Researchers from North Carolina State University and the University of Buffalo created and tested a “self-driving lab” that employs artificial intelligence (AI) and fluidic systems to increase our understanding of metal halide perovskite (MHP) nanocrystals. This self-driving lab may also be utilized to study a wide range of other semiconductors and metallic nanoparticles.
Milad Abolhasani, an associate professor of chemical and biomolecular engineering at NC State and corresponding author of a publication on the study, says “We’ve constructed a self-driving laboratory that can be utilized to enhance both basic nanoscience and practical engineering.”
The researchers concentrated on all-inorganic metal halide perovskite (MHP) nanocrystals, cesium lead halide (CsPbX3, X=Cl, Br), for their proof-of-concept presentations. MHP nanocrystals are a new class of semiconductor materials that have the potential to be used in printed photonic devices and energy technologies due to their solution-processability and unique size- and composition-tunable features. MHP nanocrystals, for example, are highly efficient optically active materials that are being considered for application in next-generation LEDs. And, because they may be produced utilizing solution processing, they have the potential to be produced at a low cost.
Solution-processed materials, which include high-value materials like quantum dots, metal/metal oxide nanoparticles, and metal-organic frameworks, are created with liquid chemical precursors.
MHP nanocrystals, on the other hand, have yet to be used in industry.
“Part of this is due to the fact that we’re still learning how to synthesis these nanocrystals in order to create all of the features associated with MHPs,” adds Abolhasani. “And, in part, because synthesizing them necessitates a level of accuracy that makes large-scale manufacture uneconomical. Both of these difficulties are addressed in our work here.”
The new technology builds on Abolhasani’s lab’s Artificial Chemist 2.0 idea, which was announced in 2020. Artificial Chemist 2.0 is a fully autonomous system that performs multi-step chemical synthesis and analysis using AI and automated robotic technologies. In practice, the system focused on tailoring the bandgap of MHP quantum dots, allowing consumers to move from requesting a bespoke quantum dot to completing the necessary R&D and starting production in under an hour.
“Our novel self-driving lab technology can dope MHP nanocrystals on demand, adding manganese atoms to the crystalline lattice of the nanocrystals,” explains Abolhasani.
Doping the material with various quantities of manganese alters the nanocrystals’ optical and electrical capabilities while also introducing magnetic features. Doping MHP nanocrystals with manganese, for example, can alter the wavelength of light emitted by the material.
“We now have even more control over the characteristics of the MHP nanocrystals,” explains Abolhasani. “In other words, the universe of possible hues that MHP nanocrystals may create has grown. It’s not just about the hue. It has a substantially wider spectrum of electrical and magnetic characteristics than other materials.”
The new self-driving lab technology also allows researchers to learn how to manufacture MHP nanocrystals to get the appropriate mix of attributes faster and more efficiently. “ https://www.youtube.com/watch?v=2BflpW6R4HI ” is a video of how the new technology works.
“Let’s assume you want to learn more about how manganese doping and bandgap tuning influence a certain class of MHP nanocrystals, such CsPbX3,” explains Abolhasani. “If you were to control for every potential variable in each experiment, there are around 160 billion different experiments. Using standard procedures, learning how those two processes—manganese doping and bandgap tuning—affect the characteristics of the cesium lead halide nanocrystals would take hundreds or thousands of trials.”
The new system, on the other hand, accomplishes everything on its own. Its AI algorithm, in particular, picks and performs its own trials. It continues to execute experiments until it knows which processes regulate the MHP’s many qualities, based on the findings of each completed experiment.
“In a practical demonstration,” Abolhasani explains, “we discovered that the system was able to gain a full grasp of how these processes modify the characteristics of cesium lead halide nanocrystals in only 60 tests.” “In other words, instead of months, we can receive the knowledge we need to develop a material in hours.”
While the paper’s research focuses on MHP nanocrystals, the autonomous system might also be used to describe other solution-processed nanomaterials, such as a wide range of metallic and semiconductor nanomaterials.
“We’re thrilled about how this technology can expand our understanding of how to regulate the characteristics of these materials,” adds Abolhasani, who also points out that the system may be used for continuous production. “So you can utilize the system to figure out the optimum way to make your desired nanocrystals, and then configure the system to produce material nonstop—and with great precision.”
“We’ve developed cutting-edge technology. We’re now searching for collaborators to help us adapt this technology to particular industry concerns.”