Mr. Bendis opened the panel by saying he had observed an indirect but substantial benefit of the symposium: new conversations, both between people working in the same agencies, and those in different agencies. “The format is giving people an unusual opportunity to hear about gaps in understanding and new ways to leverage resources more effectively within the federal technology investment portfolio that already exists,” he said. He applauded the DoE for taking the lead in bringing other agencies into the discussions.
Acting Director, Electronics and Electrical Engineering Laboratory
National Institute of Standards and Technology (NIST)
At NIST, said Dr. Rochford, “we are very broad, but we focus like a laser on measurement science.” NIST also helps develop standards and promote technology. He called the institute “fairly small,” compared to DoE labs for example, with about 2,800 employees and almost that many affiliated associates, postdoctoral students, and guest researchers from companies, universities, and metrology institutes. About 400 NIST staff work with more than 1,000 bodies to help set standards. Among the Institute’s ongoing programs are the Technology Innovation Program, the Malcolm Baldridge Quality Program, and the Manufacturing Extension Program. He said he would focus on the laboratory programs, which
were organized by discipline, including several that bring measurement science to bear on the photovoltaic technologies area.
The Goal of Measurement Traceability
Why do measurements matter? They make possible the research collaboration and commercialization of products that underpin innovation and trade.
To illustrate his answer, Dr. Rochford described the recent international comparison of solar module ratings. Published three years earlier, these ratings were based on an experiment by NREL that sent solar modules to major laboratories around the world. These modules were of four kinds: monocrystalline silicon, thin-film silicon, cadmium telluride, and cadmium-indium-selenide (CIS). On the CIS system, a measurement spread of almost 9 percent was found about a mean. “The graph has taken eight participants’ measurements and assumed that the mean is the right value. That’s not necessarily true.” He also said that the accuracy was not sufficient. “Our role is to create measurement traceability,” he said. “As the National Metrology Institute for the U.S. our job is to be able to trace measurements to the international system of units. By providing measurement traceability, we can strengthen comparisons of measurements across companies, nations, and laboratories. To facilitate trade, the companies, vendors, and other participants have to agree on what a product is, and agreement is based on measurement. Also, if you want to innovate efficiently, you have to be able to accurately measure product characteristics throughout the R&D process so you can share results and perform reproducible engineering production and even reproducible research.” For these reasons, he said, NIST provides traceability to many places, for example NREL, which does PV measurements for a variety of vendors. At the top level, NIST ensures that all the units are traceable internationally by working with international partners and other national metrology institutes.
In addition to the high-level metrology, NIST also offers develops measurement science and services. For example, the Building and Fire Research Laboratory (BFRL) seeks to improve measurements needed to certify and model net zero buildings. Recently the Institute developed a new high-speed radiometer to measure the performance of PV panels, allowing the industry to move away from over-reliance on a single standard artifact. In some industries, he said, metrology is often limited to a single “golden sample,” in the assumption that a certain test piece is of adequate quality, stability, and suited to the measurement problem, but this makes broad applicability very difficult.
BFRL, he said, is building simulation tools to improve evaluation of energy usage in buildings. Because solar energy is just one aspect of this challenge, the lab has gathered data and modeled buildings in a variety of applications to bring a better understanding of how to perform simulations. This will enable better economic and energy budget analyses of buildings.
He said that photovoltaic technologies R&D must be viewed in the context of a variety of electronics and power-conditioning systems, a broader area NIST has worked in for many years. “This isn’t your standard CMOS stuff,” he said, “this is semiconductor electronics that can convert the DC power of a solar panel into the AC power used by the house and the grid. We’ve taken a leadership role in some of those measurements and developed measurement traceability to help that sector produce more powerful conditioning systems.” In this effort, he said, NIST also leads interagency groups in semiconductor measurements with DoD, DoE, and other partners.
The Need to Guarantee PV Lifetimes
Dr. Rochford emphasized that the service life of PV modules is critical. “For PV to make sense,” he said, “you have to be able to sell a product that has a guaranteed lifetime in order to get the expected return on investment.” NREL and other labs are working on service life, which he said is “not a trivial problem.” NIST has built a device that simulates accelerated aging at known humidity and temperature, but to predict accurately the lifetime of a product, it is also necessary to understand the processes that degrade it. This requires a microscopic and nanoscopic understanding of aging. NIST has found that some tools developed in the semiconductor industry can be adapted for us in photovoltaic technologies. For example, researchers have showed that defects generated at the silicon dioxide interface of PV have a mechanism similar to electrical stresses in MOSFETs.9
Developing Next-Generation Tools
In the future, Dr. Rochford said, NIST planned to leverage some current capabilities into later-generation PV, such as using inhomogeneities or nanostructures to increase efficiency. “But clearly, to make those work,” he said, “there will have to be better understanding of some processes.” As examples, he mentioned carrier generation, carrier transport, and electron hole band diagrams at the smallest scales. “By developing measurement capabilities that can be applied to next-generation work,” he said, ”we will have the tools to better understand reliability and defects in today’s manufacturing at the same time we are advancing the learning curve.”
