CAREER: Automating Construction of Polarizable, Flexible, Nonreactive Force-Fields for Metal-Organic Frameworks & Applications to Helium and Solar Water Splitting Gas Purificat

该职业奖支持预测气体分离候选材料的计算和理论研究和教育。太阳能水分解是一种可再生能源，可用于减少温室气体排放。在太阳能水分解中，通过将水转化为氢气和氧气来捕获太阳光能量。氢是一种清洁燃料，燃烧时只产生水。在这个项目中，PI和他的学生研究团队将使用计算模型来预测用于分离太阳能水分解气体产物的高性能膜材料。PI的方法可能会为能源应用带来更高效、成本更低的制氢设备。膜材料也将用于从自然气体中净化氦。传统上，氦是通过在极冷的温度下液化来提纯的，但冷却到这种温度需要巨大的能量。PI的团队将开发计算方法来搜索最近发表的金属有机框架材料数据库，以识别具有孔径大小和其他适合氦净化特性的分子筛。他们将进一步开发计算方法来模拟由这些材料制成的膜中的气体分离。通过允许在适度温度下进行部分气体净化，这将减少氦气净化过程中的能量需求。在这个项目中开发的计算方法将在PI小组开发的软件工具中公开提供。教育活动是该项目的重要组成部分。PI将在计算材料科学方法方面培训研究生和本科生。研究生将进行博士论文研究，扩展计算方法的能力。PI和研究生将制作YouTube视频，以适合K至12年级、本科生和公众的水平解释计算材料科学。他们还将准备通过Nanohub.org传播的培训模块，向研究生和专业人士解释计算材料科学技术。PI和研究生将以科学博览会评委的身份向初高中学生进行宣传。来自代表性不足群体的学生将参与该项目。该职业奖支持预测气体分离候选材料的计算和理论研究和教育。PI和学生将研究从量子化学计算中自动参数化灵活、极化力场的策略。这些力场将用于金属-有机框架的经典分子动力学和蒙特卡罗模拟，以计算气体扩散常数和吸附等温线。实验已知的金属-有机框架晶体结构数据库将被筛选，以确定适合净化(A)天然气源中的氦和(b)太阳能水分解产生的氢气的金属-有机框架基材料。低温精馏是目前主要的氦净化方法。使用氦选择膜代替全部或部分的净化可以显著降低能量需求。太阳能水分解是一种环保、可再生的氢气源。无线太阳能水分解利用浸没在液体中的粒子，生成含有微量水蒸气的氢、氧混合物。无线太阳能水分解可以提供更高的气体生成率每体积比有线太阳能水分解包含不同的阳极和阴极室。该项目将通过确定合适的金属氧化物框架材料用于氢气-氧气-水气体分离，从而使无线太阳能水分解（以及具有小阳极-阴极间隙距离的有线太阳能水分解）得到更广泛的应用。由于用纯金属氧化物框架晶体构建机械坚固的膜是困难的，因此PI将致力于预测更容易制造的混合基质膜，其中包含嵌入机械坚固聚合物的金属氧化物框架晶体。先前的金属氧化物框架研究表明，框架的灵活性有时会影响气体扩散常数的数量级。关键的科学挑战是开发计算效率高、自动化的方法来构建精确的柔性力场。为了实现这一目标，PI的研究团队将结合(1)计算净原子电荷和其他原子性质的密度衍生静电和化学方法，(2)修改Tkatchenko-Scheffler自洽色散筛选来计算极化率和色散系数，(3)修改快速力场（QuickFF）方法来计算灵活性参数。通过计算系统特有的力场参数，该方法可以获得比一般力场更高的精度。在这个项目中开发的计算方法将在PI小组开发的软件工具中公开提供。教育活动是该项目的重要组成部分。PI将在计算材料科学方法方面培训研究生和本科生。研究生将进行博士论文研究，扩展计算方法的能力。PI和研究生将制作YouTube视频，以适合K至12年级、本科生和公众的水平解释计算材料科学。他们还将准备通过Nanohub.org传播的培训模块，向研究生和专业人士解释计算材料科学技术。PI和研究生将以科学博览会评委的身份向初高中学生进行宣传。来自代表性不足群体的学生将参与该项目。

NONTECHNICAL SUMMARYThis CAREER award supports computational and theoretical research and education in predicting candidate materials for gas separation. Solar water splitting is a renewable energy source that could be used to reduce greenhouse gas emissions. In solar water splitting, sunlight energy is captured by turning water into hydrogen and oxygen gases. Hydrogen is a clean fuel that generates only water when burned. In this project, the PI and his team of student researchers will use computational modeling to predict high performance membrane materials for separating the gaseous products of solar water splitting. The PI's approach may lead to more efficient and lower cost hydrogen producing devices for energy applications.Membrane materials will also be predicted for the application to purify helium from naturally occurring gases. Helium is traditionally purified by liquefying it at extremely cold temperatures, but enormous energy is required to cool to these temperatures. The PI's team will develop computational methods to search a recently published database of metal-organic framework materials to identify molecule sieves having pore sizes and other characteristics suitable for helium purification. They will further develop computational methods to model gas separations in membranes made from these materials. This should reduce energy requirements during helium purification, by allowing part of the gas purification to be performed at moderate temperatures. The computational methods developed in this project will be made publically available in software tools developed by the PI's group.Educational activities are an important part of this project. The PI will train graduate and undergraduate students in computational materials science methods. Graduate students will perform doctoral dissertation research extending the capabilities of computational methods. The PI and graduate students will develop YouTube videos explaining computational materials science at levels appropriate to K to 12 and undergraduate students and the general public. They will also prepare training modules to be disseminated through Nanohub.org and that explain computational