A Combined Experimental and Theoretical Investigation of the Plasma-Surface Interactions in Plasma Deposition of Hydrogenated Amorphous and Nanocrystalline Silicon Films

化学反应气体等离子体被广泛用于蚀刻和沉积薄膜，并在微电子工业中实现了一整套技术。尽管这种等离子体的广泛应用和重要性，但等离子体工艺的优化和等离子体反应堆的设计在很大程度上依赖于反复试验。迫切需要从根本上了解等离子体物理、均相和非均相化学以及等离子体反应堆中的物质传输之间的复杂耦合。特别是，化学反应气体等离子体中产生的离子和自由基与暴露在放电中的表面之间的相互作用仍然是等离子体处理技术中最不被了解的方面。对表面反应机理和动力学知识的缺乏是等离子体反应器模型预测能力的主要限制，等离子体反应器模型旨在将等离子体物理与气相和表面化学相结合。提出了一种将等离子体和表面诊断与原子模拟相结合的研究策略，以提供关于从SiH4/H2/Ar辉光放电沉积氢化非晶和纳米晶硅薄膜过程中等离子体-表面相互作用的明确结论。由于硅薄膜沉积在半导体工业中的技术重要性，它被选为典型的化学工艺。这项拟议的研究旨在确定支配等离子体沉积机制的基本表面化学反应，确定相应的反应速率，并阐明这些表面动力学过程如何影响表面结构和成分的演变。只有通过实验和模拟结果的协同分析才能获得这样的知识。为此，将使用原子尺度的计算机模拟来研究来自等离子体的硅烷分子碎片、氢原子和高能离子与沉积表面的相互作用。为了详细研究等离子体-表面相互作用的机理，基于原子间势能函数的分子动力学、分子静力学和蒙特卡罗模拟器已经被开发出来，并通过与从头计算和实验数据的比较来评估它们的有效性。此外，还将使用密度泛函理论中的从头计算来生成精确的化学反应能面，并将使用变分速率理论来计算相应的反应速率。此外，还将实施混合非晶格动力学蒙特卡罗模拟，以在真实时间尺度上对等离子体沉积过程进行全尺度动力学模拟。这些计算研究将确定暴露在化学反应等离子体中的表面上发生的表面化学反应，定量分析这些反应的能量学和动力学，并阐明等离子体沉积机制的基本步骤。计算机模拟的结果将与实验数据进行比较，模拟得到的见解将用于指导新的实验研究和设计新的等离子体沉积策略。原位表面和等离子体诊断方法将用于研究薄膜生长过程中的气相和表面现象。表面诊断将包括原位衰减全反射傅里叶变换红外(ATR-FTIR)光谱和原位椭圆偏振光谱。ATR-FTIR将是我们实验计划的核心：PI开发了这项技术，以便在等离子体环境中研究表面物理和化学。ATR-FTIR将被用来确定生长表面成分作为入射到表面的物种的通量和能量的函数。等离子体气相诊断将包括各种光谱方法，如红外和可见光发射光谱，以及视距阈值电离质谱仪，以检测和测量撞击表面的自由基能量和通量。该研究首次将实验原位等离子体和表面诊断与原子尺度动力学建模和理论表面反应分析相结合，以建立对沉积表面与化学反应等离子体相互作用的基本机理和定量理解，以及这种相互作用在沉积过程中如何演变。此外，拟议的研究将为开发准确的化学反应数据库奠定基础，该数据库可用于设备规模的等离子体反应堆建模。考虑到该领域最近的实验和理论进展，进行这样一项具有挑战性的研究工作尤其及时。等离子体处理的科学基础是多学科的，跨越了不同学科之间的传统界限，包括物理、化学、化学和电气工程。因此，拟议的等离子体-表面相互作用的基础研究为培训学生和博士后学者利用综合的实验和理论方法解决重要的技术研究问题提供了理想的教育工具。

0078711AydilChemically reactive gas plasmas are used widely for etching and deposition of thin films and enable a whole class of technologies in the microelectronics industry. Despite the wide spread use and importance of such plasmas, optimization of plasma processes and design of plasma reactors rely heavily upon trial-and-error experimentation. There is a strong need for fundamental understanding of the intricate and complex coupling between plasma physics, homogeneous and heterogeneous chemistry, and species transport in plasma reactors. In particular, interactions of ions and radicals produced in chemically reactive gas plasmas with surfaces exposed to the discharge remain among the least understood aspects of plasma processing technologies. This lack of knowledge on surface reaction mechanisms and kinetics is a major limitation to the predictive capabilities of plasma reactor models that aim to integrate the plasma physics with gas phase and surface chemistry.A research strategy that integrates plasma and surface diagnostics with atomistic simulations is proposed to provide definitive conclusions about plasma-surface interactions during deposition of hydrogenated amorphous and nanocrystalline silicon films from SiH4/H2,/Ar glow discharges. Si film deposition is chosen as a prototypical chemical process because of its technological importance in the semiconductor industry. The proposed study aims at identifying the elementary surface chemical reactions that govern the plasma deposition mechanism, determining the corresponding reaction rates, and elucidating how these surface kinetic processes affect the evolution of the structure and composition of the surface. Such knowledge can only be achieved through synergistic analysis of the experimental and simulation results.To this end, atomic-scale computer simulations will be