Lewis Structures and the Octet Rule

Take a bottle of chlorine gas and a canister of methane. Ask a simple question. Why does each molecule stop growing where it does? Cl₂ is a single Cl–Cl unit, not a chain. CH₄ has exactly four hydrogens bonded to carbon, never three or five. The answer lies in a deceptively simple bookkeeping device. It was invented in 1916 by Gilbert N. Lewis: the electron-dot structure. It tracks only the valence electrons of each atom. It insists, wherever possible, that every main-group atom finish with eight electrons around it (two for hydrogen). Lewis structures let a chemist predict connectivity, bond order, and reactive sites. All this comes from nothing but a molecular formula and a periodic table. This matters because Lewis structures are the entry point to essentially every qualitative model of molecular structure used afterward. Once you can draw a correct dot structure, you can assign formal charges. You can also choose between resonance contributors. Then you can feed the arrangement of bonding and lone pairs into VSEPR theory to predict three-dimensional shape. The octet rule itself is not a law of physics. It is an empirical summary of a fact: filled s and p subshells (eight electrons, matching the noble-gas configuration) are unusually stable arrangements for second-period atoms. But as a predictive heuristic for C, N, O, F and their neighbors, it is remarkably reliable. The big picture is this: bonding is a redistribution of valence electrons. This redistribution lowers the total energy of a system of nuclei and electrons relative to the separated atoms. Lewis's insight was that for a huge swath of ordinary molecules, this redistribution can be captured by a small set of rules. Count electrons, place lone pairs, share pairs to complete octets. You can do this without ever solving a Schrödinger equation. It is a bookkeeping theory, not a quantum-mechanical one. But its predictions about connectivity and bond multiplicity survive essentially unchanged. This holds true even when you later derive them from molecular orbital theory. This article builds the method from the ground up. It covers how to count valence electrons correctly, and the step-by-step algorithm for placing atoms, bonds, and lone pairs. It explains how multiple bonds and expanded/contracted octets arise. It also shows where the rule bends or breaks: odd-electron species, electron-deficient boron, hypervalent main-group elements below period 2. Along the way you will see the exact arithmetic that makes the rule work, not just the recipe.

Key takeaways

A Lewis structure shows only valence electrons, arranged as bonding pairs (shared, shown as lines) and lone pairs (unshared, shown as dots).
Total valence electrons for a neutral molecule = sum of group numbers; for an ion, add electrons for negative charge, subtract for positive charge.
The octet rule says most second-row atoms (C, N, O, F) are most stable with 8 electrons around them; H wants only 2 (duet rule).
A systematic algorithm — skeleton, distribute electrons, complete octets, form multiple bonds if needed — gives a unique, checkable structure.
Exceptions are real and rule-governed: electron deficiency (BF₃), odd-electron radicals (NO, NO₂), and expanded octets for period 3+ elements (SF₆, PCl₅).
Key idea: a correct Lewis structure is the electron count and connectivity that uses every valence electron exactly once while getting as many atoms as possible to a noble-gas-like electron count, with multiple bonds substituting for lone pairs whenever a central atom is short of electrons.

The core insight

Electrons are conserved and arranged, never created: the total dot-and-bond count in a correct Lewis structure must exactly equal the sum of valence electrons contributed by all atoms (corrected for charge), and the placement is chosen to give as many atoms as possible a noble-gas electron count — eight for C–F row elements, two for hydrogen — using multiple bonds when single bonds alone leave a central atom short.

