A team at the Massachusetts Institute of Technology claims that it has developed a new and faster method of evaluating the efficiency of new photovoltaic materials to avoid long-term, expensive and time-consuming laboratory sample test evaluations.
The process, scientists say, is based on a series of tools that rely on a series of relatively simple laboratory tests, combined with computer-modeled material physics properties and additional modeling of statistical methods based on Bayesian inference, Each new measurement changes the estimate of each parameter.
The system consists of three main steps: making a simple test fixture; measuring its current output at different levels of illumination and at different voltages; and accurately quantifying these changes under conditions that the team believes are all used to improve statistics model.
Tonio, research coordinator, said: 'After we got a lot of current-voltage measurements at different temperatures and light intensities, we also needed to find out if the combination of material and interface variables was consistent with our measurements.
He said: "Representing each parameter as a probability distribution allows us to interpret the uncertainty of the experiment, while also enabling us to determine which parameters are coherently covariant.
Massachusetts Institute of Technology scientists believe that the new process may shorten the development of new photovoltaic materials, but also accelerate the development of many other different materials, reducing the time to 20 to 3 years or 5 years.
The researchers said: 'We look forward to helping us increase the development of new materials more than five times through the combination of high-throughput computing, automation and machine learning.