But the design requires that this silicon layer be no more than five nanometres (billionths of a metre) deep.
The problem, however, is that the silicon layer is amorphous, rather than crystalline.
The one that actually builds silicon layer infrastructure, new chips, networking, storage.
What's left is machine-ready silicon atop a layer of insulator.
That leaves the thin layer of silicon above it intact and insulated.
But it is possible to lay down a thin layer of silicon on a flexible plastic sheet, through a process called vacuum deposition.
As the pressure builds up in these bubbles, they expand and rip the thin layer of silicon above them away from the rest of the wafer.
It's a layer of carbon nanotubes on silicon, and it might just be instrumental to getting a lot more power out of solar cells than we're used to.
The other established approach is to take two wafers of silicon, coat each with a thin layer of glass on one side (which is done by baking them in oxygen), and then fuse them together to produce a silicon-glass-silicon sandwich.
More layers of silicon dioxide and other materials are grown on the etched layer.
The system the researchers tested used a type of solar cell known as dye-sensitized solar cells, a lightweight and inexpensive type where the active layer is composed of titanium dioxide, rather than the silicon used in conventional solar cells.
One side of the silicon can then be etched down to within a fraction of a hair's width of the glass layer, leaving the other side as a support.
Gallium-nitride is manmade and not found in nature, and historically, LEDs are made by growing a gallium-nitride layer on top of a base material, which is typically sapphire or, more recently, silicon carbide.
应用推荐