Modeling oxide etching in a magnetically enhanced reactive ion plasma using neural networks

Abstract

Oxide films etched in a CHF3/CF4 plasma were qualitatively modeled using neural networks. The etching was conducted using a magnetically enhanced reactive ion etcher. A statistical 24-1 experimental design plus one center point was used to characterize relationships between process factors and etch responses. The factors that were varied include a radio frequency power, pressure, CHF3 and CF4 flow rates. The resultant nine experiments were used to train neural networks and trained networks were subsequently tested on eight experiments not belonging to the training data. A total of 17 experiments were thus conducted for modeling. Etch responses modeled are etch rate and etch profile. Root-mean-squared prediction errors of optimized models are 111 Å/min and 3.50° for the etch rate and etch profile, respectively. Interaction effects between the factors were examined from the prediction models. Besides the dc bias, the F and CF2 intensities measured with an optical emission spectroscopy were related to the etch responses. Polymer deposition effect was transparent in the etch rate for a variation in CHF3, but little for the one in CF4. Conflicting effects between CHF3 and CF4 were noticed with respect to either etch response. Pressure-induced dc bias played an important role in determining etch responses.

Virtual metrology for plasma etch using tool variables

“…The other parameters, m and η are the training iteration number (the gradient descent is iteratively completed on the dataset until an error minimum is found) and the learning rate respectively [15]. Test datasets can be used in parallel to the training to assess when networks are becoming overtrained on the training set.…”

Optimization of Principal-Component-Analysis-Applied in Situ Spectroscopy Data Using Neural Networks and Genetic Algorithms

“…The processes to fabricate test patterns have been explained in detail. 20 To systematically model the experimental data, the etch process was characterized by means of a 2 4-1 fractional factorial experiment with one center point. 21 This type of statistical experiment is often adopted as the experimental budget is limited.…”

Partial diagnostic data to plasma etch modeling using neural network

“…Actually, the oxide film is largely etched by energetic ions, which react with the oxide and produce volatile etch products. This dependency has also been characterized by means of a prediction model [14]. Fig.…”