Formal Charge and Resonance

Formal charge and resonance are not museum pieces; they remain the fastest working tools chemists reach for when a new molecule's structure is uncertain. In organic and organometallic synthesis, resonance reasoning predicts which carbon of an enolate or which nitrogen of an amidinium ion will be more nucleophilic or electrophilic, directly guiding reaction design; carboxylate resonance (two equivalent C–O bonds, each order 1.5) explains why carboxylic acids are far more acidic than simple alcohols, since the conjugate base's negative charge is delocalized rather than localized on one oxygen. In materials chemistry, resonance-stabilized anions and cations (like the guanidinium or the cyclopentadienyl anion) are chosen as ligands or counterions precisely because delocalization lowers their energy and moderates their reactivity, a principle traceable straight back to the same bond-order averaging done for nitrate. In drug design, formal-charge and resonance analysis of a candidate molecule's functional groups (amidines, carboxylates, aromatic heterocycles) predicts protonation states and hydrogen-bonding patterns at physiological pH, feeding directly into computational models of binding affinity. In spectroscopy, infrared and Raman band positions for species like nitrate, carbonate, and azide are interpreted using the averaged, fractional bond orders resonance predicts, since a bond order of 1.33 vibrates at a frequency between the single- and double-bond extremes, offering an experimental fingerprint that confirms — rather than merely illustrates — the resonance picture. Even at the frontier of computational chemistry, natural bond orbital (NBO) analysis of high-level wavefunctions routinely reports "resonance weights" for candidate Lewis structures, a direct, quantitatively rigorous descendant of the pencil-and-paper formal-charge ranking taught here.

Ask the tutor: In SCN⁻, why does structure II (with negative formal charge on sulfur) actually contribute more to the real hybrid than a pure electronegativity argument would predict?

A Lewis structure is built from two ingredients: the skeleton (which atoms bond to which) and the electron placement (where the lone pairs and multiple bonds go). Once the skeleton is fixed, there is often more than one way to distribute the remaining electrons while still satisfying the octet rule (or the known exceptions to it). Formal charge lets you compare these electron-placement choices without any knowledge of quantum mechanics beyond the Lewis picture itself. The procedure encodes a simple idea: pretend that every bond is nonpolar, splitting each bonding pair exactly in half between the two atoms that share it. Under that fictitious assumption, count how many electrons "belong" to each atom (its lone pairs entirely, plus half of every bond it participates in). Compare that number to the number of valence electrons the atom would have as a free, neutral atom. Any difference is the formal charge — a bookkeeping surplus or deficit that would appear if bonds were perfectly covalent, which of course they rarely are. The word "formal" is doing real work in the name: it flags that this is a formalism, an internally consistent counting convention, not a claim about the actual electron density measured by X-ray diffraction or computed by quantum chemistry. Formal charges become powerful when several Lewis structures are candidates for the same molecule. The guiding principle — closely related to electronegativity and to the general chemical preference for charge neutrality — is that the most realistic structure is usually the one with formal charges closest to zero, and, when charge separation is unavoidable, the one that places negative formal charge on the more electronegative atom and positive formal charge on the less electronegative one. This is not an arbitrary rule; it reflects that real bonds are polarized toward the more electronegative atom, so a Lewis structure that already "agrees" with that polarization direction is closer to physical reality even though the covalent bookkeeping treats the bond as unpolarized. Resonance enters when two or more structures, differing only in the arrangement of electrons over a fixed atomic skeleton, are equally or comparably good by the formal-charge criterion. Ozone is the classic case: whichever oxygen–oxygen bond you draw as the double bond, the molecule looks asymmetric on paper, yet experimentally both O–O bonds are identical, with a bond length between that of a single and a double bond. No single Lewis structure can capture this; the real molecule is the resonance hybrid, an average over the contributing structures, connected by a double-headed resonance arrow (↔), never an equilibrium arrow (⇌) — the molecule does not oscillate between forms in time, it simply is the blend at all times.

The formal-charge formula, applied atom by atom within one Lewis structure:

where: $V$ is the number of valence electrons the free, neutral atom has (its group number in the main-group periodic table); $N$ is the number of nonbonding (lone-pair) electrons drawn on that atom in the structure; $B$ is the total number of electrons in bonds drawn to that atom (i.e., twice the number of bonds — a double bond contributes 4 to $B$, a single bond contributes 2). Reading: this is a conservation statement in disguise. Start from the atom's native valence-electron budget $V$. In the molecule it keeps all $N$ nonbonding electrons outright, and it is credited with exactly half of each shared bonding pair, $B/2$, under the covalent fiction. If the atom ends up with fewer electrons than it started with ($N + B/2 < V$), it has a positive formal charge, having "invested" electron density into bonding beyond what an equal split would return. If it ends up with more ($N + B/2 > V$), it has a negative formal charge, hoarding more electron density than a neutral atom's fair share. A second core relationship is the sum rule, essential as an error check on every structure you draw:

