Method and device for optical critical dimension measurement

Abstract

The invention describes a method and a device for improving spectral curve matching reliability and measurement accuracy in optical critical dimension (OCD) measuring equipment. It includes performing sensitivity analysis on the theoretical spectrum of each structural parameter to be measured at all wavelength points and obtaining the distribution of sensitivity with wavelength; normalizing the sensitivity of each structural parameter to be measured with wavelength distribution; flexibly setting the weight coefficient of the system and The normalized sensitivity is systematically processed along with the wavelength distribution; in the matching process of the theoretical spectrum and the measured spectrum, the matching weight coefficient is set at each wavelength point and the theoretical spectrum and the measured spectrum are evaluated and optimized to determine the difference between the two. The degree of matching; in this way, the optimal design of matching evaluation is realized and the accuracy of the fitting value of the structural parameters to be tested is improved.

Description

Technical field [0001] The present invention relates to the measurement of optical critical dimensions of a semiconductor manufacturing process, and more specifically, to a measurement method for improving the reliability of spectral curve matching and measurement stability in optical critical dimension OCD (Optical Critical Dimension) measuring equipment. Background technique [0002] In the integrated circuit industry, with the shrinking of process nodes and the complexity and diversification of device structures, the manufacturing and production of very large-scale integrated circuits have increasingly higher requirements for the accuracy and sensitivity of measuring equipment. As an important critical dimension measurement technology, OCD (Optical Critical Dimension) measurement technology has gradually occupied a dominant position in the production of logic devices and memory at 65nm and below process nodes. [0003] With the development of the semiconductor integrated circuit...

AI Technical Summary

Problems solved by technology

[0010] OCD measurement equipment includes many hardware system parameter designs. In the optical measurement process, engineers are striving to achieve high-precision control of the system parameters of each OCD measurement equipment, reducing hardware errors (such as the accuracy limit of moving parts and precision positioning subsystems) It causes deviation to the diffracted light signal collected in the measurement, but the ideal zero error cannot be realized, and the total spectral noise formed by system noise, uncertainty and random error is inevitable

The existence of noise leads to a decrease in the matching degree between the measured spectrum and the theoretical spectrum, and at the same time causes the fitting value of the measured structural parameters to deviate from the true value, and more importantly, affects the repeatability and accuracy of industrial measurement

[0011] Due to different semiconductor manufacturing processes and changes in the structure of the device to be tested, it is generally difficult to consider the sensitivity difference of the spectral signal at each wavelength in the design process of the system parameters of the usual OCD measurement equipment

In the process of evaluating the matching between the measured spectrum and the theoretical spectrum, the weight of all wavelength points participating in the matching is used to adopt an indiscriminate matching method, which restricts the improvement of sensitivity, signal-to-noise ratio and accuracy of measurement results

Images