After the various arguments I’ve been making about being able to harvest thermal energy from the environment, it’s looking like the US army agrees with me. US patent 9640698 was issued on May 2nd 2017 (if you want to download this as a .pdf, try patent2pdf and put in the number). Here, they are using variations on MerCaT (an Hg/Cd/Te alloy) which has a low and adjustable band-gap but is a little tricky to actually deal with. As regards the structure of the device, it’s much the same as a standard 2 or 3-junction PV with some modifications to increase the effective surface area and thus the amount of energy harvested. Fairly obviously, this sort of device would be pretty useful for the military – no need to replace batteries and it just delivers power all the time, and the expense is really not important. The MerCaT materials are after all pretty expensive and hard to work with.
If you read the patent, you’ll find that, like the NASA patent on LENR, they’ve used a scatter-gun approach to their claims, and have specified (as they think, at least) every which way possible to actually make the device work with a range of materials. For the NASA LENR patent, it was pretty obvious that they were covering all the bases they could think of yet hadn’t actually made it work, since otherwise we’d be seeing commercial devices – more like poisoning the well than actually being useful in telling people what worked. It’s possible that these thermal PVs have actually been made, too, though the expense would deter them from entering the mass market which may be why we haven’t seen them around. It’s also possible that they’ve only made a few trial wafers and found it too difficult to deal with the materials, of course. As far as I can see, the only new thing in the patent is the convoluted structure they specify which increases the effective surface area, since after all the PV structure is well-known and you can use any semiconductor in that structure, and MerCaT is well-known in the literature too. MerCaT devices have been for sale for a long time. Luckily you can’t patent a law of nature, though US patents can run pretty close to that.
As it happens, I think they’ve missed a few points in the design and what’s actually happening in the semiconductors, and though they claim that there is around 500W/m² available in the 2 micron to 40 micron IR bandwidth, their design won’t harvest anywhere near that amount and they are really looking to put this in the sunlight where the energy-levels in the 2 micron range are in fact pretty high. Once the Sun goes down, they’ll be limited to a peak of around 1W/m² from the ~300K room temperature LWIR radiation, so the output of their panel will go way down at night. It will definitely still give power, though, which is the main point of the device and where this differs from a standard solar panel. Still, they haven’t really thought this through properly and though they do say the words “perpetual energy harvester” and thus give a nod to perpetual motion by inference, they aren’t admitting that it’s breaking 2LoT (or even mentioning 2LoT). That would after all be a bar to getting a patent at all, unless you supply a real device that can be proved to actually do the job. Manufacturing the device using flat layers is not however that difficult, and would take maybe a day if you already have the deposition kit and materials sitting around (as these people obviously would have, since I would assume Banpil Photonics do actually manufacture photonics devices), but making the masks and setting up the manufacture of the structures as patented would be a much longer job.
A quote from the patent, just as a teaser:
“What is claimed is:
1. A perpetual thermal energy harvester, comprising: a substrate; a buffer layer attached to the substrate, wherein the buffer layer comprises a dissimilar material system from the substrate; a first electrode; a second electrode; a absorption layer electrically connected between the first and second electrodes comprising a first material; a second material of a different type than the first material; a third material of a different type than the second material, wherein the first, the second, and the third materials selected from the group consisting of HgCdTe, HgZnTe and combination thereof material systems, and wherein the first, the second, and the third materials in the absorption layer comprises two or more p-n or p-i-n junctions comprising three-dimensional structures, wherein the two or more p-n or p-i-n junctions are in series increasing the open circuit voltage, and wherein the two or more p-n or p-i-n junctions comprising cutoff wavelengths between 2 .mu.m to 40 .mu.m; and a insulator layer attached to the absorption layer on the opposite surface from the substrate. ”
Despite the manufacturing problems that their design entails, I find it interesting that the US Army thinks that they can get around the 2LoT by using photovoltaic devices. This is probably the easiest way to get around 2LoT, since it does not actually apply to photons even though in the textbooks it is insisted that it does. I’ve pointed out the logical paradoxes this insistence entails in various previous articles.
Thanks to Lou Pagnucco on LENR Forum for noticing this and pointing to the patent. As he points out, this is the reason why we couldn’t dismiss Steorn’s energy harvester as a fraud until it had gone through the testing process. I also know from other experiments that the structure Steorn were said to be using can harvest heat energy if it’s done correctly, though the power is pretty low at room temperature and really needs to be over 50°C in order to deliver a more-reasonable amount of power (though Steorn’s claims were still at least an order of magnitude too high). Much the same for the ADGEX torch, which is why such claims need to be carefully tested rather than either accepted or dismissed without such tests. Fairly soon we’ll be finding out if my structures perform as well as I expect.