Researchers Apply Self-Assembly to NIL Master for Bit-Patterned Media
Researchers working at Hitachi Global Storage Technologies and the University of Wisconsin have applied a self-assembly approach that would sharply reduce the time required to write NIL templates. The HDD industry is moving to bit-patterned media, and the self-assembly approach would sharply reduce the e-beam writing time for a 95 mm diameter template with terabit/in2 densities.
David Lammers, News Editor -- Semiconductor International, 8/19/2008
|
A method that combines electron-beam (e-beam) patterning with a self-assembling block copolymer could prove critical to reducing the cost of making bit-patterned media for the hard disk drive (HDD) industry, smoothing the way for nanoimprint lithography (NIL) to gain market adoption in the HDD media sector.
| A self-assembly approach can improve the quality and density of a nanoimprint lithography template for bit-patterned media. |
Researchers working at the San Jose Research Center of Hitachi Global Storage Technologies (HGST, Tokyo) and at the Department of Chemical and Biological Engineering of the University of Wisconsin (Madison) said their research would sharply reduce the time required to write the NIL 1:1 templates, now estimated at nearly a full month of e-beam writing time for a 95 mm diameter disk with terabit/in2 patterns.
| Tom Albrecht, Research Manager, HGST |
The HDD industry is moving from today’s continuous grain magnetic media to patterned media. First, discrete tracks ~50-70 nm long would boost densities somewhat, an approach that probably can be handled by NIL without the use of the block copolymer. As the HDD industry moves to bit-patterned media, with ~20 nm pillars of magnetic material, the self-assembly approach will be required to create the template, according to Tom Albrecht, manager of the patterned media research program at HGST, who previously worked at IBM Corp.’s Almaden laboratory.
Because e-beam systems are slow and cannot define patterns as small as the HDD industry needs, the self-assembly approach enables feature density multiplication, boosting the aerial density of the pattern by 4× while improving the pattern quality. With density multiplication, exposure time on the e-beam tool is reduced, which makes the approach more affordable, Albrecht said.
“We use an e-beam resist to define the chemical pattern. We write with the e-beam tool, but instead of using the pattern as a template for fabrication, we use it as a template to define a chemical pattern. The BC [block copolymer] assembles on top of that chemical pattern, which corrects for problems. It improves both the quality of the pattern and enhances the resolution,” Albrecht said. While he served as manager of the program and its spokesman, Albrecht said the work was done by HGST researcher Ricardo Ruiz.
| Paul Nealey, Professor, University of Wisconsin-Madison |
Paul Nealey, a professor of chemical and biological engineering and director of the Nanoscale Science and Engineering Center at the University of Wisconsin, said the approach effectively means that the e-beam tool only needs to define every fourth spot in the pattern. Then, the block copolymer interpolates the final pattern in between the lithographic pattern.
“If we use an e-beam resist to define a chemical pattern and then use the BC on top of it, we get beautiful structures,” Nealey said. “They are defect-free, over as large an area as you care to pattern. From a quality point of view, this reduces the defectivity and you don’t have to pattern every spot.”
Nealey said he believes the technology is far enough along to be introduced by HGST to make the NIL 1:1 master. “The risk of writing a disk-sized template at terabits per square densities is huge. The writing time on the e-beam tool can be weeks to months, and the chance of it not actually working is prohibitive. Here, we nominally reduce the writing time by a factor of four and, in actuality, it is even greater than that. We believe we get at least an order of magnitude decrease in the writing time. Moreover, the quality of the chemical pattern doesn’t have to be at device-level quality,” because the self-assembly technique improves the rectification of the pattern.
Nealey said the approach uses an existing BC material: polystyrene-block methyl methacrylate (PS-b-PMMA). Nealey said that his collaborator at Wisconsin, Juan de Pablo, is the world’s leading expert on molecular simulation of the block copolymers.
“Our expertise is in the polymer physics,” Nealey said. “It is a misconception that the BC itself makes a certain morphology or structure, that it just magically gets a structure to occur on a surface. The way the self-assembly process works is to understand how this material equilibrates in the presence of this pattern. The intellectual strength is to understand the thermodynamics, the minimization of energy of these films.”
Albrecht said he is confident that the self-assemby block copolymer technique will prove to be “possible and practical, though by no means would I say that there aren’t bugs to be worked out. We believe this will actually work, and consider it quite likely that it will be brought into manufacturing at the 1-2 tB/in2 density.”
Earlier this year, HGST purchased a NIL tool from Molecular Imprint Inc. (MII, Austin, Texas) for patterned media development, said Ken Rygler, chief marketing officer at MII.
The research was described in an article that appeared this week in Science.
MII Tackles Patterned Media Opportunity
03/16/2010Obducat Rolls Sindre for Patterned Media
03/16/2010New Roles for Merchant Mask Makers
11/11/2009Metals Self-Assemble Into Nanostructures
03/16/2010
