Fast switching and modulation of light are at the heart of modern data transmission, where, among other things, information is sent in the form of modulated light beams over fiber optic cables. Miniaturizing light modulators and integrating them into chips has been possible for several years, but the light sources themselves – light-emitting diodes (LEDs) or lasers – still pose challenges for engineers.

A research team from ETH Zurich led by Professor Lukas Novotny, together with colleagues from EMPA in Dübendorf and ICFO in Barcelona, ​​have discovered a new mechanism by which small but efficient light sources can be produced in the future. The results of their research were recently published in a journal. Nature Supplies.

An unexpected venture

“To achieve this, we had to try the unexpected first,” Novotny says. For several years he and his colleagues worked on miniature light sources based on the tunnel effect. Electrons can tunnel according to the rules of quantum mechanics between two electrodes (in this case gold and graphene) separated by an insulating material. Under certain conditions, that is, if the tunneling process is inelastic, i.e. the energy of the electrons is not conserved, light can occur.

“Unfortunately, these light sources are quite low in power because the radiation emission is so inefficient,” explains postdoctoral Sotirios Papadopoulos. This emission problem is well known in other areas of the art. In cell phones, for example, the chips that produce the microwaves necessary for transmission are only a few millimeters in size.

In contrast, microwaves themselves have a wavelength of about 20 centimeters, making them a hundred times larger than a chip. To overcome this size difference you need an antenna (in modern phones it’s actually no longer visible from the outside). Similarly, in the experiments of the Zurich researchers, the wavelength of the light is much longer than the light source.

Semiconductor outside the tunnel connection

“So you might think we’re consciously looking for an antenna solution, but we’re not,” says Papadopoulos. Like other groups before them, the researchers discovered one-atom-thick layers of semiconductor material, such as tungsten disulfide, sandwiched between tunnel junction electrodes to generate light.

In principle, it can be assumed that the optimum location should be between the two electrodes, perhaps a little closer to one than the other. Instead, the researchers tried something completely different, placing a semiconductor on top of the graphene electrode — completely outside the tunnel junction.

Amazing antenna performance

Surprisingly, this seemingly illogical position worked very well. The researchers found the reason for this by changing the voltage applied to the tunnel junction and measuring the current flowing through it. This measurement showed a net resonance corresponding to the so-called exciton resonance of the semiconductor material.

Excitons consist of a positively charged hole corresponding to the missing electron and an electron bound to the hole. For example, they can be excited by the glow of light. The exciton resonance was a clear indication that the semiconductor was not directly excited by the charge carriers—after all, electrons were not flowing through it—but instead absorbing the energy created in the tunnel junction and then re-emitting it. In other words, it was acting like an antenna.

Application in nanoscale light sources

“So far, this antenna is not very good, because so-called dark excitons are created inside the semiconductor, which means very little light is emitted,” Novotny admits. “Developing it will be our task in the near future.” If researchers can make semiconductor light emission more efficient, it will be possible to create light sources that are only a few nanometers in size and therefore a thousand times smaller than the wavelength of the light they produce.

Since electrons do not flow through a semiconductor antenna, there are also no undesirable effects that usually occur at the boundaries that can reduce efficiency. “In any case, we opened the door to new applications,” says Novotny. His unexpected attempt clearly paid off. Source