Dave Auston, Executive Director for the Institute for Energy Efficiency discusses the different types of research that his organization conducts.
Dave Auston: The Institute for Energy Efficiency is two years old now and about to enter its third year. It was started by Matt Tirrell who at the time was Dean of Engineering along with John Bowers who is our faculty director, and it really was based on not just the need for an additional university-based institute of this kind because, you know, there are quite a lot throughout the country. But there was a feeling that we could do something distinctive and different than some of the other university-based institutes that are directed at energy and climate change. And the thing that does distinguish ours is that we’re a very strong program here, especially the materials aspect of the science and engineering of energy renewables and related aspects of the energy crisis.
The program material science here is extremely strong. Just to give you an illustration of that one consequence of the Institute’s work is that very soon after it was established; it pulled in a $19 million grant from the Department of Energy for work on materials or what are called energy efficient applications. That’s one of the Department of Energy’s new energy frontier research center and it encompasses three subject areas that are currently actively being pursued and researched. One is solar cell technology and there are two aspects to that. We have a very strong program here in the use of polymers, much less expensive materials and materials can in principle be manufactured much-more readily than the semi-conductors. So then crystalline semi-conductors. Then we also have a program in what we call high-efficiency, primarily multi-junction solar cells. Both of these are actually a collaboration with NREL, a very important collaboration there and that we have of our total of some 22 faculty, five are actually from NREL, and we have a very close partnership with them in that regard.
Then we also have a really fabulous program in cell-state lighting. I don’t know if you’re aware of the background, but I can give you a little bit of information there that might help you understand what we’re doing. The field of cell-state lighting, LEDs, for applications to lighting, really received a major impetus with some breakthrough work. The person on our faculty by the name of Shuji Nakamura did, it was almost ten years ago now. He threw the development of some nitride family of materials and related compounds. He was able to make the first really effective and manufacturable blue lasers and then blue LEDs. That opened up then an enormous area of opportunity of research and development and ultimately transferred into the marketplace of current LEDs for lighting which utilize a blue LED that strikes the phosphorer to produce the white light that comes out. And they of course are enormously important for energy efficiency applications.
Ben Lack: And so the research that’s being done is to find ways to lower the cost or rather increase the output of the blue lasers?
Dave Auston: It’s both. Now they’re working actually on three-color LEDs because ultimately that gives even better performance than taking a blue LED and striking a phosphorer with it. Three color meaning RGB, red, blue, green, just as you have in a TV system and in your computer monitors. That’s the better way to go. The problem with that now is that the green LEDs, while they can be made, are not very efficient. There are some issues that are not well understood with regard to the material properties there. And so our group has a major effort directed at improving the efficiency of the green LEDs, understanding just what it is that limits their efficiency. But also something called droop that we’re looking at very carefully which is that LEDs, even the blue LEDs, when their driven hard at high currents, their efficiencies decreased. And that’s not well understood either. And so there’s a big of our program that’s directed at understanding those.
And then we have a third area of research, all within this Department of Energy grant which is only a part of the Institute. I’ll come back and tell you more about the Institute in a moment. Thermal electrics is really a nifty field and it’s sort of a sleeper because it’s been neglected. It’s understood in terms of how it works for the most part. And it’s actually a concept that’s been around for decades but never really deployed in a major way because of relatively low efficiencies. But we have a group here that is using two approaches. One is a very precise layer by layer deposition of materials through something called molecular beam epitaxy where you can actually control the structure of material on an atom by atom layer level and basically synthesize materials that way. The others use nanotechnology to incorporate a particular structure in these materials that enhances thermal electric properties. And the potential here and the goal here is to be able to make efficient materials and devices that can convert waste heat directly into electricity which is what thermal electric materials do. It’s the simplest imaginable device. You have just a little chunk of very specially prepared material. You make one side hot, the other side cold and low and behold, electric current runs through the material. And already there are some applications that we’re pursuing. For example, taking waste heat and using it to cool semi-conductor circuits is one application.
There are some companies we’re working actually with BSST, a company that makes thermal electric materials. They’ve been selling these for applications actually to heating automobile seats. You can reverse these also. You can run them, it’s really nifty property of thermal electric devices is that you can run them either as coolers or meaning you’re running electric current through it and you can cool something. Or if you have source of heat, you can apply the source of heat and you can get electricity out of them. So they’re very versatile.
