Pentacoordinated Carbon

From bondslam
Jump to: navigation, search


Henry Rzepa, Imperial College London[edit]

slides: DOI:b9r9


Henry (talk) 11:53, 19 July 2017 (CEST)[edit]

The recently reported [1] crystal structure of hexacoordinate carbon now firmly identifies a species first speculated about in 1973.[2]. I here propose an iso-electronic species which leads directly to an analogous pentacoordinated nitrogen. The logic can be expressed as follows.

Kekule stretching mode N, 1264

Kekule stretching mode C, 1275

Kekule stretching mode B, 1281

Kekule stretching mode O, 1202

  1. Five-coordinate nitrogen
    The cyclopentadienyl anion is considered as a ligand for the tri-cation RC3+. The latter carbon has just two valence electrons originating from the R-C bond.
  2. The aromatic Cp1- ligand provides a further six electrons, completing the valence shell of this carbon.
  3. One can consider the species RN4+ as also having just two valence electrons, again from the R-N bond (R=Me).
  4. This can now coordinate to the (aromatic) di-anion of cyclobutadiene to form a formally five-coordinate nitrogen, again completing the valence shell of the nitrogen and again producing a di-cationic species.
  5. An ωB97XD/Def2-TZVPP calculation [3] reveals the species to be a true minimum.
  6. QTAIM analysis of the electronic topology for five-coordinate nitrogen
    A QTAIM analysis shows all five coordinations to nitrogen have a line (bond) critical point with high values of ρ®
  7. NBO analysis indicates a nitrogen Wiberg bond index of 3.66, indicating it is not hypervalent, merely hypercoordinate. The four basal C-N bonds have Wiberg orders of 0.66, and the apical N-C bond order is 0.90.
  8. ELF analysis of the electronic topology for five-coordinate nitrogen
    An ELF basin analysis shows four basal C-N basins of 1.25e each and a more conventional two-electron apical C-N basin.
  9. In all respects this five-coordinate nitrogen analogue is similar to the recently reported carbon system. Other than three boron-caged systems[4] no formal five-coordinate nitrogen species have been either speculated upon, or attempts made to synthesize them.
  10. An iso-electronic replacement of N by C results in a pentacoordinate carbon monocation[5]
  11. A further iso-electronic replacement of N by B results in a pentacoordinate neutral boron system.[6]
  12. Extension to oxygen as a tri-cation also results in penta-coordination[7].
  13. One of the normal modes for these species is the so-called Kekulé mode, which elongates two C-C bonds and shortens the other two of the basal ligand. The values (cm-1) are typical of aromatic rings such as benzene itself, suggesting the carbon ligand retains aromaticity in the manner of benzene. They show a small but interesting progression along the series O → N → C → B from 1202 to 1281 cm-1.
  1. The challenge then to the synthetic community is to see if the pentacoordinate nitrogen species can be made and characterised.
  2. A second challenge would be whether penta-coordinated oxygen could be achieved.[8]

--Henry (talk) 09:08, 10 August 2017 (CEST)[edit]

Here I report some simple searches of CSD (Cambridge structure database).

  1. A simple search specifying merely that a single carbon have at least five other bonded (coordinated) atoms leads to 5910 hits.
  2. Specifying that no C-B (any bond type) must exist in the structure reduces this to 3931.
  3. Specifying that at least five C-C bonds to the central C must exist (any bond type) gives 8 hits, of which five do appear to be errors (the other three relate to the crystal structure discussed above).
  4. Specifying that at least five C-C or C-H bonds of any type must exist gives 36 hits, of which most are again probably errors. One genuine example is the recent structure of the norbornyl cation.[9]
  5. Specifying that at least five N-C or N-H bonds of any type to a central N atom results in zero hits.
  6. Specifying that at least five O-C or O-H bonds of any type to a central O atom results in one hit, the coordination being a clear error due to disorder.

To a certain extent at least these results indicate how difficult it can sometimes be to index a structure in the CSD for the bonds it carries. I suspect the indexing is sometimes done by a human based on their own interpretations, with guidelines of course.


  1. M. Malischewski, and K. Seppelt, Angewandte Chemie International Edition, 2016, 56, 368-370, DOI:10.1002/anie.201608795 . The crystal structure can be viewed at DOI:10.5517/CCDC.CSD.CC1M71QM
  2. H. Hogeveen and P. W. Kwant, Tet. Lett., 1973, 1665. DOI:10.1016/S0040-4039(01)96023-X
  3. H. S. Rzepa, 2017, DOI:10.14469/hpc/2348
  4. H. S. Rzepa, 2017, DOI:10.14469/hpc/2747
  5. H. S. Rzepa, 2017, DOI:10.14469/hpc/2872
  6. H. S. Rzepa, 2017, DOI:cbpt
  7. H. S. Rzepa, 2017, DOI:cbpx
  8. H. S. Rzepa, 2017, DOI:10.14469/hpc/2746 and DOI:cbpx
  9. DOI:10.1126/science.1238849 .