NERSC Resources Power Advances in Solar Cell Efficiency
Improvements in a cheaper form of solar energy could one day change the dynamics of the energy market
July 12, 2021
The modern world was built with fossil fuels, and the modern economy still mostly runs on them. Now that same economy is tasked with the massive undertaking of swapping its power sources over the next generation.
However, the goal is not simply to eventually replace legacy fuels, but to do so at a speed that will help mitigate the effects of climate change. Getting to the finish line quickly enough will require more than steadily building up sustainable energy production; we need to find multipliers that change the equation.
New research out of the Materials Department at the University of California - Santa Barbara (UCSB), leveraging supercomputing resources at the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory, could help solar energy pick up the pace by improving the efficiency of a new type of solar cell. The findings were recently published in Nature Materials in a paper titled “Minimizing hydrogen vacancies to enable highly efficient hybrid perovskites.”
A New Type of Solar Cell
The solar panels you see on rooftops use silicon as their key ingredient. The steady improvement in these cells has made them cost competitive with fossil fuel sources in recent years. Additional advances in their efficiency will make them even more attractive.
New materials, “hybrid perovskites,” are poised to give solar-cell efficiency a boost. These hybrid perovskites are better than silicon at capturing some parts of the solar spectrum; combining them with silicon, in so-called “tandem cells,” results in a more efficient solar cell.
“Perovskite,” named after Russian mineralogist Lev Perovski, refers to a crystalline material with a particular structure. Within this category are “hybrid perovskites” that combine organic molecules (which contain hydrogen or carbon) with the inorganic lattice.
Twelve years ago, a particular hybrid perovskite — methylammonium lead iodide, or MAPbI3 — was tested as a solar cell, but its efficiency at turning the sun’s rays into usable power was initially quite low — around 4 percent, compared to 25 percent for a silicon cell. But scientists managed to continually improve these cells, mostly through focusing on the inorganic parts of the material. Over time, the efficiency of MAPbI3 improved to the point of being comparable with silicon cells.
A major advantage of hybrid perovskites is that they are relatively cheap to manufacture. Adding them to silicon solar cells boosts the power that the cell can produce, at relatively low additional cost. However, to make this cost-effective, the performance of the hybrid perovskite material needs to be really high. It became clear that this performance was still limited by the presence of imperfections in the material ― and identifying the source of these imperfections has been the subject of many investigations.
An Atomic Discovery with Big Implications
Professor Chris Van de Walle’s team at UCSB, conducting research led by postdoctoral researcher Xie Zhang, discovered a key difference between the high-performing and underperforming perovskite cells. Graduate students Mark Turiansky and Jimmy-Xuan Shen also contributed to this effort, which was funded by the U.S. Department of Energy (DOE) Office of Science Basic Energy Sciences program.
“Until now, the focus was on defects that occur in the inorganic part of the crystal,” Van de Walle explained. “Everybody had been assuming that the organic molecule was essentially inert.”
Through a computational modeling technique called density functional theory (DFT) and heavy computations run on NERSC’s Cori supercomputer, Van de Walle was able to spelunk in the microscopic world of molecules and examine their effect on the power conversion process. They discovered that defects related to the organic molecule can occur; in particular, missing hydrogen atoms (“vacancies”) can severely severely reduce the cell’s efficiency.
“DFT is a very powerful way to treat the behavior of electrons in a material quantum-mechanically,” said Van de Walle. “It makes the problem feasible; but it still requires a lot of CPU time.”
DFT is a method for modeling the structures of “many-body” systems, which have numerous interacting particles. The bulk of the material in the cell follows a regular crystalline structure and can be studied simply by examining one of its component parts. Studying defects, however, requires examining a much larger volume of the material.
Van de Walle’s team used a form of DFT that employs “hybrid functionals” (no relation to hybrid perovskites). This method provides a better description of the electronic structure of materials, plus more detailed insight into defect energies. The higher accuracy, combined with the task of hunting for defects, necessitates an unusually heavy computational load and NERSC supercomputing resources.
But ample processing power is only half the battle.
“Computers are extremely useful for doing accurate calculations to test a hypothesis,” said Van de Walle. “But the hypothesis still needs to be generated by our own brains.”
Now that the enervating effects of hydrogen defects have been discovered, scientists should be able to avoid these hydrogen “vacancies” in fabricating these cells. Future perovskite solar cells are likely to be more productive and consistent due to this research.
Fossil fuels created new possibilities in what human beings could do and how quickly they could do them. To replace those same fuels, while maintaining and improving the society they helped build, we will need new innovations that change what is possible in sustainable energy production. Research like Van de Walle’s, using resources like those at NERSC, can give us crucial new methods and tools to reach our clean energy goals.
NERSC is a DOE Office of Science user facility.
For more information, see this UC Santa Barbara news release.
About Computing Sciences at Berkeley Lab
High performance computing plays a critical role in scientific discovery. Researchers increasingly rely on advances in computer science, mathematics, computational science, data science, and large-scale computing and networking to increase our understanding of ourselves, our planet, and our universe. Berkeley Lab’s Computing Sciences Area researches, develops, and deploys new foundations, tools, and technologies to meet these needs and to advance research across a broad range of scientific disciplines.