Those heat transfer cements (such at the Thermon one you posted) are more thermally conductive than air for sure. But you'll note that no thermal conductivity is listed on the spec sheet. And their resistance to thermal cycling is not infinite. As a filler between materials with differing thermal expansion coefficients, it would be tough for any continuous solid material to handle that cycling while remaining perfectly void-filling.

Depending on your geometry, aluminum powder might work, if you can tolerate it locally melting on the hot body.

One possibility might be: http://www.electroscience.com/pdf/44007.pdf but I've no idea how the binders would behave at an intermediate temperature, but it might be worth trying.

TTFN
FAQ731-376: Eng-Tips.com Forum Policies

Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers

MgO would help. OK heat transfer (sure better than air). Put it in as a slurry and then slow bake to dry it out.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

What about copper anti-seize? Good to 2000*F (1093*C) and should do well with cycling. Heat transfer properties may be a little hard to find, but I guess they would be pretty good.

Not exactly a full fledged recommendation, but worth looking into.

I think Copper anti-seize would only act as a surface coating at elevated temperatures to prevent surfaces binding. It would not fill voids and conduct heat.

Filling the voids may be irrelevant at the size you stipulated. A 6-mm gap filled with any thermal compound might as well be empty, compared with the surrounding metal's thermal conductivity.

However, the binders in most copper anti-seize compounds most likely will burn off into a carbonized mess.

What's already been cited and this: http://www.electroscience.com/pdf/4492.pdf are probably the only viable candidates for filling such large voids and providing a decent thermal conductivity.

As a reminder, the ultimate thermal conductivity of a thermal compound is only part of the story, since the compound is not supposed to be lathered, but instead, is supposed to be just barely thicker than most typical voids, so about 25 um. Given that, even if the material was mustard, thermal conductivity is still pretty good: http://www.hardwaresecrets.com/article/Thermal-Com....

A 25 um thick layer of polystyrene comes in at 2.4 kW/m^2-K.

TTFN
FAQ731-376: Eng-Tips.com Forum Policies

Need help writing a question or understanding a reply? forum1529: Translation Assistance for Engineers