Rapid Data‐Efficient Optimization of Perovskite Nanocrystal Syntheses through Machine Learning Algorithm Fusion

Abstract

With the demand for renewable energy and efficient devices rapidly increasing, a need arises to find and optimize novel (nano)materials. With sheer limitless possibilities for material combinations and synthetic procedures, obtaining novel, highly functional materials has been a tedious trial and error process. Recently, machine learning has emerged as a powerful tool to help optimize syntheses; however, most approaches require a substantial amount of input data, limiting their pertinence. Here, three well‐known machine‐learning models are merged with Bayesian optimization into one to optimize the synthesis of CsPbBr3 nanoplatelets with limited data demand. The algorithm can accurately predict the photoluminescence emission maxima of nanoplatelet dispersions using only the three precursor ratios as input parameters. This allows us to fabricate previously unobtainable seven and eight monolayer‐thick nanoplatelets. Moreover, the algorithm dramatically improves the homogeneity of 2–6‐monolayer‐thick nanoplatelet dispersions, as evidenced by narrower and more symmetric photoluminescence spectra. Decisively, only 200 total syntheses are required to achieve this vast improvement, highlighting how rapidly material properties can be optimized. The algorithm is highly versatile and can incorporate additional synthetic parameters. Accordingly, it is readily applicable to other less‐explored nanocrystal syntheses and can help rapidly identify and improve exciting compositions’ quality.

Optical anisotropy of CsPbBr3 perovskite nanoplatelets

“…1 . Samples of NPLs were isolated with absorption spectra (shown in solid lines) which were previously identified in literature as 2 monolayer (ML) to 5 ML in thickness [ 19 , 53 ]. 1 ML samples, which can be prepared synthetically, evidenced by a sharp absorption feature near 400 nm, were unstable at room temperature for the time necessary to perform other data collection.…”

“…Nanoplates (NPLs) of CsPbBr 3 have a quantum well electronic structure in which the thickness dictates the band gap. These materials have been prepared with tunable quantum confinement by preparation of ensembles of 2-8 monolayers [15][16][17][18][19]. They are reported to have large oscillator strength [20] and narrow, polarized emission [21][22][23].…”

Introduction to new horizons in materials for energy conversion, optics and electronics

“…[3][4][5] Meanwhile, theoretical calculations, such as machine learning methods, have been widely applied to solve complex problems. 6,7 This collection focuses on energy conversion, optics, and electronic applications of (nano) materials and provides an overview of the most impactful experimental approaches and theoretical methods for energy conversion and storage, intending to connect different communities and identify common challenges in the field.…”

Utilizing Data-Driven Optimization to Automate the Parametrization of Kinetic Monte Carlo Models

“…However, the combination of BO and GPs in terms of a strong, data-driven surrogate model is yet relatively unexplored. It has been proven to be powerful in the field of compositional engineering, , for high throughput laboratories, for the optimization of quantum cascade detectors, and in kMC models for structural prediction …”

“…38 However, the combination of BO and GPs in terms of a strong, data-driven surrogate model is yet relatively unexplored. It has been proven to be powerful in the field of compositional engineering, 39,40 for high throughput laboratories, 41 for the optimization of quantum cascade detectors, 42 and in kMC models for structural prediction. 43 In this work, we present an innovative data-driven optimization pipeline to enable an automated parametrization of kinetic Monte Carlo models.…”