Modern life and technology would be very different if not for photolithography, a simple step in the process of creating an integrated circuit. Basically, you use a slide project to project an image on the circuit pattern on a wafer. Subjecting the wafer to a light pattern alters the chemical layer on top of the wafer and creates a mask. The mask enables certain parts of the wafer to be processed to produce the circuit.
Despite the fact that the process is relatively simple, photolithography is a crucial part of determining power consumption, speed, and the number of transistors on a chip. Because of this, many people have tried to make the process more efficient or replace it altogether – many have seen success.
Some are concerned that further advancements are limited. Current photolithographic systems are linear optical systems, meaning that the smallest feature they project is determined by the size and precision of the optics. Since all optics have a finite size that can be produced, there is a limit to the feature size that can be created, called the diffraction limit.
Getting Around the Diffraction Limit
The focused ion beam machine in use at the nanofabrication facility of the Molecular Foundry, Lawrence Berkeley National Laboratory. The yellow light is necessary because photolithography is a primary use of the facility, and shorter wavelengths of light can interfere with fabrication using such processes.
The focused ion beam machine in use at the nanofabrication facility of the Molecular Foundry, Lawrence Berkeley National Laboratory. The yellow light is necessary because photolithography is a primary use of the facility, and shorter wavelengths of light can interfere with fabrication using such processes.
Current systems manage to get around by using certain tricks, like changing the refractive index of the material the light goes through after it is released from the projection system, or by putting a cap on exposure time and double patterning, so each wafer is exposed two times with some offset between exposures. These small tweaks have given us some improvement in the arena.
Unfortunately, it appears as those we have reached our limit, and it is time for new technology to be brought to the forefront. One option that some are considering is a modification to stimulated emission depletion imaging (STED).
STED imaging
STED imaging, as it is currently used, has no value for photolithography, but it does allow us to beat the diffraction limit. With STED imaging, the use of two different lasers lets you get around the limit. And the only limits opposed on it are calculated by the power of the laser. So if you want a smaller image, use more power.
But there are two possible problems when you try to use the process for photolithography: although you can produce features that are say 6nm in size, those features will be separated by the diffraction limit, and the other problem is that you would need two masks, one as the standard circuit mask and the other to provide nodes that thin out all of the features to provide the resolution.
Some new research makes using STED actually plausible though. It works with just one mask, meaning that the alignment procedure would be no more difficult than it is now. While with normal STED imaging, people use dyes, with this photolithographic application, they would use a resist coating that will oxidize in ultraviolet light. The scientists behind the research have seen some success in the area, but it is still far too early to determine if the technology will be plausible.