Aerospace manufacturers long ago solved the problem of lubricant outgassing, contamination, and starvation. Their work is now reaping benefits in the semiconductor industry.
KEVIN AKIN DAN SHEA
Nye Lubricants Inc.
Automated assembly methods have created a conundrum for semi-conductor manufacturing. No surprise that robots, pick and place stations, conveyors, and other devices on the production line run better when lubricated. But oils and greases contribute to airborne molec-
Automated assembly methods have created a conundrum for semi-conductor manufacturing. No surprise that robots, pick and place stations, conveyors, and other devices on the production line run better when lubricated. But oils and greases contribute to airborne molecular contamination. Worse, they can give off vapors that may fog optics in high-speed inspection systems or may even contaminate wafers.
Fortunately there are ways to eliminate lubricant-related problems. Case in point: The aerospace industry has for decades qualified lubricants in mission-critical components, addressing problems such as lubricant outgassing, contamination, and starvation. Much of this research can apply to lubricants for sub-assemblies in clean rooms, laboratories, and semiconductor manufacturing facilities.
Lubricant chemistry primer
Lubricants reduce friction, the normal force of the adjoining materials multiplied by the coefficient of friction (μ), which is the resistance of the softer material to plastic deformation under shear and compression (μ = S/P). Placing a lubricant between touching surfaces reduces the coefficient of friction, thus diminishing component wear.
Solids and liquids can both serve as lubricants. Solid lubricants include graphite, molybdenum disulfide, and polytetrafluoroethylene (PTFE) powder. Solid lubricants are easily abraded and alone are generally short lived and unsuitable for long-term uses.
The most common liquid lubricants are mineral and synthetic oils. Generally, all synthetic oils withstand broader temperatures and are more chemically homogeneous than their natural counterparts. Specific synthetic oils offer additional advantages. Typically, oils get the nod for low-power devices where available torque won't over-come even the lightest grease.
Grease fits in between solid and liquid lubricants. A semisolid, it is an oil that has been immobilized with a thickening agent. Unlike oils which require tightly sealed reservoirs to prevent migration, greases rely on soap, clay, or solid-lubricant thickeners to make the lubricating fluid stay put.
Base oil is the biggest component of any lubricant. General classes of synthetic base oil chemistries include: synthetic polyalphaolefins (PAOs), synthetic esters, silicones, multiply-alkylated cyclopentanes (MACs), polyphenylethers (PPEs), and perfluoropolyethers (PFPEs). Each brings different advantages to clean rooms and semiconductor manufacturing.
PAOs generally cost the least. They are good at preventing wear and can be fortified with traditional additives. But they don't always measure up at higher temperatures or where minimal outgassing is critical.
Synthetic esters and silicones both perform better than PAOs and cost more. Because esters are highly polar, they are intrinsically good boundary lubricants. Compared to PAOs, they offer lower vapor pressure and higher thermo-oxidative stability. But esters should raise a caution flag. They react chemically with several commercial polymers and elastomers, and they can break down in the presence of acids, bases, and certain metals. This seriously degrades lubricant performance.
Silicones work significantly better than PAOs and esters. They have much wider service temperature ranges, low volatility, and good thermo-oxidative stability. But their molecules are very "compressible." This is no problem when lubricating plastics and lightly loaded metals. Silicones, though, don't resist wear as well as esters or PAOs in highly loaded, metal-on-metal situations.
MACs, PPEs, and PFPEs all high-molecular-weight materials are the latest developments in synthetic chemistry for lubrication. They cost more, but their special properties suit the demands of semiconductor fabrication. MACs exhibit exceptionally low vapor pressure, can be readily fortified with application-specific additives, and offer good thermo-oxidative stability and chemical compatibility.
PPEs have long lubricated electrical connectors. They offer low-vapor-pressure qualities and a resistance to radiation that is one to two orders of magnitude better than that of other synthetic chemistries. Though they are not as robust as some other synthetic chemistries for boundary lubrication, PFPEs are already well known in vacuum-critical applications, especially when low-vapor pressure, wide temperature, chemical inertness, and material compatibility are important.
Lubricants destined for vacuums or clean rooms must meet rigid vapor pressure standards. Vapor pressure is the force per unit area exerted by gas-phase molecules on an imaginary surface immediately above the liquid surface. This property helps those designing lubricants to compare the tendency of different oils or other lubricant components to become volatile. Volatility boosts the chance of contamination from evaporative loss. Thus, the lower the vapor pressure, the better the lubricant.
A lubricant's vapor pressure, which is measured in a vacuum, comes from Langmuir's equation: P = 17.14 G (T/M), where P = vapor pressure, Torr; G = rate of evaporation, gm/cm 2 /sec; T = temperature in °K, and M = molecular weight of the material in question.
When designing a lubricant, the surface area of the device to be lubricated (a factor of G) and the temperature of the operating environment are both givens. So the above equation underscores the importance of molecular weight, a variable linked to the selection of the base oil, in designing a low-vapor-pressure lubricant.
The best candidates for low-outgassing are base oils with high molecular weights, materials such as MACs, PPEs, and PFPEs. Further, because each molecule in a base oil can have a slightly different molecular weight, narrowing the range of molecular weights is better still. A special distillation process can remove lighter materials during lubricant manufacturing, instead of waiting for them to volatilize and contaminate on the job.
It is important to note that while most "vacuum lubricant" suppliers specify vapor pressure, this data is seldom the result of measuring each lot of lubricant. Without close control, the vapor pressure of two batches of the same lubricant may differ by several orders of magnitude. So it's best to conduct per-lot assays in determining a lubricant's outgassing potential.
It is also important that lubricants destined for clean rooms and vacuums be free of particulates. Dust, dirt, and even improperly dispersed thickeners can become entrained during manufacture of an oil or grease. In rolling-element bearings, for example, particulates or agglomerated thickener can rupture the elastohydrodynamic film that separates the balls from the raceway. Scarring of balls and raceway can result. In fact, most bearing failures today arise from lubricant problems. Generally, anything less than superclean lubricants can compromise device accuracy, repeatability, and operating life.
Ultrafiltration is a sophisticated pre and postproduction process that can ameliorate these difficulties. It homogenizes thickener in grease and removes particulate matter as small as 1 micron from oils and 35 microns from grease (for reference, a grain of sand is about 650 microns). Because various military standards define lubricant cleanliness, ultrafiltration, at a minimum, tells the user the relative sizes of engrained particles. That information can help determine if particles are big enough to harm operation.