Plasma Source Ion Implantation [1] (PSII) is a non-line-of-sight process for implanting energetic ions into solids. Ion implantation is used to improve the wear, corrosion, and fatigue resistance of metals and in the fabrication of semiconductors. In PSII, targets are placed directly in a plasma source chamber and are then pulse-biased to high negative voltages. During this pulse, electrons are repelled from the target on the time scale of the electron plasma frequency (which is short relative to the ion motion), leaving behind an ion matrix sheath. This sheath then expands and the ions uncovered by the sheath accelerate into the target with roughly normal incidence on all sides. This minimizes the sputtering of the target, thus maximizing the retained dose.
One important parameter of the PSII process is the target temperature reached during the surface modification. The desire to maximize throughput in the chamber often produces a desire to maximize the power to the target, allowing rapid achievement of high fluence. This can produce large target temperatures, which may or may not be desirable. High temperatures are desirable, for instance, in cases where diffusion of the implanted species is desired. Other investigators [2] have taken advantage of this concept to increase the modified layer in a nitrogen implant treatment of a metal. On the other hand, high temperatures are undesirable in many cases, such as for materials which temper at relatively low temperatures. For these reasons, it is desirable for process designers to be able to predict target temperatures for a wide variety of implant materials, geometries, and power densities.
An obvious solution to the issue of target temperature control is on-line measurement of the temperature. This is not as easy as one may imagine, because thermocouples cannot easily be used in a PSII chamber; they cannot cope with the pulsed bias on the target. Therefore, optical measurement is the only realistic option for on-line temperature measurement. There are several research groups who use this method, adjusting the power density to achieve a desired temperature history. For many groups, though, this is not done because optical measurement of temperature can be expensive and it only yields surface temperatures. Hence, it is still desirable to have the capability of predicting the temperatures of PSII-implanted targets.
Previous authors (notably Parry [3]) have formulated models for conventional ion implantation, but the pulsed nature of PSII requires additional considerations. In this paper a numerical model for predicting PSII target temperatures is developed. Some simple analytical models, primarily concerned with the effects of pulsing, are used to justify the assumptions of the numerical model. The results of this model are then used to predict the target temperature of a Ti-6Al-4V artificial knee joint implanted with nitrogen ions. The comparison is excellent.