Solar energy intrigues researchers with its tantalizingly immense possibilities and daunting, unconquered technological obstacles. In the following VerdeXchange News interview, Professor Harry Atwater of Caltech in Pasadena, California , gives a detailed explanation of his research into thin-film photovoltaics, which could hold the key to unlocking solar's cost efficiency for widespread utilization around the world.
Share with our readers your research into thin-film photovoltaic materials. The Caltech website states that the most important challenge facing photovoltaic solar cell technology is reduction of the cost per watt of solar power-generating capacity. In pursuit of that goal, your research focuses on silicon. Why is that?
The portfolio of efforts we have going on at Caltech may give you a imagine. We do, indeed, have an effort on thin-film silicon, and the rationale for that is that silicon is, of course, abundant, and moreover, the processing technologies for making silicon solar cells have been well developed by the mainstream, crystalline-silicon, waferbased photovoltaics industry. Therefore, a thin-film-silicon-on, foreign-substrate approach seems the most compelling among many of the "conventional" thin-film technologies. As a result, we're working on silicon rather than, say, copper indium diselenide, cadmium telluride, or other alternatives. One of the reasons for this focus is that if you take copper indium gallium diselenide—it's an area where the possibility of making small solar cells with high efficiency is actually quite appealing—it's historically been quite difficult to scale those efficiencies to large areas. But, when you think across the horizon to manufacturing scales or the manufacturing output, the cumulative output is in the range of tens to hundreds of gigawatts. Approaching the terawatt scale, we're going to face serious abundance issues with some of the components of copper indium gallium diselenide, particularly the indium and the selenium. There simply isn't enough material in those thin film components to make cells that produce renewable energy on the scale of global energy supply. So that's one reason why we've chosen silicon as a thin-film material.
We are also engaging in other research projects that we think will leverage thin-film photovoltaics. One is a recent project on silicon nanowire solar cells that allows us to grow and make cells in nanowires grown on a foreign substrate. The motivating factor for the nanowire solar cells is that nanowire enables you to break solar cell design's strict dependence between light absorption and carrier collection. There are two things that a solar cell has to do in order to be an efficient energy conversion device: absorb light and collect the carriers in an electric current across an open circuit voltage in the cell. Normally, to make a cell absorb all the light, you have to make it a certain thickness. Silicon, for example, is a relatively weak light absorber, and therefore the silicone layers have to be of wafer-scale thickness simply to absorb the light. In order to collect the carriers from a cell of that thickness, carriers generated by light in the semi-connector have to make it to the surface to be collected at the junctions and contacts. Therefore, they have to travel the entire wafer thickness. In order to do that without trapping the carriers and therefore losing external power generation (generating heat instead of power), the material quality has to be a certain level. That's the fundamental thing that is determining the cost of silicon solar cells right now: the need for the crystal growth and purification and cell fabrication that enables this long carrier-collection link. If you could somehow break this limit and collect carriers in materials that have very, very poor quality and therefore low carrier collection and diffusion life, you could use a much lower quality material and achieve a much lower cost, which thereby dramatically reduces the manufacturing costs for silicon cells relative to current fabrication, but enjoys the same performance.
Who supports Caltech's current research in photovoltaics?
We have support from BP, which supports the nanowire project. BP Solar is the largest U.S.-based solar manufacturer. BP is obviously multinational, but in terms of solar cell manufacturing, it's the largest in the United States. We support and work jointly with Spectra Lab, which makes ultra-high efficiency cells for concentrators. That's another area in which we're doing work. The Department of Energy supports us as well, both through basic energy sciences and some of the applied energy efficiency/renewable energy programs. The Global Climate and Energy Project at Stanford also supported us. And we're supported, of course, by our Caltech Center for Sustainable Energy Research.
Germany, Spain, and now the U.S. have been pressing concentrated solar power (CSP); a deal was just signed in Southern California by Solel from Germany and the LADWP to build what's said to be the largest concentrated solar power facility in the world. Are you following this CSP work? What are you working on that could contribute to applications of concentrated solar technology?
Let me be careful to distinguish a couple of different approaches. We are working on devices that are useful in CPV, concentrating photovoltaics. CSP, concentrating solar power, refers to a technology where you concentrate sunlight, usually with a mirror, ray, or another direct solar resource, onto a tower where a thermal receiver heats liquid. The liquid is then driven through a heat pump that heats water, producing hot water or steam that runs a turbine. It's a very compelling approach to solar energy, and there are a number of start-ups in that area. We are, in fact, working on cell technology with gallium nitride solar cells that would enable the combination of CPV and CSP. So, it's possible to imagine a hybrid system that has higher efficiency than either system alone. That's the approach that we think has the greatest promise.
We're not working on the thermal part—the thermal receiver and turbines. GE already builds turbines, and there are a number of large-scale solar demonstration projects. We're working on cells that would absorb the blue and ultraviolet part of the spectrum and transmit some of the red and infrared part of the spectrum to the thermal receiver so that you'd have a target system efficiency of something like 30-percent thermal efficiency and ten percent photovoltaic efficiency for 40-percent overall photon-to-electricity efficiency.
