Electricity from layered chemical inks

I saw this video of Robert Murray-Smith's channel on Youtube. He makes interesting videos about making graphene inks... in a kitchen blender

In this video he demonstrates his work on 'monotherm' strips which purportedly generates "electricity from ambient heat."

https://www.youtube.com/watch?v=wSZGUV8D19I

His work is a variant based on the US Patent attached [1].

His one cell creates a decent potential of 1.6V at minute currents (voltmeter load), but it's unclear what power these things can deliver per unit area.

I sat through his explanation of electronegativities, but still don't understand quite how these devices operate and would appreciate some insights and discussion of what's going on here. Also how does this utelise ambient heat to create work... as an engineer, I thought you always need a temperature difference to harness work?

At first I thought this voltage was electrochemical in nature, but how does it work in the absence of electrolytes?

Reference:

[1] Lovell, WC (2000), Method of making a laminated device for generating electrical energy. US Patent 6103054.

Attachment: US6103054.pdf (759kB)
This file has been downloaded 478 times

[Edited on 10-1-2015 by deltaH]

Further on into this vid, he does shed some extra light on the matter, but still not so sure exactly what's going on:

https://www.youtube.com/watch?v=aNGoXjQpU0E&src_vid=yBX8...

I am struggling to understand some of what he says... very strong accent. Did he say that this is related to MEM diodes? He says that these operate with quantum tunnelling, but these painted devices are too thick for tunneling, no?

****
Just saw from the patent that 2.4mW/in^2 can be generated at 212°F, which while weak, is much more than I would have expected.

Bear in mind that his is for a very thin strip, so in principle this can be stacked to go from an area device to a volume one and multiply the power accordingly. So for example, let's say the strip has a thickness of 0.5mm (thumb-sucked) and is 2.54 cm wide (1in) and 1m in length and you stack them until you get a height of 1m as well. Then your 1m by 1m by 2.54cm 'panel' would generate close to 5W at 212°F from practically 'thin air'... ok very hot 'thin air'

[Edited on 11-1-2015 by deltaH]