employed to study the interactions of silane molecular fragments, H atoms, and energetic ions from the plasma with the deposition surfaces. For detailed mechanistic study of plasma-surface interactions, molecular-dynamics, molecular-statics, and Monte Carlo simulators have been developed based on interatomic potential-energy functions, which have been tested exhaustively to assess their validity in comparison with ab initio calculations and experimental data. In addition, ab initio calculations within density functional theory will be used to generate accurate chemical reaction energy surfaces and variational rate theory will be employed to calculate the corresponding reaction rates. Furthermore, hybrid off-lattice kinetic Monte Carlo simulations will be implemented for full-scale dynamical modeling of the plasma deposition process over realistic time scales. These computational studies will identify surface chemical reactions that occur on surfaces exposed to a chemically reactive plasma, analyze quantitatively the energetics and kinetics of these reactions, and elucidate the elementary steps of the plasma deposition mechanism. The results of the computer simulations will be compared with experimental data and the insights gained from the simulations will be used to guide new experimental studies and design new plasma deposition strategies.In situ surface and plasma diagnostic methods will be used to study the phenomena occurring in the gas phase and on surfaces during film growth. Surface diagnostics will include in situ attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy and in situ spectroscopic ellipsometry. ATR-FTIR will be at the heart of our experimental plan: The PI's have developed this technique in order to study surface physics and chemistry in a plasma environment. ATR-FTIR will be used to determine the growth surface composition as a function of the fluxes and energies of species incident onto the surface. The plasma gas-phase diagnostics will include various spectroscopic methods, such as infrared and visible emission spectroscopy, and line-of-sight threshold-ionization mass spectroscopy to detect and measure the radical energies and fluxes impinging on the surface.The proposed research is pioneering in linking experimental in situ plasma and surface diagnostics with atomic-scale dynamical modeling and theoretical surface reaction analysis to establish fundamental mechanistic and quantitative understanding of deposition surface interactions with chemically reactive plasmas and how this interactions evolve during the deposition process. In addition, the proposed research will set the stage for developing an accurate chemical reaction database that can be utilized for equipment-scale plasma reactor modeling. Undertaking such a challenging research effort is particularly timely given the recent experimental and theoretical advances in the field.The scientific underpinnings of plasma processing are multidisciplinary and cut across traditional boundaries between different disciplines including physics, chemistry, chemical and electrical engineering. Thus, the proposed fundamental study in plasma-surface interactions provides an ideal educational tool for training students and postdoctoral scholars in addressing technologically important research problems using an integrated experimental and theoretical approach.***