In reply to Jones Beene's message of Tue, 29 Jan 2013 11:14:17 -0800: Hi Jones, [snip] >In the end - if you want to find a practical and gainful heat-to-electricity >device close to ambient, then provide the virtual sink well below ambient. >That may be difficult, but Dirac permits it and I would never argue with PAM.
..wasn't it you who first mentioned mercury based semiconductors with a very low bandgap on this list? Quite apart from that however consider that the kinetic energy of molecules tends to be distributed across all energy levels, so if energy can be withdrawn at *any* level, then that level will eventually be replenished by energy from the other levels (the sum of which will become depleted by the amount withdrawn). This is essentially what happens with wind-chill. There is a specific amount of energy required to break the hydrogen bonds between water molecules, and this is supplied by thermal energy of those molecules with sufficient kinetic energy, with the temperature of the liquid dropping to compensate for the lost energy, as the energy of the other molecules is redistributed. The implication of this is that a semiconductor with any bandgap should work, though I would think that those with a smaller bandgap would probably work faster as there is a larger population of low energy electrons than of high energy electrons. Regards, Robin van Spaandonk http://rvanspaa.freehostia.com/project.html