Hacker News new | ask | show | jobs
by mchannon 3216 days ago
As an expert on silicon and solar, the technology is doomed- here's why:

•It is totally possible, in terms of the physics, to do what they're proposing. All you have to do is take your ingot or boule of FZ or Cz Silicon, draw a vacuum on it, backfill it with hydrogen, and run your ion implantation head over it. Like a Durandal bomb, the upper surface is relatively undisturbed but a cutting effect occurs beneath the surface. Pop, off comes your perfect and thin wafer.

•This idea has been proposed before.

•There's nothing fundamentally wrong with a 3 micron thick silicon wafer. It takes some different processing steps to trap the light, but you can still make a decent solar cell out of it. It's not better than a 400 micron wafer in terms of efficiency, but not much worse. Also, by it being so thin, it's flexible and not so darn brittle. That reduces breakage but also makes handling a little more challenging.

•The idea fails when you consider the economics. First consider the market you're trying to break into- there's a huge glut of silicon wafer manufacturing capacity in the industry. Silicon manufacture is going gang busters, so it's not a commodity that's expensive to begin with. Solyndra bet on silicon prices staying high and that, more than any other reason, is why they failed. This is the same bet with a different topology.

•Bill Nye should perform a little math problem- consider the amount of electricity and hydrogen gas necessary to perform a cleave, 100% efficient, 100% yield, and multiply these quantities by their respective commodity costs. You will end up with a number in excess of the cost of a 500 micron wafer, or perilously close to one.

•I did the math five years ago and it made no sense then, when silicon wafers sold for over twice as much. Just because the existing technology is materially wasteful (it definitely is, generating a lot of kerf and going through a lot of expensive wiresaw blades and polishing processes) doesn't mean that it's economically wasteful.

It's not too late to pivot- I'm sure he'll have no difficulty raising the money in spite of these shortcomings. I'd be more than happy to steer this company in the right direction, contributing IP that actually delivers on these promises. All it takes is an e-mail.

5 comments

I agree that this is going to be a commercial failure even if funded. Additional stray observations:

- The "magic," if anything, is supposed to be that super-thin kerfless wafers could justify the industrial scale use of float zone silicon. FZ has often been the material of choice for lab scale fabrication of champion devices. Unlike ordinary boron doped p-type Cz silicon, it doesn't have dissolved oxygen traces from crucible walls, so it doesn't suffer B-O complex light induced degradation.

- High efficiency is still worth a premium, even if low-mid range quality cells are stuck in a brutal commoditized competition. (But high efficiency is not worth a large premium, hence the problems of even well-established high efficiency manufacturers like SunPower and Panasonic.)

- Super thin wafers are less sensitive to bulk recombination. Though this theoretical advantage is unneeded if you're starting with superpure FZ silicon in the first place.

- FZ silicon is only available in smaller diameters, so the cells wouldn't be drop-in replacements in the usual 60/72 cell module that starts with 156 mm cells.

- Silicon's resistance to damage by cosmic rays goes up dramatically as the cell thickness drops down to the few-tens-of-microns range. But that's not relevant for terrestrial use and compound semiconductors are still more efficient and more damage-resistant for space use.

Other companies that proposed to do kerfless wafering with ion implantation and failed to reach commercial success:

Twin Creeks Technologies

GT Advanced Technologies (later owners of Twin Creeks assets)

SiGen

1366 Technologies appears to have the closest-to-industrial kerfless process at present, though it's not as dramatic as as some kerfless technology concepts. I hope that their partnership with Hanwha Q Cells goes to full scale production before the money runs out.

1366 Technologies's full-scale plant in Western New York has yet to break ground. The state finally broke ground two days ago on some minor infrastructure work for the site, but 1366 might be held up by financing issues due to worries that Trump won't follow through on a DOE $150M loan guarantee.
I think their interest is clear enough, and not long term...

“After each closing, funds tendered by investors will be available to the company and, after the company has sold $7,000,000 worth of Common Stock, selling securityholders will be permitted to sell up to $3,000,000 worth of Common Stock.

> Like a Durandal bomb

Sorry, can you explain this metaphor?

> Just because the existing technology is materially wasteful (it definitely is, generating a lot of kerf and going through a lot of expensive wiresaw blades and polishing processes) doesn't mean that it's economically wasteful.

Always wondered- why not EDM? Too slow? Hydrogen injection? Tungsten/copper contamination?

One more:

* Silicon is not a rare material. It's one of the more common elements in Earth's crust. The price is largely in processing and fabrication, meaning it's going to continue to fall like a rock due to economies of scale and optimization.

Silicon of the requisite purity is not just lying around. Purifying it takes a significant amount of energy, which puts a floor under the price.
Energy prices have fallen continually through human history. As our demand for energy increases we find better and better ways of harnessing available energy. This wafer technology itself could lead to reducing energy costs to a point where silicon refinement becomes economical for the production of further wafers.

The availability of Silicon is not a limitation energy prices are.

energy for industrial use costs more today than it did in 1960: https://www.eia.gov/totalenergy/data/annual/showtext.php?t=p...