Like the circuitry on integrated chips, the periodic stripes etched on laser diodes are shrinking. It is intuitive that reducing feature dimensions is simply a matter of reducing the illumination wavelength of a lithography laser. However, operating at shorter wavelengths has never been a trivial task, and decreasing them further won't be so easy. One alternative is to apply the evanescent near-field region of light to interferometric lithography techniques used to produce the periodic stripes on distributed feedback and distributed Bragg reflector laser diodes. As reported in the April 1 issue of Applied Optics, a group of researchers at the University of Canterbury applied this principle in lithography simulations. They directed light from a 450-nm source through a 270-nm-period chrome grating embedded in material with a refractive index of 1.6. The light formed high-contrast 135-nm-period intensity patterns in photoresist. The evanescent near field is dominated by nonpropagating light that decays within a few dozen nanometers from the source. Applying the evanescent region of a 450-nm light source in interferometric lithography simulations produced high-contrast, 135-nm-period intensity patterns in photoresist. Courtesy of the University of Canterbury. Although the evanescent near field encompasses the shortest wavelengths of a light source, it does not propagate and decays a few dozen nanometers from the source. However, recent theoretical work postulates negative refractive index materials that amplify the evanescent field (see "Radical Lens Theory Repeals Diffraction Limit," Photonics Spectra, March 2001, page 20). Interestingly, the university reported a parallel result. "Our [evanescent interferometric lithography] paper shows enhancements by up to a factor of five, which would lead to reduced exposure time for lithography -- and throughput is all-important," said Richard Blaikie, who is leading the research. Within a distance of 20 nm beyond the grating, the intensity typically increased by a factor of three for wavelengths above the grating's 432-nm cutoff -- particularly those between 450 and 460 nm. However, within a distance of 10 nm below the grating, the 454-nm wavelength showed a peak intensity more than five times as great as wavelengths below cutoff. Peak intensities for sub-cutoff wavelengths trailed incident intensity by about 10 percent. "This [enhancement] is somewhat counterintuitive, as one usually expects light transmission to reduce for subwavelength apertures," Blaikie said. "However, for gratinglike structures, there is always good transmission for one polarization, and if the grating material is chosen correctly, then surface plasmon excitations on the exit side of the grating can cause the near-field intensity to be enhanced." Potential drawbacks include a reduced depth of field for enhanced wavelengths, but Blaikie noted that the pattern's limited depth of field can be "tuned" to match the resist thickness and avoid substrate reflection problems that lead to unwanted interference. Another drawback is the hard contact between mask and substrate, which can cause mask contamination. However, Blaikie countered that modern cleanrooms have much tighter control of particulate count, which makes the issue worth revisiting.