Wade's Rules for Cluster Electron Counting

Wade's rules remain the everyday sorting tool for an entire branch of structural inorganic chemistry that has moved well past the original boron hydrides. Carborane chemistry, where one or more $\{\text{BH}\}$ vertices are replaced by isolobal $\{\text{CH}\}$ units, uses exactly the same SEP counting (with the vertex electron contributions adjusted per the Isolobal Analogy and Cluster Extensions) to predict cage shapes; the resulting closo-carboranes and their nido-derived "dicarbollide" anions are workhorse ligands in organometallic and medicinal chemistry, including boron neutron capture therapy agents that exploit the intense, localized neutron absorption of $^{10}\text{B}$ packed into a stable, biologically tolerable polyhedral cage. Metallaborane and metallacarborane clusters, in which transition-metal fragments such as $\{\text{Fe(CO)}_3\}$ or $\{\text{Cp*Rh}\}$ occupy vertices of the same deltahedra, are counted with Wade–Mingos electron bookkeeping and underpin a large swath of cluster organometallic chemistry, some of it relevant to homogeneous and heterogeneous catalyst design where metal clusters model surface active sites.

Zintl ion clusters of the heavier p-block elements — species like $\text{Ge}_9^{2-}$, $\text{Sn}_9^{4-}$, or $\text{Pb}_{10}^{2-}$ formed by reducing group 14 elements with alkali metals in liquid ammonia or ionic liquids — obey the same closo/nido/arachno logic, because a naked main-group atom carrying a lone pair pointing away from the cage is isolobal with a $\{\text{BH}\}$ vertex once the lone pair is set aside as the "exo" orbital. This has made Wade-type counting a standard diagnostic in the surging field of Zintl cluster and polyanion chemistry relevant to intermetallic materials and thermoelectric compound design. In materials science, boron-rich solids and boride ceramics (used for their extreme hardness and neutron-absorbing properties in reactor shielding) are frequently rationalized starting from the icosahedral $\text{B}_{12}$ unit, whose closed-cage stability is itself an SEP = 13 = n+1 closo count — the same arithmetic that predicts the shape of a simple molecular anion in solution also explains the persistent structural motif inside extended solid-state boron networks. Finally, the electron-counting habit of mind that Wade's rules instill — always ask "how many electron pairs are actually available for this bonding job" before drawing a structure — generalizes directly to the Isolobal Analogy and Cluster Extensions used to connect organometallic fragments, main-group clusters, and even surface science across what would otherwise look like unrelated corners of the periodic table.

Ask the tutor: How does the skeletal electron count change when a $\{\text{BH}\}$ vertex is replaced by a transition-metal carbonyl fragment like $\{\text{Fe(CO)}_3\}$ in a metallaborane cluster?

Ordinary molecules are usually described with localized two-center two-electron (2c–2e) bonds: each bond is "owned" by a specific pair of atoms, and the molecular shape follows from VSEPR reasoning around each center. Boron clusters break this picture at the outset, for a structural reason: boron has only three valence electrons and four valence orbitals, so it is chronically electron-poor. A boron atom in a $\text{BH}$ cluster vertex uses one orbital and two electrons for an external, localized $\text{B–H}$ bond (radial, pointing away from the cage), leaving three roughly tangential orbitals (one radially-inward-pointing hybrid plus two tangential p-orbitals) and only one electron per vertex to build all of the cage bonding. There simply are not enough electrons to make normal 2c–2e bonds along every polyhedral edge — an icosahedron has 30 edges but a $\text{B}_{12}\text{H}_{12}^{2-}$ cage has only 13 pairs of skeletal electrons to spend. The solution nature adopts is exactly the one used in the boron hydrides individually: multi-center bonding, in which a single pair of electrons occupies a molecular orbital delocalized over three, four, or more cage atoms simultaneously, distributing the scarce electron density across the whole polyhedral surface rather than concentrating it on any one edge. Because the bonding is a property of the whole cage rather than of individual edges, the natural framework is molecular orbital theory applied to the entire deltahedron (a polyhedron whose faces are all triangles) rather than a localized Lewis structure. Each vertex $\text{BH}$ unit contributes three frontier orbitals directed into the cage interior: one large hybrid orbital pointing radially toward the cage center, and two smaller orbitals tangential to the cage surface. When $n$ such vertices combine to build a closed deltahedral cage, group theory shows that these $3n$ atomic-like fragment orbitals sort into exactly $n+1$ strongly bonding combinations (which lie well below the rest in energy, forming a nonbonding-to-antibonding "gap") and $2n-1$ orbitals that are nonbonding or antibonding with respect to the cage skeleton. This $n+1$ figure is topological — it comes from the way the radial orbitals combine into one fully-symmetric, all-bonding combination plus $n-1$ further combinations matching the frame's rotational and vibrational symmetry, joined by additional bonding combinations from the tangential set — and it is essentially independent of which particular deltahedron (tetrahedron, octahedron, icosahedron, or any of the intermediate shapes) is used. It is this invariance that lets one write a single rule spanning every closo borane from $\text{B}_5\text{H}_5^{2-}$ to $\text{B}_{12}\text{H}_{12}^{2-}$. If a cluster supplies exactly $n+1$ pairs of skeletal electrons, all $n+1$ bonding MOs fill and no more, giving maximum bonding stabilization with a complete, closed deltahedral cage: this is the closo case. If there are more electrons than that — $n+2$, $n+3$, or $n+4$ pairs — the extra pairs must occupy orbitals that would be strongly antibonding on the closed $n$-vertex cage. The cluster avoids this penalty by opening up: it adopts the geometry of a larger closed parent deltahedron (with $n+1$, $n+2$, or $n+3$ vertices respectively) with one, two, or three vertices left vacant. Removing a vertex removes exactly the antibonding orbitals that the extra electron pairs would otherwise have had to occupy, so the electron count and the geometry are locked together: nido clusters are one-vertex-short fragments of a closed $(n+1)$-vertex deltahedron, arachno clusters are two-vertex-short fragments of an $(n+2)$-vertex deltahedron, and so on. The vertices removed are always the ones of highest connectivity in the parent polyhedron, because removing a high-connectivity vertex eliminates the most antibonding character per vertex lost.