And all the other things we’re doing are also, as I said, are things that we don’t choose the research unless we believe it has some really important impact, potential impact. Now, of course, some of it’s hit and miss research. Not everything is going to ring the bell and become viable commercially. Our approach is to really understand the material properties because we believe it’s essential to do so to make the advances that are necessary.
Altogether we have six areas in addition to what I just told you, we have a program in what’s called the Silicone Photonics which is led by John Bowers, our director, and it’s recently actually just made a major breakthrough whereby the… Well, let me, first of all, tell you what John has done. As you know, there’s a major challenge in electronics, making microprocessor, for example. Each year they get faster and smaller, more transistors on a chip. One of the really difficult and most challenging aspects of that work is that as you make things smaller and faster, the heat dissipation becomes greater and greater and becomes a major problem, both in terms of keeping the chips cool but also there’s another problems that arises in that the signals, because they’re faster, really don’t propagate well on regular wires. So connecting the chips and even getting the signals around the microprocessor chips at high speed in these very, very dense microcircuits is one of the major challenging issues in silicone technology today.
The parallel with that, you have a whole technology of optical communications, fiber optics, lasers and things of that sort which work at a very efficient level, much, much lower heat dissipation at much, much higher speeds. A lot of people have tried to marry these two technologies, as to bring electronics into silicone microcircuits, but it’s really been tough because all of the materials technology for the optical work is in compact semi-conductors, things like gali marsenai and those families of materials. But they’re not compatible with silicone. You can’t grow them in the same way at the same time.
What John has done is just a fabulous expedient approach. He treats the silicone as the home for the microprocessors, and he takes the gali marsonite compound semi-conductor materials for the optic and then he bonds the two together. He makes the lasers, the detectors, the wave guides and things in the compound semi-conductors and basically just bonds them to the silicone microcircuits and has been able to really, really advance that field in a very important way. It’s a lot more difficult than I’d just described to you, but it’s a very, very effective approach. And one example of that, it was just announced, actually, a few weeks ago through a partnership that John with Intel, is a 50-gigabyte interconnect that Intel has developed and is now gong to release for sale. And it’s just like a USB connector on each end, although it’s not a standard USB, obviously, because that’s limited to hundreds of megabytes per second in terms of transfer.
This is, however, in terms of ease of use the same sort of thing. On the next generation of computers, you’ll have a connector that you can plug this into or if you want to connect two computers or this to an external drive or something, and you can transfer data rates of nominally of 50-gigabits per second which is just phenomenal. It’s an enormous leap forward. And it’s being possible because of this marriage of silicon microprocessor technology with optical, what we call, platonics which is the optical circuitry that is based, as I said, in an entirely different material system. Bringing those two together has been phenomenally successful.
Ben Lack: Definitely a game changer.
Dave Auston: Oh, yeah. Definitely a game changer. And you’re going to see this, for example, the next generation of supercomputers which that are just being designed now. In fact, the Department of Energy is planning to spend, I heard a presentation on this recently, anywhere from two to three billion over the next 10 years to develop what they called the exoscale computers. These are computers that will run at 10 to the 18 floating-point operations per second. That’s a thousand times faster than the current state of the art computers. And they will, out of necessity, deploy this marriage of optical and electronic circuitry. They just can’t do it any other way. They can’t do it with conventional silicon circuitry to get to that increase in speed. So you’re going to see a lot of things happening over the next few years of this kind.
It has consequences for energy efficiency especially because, as I said, the heat dissipation, and it’s not just about the problem of designing these circuits so that you get the heat out. It’s also in a supercomputer; the total amount of electrical energy needed is enormous. You can just, for example, if you were just to naively try to make a supercomputer that’s a thousand times faster, you’d be talking about 125 megawatts of electric power going into a single supercomputer which is totally ridiculous. You can’t do it. So the targets that DOE and others have set for developing this next generation of supercomputers requires a five-fold increase in energy efficiency. And the technology I just described to you is going to a key to achieving that goal. It’s just impossible otherwise. So we’re pretty excited about that.
And then we also have a program in databases, smart databases which is the largest, in fact I think it was the largest Google grant last year to develop schemes for managing databases in a much more energy efficient way. Database and obviously Google being one of the biggest users of such databases require enormous amounts of energy. Every time you and the hundreds of millions of other people are searching the web, are using Google and other browsers and search engines, somewhere there are all these discs spinning, all this data has been analyzed, parsed and is sent back to you. The energy required to run those is enormous. In fact, Google now locates its databases in locations that are close to the lowest possible electricity that they can find. In fact, it’s usually near hydroelectric sources. Again, that’s one of the lowest costs that you can get around the country. So we have a program of that kind. It’s headed up by Fred Chong, a young, very bright computer scientist here at UCSB.
