Effects of High-Dose Silicon Implants

Underlying the software tools that make TCAD possible are physical models. Of course, the accuracy of these models directly determines the usefulness of the tools. Currently, the models used in circuit simulators like SPICE and device simulators like PISCES are well-developed and quite reliable. Unfortunately this is not the case for process simulators. There are several reasons for this. First of all, process simulation encompasses many different procedures, each of which involves different kinetics. Furthermore, within a particular process, the actual physical mechanisms may involve multiple variables which may have very different characteristics. For example, in diffusion, one may need to solve continuity equations not only for dopants, but for point defects as well, and the time scale for motion of the point defects are orders of magnitude smaller than for the dopants. This makes the solution for dopant diffusion much more complex and time-consuming. At Stanford, our approach has been to implement physically-based models into our simulators like SUPREM and PISCES. The success of TCAD as a design tool ultimately relies on the accuracy of the underlying models.

Robert Huang's research focuses on the physical interaction between defects generated by ion implantation and dopant diffusion. Using clever test structures fabricated in our own Integrated Circuits Laboratory, we have characterized the effect of implant dose, implant energy, anneal temperature, and anneal time on the diffusion of a buried boron marker layer[1],[2]. The results from our experiments are currently being used to evaluate whether or not the current equilibrium diffusion models in SUPREM-IV are adequate to describe implant-damage enhanced diffusion. This research is of particular importance since dopants themselves are typically introduced into the wafer through ion implantation, thereby introducing damage which not only affects their own diffusion but also the diffusion of any previously introduced dopant in the substrate. The goal of this research is to characterize the effects of implant damage on dopant diffusion, to implement new diffusion models if necessary into SUPREM-IV, and to develop processes which will limit the detrimental effects of implant-damage enhanced diffusion.

Ref. 1 R. Y.S. Huang and R. W. Dutton, "The Effect of Amorphizing Implants on Boron Diffusion," Proceedings of the Third International Symposium in Process Physics and Modeling in Semiconductor Technology, ed. by G. R. Srinivasan, K. Taniguchi, and C. S. Murthy, pp. 46-53, May 19-21, 1993.

Ref. 2 Robert Y.S. Huang and Robert W. Dutton, "The Effects of High-Dose Silicon Implants on Boron Diffusion," TECHCON '93 Extended Abstract Volume, pp. 59-61, Atlanta, GA, September 28-30, 1993.

Robert Huang (huang@gloworm.Stanford.EDU)
AEL 231
Integrated Circuits Laboratory
Stanford University
Stanford, CA 94305-4055