Optical Proximity Correction Method

Abstract

The invention discloses an optical proximity correction method. The method is characterized in that first optical proximity correction is executed on a graph to be corrected to obtain a first correction graph, the first correction graph is verification by employing a model having a largest space value and a model having a smallest linewidth value selected in a process window model corresponding to the processes, a point with easy deviation is determined, the point is performed with target correction, second optical proximity correction is executed to obtain a second correction graph, the process window model is used for examining the second correction graph, performing modification on the first correction graph during optical proximity correction, thereby the point with easy deviation is timely discovered and corrected, redundant operation data volume for the subsequent optical proximity correction is reduced, data operation time of a mask layer is shortened, the cycling correction for redetermining the optical proximity corrective target graph is not required, a period for the optical proximity correction is shortened, and production efficiency is increased.

Description

Technical field [0001] The invention relates to the technical field of semiconductor manufacturing, and in particular to an optical proximity correction method. Background technique [0002] A mask is a necessary device in the photolithography process of integrated circuit manufacturing. The mask mainly includes a transparent glass substrate and a pattern to be exposed composed of a non-transparent material (generally chromium) covering the glass substrate. During photolithography, the mask is placed between the radiation source and the focusing lens. The light emitted by the radiation source passes through the mask, passes through the lens, and irradiates the surface of the wafer with photoresist spin-coated on the surface of the wafer. The photoresist is selectively exposed, so that the pattern on the mask is mapped to the photoresist layer. [0003] When the light emitted by the radiation source passes through the mask, it is affected by the edge of the pattern to be exposed (...

AI Technical Summary

Problems solved by technology

[0004] At present, the normal condition model (normal condition model) is usually used for optical proximity correction, that is, to measure the edge position error (Edge Placement Error, referred to as EPE), while the process window (ProcessWindow, referred to as PW) model (PW condition model) is usually used for Post OPC verification after optical proximity correction. Some target points under the model under normal conditions become weak points under the model of process window conditions. Points are areas that are prone to deformation and distortion. These points need to adjust the correction target, then perform optical proximity correction again, and then perform optical proximity correction calibration again, and so on. Weaknesses are eliminated, and the process of removing these weaknesses will cause a longer cycle time in the optical proximity correction process, which will affect the efficiency of production, and the production cost will be higher

Images