At the University of Maryland, scientists have achieved a breakthrough in using visible light to make tiny integrated circuits. Although the technology may not become commercially available for more than a decade, it may eventually make it possible to produce computer chips that are smaller, faster and cheaper. A new lithography process could take today's computer chips such as the one pictured here into the realm of smaller, faster and cheaper. Called RAPID lithography, the technique uses visible light to attain lithographic resolution comparable to, and maybe even better than, the resolution attainable with shorter-wavelength radiation. Essential to the process is a special photoinitiator that may be excited by one laser beam and deactivated by another. The advance could one day replace photolithography in the manufacture of integrated circuits. In photolithography, each computer chip is built up in layers, with conductive materials piled on a chip and coated with chemicals that harden when exposed to light. The light shined through a mask, or stencil, projects a detailed pattern onto the photoresist, which hardens where it is exposed. The unhardened areas and underlying metals are then etched away with a chemical. The remaining photoresist is etched away using different chemicals, leaving an underlying layer of metal with the same shape as the mask. Building multiple circuits on a chip has meant forming ever-smaller circuits with shorter and shorter wavelengths of light, or even charged particles. The new technique, however, uses visible light, which is far less expensive to generate, propagate and manipulate and does not require high-vacuum conditions. In their work, the researchers used a beam of ultrafast pulses for the excitation step and a continuous laser for deactivation. They found recently that with materials deactivation, however, they could use ultrafast pulses for both the excitation and deactivation of the molecules. The discovery resulted in findings that higher exposures can lead to smaller features, leading to what they call proportional velocity (PROVE) dependence. “PROVE behavior is a simple way to identify photoinitiators that can be deactivated efficiently, which is an important step toward being able to use RAPID in an industrial setting,” said John Fourkas, chemistry professor at the university. They found that they were able to demonstrate a photoresist for which the resolution was independent of the exposure over a range of exposure times when combining a PROVE photoinitiator with one that had a conventional exposure dependence. Their work appeared in Nature Chemistry, March 2011 (doi: 10.1038/nchem.965).