Use of Optical Emission Spectroscopy Data for Fault Detection of Mass Flow Controller in Plasma Etch Equipment

To minimize wafer yield losses by misprocessing during semiconductor manufacturing, faster and more accurate fault detection during the plasma process are desired to increase production yields. Process faults can be caused by abnormal equipment conditions, and the performance drifts of the parts or components of complicated semiconductor fabrication equipment are some of the most unnoticed factors that eventually change the plasma conditions. In this work, we propose improved stability and accuracy of process fault detection using optical emission spectroscopy (OES) data. Under a controlled experimental setup of arbitrarily induced fault scenarios, the extended isolation forest (EIF) approach was used to detect anomalies in OES data compared with the conventional isolation forest method in terms of accuracy and speed. We also used the OES data to generate features related to electron temperature and found that using the electron temperature features together with equipment status variable identification data (SVID) and OES data improved the prediction accuracy of process/equipment fault detection by a maximum of 0.84%.

Dual-Frequency RF Impedance Matching Circuits for Semiconductor Plasma Etch Equipment

The change in electrode impedance of semiconductor equipment due to repetitive processes is a major issue that creates process drift. In the current plasma etch chamber with a dual-frequency power system, the high-powered radio frequency (RF) source contributes to the enhancement of the plasma density, and the low-frequency bias power at the bottom electrode is adopted to enhance the injected ion energy in the plasma. The impedance control of the top electrode in dual-frequency capacity coupled plasma limits the impedance matching capability of the RF matching system because it only considers the high-frequency RF source. To control the precise impedance in dual-frequency semiconductor equipment, independent impedance control is required for each frequency. In this study, the impedance corresponding to a specific frequency was independently controlled using L (inductor) and C (capacitor). A 60 MHz stop filter and VVC were used to control 2 MHz impedance at a specific point, and a 2 MHz stop filter and VVC were used to control 60 MHz impedance. In the case of 2 MHz impedance control, the 2 MHz impedance changed from 10.9−j893 to 0.3−j62 and the 60 MHz impedance did not change. When controlling the 60 MHz impedance, the 60 MHz impedance changed from 0.33 + j26.53 to 0.2 + j190 and the 2 MHz impedance did not change. The designed LC circuits cover the impedance of 60 and 2 MHz separately and are verified by the change in the capacitance of the vacuum variable capacitors implemented in the RF impedance matching system.

Fabrication and repair of GaN nanorods by plasma etching with self-assembled nickel nanomasks

High crystal quality GaN nanorod arrays were fabricated by inductively coupled plasma (ICP) etching using self-organized nickel (Ni) nano-islands mask on GaN film and subsequent repaired process including annealing in ammonia and KOH etching. The Ni nano-islands have been formed by rapid thermal annealing, whose density, shape, and dimensions were regulated by annealing temperature and Ni layer thickness. The structural and optical properties of the nanorods obtained from GaN epitaxial layers were comparatively studied by high-resolution X-ray diffraction (HRXRD), Raman spectroscopy and photoluminescence (PL). The results indicate that damage induced by plasma can be successfully healed by annealing in NH3 at 900 °C. The average diameter of the as-etched nanorod was effectively reduced and the plasma etch damage was removed after a wet treatment process in a KOH solution. It was found that the diameter of the GaN nanorod was continuously reduced and the PL intensity first increased, then reduced and finally increased as the KOH etching time sequentially increased.

Multiscale Modeling and Recurrent Neural Network Based Optimization of a Plasma Etch Process

In this article, we focus on the development of a multiscale modeling and recurrent neural network (RNN) based optimization framework of a plasma etch process on a three-dimensional substrate with uniform thickness using the inductive coupled plasma (ICP). Specifically, the gas flow and chemical reactions of plasma are simulated by a macroscopic fluid model. In addition, the etch process on the substrate is simulated by a kinetic Monte Carlo (kMC) model. While long time horizon optimization cannot be completed due to the computational complexity of the simulation models, RNN models are applied to approximate the fluid model and kMC model. The training data of RNN models are generated by open-loop simulations of the fluid model and the kMC model. Additionally, the stochastic characteristic of the kMC model is presented by a probability function. The well-trained RNN models and the probability function are then implemented in computing an open-loop optimization problem, in which a moving optimization method is applied to overcome the error accumulation problem when using RNN models. The optimization goal is to achieve the desired average etching depth and average bottom roughness within the least amount of time. The simulation results show that our prediction model is accurate enough and the optimization objectives can be completed well.