Bridging Length and Time Scales in Catalytic Reaction Systems

研究：动力学蒙特卡罗(KMC)模拟已经成为一种优秀的计算工具，适用于从材料生长、催化、DNA/表面相互作用、图像处理到生化工程和生物信息学的代谢路径建模等各种问题。通常，KMC模拟限于较短的长度和时间尺度，而器件大小和形态特征通常涉及更大的空间和时间尺度。PI一直在开发系统的、层次化的粗粒度随机模型，称为粗粒度MC(CGMC)，它能够以显著较低的计算代价描述比传统KMC模拟更大的长度尺度，同时仍然包含微观特征和正确的噪声。因此，它们提供了一种数学和计算范例，除了CGMC工具目前不能准确地处理复杂的化学物质之外，它对许多应用都有潜在的影响。在这个项目中，PI计划为表面反应系统开发必要的多尺度使能技术。这些任务脱离了现有的技术，需要结合以下几个方面：(A)估计表面反应参数和势能面的密度泛函理论，(B)模拟表面扩散的分子动力学，(C)新颖的多级自适应网格CGMC，以及(D)将这种多级CGMC集成到混合多尺度反应器模拟中，以便能够对真实的反应流动长度尺度进行建模。这将应用于模拟反应系统和通过高速扫描隧道显微镜(STM)观察到的铂上氢氧化的时空模式。智能的优点来自于反应多尺度框架的开发，该框架将具有将微观参数(如分子间势、微观速率和波动)与介观尺度现象(如时空图案和局部反应速率)联系起来的能力。更广泛的影响：粗粒化方法在许多应用中都具有潜力，从单一催化剂晶体，到便携式“绿色”设备的微型化学系统，到氢基燃料电池，到基于微孔薄膜的膜反应器，到先进材料制造，再到微磁学。因此，它可以使工业和环境受益。在教育方面，PI计划(1)开发一个关于介观建模的短期课程，(2)通过网络广泛传播研究材料，并为学生提供教程示例和代码。

Research: Kinetic Monte Carlo (KMC) simulations have emerged as an excellent computational tool for diverse problems ranging from materials growth, to catalysis, to DNA/surface interactions, to image processing, and to modeling of metabolic pathways for biochemical engineering and bio-informatics. Generally, KMC simulations are limited to short length and time scales, while device sizes and morphological features often involve much larger spatial and temporal scales. The PI has been developing systematic, hierarchical coarse-grained stochastic models, referred to as Coarse-Grained MC (CGMC), which are capable of describing much larger length scales than conventional KMC simulations at significantly lower computational cost, while still incorporating microscopic features and the correct noise. They thus provide a mathematical and computational paradigm with potential impact on numerous applications, except that CGMC tools cannot currently handle complex chemistry accurately.In this project the PI plans to develop the necessary multiscale enabling technology for surface reaction systems. The tasks depart from the available technology and entail a combination of: (a) density functional theory for estimation of surface reaction parameters and potential energy surfaces, (b) molecular dynamics to model surface diffusion, (c) novel multilevel, adaptive mesh CGMC, and (d) the integration of this multilevel CGMC in hybrid multiscale reactor simulations to enable modeling of realistic reacting flow length scales. This will be applied to model reaction systems and to the spatio-temporal patterns observed in hydrogen oxidation on platinum via high speed scanning tunneling microscopy (STM). The Intellectual Merit derives from the development of a reaction multiscale framework that will have the ability to link microscopic parameters, such as intermolecular potentials, micrscopkic rates, and fluctuations, to mesoscopic scale phenomena such as spatio-temporal patterns and local reaction rates. Broader Impacts: The coarse-graining methodology has potential in numerous applications, from single catalyst crystals, to microchemical systems for portable, "green" devices, to hydrogen-based fuel cells, to microporous films based membrane reactors, to advanced materials fabrication, to micromagnetics. It can thus benefit industry and the environment. On the educational side, the PI plans to (1) develop a short course on mesoscopic modeling and (2) widely disseminate the research material via the web with tutorial examples and codes for students.