Scientists trying to build a “perfect” solar cell accidentally discovered that its flaws were the key to making it better
AI-madeFor decades, the guiding principle of solar cell research has been simple: build a cleaner, more perfect interface between materials, and efficiency will follow. Remove every atomic flaw, and the energy flows better.
Then a team at the University of Paderborn ran simulations on one of Europe’s most powerful supercomputers—tracking hundreds of atoms at femtosecond intervals—and the data told a different story. What they found at the High-Performance Computing Center Stuttgart quietly contradicted a core assumption the field had been building on for years.
Why solar cells still waste so much energy
Modern solar panels convert only about 22% of incoming sunlight into electricity on average. That number has improved steadily, but a hard physical ceiling keeps pushing back.
“The current direction of the research was going toward designing ever-more-perfect interfaces and removing defects at all costs,” he said.
The problem lies in how silicon handles photons. High-energy photons – violet light, for example – carry around 3 electron volts of energy. Silicon can only convert roughly 1.1 eV into electricity. The rest escapes as heat, and that’s not just wasted energy; the heat also degrades the cell over time, shortening its useful life.
The stakes are concrete. Germany generated less than 1% of its electricity from solar in 2000, and by 2025 that share had climbed to around 15%, driven by subsidies and falling hardware costs. As energy markets grow more volatile, the pressure to squeeze more out of every photon keeps building.
Enter tetracene: a molecule-thin shortcut around silicon’s limits
One promising workaround involves layering a different material on top of silicon. Researchers at the University of Paderborn, working with collaborators at Helmholtz Zentrum Berlin, have been studying tetracene — an organic semiconductor — as a potential top layer for solar cells.
Tetracene does something silicon can’t: it performs singlet fission. When a high-energy photon hits tetracene, the resulting exciton — an excited electron paired with the “hole” it leaves behind — splits into two lower-energy excitons instead of dumping excess energy as heat. Those two excitons can then transfer into the silicon layer beneath, where most of their energy gets converted into electricity.
In theory, this approach could boost efficiency by a factor of 1.4. The catch is engineering an interface between tetracene and silicon that actually lets those excitons through.
How the Hawk supercomputer made the invisible visible
To study that interface, the team turned to HLRS’s Hawk supercomputer in Stuttgart, using a method called ab initio molecular dynamics (AIMD), which simulates how individual atoms and their electrons interact over time.
The simulations tracked hundreds of atoms simultaneously, advancing in femtosecond increments—each step a millionth of a billionth of a second. That level of detail is computationally far beyond what standard hardware can handle. Researchers could watch excitons move — or fail to move — across the tetracene-silicon interface at the atomic scale. The original goal seemed straightforward: find the flaws, understand them, eliminate them.
The surprise hiding in the imperfections
That’s when the data contradicted the plan.
In atomic systems, atoms that aren’t fully bonded to their neighbors leave what are called “dangling bonds.” The field’s long-standing assumption is that dangling bonds are defects—sources of inefficiency that degrade electronic interfaces and should be minimized. The AIMD simulations said otherwise. Silicon dangling bonds at the tetracene interface didn’t impede exciton transfer. They actively helped it.
This isn’t entirely without precedent. As Prof. Uwe Gerstmann noted, semiconductor physics already uses strategic defects — donors and acceptors — to build diodes and transistors. The idea that a flaw could be functional isn’t new to physics. But it was new to this corner of solar research.
Lead author Dr. Marvin Krenz described the finding as a direct challenge to the field’s current direction. “The current direction of the research was going toward designing ever-more-perfect interfaces and removing defects at all costs,” he said. “Our paper might be interesting for the larger research community because it points out a different way to go.”
Designing solar cells that are ‘perfectly imperfect’
The team’s next step is to use future computing time to deliberately engineer interfaces with optimized dangling bonds—designing imperfection with precision rather than avoiding it entirely.
The goal isn’t a sudden leap in performance. Prof. Schmidt frames it as continuing a steady trajectory: solar cell architectures have seen roughly 1% average annual efficiency gains over recent decades, and strategic imperfections, he suggests, could help sustain or even accelerate that pace.
The results were published in Physical Review Letters in 2024. Whether other research groups move to test or build on the finding will determine how quickly this reframing spreads across the field. If deliberate dangling-bond engineering proves reproducible in physical prototypes—not just simulations—it could shift how the entire field thinks about interface design. The pursuit of perfection may turn out to need a few well-placed flaws.
Kelly is an experienced writer with 15 years of experience exploring the big stories that shape our world, from tech breakthroughs and space exploration to climate, energy, and the fascinating quirks of science. She has a talent for turning complex ideas into sharp, memorable insights that stay with readers long after they’ve finished reading.
