When considering where solar energy is heading, Tobias Hanrath finds it helpful to look backward.

“If you’d asked in 2012, I think many people would have said all of the interesting materials in photovoltaics have been discovered,” said Hanrath, the David Croll Professor in Engineering in the Smith School of Chemical and Biomolecular Engineering in Cornell Engineering. “Perovskites are a clear example that that’s not the case. They kind of came out of nowhere.”

Researchers such as Hanrath from across a range of disciplines at Cornell are working to develop and commercialize this next generation of solar technology, and exploring how best to implement it. The university – like New York state itself – is undergoing something of a solar boom, with interdisciplinary projects underway in almost every college – from incorporating solar panels into farmland via agrivoltaics, to designing biologically inspired solar materials that can cling to different types of architecture, to studying the impact the state’s solar expansion is having on the local workforce.

“For a long time, there weren’t many of us working on solar,” said John Marohn, professor of chemistry and chemical biology in the College of Arts and Sciences. “And that’s really changed.”

‘Come on, guys, get working’

Perovskites – discovered in 1839 and named after Russian mineralist Lev Perovski – are a class of compound minerals that over the last decade have become the most exciting alternative to silicon, which since the mid-20th century has been the material of choice for manufacturing photovoltaic cells, due to the fact that silicon is durable and efficient – up to a point.

Credit: Ryan Young/Cornell University

As good as it is, even silicon is at the mercy of the Shockley-Queisser limit, a theoretical law that states solar cells can only attain, at most, 34% efficiency – the rest of the energy gets reflected back to the sun or turned into heat. Silicon maxes out at less than 30%, although that percentage can be boosted when other materials are stacked atop it, in so-called tandem solar cells.

“A lot of people I talk to about solar cells say, ‘It’s only 20%? What about the other 80%? Come on, guys, get working,’” Marohn said. “We’re like, ‘Well, you know the limit is 34%, so 20% is more than halfway there, right?’”

But silicon presents a number of far greater challenges. The material actually doesn’t absorb light very well, so it needs to be made in thick, pristine layers: long lasting, yes, but difficult to recycle. The biggest problem is that the production of silicon requires enormous amounts of energy, in the form of extreme pressure and superheating from carbon fuels, which drain resources and take a toll on the environment.

Over the years, several alternatives have been studied. First were thin films, such as cadmium telluride (light and small, but full of toxic metals and not very efficient). There were also dye-sensitized solar cells and inorganic semiconductor nanoparticles called quantum dots, which have strong light-absorbing properties. Next came organic photovoltaics, which can be made under more mild conditions and are currently being studied by Marohn and his colleagues.

Perovskites are essentially the offspring of organic photovoltaics, according to Fengqi You, the Roxanne E. and Michael J. Zak Professor in Energy Systems Engineering in Cornell Engineering. Their crystal structure is different and includes inorganic elements, including trace amounts of metals, such as gold, but without nearly the level of toxicity found in many thin films. But perhaps the greatest advantage of perovskites: They are made in solution and can be fabricated in identical batches with low-cost methods, such as inkjet printing and roll-to-roll processing.

“The inputs – the capital input, the energy, the labor – for making these perovskite products are much lower than silicon,” You said. “For silicon, you need high temperature, high pressure. Not everyone can afford that equipment. Those are so expensive and so strict. But making perovskites is like cooking.”

‘A frankly crazy material’ 

Cornell faculty have been responsible for notable advancements in perovskite research. Qiuming Yu, Ph.D. ’95, professor of chemical and biomolecular engineering (Cornell Engineering), created what are called 2D lead halide perovskites, in which the atoms are stacked in sheets, like phyllo dough; and Zhiting Tian, assistant professor of mechanical and aerospace engineering (Cornell Engineering), has shown how perovskites and their analogues can be used in thermoelectric devices to capture waste heat.

For more than a decade, You has been studying the performance and design of perovskite photovoltaics and quantifying their environmental and financial benefits. Recently, his team demonstrated how a low-cost, environmentally friendly solution can be used to recycle all valuable components from perovskite photovoltaic waste, including everything from the charge-transport layers, substrates, cover glass and electrodes.

