Plasma Deposition and Polymerization
In addition to being essential to the etching process, high-density plasmas at somewhat higher pressures are also needed for deposition of insulating layers and metal contacts in the processing of semiconductors. In the deposition of SiO2 and Si3N4 dielectric layers, the same problems of oxide damage, high-aspect-ratio trenches, and RIE lag (the influence of neighboring structures) are likely to be encountered as in the etching process. A new problem is the formation of dielectrics of desirably low permittivity (ε < 2.5); no suitable material has yet been found. This presents an opportunity for advanced research, perhaps yielding a new dielectric material that can be formed only in a plasma environment.
The formation of barrier coatings, for instance, in automobiles, food packaging, or pharmaceutical capsules, has great potential as a widespread plasma application. In food containers, such as potato chip bags or plastic soft drink bottles, the problem is to prevent leakage of O2 into the container or CO2 out of it. This can usually be done by fluorinating the surface using CF4 plasma. Treatment is also required to improve the adhesion of paint so that printing can be done on plastic containers. Reliability and aging are also improved. Examples from the automobile industry include gas tanks and bumpers. Plasma treatment of the inside of a plastic gas tank with CF4 can slow the permeation of gasoline through the walls. However, methanol cannot be contained as easily, and, should there be a trend toward the use of methanol mixtures, a new material or process would have to be developed. Problems that could lead to fruitful lines of research include (1) efficient creation of suitable plasmas, including line sources that could process continuous webs of material; (2) plasma production in complicated geometries, such as the inside of a plastic bottle; and (3) the development of new materials that would make plastics more easily recyclable.
Plasmas are commonly used to treat textiles and lignocellulosics (paper and wood) to improve their wettability, dyeability, adhesion, or optical properties, such as ultraviolet transparency. For instance, paper towels or diapers can be treated to improve their water absorption, filter paper to decrease water absorption, and synthetic or wool fabrics to improve the adhesion of dyes. The process involved is not one of coating but of actual chemical surface modification; for instance, the addition of a hydroxyl (OH) group will usually improve wettability. Plasmas of hexymethyldisulfoxide (HMSDO) or tetramethyl tin (TMT) are commonly used for this purpose. The physical and chemical processes occurring on the fiber surfaces are not well known, and there are opportunities for research on the composition of the precursor radicals formed in the plasma using such standard diagnostics as electron spectroscopy for chemical analysis. Once the mechanism is better known, the industrial tools for treating such materials can be optimized for speed, economy, and rate of degradation. The scalability of the process to handle large volumes needs to be shown, as well as the advantage of plasma processing over wet chemistry in terms of the volume of liquid waste generated.
Optical Coatings and Photonics
Plasmas have been used for some time to deposit coatings on optical elements such as lenses and filters, as well as for more mundane applications such as reflectors on automobile bumpers or highway signs. A newer use under development is the manufacture of multilayer optical fibers. Furthermore, a more glamorous and potentially more important application looming on the horizon is the fabrication of integrated circuits containing photonic elements—a necessary step in the move toward optical computing. To handle photons on a chip, a polymer such as polymethylmethacrylate (PMAA) can be spin-coated onto a patterned silicon wafer, and plasma deposition can be used to create a graded-index structure that can then be patterned and etched. This is obviously a new research direction with many problems to be overcome but with a large payoff.
One of the most commonly used plastics is methyl methacrylate (MMA), commonly known as Plexiglas or Lucite. In this material, the long polymer chains are arranged in a linear (not cross-linked) manner, resulting in a comparatively weak material. On the other hand, plasma polymerized methyl methacrylate (PPMMA) has a dense, highly cross-linked structure and can be used to strengthen the surfaces of plastic containers, textiles, or even metal automobile parts. This can be done by exposing the surface to an MMA or HMDSO plasma. Another advantage of the plasma polymerization process is the retention of the original monomer structure, whereas other processes tend to change it. It is not understood how the cross-linked structure is formed, what the precursors are, and what plasma parameters will optimize the process. Though there is not yet any commercial application of this process, it is clear that further research may lead to the improvement of many manufactured products.
A ROLE FOR NRL
Deposition of new materials is a suitable use for the excellent laboratory ECR equipment in NRL's ion/plasma processing group. As mentioned above, the development of a low-permittivity dielectric for integrated circuits is a challenge—such a development would be a significant contribution to ULSI technology. Formation of III-V compounds such as GaN, using ECR as a deposition source, may well be a rich field of investigation. Materials developed here could then be incorporated into unique circuits made in the fabrication line of the Surface and Interface Sciences Branch in the Electronics Division. Production of such one-of-a-kind devices could justify the maintenance, though not the upgrading, of a fabrication facility at NRL.
The extensive diagnostics capabilities at NRL in the Chemical Vapor Processing Group and the Surface and Interface Sciences Branch laboratories could be applied to the problem of understanding the precursor radicals in the deposition and polymerization processes. Though it has great potential, plasma polymerization is not well understood, and NRL 's personnel are capable of making significant progress in this direction also.