Machine learning and nano-3D printing produce nano-architected materials | Hacker News Reader

Machine learning and nano-3D printing produce nano-architected materials

news.engineering.utoronto.ca

58 pointsgmays6 comments

The material is very cool, but I'm almost more interested in the progress in 3D nanofabrication this shows - "printing" material in 3D by selective solidification tends to be pretty slow, as you need to move around like a 3D printer head instead of one-shotting an entire layer like in 2D lithography.

The fact that they have made a millimeter scale version of their design instead of a small demo that looks cool under an electron microscope is very impressive! The method is pretty clever too - 2 photon absorption works only where the light is focused (and therefore high intensity). They use a lens array to create 49 focus points, allowing them to parallelize the printing of the repeating grid!

Scalable 3D fab could have fun implications for materials like this, and for chip-scale photonics.

"Nano-architected materials are made of tiny building blocks or repeating units measuring a few hundred nanometres in size — it would take more than 100 of them patterned in a row to reach the thickness of a human hair."

I am perpetually shocked that a structure (hair) that we can actually see is so close to nanoscale.

(It's also noteworthy that our sense of touch in fact extends into the nanoscale: https://www.nature.com/articles/srep02617)

You can use a light microscope to visualize things down to about 500 nm, somewhat dependent on shape (you can see down to about 100 nm, but really only that something is there - whereas nanocubes with 200nm side length had some definition when I tried it).

Which is pretty shocking when you realize a nanometer will hold about 10 hydrogen atoms (give or take the fact that the electron cloud repulsion isn't well defined), meaning you're in fact only talking about 5000 hydrogen atoms at the same packing for that 500 nm linear distance.

Which puts the idea of "Plenty of space at the bottom" by Feynmann in an odd contrast because it's like...well, that doesn't seem so far away when my basic eyeball and some optics can operate plausibly see down to that scale directly - photons are interacting with things and passing straight to it.

I was thinking recently how there used to be about a lot of talk about waldos and nanotechnology. I think the idea was to build a Waldo that can assemble a 1/10th scale Waldo, repeat and voila you are soon building nano scale contraptions. It’s a very intriguing idea, seemingly not practical though.

Its operation at each scale would be very different, and it's not obvious how many different layers we'd need before we'd be at the smallest possible, but maybe.

Technically we already can go from human scale to nano or even atomic scale in "one operation". i.e. a scanning tunneling microscope you etch a tungsten wire in acid and then can go and look at individual carbon atoms at room temperature.

An atomic force microscope is even more direct - while not absolutely "atomic" - you can definitely with a sharp enough tip push things around on the nanoscale.

The problem is definitely as you note: things do in fact behave very differently when you're doing that.

Reply on news.ycombinator.com