Metalorganic vapour phase epitaxy
The metal- organic vapor phase epitaxy (german metal organic chemical vapor phase epitaxy MOVPE, also organo- metallic vapor phase epitaxy OMVPE ) is an epitaxial process for producing crystalline layers. It is identical to the metal organic chemical vapor deposition ( engl. metal organic chemical vapor deposition, MOCVD ) with respect to the equipment used, the terms MOVPE, MOCVD and OMVPE be used in compound semiconductor area usually for the same processes. The MOCVD deposition refers to any of the method, the MOVPE and OMVPE epitaxy only, that is, the ( a ) crystal growth on a crystalline substrate. In contrast to molecular beam epitaxy (MBE) is the growth of the crystals are not in high vacuum but in the low vacuum ( 20 to 1000 hPa) instead.
MOVPE is the most important manufacturing method of group III-V compound semiconductor, in particular for gallium nitride ( GaN) based semiconductor, which today represents the most important basic material for blue, green and white LEDs.
As the starting materials for compound semiconductors are often metals, they can not be introduced at low temperatures in the elemental form in the gas phase. Therefore, the starting materials in the form of organometallic compounds (eg trimethylgallium ) and hydrides provided ( eg, ammonia, phosphine, arsine ) are available at this epitaxy. The advantage of these compounds is a moderate vapor pressure at room temperature so that it evaporates close to standard conditions and can be transported through pipelines.
The organometallic compounds are used in so-called bubblers kept ( by the construction principle, gas washing bottles) and form in a saturated vapor above the liquid or solid, which (usually hydrogen or nitrogen used to be argon) into the reaction chamber (reactor ) is transported by a flowing carrier gas. The bubblers are used in thermostats, by which they are kept at a constant temperature in order to obtain a defined constant vapor pressure of Metallorganikums. About the total pressure and the flow rate of the carrier gas, the molar flow of Metallorganikums can be determined:
The overall reaction formula of trimethylgallium ( CH3) 3Ga and ammonia NH3 in the growth of gallium nitride can be used as
Be written. However, this reaction is an oversimplification of the actual situation before and during the crystal growth. To find pre-reactions between the starting materials rather than in the gas phase, and often form unreactive adducts. The possible individual reactions are varied depending on the nature of the starting materials and carrier gases used, and can be predicted only vague partly because of the elusive catalytic properties of the different surfaces of the MOCVD reactor.
By the operating principle, the presence of large amounts of carbon and hydrogen, small amounts of these materials will be always built with the semiconductor crystal. Hydrogen passivated often necessary for the p- type conduction acceptors, but can usually simply by annealing in an inert atmosphere or in a vacuum to remove. Carbon is usually not disruptive and is used in gallium arsenide growth target for p-type doping.
The layer growth, the reactants diffuse from the gas stream to the substrate surface, where the mounting takes place in the crystal. At low temperatures, the incorporation of the reactants is determined by their decomposition. This is the kinetically controlled region. Since the decomposition of the starting materials or surface reactions have an exponential dependence on the temperature, the growth rate is strongly temperature dependent in this area and therefore difficult to control. At higher temperatures, the growth again, so is limited by the supply rate of diffusion. However, the diffusion is not temperature dependent in the first approximation. Therefore, it is usually carried out in the diffusion- controlled area. At higher temperatures occur enhances the growth inhibitory prereactions on or is the vapor pressure of the semiconductor so high that the growth rate is reduced again (desorption). This reduction in growth rate also has an exponential dependence on temperature. Therefore, this area is also difficult to control and is avoided.
The surface processes during growth play a more decisive role. The processes can be divided into the transport of reactants to the surface, chemical reactions and adsorption on the surface, surface kinetic processes and desorption as well as the transport of the reactants. The aim is, as already mentioned above that the growth diffusion- limited, so take place transport limited. Growth is then only limited by the diffusion of the raw materials to the substrate, or by the removal of the products from the substrate. In the kinetically defined area may be disabled, for example, the desorption of the reaction products. Then the remaining reactants can be expected with increased carbon incorporation due to the incomplete transported away. It is also important to achieve a step-growth sufficient mobility of the starting materials on the surface of the normally desired smooth surfaces.
Essential to the growth process is in addition to the temperature and the total pressure in the reactor the partial pressure of the reactants employed and the partial pressure. This is, inter alia, for determining the stoichiometry and the growth mode, i.e., whether an island growth or step-growth takes place. Thus, for these parameters, the growth rates of different crystal facets but also the incorporation of impurities influence. If a strained ternary layer grown beyond so this may vary depending on material combination and growth parameters in the Frank - van der Merwe mode as a two-dimensional layer in the Stranski - Krastanow growth mode and wetting layer (german wetting layer) followed by three-dimensional island growth or directly grown as three-dimensional islands in Volmer - Weber growth mode. Taking advantage of the Stranski - Krastanow mode are nowadays often self-organized quantum dots, preferably grown in the system In ( Ga) As / GaAs for applications such as quantum dot lasers.
Pros and cons of MOVPE
With the MOVPE is important for the functioning of the components semiconductor crystal layers grow reproducible to less than a monolayer accurate ( <2.5 Å). Typical growth rates of 0.1 nm / s to 1 nm / s and thus is higher than that of the MBE. The method was developed in the 1980s, especially because of the easy way phosphorus- based semiconductor crystals, such as indium phosphide conveyed to grow. Until then, the MBE is not or only partially this was possible. In the 1990s, the blue LED was the realization on the basis of gallium nitride and to a lesser extent by the growing market for GaAs and InP -based devices for dispersion- or low-attenuation data communications around 1310 and 1550 nm using fiber optic cable and microwave applications for mobile phones and military applications ( radar) a boom for the MOVPE technique triggered. Especially GaN can not be with the MBE in sufficient quality and quantity of produce for LEDs. Due to the easy scalability of the systems and processes ( from simple 2- inch single wafer systems up to 95 × 2-inch or 25 × 4- inch wafers ), it is ideal for mass production. By eliminating high vacuum equipment, as required in the MBE, MOVPE technique is relatively inexpensive and easy to maintain.
Main cost factors are the expensive high-purity starting materials and the low compared to the MBE material efficiency. By working with compounds of elements always a relatively large amount of impurities (C, O, H ) is incorporated into the crystal, in contrast to the MBE and can therefore not as pure semiconductor crystals such as the MBE prepared.