Working in a vacuum

By Richard Halstead
President
Empire Magnetics Inc.
Rohnert Park, Calif.

Edited by Leland Teschler

Engineers faced for the first time with operating a motor in a vacuum are often surprised at some of the potential problems. Outgassing, contamination, and cooling are all areas of concern. And operation in a vacuum is becoming a hot topic. The trend is for large sections of semiconductor processing lines to operate in a vacuum. This mode of operation is now seen as a means of preventing the transport of contaminating particles between process chambers. Moreover, keeping large parts of a fab line in a vacuum eliminates the time spent pumping down individual process chambers to the required vacuum levels. The fab process thus becomes more of a continuous flow as opposed to a batch-mode operation.

Leading-edge technologies in semiconductor manufacturing are spinoffs of development work originally done to create the beam lines found at National Laboratories. The long vacuum chambers and equipment used to make these lines will probably be the prototypes for wafer fab lines of the future.

There are now motors that address the problems associated with operating in a vacuum. Some of these designs came out of joint efforts with Lawrence Berkeley National Labs aimed toward vacuum and space applications.

Automation requirements
Historically, designers have located motors outside vacuum chambers. In this type of control solution, the drive mechanism transmits its motion through the vacuum chamber wall using sealed couplings.

But there are a number of disadvantages to this traditional vacuum-motion-control approach. One is that it limits design options. For instance, it is difficult to implement an X-Y stage (where one stage moves on top of the other) inside a vacuum chamber when the motors sit outside the chamber. The mechanical components used to transfer motor power greatly restrict the design possibilities. Further, this approach compromises the accuracy, repeatability, and resolution of the positioning system inside the vacuum chamber.

On the other hand, placing the motor directly in the chamber simplifies things. There are a larger number of possible physical arrangements; motor cables can be routed in a number of ways without inhibiting the chamber activity. Directly coupling the motor to the load also vastly improves accuracy and other system specifications. Finally, it is noteworthy that mechanical feed-throughs for the vacuum system are often as expensive as a vacuum motor in its entirety. So a system that features an internally placed motor often costs less than the alternative with the motor outside.

For all these reasons, more and more vacuum applications use specialized motors that take into account the special environmental conditions of a vacuum. Note that standard motors are inappropriate here. In general, standard motors will not survive in a vacuum of 10^-4 Torr or lower. The primary reason is that lubricants in the bearings will vaporize and the insulation materials of the motor and cable will evaporate, a phenomenon known as "outgassing." Outgassing within a vacuum chamber is obviously bad: In addition to destroying the motor, the vaporized materials condense on precision optical components and delicate mechanical devices, fouling the application.

Some materials like petroleumbased bearing grease vaporize so quickly that they literally create clouds of vapor in the vacuum chamber. Other materials vaporize more slowly, but can become an application nightmare. Silicone is one of these nightmares. Once a vacuum chamber has been contaminated with silicone, it is nearly impossible to clean it all out. It will continue to spread to anything and everything placed in the chamber. Additional materials notorious for outgassing include paper slot liners, conformal coatings, winding insulation, and many kinds of cements or glues.

Motor cooling in a vacuum is also a problem. Conventional motors cool primarily by convection into the air that normally surrounds them, and to a lesser degree by conduction through the mounting surface. When a motor is energized in a vacuum, convection cooling is obviously nonexistent. Heat dissipates primarily by conduction from the motor to the mounting structure. As a result, there must be provisions for heat exchange and higher operating temperatures when running a motor in a vacuum.

Leakage is another difficulty. Even if a standard motor does not outgas contaminants in a vacuum, it is likely that the vacuum process will be hampered with the motor present. For example, a stepmotor that is not properly treated will "leak" air molecules long after vacuum is applied. Leakage is the slow release of captured or clinging gas molecules from minute cracks in the motor laminations, windings, bearings, and metal surfaces. Leakage is a problem when the surfaces of these porous materials are not treated, resulting in unacceptably long pumpdown time or an inadequate vacuum level. This leakage is a mechanical problem as opposed to the materials problem of outgassing.

Finally, high-voltage exposed conductors on motors in a vacuum can create corona effects. At certain vacuum levels, the rarified air will easily ionize, and current will flow between unprotected high voltage conductors (this principle was the basis of the well-known electron tube).

