| Currently, the lithography technology is still the main means of VLSI fabrication. As the critical dimension of integrated circuits continues to shrink, the demand to improve lithography resolution is also increasing. They are the most effective ways to improve the lithographic resolution by shortening the exposure wavelength and increasing the numerical aperture, but it will increase the difficulty for design of optical system, selection of optical materials, optical coatings, photolithography process and environmental conditions. Therefore, under existing technological conditions, through the theoretical analysis of lithography imaging, it is an effective way to improve lithography resolution by optimizing process parameters.This paper focuses on the impact of illumination pupil shape in lithography illumination system on the imaging performance. The obtained image will be very different by using incident light irradiated on the mask with different directions. An excellent imaging performance may be received by using the incident light in certain directions, while the incident light in the other directions will cause the image quality becoming worse. The purpose of this study is to find the best lighting pupil distribution for different types of masks, so that the imaging performance could be improved to meet actual demand. The main research work includes the following aspects:(1) The establishment of lithography imaging model. The formula for calculating the scalar and vector aerial image is derived in detail by using scalar and vector diffraction theory. The software, that can quickly and accurately calculate aerial image intensity, and also can effectively analyze the imaging performance, is written using Matlab GUI.(2) Source optimization. The design of source optimization is implemented by using our written software. The variables need to be optimized are intensity distribution and polarization state. The objective functions that describe imaging fidelity and process window separately are proposed. It is so hard to calculate the solutions of these two functions directly that low computational efficiency is shown. We propose a method to characterize the imaging fidelity and process window through characters of aerial image. Due to the lack of experimental conditions, imaging performance obtained by optimized illumination mode will be analyzed by taking advantage of commercialsoftware Prolith in this paper, and simulation results show that a great improvement of the process window is achieved by adopting the optimal source shape.(3) The analysis of aberration of object lens. The impact of object lens aberration on imaging performance is analyzed theoretically, and verified by simulation. The aberration sensitivity matrix corresponding to different L/S patterns is calculated, and the degree of impact of each term in Zernike function on major imaging performance is analyzed for different L/S patterns. By theoretical analysis, the expression of each Zernike aberration term is derived in the form of different off-axis illumination, and the effect of each Zernike aberration term on imaging performance is compared by simulations with different off-axis illumination point. The optimal source pupil distribution that makes the impact of each aberration term on imaging performance minimum is found.(4) Source sensitivity analysis. For different types of structures on the reticle, a number of the reference positions are selected. The change of objective function due to the changes of point source intensity on illumination pupil plane is calculated under certain illumination mode, i.e., source sensitive matrix. Source sensitivity matrix could reflect which region on the pupil plane will improve imaging performance of different positions on the reticle, and can also be used to verify the validity of certain illumination mode on imaging performance for different patterns on the reticle.(5) Source error analysis. The errors that may exist in our actual lithography system are introduced in detail. The degeneration of lithographic performance caused by various types of errors is analyzed through theory and simulation experiments. Depending on the degree of influence, the error tolerance can be obtained for various types of errors, which may provide some guidance for designers of lithographic systems. In addition, illumination pupil distribution that makes the deviation of lithographic performance caused by various errors minimum is analyzed, which has obvious effect on reducing the impact of errors on imaging performance. |
