InChI for inorganic chemists: Organometallics and molecular inorganics
Gerd Blanke1, Sonja H. Herres-Pawlis2, Jonathan M. Goodman3, Ulrich Schatzschneider4, Andrey Yerin5, Richard Hartshorn6, Felix Bänsch7, Nauman Khan2, Djordje Baljozovic2, Jan Brammer2, Frank Lange2
1StructurePendium Technologies GmbH, Essen, Germany
2Institute of Inorganic Chemistry, RWTH Aachen University, Aachen, Germany
3Yusuf Hamied Department of Chemistry, University of Cambridge, CB2 1EW, UK
4Institute of Inorganic Chemistry, Julius Maximilians University Würzburg, Germany
5Advanced Chemistry Development, Inc.
6University of Canterbury, New Zeeland
7Beilstein Institut, Frankfurt(Main), Germany
The current standard InChI (v1.07) was developed with organic molecules as the focus. The algorithm generates identifiers for organometallics, which are useful even though they neglect aspects of these structures. In the recent years more interest has been articulated to fill this gap in the identification of inorganics and organometallics. For example, the NFDI4Chem initiative is intending to use InChI for the identification of inorganic compounds as well.
A revision of the InChI is being developed which includes more information about the organometallics and molecular inorganics. This requires an improved bond recognition and a more sophisticated stereochemical model. Beside single, double, and triple bonds other bond types like coordinative or haptic bonds must be processed to ensure that the same compound drawn in different ways is identified by the same InChI: for example, ferrocene can be drawn in multiple ways but should have just one InChI. Which depictions of a compound are identical and which are out of scope? The stereochemistry recognition of InChI must be extended to the more complex relationships in inorganic compounds. While tetrahedral stereocenters may lead up to 2n stereoisomers (with n = number of stereocenters), just one octahedral center leads to 30 isomers. Not all approaches to these issues are equally good, and community comments on the proposed InChI methods are very welcome. What will make the InChI even more useful for scientists using organometallic and molecular inorganic entities?