Amazing 50 foot crystals found

[Simon F]

Crystals tend to have hard edges, because crystalline structures can only have edges at particular angles. Thus icicles are not likely to be singular crystals. Ice crystals tend to be microscopic.

Have you ever seen a natural diamond? They can sometimes just look like a blob of, say, glass. I wouldn't say it's a necessary condition that crystals have well defined edges (though, obviously, they look nicer)

 

[Rainbow Man]

The silicon isn't cut into a cylinder from a hard edge; it's grown as a cylinder.

So it IS manufactured to be round.. Why don't they make em square instead? ICs generally have four corners after all. Surely that would be more efficient?



Peace.

 

[Basic]

The round shape comes automatically from the manufacturing process, it would be a lot harder to get any other shape. The easiest would probably be to saw off the edges of a round wafer, which makes it kind of pointless.

 

[nutball]

So it IS manufactured to be round.. Why don't they make em square instead? ICs generally have four corners after all. Surely that would be more efficient?

If you read the link it should be reasonably obvious why the resulting ingot is cylindrical rather than a cuboid.

 

[Mize]

So it IS manufactured to be round.. Why don't they make em square instead? ICs generally have four corners after all. Surely that would be more efficient?



Peace.

It is grown round. This is the natural macroscopic shap of the growing process described in the link above (even though silicon has an FCC - face-centered-cubic - microstructure). To grow a square or rectangular crystal would likely require spatially-temperature-controlled mold. like those used in growing turbine blades, and a real-time x-ray system to detect a transition to polycrystallinity (i.e. bad part). Also, it would be extremely difficult to grow it with a specific lattice orientation in a mold.

Suffice it to say they've spent billions finding cost-effective ways to grow silicon already.

 

[bloodbob]

Really?

Thats a nice picture of amorphous ice you have there simonF. Well actually seeing as it is outdoor its probably polycrystaline just with defects belong the diffraction wavelength. If you read up how they make ice for clear ice scultpures they actively try to stop the formation of crystals because well it causes light to scatter and the ice to become opaque so they rapidly cool it and constantly vibrate the center which of course the unfrozen liquid is so that it can't crystalise.

 

[Simon F]

Thats a nice picture of amorphous ice you have there simonF. Well actually seeing as it is outdoor its probably polycrystaline just with defects belong the diffraction wavelength.

I'll take your word for it. So (quickly reading up on it), are you saying that would be due to it being frozen too rapidly or because there would be too many impurities in the water?

 

[bloodbob]

I'll take your word for it. So (quickly reading up on it), are you saying that would be due to it being frozen too rapidly or because there would be too many impurities in the water?

A little from A and a little from B. You can see the bubbles in the center of the thing.

 

[AlNom]

I seem to remember that they're also picky about the orientation of the lattice structure, such that the cut wafers expose a specific lattice edge.

Yup, they are indeed picky, particularly with respect to doping and etching.

Just for reference, here's a view of the silicon lattice network (with some other-coloured impurities):

Depending on the orientation, you can actually get a view where you see uniform hexagonal lattices. For doping, it's a bit of a problem because most dopants would just fly through!

In the case of etching, the etchants can follow along the angle of the lattice rather than universally eating away vertically/anisotropically (as the etchants are introduced as a top-layer bath). All for different structures for different properties or insulation layers and such.... bunch of things.