Optical versus X-ray lithography for future device fabrication

Abstract

The lithographic requirements of present and future DRAM memories are described concerning feature size, overlay, chip size, defect size and time schedule. The state-of-the-art performance and the future potential of optical and X-ray lithography will be discussed in reference to these requirements, from a theoretical as well as a practical viewpoint including economic aspects.In optical lithography, the deep UV technology has evolved fast in recent years. Presently, there is no doubt that DUV lithography is the main road for 64 M DRAM production with 0.35 um minimum features. The future potential of optical lithography is analyzed, revealing that - although resolution may be enhanced down to 0.2 um by reducing λ/NA - focus depth and field size are severe limiting factors. Key issues for the extension of optical lithography into the future are device planarization and top surface imaging resists.In X-ray lithography recently, highly sensitive resists with sufficient process tolerance and stepper prototypes with vertical stages for synchrotron radiation became available. On the other hand, the reliability of powerful X-ray sources and the viability of 1 x masks still has to be demonstrated. Whereas the resolution potential of X-ray lithography will easily cover the 1GBit DRAM requirements, overlay issues will truly limit its performance due to mask and wafer distortion and e-beam pattern placement accuracy.Comparing the physical and technological limits of both competing lithographies, there is a window of two DRAM generation in favor of X-ray, which is obviously slowly decreasing with progressing time. However, an economic comparison of both techniques reveals that there may be an advantage for X-ray. Thus, an entry of X-ray lithography driven by economic rather than technological reasons cannot be excluded today.