Interview: Dr. Kathryn W. Guarini
2004 Peter Mark Award Recipient
MARRIAN: Good afternoon, I am Christie Marrian, the President-Elect of AVS. As part of the Society's Historical Archive Series today I will be talking with Dr. Kathryn Guarini, who is this year's Peter Mark Award winner. It is November 17, 2004 and we are at the 51st AVS Symposium in Anaheim, California. Kathryn, it is a real pleasure to have you with us today. Congratulations on the Peter Mark Award; fantastic! I would like to talk to you about the work behind the award and your current path with IBM. But first of all could you briefly describe your background and introduce yourself?
GUARINI: Sure. I am Kathryn Guarini, and I am currently employed at IBM. I am at the IBM T.J. Watson Research Center as a research staff member and manager in the Silicon Technology Research area. I have been at IBM for a little over five years and before I joined IBM, I got my Ph.D. at Stanford in Applied Physics.
MARRIAN: How did your earlier work on scanning probe lithography lead to and motivate your research on self assembly?
GUARINI: My thesis research at Stanford was done under the direction of Dr. Cal Quate. Cal is well known as being co-inventor of the atomic force microscope. At the time in the mid 90's there was a lot of interest in the whole semiconductor industry in finding a next generation of lithography solutions, the so called "next generation." There was the perception that optical lithography was reaching fundamental limits and we needed to find a new solution to enable us to pattern at dimensions beyond the resolution and capabilities of the current systems. At the time Cal was looking at using scanning probes themselves to do lithography to pattern nano-features. So my work on that was to use a scanning probe with a very sharp tip in close proximity to a sample and field emit electrons and use that to expose features. I think we made good progress in showing resolution capabilities and some feasibility of that process, but I think an overriding concern of that technique was the challenge of throughput. The scanning probe system, like the electron beam system, is serial patterning and it is slow, notoriously slow. While we typically measure throughput for conventional lithography in number of wafers per hour, we routinely measure throughput for scanning probe lithography in number of pixels per day. So you have a new challenge. In Cal's group, we ran many, many probes in parallel to increase the throughput. We had some success with that but it is still a very, very daunting challenge. And when I learned more about materials that self organize into nanometer-size structures so the same dimensions that we were patterning with the scanning probe could happen naturally using natural forces, and you could access these dimensions in a highly parallel way, I thought that that was a very promising avenue of research and that was one of the things that excited me when I learned more about the material.
MARRIAN: We heard a lot about self-assembly at this conference and conferences in the past, but very little about actual devices where self-assembly is utilized. Could you describe this part of your research and how this project came about?
GUARINI: Sure, it is actually that significant interest of mine to explore the applications themselves beyond the fundamentals of materials with self assembling properties. I think that great fundamental work has been done in the past at Universities and some industrial labs looking at how do materials self organize into regular arrays of small features. But as you say, we have not seen much evidence of really applying that to making things that we really care about. So that was the approach that we took in a project that we started about five years ago in collaboration with Dr. Chuck Black at IBM. He was in the physical sciences department, his department is more exploratory research at IBM, and I am in the Silicon Technology, more applied devices. It was a great collaboration. We were looking at taking some of the fundamental basic physics and chemistry and material science and applying those to real life problems. So I can give you a couple of examples of ones that we focused on the last few years. One that is good to look at is the decoupling capacitor. Decoupling capacitors are critical for integrated circuits today. They are used to stabilize the voltage supply. The problem is they take up a lot of space on a chip. So what we did was use a self assembling polymer material to form a nano-structured integrated circuit electrode and built a metal oxide semiconductor thin film capacitor that behaved just like a conventional capacitor used as a decoupling capacitor, but able to have much higher capacitive density, effectively taking up much less space on a chip. It is a very simple type of device, but it shows the utility of these new materials for solving existing problems, rather than really completely revolutionizing the technology or the way that we build devices. It's a way of enhancing the functionality or performance of the device.
A second example is one that we more recently showed looking at nanocrystal flash memory devices. Flash memories are pervasive today. They are in our digital cameras, handhelds, and so forth, and we know that there is a push in the industry to be able to pack more flash on to a single memory card and to do that you need to shrink down the size of each memory bit, each flash element, and it is becoming very hard to do that. So the nanocrystal flash device is something that has been proposed in the past as potentially being more scalable. What we did was to take, again, a self assembling system of material that organizes into nanometer scale features by itself, all by itself, without manual intervention, and use that to form regular arrays of nanocrystals, silicon nanocrystals. Then we built the device from the next step making the device and seeing how to operate it and improving that. In fact this material can be compatible with conventional processes for fabricating devices and can add some performance or functionality to the electron device itself.
MARRIAN: Great, does this type of assembly process have a future in manufacturing? Could you hypothesize on specific problems in which assembly could be the solution
GUARINI: I think that it has a future but again, in only certain areas. I don't see it, at least not in the short term, taking over the way of manufacturing. I think that that is unrealistic in the short term. There is a huge infrastructure that has built up in the last few decades as to how we build devices and there is a reason for that great expertise in the materials that we use or the way of building it. But I do think that there are some key elements in the devices that I just described in the few examples and there are others that we can think of that are also a mixed application where a self assembling material can be utilized where the conventional patterning methods simply can't access those dimensions or the density or the uniformity where we can utilize that and use it in a very traditional sense rather than completely changing the way we build it. I think the concept of being able to truly self assemble your integrated circuit is science fiction today. I don't think we are going to imagine shaking up transistors and wires in a beaker and imagine that they are going to spill out to form a functional circuit; the complexity is enormous. But I do think that if we take a step back from that, and say we have a new structure today and how can we aid that, maybe it is making nanocrystals of a uniformity and size distribution that help the decoupling capacitor, and I think that there are others that we have in mind and still others one can think about.
