"In regard to sustainability, not all PVD processes are equal."

In the recently concluded SVC TechCon 2021, PVD's role in shaping our sustainable future was a main focus. We strongly concur with and support our industry leaders in raising the awareness and promoting products/practices for this common goal.

A related subject to consider is the efficiency of PVD processes themselves -- not all PVD processes are equal in consumption of natural resources. During our R&D, we realized the need of an objective measure of PVD processes in regard to the efficient use of natural resources. Here, we introduce a figure of merit, M, to account for the efficiency of a PVD process: 

       M = R·V·(1-Nu) / W ,


  •   R is deposition rate in unit of nm/min;
  •   V is vapor capture: % of material vapor captured by the work pieces (or work- piece carriers);
  •   Nu is the thickness non-uniformity of the deposited layer;
  •   W is the electric power (in unit of kW) applied to the source for vaporizing the material.

For a given job, there can be various processes that meet the technical requirements. These different processes, however, can have different M values: one process may consume more (or less) precious natural resources than another.

As an example, we consider the deposition of a copper layer, by means of DC magnetron sputtering, on 16-inch-diameter substrates in the fabrication of an optoelectronic device [1]. Assume that the production requires: 1) 6 substrates are coated in a single batch; 2) the thickness uniformity is better than 1%; and 3) the deposition rate is no less than 20 nm/min. In the below, we analyze two processes.

In a conventional process design, the 6 substrates are mounted on a 6-planet fixture that undergoes a double rotation; the center of an 8-inch magnetron cathode is placed outside the circle-of-orbit of the planets. The deposition rate of 20 nm/min can be reached with 4000 watts of power applied to the cathode. The thickness uniformity is 0.4%; the vapor capture is 25.6%, which is typical of planetary rotation processes.

In an alternative and innovative process, pioneered by Vacuum Process Technology (Plymouth, MA), the 6 substrates are mounted on the same fixture. The center of the cathode (of the same kind), however, is placed inside the circle-of-orbit. The same deposition rate of 20 nm/min is obtained with 2350 watts of power to the cathode. The thickness uniformity is 0.3%; the vapor capture is 43.9%. 

The second process, with an M index of 3.74, is far more efficient than the first process which has an M index of 1.27. While both processes satisfy the production needs, the second process does so by consuming 40% less power and 42% less source material than the first process. 

As illustrated by this example, different processes can have startling difference in consumption of energy and raw materials, thus in their environmental impact. These differences are important but not always easy to quantify. Advanced numerical modeling is necessary if one wants to incorporate these aspects into PVD engineering. Tin Model LLC's latest software release, V-Grade 5S Pro, contains a set of tools for the purpose. We are proud that our simulation software is the first commercial product that makes these comparisons straightforward.

Coincidentally, M index also affects costs of manufactured goods: if everything else is equal, a higher M indicates a lower manufacturing cost because less power and raw material is consumed for the same end products.

Please contact us if you have comments or questions about this communication, or about our products. 

Phone: (857)498-9723 

*Reprint of our paper presented at the SVC TechCon 2021 is now available for download.
Copyright © 2015-2021 Tin Model LLC. All rights reserved.