where: the sum runs over every atom in the species, and $q$ is the overall charge of the molecule or ion (zero for a neutral molecule, $-1$ for an anion such as NO₃⁻, $+1$ for a cation such as NH₄⁺, and so on). Reading: formal charge is invented electron bookkeeping, but it must still conserve total electron count. Since the Lewis structure as a whole has exactly the right number of electrons to give the species its actual charge, the formal charges — being redistributions of that same electron count relative to neutral atoms — must sum to that same overall charge. If your computed formal charges do not sum to $q$, you have miscounted electrons somewhere (wrong total valence electron count, a lone pair drawn twice, a bond drawn with the wrong multiplicity), and the structure must be redrawn before it is used for anything else. A third relationship, used specifically for resonance hybrids built from $n$ equivalent contributing structures, gives the average bond order:

Reading: when resonance structures are true energetic equivalents (same skeleton, same energy, just electrons shuffled by symmetry), the real molecule does not "pick one"; every bond position that varies among the structures ends up with an averaged, generally fractional, bond order, and every atom that varies in formal charge ends up with an averaged, generally fractional, charge.

Example 1 — Foundational: carbon dioxide, CO₂, confirming an all-zero structure. Total valence electrons: C contributes 4, each O contributes 6, giving $4 + 6 + 6 = 16$ electrons, i.e. 8 pairs. The standard Lewis structure is O=C=O, with two lone pairs on each terminal oxygen and no lone pairs on carbon. For carbon: $V = 4$ (group 14). It has no lone pairs, $N=0$. It participates in two double bonds, each contributing 4 bonding electrons, so $B = 4 + 4 = 8$.

For each oxygen: $V = 6$ (group 16). Each has two lone pairs, so $N = 4$. Each is doubly bonded to carbon only, so $B = 4$.

Check the sum rule: $FC_C + FC_{O(1)} + FC_{O(2)} = 0 + 0 + 0 = 0$, matching the neutral molecule's charge $q=0$. Every atom achieves an octet (carbon: two double bonds = 4 bonding pairs = 8 electrons around it; each oxygen: one double bond, 4 bonding electrons, plus two lone pairs, 4 more electrons = 8 total). All formal charges are zero and all octets are satisfied, so this structure is essentially definitive — no resonance is needed, and this is exactly why CO₂'s bond length (116 pm) matches a genuine C=O double bond rather than something intermediate.

Example 2 — Applied: nitrate, NO₃⁻, showing resonance among three equivalent structures. Total valence electrons: N contributes 5, each O contributes 6 (three oxygens), plus 1 extra electron for the overall $-1$ charge:

Nitrogen is central, bonded to three oxygens. To give nitrogen an octet without exceeding it (nitrogen, a period-2 element, cannot exceed 8 electrons around it), exactly one N–O bond must be a double bond and the other two must be single bonds, with each oxygen carrying enough lone pairs to complete its own octet. Take the double bond to O(1), single bonds to O(2) and O(3):

• Nitrogen: $V=5$, no lone pairs ($N=0$), bonded via one double bond (4 electrons) and two single bonds (2 + 2 = 4 electrons), so $B = 4+2+2 = 8$.

• O(1), the doubly bonded oxygen: $V=6$, two lone pairs ($N=4$), one double bond ($B=4$).

• O(2) and O(3), each singly bonded: $V=6$, three lone pairs ($N=6$), one single bond ($B=2$).

Sum check: $FC_N + FC_{O(1)} + FC_{O(2)} + FC_{O(3)} = 1 + 0 + (-1) + (-1) = -1$, matching $q = -1$ for the nitrate ion — the bookkeeping is consistent. This structure is chemically sensible (positive charge on nitrogen, the least electronegative atom present; negative charge on the singly bonded oxygens, the most electronegative). But nothing distinguishes O(1) from O(2) or O(3) — any of the three oxygens could equally well be the doubly bonded one, giving three resonance structures of identical energy, related by simply relabeling which oxygen carries the double bond (a genuine symmetry of the ion, $D_{3h}$). Averaging over the three equivalent contributors: each N–O bond is a double bond in one structure out of three and a single bond in the other two, so the averaged bond order at every N–O position is

Each oxygen carries formal charge $0$ in one structure and $-1$ in the other two, so the averaged charge on every oxygen is

and nitrogen's charge, being $+1$ in all three structures, stays exactly $+1$. Experimentally, all three N–O bonds in NO₃⁻ are indeed identical, at 124 pm — intermediate between a typical N–O single bond (~140 pm) and N=O double bond (~120 pm) — exactly consistent with a bond order near 1.33, and the ion's $D_{3h}$ symmetry (a threefold axis with three equivalent mirror planes) forbids any single Lewis structure with one unique double bond from being the true picture.

