Power/Performance Bits: Aug. 10

Flexible electrodes for thin films
Researchers from the University of Queensland and ARC Centre of Excellence in Exciton Science (University of Melbourne) developed a material for flexible, recyclable, transparent electrodes that could be used in things like solar panels, touchscreens, and smart windows.

Eser Akinoglu of the ARC Centre of Excellence in Exciton Science said, “The performance of the material is excellent, the transmission of above 90% and high electrical conductivity rivals the ITO benchmark.” ITO, indium tin oxide, is the standard material used in thin-film displays.

The dielectric/metal/dielectric (D/M/D) nanomesh electrodes are produced through nanosphere lithography, a deposition method which involves evaporating the desired combination of materials into a nanoscale pattern.

“In principle, you should be able to integrate this technology into industrial roll-to-roll printing,” Akinoglu noted.

The nanomesh electrodes produced through this approach showed precisely controlled perforation size, wire width and uniform hole distribution, yielding high transmittance, low sheet resistance (which minimises loss of voltage) and flexural endurance.

“We offered a strategy to make the shadow area of the metallic nanomesh highly transparent, by integrating D/M/D structures to the nanomesh system. The nanomesh transparent films with D/M/D layered structure have not been studied before. The simple and cost-effective nanosphere lithography technique can be applied to fabricate diverse layered nanomesh materials,” said Tengfei Qiu of the University of Queensland.

Additionally, for certain applications, the electrodes demonstrated recyclability, said Akinoglu. “It means that if you make a device like an electrochromic window, which may deteriorate in functionality after its life-span, you can take it apart, flush rinse the electrodes, and reuse them for another device.”

Next work for the researchers involves expanding to a larger, commercially viable capacity. “You want to get the transparency higher, you want to get sheet resistance lower and you want to get the endurance for mechanical stress and flexibility higher,” Akinoglu said. “And you want to be able to fabricate it on a large-scale area, at a low cost.”

Printing perovskite LEDs
Scientists from Helmholtz-Zentrum Berlin (HZB) and Humboldt-Universität zu Berlin used inkjet printing to produce LEDs from metal halide perovskite semiconductors.

Most frequently known for its use in solar cells, hybrid perovskites show promise in other areas, too. “They can be used to manufacture all kinds of microelectronic components by modifying their composition,” said Prof. Emil List-Kratochvil, head of a Joint Research Group at HZB and Humboldt-Universität. “They can be produced from a liquid solution, so you can build the desired component one layer at a time directly on the substrate.”

Perovskites have been used to make printed solar cells before. Similarly, LEDs can be printed from organic semiconductors, but with lower luminosity.

“Until now, it has not been possible to produce these kinds of semiconductor layers with sufficient quality from a liquid solution,” said List-Kratochvil. “The challenge was how to cause the salt-like precursor that we printed onto the substrate to crystallise quickly and evenly by using some sort of an attractant or catalyst.” The team chose a seed crystal for this purpose: a salt crystal that attaches itself to the substrate and triggers formation of a gridwork for the subsequent perovskite layers.

Using this method, the researchers created printed LEDs that possess higher luminosity and better electrical properties than could be previously achieved using additive manufacturing processes.

Beyond LEDs, the researchers believe hybrid perovskites will be the future for micro- and optoelectronics. “The advantages offered by a single universally applicable class of materials and a single cost-effective and simple process for manufacturing any kind of component are striking,” said List-Kratochvil.

AI for WiFi
Researchers from the Universitat Pompeu Fabra propose a way to use machine learning to improve WiFi networks in areas where a network has a large number of access points, such as corporate campuses and universities.

“In this study, we look at how stations (PCs, tablets, mobile phones, etc.) may themselves decide dynamically which of the different access points available in their coverage area is offering the best service for their needs using reinforcement learning techniques,” said Marc Carrascosa and Boris Bellalta, researchers with the Wireless Networking Research Group at Universitat Pompeu Fabra.

Rather than trying to get a top-down view of the network and connected devices, in the proposed system, each station is independent and takes decisions dynamically based on the quality of service offered by the WiFi network over time.

“For this learning, as a basis we use an algorithm called ε-greedy, which alternates between choosing access points at random to obtain information (exploring), and choosing the best access points used based on this accumulated information (exploiting),” said Carrascosa and Bellalta. “Thus, the more information, the better decisions we take, considering, however, that there is a compromise between the time a station can devote to learning and the time it disposes of to use what it has learned successfully.”

To limit learning time, the researchers developed the ε-sticky algorithm, which includes the concept of emotional attachment. Once a suitable access point has been identified, even if it ceases to be later, the station does not immediately discard it to look for another new one again in the hope that in the future it might give the same good service.

The team said that their work could reduce service disruptions and network instability, and not just for the users of the algorithm. Only a few stations would need to implement it for the whole network to benefit.