Takayuki Iwamatsu

Evaluation of phase and transmittance error on deep-UV halftone phase-shift mask

The effect of phase shift and transmittance fluctuation in a mask plate have been studied. The differences of these optical properties of halftone phase shift masks result in critical dimension(CD) error on a wafer so that these fluctuation in a plate reduce the process window across the exposure field. In considering CD error budget, such factors as phase shift and transmittance has to be taken into account. To estimate this budget, a set of test masks were fabricated, in which phase shift and transmittance are varied, and the exposures using these masks under the same conditions were performed.

Lithographic performance of SiNx single-layer halftone mask

An aging process that makes SiNx single-layer halftone film stable for DUV (248 nm) exposure has been established. The light irradiation with a low pressure mercury lamp was used to age the SiNx halftone film from the tendency of the transmittance change caused by the DUV exposure. Taking account of the optical constants shift during aging process, a SiNx halftone film with transmittance T equals 9.3%, phase shift angle (theta) equals 178 degrees was obtained. At the SiNx film, no transmittance change was observed after 2800 J/cm2 DUV exposure. Using the mask, 0.2 micrometers hole patterns were obtained with above 1.0 micrometers depth of focus (DOF).

Deep-UV attenuated phase-shift mask for a quarter-micrometer photolithography

A set of deep UV attenuated phase shift masks (PSMs) was fabricated to clarify the correlation between PSM optical characters and the printing performance. A SiNx single-layer film was adopted as a phase shifter which works as an attenuator simultaneously. We fabricated nine PSM test masks in which both phase shift and transmittance were varied. In these masks, phase shift ranges from 176 to 205 degrees and transmittance ranges from 6.4 to 13.1%. The target feature size of the experiment was set to that of 256 M DRAM. Line/space patterns under 0.25 micrometer and the contact holes under 0.3 micrometer have been evaluated. The evaluation was made using an exposure-defocus window analysis from measured aerial images. Each mask was compared by the calculated depth of focus (DOF) at 10% exposure latitude. The experimental results show that shifter optical characters do not play an important role in case of line/space patterns. However, the phase shift error leads to serious reduction of DOF for hole patterns. In using high transmittance PSM, mask bias is inevitable to print under -0.30 micrometer hole patterns because of the side-lobe peak. To obtain the maximum DOF, mask bias should be decided with meticulous care.