Pellicles bedevil 157nm despite overall progress
07/01/2002
Nearly 140 industry experts from IC manufacturers, tool and materials suppliers, and academia attending International Sematech's (ISMT) recent Technical Data Review were told that 157nm lithography is making significant progress, leading to manufacturing insertion in 2005. Among the reported data was information related to solutions for intrinsic birefringence (IBR) of CaF2 — the major constituent of 157nm lens systems. ISMT reports that annealed <111> crystals have now met the specifications for homogeneity and stress birefringence. Reduction in stress birefringence for annealed <100> crystals has also been achieved.
One issue that continues to bedevil 157nm progress, however, is how to develop soft pellicles for photomasks that can retain their transparency while being irradiated. It is hoped that — by studying the mechanisms of photochemical darkening, through a project commissioned by ISMT in conjunction with several universities — polymer materials that can withstand 157nm radiation will be developed. Until a solution for soft pellicles is achieved, hard pellicles — modified fused silica sheets ~800μm thick that transmit about 95% of light at 157nm — will have to be pursued as an interim measure. Research activities are coordinated worldwide by ISMT's Hard Pellicle Task Force. The group reports a tenfold reduction in image distortion over previous results when a hard pellicle is part of the exposure system.
Asked to comment on the pellicle challenge facing 157nm lithography, Griff Resor, president of Resor Associates, noted that each hard pellicle introduced into the exposure system contributes to imaging errors arising from sag, thickness errors, and deformation (caused by mounting forces). "Every element other than the hard pellicle is now fabricated to a tolerance of 1/50th of a wave. It appears unlikely that this tolerance can be reached on a hard pellicle window," Resor said.
Even if hard pellicles achieve their lofty tolerance goals, another challenge remains. "Dedicated metrology tools with the capabability to measure wavefront distortion caused by mounted hard pellicles do not yet exist, at either the 193 or 157nm device generation," says Andrew Merriam, chief scientist at Actinix and an attendee and presenter at the conference.
"Such metrology tools should also verify the gross thickness of the pellicle for aberration correction, and the quality of the antireflection coatings applied to the surfaces." Actinix is looking into the possibility of producing such an at-wavelength, interferometer-based tool. —D.V.