A final point of involvement with PV manufacturing is solar-related documentary standards. In the U.S., he said, these are not handed down by the government, but produced through a collaborative process. NIST’s role is to send experts to those making the decisions. “Because we are technology agnostic and are not trying to make a profit,” he said, “we can provide participants with unbiased
9A metal-oxide-semiconductor field-effect transistor, or MOSFET, is a device used to amplify or switch electronic signals.
technical information and assessment.” NIST intends to start a mapping exercise in 2009 to look at any gaps in the documentary PV standards and advise on where additional work may be needed.
He concluded with the comment that NIST is also active in developing standards for “smart grid, an area where the documentary standards effort is just huge. This requires a lot of input from the PV community, because the whole point of smart grid is to be able to use intermittent and renewable sources. As the PV community grows, we’d be very interested in any type of consortium on standards where we think we can help.”
Thomas W. Peterson
NSF Directorate of Engineering
Dr. Peterson said he would discuss three aspects of the National Science Foundation (NSF). First, he would describe some of its renewable energy research, conducted primarily in the engineering (ENG) and mathematics/physical sciences (MPS) directorates, and also throughout the foundation. Second, he would discuss specific examples of PV research, and again primarily in two directorates. Third, he would end by talking specifically about translational research and engineering and how that supports renewable energy.
He said that NSF is not strictly a mission agency, but rather has the broad mission of supporting basic research and education in science and engineering. At the same time, its portfolio of programs does contain extensive energy investments. One example of renewable energy research was the Green Gasoline Project, developing direct conversion of cellulosic materials to hydrocarbons or feedstocks for gasoline. This project, based at the University of Massachusetts at Amherst, uses fast catalytic pyrolysis to generate hydrocarbon feedstocks directly from cellulose.
Liaisons with Industry
The NSF, Dr. Peterson said, had an extensive program that emphasized aspects of energy manufacturing, systems engineering, and industrial engineering related to production of alternative energy devices. This program supported a project whose goal is to extract energy from ocean waves via linear direct drive generator buoys. This is done in collaboration with a small company in a new program called GOALI, Grant Opportunities for Academic Liaison with Industry.
There are also renewable energy programs in a number of NSF’s Engineering Research Centers (ERCs), which he called “one of the jewels in the crown.” These are large-scale operations involved multiple institutions, almost all of
which are partnerships with industry. Four projects have energy components: North Carolina State University is the lead institution for a project on smart grid issues; Iowa State University on biorenewable chemicals; Rennsalaer Polytechnic Institute on smart lighting; and the University of Minnesota on compact and efficient fluid power.
Another program that supports energy research is outside the disciplinary divisions: Emerging Frontiers in Research and Innovation (EFRI). The objective of EFRI is to support high-risk, potentially high-return research. Unlike most of the foundation’s proposals, which are unsolicited, EFRI designates its solicitations for specific target areas. During its four-year history, EFRI has focused on a number of topics, but in the last two years, half have been energy related. The energy programs have supported work on resilient and sustainable infrastructures, hydrocarbons from biomass, and for 2009, both renewable energy storage and engineering sustainable buildings.
Within ENG and MPS, the following projects focus specifically on PV research:
• Center for Powering the Planet, headed by Harry B. Gray at Caltech; developing components for solar water splitting. The objective is to generate hydrogen that can be used in fuel cells.
• Nanoparticle Catalyst for Fuel Cells, headed by P. Strasser, University of Houston; developing nanoparticle catalysis for application to fuel cells.
• Solar Cell Material Surface Structure Observed, Angus Rockett, University of Illinois at Champaign-Urbana; coordinates studies of the material properties of surfaces used for solar cells. It seeks insights into the nature of the semiconductor junction to help explain why some solar cells work and some fail.
• Renewable Energy Materials Research Science and Engineering Center (MRSEC), Craig Taylor, Colorado School of Mines; materials research on PV materials and fuel-cell membranes, from fundamental physics of electronic excitations to applied aspects of PV cell efficiency.