MARRIAN: Looking further into the future do you think, Kathryn, can we get beyond materials, passive and active devices to get to circuits and chips or is that just a pipe dream?
GUARINI: Someday maybe, I don't know that we know. We don't see a clear path on how to get there at this point.
MARRIAN: OK, I understand that you are now a manager, congratulations on that! Could you tell how that happened and was it an easy decision to make?
GUARINII: It was a little over a year ago when I became manager of the group. It is a group in the Silicon Technology Research department, the same area I had been working in as a researcher. It was a hard decision, for me, because when you take on a new position you cannot always continue all the things you did before and I enjoyed doing research and getting in the lab and getting my hands dirty and getting things done, and while I continue to do some technical work, I certainly don't have the time or the bandwidth to do as much as I would have liked. On the other hand, I now manage a very large group of people who are extremely capable and as a group we get a lot more done than I would have gotten done on my own. And I really enjoy the people part of management, helping to contribute to technical discussions as well as helping employees with career development and progressing in their roles, as well. So that is a part of the job that is new and that I really enjoy.
MARRIAN: What kind of specific skills do you think that a researcher needs to make that transition to being a manager?
GUARINI: I think that part of being a manager is inherent; you have to have inherent people skills to be very good, because part of the job of a manager is to help your employees to be successful and I take that very seriously. It is not your whole job, especially within technical areas. As a first line manager I have direct reports who do technical work. Managers are involved in the technical work itself; in some program management and some setting technical direction, technical milestones, making technical decisions. So I think that there are two sides of it: having the credibility, and having some background in technical areas where you can make a contribution there, and then also being able to be a good communicator and sensitive to issues that arise that may not be technical in nature. I think that is what makes a good manager today.
MARRIAN: I think that you're in an IBM research video that has just been released. You make some very interesting comments about work/life balance. Could you share some of those with us today?
GUARINI: Sure, work/life balance is hard. I think anyone, whether you have children or pets or elderly parents or relatives, whether it is trying to maintain an outside interest and a career, it is just really hard to balance all that. In my case, I am married and have a two-year-old daughter. It is fun and I wouldn't give it up for anything in the world, but it is a balancing act or a juggling act in some senses. So I don't pretend to have it figured out, there are not enough hours in the day to do all the things I would like to do, either for work or for family. But I think there are some things that have made it easier. For instance we have, at IBM, fairly flexible work hours, both as an institution and within my management structure. They are very supportive; your child is sick and you need to work at home, you can do that. We have the capabilities and the ability to be connected all the time, to be able to connect even though you are not physically there. So I appreciate that flexibility. We also have a daycare center that is on site where I work, so I bring my daughter when I go to work in the morning and I drop her off, so I have a little more time to spend with her than I would have otherwise. I am close by if there is an emergency, so I appreciate that. It doesn't make it easy but it makes it easier. So I think that, at this point, I have figured out how, at this stage of my life, to balance some of these things, but you still have to set limitation for yourself and the goal is to figure out what you would be comfortable with. In my case I don't stay as late as I used to, but after my daughter goes to bed I will often take my laptop out and do some work. So I think you figure out ways of balancing the things that we need to do.
MARRIAN: I know you are very active in promoting a career in physical sciences to women and minorities. How can the professional society like AVS do more in this regard?
GUARINI: I think women and minorities are continuing to be dramatically under-represented in the fields of science and technology. And it is, in my mind, clearly not for lack of ability. If you look at the data, the test scores and so forth, the women and the girls in high school and college are doing as well as the boys and men. It is that they are choosing not to pursue it at some point in their careers, whether is it early on when they make that choice or at some point later. I think that there are a number of things we need to address. As an example, I think that we need to target children when they are young. I think people make decisions early in their lives that will have repercussions later. We know in science and math, if you don't take the core curriculum--you don't take the math in high school, you limit your options. Apparently there is a disproportionate number of girls and minorities who select out of some of these options. So it does not become a career option when they are in college or beyond, because they have not taken those prerequisites. I think that one point to address is to reach out to students when they are in elementary school and middle school, when they are impressionable, and show them the excitement of science; not to get them to do calculus when they are in fourth grade, but to get them excited about the potential and the challenge from problem solving and what is fun about it. We have some programs at IBM, such as our Family Science program, where we bring in fourth, fifth and six graders along with their parents to do fun science things, whether it is building a web page or teaching them a life science, and it is great. I mean you would be amazed at how excited that fifth grader is about digital electronics, where if you ask most college students they would roll their eyes. So you get them early and they realize that excitement; so I think that is one point is get them early. Second is teachers. Teachers at the high school and college level have a huge influence on whether students will continue. I think helping to educate teachers and to enable them to bring some of the applications into the classrooms is important. A lot of times we tell students you have to wait until you get to college or into graduate school to get to do the fun stuff. You have to learn the math, learn the calculus, learn physics, but they don't know why. "Why do I have to study math, I don't know what I am going to ever use it for?" But if you can bring some exciting applications, what you are going to do with it, into the school at an earlier stage, people will say "oh that's why" and it can motivate the learning. I think that one other point is to mentor--to have people in academic and industrial roles to reach out to students and to less senior people in the technical fields and help. We know it is difficult and we know it is challenging and there are a lot of challenges, both technical and otherwise. Mentoring helps to promote a sense of camaraderie and excitement about what we do. I think that those are things we all can do but I think the societies in particular can do these things.
MARRIAN: Congratulations again on winning the Peter Mark Award and thank you very much for taking time to be with us today to contribute to the AVS historical archives.