Properly designed vacuum motors address these vacuum chamber/motor design problems. Outgassing, for instance, can be prevented by the careful selection of appropriate materials. Teflon is commonly used in vacuum applications as it is quite stable, has good temperature qualities, and is readily available at a reasonable price. Most metals are acceptable for use in a vacuum (exceptions include cadmium and zinc), but stainless steel is particularly good here.

Design for vacuum
In general, many commercial plastics have outgassing rates that are acceptable at up to 10^-4 Torr, but lubricants used in this range usually need to be selected carefully. A vacuum in the 10^-7 Torr range eliminates most natural materials, and only a limited number of plastics are usable. At this pressure, vacuum lubricants are essential. At 10^-9 Torr, most plastics are excluded and dry lubricants are a must.

Outgassing can also result from a lack of motor cleanliness. Motor materials are subjected to a variety of contaminants during the manufacturing process. Trace materials are always left behind. Steel is exposed to cutting oils, plastics are lubricated when extruded from dies, and epoxies get mixed with solvents. Additionally, human hands will leave behind a residue of oil when they touch motor parts.

Different applications require varying degrees of vacuum purity. But without proper component cleaning, there will definitely be outgassing of various contaminants into the vacuum chamber.

Special measures are called for to handle rigid specifications of chamber contamination. For example, Empire Magnetics puts its vacuum motors destined for such applications through a proprietary extraction and cleaning process. This process gets into the deep crevices that vapor degreasing cannot reach, accelerating molecular changes in contaminants that would otherwise outgas, rendering them inert. In general, the sensitivity of vacuum environments usually requires the use of motors constructed of nonvolatile materials, vacuum baked, processed to extract contaminants, and then sealed.

High temperatures can accelerate outgassing. So it is desirable to select a motor and drive voltage that minimize generated heat. High-voltage PWM (pulse-width-modulated) drives heat motors more than lowvoltage linear drives. Bipolar drives using all of the copper in the motor at one time generate less heat than unipolar drives that energize half the copper. Other systems that produce less heat include those that automatically reduce standby currents or servo systems that cut back current when the motor is not moving.

Sensors that include thermistors, thermocouples, or RTDs (resistance thermometer device) can monitor motor temperature. The information can be used to modulate power to the motor to keep temperature within the safe operating range. Cold plates or cooling jackets are a possibility in cases involving substantial generated power and a consequent rise in temperature.

One example of a successful cooling jacket was designed and used by Hughes Aircraft. The cooling medium was vapor from liquid nitrogen. It was controlled by a feedback system that monitored the motor temperature. This system held the temperature to a comfortable operating range, though the motor was heavily loaded for three months of continuous satellite testing.

Leakage in a vacuum/motor application can be overcome by eliminating the cracks, crevices, and other areas that trap gasses within the motor. If the motor is not properly treated before installation in a vacuum, the laminations, windings, bearings, and even metal surfaces will release air molecules trapped in surface cracks. The motor will then need a significant period in a vacuum to release all of its trapped air, perhaps lengthening the time needed to reach a desired vacuum level.

Selecting fine machine finishes that hold less air will minimize leakage. For instance, porous metals typically require cleaning and sealing, and machined metals are preferable to castings. If castings are unavoidable, they should be modified with a machined finish. Even still, additional cleaning may be required depending on the level of vacuum required. To further reduce leakage, the preferred technique is to turn blind screw holes into through-holes and treat all surfaces with appropriate sealers. To prevent the corona effect that can be generated by high voltage, exposed conductors must be insulated with appropriate materials to prevent arcing.

Motor grades
Because each vacuum application has differing requirements, there are a number of different vacuum-rated motor grades available. Three of the most common types are commercial, standard, and laboratory grade.

Commercial grade motors will survive in a vacuum down to 10^-7 Torr, provided the motor does not exceed its rated operating temperature. These motors aren't specially treated to reduce outgassing but are built with materials that will not evaporate quickly in a vacuum.

Standard grade motors are built with the same materials as commercial grade motors, but are cleaned in an ultrasonic cleaner and vapor degreaser. They are also vacuum-baked and sealed to minimize recontamination.

Laboratory grade motors are cleaned and baked like the standard grade, but also undergo an extraction process to remove contaminants. Motor windings often use higher-temperature insulation, which allows winding temperatures 50C higher than the other grades. These motors should be handled only under clean-room conditions, with new nylon gloves (Even clean rubber gloves would contaminate the motor).