The skeletal electron pair count for a neutral or anionic borane $\text{B}_n\text{H}_{n+x}^{q-}$ (or, more generally, any cluster built of vertex units each contributing orbitals and electrons in the same pattern as $\text{BH}$) is obtained from a total valence electron count, with each terminal (exo) bond subtracted off before halving:

where:

• $V_{\text{total}}$ is the sum of all valence electrons in the cluster: 3 per boron, 1 per terminal hydrogen (each $\text{BH}$ vertex contributes 4 electrons total, of which 2 go to the exo B–H bond), plus the charge contribution ($+1$ electron per unit negative charge, $-1$ per unit positive charge),
• $n_{\text{exo B–H}}$ is the number of terminal, non-bridging B–H bonds (extra bridging hydrogens do not consume a skeletal pair — a bridging H contributes its single electron directly into the skeletal count),
• the division by 2 converts electrons into electron pairs, since bonding MOs are filled two electrons at a time.

For the common family $\text{B}_n\text{H}_{n+x}^{q-}$ with one terminal H per vertex plus $x$ extra (bridging or additional terminal) hydrogens, this collapses to the frequently quoted shortcut:

where:

• $n$ is the number of boron vertices,
• $x$ is the number of hydrogens in excess of one-per-boron (i.e. total H count minus $n$),
• $q$ is the magnitude of the anionic charge (subtract for cationic charge).

The classification rule compares SEP to $n$:

where in every line $n$ is the number of vertex atoms actually present in the observed cluster (not the parent). Reading these four lines together: the parent polyhedron used to name the shape always has $\text{SEP}$ vertices in total — the "$+1$" tells you directly how many more vertices the imagined closed cage has than the atoms you can see, and hence how many vertices are missing.

Example 1 (foundational): $\text{B}_6\text{H}_6^{2-}$. Identify $n = 6$ boron vertices, each already carrying exactly one terminal hydrogen, so $x = 0$ (six H total, six borons, no excess), and charge $q = 2$ (dianion, so $+2$ electrons).

Compare to $n = 6$: $\text{SEP} = n+1 = 7$. This is the closo condition, so $\text{B}_6\text{H}_6^{2-}$ adopts the complete 6-vertex deltahedron, the octahedron — matching the experimentally observed regular-octahedral $\text{B}_6\text{H}_6^{2-}$ ion. Cross-check by full valence count: $V_{\text{total}} = 3(6) + 6 + 2 = 26$ electrons; subtract $2n = 12$ for the six exocyclic B–H bonds leaves $14$ skeletal electrons, i.e. $7$ pairs — the same answer, confirming the shortcut formula is self-consistent with the definitional route.

Example 2 (applied): $\text{B}_5\text{H}_{11}$, an unsymmetrical, non-round test of the rule. This species has 5 borons and 11 hydrogens — deliberately not the more commonly quoted $\text{B}_5\text{H}_9$, to test the formula on an asymmetric, less "special" case. Here $n=5$; excess hydrogens $x = 11 - 5 = 6$; the molecule is neutral so $q=0$.

