D. Olynick, D. Staaks, S. Rangnekar, D. Tierno, S. Dallarto
Molecular Foundry, LBNL,
Keywords: bit patterned media, plasma etching, nanomanufacturing
Summary:Requirements for bit patterned media (BPM) fabrication are pushing the limits of lithography and plasma etching. In the next 10 years, bit densities are slated to approach 10 TB/in2 with feature half-pitches close to 5 nm. For high throughput manufacturing, nanoimprint lithography will be utilized to pattern the disks. The nanoimprint template manufacturing process depends on a multitude of plasma etching steps culminating in quartz etching. Initial dimensions will be determined by the capabilities of block copolymer lithography while the ultimate dimensions will be defined using either double or quadruple patterning. Here, I will discuss our studies of the double patterning process. Figure 1 shows a schematic of a sample double patterning process. After creating the starting material stack, there are up to 8 etching steps and 1 critical deposition step. We have investigated low-temperature plasma etching and atomic layer etching for several steps. Benefits of cryogenic temperature plasma etching can include enhanced selectivity, better profile control, and novel passivant formation whereas atomic layer etching gives excellent thickness control. I will cover discuss cryogenic etching of chromium which can be used as a mask for the quartz template etching. In addition, I will compare oxide etching with atomic layer etching and cryogenic processes. Figure 2 shows etching rate studies of the chromium etching process. Spectroscopic ellipsometry was optimized to quickly collect data on these thin, absorbing metal films. Rates can be slowed and critical dimension controlled by working at the extremes of oxygen flow and lower temperature. Figure 3 shows 15 nm HP features transferred first into chromium and then oxide. The process appears scalable to smaller features. Both low temperature and atomic layer etching has been studied for the oxide patterning step. Ultimately, these studies shows that while small feature patterns are attainable but that significant R&D is critical for obtaining features down to 5 nm.