IT COULD be the world鈥檚 smallest fence: a chain of five crude, squarish links is strung between posts just 40 micrometres apart. This microscopic barrier marks the threshold of a new world of fabrication.
The interlocking links of the chain would be impossible to make using conventional microfabrication techniques. Instead, they were conjured into existence by waving a magic wand 鈥� or rather, a laser beam. This method, called two-photon 3D microfabrication, is just starting to make its mark. Fast, cheap and flexible, it can create multilayered microchips, custom-designed crystals, pre-assembled propellers 鈥� tiny objects with fantastical curves, interlocking parts and 3D structures, all qualities difficult or impossible to achieve with traditional techniques. 鈥淚t鈥檚 like making Frank Gehry-type buildings instead of flat warehouses,鈥� says Malvin Teich, an engineer at Boston University.
The 鈥渇lat warehouses鈥� he means are the silicon computer chips made by photolithography, today鈥檚 standard in microfabrication. Photolithography etches the circuit onto the silicon by shining ultraviolet light through a flat mask perforated with the pattern of the chip circuitry. The technique can produce multilayer constructions, by depositing successive layers of silicon and using different masks to etch each layer, slowly building up a 3D structure. But you could never build a curvy Bilbao Guggenheim in this way.
Advertisement
Two-photon microfabrication is a different story. 鈥淵ou could do this on top of a desk, if you had the right lasers,鈥� says John Fourkas, Teich鈥檚 collaborator at the University of Maryland in College Park. In fact, Fourkas is pretty much describing his lab, which is essentially just a big table with a laser hooked up to a computer and a few lenses. The laser scans through a drop of liquid resin on a slide, hardening the resin in a 3D pattern so that when the excess liquid is washed off it leaves behind a microscopic plastic structure.
The resin works a bit like the material in white dental fillings. When you get a filling, the dentist shines an ultraviolet light onto the mixture to fix the substance into a hard, acrylate polymer. In two-photon 3D microfabrication, the laser does essentially the same thing: energy from the beam excites molecules of a special dye, each of which then offloads one of its electrons onto a nearby acrylate molecule. That triggers a chain reaction in which the acrylate molecule grabs another acrylate molecule and passes on the electron, the second acrylate grabs a third, and so on, linking together a soup of small molecules to make a sturdy chain of plastic.
But there is a twist: a single photon from the laser doesn鈥檛 have enough energy to excite a dye molecule, but two photons have exactly the right amount. Only at the focus of the laser are the photons dense enough to allow the dye to absorb two at a time, triggering polymerisation. So the resin hardens only at the laser鈥檚 focus, while the rest of the beam passes through with no effect.
This is what makes two-photon microfabrication so powerful. The beam鈥檚 focus is fine enough to create details just a few tens of nanometres across, smaller than a single wavelength. The laser can focus at any depth within the transparent drop of resin, and it can scan smoothly or skip, to generate microscopic structures with separate components, including such finely wrought 3D creations as the microscopic chain. The fence was created in 2003 by Joe Perry, a chemist working with two-photon microfabrication, and Kevin Braun, then Perry鈥檚 student at the University of Arizona in Tucson, to demonstrate the potential of the technique.
As well as reaching into the third dimension, this process has two more advantages over photolithography: it is faster and cheaper. The elaborate masks required for photolithography cost tens of thousands of dollars each and require several weeks to make. In comparison, two-photon microfabrication is ridiculously cheap: you just fit an off-the-shelf laser with the appropriate software, and then the only costs are for raw materials.
鈥淧eople know how to make things on the microscale, but not arbitrary things,鈥� says Fourkas. 鈥淭his is the first technique that can operate on that microscale and you can make anything you want.鈥� So if you could wave a laser wand and make anything at all, what would you choose?
鈥淭hree-dimensional optical information storage,鈥� says Teich. A cube could store information much more compactly than flat CDs and DVDs, for example, perhaps holding thousands of movies instead of one or two.
鈥淢icro-MRI probes,鈥� says Fourkas. He and his collaborators think that instead of putting the patient inside an MRI machine, doctors could put the MRI inside the patient. You could thread tiny MRI coils through a diseased heart, for example, to give a far more detailed picture than external MRI scans.
Digital light processors, scaffolding for artificial organs, microfluidic 鈥渓ab on a chip鈥� devices鈥� all these come up on the wish lists of researchers who have seen laser sculpture in action. The polymer resin should be able to meet the physical demands of each, because the recipe is quite flexible, as long as it contains the light-sensitive dye. Metal parts or semiconductors could be plated on afterwards to make electrical circuits, and perhaps eventually whole 3D chips. Adding a third dimension to computer chips could tremendously increase their processing power.
At Focal Point Microsystems in Atlanta, Georgia, Perry hopes to turn promise into commerce. The company, founded by Perry and other researchers at the Georgia Institute of Technology, is developing a ready-made laser-machining device to sell to researchers. The next step is to take the technique beyond research. At the moment, although two-photon microfabrication is fast and cheap compared with photolithography when making one-off designs, it loses out in mass production. Once the photolithographic mask is made, the printing process can turn out thousands of components in the time it takes a laser to build up a single structure. The speed needs to increase.
鈥淚t鈥檚 like making Frank Gehry buildings instead of flat warehouses鈥�
One idea is to use moulds. You laser-sculpt the prototype and then form a mould around it. A single prototype can form many moulds, which can each be used again and again. You might think that micron-sized objects would be too delicate to be torn out of a reusable mould, but if the material of the mould is soft enough they won鈥檛 break. Fourkas and colleagues have been experimenting with polymethylsiloxane, more familiar as bathtub caulk, which is so rubbery that it bends around just about anything without damaging it. The group has been able to cast details as delicate as micron-sized needles, as well as overhangs and cavities.
Commercial applications may be only a few years away. And after the architectural revolution of microfabrication, 2D microdevices will probably seem as primitive as a flat-roofed shack.
