Enhanced Aggregation-Induced Emission Activity of Metal–Organic Frameworks by Using Machine Learning Technology

Abstract

Metal−organic frameworks (MOFs) with aggregation-induced emission (AIE) activity show a high emission intensity, high sensitivity, and high resolution in biological imaging and identification technologies. However, their AIE activity is controlled by various Eu precursors' components and synthesis process parameters, and traditional research methods are hard to deal with these complex multiple parameter systems. In this work, we utilize two machine learning technologies to optimize the synthesis process parameters of Eu-MOFs and analyze their synthesis mechanism. First, we choose gradient boosting decision tree (GBDT) regression as the best fitting model. Second, on the basis of the SHapley Additive exPlanation (SHAP) calculation method with the PL/UV intensity ratio regarded as the evaluation standard, we demonstrate that the Eu-precursor concentration (1.91 × 10 7 ) and synthesis time (1.73 × 10 7 ) dominate in the synthesis systems. Meanwhile, these two parameters show synergic and antagonistic effects on the PL/UV intensity ratio, respectively. Finally, we employ a greedy random walk method to work out that "142-0.83-4.1" should be the best optimization process parameters, and the corresponding sample shows a high photoluminescence quantum yield (PLQY) with a value of 7.65% in the solid state. More importantly, the screen-printed pattern exhibits bright red fluorescence under UV light.

Insights into Isotherm Step and Pore-Filling Mechanisms for Water Adsorption in Covalent Organic Frameworks

“…However, this can be made amenable with computational tools. 18 Recently, Zhang et al 19 developed a machine learning (ML) model that could identify top-performing MOFs as AWHs from the CoRE MOF database, trained on 344 experimental data of 285 MOFs. As mentioned earlier, the location of the isotherm step is a crucial factor in identifying COFs as AWHs.…”

Classifying and Predicting the Thermal Expansion Properties of Metal–Organic Frameworks: A Data-Driven Approach

“…5−7 In the past few years, machine learning (ML) methods have facilitated the analysis of large quantities of data generated for MOFs and shortlisted top-performing MOFs for various applications including CO 2 capture, 8 gas sensing, 9 methane storage, 10 paraffin/olefin separation, 11 and water harvesting. 12 Recently, the crystal graph convolution neural network (CGCNN) 13 technique has been exploited for MOFs by extracting atom-based graph features. 14,15 Furthermore, MOFTransformer 16 and Moformer 17 were developed using large language models, which leverage upon pretraining and fine-tuning procedures to produce significantly more accurate predictions.…”

Leveraging Machine Learning for Metal–Organic Frameworks: A Perspective

“…The recent ML studies by Jiang and co-workers demonstrated that the incorporation of chemical descriptors including atom types and densities in MOFs would improve the ML prediction accuracy for C 3 H 8 /C 3 H 6 separation and water adsorption. 17,42 A commonly used descriptor of MOF chemistry is the revised autocorrelation functions (RACs). 43 The RACs possess two key advantages: (1) decomposing a MOF into subgraphs of respective metal clusters, organic linkers, and functional groups, hence capturing subtle chemical details and (2) considering MOF hierarchy that couples atomic-level properties (i.e., atom 11 thermal stability 30 and synthesizability, 31 colors, 44 and oxidation states of metal centers in MOFs).…”

“…17 Starting from experimentally measured water adsorption isotherms in 285 MOFs, Zhang et al developed ML models for atmospheric water harvesting; the transferability of the ML models was validated by out-of-sample predictions in newly reported MOFs, and finally the ML models were applied to screen ∼8,000 CoRE MOFs and identify top-performing candidates. 42 For gas diffusion in MOFs, Krokidas et al benchmarked a ML model to predict the diffusivities of 12 different gases in 72 variants of zeoliticimidazolate frameworks (ZIFs). 69 Daglar and Keskin constructed ML models to predict gas diffusivities and permeabilities in MOF membranes and MOF/polymer mixed-matrix membranes.…”