Compare to $n=5$: $\text{SEP} - n = 3$, so $\text{SEP} = n+3$, which is the arachno condition. The parent closed deltahedron has $n+2 = 7$ vertices — a pentagonal bipyramid — with two vertices removed from that seven-vertex parent (specifically, two adjacent equatorial-type high-connectivity vertices), leaving the very open, web-like framework characteristic of arachno-boranes. This matches the known structure of $\text{B}_5\text{H}_{11}$, which is markedly more open and reactive than the nido cluster $\text{B}_5\text{H}_9$ (itself $n=5$, $x=4$, $\text{SEP} = 5+2=7=n+2$, nido, parent 6-vertex octahedron minus one vertex — a square pyramid, matching the well-known square-pyramidal $\text{B}_5\text{H}_9$ cage). Cross-check by full electron count: $V_{\text{total}} = 3(5)+11+0 = 26$; subtract $2n=10$ for five exocyclic B–H bonds, leaving $16$ skeletal electrons distributed among cage bonds and the extra bridging/terminal hydrogens, i.e. $8$ pairs — matching the shortcut result exactly.

Example 3 (exam/competition level): a heteroborane anion, $[\text{C}_2\text{B}_9\text{H}_{13}]^{-}$, testing vertex-type substitution. This tests whether the rule extends past pure boron cages, foreshadowing carborane chemistry. A $\{\text{CH}\}$ vertex is isolobal with $\{\text{BH}\}^{-}$: carbon brings one more valence electron than boron, but as a neutral vertex it contributes one more skeletal electron than a neutral $\{\text{BH}\}$ vertex contributes (each vertex still uses one orbital/2 electrons for its exocyclic C–H or B–H bond and leaves 3 frontier orbitals; a $\{\text{CH}\}$ vertex leaves 2 skeletal electrons instead of $\{\text{BH}\}$'s 1). Total vertex count $n = 2 (\text{C}) + 9 (\text{B}) = 11$. Count valence electrons directly, vertex by vertex, rather than trying to force the shortcut formula (which was built for all-boron cages):

• 2 carbons at 4 valence electrons each: $8$
• 9 borons at 3 valence electrons each: $27$
• Hydrogens: 11 vertex atoms need 11 terminal H (one per vertex) for the "base" cage, and the formula shows 13 total H, so there are $13 - 11 = 2$ extra (bridging) hydrogens contributing 1 electron each: $13$ total H electrons
• Charge: anionic, $+1$ electron

Subtract $2n = 22$ for the eleven exocyclic terminal C–H/B–H bonds (one per vertex, the "base" 11 hydrogens; the two extra bridging hydrogens are not subtracted here since they were already counted as skeletal-electron contributors, not exocyclic-bond electrons):

A non-integer pair count signals a bookkeeping slip worth pausing on: the total electron count $V_{\text{total}}=49$ is odd, which cannot correspond to a normal closed-shell cluster, so the assumed formula must be re-examined — this is exactly the kind of internal-consistency check the rule offers for free. Revisiting the hydrogen count: $[\text{C}_2\text{B}_9\text{H}_{13}]^{-}$, sometimes called the "dicarbollide precursor," in fact has one C–H and one B–H per vertex (11 exocyclic H) plus 2 bridging H, totaling 13 — consistent with what was assumed — so the electron count itself must be re-examined for a units slip: each neutral $\{\text{CH}\}$ vertex supplies $4(\text{C}) + 1(\text{H}) = 5$ valence electrons total, of which 2 go to the exocyclic C–H bond, leaving 3 skeletal electrons per carbon vertex (not the "2 skeletal electrons per {CH}" heuristic used loosely above, which conflated valence count with isolobal-replacement shorthand). Recomputing cleanly and only from first principles: each $\{\text{CH}\}$ vertex $= 4+1=5$ total electrons; each $\{\text{BH}\}$ vertex $=3+1=4$ total electrons; two extra bridging H contribute $1$ electron each; the anionic charge contributes $1$ electron:

which reproduces $49$ — so the electron count was right, and the apparent problem is that $[\text{C}_2\text{B}_9\text{H}_{13}]^{-}$ as conventionally drawn is a nido cluster missing one vertex from an 11-vertex-parent count, and real dicarbollide-type clusters of this hydrogen/charge combination are typically written with an even total; the resolution taught in carborane chemistry is that the precise protonation state (13 vs. 12 H) fixes the parity, and a chemist encountering a half-integer SEP should treat it as a flag to recheck the formula (most likely an extra or missing bridging hydrogen or a charge typo) rather than as a valid cluster — precisely the diagnostic value of doing the sum explicitly rather than quoting a memorized shape.