DART: Unlocking Coordination Chemistry Beyond the Cambridge Structural Database

Computational high-throughput screening has become an essential tool in modern research, accelerating the discovery of novel compounds with tailored chemical reactivity and electronic properties. Metal complexes, with their versatile combinations of metal centers and ligands, offer vast potential for tuning their properties across a wide range of applications. This flexibility can be harnessed by computational algorithms that use ligands and metal centers as modular building blocks to create new molecules for exploration. However, there remains a need for a unified platform that integrates ligand databases, filters, and assembly algorithms into a single user-friendly framework, but also incorporates customizable ligand filters to precisely target novel metal complexes for specific applications. Here, we introduce DART (Directed Assembly of Random Transition metal complexes), an intuitive, end-to-end computational platform for automating the targeted assembly of monometallic complexes. Central to DART is the MetaLig database, containing 41,018 unique ligands extracted from the Cambridge Structural Database, each with automatically assigned formal charges. DART can generate up to 500 billion unique neutral molecular complexes in octahedral and square planar 3D structures by selecting ligands from the MetaLig database. Its assembler module enables the exploration of user-defined chemical spaces through customizable ligand filters, allowing DART to overcome the limitations of empirical databases, which are shaped by historical research trends and restricted to crystallizable structures, thereby enabling the exploration of novel chemistries. To demonstrate DART’s potential, we present a case study modeling intermediates in the oxidative addition step of transition metal-catalyzed C–C cross-coupling reactions. Within minutes, we assembled 620 square planar Pd and Ni complexes with P,N-type ligands. Subsequent density functional theory calculations revealed a broad spectrum of electronic properties for these intermediates, showcasing DART’s ability to facilitate targeted molecular design for complex research challenges.