The aqueous-based recycling approach can also rejuvenate degraded perovskites, and You’s team’s research (published Feb. 12 in Nature) found it “reduces 96.6% resource depletion and 68.8% human toxicity (carcinogenic) impacts associated with perovskite photovoltaics compared to the landfill treatment. With recycling, the levelized cost of electricity decreases by 18.8% for utility-scale systems and by 20.9% for residential systems.”

Credit: Ryan Young/Cornell University

While lead halide perovskites were commonly researched in the 1970s and ’80s and were explored for transistor applications in the ’90s, it was only in the past 15 years that they became truly competitive for solar applications, thanks to a growing body of research.

From 2013 to 2020, perovskites rose from roughly 14% to 26.7% efficiency, according to the National Renewable Energy Lab, a remarkable leap compared to other materials.

There is also something deeply confounding about perovskites, Marohn said.

“It’s a frankly crazy material,” he said. “If, for example, you do Raman spectroscopy on one of these perovskites, it looks like a liquid; the atoms are jiggling around like crazy. And then you go to the X-ray diffractometer, and it looks like a solid. The atoms are all ordered. So it’s totally schizophrenic.

“The thing that you learn about silicon is that it has to be extremely pure,” he said. “These perovskites are a million times more defective, and yet it is still a very good solar cell. It just has no right to work this well.”

However, there is a catch: Perovskites only perform well in tiny, centimeter-scale amounts. 

Since 2021, Marohn has been working with Lara Estroff, the Herbert Fisk Johnson Professor of Industrial Chemistry in Cornell Engineering, and collaborators at Johns Hopkins University, the University of Michigan and several national labs on a Department of Energy-funded project that uses artificial intelligence to reveal chemical formulations and processing conditions that reliably grow durable perovskites with optimal performance.

“Our team on the perovskites intentionally jumped into a less crowded part of the field, and really asked this question about understanding how the perovskite materials grow as a function of their processing conditions, and how the processing conditions, which we know impact the quality of the film, then impact the electronic properties and ultimately the device performance,” said Estroff, who is also director of Materials Science and Engineering. “Getting the perovskites commercialized is going to rely on understanding how these materials grow.”

‘In it for the crystals’

While Estroff’s team focused on using in situ techniques for watching the crystals grow in real time and understanding how the chemistry of the solutions impacted the growth mechanisms, Marohn’s group has been developing tools to characterize the perovskites’ electronic properties, specifically a long-standing puzzle about how electron charges recombine in organic materials.

In a solar cell, sunlight knocks electrons loose from the material’s atoms, and those free-moving electrons become the electronic current that is extracted from the cell. The longer it takes for the electrons to return to the atoms, the better the cell’s performance.

Marohn’s team needed a way to capture the movement of electrons and ions, and measure the charge density and conductivity, so the researchers essentially built microscopes that can take time-lapse photos at nanometer-length scales in one billionth of a second.

Compounding these difficulties is the fact that perovskites are so poorly behaved, with so many unknown variables controlling the crystal nucleation and growth, the researchers have struggled to duplicate one lab’s perovskite recipe in a another’s lab.

“The measurements are hard. The materials are irreproducible. So it’s been, frankly, quite a fight to get to where we can make materials and have them be the same every day, every week,” Marohn said. “We had really promising preliminary data. But we need to get a measurement that I can take to the bank. Now we’re finally at that point.”

Despite the technological challenges, perovskites remain attractive for their solution processability, their flexibility and for the types of substrates they can be deposited onto, Estroff said, but there is a caveat: “If we can’t come up with a way of scaling up the reproducible fabrication, it’s never going to make it in the market.”

When that could happen for an all-perovskite solar cell that can be produced on an industrial scale is still an open question.

“The number, in general, that the material science community bandies around is that it’s roughly a 20-year horizon to take a new material from discovery to something that you can buy,” said Marohn, noting that perovskites have already begun to show up commercially as a top layer in tandem solar cells. “In that transition from lab curiosity to the market, they have somewhat beat the 20-year rule.”

As exciting as the commercial potential is, that isn’t what keeps Estroff coming back to perovskites.

“It’s a very hard field to keep up with. I’m doing this because I love understanding how crystals grow,” she said. “I’m glad that I’m doing work that has potential real-world implications and impact. But, yeah, I’m in it for the crystals.”