MIT and Princeton University researchers have harnessed the process of singlet exciton fission as a way to produce two electrons from a single incoming photon of light in silicon solar cells. [Image: Courtesy of the researchers/MIT News Office]
Silicon photovoltaics (PVs) dominate the solar-energy landscape, with dramatic decreases in cost in recent years spurring equally dramatic gains in use. But these solar-cell workhorses face a built-in theoretical efficiently limit of around 29%. One reason is that, in silicon, every incident solar photon can produce no more than a single electron to contribute to the generated photocurrent.
Now, a research team from the Massachusetts Institute of Technology (MIT) and Princeton University, USA, has used some clever materials science and fabrication to coax silicon solar cells into kicking out two electrons when struck by a single high-energy photon (Nature, doi: 10.1038/s41586-019-1339-4). While the technique currently works only in the blue–green part of the spectrum, the researchers say it could boost the maximum theoretical efficiency of silicon solar cells from 29% to 35% out of the box. And additional work to refine the system could push efficiencies higher still.
From singlet to triplet
The key to this photonic slight-of-hand lies in a process called singlet excition fission. When a photon strikes a semiconductor material, it generates a bound electron–hole pair, or quasiparticle, called an exciton. In certain molecular semiconductors, excitons with a singlet spin state can divide into pairs of excitons with triplet spin states, with each triplet exciton having half the energy of the parent singlet exciton. That raises the potential of generating two excitons—and, thus, two photocurrent electrons—from a single photon.
Some of the MIT members of the research team were, in fact, able to realize that two-for-one potential in organic solar cells in work published six years ago (Science, doi: 10.1126/science.1232994). The problem is that the most common solar-cell material, inorganic silicon, doesn’t support singlet exciton fission, and is limited to one-electron-for-one-photon accounting.
A heterostructure approach
The MIT–Princeton team wanted to see if it could find a workaround to this limitation of silicon. One possible approach was to marry silicon with another material that could support singlet exciton fission, in a heterostructure. In such a scheme, an incident photon in the singlet-exciton-fission material would create two triplet excitons; those excitons could pass into the silicon as two electron–hole pairs, and be converted into two electrons for photocurrent, just as if they had been created by two separate photons.
A promising candidate for the singlet-exciton-fission material in such a setup was the molecular semiconductor tetracene. It turns out that tetracene, when struck by light in the blue–green part of the spectrum, generates singlet excitons that can split into pairs of triplet excitons with an energy almost ideally matched to silicon’s band gap. But efforts to create such tetracene–silicon hybrids didn’t produce the hoped-for photocurrent in the silicon. That’s probably, according to the researchers, because the abrupt transfer of energy at the boundary between the two materials led to immediate recombination of the charge carriers at the silicon surface, before electrical energy could be harvested.
Finding a go-between
As so often is the case, solving this boundary problem boiled down to finding the right mediator. The researchers reasoned that a thin interlayer of material at the tetracene–silicon boundary could passivate the junction, allowing excitons to tunnel through the interlayer and into the silicon layer and preventing immediate recombination of the electron–hole pairs.
To pull off the trick, the team needed a dielectric material that could be grown with angstrom-level thickness control, so that the thickness could be tuned to maximize the tunneling probability. They settled on a layer of hafnium oxynitride (HaOxNy), an insulating material that could be grown on the silicon surface via atomic layer deposition, allowing the requisite thickness control.
Smart efficiency increase
Putting all of these threads together, the team created silicon–HaOxNy–tetracene heterostructures with various thicknesses of the HaOxNy layer, and used photoluminescence measurements to determine the efficiency of the exciton energy transfer from the tetracene to the silicon. They found that an interlayer thickness of 8 angstroms constituted a sort of sweet spot for the system. At that interlayer thickness, the heterostructure was able to deliver a peak exciton yield in the silicon of as high as 133% from the tetracene—around 56% from the transfer of singlet excitons, and 76% from the transfer of triplet excitons.
Because the tetracene creates the right triplet excitons only in the blue–green region, and because only part of the energy is converted into triplet excitons, the maximum theoretical efficiency of the hybrid system is only 35%. That’s still a smart increase from the 29% maximum theoretical efficiency of silicon solar cells alone.
The researchers believe they can get to still higher efficiencies by tweaking the thickness of the silicon layer and through tighter control on materials, fabrication and processing. The ultimate goal is to get the most energy bang per photon from a single silicon cell. “Fundamentally,” corresponding author Marc Baldo of MIT noted in a press release accompanying the work, “we’re kind of turbocharging the silicon cell. We’re adding more current into the silicon, as opposed to making two cells.”