Collaborative Research: Atomic Displacement Engineering of Post-epitaxial Thin-films (ADEPT)

非技术描述：现代外延半导体技术的出现导致了光电子和电子器件制造的范式转变。具体地说，将半导体层厚度控制在单个原子层内，并在两种不同的半导体材料之间创建多个突然的界面，从而实现了一系列结构，这些结构是现代激光器、探测器和晶体管中的关键元件。然而，到目前为止，对异质外延半导体组件中化学成分的精确控制仅限于增长方向-半导体结构演变的下一次飞跃将是实现对所有三个维度的成分的精确控制，为量子计算和密码学等新技术开辟光电子基石的道路。该项目试图通过利用纳米压痕产生的机械力来横向控制用现有外延技术制造的半导体异质结中的化学成分来实现这一目标。该项目将纳米尺度的化学成分和结构的实验测量与原子扩散的多尺度计算机模拟相结合。该团队使用部分由专门定制的实验提供信息的计算机模型，探索了广泛的参数空间，并确定了提高效率和有力地推动原子扩散所需的必要操作条件。如果成功，该项目的结果可能为创造具有广泛潜在应用的新型半导体结构铺平路线图。该项目作为一个多学科之家，为两个参与机构的研究生和本科生提供培训。针对代表性不足群体的本科研究机会是一个特别强调的问题，在新墨西哥大学确定的学生被招募到宾夕法尼亚大学的暑期研究机会中。技术描述：在这个项目中，研究团队调查了一种新的战略，以横向定义外延生长的III-V半导体系统中的纳米级组成模式。这种多步骤的策略被称为后外延薄膜的原子置换工程，或ADEPT，它与传统的异质外延方法相结合，可能为创造高度可控的三维量子势垒配置提供一条实用的途径。ADEPT方法使用空间图案化应力场和热退火来驱动由两个不同大小的可移动原子组成的化合物半导体合金薄膜中的扩散，并对弹性应力进行扩散响应。通过将可重复使用的、预先图案化的纳米柱阵列压在半导体衬底上来施加应力场。值得注意的是，“压印”策略不依赖于难以控制的亚相的形核和生长，适用于存在两个不同大小的可移动原子物种的任何材料体系。该团队最近在SiGe中演示了这种熟练的方法，并在此研究了其在III-V异质外延系统中的更广泛应用，即InGaAs和GaAsSb，这两种系统都在InP上得到支持。与SiGe不同，这些材料中的原子扩散率没有得到充分的表征，特别是作为温度和弹性应力的函数，从而阻碍了预测性建模和充分探索熟练过程的能力。该研究活动试图通过两步计算机辅助的实验设计方法来克服大的工艺参数空间和有限的实验吞吐量所带来的障碍。在第一阶段，使用一系列相对“简单”的实验来开发数据辅助模型，这些实验专门用来为III-V材料中互扩散的参数化模型生成信息。在第二阶段，参数化模型被用来探索多维参数空间，并确定合适的条件来执行熟练过程的实验演示。这一奖项反映了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。

Nontechnical Description: The advent of modern epitaxial semiconductor technologies has resulted in a paradigm shift in optoelectronic and electronic device fabrication. Specifically, the control of semiconductor layer thickness to within a single atomic layer and the creation of multiple, abrupt interfaces between two different semiconductor materials have led to the realization of a host of structures that are key elements in modern lasers, detectors and transistors. However, to date, precise control of chemical composition in heteroepitaxial semiconductor assemblies has been limited to the growth direction - the next leap in the evolution of semiconductor structures will be to realize precise control of composition in all three dimensions, opening routes to optoelectronic building blocks for new technologies, such as quantum computing and cryptography. This project seeks to achieve this goal by using mechanical forces induced by nano-indentation to laterally control chemical composition in semiconductor heterostructures created with existing epitaxial technology. The project integrates experimental measurements of chemical composition and structure at the nanoscale with multiscale computer modeling of atomic diffusion. Using computer models that have been partially informed by specifically tailored experiments, the team explores a broad parameter space and identifies the necessary operating conditions needed to efficiency and robustly drive atomic diffusion. If successful, the outcome of this project could pave the roadmap for creating a new class of semiconductor structures with broad potential applications. The project serves as a multidisciplinary home for training graduate and undergraduate students at both participating institutions. Undergraduate research opportunities for underrepresented groups is a special emphasis whereby students identified at the University of New Mexico are recruited into summer research opportunities at the University of Pennsylvania.Technical Description: In this project, the research team investigates a new strategy for laterally defining nanoscale compositional patterns in epitaxially-grown III-V semiconductor systems. This multistep strategy is referred to as Atomic-Displacement Engineering of Post-epitaxial Thin-films, or ADEPT, which, coupled with traditional heteroepitaxial methods, may provide a practical pathway for creating three-dimensional quantum barrier configurations with a high degree of controllability. The ADEPT approach employs spatially patterned stress fields and thermal annealing to drive diffusion in a compound semiconductor alloy film comprised of two mobile atomic species with different sizes and diffusional responses to elastic stress. The stress fields are applied by pressing a reusable, pre-patterned array of nanopillars against the semiconductor substrate. Notably, the "press-and-print" ADEPT strategy does not rely on nucleation and growth of sub-phases that are difficult to control and is applicable to any material system in which two mobile atomic species of different sizes are present. The team has recently demonstrated the ADEPT approach in SiGe and here investigates its broader application to III-V heteroepitaxial systems, namely InGaAs and GaAsSb, both supported on InP. In contrast to SiGe, atomic diffusivities in these materials is not fully characterized, particularly as a function of temperature and elastic stress, preventing predictive modeling and the ability to fully explore the ADEPT process. The research activity seeks to overcome the obstacles posed by a large process parameter space and limited experimental throughput with a two-step, computer-aided experimental design approach. In the first stage, data-assisted models are developed using a sequence of relatively "simple" experiments specifically designed to generate information for parameterizing models for interdiffusion in III-V materials. In the second stage, the parameterized models are used to explore the multidimensional parameter space and identify suitable conditions for performing experimental demonstrations of the ADEPT process.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.