Can lasers or some type of energy beam print matter acuritly ,controled by a computer, at the 10nm level.?

Answer 1

No, an energy beam can’t print matter, you would need to use a particle beam. Something like ion implantation for semiconductor doping might work, but that has to be done in vacuum. Try an ink jet printer, and modify it by making the beam path much longer so that you could steer off most of it and shield portions of a droplet. You would have to build electrodes to accelerate the droplets to travel much further, too. You might be better off by using lasers to REMOVE the matter you didn't want, sort of like sculpting away the excess material to make what you want. Don’t count on being able to work in more than one kind of material at a time, though.

24 DEC 06, 0411 hrs, GMT.

Answer 2

Yes, there are all sorts of interesting ways to “print” matter at the ten nanometer level, some of them practical for small quantities but impractical or impossible right now for large quantities.

The usual way borrows from photolithography, the mechanism routinely used to print newspapers, magazines, and any other printed matter in copies greater than about a hundred or so. Photolithography involves creating a mask through which the matter is deposited. Where you don’t want to deposit matter, the mask absorbs and traps it. Then you remove the mask, along with the trapped matter. What’s left behind is the matter you “printed.” This is basically the procedure used to produce integrated circuits, repeated several times as each new material is deposited through the apertures in each new mask.

There is currently no limit to the size of the features that can theoretically be produced with photolithography. The photoresist material (that’s what the mask is made of) consists of photosensitive polymers whose molecular chains can be “cross linked” by exposure to light of sufficient energy. The exposed molecules behave different, chemically, than the unexposed molecules. Depending on whether a positive photoresist or a negative photoresist is used, either the exposed or the unexposed molecules will preferentially dissolve in a specific solvent, leaving the other molecules behind. Even large polymer molecules are pretty small critters, way smaller than a nanometer, which is why there is currently no limit to how small you can theoretically make the features. The problem comes when it is time to actually expose the photoresist coating to light to make the desired feature pattern.

The size of features, in this case 10 nanometers, makes a big difference in what is not only practical but possible. Without going to some very special optical techniques, it is generally not possible to optically define features with wavelengths of light longer than about half the feature size. That means 10 nm features generally require 5 nm wavelengths or less to define them. Such wavelengths are in the far ultraviolet, approaching x-rays, which makes them difficult to focus optically and even more difficult to generate efficiently with a laser.

The only other candidates are an electron beam or an ion beam. It is relatively “easy” to focus an electron beam down to a very small spot, and only slightly more difficult to focus ion beams down to very small spots. Actually, because electrons do not occupy space (they behave as “point” objects) there is no theoretical limit to how small a spot you can focus an electron beam down to. There may be some quantum mechanical limits because of the Uncertainty Principle, and there may be some practical limits because of the engineering required, but electrons are routinely focused to spots way smaller than one nanometer.

Focusing electrons or ions is not the problem. Creating a useful photomask is THE big problem. Photomasks start with drawings that can be easily inspected with the naked eye. Lens are then used to reduce the size of the image, usually in several sequential steps, to make a photomask. That is, the first lens may reduce the original image by a factor of 100 to make an intermediate mask. Then this mask is imaged to reduce it in size by another factor of 100. And so on until the final image size is reached. However, I am not aware of ANY production optical system that can make a photomask with one nanometer features today.

An alternative to optically reducing and exposing an image is to scan the image and reproduce it at a smaller size by scanning and modulating a small spot on and off, the same way a television works. The small spot can be light, electrons, or ions if it can be focused small enough. The disadvantage to this approach is the time it takes to perform the scan, which can run into days or weeks depending on the actual area scanned and the resolution desired. Still, it’s better than nothing.

For research purposes (just to try an idea out) you can scan and modulate ion beams to deposit matter where you want it in the quantities you want. This means it is very easy to deposit material with ten nanometer features using ions. It just takes almost forever to build anything of practical dimensions.

Laser beams are routinely used to move atoms and molecules around, but again in very small numbers. So using a laser to directly deposit matter is impractical for now, even if the laser could be focused to a small enough spot to create ten nanometer features.

There are other possibilities, such as contact laser ablation which vaporizes a material in contact with another material, driving the vaporized material into the surface of the target material. This is sort of like the thermal printers popular with calculators and some cash registers that make a paper tape copy of the calculations or transactions, except the spots are much smaller and the energy (heat) from the laser much more intense.

But always, there is the practical problem of how to make zillions of copies for the world market. One copy we do right here in the lab. For two or more, please consult the factory (as soon as it exists). BTW, a conventional wafer fab capable of producing fifty nanometer features on large wafers costs more than a billion dollars today.