The semiconductor industry has witnessed an exponential growth in the past ten years. Technological advancement in IC material and design has produced multiple generations of ICs where each new generation has smaller and more complex circuits than the previous generation. With each new generation of ICs, the number of interconnected transistors increases while the geometry size of the IC decreases. This IC miniaturization has increased the complexity of IC design and fabrication. To achieve even smaller transistors with a size ranging in nanometers, the development of higher resolution lithography technology is required.
Extreme Ultraviolet Lithography (EUVL) is an innovative technology that provides a UV beam with a wavelength of 13.5nm, enabling chip manufacturers to expose fine circuit pattern with half-pitch below 20nm, that cannot be produced with the help of conventional lithography techniques. The unique aspect of EUV is that it is absorbed by most of the materials. Thus, in EUVL, the reflective optic arrangement is used rather than refractive. The biggest challenge for this technology, however, is to generate a powerful enough EUV with a short wavelength range of 13.5nm. Currently, the system uses high energy laser which is directed on the microscopic droplets of tin and converted into plasma, providing EUV light which is in turn converted into a focused beam with the help of an optical arrangement.
Although EUVL seems like a likely candidate to replace optical lithography, it is not being used on large scale. Despite industry backing, vast resources and billions of dollars in funding, the challenge of generating a powerful enough EUV beam of extremely short wavelength remains a big hurdle.
Some recent developments indicate a better future for this technology. Samsung, the first player likely to use EUVL, gave the first detailed look at its long-awaited 7nm process at the 2018 Symposia on VLSI Technology and Circuits. Samsung may start manufacturing actual products using EUVL as early as 2019. Intel, previously at par with Samsung in using EUVL, appears to have run into some bottlenecks and has delayed volume shipment on 10nm into 2019. In July 2018, ASML said it was on track to ship 20 EUV systems in 2018 and expected to ship at least 30 more in 2019.
However, some problems that still need to be addressed are:
- Readiness of commercial resists
- EUV stochastic-induced effects
- Unavailability of commercial tools for actinic patterned mask inspection
Industry leaders have been researching EUVL for nearly three decades. Let’s see how much longer it will take to make this technology economically and technically feasible for large-scale production.
Author: Vikas Kumar