Example 3 — Competition-level: thiocyanate, SCN⁻, resolving genuinely non-equivalent resonance structures. Thiocyanate has the skeleton S–C–N (sulfur and nitrogen both terminal, carbon central), and total valence electrons:

Three distinct Lewis structures are possible, differing in how the 8 pairs are distributed as bonds versus lone pairs, all satisfying every atom's octet:

Structure I: S=C=N⁻ pattern — a cumulated double bond arrangement: S=C double bond, C=N double bond.

• S: $V=6$, two lone pairs ($N=4$), one double bond ($B=4$): $FC_S = 6-4-2 = 0$.
• C: $V=4$, no lone pairs, two double bonds ($B=8$): $FC_C = 4-0-4=0$.
• N: $V=5$, two lone pairs ($N=4$), one double bond ($B=4$): $FC_N = 5-4-2 = -1$.

Sum: $0+0+(-1) = -1$ ✓.

Structure II: ⁻S–C≡N — single bond to sulfur, triple bond to nitrogen.

• S: $V=6$, three lone pairs ($N=6$), one single bond ($B=2$): $FC_S = 6-6-1 = -1$.
• C: $V=4$, no lone pairs, one single bond + one triple bond ($B = 2+6=8$): $FC_C = 4-0-4=0$.
• N: $V=5$, one lone pair ($N=2$), one triple bond ($B=6$): $FC_N = 5-2-3=0$.

Sum: $-1+0+0=-1$ ✓.

Structure III: S≡C–N²⁻-type — triple bond to sulfur, single bond to nitrogen (formally allowed by electron count, but chemically implausible).

• S: $V=6$, one lone pair ($N=2$), one triple bond ($B=6$): $FC_S = 6-2-3=1$.
• C: $V=4$, no lone pairs, $B=8$: $FC_C=0$.
• N: $V=5$, three lone pairs ($N=6$), one single bond ($B=2$): $FC_N=5-6-1=-2$.

Sum: $1+0+(-2)=-1$ ✓ (bookkeeping is internally consistent, but the physical judgment below rules this one out). Now rank using the electronegativity criterion. Nitrogen (electronegativity ≈ 3.0) is more electronegative than sulfur (≈ 2.6). Structure III piles up charge magnitude (total $|FC|$ summed $=1+0+2=3$) and puts $-2$ on nitrogen while leaving sulfur positively charged — the wrong direction relative to sulfur's lower electronegativity being expected to bear less negative charge, not positive, and the large charge magnitude itself is a red flag. Structure III is essentially never a real contributor. Between I and II: total $|FC|$ is $0+0+1=1$ for structure I and $1+0+0=1$ for structure II — tied on magnitude. The deciding factor is where the negative charge sits: structure I places $-1$ on nitrogen (more electronegative, ✓ favorable), while structure II places $-1$ on sulfur (less electronegative than nitrogen, but sulfur is a large, polarizable, "soft" atom that can comfortably stabilize negative charge despite its lower electronegativity — an important refinement, since electronegativity is not the only determinant of charge-bearing comfort; size and polarizability matter for period-3+ atoms). In practice, both contribute to the resonance hybrid, but spectroscopic and computational evidence (bond lengths intermediate between single and triple bond values, with the C–S distance notably shorter than a pure single bond) shows structure II contributes more than a naïve electronegativity-only argument would suggest, precisely because sulfur tolerates negative formal charge well. Real thiocyanate is best described as a hybrid dominated by structures I and II, with S≡C–N-type structure III a negligible minor contributor. This example is the standard illustration that "minimize formal charge magnitude" and "put negative charge on the most electronegative atom" can trade off against real chemical factors like polarizability, and a careful chemist weighs both.

Resonance as a formal concept was developed somewhat later, principally by Linus Pauling in the 1930s, who systematized the idea that a single classical structure can be inadequate and that a weighted average of several valence-bond structures (the "resonance hybrid") gives a better description of molecules like benzene, the carboxylate ion, and ozone. Pauling's The Nature of the Chemical Bond (1939) made resonance a standard tool of structural chemistry, connecting it explicitly to bond-length and bond-energy data — a direct forerunner of the Delocalized Bonding Introduction treatment using molecular orbital theory, which recovers the same delocalization physically rather than as an average of separate localized pictures.