NREL, First Solar Celebrate Nearly 30 Years of Collaboration on Cadmium Telluride Solar Cell Research

The National Renewable Energy Laboratory (NREL) has more than 550 unique, active partners. One such partner is First Solar Inc.—a relationship that dates back to the early ‘90s. NREL and First Solar, then called Solar Cells Inc., have been collaboratively breaking ground on thin film solar technology for more than two decades, helping NREL fulfill its goal as a Department of Energy (DOE) national laboratory of commercializing technology through partnerships, and contributing to First Solar’s success in development, manufacturing, and operation of photovoltaic (PV) power plants with technology that still bucks industry trends.

In early July, First Solar made headlines when it announced its intention to construct a $680 million solar panel manufacturing factory in Ohio, bringing with it an estimated 500 jobs. The announcement came just seven months into the inaugural year of the Biden administration, whose intersecting priorities to “build a modern, sustainable infrastructure and an equitable clean energy future” and “ensure the future is ‘made in all of America’ by all of America’s workers” have been well supported by Secretary of Energy Jennifer Granholm.

NREL is on the job, working consistently to enable such a future for nearly 45 years. When Solar Cells Inc. came along in the early ‘90s, the collaboration centered around the reliability, stability, and efficiency of the thin film cadmium telluride (“CdTe” for short) technology that it was using in its solar panels, also called modules. Then and now, the vast majority of solar modules are of crystalline silicon construction instead of cadmium telluride.

The Early Days of the NREL–First Solar Partnership
In 1991, NREL was designated a national laboratory by President George Bush and renamed from its original title, the Solar Energy Research Institute. It was a different time, and unique approaches were taken with research related to CdTe technology.

NREL Group Manager of Materials Reliability and Durability Teresa Barnes said, “in those early times, it was scrappy little startup companies and it was a bunch of old friends working together. The CdTe community has always been very, very small. It was primarily in the U.S., and it was a much smaller research community than any other solar technology that I’ve ever worked in.”

Larry Kazmerski, a retired NREL senior scientist who headed materials and characterization for solar research at NREL at the time, said, for the duration of his time working on Solar Cells Inc. projects, there was never a nondisclosure agreement in place. It was just trust.

“You have to realize they didn’t come to NREL and say, ‘Hey we have a problem. Can you solve it?’” he said. “They would bring people out, and there would be discussion, and people would be looking at it in the lab together. It was a fairly open policy. They could come in and work together without much of a problem.”

Kazmerski saw a larger dependence on NREL, especially in the solar PV area with DOE, industries, and universities. It was more of a partnership of all players. During his time at NREL, Kazmerski recalled two NREL researchers who played key roles in moving CdTe research forward. One was Tom Surek, who pushed university, industry, and national laboratory partnerships, and who Kazmerski called the kingpin of the PV program management side. The other was Ken Zweibel, who brought his physics degree to the program management side of the house and was key in selling the laboratory on the value of CdTe. Kazmerski said they were blessed with program managers who were highly technical.

Reliability of CdTe solar cells was critical early on. Kazmerski said First Solar did not want to release anything until it was ready. They knew they had to have reliable modules. They tested everything to death and worked with NREL to validate efficiency and reliability. Barnes said First Solar is one of the best in the business for understanding reliability and durability, and NREL has learned a lot from the partnership and how it affects product development.

In fact, NREL is still learning from that early success today, thanks to one especially noteworthy achievement that planted roots two and a half decades ago. Twenty-five years ago, NREL installed a 0.6 kW Solar Cells Inc. PV system with 12 modules for a long-term test at the Outdoor Testing Facility. Those modules are still operating today.

Dirk Jordan, a senior engineer at NREL who studies the long-term performance of PV systems and analyzes the physics and chemistry of changes, said, “it’s the longest running system we have. It showed the degradation of a pretty reasonable 0.5% per year, which is in line with what silicon systems do.”

A Few Key Milestones in the Journey
NREL and First Solar have crossed a series of thresholds together and hit numerous milestones along the way. NREL Director of the Power Systems Engineering Center Ben Kroposki said producing a CdTe cell that converted light to electricity with 10% efficiency was a major accomplishment Further research and development has led to today’s CdTe solar cells with efficiencies above 20%.

“When they were first coming out, they did not have very high-efficiency modules, but in working with them—and this was pretty early on—they always had a good idea of how to mass produce panels where they were basically making glass and depositing the CdTe on the glass,” he said. “They had a breakthrough working with some of the materials people at NREL that pushed efficiencies over 10%. That’s one of the big milestones: once they crossed over 10% efficiency and could start mass producing using their production techniques. That was a game-changer for these large-scale systems.”

Scalability has always set the First Solar partnership apart. A partnership goes two ways and where NREL was able to help test, validate, and provide fundamental research for First Solar, First Solar was able to teach NREL a thing or two as well.

