Integration of Computation, Experiment, Simulation and Data to Predict Defect Properties in Semiconductor Thin Films

技术描述：该项目旨在为描述复杂材料系统的加工过程的新框架奠定基础。它是通过将计算、实验测量、模拟和数据集集成到材料加工过程中的动力学演变的预测描述符中来实现的。具体的焦点/例子是对外延薄膜中位错网络的结构、电子和光学性质的预测模拟的发展，最终目标是加速此类系统中器件级材料的开发。该项目首先扩展了现有的模拟器，该模拟器是为预测GeSi/Si系统在生长和/或热处理过程中位错的产生而开发的。进化生物学和生物信息学领域发展起来的系统发育图谱方法，通过优化实验和模拟数据集的匹配，来精炼模拟器中的材料参数和基本力学假设。最初的一个重要目标是定义一个必要的“最小数据集”，以充分描述演化系统的能量和动力学参数，并使定量预测成为可能。下一步，该项目将把这些方法扩展到一种新的材料系统，具有显著不同的位错产生机制，应变III-氮化物膜。这既可以评估模拟器方法的一般性，也可以促进对III-氮化物系统中位错产生的理解。最后，模拟器正在扩展，以便能够预测与观察到的/预测的缺陷密度相关的光学和电子参数。这涉及到对特定缺陷配置的电子性质的第一性原理计算，从而实现了额外一代的模拟器，该模拟器将预测的缺陷密度和随后的(光)电子活动与观察到的位错材料的光电和器件属性相关联。非技术描述：该项目正在开发新的框架，以实现对薄膜晶体材料生长和加工过程中缺陷产生的预测模拟，以及这些缺陷对这些材料的性能参数的影响。这可以在整个加工周期之前预测设备质量的材料合成条件，从而加快设备开发周期。这最终促进了国家需要加快从新材料发现向可制造技术的过渡。将要开发的方法使用来自模拟器的预测与广泛的实验数据集的比较来改进和“训练”模拟器。它既提高了模拟器对其所训练的给定系统的预测能力，又加速了其应用向新材料系统的转移。从事这一项目的学生正在接受关于优化计算、实验和数据管理整合的愿景的培训，从而帮助建立一支接受过这些方法培训的队伍。研究团队正在使研发社区能够访问所产生的工艺模拟器，并正在响应选定的特定更新或增强的请求。该项目由材料研究部门(DMR)的电子和光子材料计划(EPM)以及计算和数据驱动材料研究计划(CDMR)共同资助。

Technical Description: This project seeks to develop the foundations for a new framework for describing the processing of complex materials systems. It is being achieved by integrating computation, experimental measurements, simulation and data sets into a predictive descriptor of kinetic evolution during materials processing. The specific focus / example is the development of predictive simulation of the structural, electronic and optical properties of dislocation networks in epitaxial films, with an ultimate goal of accelerating the development of device-quality material in such systems. The project is first extending an existing simulator developed for predicting dislocation generation in the GeSi/Si system during growth and/or thermal processing. Phylogenetic mapping methods, developed for the fields of evolutionary biology and bioinformatics, are employed to refine materials parameters and fundamental mechanistic assumptions in the simulator by optimizing the matching of experimental and simulated data sets. An initial important goal is to define a "minimum data set" necessary to adequately describe the energetic and kinetic parameters of the evolving system, and enable quantitative predictions. Next, the project is extending these methods to a new materials system, with significantly different dislocation generation mechanisms, strained III-nitride films. This enables both assessment of the generality of the simulator method, and should advance understanding of dislocation generation in the III-nitride systems. Finally, the simulator is being extended to enable prediction of optical and electronic parameters that correlate to the observed /predicted defect densities. This involves first-principle calculations of electronic properties of specific defect configurations, enabling an additional generation of simulator that correlates predicted defect densities and consequent (opto)electronic activity to observed optoelectronic and device properties of the dislocated material.Non-Technical Description: This project is developing new frameworks for enabling predictive simulation of defect generation during growth and processing of thin-film crystalline materials, and the effects of these defects upon the performance parameters of these materials. This enables accelerated device development cycle times by predicting conditions for device-quality materials synthesis ahead of full processing cycles. This ultimately contributes to the national need to accelerate the transition from new materials discoveries to manufacturable technologies. The method to be developed employs the comparison of predictions from the simulator to extensive experimental data sets to refine and "train" the simulator. It both improves the predictive capability of the simulator for the given system in which it is trained, and accelerates transfer of its application to new materials systems. The set of students working on this project are being schooled in this vision of optimizing integration of calculation, experiment and data management, thus helping to establish a workforce trained in these methods. The research team is making the resultant process simulator accessible to the research and development community and is responding to selected requests for specific updates or enhancements.This project is co-funded by the Electronic and Photonic Materials Program (EPM) and the Computational and Data driven Materials Research Program (CDMR), both in the Division of Materials Research (DMR).