MIT boffins claim they reached higher efficiency for thermal photovoltaics:
http://web.mit.edu/press/2009/thermoelectric.html
Their paper:
http://link.aip.org/link/?JAPIAU/106/094315/1
Citation:
The maximum load power occurs at 27.5 mV, where the calculated efficiency is 26% (52% of the Carnot limit). The maximum efficiency occurs at 42 mV, where the calculated efficiency is 43% (86% of the Carnot limit).
Well that’s for the simulation. Technology Review (a few months old) states a different figure for the prototypes:
http://www.technologyreview.com/energy/21981/
Citation:
The prototypes aren’t this efficient: they convert about 10 to 15 percent of the heat that they absorb from the glass-factory exhaust into electricity, which DiMatteo says is enough to make the devices economical. (The expected efficiency of TPV devices is also much higher than efficiencies anticipated for thermoelectric devices, which directly convert heat into electricity.)
The cooperation involved:
http://www.mtpvcorp.com/
Doesn’t match the efficiency of turbines yet, but might surpass them someday.