⚡DC Sputtering (WIP) (UWaterloo)
Preface
Although Waterloo's DC sputtering efforts began in 2024, the majority of work between Fall 2024 and Spring 2025 went undocumented. This page details the state and lessons learned of Waterloo's sputtering machine from the Summer of 2025 onwards. DC Sputtering was chosen primarily due to the relative accessibility of second-hand high voltage DC power supplies, although our long term goal is to move to an RF power supply.
Contacts:
Skye Koh (@skyekoh)
Eric Jessee (@eejay)
Current Specifications (April 2026)
Machine:
Chamber base pressure: 5 to 10mTorr ✅
(Minimum pressure achievable by roughing pump)
Max power: ~20W DC
Magnets will overheat after this threshold with sustained use
Thin Films:
Copper Oxide / Copper
Minimum resistance achieved: ~20Ω
Design
Power Supply
In choosing a DC power supply, the voltage of the power source must be high enough to provide sufficient force to accelerate electrons until they carry enough kinetic energy to ionize the sputtering gas molecules on collision.
Our initial solution was to re-purpose existing power supplies designed for electrophoresis processes. On paper, these power supplies are capable of delivering the voltage and current required for sputtering, and are ubiquitous on the second-hand market. They are also generally designed with built in ground leakage and arc fault protections, which at first glance appear to be welcome safety features. For these reasons, in order to minimize cost and increase safety for initial testing, an electrophohresis supply was purchased second hand, pictured below. The drawbacks of this type of supply are discussed in the Lessons Learned section.
Vacuum Chamber Design
To minimize cost as well as design/manufacturing complexity, the chamber is comprised of a borosilicate glass bell jar, aluminum baseplate, Viton gasket, two electrical feedthroughs, and two KF inlet ports. The baseplate was manufactured using a milling machine in the school's student machine shop, although a drill press would suffice.
The rest of the vacuum setup included a roughing pump connected by a KF hose, a vacuum gauge, and an Argon inlet controlled manually by an isolation valve and needle valve. KF connections were possible to minimize leaks. The overall schematic is shown below.
Electrical Feedthroughs
The redesigned electrical feedthroughs were a significant breakthrough in reducing the pressure of the chamber by mitigating potential leak paths. These custom feedthroughs are made entirely from off-the-shelf low-cost components, comprised of a threaded rod, hex standoff, PTFE spacer, and Viton gasket. When the argon gas line is isolated from the chamber through the right-angle valve the chamber achieves a vacuum of 5 to 10mTorr, which is within the range of the maximum vacuum the roughing pump can pull, proving the effectiveness of the feedthrough.
Source Design
For ease of manufacturing, the size of the source was reduced to a diameter just over 30mm. The ground shield was improved to include a lip that partially covers the target to improve the plasma confinement and ensure only the target material is sputtered. Additionally, the magnet enclosure is made from mild steel to act as a pole piece, which provides a path for the magnetic flux to create an unbalanced magnetron, improving plasma confinement on the target.
Because the goal of this source design was primarily to validate the design of the chamber, confinement of the plasma, and use of the DC power supply, the source design does not include active thermal management for the magnets. The next iteration will be designed with active cooling, either through air cooling or water cooling.
Testing and Deposition Results
The sputtering source achieved stable plasma confinement using an electrophoresis power supply with minimal arcing between 10-200mTorr.
Copper was the only metal we sputtered, to varying degrees of success. Many trials exhibited high resistivity, indicating impure films and the presence of copper oxide. The lack of turbopump or Argon mass flow controller on our machine limits our ability to deposit high-quality films.
Lessons Learned
Power Supply
We were able to carry out many experiments with this supply, and were able to learn a lot about the characteristics of plasma formation and maintenance. Unfortunately, one of the main features that the power supply was chosen for (ground and arc fault protection) ended up being a major limitation, making it very difficult to strike and maintain plasma.
As such, an electrophoresis power supply is unsuitable for this application, as the in-built safety features are not designed to tolerate plasma behaviour, and make it very difficult to use. In order to continue with the DC sputtering process, a custom power supply must be constructed, and other engineering controls should be put in place to account for the lack of arc and ground fault protection.
Bill of Materials
Most raw materials were purchased from the UWaterloo Engineering Machine Shop. However, equivalents can be found on McMaster-Carr or a local metal supermarket.
1/2" 6061 Aluminum Chamber Baseplate
Raw Material
$24.81
Baseplate
Borosilicate Glass Vacuum Bell Jar
EBay
$205
Swap out for an ordinary thick borosilicate glass jar (at your own risk)
Cold Rolled Steel Magnet Enclosure
Raw Material
$3.14
Pole piece
6061 Aluminum Ground Shield
Raw Material
$3.86
Sputter only target material + plasma confinement
Feedthrough Hardware
McMaster-Carr, Various
~$30
Electrical vacuum feedthroughs
Total (Excluding roughing pump)
$574.87
Next Steps
Vacuum
Turbopump
Our team has just acquired an Edwards EXT70H Turbomolecular Pump and EXDC160 Drive Controller. Integrating this into our system will require a full resdesign of the vacuum chamber, as well as a custom control system and new power supply for the Turbopump.
Implement Mass Flow Controllers
Power Supply
Design a new RF or DC Power Supply
Sputtering Source
Active cooling (air cooled or water cooling)
Last updated