The resolution of electron beam lithography is limited by the proximity effect, which is due to electron scattering in the resist and substrate, the beam profile and diffusion-type effects during development and post processing. Usually the radial dose distribution in the resist is used to correct for the proximity effect.
We propose a method to summarize all resolution limiting effects in one function (total process function). This function is calculated by Monte-Carlo simulation of the radial dose distribution in the resist produced by a pencil beam and subsequent 2D-convolutions with the beam profile and a process function. The suggested approach is useful to short-range and long-range proximity effect correction and to assess the quality of the lithographic process.
Karl E. Hoffmann1, Michael Asteiner2, Peter Speckbacher2Show Affiliations
- University of Applied Sciences Rosenheim, Hochschulstrasse 1, D-83024 Rosenheim, Germany
- Johannes Heidenhain GmbH, Dr.-Johannes-Heidenhain-Strasse 5, D-83301 Traunreut, Germany
Proximity effect correction software in electron beam lithography needs the pattern layout and the total process function (TPF) as an input. This function is determined by electron scattering in resist layer and substrate, the beam blur and diffusion effects during development and post processing. Measuring the TPF is difficult. A three-step method is therefore suggested to calculate the TPF for thin resist layers: Monte Carlo simulation of electron scattering followed by 2d-convolutions with the beam blur and a diffusion function. The convolution algorithm will be described in detail. Experimental data taken from literature are used to verify this calculation method. The method is also useful to assess the quality of the lithographic process.Go To Microelectronic Engineering
The diagram shows typical simulation results for PMMA (150 nm) on Si Substrate with 30 keV primary beam energy. The beam shape is assumed to be Gaussian with FWHM = 50 nm. The experimental data (fictional, but close to real) exhibits a process effect (Gaussian with diffusion length 50 nm). This PMMA process would not be optimal.