Where this lives today

Lewis structures remain the first and fastest diagnostic tool for predicting reactivity across chemistry, biochemistry, and materials science. A medicinal chemist sketching a new drug candidate uses dot structures to locate lone pairs available for hydrogen bonding or metal coordination, and to flag electron-deficient centers (boron-containing proteasome inhibitors like bortezomib exploit exactly the electron deficiency taught by BF₃-type reasoning) or radical centers relevant to metabolic oxidation. In atmospheric chemistry, the odd-electron Lewis structures of NO, NO₂, and ClO explain why these species are so reactive as radicals, directly underpinning models of stratospheric ozone depletion and photochemical smog. Materials chemists drawing hypervalent silicon and phosphorus intermediates (in silicone polymer synthesis or in flame-retardant phosphate esters) rely on expanded-octet reasoning to anticipate five- and six-coordinate geometries before ever running a computation. In inorganic and organometallic chemistry, Lewis structures are the entry point for electron counting at metal centers (extended, with formal charge and oxidation-state bookkeeping, into the 18-electron rule for transition-metal complexes), and for choosing among resonance contributors in ions like carbonate, nitrate, and sulfate that appear throughout geochemistry and biochemistry. Even modern quantum chemistry software reports "natural bond orbital" analyses that are, at heart, a rigorous re-derivation of exactly the same shared-pair and lone-pair picture Lewis proposed by hand in 1916.

Ask the tutor: Why does nitrogen dioxide (NO₂) refuse to form a symmetric Lewis structure the way carbon dioxide does, and what does that tell us about its odd-electron reactivity?

The physical picture

Lewis's central claim is that a chemical bond is a shared pair of electrons localized (in the simplest picture) between two nuclei, and that this sharing is favorable because it lets both atoms simultaneously "see" a more complete valence shell. For main-group elements in the second period, "complete" means eight electrons in the n = 2 shell — two in the 2s orbital and six in the three 2p orbitals — mimicking the electron configuration of the nearest noble gas, neon. Hydrogen and helium instead fill with only two electrons, since the n = 1 shell holds only the 1s orbital: this is often called the duet rule. The reason eight (or two) is special is not mysterious once you recall orbital counting. Each atom in the second period has one 2s and three 2p valence orbitals available — four orbitals total, each holding two electrons with opposite spin, for a maximum of eight electrons. When all four orbitals are used, either as bonding pairs (shared with a neighbor) or lone pairs (kept entirely on one atom), the atom has no more room and no partly filled valence orbital exposed to attack. This is what makes octets kinetically and thermodynamically privileged for C, N, O, F, and (usually) their heavier period-2 neighbor Ne — although the true energetic argument, developed rigorously in molecular orbital theory, is that bonding and nonbonding orbitals are all filled while antibonding orbitals are all empty. Building a Lewis structure is bookkeeping with a physical justification: every dot you place represents a real valence electron, contributed by a real atom, and the final arrangement must minimize the energy by maximizing bonding-pair stabilization while completing octets. The skeleton (which atoms bond to which) is decided first — normally the least electronegative nonhydrogen atom sits central, since it can form the most bonds — and then electrons are distributed outward to inward, lone pairs before multiple bonds, exactly as detailed in the derivations below.

Figure

Figure 1 — Building CO₂ step by step: skeleton, lone pairs, then double bonds

Figure

Figure 2 — Electron count comparison: SO₂ (expanded/resonance) vs CH₄ (simple octet) vs BF₃ (electron-deficient)

The core ideas

The entire method rests on one conservation law and one stability criterion.

N_{\text{total}} = \sum_{\text{atoms}} G_i \;\pm\; q

where: $N_{\text{total}}$ is the number of valence electrons to be drawn as dots/lines; $G_i$ is the main-group number (1–8, i.e. number of valence electrons) of atom $i$ ; $q$ is the ionic charge, with $+q$ electrons added for an anion of charge $-q$ and $q$ electrons subtracted for a cation of charge $+q$ .

Reading: this equation is nothing but conservation of charge and electron number. It says a Lewis structure must never invent or destroy electrons — whatever valence electrons the free atoms would have (adjusted for the ion's charge) is exactly what appears as dots and bonds in the finished diagram. Every subsequent step is just placement of this fixed total.

N_{\text{total}} = 2B + 2L

where: $B$ is the number of bonding pairs (shared electron pairs, drawn as lines), $L$ is the number of lone pairs (unshared electron pairs, drawn as dot pairs), and the factor of 2 converts pairs to electrons.

