There are cases where hydrogen forms two covalent bonds (of order 0.5 I assume) in certain molecules and ions such as $\ce{H3^+}$, $\ce{CH5^+}$, $\ce{B2H6}$, $\ce{HF2^-}$, and hydrogen-bridged cations (below is an example from wikipedia about hydrogen-bridged cations), but are there cases where a coordination number of 3 or higher is achieved, even theoretically? Aside from maybe the more salt-like compounds, e.g. $\ce{NaH}$.

enter image description here

  • 6
    $\begingroup$ If the illustration reproduced is from a publication, or a book: Don't forget to state the source of the illustration used. For one, credit to whom credit is due. For two, the publication where this illustration was used may provide context, and serve as source for inspiration among future readers of your question. $\endgroup$
    – Buttonwood
    Commented Feb 21, 2023 at 13:50
  • 1
    $\begingroup$ I would not include $\ce{CH5+}$ or rather $\ce{[CH3(H2)]+}$ in the mix. en.wikipedia.org/wiki/Methanium Also, trivalent is the wrong word. You are looking for tricoordinate. Valence is a fixed value for each element; and that value for hydrgen is one. Valency is also a very ambiguous word, many people use it for different things, and in combination with other terms (e.g. hypervalent) it has become quite muddy. Since you obviously look for a coordination number, use that word to avoid ambiguity. $\endgroup$ Commented Feb 22, 2023 at 19:06
  • $\begingroup$ @Martin-マーチン, I agree I probably shouldn't have used the term "trivalent", but I don't see how methanium is unique from my other examples with 2e3c and 4e3c bonds. $\endgroup$
    – EpichinoM2
    Commented Feb 22, 2023 at 20:01
  • $\begingroup$ Well, there certainly is room for debate. And yes, looking back at all of those, they all seem rather different and unique. I guess I cannot come up with my thought process (again) from my last comment. $\endgroup$ Commented Feb 22, 2023 at 21:50
  • 1
    $\begingroup$ There are hundreds of metal complex structures with tricoordinated hydrogen, it's nothing special or exotic these days (e.g. search for μ3-H). $\endgroup$
    – andselisk
    Commented Feb 26, 2023 at 13:13

1 Answer 1


Magnesium hydride, with the rutile structure, has each hydrogen atom bonded to three magnesium atoms, but this is generally considered predominantly ionic (see here for a discussion of the structure and ionicity of alkali and alkaline earth hydrides).

A better candidate for covalently bound, three-coordinate hydrogen is the iridium complexes given by Chodosh et al.[1]. The formulae of these complexes are described as follows:


$\ce{L: PPh3,P}\text{(cyclohexyl)}_3$

$\ce{L':}\text{ pyridine, acetonitrile}$

In these complexes the three iridium centers are each bonded to two outer hydride ligands and rwo neutral organic ligands, with a central hydride ligand joined to all three iridium centers holding the complex together.

The introductory section of the article (available in front of the paywall) mentions several other, previously reported examples.

Three is actually not the maximum. Cousins et al. [2] have obtained four-coordinate hydrogen incorporated into alkali metal phenolate oligomers, the hydrogen taking the place of a lithium or sodium atom and thus showing alkali-metal-like characteristics.


  1. Daniel F. Chodosh, Robert H. Crabtree, Hugh Felkin, Sheila Morehouse, and George E. Morris (1982). "Trinuclear iridium cluster containing a tricoordinate bridging hydrogen ligand: structural and chemical studies". Inorg. Chem. 21, 4, 1307–1311. Publication Date:April 1, 1982 https://doi.org/10.1021/ic00134a005

  2. David M. Cousins, Matthew G. Davidson and Daniel García-Vivó (2013). "Unprecedented participation of a four-coordinate hydrogen atom in the cubane core of lithium and sodium phenolates". Chem. Commun. 49, 11809-11811. https://doi.org/10.1039/C3CC47393G.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.