Researchers in chemistry, including a scholar from Oregon State University, have made significant progress towards advanced optical computing and memory through the identification of luminescent nanocrystals that can swiftly switch between light and darkness.
According to Artiom Skripka, an assistant professor in the OSU College of Science, “These remarkable switching and memory traits of the nanocrystals might eventually play a crucial role in optical computing. This method uses light particles, which travel faster than anything else in the universe, to quickly process and store data. Our discoveries could propel advancements in artificial intelligence and broader information technologies.”
The research, featured in Nature Photonics, was conducted by Skripka along with collaborators from Lawrence Berkeley National Laboratory, Columbia University, and the Autonomous University of Madrid. The focus was on a type of material called avalanching nanoparticles.
Nanomaterials are minuscule particles measuring between one billionth and one hundred billionths of a meter. Avalanching nanoparticles uniquely exhibit extreme non-linearity in their light emissions, meaning they can produce a substantial increase in light intensity with a minimal rise in the intensity of the laser used to excite them.
The research concentrated on nanocrystals made of potassium, chlorine, and lead, with neodymium added. Alone, the potassium lead chloride nanocrystals do not respond to light; instead, they act as hosts that allow the neodymium guest ions to manage light signals more effectively, making them valuable for optoelectronics, laser technology, and various optical uses.
Skripka explained, “Typically, luminescent materials emit light when stimulated by a laser and remain unlit when inactive. However, we were intrigued to discover that our nanocrystals exhibit unusual behavior; under specific conditions, they can alternate between being bright and dark under the same laser excitation wavelength and power.”
This phenomenon is known as intrinsic optical bistability.
“If the crystals start off dark, we need stronger laser power to activate them and observe their light emission. Once they begin emitting, they can maintain that emission at lower laser powers than initially required,” Skripka added. “It’s similar to pedaling a bike; once you get it moving, you don’t need to pedal hard to keep it going. Their luminescence can be switched on and off very sharply, like pushing a button.”
The capacity for low-power switching in these nanocrystals aligns with global initiatives aimed at lowering energy consumption related to the increasing demand from artificial intelligence, data centers, and electronic devices. Not only do AI functions require significant computational power, but they also encounter restrictions tied to existing hardware—issues that this new research could help overcome.
Skripka remarked, “By integrating photonic materials with intrinsic optical bistability, we could achieve quicker and more efficient data processors, thereby improving machine learning algorithms and data analysis. This could also lead to more effective light-based devices used in telecommunications, medical imaging, environmental monitoring, and connections for optical and quantum computers.”
Moreover, he noted that this study supports ongoing efforts to create powerful, versatile optical computers based on the interaction of light and matter at the nanoscale, highlighting the critical role of fundamental research in fostering innovation and economic progress.
While our findings are promising, Skripka concluded, “We need further research to tackle challenges like scalability and compatibility with existing technologies before our discovery can be applied practically.”
This research was supported by the U.S. Department of Energy, the National Science Foundation, and the Defense Advanced Research Projects Agency, with leadership from Bruce Cohen and Emory Chan at Lawrence Berkeley, P. James Schuck at Columbia University, and Daniel Jaque at the Autonomous University of Madrid.
