Chuck Black has been around long enough to know microprocessor fabrication is a big business for IBM. Now he wants to see it happen on a smaller scale.
As one of Big Blue’s research staff, Black and his colleague Katheryn Guarini
Monday said they had discovered a nanotechnology self-assembly technique whereby certain types of polymer molecules could organize themselves and help improve the chipmaking process. Nanotechnology uses tiny materials measured in nanometres, which are approximately about the size of four water molecules placed side by side, or about one one-thousandth the length of a single-cell bacterium.
Black said he realizes the discovery is not going to change chip manufacturing overnight.
“”It’s going to be suddenly that we all stop making chips one way,”” he said. “”We’re always introducing new things and there’s always a bit of pain getting them through, but the pain is understood and that’s how we’re going to get to better performance.””
Black, who was due to present the research findings at the IEEE International Electron Devices Meeting (IEDM) in Washington, D.C., took some time to explain the process to ITBusiness.ca.
ITBusiness.ca: Why did this particular process seem to apply well to processor fabrication?
Chuck Black: The thing that’s really caught our particular process which we’re using is that it relies on polymer molecules for self-assembly. Polymers are used all the time in chipmaking. They often go by a different name — people call them resists or photo-resists — and they’re the materials which are used to template and pattern all of the circuit elements in microelectronics. So how those particular polymers work is light is shined on them, and that’s what allows them to define the transistors, the wires and all those things. That’s all known as a process known as lithography.
The key for us is these polymers, which undergo self-assembly, have a special name. They’re called die-block co-polymers. They make these beautiful patterns, and in terms of microelectronics they would be used all the time. The industry in some ways is very conservative because trying to make new materials in chipmaking is very difficult. You don’t want to wreck the transistors. But here we thought they’re polymers, so they’re eminently acceptable, and yet they have this special property to form patterns on these extremely small dimensions. That’s what we’re excited about, that we’ve found a nanotechnology process and combine it with all fantastic tools and processes of semiconductor manufacturing and being able to mesh them together.
ITB: What are you making them form exactly?
CB: They’re very small — they look like dots with extremely uniform size and spacing from each other. In this case they’re 20 nm large and they’re spaced 40 nm apart. This is a perfectly beautiful, hexagonal pattern. This turns out to be quite attractive for building a nanocrystal memory device. But we have other ideas for other things to build as well.
ITB: Is there a way you would see IBM preparing their workflow and process to take advantage of these techniques for when they become commercialized?
CB: Putting polymers onto wafers is something we do all the time. You don’t have to buy any new tools than they have now. All of the materials are allowable in the lab. In contrast to other possible nanotechnology applications, this one could very easily find its way in, because there’s no new infrastructure required, and the materials are actually commercially available, and have been for a long time. They have very grungy industrial uses (laughs), and no one’s applied these to microelectronics before, but they’re used in all kinds of things.
ITB: If these techniques were used, would we see a faster time-to-market for the processors, or better performance of the chip?
CB: The main place where this is helpful is that it provides a straightforward and easy way to access extremely small dimensions. It does it in a uniform way because the dimensions are defined, in some sense, by nature and the size of the molecules. We don’t ever imagine you being able to use this technique to build all elements of a chip. At least, I don’t see a route to that happening today, but you appreciate the enormous complexity in building those things. Our idea is that in the process of the hundreds of steps in making one of those — if there’s a few steps where you need high-resolution dots if you can use something that gives it to you naturally, that’s an enormous savings in complexity. That could translate into a savings in time, as you say, or a savings in cost or those types of things.
ITB: As we start to apply nanomaterials into the chipmaking process, how do you see that affecting the skill sets needed for those intimately involved in chip production?
CB: I can give you one interesting example from our lab here. On a smaller scale, to produce the devices we’ll be presenting, we introduced this self-assembly process into our smaller scale research lab here. It’s a decent lab — it doesn’t make products but it has all the tools — we were proud of ourselves to take this process and introduce it to those people. It’s now in our lithography sector, and I think that’s the perfect place for it. You need to take a polymer, coat it onto a wafer, put it in an oven and put it into a chemical developer. I have a great chart where I have lithography on the left side and self-assembly on the other, and they’re identical. That is the real thing that gets my attention about self-assembly: you put it in an oven, you go away, the polymer molecules kind of find each other and in the end they give you something that you want.