Solar cells are electrical devices that are most utilized as we slowly shift into a world powered by renewable energy. OPVs (Organic photovoltaics) are a class of solar cells that are based on a light-absorbing organic molecule which is combined with a semiconducting polymer.
OPVs are made out of cost efficient, lightweight materials, and benefit from good safety and easy production. However, their ability to convert light into electricity or their PCEs (power conversion efficiencies) are still too low for full-scale commercialization.
The PCE will depend on the organic and polymer layer. In the past, chemists have experimented with various combinations of these by trial-and-error which leads to a lot of wasted time and effort.
Now, a team of researchers from Osaka University have used the power of computers to automated this search process for well-matched solar materials. Soon, this could lead to more efficient devices. The researchers’ study was reported in The Journal of Physical Chemistry Letters.
“The choice of polymer affects several properties, like short-circuit current, that directly determine the PCE,” study first author Shinji Nagasawa said. “However, there’s no easy way to design polymers with improved properties. Traditional chemical knowledge isn’t enough. Instead, we used artificial intelligence to guide the design process.”
Informatics is capable of understanding lage, complex data sets by detecting statistical trends that cannot easily be found by human experts. The team has gathered data from 1,200 OPVs from around 500 studies. By using Random Forest machine learning, they have built a model that combines the band gap, molecular weight, and chemical structure of the previous OPVs, together with their PCE, to forecast the efficiency of potential new devices.
Source: Phys
Random Forest has found an improved correlation between the properties of the materials and their actual performance in OPVs. To make use of this, the model was used to automatically screen the prospective polymers for their theoretical PCE. The list of the top choices was then reduced down based on chemical intuition about what can be synthesized in practice.
With this strategy, the team was able to create a new, previously untested polymer. In the event, a practical OPV based on this first try has proved that is was less efficient than expected. However, this model has provided useful insights into the structure-property relationship. Its predictions can be improved by adding more data, like the polymers’ solubility in water, or their backbones’ regularity.
“Machine learning could hugely accelerate solar cell development, since it instantaneously predicts results that would take months in the lab,” co-author Akinori Saeki says. “It’s not a straightforward replacement for the human factor — but it could provide crucial support when molecular designers have to choose which pathways to explore.”
Article Source:
:fill(transparent,1)/172701220-56a59ec73df78cf77289194f.jpg)
