同步辐射光谱预测方法(Bethe-Salpeter程序)的发展及其应用

X-ray spectroscopy is an essential characterization tool to scientists of all disciplines. Experiments employing x-ray measurements have led to 20 Nobel Prizes in Physics, Chemistry and Medicine. While synchrotron-base experiments are indispensable, gaining a complete understanding from data is non-trivial. Experiment and theory must work together. Presently, there is no first-principles code that can produce accurate x-ray absorption spectra for large-scale systems. Our Bethe-Salpeter code can generate accurate X-ray Absorption Spectra, X-ray Emission Spectra, and Non-Resonant Inelastic X-ray Scattering spectra , but at high computational cost. We propose three efficiency improvements to our code that will increase its speed by a factor of 1000 allowing us to perform calculations on systems 100 times larger than currently feasible. First, we will replace Abinit by Quantum Espresso for the initial DFT stage of the calculation. Second, we will make the entire code compatible with ultrasoft pseudopotentials. Third, we will incorporate a k-space interpolation scheme for performing fast intergrations over the Brillouin zone. The resulting code will be world-leading. This project represents a simple means of greatly amplifying the scientific productivity of the Chinese national lightsources. To demonstrate the capabilities of the improved code we perform large-scale spectral calculations on systems with energy and biological applications. Specifically, we investigate the structure of bonds in water with the goal of accurately studying chemicals in an aqueous environment, we probe the effect that oxygen vacancies have on the water splitting efficiency of hematite, and we perform spectral characterizations of decay products of the oLED molecule Alq3 in an effort to understand the device failure process.

X射线光谱法是应用于不同学科的重要分析工具，对其进行理论预测的重要性不言而喻，但是目前基于第一性原理代码预测大尺度体系的精确的X-射线吸收光谱的方法还不成熟。我们研发的Bethe-Salpeter程序可以预测X射线的吸收、发射、散射光谱，但目前程序的计算成本很高。本项目书提出了三个途径以提高程序的效率，首先，我们将用Quantum Espresso取代ABINIT作为我们程序中更有效的DFT的计算手段；其次，我们将改进我们的程序以兼容超软赝势；第三，我们将k空间插值手段用在布里渊区的积分过程以提高程序的效率，从而可以将当前程序的速度提高1000倍，可预测体系的大小也可以提高100倍。本项目同时将新方法应用于不同的复杂体系：例如研究赤铁矿中氧空位对其分解水效率的影响，再如研究OLED分子Alq3的衰变机制。本项目的顺利实施将获得国际领先的光谱预测程序，在理论方面增强中国光源领域的科技创新能力

整个项目按照预定的计划顺利完成，在本项目的资助下，在J. Phys. Chem. C等本领域一流学术期刊发表论文7篇。在本项目的资助下，共培养1名博士生，3名硕士研究生。.X 射线光谱法是应用于不同学科的重要分析工具， 对其进行理论预测的重要性不言而喻，但是目前基于第一性原理代码预测大尺度体系的精确的 X-射线吸收光谱的方法还不成熟。我们研发的 Bethe-Salpeter 程序可以较快预测 X 射线的吸收、 发射、 散射光谱。 项目管理合理，实施进度顺利，顺利完成了项目的研究目标，具体包括：1. 基于OCEAN代码，发展高效的用于同步辐射光谱的预测程序，促进中国同步辐射预测技术的发展；2. 目前OLED设备寿命太短，受限于通过水解反应分解AlQ3。本项目使用分子动力学模拟这个反应和确定衰变产物；3. 研究赤铁矿，计算X射线吸收光谱，确定最有利的氧空位的配置，验证它们在光电化学电池设备中的机理；4. 研究硅纳米柱，计算X射线吸收光谱，研究异质结界面处重组机制，以减少界面重组并提高硅电池的效率。.