Ph.D. Student: Carlos Kazuo Inoki, M.Sc.
Advisor:J. A. Zuffo/A. C. Seabra
Optical lithography is a standard tool to transfer patterns for the microelectronics industry. Its major drawback is the ultimate resolution that is limited by the light wavelenght (UV or DUV). The electron beam lithography (e-beam) has high resolution capability but it has other problems. One important problem comes from the backscattered electrons that create a proximity effect on exposed areas with high density of structures. There are two approachs to control this problem, one is take a careful calculatation of the energy dose necessary to shape the right pattern, and another is to minimize this proximity effect [1]. This last approach can be done using electrons with low energy (below 2.0 kV) for the exposure [2]. This range of energy has some advantages, an improved sensitivity of the resist and a low probability of inducing damages on the device. Another aspect is that with the development of micromachining technologies it's possible to build small e-beam columns on a silicon waffer (centimeter size). An array of microcolumns can work as an array of small beam writers working in parallel [3]. These arrays work with low energy electrons (10-1000 V). One problem with low energy e-beam lithography is the low penetration depth of the electrons. For an energy of 1.0kV, nearly 92% of all energy is absorbed by the first 500Å , that results on an surface imaging [4]. At our laboratory we have a Phillips scanning electron microscope (SEM) model 515 adapted for low energy (0.3-30.0 kV) modified with a Raith ProxyWriter system to work as an e-beam writer. A SEM can resolve a very small spotsize, typically tenths of nanometer or better. The limitation that comes from this system is the low beam current (on typical working condition its current is in the range of a few tenths of pA) and consequently low throughput for the lithography. The main application of this system is to produce prototypes of semiconductor devices in a mix & match lithography process. Another point is to explore its capability to work on a low energy range (1.0kV) to perform studies of resists. This research is connected with other efforts that studies the resist sylilation process.
[1] R. Rau, J. H. MacClellan, and T. J. Drabik, J. Vac. Sci. Technol. B14, 2445 (1996).
[2] R. B. Houli, V. Umansky, and M. Heiblum, Semicond. Sci. Technol. 8, 1490 (1993).
[3] C. W. Lo, M. J. Rooks, W. K. Lo, M. Issacson, and H. G. Craighead, J. Vac. Sci. Technol. B13, 812 (1995).
[4] P. A. Peterson, Z. J. Radzimski, S. A. Schwaim, and P.E. Russel, J. Vac. Sci. Technol. B10, 3088 (1992).