Vacuum on the surface of the brakes implies you have to pump the heat through diffusion, not convection.
On the other hand
I'm not totally sure everybody understood me: one of the most corrosive materials in the world is monoatomic oxygen (when oxygen molecules reach 3 to 5 ev, they dissociate from O2 to O).
You can check here for its effects on the shuttle, where the light it creates when it impacts the shuttle can be easily observed:
http://www.spenvis.oma.be/spenvis/help/ ... osion.html
http://pdf.aiaa.org/jaPreview/JSR/1988/PVJAPRE25965.pdf
Oxygen glow on the engine pods of the Shuttle
Of course, the effect on the shuttle materials is tremendous: the oxygen flow impacts the vehicle at 18.000 kph, magnifying its effect. Carbon fiber is particularly affected by it: losses of up to 13% of material in a few days have been measured when in space. That was one of the problems that kept the Galileo vehicle from being "parked" for too long at low earth orbit: the craft was "melting" away.
This free oxygen is created in many instances, one of them by high temperatures, like in the brake pads of F1 cars.
For example, you could check the way Bessemer converters are handled: you literally throw lumps of clay to the bottom of the oven to avoid holes in it, holes that are caused on the refractary materials by heat and free oxygen. These holes, of course, would be catastrophic: the molten steel escaping from the Bessemer converter would explode on contact with any humidity. This was the cause of many accidents in the beginnings of Bessemer oven use. Next time you visit a Bessemer installation, talk with the guy in charge of the oven.
The monoatomic oxygen erodes the brake pads, I assume. That's why I suggest to investigate if its effects can be avoided.
Dave thinks that if you put the pads in a vacuum they could not be cooled. I do not agree: there are many instances of materials cooled by diffusion.
For example, that's the way the heat sink in your PC works. It's irrelevant that the fins themselves are cooled by convection and radiation: the CPU is cooled by diffusion through contact with the base of the fins.
Finally, to clarify my suggestion, if you somehow are able to avoid air in contact with the pads, you could use materials that react with oxygen, that, by chance, are the most useful to transmit heat and that
are the ones that keeps better their strength when heated.
I hope somebody gets it.