Reading: every valence electron ends up in exactly one of two roles — shared between two nuclei (bonding) or confined to one nucleus (lone). This equation is the accounting identity used to check that a finished structure has used every electron.

\text{Electrons needed} = \sum_{\text{atoms}} O_i - N_{\text{total}}

where: $O_i$ is the octet target for atom $i$ (8 for period-2 nonmetals C–F, 2 for H, sometimes more for expanded-octet elements); the difference tells you how many bonding pairs must be shared beyond single bonds — i.e., how many multiple bonds are required — because sharing a pair lets it count toward two atoms' totals simultaneously rather than one.

Reading: this is the "shortfall" equation. If simply drawing single bonds and filling remaining electrons as lone pairs leaves some atom (usually the central one) short of its octet, the shortfall must be resolved by converting a terminal atom's lone pair into a second (or third) shared pair — a multiple bond — because a shared pair counts toward both atoms' octets at once, which is strictly more efficient with a fixed electron budget than a lone pair, which counts toward only one.

Derivation AThe stepwise algorithm, with the arithmetic made explicit

Goal. Given a molecular formula, produce the unique (or resonance-averaged) Lewis structure consistent with the fixed electron count and with the maximum number of atoms reaching octet (or duet), using the shortfall logic above.

Step 1 — Count total valence electrons. Sum group numbers of every atom, then adjust for charge. This number is fixed and non-negotiable; every later step must terminate with exactly this many electrons drawn.

Step 2 — Choose the skeleton. Identify the central atom(s): usually the atom that (a) is not hydrogen (H only ever forms one bond) and (b) has the lowest electronegativity among the remaining atoms, since less electronegative atoms tolerate being surrounded by more bonds/negative density from neighbors. Connect every terminal atom to the center with a single bond. This uses $2\times(\text{number of terminal atoms})$ electrons immediately, since each single bond is one shared pair.

Step 3 — Complete terminal-atom octets first. Terminal atoms (other than H) still need lone pairs to reach 8. Add lone pairs to terminal atoms until each has 8 electrons total (bonding + lone). This step consumes electrons greedily on the outside of the molecule because terminal atoms cannot "borrow" electron density the way a central atom can via multiple bonding.

Step 4 — Assign leftover electrons to the central atom. Whatever electrons remain after Step 3 go on the central atom as lone pairs.

Step 5 — Check the central atom's total. If it is short of an octet, promote lone pairs to bonding pairs. For every 2 electrons the central atom lacks, take one lone pair from an adjacent terminal atom and turn it into a second (or third) bond to the center — this creates a double or triple bond. Each such promotion adds 2 electrons to the central atom's count without changing the total electron count, because the pair is now simply shared instead of lone; it still counts once towards conservation but now counts toward two atoms' octets simultaneously.

Step 6 — Verify. Recount: total dots + (2 × lines) must equal $N_{\text{total}}$ from Step 1, and every atom (except recognized exceptions) must show 8 electrons around it (2 for H). This closes the loop between the conservation equation and the octet criterion. Applying this to CO₂ (from Figure 1): $N_{\text{total}} = 4 + 2(6) = 16$ . Skeleton O–C–O uses 4 electrons (2 bonds). Step 3 gives each O three lone pairs (6 electrons each), consuming $2\times 6 = 12$ more, total so far $4+12=16$ — all electrons used, but C shows only 4 electrons (2 bonding pairs), a shortfall of 4. Step 5: promote one lone pair from each oxygen into a second bond to carbon, giving two C=O double bonds. Now C has 4 bonding pairs (8 electrons), each O has 2 bonding pairs + 2 lone pairs (8 electrons), and the total electron count is unchanged at 16 (still $2\times2\text{ bonds}\times2 + 2\times2\text{ lone pairs}\times2 = 8+8=16$ — check the pair count: 4 bonding pairs total (8 e⁻) + 4 lone pairs total (8 e⁻) = 16). Every criterion is satisfied simultaneously.

