| Paving the Way to Nanoelectronics 16 nm and Smaller |
![]() |
| Wednesday, 30 March 2011 00:00 | ||||
|
As the nanoelectronics industry pushes towards feature sizes of 22 nm and smaller, conventional single-exposure refractive lithography systems used to print circuit patterns onto computer chips will no longer be feasible. Extreme ultraviolet (EUV) lithography, utilizing reflective optics and 13-nm-wavelength light to print chips, is the leading candidate to meet the industry’s future needs. Despite strong progress in EUV lithography over the past decade, significant challenges remain, including defect-free mask fabrication (see Science Highlight Investigating Extreme Ultraviolet Lithography Mask Defects), and the development of ultrahigh-resolution photoresist—a light-sensitive material used to form a patterned coating—that simultaneously supports low line-edge roughness (LER), high sensitivity, and sub-22-nm resolution. Using the SEMATECH Berkeley Microfield Exposure Tool
The SEMATECH Berkeley MET at ALS Beamline 12.0.1.3 (from right to left): A rack of custom-made control electronics which allow researchers to meticulously manipulate the tool. The MET is located within a thermally controlled enclosure (shiny metal paneling) that can maintain temperatures to 1/100th of a degree centigrade inside of the projection optics chamber. A researcher inspects the photoresist on exposed film in the post-exposure clean room following development. Development of EUV photoresist is particularly challenging because it relies on the availability of ultrahigh resolution lithography tools well in advance of when they would be required for high-volume semiconductor manufacturing. Tools such as the MET allow researchers to study nanoscale patterning at EUV wavelengths, and so play a crucial role in the advanced materials development required to achieve high-volume fabrication of nanoscale electronics. The relative simplicity of these microcfield tools enables them to provide higher-resolution capabilities than current full-production-scale tools. For example, preproduction tools just now being delivered still significantly underperform in terms of resolution when compared to the MET, which came online late in 2003. The most unique attribute of the SEMATECH Berkeley MET is its use of a custom-coherence illuminator made possible by its implementation on a synchrotron beamline. Using conventional illumination, the resolution limit of the 0.3-NA optic MET is approximately 25 nm; however, with custom-coherence the system can achieve the wavelength-induced resolution limit of 12 nm. Moreover, the high spectral purity of the ALS light minimizes image degradation, which can be caused by longer ultraviolet wavelengths present in current plasma-based EUV light sources under development.
Exposures of 45-nm lines and spaces using both conventional illumination and the pseudo phase shift mask method, which results in twice as many 22.5-nm lines and spaces. Another practical problem for EUV photoresist testing at these small dimensions is that masks supporting these resolutions must also be available. To decouple resist research from mask development, a process unique to the ALS implementation of the MET called pseudo phase shift mask is used, allowing conventional binary amplitude masks to act as phase shift masks, causing line space features to print at one half the size they are on the mask. Using this method, high-quality 16-nm line space printing with 2-nm LER has been achieved in an experimental photoresist. Although still lacking in sensitivity and edge roughness, these results represent a significant breakthrough in single-exposure projection printing resolution.
Using an experimental photoresist provided by Inpria Corporation, 16-nm lines and spaces were printed on the MET using pseudo phase shift mask mode. These results were obtained in collaboration with Inpria Corporation, who has developed a photosensitive spin-on inorganic ultrathin imaging film (resist). The inorganic nature of the film enables high EUV absorption in the ultrathin film, minimizing stochastic effects. Additionally, the film does not rely on thermal chemical amplification, thereby reducing diffusion and image blur during post-exposure processing. The downside of the lack of chemical amplification is decreased sensitivity. The sensitivity of the resist shown in the image above, for example, was 70 mJ/cm2, which is 4–7 times slower than industry requirements.
Research conducted by P. Naulleau, C. Anderson, L.-M. Baclea-an, P. Denham, S. George, K. Goldberg, N. Smith, G. Jones, and S. Rekawa (Berkeley Lab); B. Rice, S. Wurm, C. Koh, and W. Montgomery (SEMATECH); B. McClinton and R. Miyakawa (University of California, Berkeley); and T. Wallow (Global Foundries). Research funding: SEMATECH. Operation of the ALS is supported by the U.S. Department of Energy, Office of Basic Energy Sciences. Publications about this research: P. Naulleau, C. Anderson, L.-M. Baclea-an, D. Chan, P. Denham, S. George, K. Goldberg, B. Hoef, G. Jones, C. Koh, B. La Fontaine, B. McClinton, R. Miyakawa, W. Montgomery, S. Rekawa, and T. Wallow, “The SEMATECH Berkeley MET pushing EUV development beyond 22-nm half pitch,” Proc. SPIE 7636, 76361J (2010); P. Naulleau, C. Anderson, L. Baclea-an, P. Denham, S. George, K. Goldberg, G. Jones, B. McClinton, R. Miyakawa, I. Mochi, W. Montgomery, S. Rekawa, and T. Wallow, “Using synchrotron light to accelerate EUV resist and mask materials learning,” Proc. SPIE 7985, 798509 (2011). ALS Science Highlight #224
|



