Systematic Studies of Plasma Reactions on Dynamic Surfaces Using a Novel Rotating Substrate Principal Investigator: Vincent M. Donnelly Institution: University of Houston, Houston TX
A novel approach will be taken to study the important interactions of gaseous plasmas with plasma reactor walls. Plasma discharges are used to deposit thin films, and most importantly, to etch fine features in silicon integrated circuits. Control of these plasma etching processes will be critical in enabling future top-down nano-technology in the coming decades. Many of the scientific and technological challenges in understanding and controlling plasma etching and deposition processes involve the complex chemistry occurring at the plasma/reactor wall boundary. Reactant radical species such as chlorine atoms that are generated in the plasma by collisions of energetic plasma electrons with the process gas chlorine (Cl2) are required to promote selective and directional etching of electronic materials including silicon. These reactants reach a concentration in the plasma that is established in part by loss reactions at the plasma-wall boundary, which produce less reactive products. A good example would be the formation of Cl2 from combination of two Cl atoms. The rate of surface reactions depends on the nature of the surface, which in turn depends on the composition of the materials being etched and the duration of the process. Consequently, reactant concentrations drift over time, making it difficult to control plasma processes.
A basic knowledge of even simple wall reactions is mostly lacking, in large part because of the challenge in applying reliable analytical methods under hostile plasma conditions. Here we will bring the established surface-science diagnostic techniques of mass spectrometry and Auger electron spectroscopy to the plasma environment, allowing reactions at the plasma-wall boundary to be identified and studied. This is accomplished by substituting a cylindrical substrate for a small, hollow section of the reactor wall. A chamber housing these diagnostic tools is sealed to the other side of the reactor wall. The substrate is rotated rapidly; consequently, a portion of its surface is periodically in the plasma and then faces the diagnostic probe a very short time thereafter. Large vacuum pumps attached to the hollow plasma wall and on the diagnostics chamber remove the gas that leaks from the plasma. In this manner, the plasma gas and charged species are prevented from distorting the diagnostic methods, and the products formed on and evolving from the plasma wall can be isolated and identified. The time between plasma exposure and analysis can also be varied by changing the substrate rotation speed, allowing reaction rates to be extracted. The proposed work will yield basic knowledge of plasma surface interactions and will also provide critical information for improving control of advanced plasma processes that will be called upon for fabrication of future nano-devices.
The proposed work will provide rich scientific and educational payoffs, as well as technological advances. The basic knowledge that will emerge from these studies, and in particular, the new method for isolating such complex reactions has broad implications for and potential impact on diverse areas including wall reactions in plasma fusion reactors, catalysis, combustion, and atmospheric chemistry, as well as basic surface science. Several outreach activities are planned, including a collaboration with the Coalition for Plasma Science to increase public awareness for societal benefits of plasma processing of semiconductors.
