Following on from his previous post, Neil Eccles has asked me to explain how Materials Registration would handle nickel organolithium complexes such as the ones from these links:

https://onlinelibrary.wiley.com/doi/10.1002/anie.202209797

https://twitter.com/AndryjBorys/status/1554881904930951171

These feature 3 center, 2 electron bonding and 2 center coordinate (dative covalent) bonding.

Materials Registration computes identifications based on a valid 2D structural formula, usually entered and stored in molfile format. Molfile format has the ability to represent 2 different types of zero-order bonds: hydrogen bonds and coordinate bonds. Materials Registration ignores hydrogen bonds when computing identifications, but it does take note of coordinate bonds.

For example, this nickel organolithium complex drawn with hydrogen bonds generates a COMPOUND identification based on all the 2D structural formula ignoring the hydrogen bonds.

In the current release, Materials Registration does not compute identifications for structures containing coordinate bonds but a future release will. For example, diborane drawn with coordinate bonds does not have a computed identification (the structure can be drawn without a direction arrow on the coordinate bond and it will be considered equivalent: the arrow has been added to make it easier to see where the coordinate bonds are):

Contrast with diborane drawn using hydrogen bonds which does have a computed identification:

This brings me to a final point which provokes a question. The adduct of ammonia and boron trifluoride can be drawn using a charge separated single bond or a coordinate bond:

or:

Currently, these would be considered two different substances based on the 2D structural formula, but should they be considered equivalent and if so, which style is preferred?