Shinku

Show abstractHide abstract

The dissipation of rf-power in a planer diode type rf-sputtering arrangement was measured experimentally, and the results were compared with numerical calculations based on a simple theory related to the physical phenomena in the sputtering. The quantity of rf-power dissipation divided in to the target, the substrate table and the vacuum vessel wall were obtained, respectively, by measuring the temperature rise of cooling waters and their flow rates.
For a copper target plate, the power dissipation at the electrode was much higher than that at the substrate table because of the good thermal contact between the cooling water and the target. On the other hand, when a quartz plate was used, the temperature of the substrate table depended markedly on the thermal contact of the plate with the rf-electrode; for a poor contact, the power dissipation at the rf-electrode was less than that at the substrate table, which meant that the target plate was heated to a higher temperature giving rise to radiation heating of the substrate table. However, with a intimate contact, the power dissipation at the electrode side was larger than that at the substrate table, resulting in the lower substrate temperature.
From these experimental observation and the theoretical approach, it is concluded that the primary factor to affect the temperdture of the substrate is the heat radiation from the target plate and that the second is the secondary electron bomardment from the target. The lowest power dissipation at the substrate table so far obtained was 18% of the total rf-power input with a direct water-cooling type copper target.

View full abstract

Download PDF (1230K)