Substrate floating potential characteristics in planar magnetron and ht sputtering systems

Abstract

The floating potential to ground Vfg and the ground to plasma potential Vgp have been measured in a magnetron and ht (non-magnetic field) sputtering system. A disc cathode (75 mm dia), made of Cu and water-cooled, was used for the ht arrangement and adapted to provide a magnetic tunnel field (200 gauss) of 52 mm mean dia and 10 mm width for the magnetron. The anode was grounded in both systems. The discharge vessel was evacuated by a silicone oil diffusion pump with an LN2 trap and an LN2 cold finger in the vessel; gas was admitted via an LN2 trap. Experiments were made at three pressures, 5, 50 and WO mtorr in Ar and air with two cathode/receiver gaps of 30 and 60 mm. Wire probes were placed at the work plane centre and edge of the electrode gap. The former was used to determine Vgp from V/I plots and the latter Vfg. The value of Vfg at the receiver centre was measured with a Cu-plate probe resting on a glass slide which simulated the conditions at the surface of a non-conducting substrate or thin film. The substrate floating potential to plasma Vfg was found from the addition of Vgp and Vfg both of which were negative. The difference between Vfg and Vgp is the voltage drop Vfg across an insulator or dielectric film resting on the anode support. For the applied voltage and current regions of 300–600 V and 11 mA cm−2 for the magnetron and 1–2 kV and 1–2 mAcm−2 for the ht system, Vgp was < −10 V and Vfg < −25 V. As the applied voltage was raised ⩾2 kV in the ht system and the pressure reduced < 50 mtorr Vfg rose sharply becoming proportional to the applied voltage and attaining nearly 1 kV as discharge blackout was approached. The negative bias was highest in air and at a small cathode/receiver gap. The high negative values of Vfg obtained in the ht system were attributed to negative charge accumulation from oxygen ions released from the cathode and accelerated by the electric field. Oxygen was present in the air discharge and probably as an impurity in the Cu-cathode. The effects are discussed of receiver bombardment by secondary electrons released from the cathode and accelerated in the cathode dark space. It is shown that negative bias from this source could not exceed −200 V, even in the complete absence of positive ion extraction.