠。当n20时,所有的最稳定结构都是两层,对于34或38,最稳定结构都是为三层。当n=55时,最稳定结构都是为四层。部分预测的结构都在扫描透射显微镜的实验中得到了验证。在最后一章中,对博士期间的工作做了一下总结,并对以后的工作做了一下展望。
[Abstract]:In the first chapter, I briefly introduce the basic theory of the first principle calculation in the first part (1.1-1.4), with the emphasis on the density functional theory (DFT) based on the electron density. due to the reliability of the first principle calculation, the great improvement of the computing speed of the computer performance and the popularization and application of the parallel computer, the first principle theory is used for calculating, and the stable structure of the system is predicted from the chemical composition of the system, In the second part (1.5), a global optimization algorithm is presented, which focuses on the adaptive genetic algorithm developed by the Kai-MingHo group. In the second chapter, the geometry and electronic structure of TiO _ 2 and the band-edge high and low order of the valence band of the conduction band of different TiO _ 2 phases are briefly introduced. The three methods of photocatalytic hydrogen production, the electrocatalytic hydrogen production of photovoltaic, the hydrogen production of the photolysis water and the photoelectrocatalytic hydrogen production are discussed. The principle of photocatalytic reaction and three key factors to improve the efficiency of photocatalytic reaction are described: (1) light absorption efficiency, (2) photo-generated carrier lifetime, and (3) surface reaction activity. In the third chapter, we have studied the relation between the surface tension stress and the surface reaction activity, and explained the contradiction between the theoretical prediction and the experimental observation for a long time. In the past 20 years, the density functional theory has always predicted that the surface of the anatase phase (Anatase) TiO _ 2 (001) and its (1-4) reconstruction are highly active, thus inducing a large number of experimental work to synthesize a high proportion of the Anatase (001) surface, but the experimental results of the recent scanning tunneling microscope are exactly the same as that of the same. The experiment shows that this surface is very inert, in addition to the defect structure. In this chapter, we calculated the first principle of the system on the surface of the anatase phase TiO _ 2 (001)-(1X4). Our calculations show that the tensile stress of the surface plays a critical role in surface activity, and the previously theoretical predicted surface activity is derived from the intrinsic surface stress, while the surface stress is easily released by a defect or the like while inhibiting the high activity of the surface. Therefore, the synthesis of the surface of the anatase phase TiO _ 2 (001) with a small number of surface defects is of great importance to the improvement of its catalytic activity, and our results show that this can be achieved by the adsorption of water molecules. Our study not only explained the contradiction between the experimental observation and the theoretical prediction on the surface activity of the anatase phase TiO _ 2 (001), but also revealed the essence of the origin of the surface activity, and the conclusion can be extended to other oxide surfaces. In the fourth chapter, we have systematically studied the various defects in the surface of anatase TiO _ 2 (001)-(1-4). The presence of defects is critical to the surface activity of oxides, and therefore it is also important to understand the geometry of the defect and its electronic structure. The surface defects of the anatase TiO _ 2 (001)-(l X4) have high activity, however, due to the appearance of surface reconstruction, the defects in different experiments have complex characteristics, which makes it very challenging to determine the atomic structure of the defect. In this chapter, we used the adaptive genetic algorithm (AGA) and the density functional theory (DFT) to study the surface defects of TiO _ 2 (001)-(1X4). For different Ti-O ratios, we give an atomic structure with a lower energy defect, and study the electronic structure by using the hybrid functional method, and obtain the stable conditions of various defects in the real environment. We have successfully found different oxygen vacancies (Ov) and titanium interstitial (Tiimi) structures, which are distinct from the conventional defects in both the charge local level, the magnetic and scanning tunneling microscope's dark and dark characteristics. Our calculation results in a successful interpretation of the relevant experiments. In the fifth chapter, we study the structure of Pt transition metal nanoparticles on MoS2 monolayer. The synthesis of nano-particles containing a small amount of Pt metal atoms on different substrates is of great significance for designing a high-efficiency and low-cost catalyst. In ord to understand that mechanism of the behavior of the nucleation and epitaxial growth of the pt metal atom on the MoS2 monolayer. The atomic structure of Ptn (n55) metal nanoparticles is predicted by the adaptive genetic algorithm, and the growth behavior and the electronic structure of the structure are analyzed with the density functional theory. The magic number of the Pt metal particles on the MoS2 monolayer is given as 7,9 and 11. The structure we've got is almost single-layer, hexagonal, and the different layers are HCP stacks. When n20, all of the most stable structures are two layers, and for 34 or 38, the most stable structure is three layers. When n = 55, the most stable structure is four layers. Some of the predicted structures were verified in the experiments of scanning the transmission microscope. In the last chapter, a summary of the doctor's work and a look at the future work.
【学位授予单位】：中国科学技术大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：O485

【相似文献】