Using computers and data mining analysis, research teams of Argonne and Cambridge were able to curb their research and result time for new light-absorbing organic dye molecules for solar power generation to a matter of few months, which otherwise would have taken them years to do with old manual curation of databases. (Photo Credit: Dave Weaver/Shutterstock/Argonne National Laboratory)
- Researchers from the US and UK have been able to identify 5 high-performing, low-cost dye materials from a list of 10,000 potential candidates thanks to new age computing and data mining approaches
- Out of 10,000 probable dye candidates, the team managed to bring down the list to 3,000 by eliminating organometallic and organic molecules
- Screening the dyes for carboxylic acid components, and with the help of Theta supercomputer, the team further reduced the list to 300 candidates, and then to 30
- Final 5 were discovered using computationally intensive density functional theory calculations on the supercomputer
A team of scientists from the US and the UK employ supercomputing with data science and experimental methods to identify promising materials for dye-sensitized solar cells (DSSC). These cells can be manufactured with low-cost, scalable techniques, they say, to be able to achieve competitive performance to price ratios.
From a pool of nearly 10,000 candidates for fabrication and device testing, the scientists of the US Department of Energy’s (DOE) Argonne National Laboratory (ANL) and the University of Cambridge in England identified 5 high-performing, low-cost dye materials.
The team led by Argonne Materials Scientist Jacqueline Cole employed data mining tools to eliminate organometallic and organic molecules and brought down the list from 10,000 to 3,000 potential dye candidates. The second step involved screening the dyes that contained carboxylic acid components, and finally using the Theta supercomputer at the Argonne Leadership Computing Facility (ALCF). With Theta, they conducted electronic structure calculations on the remaining candidates to determine the molecular dipole moment or degree of polarity of each individual dye.
From the list of 300 dyes, the researchers examined optical absorption spectra to bring it further down to roughly 30 dyes. Computationally intensive density functional theory (DFT) calculations on the supercomputer helped reach the final stage with 5 dyes for which Cole reached out to original dye developers to send new sample dye to be investigated.
The researchers then discovered that once these dyes are embedded into a PV device, these achieved power conversion efficiencies roughly equal to that of the industrial standard organometallic dye.
“The advantage of this process is that it takes away the old manual curation of databases, which involves many years’ worth of work, and reduces it to a matter of a few months and, ultimately, a few days,” said Cole.
Advanced Energy Materials published a paper based on this study in its February 1, 2019 issue.