Yes you have it right, that's the word "HAIR" written on top of an actual human hair. Crazy!!
How did they do it??
The substrate, in this case the hair, is covered in a acrylic-based liquid resin. The resin has the property that when it absorbs light of a certain frequency it hardens into a solid polymer. But there is a threshold, the resin hardens only if the intensity of light is strong enough to start the polymerizing chain reaction.
Now we point a laser at this setup. Add a lens to focus the laser into a tiny (few microns thick) spot. Now we can set the laser power low enough that only the intensity of the spot is high enough to harden the resin but not the rest of the beam. Just like how sunlight is not strong enough to burn paper, except at the tiny spot at the focus of the lens.
The laser is moved with computer controlled precision to be able to make any structure we want. Finally we just wash the substrate and all the liquid resin is removed and the hard structure is left behind. That was almost easy, right??
But there is a problem.
In 2D, this method works pretty well. Using this, we can make a very detailed pictures on a flat canvas. In fact this is the principle of writing information on CDs. But for patterns/structures in 3D, we can only have fine details up to 1 mm. Smaller than 1 mm, however the details get fuzzy because the spot is not small enough in the direction along the laser light.
So we have to add another trick, a simple quantum trick.
Now we choose a different laser, one whose frequency is half that of the one required to harden the resin. Which means that the process which required 1 photon, now needs 2. Two photon absorption needs much higher intensity to occur than single photon absorption. This means that only a really small, really dense part of the spot has enough intensity to harden the resin. That is what brings the resolution down to tens of nano meters in 3D. You can read more about this here.
|Amazing level of detail... possible using just a computer, a laser and a chemical, in a tabletop optical setup|
Although creating random nano-structures is cool by itself, it has the potential to be more. The first obvious idea is cubic CDs. 2D nano-structures can hold all the data that we write in CDs today, imagine if we had a cubic CD with information encoded in 3 dimensions... 3DCDs that could hold thousands of times more information than 2D storage media. The technology to do this is so simple that you could be burning cubical or spherical CD's on your computer, sooner than you think.
On that train of thought, why not 3D chips and microprocessors? Computers are getting more powerful and smaller, yet naysayers emerge everyday predicting the end of Moore's Law. There is only so much computing power you can cram into a 1in x 1in microprocessor chip, right??
Unless... you start building up??