Wafer CMP material removal rate prediction method of GMDH neural network

Abstract

The invention relates to a wafer CMP material removal rate prediction method of a GMDH neural network. The method comprises the following steps: (1) obtaining a polishing sample data set after abnormal value removal; (2) analyzing samples in the polishing sample data set, and determining b effective process variables; (3) extracting the mean value, the standard deviation, the skewness and the kurtosis of each effective process variable to obtain 4 * b feature vectors; (4) screening the correlation between the 4 * b feature vectors and the corresponding MRR values, and determining m feature vectors as input feature vectors of the GMDH neural network model; (5) performing normalization processing on a data set formed by the m feature vectors to obtain a training feature set; (6) adopting a binary quadratic Volterra polynomial regression model, taking an input feature value in the training feature set as an input layer, taking a correspondingly output MRR value as an output layer, and obtaining a trained GMDH network model; and (7) inputting m characteristic values serving as inputs in a to-be-predicted sample into the trained GMDH network model, and outputting a predicted MRR value.

Description

technical field [0001] The invention belongs to the technical field of semiconductor material prediction methods, and relates to a method for predicting the removal rate of wafer CMP materials by a GMDH neural network. Background technique [0002] Chemical-mechanical polishing (CMP) is the mainstream downstream process of wafer manufacturing; the purpose of this process is to overcome the problem of multi-layer metallization of wafers. CMP is the passivation and etching of the wafer material by slurry chemicals, that is, the wafer surface is planarized by sliding its surface on the slurry particles under downward pressure. Wafer CMP processes are very complex and involve multiple chemical and mechanical phenomena, such as surface dynamics, electrochemical interfaces, contact mechanics, stress mechanics, fluid dynamics, and tribochemistry. [0003] In the CMP process of the wafer, MRR is an important index to measure the performance in the process (MRR value is the material...

AI Technical Summary

Problems solved by technology

[0007] In order to solve the problem that the material removal rate in the CMP technology cannot be accurately obtained in the prior art, the present invention provides a GMDH neural network wafer CMP material removal rate prediction method, which is based on a complex system combining physical knowledge and statistics. model method; specifically, the adaptive model method of the GMDH neural network is used to ensure that the MRR value is within the normal range to improve the accuracy of the removal rate through accurate prediction of the MRR value. The range of MRR value is 140-170nm / min, 50-110nm / min. If the predicted value does not meet this range, adjust the process parameters in time, such as replacing consumable materials such as new dressers and polishing pads in time

Images