To establish and maintain contact with the mainstream of processing science, NRL should form collaborations with industry, universities, and other laboratories. NRL has had experience with setting up Cooperative Research and Development Agreements and could use these as mechanisms for accomplishing the goal of increased collaboration. (Chapter 2, Chapter 3, Chapter 4, Chapter 5, Chapter 6, Chapter 7 through Chapter 8 identify current academic, industrial, and federal laboratory efforts in the field.)
It is the judgment of the panel that NRL has sufficient facilities in place to enable it to play a key role in many important issues pertaining to the plasma processing of semiconductors in ultralarge-scale integrated manufacturing. It is also the panel's judgment that the first priority for NRL should be to intensify interactions with the outside world to understand the current issues and to gain insight on where to focus its efforts. NRL personnel should increase their involvement with the materials processing community by attending relevant conferences, presenting contributed papers, and seeking invitations to give invited papers.
The panel has the following specific conclusions on how NRL's capabilities could support the research program recommended above.
The chemical vapor processing group is currently the best known and most active one in this field. Although it would be tempting to build the NRL materials processing effort around this group, a diversified program should not be dominated by any one group. This group should be one of several strong and active groups.
This same group can provide in-depth treatment of surface chemical and gas-phase chemical issues. For example, selectivity is a key issue in gate conductor etching, and the group may be able to provide novel insights.
A sophisticated diagnostics laboratory set up as a user facility will probably not attract interest from industry. Equipment manufacturers tend to do their testing in-house, partly to protect proprietary information.
The NRL microfabrication facility can be used to process special devices such as measurement structures needed for sidewall passivation studies. NRL should not try to upgrade this facility, since it is highly unlikely that it can ever become a state-of-the-art fabrication line.
The current emphasis on electron-cyclotron resonance (ECR) should be continued as an initial phase of NRL's entry into plasma processing. To complement the NRL diamond deposition work, it may make sense to focus on plasma-enhanced chemical vapor deposition for the present. NRL's effort is not big enough to be spread over many different applications. Its strongest asset seems to be its high degree of complementarity. This could be leveraged best by focusing on one key problem to make a significant impact.
The ECR work of the ion/plasma processing group on diamond deposition is an ideal complement to the diamond work in the chemical vapor processing group. The ECR tool can be easily adapted to other materials and should be able to make significant contributions rapidly. At a later time, the use of alternate high-density plasma tools may be studied for diamond film deposition.
The real-time diagnostics used in the Electronics Division could be important for the understanding of dry etch damage issues. Dry etch damage is clearly an important issue in gate conductor etching. This effort would benefit enormously by becoming part of a well-aimed overall effort on optical diagnostics, working together with the chemical vapor processing group.