Resolving Abnormal Target Erosion in High Frequency Magnetron Discharge

Michigan State University, East Lansing MI

Investigators

Abstract

Radio frequency (RF) magnetron sputtering is an essential technology for manufacturing a broad variety of high-quality thin films. The ceramic targets, the source of the thin film coating material, consumed during the RF sputtering process are expensive, thus efficient target consumption is highly desirable in lowering manufacturing costs. Incomplete understanding of the ceramic target etching mechanisms occurring during RF magnetron sputtering results in poor target use rates (~30%). This award supports fundamental research aimed at understanding RF magnetron discharges and how they induce target consumption. New knowledge obtained on the complicated RF discharge processes will contribute to plasma science and technology, and will ultimately provide guidance on efficient RF magnetron design. Results from the research are expected to double the use rates of target materials. This will have a positive impact on many industrial fields including large-area optical coatings, energy storage, displays, solar energy, and semiconductor devices. The annual market of these fields has reached over $500 billion and is growing continuously. This project will strengthen university-industry collaborations and economic competitiveness of the U.S. by greatly reducing manufacturing costs of thin films. It will also contribute to workforce development by attracting and training graduate and undergraduate students in science and engineering. RF magnetron sputtering is essentially determined by the non-uniform electromagnetic fields that vary with time and in space. Charged particles that lead to the sputtering of the target undergo complicated interactions with the non-uniform fields. Understanding high-frequency plasma discharges under the confinement of magnetic fields is a challenge and holds strong scientific appeal. Despite the scientific community's strong interest in high-frequency plasma discharges, knowledge in RF magnetron sputtering is very limited, which results in unsatisfactory RF magnetrons. In fact, current magnetrons are designed for achieving optimum performance in direct current (DC) sputtering, which is used for target materials that are electrically conducting like metals. When this same magnetron is used for RF sputtering of insulator targets, it is extremely ineffective and produces an abnormal erosion profile completely different from that in DC sputtering. Specifically, the intensively etched region under DC sputtering is the least eroded under RF sputtering. To understand the mechanisms of the abnormal RF target erosion and fill the knowledge gap, the research team hypothesizes a 'localized charge effect', which assumes that electrons in RF magnetron discharges preferentially accumulate and stay immobile on the insulator target surface near the magnet poles to attract ions to sputter these regions. This hypothesis is unexpected from the existing knowledge on magnetron sputtering and is thus potentially transformative. The research includes two major tasks: 1) modeling of RF magnetron discharges, and 2) verifying the 'localized charge effect' by designing a highly efficient RF magnetron through modeling and experimentally testing.