Reliable predictions of unusual molecules

Quantum chemistry can today be employed to invent new molecules and investigate their properties and chemical bonding. However, the predicted species must be viable in order to be synthesized by experimentalists. In this perspective article we describe the technology of reliable theoretical predictions and show how understanding of chemical bonding in studied chemical systems could help to design new molecular structures. We also provide a short overview of successfully predicted and already produced (in some cases) planar hypercoordinate species to demonstrate that the consistent theoretical prediction of viable molecules with unusual structures and properties is now a reliable tool for exploring new, yet unknown molecules, clusters, nanomaterials and solids.