Combined Machine Learning, Computational and Experimental Analysis of the Iridium(III) Complexes with Red to Near-IR Emission

Various coordination complexes have been the subject of experimental or theoretical studies in recent decades because of their fascinating photophysical properties. In this work a combined experimental and computational approach applied to investigate the optical properties of monocationic Ir(III) complexes. In result, an interpretative machine learning-based Quantitative Structure-Activity Relationship (QSAR) model was successfully developed, which can reliably predict the emission wavelength of the Ir(III) complexes and provides foundations for theoretical evaluation of the optical properties of Ir(III) complexes. A hypothesis was proposed to mechanistically explain the differences in emission wavelengths between structurally different individual Ir(III) complexes. To the best of our knowledge, this is the first attempt to develop predictive machine learning (QSAR) model for the optical properties of Ir(III) complexes. The efficacy of the developed model was demonstrated by high R2 values for the training and test sets of 0.84 and 0.87, respectively, and by performing the validation using y-scrambling techniques. A notable relationship between the N-N distance in the diimine ligands of the Ir(III) complexes and emission wavelengths was revealed. This combined experimental and computational approach shows a great potential for rational design of new Ir(III) complexes with desired optical properties. Moreover, the developed methodology could be extended to other octahedral transition-metal complexes.