A new way to cut ultra thin solar cells


Cutting down on the silicon used in solar panels has been a hotly pursued goal for many solar startups. And now a new player is entering the ring: meet Twin Creeks Technologies, which is launching a machine that will dramatically cut the use of silicon in solar cells.

The machine, called Hyperion, creates silicon wafers at 20 microns thick, compared with the typical 200 microns, said Siva Sivaram, CEO of Twin Creeks. Being able to use the same amount of silicon to make more cells means a cut in capital equipment cost, which is measured in cents or dollars per watt. Putting Hyperion to work could reduce the capital equipment expenditure by 50 percent for a vertically integrated company, which makes everything from silicon to solar panels, Sivaram said. Using Hyperion in a large factory — that  means 100 MW of annual production capacity or more — will lead to a cell production cost of 40 cents per watt, he said.

Wafers are flat, processed slices that come from a block of silicon and serve as the base material for solar cells. The solar panels that you see on the rooftop today are made up of a series of solar cells lying side by side.

Twin Creeks is one of many startups and established solar companies looking at ways to use less silicon to make solar cells. While the price of solar cells has fallen quickly in recent years thanks to an oversupply problem, they need to fall even further to make solar more cost-competitive against power generated by coal or natural gas. The U.S. Department of Energy has been funding next-generation technologies in order to achieve its goal of bringing down the cost of building a solar power project to $1 per watt – without government subsidies – by 2020. Right now, large-scale projects built for utilities still cost a few times that, and smaller systems such as those for residential rooftops cost several times more.

Other startups are working on making thinner silicon wafers or cheaper wafers in conventional thickness and these firms include 1366 Technologies and Astrowatt. Twin Creeks and Astrowatt both are working on creating thinner slices of monocrystalline silicon – which has a crystal structure that makes it possible to go thinner without compromising its performance. 1366, on the other hand, aims to make polycrystalline silicon wafers much more cheaply by eliminating several steps (polycrystalline silicon wafers can’t go as thin or else they’ll break).

Monocrystalline silicon wafers are more expensive than the polycrystalline variety but they make solar cells that can convert a greater amount of sunlight into electricity. SunPower makes the most efficient monocrystalline silicon cells today.

How does it work?

Twin Creeks takes a thick block of silicon and put it in Hyperion, which bombards the silicon block with hydrogen ions down to the depth of 20 microns. The ions in effect create a bubble layer, and when the wafer is moved to a furnace and heated, the bubble expands and separates the 20-micron top layer from the rest of the silicon block. The remaining silicon block is then used again and again. “The only way we continue to get more value is for those materials to become more productive,” Sivaram said.

Hyperion is Twin Creeks’ first offering since it decided to change its business plan from making solar cells and assembling them into panels to selling the equipment for making silicon wafers. The San Jose startup, founded in 2008, has raised $93 million from investors including Crosslink Capital, Benchmark Capital, Artis Capital Management and DAG Ventures.

Twin Creeks previously wanted to build a cell and panel factory in Mississippi and secured a $50 million loan commitment from the state. The company has since drawn $30 million in loans from the state and built a factory that now serves a lab for Twin Creeks’ customers to try out the company’s equipment and figure out how it will fit into their own factories. The Mississippi plant is able to produce 25 MW of cells per year, and the startup plans to expand that to 100 MW.

The company broke ground on a 100 MW cell and panel factory in Malaysia in 2010, but has since put that plan on hold. The plan is to complete that factory with a customer, Sivaram said.

Photos courtesy of Twin Creeks Technologies


tom delconte

The five disadvantages of solar power

It’s intermittent. Solar energy is only available when the sun is shining.

It’s low efficiency. It requires large areas of land.

It’s not completely free of environmental impact. Ultrapure polysilicon is now scarce.

It’s high cost.

You must climb your roof to wash it off- just ask Ed Begley, Jr.!


Thanks tom – working for the Koch brothers by any chance?

To answer your post:
Intermittent – energy storage batteries/thermal storage – search Gemasolar

Low efficiency – True (1/2 coal steam turbine), but also requires no fuel. But land area? – how does that compare with the area needed for strip-mining or tar sands, or is that somehow irrelevant?

Not free of environmental impact – nothing is (including me typing this), the point is its relative impact. Polysilicon is scarce? And yet prices keep falling…

High cost – solar systems are actually at price parity in parts of the developing world and remote regions of Australia, further, cell costs have fallen substantially in just the past 5 years dropping below $1/Watt

Cleaning – gasp cleaning the cells twice a year, sounds worse than cleaning your gutters! Otherwise suffer a 5-10% power loss.

Ucilia Wang

You’ve gotten some info wrong: pure polysilicon isn’t scarce and solar doesn’t always need a lot of land. Anyway, I don’t think there is any power generating technology that doesn’t have advantages and disadvantages. And that’s why some people think it’s a good idea provide solar manufacturing services — to help address a few of the issues you mentioned.


Have you ever looked at an intraday electricity demand chart? The curve matches up very nicely with intraday solar power production. Storage doesn’t become an issue until penetration is >40%.

Comments are closed.