• Optoelectronic Processes in Materials for Solar Energy Conversion, University of Central Florida; nanosystems of conducting polymers and fullerenebased material; single-particle spectroscopy studies reveal that the states of the aggregates affect material function.
• Self-assembled Biomimetic Antireflective Coatings, University of Florida; novel templating nanofabrication platform to mass-fabricate broadband coatings for solar cells. Coatings mimic antireflective moth eyes and super-hydrophobic cicada wings.
• Nanostructuring of Silicon Surfaces for Photovoltaic Devices, Georgia Tech; molecular lithography is used to pattern silicon substrates with features that depend on the size of the molecules used, instead of on the lithography tools. This technique is thought to herald a breakthrough in manufacturing efficiency and cost.
FIGURE 10 ENG investment in photovoltaic research.
SOURCE: Thomas Peterson, Presentation at July 29, 2009, National Academies Symposium on “State and Regional Innovation Initiatives—Partnering for Photovoltaics Manufacturing in the United States.”
He showed a rough graph of what ENG had invested in PV. Compared to the total DoE investment it is a small number, he said, but growing. He expected it to be over $6 million in FY2009.
Increasing Emphasis on Translational Research
NSF places heavy emphasis on translational research—again, mostly within ENG. In the realm of renewable energy, this means both basic research and research that has industrial and commercial potential. It is almost always interdisciplinary, involving primarily teams of universities, but also of companies and government agencies. By definition, translational research relies on partnerships and is expected to deliver clear benefit to society. Programs considered to support translational research include the Science and Technology Centers, the Engineering Research Centers, GOALI, MRSEC, the SBIR program, EFRI, and the Industry/University Cooperative Research Centers (I/UCRC). The emphasis of these programs spans the squares of Pasteur’s diagram, from discovery mode through involvement with small and large businesses.
While the majority of NSF funding continues to support basic research, the I/UCRCs represent an increasing emphasis on application and commercialization of knowledge. The basic model for them, he said, was to enable discovery and innovation through collaboration. “The model works almost like a research franchise,” he said. “The NSF seed money is small and intended to act as catalyst, while the foundation takes a supportive role throughout the life of the center. The
FIGURE 11 Filling gaps.
SOURCE: Thomas Peterson, Presentation at July 29, 2009, National Academies Symposium on “State and Regional Innovation Initiatives—Partnering for Photovoltaics Manufacturing in the United States.”
I/UCRCs consist of one or several universities, but they are funded primarily by industry, and its Industry Advisory Committee is the critical component. It is a specific management and structural model with independent evaluation tools.” Over the past two decades, the program has supported some 35 to 50 centers each year, with a total of about 100 sites throughout the country.
A specific example of I/UCRCs is the Silicon Solar Consortium, or SiSoC. It consists of four universities (North Carolina State University, Georgia Tech, Lehigh University, and Texas Tech University), several national labs, and 15 industry partners. Its objective is to reduce costs and increase performance of silicon PV material, PV cells, and PV modules while developing novel breakthrough designs and processes. Two main research foci are metrology and wafer breakage. Being an academic program, the primary technology transfer mechanism is to educate graduate students who can later move into the industry.
Dr. Peterson closed with the following summary:
• The NSF has a broad renewable energy portfolio, which includes photovoltaic technologies.
• Although the mission focus of NSF is basic research, the ENG research portfolio includes strong translational research programs.
• I/UCRC is an important element of the foundation’s growing commitment to translational research.
• SiSoC, a multiuniversity, multicompany research program, is an excellent example of NSF’s approach to translational research in photovoltaic technologies.
Colorado State University
Dr. Sites opened by saying that he would discuss how research universities can contribute most effectively to PV manufacturing in the United States. He expressed no doubt that the universities, after many years of fundamental contributions, were well positioned to continue their contributions into the future, both to manufacturing research needs and to the broader development of the PV industry.
University researchers are well suited for the advancement of knowledge in this and any other field, he said. They have a high level of intellectual vitality; they tend to be creative by nature, they are well versed in the literature, and also skeptical, and interactive. At the same time, there are various challenges in combining the respective cultures of academia and industry. It was their mutual responsibility to do so as effectively as possible.
Some Challenges in Working with Industry
Potential challenges for the university working in partnership with industry can be seen in their different understandings of research and education. One university approach of high value to industry is foundational research. Foundational research, he said, is basic research directed toward specific problems. Some examples in photovoltaic technologies are:
• Analysis of optical losses.
