5.1 Atomic layer deposition  (Page 3/3)

ALD may be carried out in a vacuum system using an ultra-high vacuum with a movable substrate holder and gaseous valving. In this manner it may be also equipped with an in-situ LEED system for the direct observation of surface atom configurations and other systems such as XPS, UPS, and AES for surface analysis.

A lateral flow system may also be employed for successful ALE deposition. This uses an inert gas flow for several functions; it transports the reactants, it prevents pump oil from entering the reaction zone, it valves the sources and it purges the deposition site between pulses. Inert gas valving has many advantages as it can be used at ultra high temperatures where mechanical valves may fail and it does not corrode as mechanical valves would in the presence of halides. This method is based on the fact that as the inert gas is flowing through the feeding tube from the source to the reaction chamber, it blocks the flow of the sources. Although in this system the front end of the substrate receives a higher flux density than the down-stream end, a uniform growth rate occurs as long as the saturation layer of the monoformation predominates. This lateral flow system effectively utilizes the saturation mechanism of a monolayer formation obtained in ALE. Depending on the properties of the precursors used, and on the growth temperature, various growth systems may be used for ALE.

Requirements for ald growth

Several parameters must be taken into account in order to assure successful ALD growth. These include the physical and chemical properties of the source materials, their pulsing into the reactor, their interaction with the substrate and each other, and the thermodynamics and volatility of the film itself.

Source molecules used in ALD can be either elemental or an inorganic, organic, or organometallic compound. The chemical nature of the precursor is insignificant as long as it possesses the following properties. It must be a gas or must volatilize at a reasonable temperature producing sufficient vapor pressure. The vapor pressure must be high enough to fill the substrate area so that the monolayer chemisorption can occur in a reasonable length of time. Note that prolonged exposure to the substrate can cause the precursor to condense on the surface hindering the growth. Chemical interaction between the two precursors prior to chemisorption on the surface is also undesired. This may be overcome by purging the surface with an inert gas or hydrogen between the pulses. The inert gas not only separates the reactant pulses but also cleans out the reaction area by removing excess molecules. Also, the source molecules should not decompose on the substrate instead of chemisorbing. The decomposition of the precursor leads to uncontrolled growth of the film and this defeats the purpose of ALD as it no longer is self-controlled, layer-by-layer growth and the quality of the film plummets.

In general, temperature remains the most important parameter in the ALD process. There exists a processing window for ideal growth of monolayers. The temperature behavior of the rate of growth in monolayer units per cycle gives a first indication of the limiting mechanisms of an ALD process. If the temperature falls too low, the reactant may condense or the energy of activation required for the surface reaction may not be attained. If the temperature is too high, then the precursor may decompose or the monolayer may evaporate resulting in poor ALD growth. In the appropriate temperature window, full monolayer saturation occurs meaning that all bonding sites are occupied and a growth rate of one lattice unit per cycle is observed. If the saturation density is below one, several factors may contribute to this. These include an oversized reactant molecule, surface reconstruction, or the bond strength of an adsorbed surface atom is higher when the neighboring sites are unoccupied. Then the lower saturation density may be thermodynamically favored. If the saturation density is above one, then the undecomposed precursor molecules form the monolayer. Generally, ideal growth occurs when the temperature is set where the saturation density is one.

Advantages of ald

Atomic layer deposition provides an easy way to produce uniform, crystalline, high quality thin films. It has primarily been directed towards epitaxial growth of III-V (13-15) and II-V (12-16) compounds, especially to layered structures such as superlattices and superalloys. This application is due to the greatest advantage of this method, it is controllable to an accuracy of a single atomic layer because of saturated surface reactions. Not only that, but it produces epitaxial layers that are uniform over large areas, even on non-planar surfaces, at temperatures lower than those used in conventional epitaxial growth.

Another advantage to this method that may be most important for future applications, is the versatility associated with the process. It is possible to grow different thin films by choosing suitable starting materials among the thousands of available chemical compounds. Provided that the thermodynamics are favorable, the adjustment of the reaction conditions is a relatively easy task because the process is insensitive to small changes in temperature and pressure due to its relatively large processing window. There are also no limits in principle to the size and shape of the substrates.

One advantage that is resultant from the self-limiting growth mechanism is that the final thickness of the film is dependent only upon the number of deposition cycles and the lattice constant of the material, and can be reproduced and controlled. The thickness is independent of the partial pressures of the precursors and growth temperature. Under ideal conditions, the uniformity and the reproducibility of the films are excellent. ALE also has the potential to grow very abrupt heterostructures and very thin layers and these properties are in demand for some applications such as superlattices and quantum wells.

Bibliography

• D. C. Bradley, Chem. Rev ., 1989, 89 , 1317.
• M. Ritala and M. Leskel, Nanotechnology , 1999, 10 , 19.
• M. Pessa, P. Huttunen, and M. A. Herman, J. Appl. Phys ., 1983, 54 , 6047.
• T. Suntola and J. Antson, Method for producing compound thin films , U.S. Patent 4,058,430 (1977).
• M. A. Tischler and S. M. Bedair, Appl. Phys. Lett ., 1986, 48 , 1681.