the feed MANY MINDED · THE BRIEF
GOODS · forward · impact 3/5 · 2026-07-18

AI-blended solvents lift lab QLED efficiency and stretch lifetime 40-fold

A machine-learning model designed solvent blends that roughly doubled QLED efficiency and extended device lifetime more than 40 times in the lab.

A research team has built an AI platform that inversely designs solvent properties to achieve uniform quantum-dot packing during QLED fabrication. In lab-made devices, the optimized process roughly doubled efficiency and extended operational lifetime by more than 40 times compared with conventional single-solvent devices. The work, led by Professor Jeonghun Kwak of Seoul National University and Professor Jaehoon Lim of Sungkyunkwan University, was published online July 15 in Reports on Progress in Physics, with Beomsoo Chun as lead author.

The method started by fabricating quantum-dot films with five representative solvents and measuring surface uniformity using atomic force microscopy. A machine learning model was trained on solvent properties such as vapor pressure, viscosity, density, and dielectric constant alongside the resulting film morphology. Because no single solvent had all the optimal properties, the team blended multiple solvents to match the model's predictions, then verified the improved films.

The abundance angle sits in manufacturing. Display quality has long been limited by how evenly quantum dots deposit; better packing means brighter, more efficient screens that last far longer. Longer-lived, more efficient panels mean fewer replacements and lower energy draw, which over time can push down the cost of high-quality displays and reduce their lifetime footprint.

The honest limits: these are laboratory-fabricated devices, not production lines, and the leap from bench to commercial yield is where many display advances stall. The researchers describe extending the approach to OLEDs and solar cells as an expectation rather than a demonstrated result. The announcement was provided by Seoul National University.

Source: Phys.org