• Minimization of forward-current losses.
• Control of uniformity and stoichiometry during fabrication.
• Identification of unintentional energy barriers.
• Correction of degradation problems.
When foundational research is done in universities, it typically includes such activities as development of unique fabrication and measurement equipment,
quite time-consuming and fundamental activities. In the university setting, such research, in addition to looking toward the goals of better manufacturing, must allows for thorough exploration of questions that are encountered during a project. Because much of a university’s PV funding now comes through industrial subcontracts, the breadth and independence of university work may conflict with industry’s emphasis on short-term manufacturing goals. A related issue is that a student in training may soon be working for a competitor of a company that is funding research at that university.
Similarly, a university’s understanding of its educational function differs from that of industry. The education of graduate students is a primary function of the university, he said, and should be encouraged on fairly large scale. PhD students in particular tend to be sources of creative ideas of value both to their advisors and, later, to their employer should they proceed to a career in PV manufacturing. At the same time, their education does not proceed at an industrial pace. Ph.D. students, to do their job right, need stability and time to do their research, which typically takes three or four years.
In addition, the value of academic freedom is central to graduate education, and universities want to preserve for their graduate students the ability to disseminate the results of their research. They also prefer a thesis schedule that can proceed smoothly over those three or four years.
Suggestions for Working Together
“I would make a couple of suggestions,” Dr. Sites said. “The country should fund a solid base of university PV research and its students directly. This does not preclude the possibility of additional industrial contracts at universities; in fact, those investments can bring positive synergies for the funder. Also, we should expand opportunities for research students at our universities to do part or all of their research training at a national lab.”
One thing a university culture encourages, he said, is information exchange and collaboration with other research institutions—in part, because a single university rarely has the resources and expertise to fully investigate a complex problem on its own. A collective approach to PV research problems involving universities, national labs, and industry will cross-fertilize and synthesize new ideas that may elude individual investigators. “The sum is likely to be greater than the parts,” he said. There are already strong regional collaboration centers and nationwide networks of researchers in specific technologies. “I would suggest constructing proposal solicitation to encourage collective and consortium submissions. Also, we should support topical workshops in different PV technologies, meeting regularly. A lot of collaborations form when people can get together and chat about their work individually.”
Universities are also qualified to bring leadership to large PV manufacturing questions, he said. “We all do a fair amount of strategic planning. Universities
have expertise that can help at the national scale. At same time, faculty are skilled at evaluating people and research proposals. We tend to have that skepticism, a healthy thing to integrate into review.” He acknowledged the potential for conflicts of interest when most faculty are also submitting proposals, a conflict that is hard to avoid, especially when the number of qualified reviewers is small. “My suggestion is that we lighten up on this a little,” he said. “I know the NSF manages this fairly well, as do journal editors. We should manage rather than try to eliminate conflict of interest.”
For the PV manufacturing industry, he suggested that an issue of vital importance is its relationship to other countries. The United States has fallen behind as a manufacturer of PV over the past 20 years, he said; it has also fallen behind in foundational research, particularly with respect to the European Union. “So as we build up our industry in the United States, there is a question: Do we view other countries as partners, or competitors, or both? How can we most effectively come to terms with them?”
Filling a Research Gap
Dr. Sites raised a final specific issue, a gap in our research emphasis. “We have a habit of investing in highly fundamental research, with pretty long-term horizons. We also tend to invest in industrially driven research. But between them lies a gap: foundational research related to current technologies.” He suggested moving effort and money into this gap to advance PV technologies for the 2015 time frame.
He gave the rationale for this change by describing PV in terms of three generations. During the first generation, he said, crystalline silicon was dominant, producing high efficiency at reasonable cost. The second generation has included thin films, with low cost and reasonable efficiency. The third generation is “a little more vague”—either very low cost or very high efficiency. “I would be skeptical when people try to change that ‘or’ into an ‘and.’ In any case, all these generations need foundational research and they also need critical review. Directionality—that is, the idea that new generations will replace the old ones, which then die out—cannot be assumed. We need all three generations at once, each of them a comprehensive whole with continuing possibilities.”
He closed by mentioning a difficulty for some universities in forming PV partnerships. “We’ve been burdened a bit with cost-share requirements from some parts of DoE,” he said. “This requirement, typically 20 percent, can be a disincentive for some of the more creative faculty who might otherwise work in a PV area. I hope we can take another look at that policy.”