friction on the nanoscale

Nanotribology: study dealing with the design, friction, wear, and lubrication of interacting surfaces in relative motion (like bearings or gears) on the nano-scale

Studies in the causes of macro-scale friction seem to reveal that it is do primarily to surface defects: vacancies, impurities. In other words, increased waste heat and wear and tear of machine components can be traced to the bumps and hollows in and on the components. Lubricant is used to decrease as much as possible the influence of these defects on the performance of the machine. When you move down to the atomic scale though, because you're dealing with individual atoms, those bumps and hollows that you found on the macro-scale would be other atoms. So, if your machine is kept in a relatively sterile environment (clean vacuum? or vacuum encased in a nanoshell of tightly woven material?), the effects of friction will be almost nothing and the use of lubricant is unnecessary. In fact, lubricant would merely clog up the parts of the nanomachine as the molecules of the lubricant could be as large as the components of the nanomachine.

If you picture friction effects in the context of macro-scale objects, those in our common reality, these bumps and hollows in substances as parts of machines will cause parts to move in an unwanted fashion slower and faster than if the bumps and hollows were not present. When, in our physics class, we released a ball from the top of a ramp, having it level out and then climbing a small hill, continue on that hill for a while, and then descending once again to its original low point, we noted that although the introduction of the bump (hill) in the path of the ball did not result in the ball losing kinetic energy relative to a ball that just continued along a straight path after descending a ramp of similar height to the first. Rather, the time the ball going across the bump takes is longer that of the ball on a path without a bump. This is because for a period of time the ball is travelling at a slower speed (when it is on the bump or hill). We can from this also conclude that if we introduce a small depression in the path of area after the ramp, the ball will reach the end of the ramp faster. Therefore when macroscale objects are rubbed or dragged against one another, generally speaking (a pencil tip across a rugged desk surface), even if constant force is being applied to each one, the velocities of each will be fluctuating ever so slightly.

Surface impurities add up and can lead to the breakdown of a machine. Given a wheel that spins around an axle, if axle has many surface impurities the wheel will not be distributing equal force around the axle, but rather on spots where there are higher bumps, less or no force on hollows. The bumps will get much hotter than the hollows, or given an axle with no surface impurities: much hotter than any given part of the axle, because each part is getting an equal amount of force applied to it, and therefore all parts should be at relatively the same temperature. When you have machines working extremely fast, lots of head needs to be dissipated from them. If there are lots of surface defects in a machine working really fast, parts of it could reach temperatures such that the materials it is composed of could warp, burn, come apart or do other bad things we don't want. Lubricants, our way of trying to deal with friction on the macro-scale, can only go so far: many can easily rub off, dry up, or burn off. On the nano-scale, no lubricants at all are required as friction as we know it on the macroscale doesn't even apply.

The only elements on the nanoscale that could be considered surface 'defects' are the exact spherical shapes of atoms and molecules themselves. If there are surface impurities in a nanomachine (aside from the shape of the atoms), that means lots of molecules are either attached to wrong places in the machine (bumps), or large chunks of the machine are missing (hollows). In either case the machine simply wasn't built correctly or something destroyed it. Any minor disturbances of the structure of nanomachines (unshielded) will result in the malfunction or no function of the machine. Although errors (incorrect construction, or damage) in nanomachines result in the devastation, if constructed correctly the only friction present will be a result of only the atoms of the components rolling over one another. If protected from the invasion of foreign molecules in the vicinity of a nanomachine, our conventional idea of 'hollows and bumps friction' will not be a problem. As compared with crude macro-technology this by magnitudes more efficient.

In nanosystems by K. Eric Drexler, frictional forces in mechanical systems are described to "convert mechanical energy into thermal energy in spatially inhomogeneous fashion, causing thermal gradients.". In other words, the objects in question develop 'hot spots' which earlier I've described the bad side of.

So, the hollows and bumps found in macro-scale objects that cause great friction and in turn hot spots and structural failure at high speeds, nanomachines lack the existence of these bumps and hollows and because of the this they will experience very minimal friction, only that from the sliding of organized groups of pre-arranged atoms. This greater efficiently allows greater speeds and therefore the ability to do 'better stuff' in a shorter time.

Sources:
http://www.aps.org/BAPSMAR98/abs/S5070007.html
http://stm2.nrl.navy.mil/~wahl/6176.htm
Nanosystems by K. Eric Drexler

main