Stable and Highly Emissive Infrared Yb‐Doped Perovskite Quantum Cutters Engineered by Machine Learning

Abstract Quantum cutting (QC) allows the conversion of high‐energy photons into lower‐energy photons, exhibiting great potential for infrared communications. Yb‐doped perovskite nanocrystals can achieve an efficient QC process with extremely high photoluminescence quantum yield (PLQY) thanks to the favorable Yb 3+ incorporation in the perovskite structure. However, conventionally used oleic acid–oleylamine‐based ligand pairs cause instability issues due to highly dynamic binding to surface states that have curbed their potential applications. Herein, zwitterionic type C3‐sulfobetaine 3‐(N,N‐Dimethylpalmitylammonio)propanesulfonate molecule is utilized to build a strong binding state on the nanocrystals’ surface through a new phosphine oxide synthesis route. Leveraging machine learning and Bayesian Optimization workflow to determine optimal synthesis conditions, near‐infrared PLQY above 190% is achieved. The high PLQY is well maintained after over three months of aging, under high‐flux continuous UV irradiation, and long continuous annealing. This is the first report of highly efficient and stable perovskite quantum cutters, which will drive the study of fundamental physics phenomena and near‐infrared quantum communications.