“I think First Solar is always our example of how important it is to have a really manufacturable, scalable, steady process and to the importance of reproducibility,” Barnes said. “Because they got so good at scaling those results as they came through, they can now take that latest and greatest and scale it very quickly. For us, it was learning the lesson of not just going for that one, record cell, but understanding what the manufacturable technology limits are and having a process window that’s big enough to actually work with.”

Matt Reese, an NREL senior scientist and CdTe lead, recalled when, several years ago, the NREL group wanted to identify the sources of voltage loss in CdTe cells compared with competing technologies. The group developed experimental models and determined the problem was coming from the copper used to dope the CdTe cells. “Copper moves around a lot and is one of the biggest causes of instability and limits the efficiency,” Reese said.

First Solar found out NREL was working on this, and told NREL they would like the research to be conducted as precompetitive, meaning it would be done much more openly and for the purpose of developing new commercially applicable technologies.

“It turned out we were able—for the first time and using single crystals—instead of getting 0.85 volts, we were able to get out more than a volt,” Reese said. “We understood what was wrong with the system. Once we had this demonstration, First Solar was very excited and started working on its copper replacement process.”

Nature Energy published a paper on crossing the 1-volt barrier with single crystals. Three years later, NREL and First Solar published a paper together, again in Nature Energy, showing that manufacturable thin-film CdTe devices doped with arsenic have efficiencies on par with copper (Cu)-doped devices and a much lower long-term degradation rate. This new technology, called “CuRe” (Cu Replacement), is currently being implemented in First Solar manufacturing lines, paving the way for CdTe cells to become even more efficient and reliable.

Through the years, as technology has advanced, First Solar has occasionally licensed NREL-developed technology, the earliest on record coming in 2007. Reese said having First Solar succeed is a win and it means they are doing something right. Kazmerski noted from the early days how those at NREL also felt any success from industry reflected well on them.

NREL and First Solar continue to work together closely. “NREL recently designed and built a successful vapor transport deposition tool like the one used by First Solar to deposit CdTe films,” said David Albin, another lead CdTe researcher at NREL. “This increases the relevancy of NREL’s work for First Solar. In addition to a long-standing Cooperative Research and Development Agreement (CRADA), we have recently partnered with them in two different cost-share projects that are studying the compatibility of NREL intellectual property with their process.”

Gang Xiong, director of the First Solar California Technology Center, has witnessed the NREL–First Solar collaboration on CdTe research for a decade. “The collaboration is fruitful as we can leverage the complimentary strength of both parties to advance the CdTe solar cell technology together,” he said.

“This partnership truly does add up to more than the sum of its parts. First Solar is really good at screening innovative ideas rapidly and driving toward perfection in our R&D labs and our manufacturing lines. NREL is a great source of idea generation, and they have a lot of expertise in characterizing and understanding material/device performance-related issues. We openly share our R&D priorities with each other.”

How NREL and First Solar Helped Shape the Solar Industry
The significance of the partnership and the work jointly produced has had an impact on the solar industry beyond simply taking the last survivor of the initial wave of early CdTe companies to the top. It has established a blueprint for introducing and scaling up new technologies and helped push the boundaries and test the limits of what’s possible.

“We have a market of silicon that is always changing and improving and of course CdTe,” Kazmerski said. “If it wasn’t for CdTe, maybe we wouldn’t even be looking at perovskites now. People started to look for alternative thin films that are better. These guys are always keeping on top of things to see if they can’t themselves come into this and be a player or find out what the competition is doing. First Solar is here to stay with their CdTe, but they have shown that a thin film can be very, very competitive. The perovskite people are starting to mimic what First Solar did to show that these modules are really stable.”


“Early on in the game, they decided to implement full-scale recycling, so they really understand the process end-to-end,” he said. “Only now are the crystalline manufacturers realizing they have to get into this recycling game and start to think about that.”

Barnes acknowledged the success of First Solar and called attention to the importance of having some healthy competition. She also said the fact that they are one of the few U.S. companies still in the business is also important, but First Solar has been looking to advance CdTe technology beyond just what it is using in its modules.

In March 2021, after a series of workshops begun in 2017, Colorado State University, the University of Toledo, NREL, and Tempe, Arizona-headquartered First Solar kicked off an alliance called the U.S. Manufacturing of Advanced Cadmium Telluride (US-MAC) photovoltaics consortium, designed “To further reinforce U.S. leadership in solar technology.”


The press release from the University of Toledo read, “US-MAC will work with its members to advance foundational science and engineering, stimulate innovation, and capitalize on shared resources and expertise. It aims to support increased production volume, optimize performance, and to diversify, integrate, and support the success of domestic firms.” US-MAC includes a total of 10 industry partners (including First Solar) interested in furthering the success of CdTe PV.

“This partnership is one of the best examples of why the lab was set up originally and why it was developed as it was,” Kazmerski said. “It showed that the lab could really help a U.S. industry.”