Immersion lithography
From Wikipedia, the free encyclopedia
In photolithography, immersion lithography is a resolution enhancement technique that interposes a liquid medium between the optics and the wafer surface, replacing the usual air gap. This liquid has a refractive index greater than one. The wavelength in the liquid is reduced by a factor equal to the refractive index. With the 193 nm (nanometer) wavelength produced by ArF excimer lasers, the liquid currently used is ultra-pure, degassed, deionized water. At a refractive index of 1.44, the reduced wavelength in water is 134 nm. Equivalently, the optical resolution is increased since the immersion fluid allows using lenses with a higher numerical aperture.
Contents |
[edit] Applications of immersion lithography
As of 2005, it is expected that immersion lithography will be used in 2009 to print 45 nm lines and spaces <ref> M. LaPedus, "Litho race," EE Times, October 21, 2005. </ref>. As of 2006, immersion lithography is expected to be introduced in the 45 nm node by most chipmaking companies, including IBM, Toshiba, STMicroelectronics and others.
Following its aggressive introduction, it is speculated that enhancements will be used to prolong the use of the technology to smaller features. Such enhancements include the use of higher refractive index materials in the final lens, immersion fluid, and photoresist.
[edit] Resolution limit of immersion
A metric commonly used to quantify resolution in photolitography is numerical aperture (N.A.). The numerical aperture is estimated as the product of the refractive index of refraction and the maximum angle of propagation of light, relative to the surface. The numerical aperture has a maximum possible value equal to the refractive index. In an air-based optical system, the numerical aperture has a maximum value of 1.0. The main motivation for using immersion is the ability of the numerical aperture to go above 1.0, thus enhancing the resolution without changing the light source. The absolute resolution limit is given by a quarter-wavelength divided by the N.A. The current maximum N.A. is 1.35 for water-based immersion.
Beyond this limit, a conservative way to extend the technology to finer features is to pattern features in between features previously patterned by immersion lithography (also known as double patterning).
[edit] High-Index Materials
As noted above, the refractive index is a key factor in determining resolution. The practical upper limit for current lens materials is around 1.4 <ref> A. Grenville, "Prospects for Immersion Lithography at the 45nm Half-Pitch and Beyond," Future Fab International, Issue 20, pp. 76-79 (2006). </ref>. By using LaF3 with a refractive index n=1.67 <ref> D. Ristau et. al., Appl. Opt. vol. 41, pp. 3196-3204 (2002). </ref> as the final lens material, along with a recently demonstrated immersion fluid with refractive index n=1.64 <ref> A. Hand, "High-Index Fluids Look to 2nd-Generation Immersion," Semiconductor International, April 1, 2005. </ref>, or doped water <ref> B. W. Smith et. al., Proc. SPIE vol. 5377, pp. 273-284 (2004). </ref>, a numerical aperture of 1.5 can be reached. This is the minimum value needed for 32 nm line-space resolution. IBM has demonstrated line-space features of just under 30 nm, using lens, fluid, and photoresist indices of about 1.6, 1.6 and 1.7, respectively <ref>"IBM Researchers Develop 29.9 nm Chip-Manufacturing Process", IT News Online, 2006-02-20. Retrieved on 2006-07-09.</ref>. This is expected to be the typical design rule for the 22 nm node.
[edit] Polarization Constraints
As numerical apertures increase, the degree of polarization of the light becomes critical to the image quality. This is because the interference of light inside the photoresist becomes polarization-dependent. Specifically, the imaging of straight lines near the resolution limit is best done with light polarized parallel to the lines. This requires special illumination preparation which is available on the most advanced lithography systems. Imaging of holes or islands is more problematic unless the holes or islands are closely spaced near the resolution limit in one direction and widely spaced in the orthogonal direction; otherwise, polarization effects will be a hindrance rather than a benefit. The highest hole or island density is achieved by effectively superimposing two orthogonal line-space images. Preferably, these images will be polarized parallel to the lines. Due to these imaging constraints, integrated circuit (IC) layouts utilizing dimensions near the resolution limit will be required to be 'lithography-friendly'.
[edit] Defect concerns
Other considerations which are important to immersion lithography systems are the elimination of bubbles in the immersion fluid, temperature and pressure variations in the immersion fluid, and immersion fluid absorption by the photoresist <ref> M. Switkes et. al., J. Vac. Sci. & Tech. B vol. 21, pp. 2794-2799 (2003). </ref>. Degassing the fluid, carefully constraining the fluid thermodynamics and carefully treating the top layer of photoresist are key to the implementation of immersion lithography. Defects intrinsic to immersion lithography have been identified <ref> U. Okoroanyanwu et. al., "Defectivity in water immersion lithography," Microlithography World, Nov. 2005. </ref>. Reducing particle generation due to the water dispensing unit should help reduce the incidence of defects. Water also has been shown to extract acid from photoresist <ref> J. C. Taylor et. al., SPIE vol. 5376, pp. 34-43 (2004). </ref>. Specifically, photoacid generators (PAGs) are extracted into the water, which produce acid upon radiation exposure. This must be managed to ensure the lens is not corroded by the acid or contaminated by the extracted agents, and the photoresist is not chemically altered to the point of being defective.
[edit] Photomask impact
The higher numerical apertures allowed by immersion lithography make contact hole imaging less sensitive to mask critical dimension (CD) errors. <ref> D. Kang et. al., Proc. SPIE vol. 4404, pp. 170-179 (2001). </ref>. The mask error enhancement factor (MEEF) can therefore be reduced. As the minimum resolvable half-pitch linewidth decreases, features on the photomask will eventually approach subwavelength sizes. Subwavelength features no longer obey the laws of classical imaging optics but need to be rigorously analyzed using electromagnetic theory (see for example, <ref> C-W. Chang et. al., Laser Physics Letters vol. 2, pp. 351-355 (2005). </ref>). In addition, features around a half-wavelength in size exhibit very strongly polarizing and attenuating behavior <ref> F. T. Chen, SPIE vol. 4889, pp. 1313-1323 (2002). </ref>. One way to delay this outcome would be to increase the magnification of the photomask image relative to the wafer image.
[edit] Immersion at 157 nm
It is possible to use immersion lithography with an F2 excimer laser, which produces a 157 nm dry wavelength. However, it would be necessary to consider the mechanisms by which these wavelengths interact with the materials used in photolithography. Absorption is greatly enhanced, and the ionization potential is exceeded. The success of 157 nm immersion would depend on the use of materials with minimal absorption. Water itself is not transparent enough to serve as an immersion fluid for this wavelength. It is also important to make sure that when the wavelength is divided by the refractive index, it gives a sufficiently smaller value than 118 nm, which is the 193 nm wavelength divided by the refractive index of the next-generation immersion fluid at that wavelength (n=1.64). This is the measure of the degree of resolution enhancement.
[edit] Future of immersion lithography
Immersion lithography tools and processes are naturally expected to cost more than dry lithography tools and processes. Largely for this reason, most likely the technology will be extended to the 32 nm node and beyond as well. An ArF water-based immersion tool is ideally suited for half-pitches of 100 nm down to 40 nm. However, it is expected that at some point below 40 nm, photoresists will limit further scaling <ref> U. Okoroanyanwu and J. H. Lammers, Future Fab International, Issue 17 (2004). </ref>.
[edit] See also
[edit] References
<references />
