Simulation of Chemical Vapor Deposition Processes

Abstract

Chemical vapor deposition (CVD) processes constitute an important technology for the manufacturing of thin solid films. Applications include semiconducting, conducting, and insulating films in the integrated circuit industry, superconducting thin films, antireflection and spectrally selective coatings on optical components, and anticorrosion and antiwear layers on mechanical tools and equipment. Compared to other deposition techniques, such as sputtering, sublimation, and evaporation, CVD is very versatile and offers good control of film structure and composition, excellent uniformity, and sufficiently high growth rates. Perhaps the most important advantage of CVD over other deposition techniques is its capability for conformal deposition—i.e., the capability of depositing films of uniform thickness on highly irregularly shaped surfaces. In CVD processes a thin film is deposited from the gas phase through chemical reactions at the surface. Reactive gases are introduced into the controlled environment of a reactor chamber in which the substrates on which deposition takes place are positioned. Depending on the process conditions, reactions may take place in the gas phase, leading to the creation of gaseous intermediates. The energy required to drive the chemical reactions can be supplied thermally by heating the gas in the reactor (thermal CVD), but also by supplying photons in the form of, e.g., laser light to the gas (photo CVD), or through the application of an electrical discharge (plasma CVD or plasma enhanced CVD). The reactive gas species fed into the reactor and the reactive intermediates created in the gas phase diffuse toward and adsorb onto the solid surface. Here, solid‐catalyzed reactions lead to the growth of the desired film. The application of new materials, the desire to coat and reinforce new ceramic and fibrous materials, and the tremendous increase in complexity and performance of semiconductor and similar products employing thin films are leading to the need to develop new CVD processes and to ever increasing demands on the performance of processes and equipment. This performance is determined by the interacting influences of hydrodynamics and chemical kinetics in the reactor chamber, which in turn are determined by process conditions and reactor geometry. It is generally felt that the development of novel reactors and processes can be greatly improved if simulation is used to support the design and optimization phase. In the last decades, various sophisticated mathematical models have been developed, which relate process characteristics to operating conditions and reactor geometry. When based on fundamental laws of chemistry and physics, rather than empirical relations, such models can provide a scientific basis for design, development, and optimization of CVD equipment and processes. This may lead to a reduction of time and money spent on the development of prototypes and to better reactors and processes. Besides, mathematical CVD models may also provide fundamental insights into the underlying physicochemical processes and may be of help in the interpretation of experimental data.

Multi-scale modeling of chemical vapor deposition processes for thin film technology

“…A full description of the model equations used in CVD gas flow and transport modeling can be found in Refs. [19][20][21]. In general, such models include:…”

“…[22,[31][32][33][34]. In the absence of experimental data, thermo-chemical properties may be obtained from ab initio molecular structure calculations [35], and transport properties may be calculated from kinetic theory [19,33,36]. The required Lennard-Jones parameters for many CVD gases can be found in e.g.…”

“…Refs. [19,22,26,32,33]. From the transport properties of its constituent species, the properties of a gas mixture can be calculated using semi-empirical mixture rules [19,37].…”

Ein Mehrskalenansatz zur numerischen Simulation und Analyse der metallorganischen Gasphasenepitaxie

“…Bei der Erprobung von solchen Verfahrensvarianten wird derzeit auf "trial and error"-Methoden vertraut, d. h. es werden bestehende Anlagen an die sich verändernden Prozesse angepasst [2] [7] und den rasanten Zuwachs an Computerleistung können heute auch größere Systeme und Reaktionsnetzwerke mit guter Genauigkeit "ab initio" berechnet werden [8].…”

Dimensionless Numbers Expressed in Terms of Common CVD Process Parameters

“…No turbulence is expected for a Grashof number well below lxl0 +5 [13]. The ratio Gr/Re x is often used to describe the occurence of buoyancy effects.…”