MIT makes 36nm lines with “interference” litho step
A team of researchers at MIT have produced 36nm-wide lines using interference patterns and a photochromic material, and say the technique could be extended down to patterns on the scale of individual molecules.
The trio’s work (Rajesh Menon of the Research Laboratory of Electronics, with grad students Trisha Andrew and Hsin-Yu Tsai), published in the April 10 issue of Science, is based on “absorbance modulation,” which involves interference patterns that use different wavelengths of light to reinforce or cancel each other out. The process requires a photochromic material—one that changes color and opacity in response to light, properties that must remain stable after initial exposure to the light. (Through trial and error the team settled on “a comparatively photostable class of thiophene-substituted fluorinated cyclopentenes.”) Bright lines at one wavelength coinciding with dark lines at another wavelength form a “banded layer” of extremely narrow lines of clear material interspersed with opaque material. This then serves as a mask through which the first wavelength shines, acting as a “nanoscale writing beam” to pattern the photoresist underneath.
Their setup consisted of: a silicon substrate spin-coated with 200nm of anti-reflection coating (ARC), 200nm of photoresist, an 8nm-thick PVA barrier layer, and 410nm of the photochromic layer. Following exposure, removal of the PVA and photochromic layers, and baking (120??C for 90sec) and developing (TMAH for 60sec), the patterns were spin-coated with 2nm of palladium/gold alloy and inspected in a SEM. Average linewidth was seen to be about 36nm, roughly a tenth of the originating 325nm wavelength light, and also were spaced by 350nm, to the period of the second wavelength (633nm). Despite some line-edge roughness and variation caused by high-frequency noise in the second standing wave, “these results clearly demonstrate the feasibility of deep sub-wavelength localization of light using absorbance modulation,” the researchers determined.
In a statement, Menon noted the process could have significant application in semiconductor manufacturing, as well as various fields that use nanoscale patterning, such as photonics, fluidics, nanoelectronics, and nano-biological systems. A company already has been formed to develop the technology; Menon projects commercial production could be achieved “within five years.”
The work was partially funded by grants from maskless litho startup LumArray (cofounded by Menon in 2004), MIT’s Deshpande Center for Technological Innovation, and DARPA. —J.M.