Chemical Shrink Process Research Articles

Extension of 193nm Lithography by Chemical Shrink Process

Extension of 193nm Lithography by Chemical Shrink Process

Parallelizable Simulation of Material Effects of Chemical Shrink Process of Nanolithography

A parallelizable simulator of the chemical shrink process of nanolithography is developed to investigate the process dependence on material properties. The chemical shrink process model employed for the simulator includes the cross-linking reaction of the resolution enhancement of lithography assisted by chemical shrink (RELACS) material, the inhibitor deprotection reaction of the chemically amplified (CA) resist, and the photoacid diffusion and trapping in the CA resist and RELACS material. It is found that, in the descending order of their effectiveness on amount of shrinkage, the photoacid trapping and diffusion mechanisms of the RELACS are the most effective ones, the photoacid trapping mechanism of the CA resist is second, the photoacid diffusion mechanism of the CA resist and the cross-linking reaction of the RELACS are third, and the deprotection reaction of the CA resist is the least effective. The interpretations of the dependence of the shrinkage amounts on these mechanisms are also provided on the basis of the simulations.

Newly Developed Resolution Enhancement Lithography Assisted by Chemical Shrink Process and Materials for Next-Generation Devices

We have newly developed a resolution enhancement lithography assisted by chemical shrink (RELACS) material for ArF lithography. Several process performances were evaluated for 65 nm nodes and next-generation devices. The principle and procedure of the RELACS process is similar to those developed previously for KrF lithography. The extent of cross-linking reaction and the mobility balance of chemical components at the boundary between the resist and RELACS film is adjusted to ArF resist chemistry. The novel RELACS material causes variation in shrinkage from 10 to 50 nm by controlling process conditions. The shrinkage amount is independent of pattern pitch and lithography conditions, i.e., dose and focus. We confirmed that the pattern profile, lithography margin, critical dimension (CD) uniformity, etching resistance, and pattern defects were not deteriorated by the RELACS process with distilled water. We believe that the novel RELACS process and materials are extremely useful for 65 nm nodes and next-generation devices.

Reaction-diffusion mechanisms for the chemical shrink process of nanofabrication

The effectiveness of reaction-diffusion mechanisms on the formation of nanoscale polymer films in the chemical shrink process is investigated by a two-dimensional model. The model includes three mechanisms: the catalytic cross-linking reaction of water-soluble polymers and cross-linkers, the diffusion of photoacids, and the trapping of photoacids by the cross-linked polymers. It is found that, although the cross-linking reaction is the vital mechanism for the chemical shrink process, it is much less effective on the formation of nanoscale polymer films than the photoacids diffusion and trapping mechanisms.

Fabrication of 65-nm Holes for 157-nm Lithography

We investigated the use of hole shrink processes along with a bilayer process as a fine-hole-pattern replication technique for 157-nm lithography. A siloxane-type resist and an organic underlayer were used in the bilayer process. The lithographic resolution limit was 90 nm in diameter with a 0.85-numerical-aperture 157-nm microstepper using an attenuated phase-shift mask. Using the bilayer process, we successfully transferred 85-nm-diameter hole patterns in the resist into a 400-nm-thick tetraethylorthosilicate (TEOS) layer. A thermal flow process and the Resolution Enhancement Lithography Assisted by Chemical Shrink (RELACS) process were used as the hole shrink processes to enhance the resolution. The 120-nm-diameter hole patterns in the resist were shrunk to 80 nm by the thermal flow process and to 65 nm by the RELACS process. In the RELACS process, the diameters of the shrunk hole were less dependent on the hole pitch than in the thermal flow process.