How applicable is your research at Caltech to the market opportunities for renewables?
Photovoltaics has a very bright future. The mainstream silicon industry, as you know, is expanding very rapidly. The portfolio of cell research ideas that we're working on, in addition to the gallium nitride cells, includes conventional three-five compound semi-connector cells, gallium indium phosphite, and gallium arsenide multi-junction cells. We've developed an approach to lower the cost of those cells. We're working together with Spectra Lab, which is the world solar efficiency leader, out in Sylmar, CA. They are now the principal cell supplier to the concentrating photovoltaic industry, in addition to supplying these very high efficiency cells for their first market, which was in satellite power. Our work together with Spectra Lab is aimed at making four-junction solar cells in the most efficient solar materials: gallium indium phosphite, gallium arsenide, and indium gallium arsenide. Our work, particularly here at Caltech, is aimed at lowering the costs of the substrates.
The other areas of your research include second-generation biofuels and cellulosic ethanol. What are your priorities in terms of research and development of those renewables?
Our research is really aimed at solar fuels and trying to develop the inorganic analog to a photosynthetic process—like an inorganic leaf. Solar photosynthesis has a fundamental limiting efficiency at around ten percent. The photon to stored energy—in the case of photosynthesis, it's stored in hydrogen and then carbohydrate-generation in the leaf—generates ATPs from ADP, which are used to sustain the plant. But, of course, we know how to make solar cells, and the typical biomass conversion yields for starch and corn are a few tenths of a percent. For the best materials, such as switchgrass, it's a few percent. Blue-green algae, for bacteria, also stays in that small range of a few percent. But everything in that area is capped by the limits of organic photosynthetic processes to around ten percent. On the other hand, we know how to make solar cells now whose efficiencies are as high as 40 percent, so we can generate photoelectric chemical cells where the primary energy absorber looks a lot like a solar cell, but the surfaces are photoelectric chemical electrodes that generate hydrogen and oxygen rather than having wires attached to them that generate electrical power.
So that's the idea behind solar fuel generation: we're taking inorganic structures made of semi-conductors and metal catalysts that look like solar cells—membranes of semi-conductor insulating materials that generate electrochemical potential large enough to reduce and oxidize water. That is a very different approach. And to set it in context, first-generation and second-generation biofuels are clearly viable. There's a lot of debate, but I think there are fundamental limits to the energy yields of those materials. I would characterize our work on solar fuel synthesis as a grand challenge; it's almost like a moon shot. We're building for the highest-possible conversion efficiency, the most deployable solution, the smallest land area at terawatt-scale supply for fuel generation compared to any conventional biomass process.
The biggest challenges are not in making the photo absorber; as I mentioned, solar cells are very efficient. The big challenge lies in the catalysis. Where are the cheap and abundant catalysts? Currently we use platinum, and platinum is obviously completely un-scaleable to terawatt energy scale significance. We can't deploy photoelectric chemical cells for solar fuel generation at the scale of three percent of the U.S. land area if they have platinum catalysts on the surfaces. We have to find other catalysts made of transition metals, inorganic interfaces between the sun conductor and the electrolyte. Scientists often take cues from nature, but we also have to work with inorganic materials. It's a very different approach from conventional biomass, one that entails more risk and has fundamental scientific challenges, but which has a very high payoff and energy conversion efficiency at a large scale.
As you observe what's happening in the applied commercial world and the dialogue in the political world from your perch at Caltech, what insights can you offer about the trends, actions, and conversations currently taking place?
I feel like, in the last year, for a set of complex and inter-connected reasons, American society has suddenly become very conscious about the need for renewable energy generation. However, don't think that most people appreciate the scale that's needed to meet the growing demand for renewable energy. People know that it's important, but they think they are doing enough when they go out and buy a Prius, turn down their thermostat in winter, and reduce their air conditioning in summer. In order to make an impact on greenhouse gas emissions and non-renewable energy generation, we have to do so much more than that. The next big challenge for engineers and scientists is to communicate the scale of the problem to the world. The renewable energy industry is going to become, and I tell my students this, the largest employer in the United States. It will become the biggest industry in the world. It will become the largest consumer of raw materials and primary energy input to generate renewable energy generation devices.
That's the scale that it's going to be if we're going to have an impact. I tell my own children, my students, and anybody who will listen that it's the most important problem of the 21st century, and it's going to be the biggest issue in their lifetimes, the thing that will determine so much about the development of our world. That's what is so exciting about being in this field at a place like Caltech—it's a place where advances in fundamental science can leverage enormous impact in society. Sometimes scientists can do something really elegant and it has no impact on the world—which is personally gratifying for the scientists but isn't otherwise recognized anywhere outside of the scietific community. This is a case where there is tremendous leverage in the real world, and I think that's a message that resonates very widely.