PORTLAND, Ore.—As photolithography research moves toward extremely short wavelengths of ultra-violet light, one group at the University of Maryland is proposing multi-photon photoresists that allow visible light to achieve nanoscale resolution that is inversely dependent on exposure time.
"Most approaches to getting higher resolution with photolithography involve using light of ever shorter wavelengths," said professor John Fourkas. "Our goal is to use visible light to produce nanoscale features."
The new multi-photon technique, called Resolution Augmentation through Photo-Induced Deactivation (RAPID), uses one laser to initiate exposure in the photoresist and a second to complete it, allowing full exposure of only the nanoscale overlapping areas of the two focused beams.
"If we take a laser beam and focus it through a microscope objective," said Fourkas. "We can confine absorption to this very tiny region right at the focal volume of the laser."
The researchers have already perfected the technique for use in the selective polymerization of 3-D materials on-chip. Using what is called Multi-photon Absorption Polymerization (MAP), the team has fabricated tiny inductors on chips. RAPID is a follow-on effort to use multi-photon absorption with photoresists to achieve nanoscale resolution with focused visible light, delaying or possible eliminating the need to move to extreme ultra-violet light sources (EUV).
Microinductor created using multiphoton absorption polymerization (MAP) followed by selective metallization.
The technique works at normal atmospheric pressure, unlike EUV with requires processing in a vacuum. Instead, a special photo-initiator in the resist is activated by one laser, then deactivated by a second, realizing a phenomenon the researchers call proportional velocity (PROVE), which yields smaller features for higher exposures.
Next the researchers plan to test their technique of the wafer scale, in contrast to the point-by-point demonstrations they have given so far. The team estimates that RAPID will be ready for commercialization in about 10 years.
There is a company called LumArray out of MIT has been trying to apply similar type of technology to their maskless litho systems for almost 10 years. And is selling working systems now. Prof. Fourkas' approach requires two high a light intensity it will be difficult to achieve high throughput. In addition, a similar SEM picture has been posted outside Prof. Fourkas' lab at UMD for about 10 years, how come it has just mad news?
OK, it turns out that the yellow cleanroom light will not expose this special photoresist. Here is what professor Fourkas says: "We do everything under yellow cleanroom lights, so it's not highly sensitive to it. We tend to prepare resists fresh and use them within a day or so, and so it's possible that there could be a longer-term effect if they aren't stored in the dark (like any other photoresist....)"
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.