CAREER: Accurate, Reliable, and Routine First-Principles Prediction of the Structure and Stability of Molecular Crystal Polymorphs

小罗伯特·A·迪斯塔西奥教授康奈尔大学的这项研究得到了化学系化学理论、模型和计算方法项目的支持，以开发能够准确和可靠地预测分子晶体的结构和稳定性的方法。分子晶体是普遍存在的、用途广泛的材料，用于替代能源和环境科学、制药和医学，以及技术和工业。对这些应用至关重要的是，分子晶体通常具有许多可访问的多晶型--在稳定性上几乎相同但显示出截然不同的物理和化学性质的替代结构。当用于提高器件性能或实现替代能源应用中的新功能时，这些独特的特性可能非常有益。多晶型的存在也可能具有破坏性和潜在的灾难性影响，例如，当药剂意外地转变为未知的(和不活跃的)多晶型，从而减少了抗病(和潜在的挽救生命)形式的数量。Distasio和他的研究小组正在开发准确可靠的方法，根据量子力学和统计力学的定律，使用计算机模拟来预测分子晶体多晶型的结构和稳定性。在这样做的过程中，DiStasio正在开发一种计算框架，以解决在新的能源解决方案和制药试剂中使用分子晶体多晶型的新机会。DiStasio研究小组还在开发一个交互式分子动力学(MD)包，它允许所有级别的学生(K-12、本科生和研究生)将化学中遇到的抽象数学概念可视化，并建立关于观察到的性质和物质微观结构之间复杂相互作用的物理和化学直觉。通过开发和利用结合量子力学和统计力学、数值分析和高性能计算的最先进方法，DiStasio和他的研究小组正在创建一个理论和算法框架，该框架利用稀疏性对整个生物学中感兴趣的复杂和大规模凝聚相系统进行高精度的从头算分子动力学(AIMD)模拟，化学、物理和材料科学。这个框架同时考虑了电子(包括复杂的交换-关联效应)和原子核的量子力学性质，以及现实的实验条件(即有限的温度和压力)，因此能够从第一原理准确、可靠和常规地预测所有热力学(和动力学)相关的分子晶体晶型的结构和稳定性。这些研究工作为加强对复杂的多态能量格局的基本理解和基于预测第一原理的晶体结构预测的常规使用铺平了道路，这两者对于获取分子晶体巨大且基本上未被探索的潜力都是至关重要的。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Professor Robert A. DiStasio Jr. of Cornell University is supported by an award from the Chemical Theory, Models and Computational Methods program in the Division of Chemistry to develop methods which enable accurate and reliable predictions of the structure and stability of molecular crystals. Molecular crystals are ubiquitous and versatile materials that are used in alternative energy and the environmental sciences, pharmaceuticals and medicine, as well as technology and industry. Of crucial importance to these applications is the fact that molecular crystals often have many accessible polymorphs—alternative structures that are nearly identical in stability yet display drastically different physical and chemical properties. When used to improve device performance or enable novel functions in alternative energy applications, these distinct properties can be very beneficial. The existence of polymorphs can also have devastating and potentially catastrophic effects, e.g., when a pharmaceutical agent unexpectedly converts into an unknown (and inactive) polymorph, and thereby reduces the amount of the disease-fighting (and potentially life-saving) form. DiStasio and his research group are developing accurate and reliable methods which use computer simulations to predict the structures and stabilities of molecular crystal polymorphs based on the laws of quantum and statistical mechanics. In doing so, DiStasio is developing a computational framework that addresses new opportunities to use molecular crystal polymorphs in novel energy solutions and pharmaceutical agents. The DiStasio research group is also developing an interactive molecular dynamics (MD) package which allows students at all levels (K-12, undergraduate, and graduate) to visualize the abstract mathematical concepts encountered in chemistry, and to build physical and chemical intuition regarding the complex interplay between observed properties and the microscopic structure of matter.By developing and utilizing state-of-the-art methods that combine quantum and statistical mechanics, numerical analysis, and high-performance computing, DiStasio and his research group are creating a theoretical and algorithmic framework which exploits sparsity to enable highly accurate ab initio molecular dynamics (AIMD) simulations of complex and large-scale condensed-phase systems of interest throughout biology, chemistry, physics, and materials science. This framework simultaneously accounts for the quantum mechanical nature of the electrons (including sophisticated exchange-correlation effects) and the nuclei, as well as realistic experimental conditions (i.e., finite temperatures and pressures), and therefore enables accurate, reliable, and routine predictions of the structures and stabilities of all thermodynamically (and kinetically) relevant molecular crystal polymorphs from first principles. These research efforts pave the way towards an enhanced fundamental understanding of complex polymorphic energy landscapes and the routine use of predictive first-principles based crystal structure prediction, both of which are crucial to accessing the vast and largely unexplored potential of molecular crystals.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.