CDS&E: First Principles Prediction of Thermal Radiative Properties of Dielectric Materials

该项目由材料研究部的凝聚态物质和材料理论项目以及化学、生物工程、环境和运输系统部的计算和数据支持科学与工程和热传输过程项目资助。热辐射在广泛的能源和热管理应用中起着关键作用，包括航天器、太阳能电池和被动辐射冷却。这些应用通常需要不同的选择性辐射特性：太阳能电池需要高吸收量的太阳光，而辐射冷却需要低吸收量的太阳光和在大气透明窗口中高发射红外光。通过反射阳光，同时将红外光辐射到太空，辐射冷却涂料已经被证明可以在不消耗任何能量的情况下将表面冷却到低于环境温度。筛选和设计这种材料需要了解热辐射特性如何依赖于材料的原子结构。然而，用于此目的的方法和软件工具通常是缺乏的，经验的试错方法仍然是主流。因此，该项目的目标是加强理论和模拟方法，可以从原子结构中预测材料的热辐射特性，并随后开发和部署一个开源代码，帮助其他研究人员建模他们自己的辐射材料。此外，PI将使用这些工具来了解粒子基纳米复合材料中超高效辐射冷却的原子起源，并利用机器学习对包括氧化物、碳酸盐和硫酸盐在内的大量材料进行高通量筛选，旨在发现更好的辐射冷却材料。这项工作将节约能源，对应对气候变化具有重大意义。同时，该项目将纳入教育和外联工作。除了扩大关于辐射材料的研究生和本科生课程外，它还将提供技术上有吸引力的题目，以扩大妇女和代表性不足的群体在工程和科学领域的参与。本研究的目标是开发计算热辐射特性的第一性原理方法，部署一个开源代码，并实现颗粒基质辐射冷却涂料的高通量筛选。在广泛的能源和热管理应用中需要量身定制的热辐射性能。然而，目前还没有从第一性原理预测电介质材料红外辐射特性的开源代码，这阻碍了从原子结构对辐射特性的理解和新型辐射材料的设计。同时，尽管在辐射特性的第一性原理预测方面取得了令人鼓舞的进展，但还需要包括其他重要的声子散射过程以及声子重整化。这些工具将对诸如选择辐射冷却材料等应用极为有益，这些材料目前是在经验试错的基础上进行研究的。在这个项目中，PI将通过计算和数据支持的方法来解决这些紧迫的研究需求。具体的研究任务有三个：(1)通过结合声子重整化、声子-电子散射、缺陷、杂质和边界声子散射，增强对四声子散射以外热辐射特性的第一性原理预测能力；(2)开发和部署热辐射性质第一性原理计算的开源代码；(3)结合第一性原理预测、蒙特卡罗模拟和机器学习，实现电介质颗粒-聚合物基质辐射冷却涂料的高通量筛选。该项目有望从第一性原理预测介电材料的热辐射特性，并使研究人员能够通过开源代码筛选或设计热辐射材料，达到前所未有的精度。它有可能改变目前的试错实践，不仅适用于辐射冷却纳米复合材料，也适用于许多其他重要的辐射材料，如热障涂层、热光伏发射器和空间任务涂层。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This project is funded by the Condensed-Matter-and-Materials-Theory program in the Division of Materials Research and by the programs in Computational and Data-Enabled Science and Engineering and Thermal Transport Processes in the Division of Chemical, Bioengineering, Environmental, and Transport Systems.Non-technical summaryThermal radiation plays a key role in a broad set of energy and thermal-management applications, including spacecraft, solar cells, and passive radiative cooling. These applications often require distinct selective radiative properties: high absorption of sunlight is needed for solar cells, while low absorption of sunlight and high emission of infrared light in the window of atmospheric transparency are desired for radiative cooling. By reflecting sunlight while radiating infrared light to space, radiative-cooling paints have been shown to cool surfaces to below the ambient temperature without any energy expenditure. Screening and designing such materials call for an understanding of how thermal radiative properties depend on the atomic structures of materials. However, methods and software tools for this purpose are generally lacking, and empirical trial-and-error approaches are still the mainstream. Therefore, the objectives of this project are to enhance theoretical and simulation methodologies that can predict thermal radiative properties of materials from their atomic structures and subsequently to develop and deploy an open-source code that will help other researchers model their own radiative materials. Moreover, the PI will use these tools to understand the atomistic origins of ultra-efficient radiative cooling in particle-matrix nanocomposites and employ machine learning to pursue high-throughput screening of a large number of materials including oxides, carbonates, and sulfates, aiming to discover better radiative-cooling materials. The work will lead to energy savings with significant promise for combating climate change. In parallel, this project will incorporate education and outreach efforts. Besides expanding the graduate and undergraduate curriculum on radiative materials, it will provide technologically attractive topics to broaden the participation from women and underrepresented groups in engineering and science.Technical summaryThe goals of this research are to develop first-principles methods for calculating thermal radiative properties, deploy an open-source code, and enable high-throughput screening of particle-matrix radiative cooling paints. Tailored thermal radiative properties are demanded in a broad set of energy and thermal-management applications. However, no open-source codes are available to predict infrared radiative properties of dielectric materials from first principles, hindering the understanding of radiative properties and the design of new radiative materials from atomic structures. Meanwhile, although encouraging progress has been made in first-principles prediction of radiative properties, additional important phonon-scattering processes as well as phonon renormalization need to be included. Such tools will be extremely beneficial for applications such as selecting radiative-cooling materials, which are currently studied on an empirical trial-and-error basis. In this project, the PI will address these urgent research needs via computation and data-enabled approaches. There are three specific research tasks: (1) enhancing the capabilities of first-principles prediction of thermal radiative properties beyond four-phonon scattering, by incorporating phonon renormalization, phonon-electron scattering, and phonon scattering with defects, impurities, and boundaries; (2) developing and deploying an open-source code for first-principles calculations of thermal radiative properties; and (3) coupling first-principles predictions, Monte-Carlo simulations, and machine learning to enable high-throughput screening of dielectric particle-polymer-matrix radiative-cooling paints. The project is expected to achieve unprecedented accuracy in predicting thermal radiative properties of dielectric materials from first principles and enabling researchers to screen or design thermal radiative materials via an open-source code. It has the potential to change the current trial-and-error practice not only for radiative-cooling nanocomposites but also for many other important radiative materials such as thermal barrier coatings, thermophotovoltaic emitters, and coatings for space missions.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.