Skip to content

Dynamic Reverse Pulsing – What About Duty Cycle?

Posted October 16, 2023 by Gayatri Rane

A previous blog showed that Dynamic Reverse Pulsing (DRP) mode offers several benefits to reactive sputtering of SiO2 through the reduction of substrate heat load by about 12% and delivered a 10% increase in deposition rate when compared to conventional bipolar pulsed (BP) dual-magnetron sputtering. DRP mode halves the power applied to each magnetron and shares the pulsing with an explicit anode, as opposed to bipolar mode in which the two targets alternate as cathode and anode with a 50/50 duty cycle (Figure 1). DRP mode maintains a high duty cycle such that the polarity is reversed on the anode only for a short time, about 5 to 30% with the goal being to sufficiently discharge target electrical charge buildup. 

A diagram of a power supply system

Description automatically generated
Figure 1: DRP vs Bipolar Pulsed Mode

Studies indicate lower plasma spread, resulting in reduced heating, towards the substrate with DRP has been attributed partly to the effect of the magnetic field on electron movement and plasma confinement. In Bipolar (BP) mode, where the targets alternate as anode and cathode, the anode is magnetic so that electrons are magnetically shielded and can reach the anode only along field lines. The secondary electrons emitted from the cathode are, therefore, concentrated in the cathode vicinity increasing the plasma density in front of the substrate. In contrast, in the case of DRP mode, if the non-magnetic anode intersects with the magnetic field line of the target, it will effectively collect the electrons so that they are not lost to the plasma and thus do not contribute to substrate heating1.

The roles of both the duty cycle and the anode are of particular interest when it comes to DRP mode. We studied the effect of different duty cycles on process and film properties. Tests were performed in an industrial drum coater with dual rotary targets and an anode placed between the targets, as shown in Figure 2.