materials science techniques to graduate students and professionals. The PI and graduate students will perform outreach to middle and high school students by being science fair judges. Students from under-represented groups will be involved in the project. TECHNICAL SUMMARYThis CAREER award supports computational and theoretical research and education in predicting candidate materials for gas separation. The PI and students will research strategies to automatically parameterize flexible, polarizable force-fields from quantum chemistry calculations. These force-fields will be used in classical molecular dynamics and Monte Carlo simulations of metal-organic frameworks to compute gas diffusion constants and adsorption isotherms. A database of experimentally known metal-organic framework crystal structures will be screened to identify metal-organic framework-based materials suitable for purifying (a) helium from naturally occurring gas sources and (b) hydrogen gas from solar water splitting. Cryogenic distillation is currently the primary helium purification method. Using helium-selective membranes instead for all or portions of this purification could dramatically reduce energy requirements. Solar water splitting is an environmentally friendly and renewable hydrogen gas source. Wireless solar water splitting uses particles immersed in liquid to cogenerate a hydrogen and oxygen gas mixture with trace water vapor. Wireless solar water splitting could offer higher gas generation rates per volume than wired solar water splitting containing distinct anode and cathode compartments. This project will enable more widespread use of wireless solar water splitting (and wired solar water splitting with small anode-cathode gap distance) by identifying suitable metal-oxide framework-based materials for hydrogen-oxygen-water gas separations. Because constructing mechanically robust membranes from pure metal-oxide framework crystals is difficult, the PI will aim to predict easier-to-fabricate mixed matrix membranes containing metal-oxide framework crystals embedded in mechanically robust polymers. Prior metal-oxide framework studies showed framework flexibility that sometimes impacts gas diffusion constants by orders of magnitude. The key scientific challenge is to develop computationally efficient, automated methods to construct accurate flexible force-fields. To achieve this, the PI's research team will combine (1) the Density Derived Electrostatic and Chemical method for computing net atomic charges and other atomic properties with (2) a modification of Tkatchenko-Scheffler self-consistent dispersion screening to compute polarizabilities and dispersion coefficients with (3) a modification of the quick force-field (QuickFF) method for computing flexibility parameters. By computing system-specific force-field parameters, this approach should achieve higher accuracy than generic force-fields. The computational methods developed in this project will be made publically available in software tools developed by the PI's group.Educational activities are an important part of this project. The PI will train graduate and undergraduate students in computational materials science methods. Graduate students will perform doctoral dissertation research extending the capabilities of computational methods. The PI and graduate students will develop YouTube videos explaining computational materials science at levels appropriate to K to 12 and undergraduate students and the general public. They will also prepare training modules to be disseminated through Nanohub.org and that explain computational materials science techniques to graduate students and professionals. The PI and graduate students will perform outreach to middle and high school students by being science fair judges. Students from under-represented groups will be involved in the project.