Discovery of Dual Ion-Electron Conductivity of Metal-Organic Frameworks via Machine Learning-Guided Experimentation.

Identifying conductive metal-organic frameworks (MOFs) with a coupled ion-electron behavior from a vast array of existing MOFs offers a cost-effective strategy to tap into their potential in energy storage applications. This study employs classification and regression machine learning (ML) to rapidly screen the CoREMOF database and experimental methodologies to validate ML predictions. This process revealed the structure-property relationships contributing to MOFs' bulk ion-electron conductivity. Among the 60 conductive compounds predicted, only two p-type conductive MOFs, [Cu(3)(μ(3)-OH) (μ(3)-C(4)H(2)N(2)O(2))(3)(H(3)O)]·2C(2)H(5)OH·4H(2)O (1) and NH(4)[Cu(3)(μ(3)-OH)(μ(3)-C(4)H(2)N(2)O(2))(3)]·8H(2)O or (2) (C(4)H(2)N(2)O = 1H-pyrazole-4-carboxylic acid), were validated for their coupled electron-ion behavior. MOFs utilize earth-abundant copper and pyrazoles as ligands, demonstrating significant potential following thorough electrochemical characterization. Further analysis confirmed the critical role of strong σ-donating, π-accepting, and redox-active ligands in promoting electron mobility. In-depth structural investigations revealed that the presence of the O-Cu-N chain significantly influences conductivity, outperforming MOFs with only Cu-N or Cu-O bonds. Additionally, this study highlights how higher ionic conductivity is correlated with the ion mobility through linkers in 1 or the presence of ammonium ions in 2. These structure-property relationships offer valuable insights for future research in using ML coupled with experimentation to design MOFs containing earth-abundant reagents for ion-electron conductivity without employing a host-guest MOF strategy.