Accelerated room temperature synthesis of desired cesium lead halide perovskite nanocrystals via automated microfluidic meta learner

Abstract

Cesium lead halide perovskite nanocrystals (CLHP NCs) have demonstrated their unique optical properties in next-generation lighting and display devices through controlling the size, shape, and composition. However, achieving their target-specific synthesis with numerous reaction parameters via traditional methods requires long and tedious optimization. Herein, a fully automated microfluidic meta learner (AMML) coupled with coiled flow inverter reactors (CFIR) is developed for facilely synthesizing desired CLHP NCs. CFIR offers a strong support in reagent mixing for AMML, producing high-quality NCs with negligible by-products. By concatenating the model-agnostic meta-learning and feedback control, AMML provides quick and accurate locating of target emission CsPb(BrxI1-x)3 NCs in the wide range from 519 nm to 636 nm based on 3 data points, even in the no-prior-knowledge part. AMML exhibits excellent screening ability (25 experiments in 110 min) of reaction parameters for controllable synthesis of CsPbBr3 NCs. Based on the optimal results, a white light-emitting diode is fabricated with excellent stability and a wide color gamut of 121 % NTSC. AMML offers a novel paradigm for the large-scale synthesis of desired NCs and can be applied to other NCs for intelligent and sustainable discovery.