The Princeton Plasma Physics Laboratory (PPPL) of the U.S. Department of Energy (DOE) has joined the industry’s efforts to expand the process and develop new ways to produce more powerful and efficient chips. As part of the first PPPL study conducted in collaboration with Lam Research Corp., a global supplier of chip manufacturing equipment, laboratory scientists accurately predicted the key stage of chip manufacturing on an atomic scale using simulation.
Laboratory specialists have modeled the so-called “atomic layer etching” (ALE). The purpose of ALE is to remove individual atomic layers from the surface at a time. The process is used to etch complex three-dimensional structures with critical dimensions, thousands of times smaller than a human hair, into a film on a silicon wafer.
“The simulation basically coincided with the experiments as a first step and could lead to a better understanding of the use of ALE for etching at the atomic scale,” said Joseph Vella, a researcher at PPPL. The improved understanding will allow PPPL to investigate such things as the degree of surface damage and the degree of roughness that occurs during ALE, and it should start with a fundamental understanding of atomic layer etching.”
The model simulated the sequential use of chlorine gas ions and argon plasma to control the etching process of silicon on an atomic scale. Plasma, or ionized gas, consists of a mixture of free electrons, positively charged ions and neutral molecules. The temperature of the plasma used in the processing of semiconductor devices is close to room temperature, unlike super-hot plasma used in fusion experiments.
“An unexpected empirical finding of Lam Research was the understanding that the ALE process becomes especially efficient when the ion energy is significantly higher than at the beginning of the work,” said David Graves, professor of chemical and biological engineering at Princeton. “So the next step in the simulation is to see if we can understand what happens when the ion energy is much higher, and why this process is so good.”
In the future, “the semiconductor industry as a whole is considering the possibility of significantly expanding the range of materials and types of devices used, and such an expansion will also require processing with atomic-scale accuracy,” Graves said.
The scientists highlighted the results of their work in an article published in the Journal of Vacuum Science & Technology B.