Then we also have a program in energy efficient buildings. And our particular focus there is on the development of smart systems for managing buildings. That’s headed up by one of our faculty, Igor Mezic. And Igor is doing some really neat things to develop the software protocols and systems to take massive amounts of data because of the things that’s not well understood here is how to manage the enormous amounts of data that are generated in a commercial setting. For example, in a commercial building where you might have a few hundred thousand square feet of designable space, with each room heavily monitored. Not just for temperature and air flow and what the external conditions are, but also any equipment in that room that’s drawing energy and dissipating heat. And also one of the neat things that Igor is doing, he’s also monitoring occupancy. It turns out that that’s really important.
People on average generate even in a passive mode just sitting about 150 watts of energy. If you’re more active that can be 200 or if you’re a really big guy, can be even more again. And you pack a whole lot of people in a room, and everybody’s experienced this, the temperature goes up, and that has to be tracked also. But when you put all of these censors in buildings which people are doing and you generate all this data. It’s not absolutely clear how you should take that data and use it to then adjust, not just the heating, ventilating and air conditioning system, but also such things as opening or closing windows. Modern buildings now have these adjustable blinds that control sunlight coming in rooms and things like that, controlling also just the source of energy into the building. And so that’s what he’s doing. He’s developing that in partnership with a number of industries including United Technologies and others.
And lastly we have a group that’s working on policy issues. We have a whole school here at UCSB called Bren’s School of Environmental Science and Management, and it’s main thrust is actually in policy in addition to our economics department. It’s, for example, been doing studies on carbon pricing, things of that sort. And because we’re a campus-wide institute, we try to bring and do bring the policy people into contact with the people doing the science and engineering because we think that is a good two-way communication channel. People in science and engineering need to know where the policy issues are and where they’re headed. But even more important, the policy people need to know where the science and engineering is headed and so we do a lot of that also.
And so that’s sort of the programs that we have. But then we also do a lot of things in terms of outreach. We have once a year a major summit conference we call the Santa Barbara Summit on Energy Efficiency and it features very high level speakers and attendees and topics. We’ve had, for example, Arun Majumdar who’s now currently the head of ARPA-E, people from Google, people from all sorts of places. We bring in people from, for example, Randy Komisar, from Kleiner Perkins, big VC entity in the Bay area, bring them together to focus on some of the major issues, mostly around the science and technology, but frequently with the policy component as well. And that’s been a very successful meeting. We do that once a year.
We also have much smaller, more focused, what we call, technology forums. We did one, for example, recently on light-emitting diodes, and the central focus of it was bringing together people from industry, government, and university. Most of it was really industry reps, people especially in decision-making roles in industry. Focused on the single question, what needs to be done and how do we do it to accelerate the deployment of light-emitting diodes into the marketplace. The technology is terrific. The cost is still high. There’s still skepticism among consumers about lifetime or liability. Different colors, although most people really find the color of LEDs to much more attractive than compact fluorescents. And as you know, compact fluorescents also are toxic because of the mercury whereas LEDs are not. There’s no toxicity in the materials used, but the cost threshold is still a big issue. And everybody knows that once you ramp up the production levels, that cost will come down, way down but getting over that hump is tough. And that was the focus of that. We’re just organizing a new workshop on efficient buildings, and energy-efficient buildings. The theme of it will be how to not only develop these systems for managing these smart buildings, energy-efficient buildings, but also to develop some of the first protocols for standardizing the software will ultimately be used in these cases because you’re probably aware one of the big challenges is, when you have a new field like this, everybody sort of rushes in. That’s great because you want lots of activity, but people start doing their own thing and then you end up with all kinds of problems or compatibility. You’ll have a particular software package that will work with a certain set of components and regimens and goals and another won’t. And universities being a neutral party in this, we don’t have any vested commercial interest in this are a good place to actually bring those parties together and to start the dialogue that will ultimately lead to compromises so that some standards can be established.
Tags: ARPA-E, Arun Majumdar, blue lasers, blue LED, cell-state lighting, climate change, daily energy news, daily energy report, dave auston, energy news, energy report, institute for energy efficiency, Kleiner Perkins, LED, Light Emitting Diode, microcircuits, microprocessor chips, molecular beam epitaxy, NREL, phosphorer, Randy Komisar, semi-conductors, silicon microprocessor technology, solar cell technology, thermal electric material, Thermal electrics