Go deeper · Derivation B · 2 minWhy multiple bonds — not extra lone pairs — resolve an octet shortfall

Derivation BWhy multiple bonds — not extra lone pairs — resolve an octet shortfall

Goal. Show algebraically why an electron-poor central atom is fixed by converting a neighboring lone pair into a shared pair, rather than by simply adding more lone pairs (which the fixed total electron count forbids), and quantify the shortfall for a general AXₙ fragment.

Step 1 — Set up the electron budget for a general central atom A bonded to $n$ terminal atoms X, ignoring lone pairs on A for now. Let $G_A$ be A's group number (valence electrons) and $G_X$ each terminal atom's group number. Total valence electrons:

N_{\text{total}} = G_A + nG_X

Step 2 — Electrons used forming the $n$ single bonds and completing each X's octet. Each single bond uses 2 electrons (shared). Each terminal atom X (assumed to want an octet like O, N, F) needs $8 - 2 = 6$ more electrons as lone pairs if singly bonded (one shared pair already contributes 2 of its 8). So electrons committed to bonds + terminal lone pairs:

E_{\text{used}} = 2n + 6n = 8n

This is just $n$ complete octets built entirely from A–X single bonds and terminal lone pairs, with A itself not yet counted.

Step 3 — Electrons left for the central atom A.

E_A = N_{\text{total}} - E_{\text{used}} = G_A + nG_X - 8n

Because each A–X bond already gives A 2 electrons (one shared pair counts toward both atoms), A's total electron count around itself is

\text{(A's count)} = 2n + E_A = G_A + nG_X - 6n

Step 4 — Compare to A's octet target (8) to get the shortfall.

\text{Shortfall} = 8 - (G_A + nG_X - 6n) = 8 + 6n - G_A - nG_X

If the shortfall is positive, A does not reach an octet using single bonds alone. Since the total electron count $N_{\text{total}}$ is fixed (Step 1's equation cannot be revisited — no electrons can be manufactured), the only way to raise A's local count without changing the global total is to reassign an electron pair that is already counted: take one lone pair sitting entirely on a neighboring X and make it a second shared pair between A and that X. This converts one A–X single bond into a double bond. The pair's location changes; its existence does not. Each such promotion raises A's count by exactly 2 (it now "sees" that pair too) while leaving X's total unchanged at 8 (X trades one lone pair for one bonding pair, both of which count fully toward its own octet).

Step 5 — Apply to CO₂ as the check. Here A = C ( $G_A = 4$ ), X = O ( $G_X = 6$ ), $n = 2$ :

\text{Shortfall} = 8 + 6(2) - 4 - 2(6) = 8 + 12 - 4 - 12 = 4

A shortfall of 4 electrons means two promotions (each promotion supplies 2 electrons to C), i.e., two lone pairs must become two additional shared pairs — exactly the two C=O double bonds found in Step 5 of the first derivation. The algebra and the diagram agree exactly.

Limiting check. For CH₄, treat H not as an octet-seeking X but note H needs no lone pair at all (duet already satisfied by 1 shared pair), so the generic formula above (built assuming X needs 6 extra electrons) does not apply; redo directly: $N_{\text{total}}=4+4(1)=8$ , four C–H single bonds use all 8 electrons, C already shows $2\times4=8$ electrons — shortfall is zero, correctly predicting no multiple bonds are needed, consistent with the sp³, four-single-bond structure of methane.

Worked examples

Example 1 (foundational) — Ammonia, NH₃. Count valence electrons: N contributes 5, each of three H contributes 1, total

N_{\text{total}} = 5 + 3(1) = 8.

Skeleton: N is central (less electronegative than nothing else is present besides H, and H cannot be central since it forms only one bond), three N–H single bonds. Each bond uses 2 electrons: $3 \times 2 = 6$ electrons used, leaving $8-6=2$ electrons, i.e., one lone pair, placed on N. Check N's total: 3 bonding pairs (6 electrons) + 1 lone pair (2 electrons) = 8 — octet satisfied. Each H has 1 bonding pair = 2 electrons — duet satisfied. Total electrons drawn: $3$ bonds $\times 2 + 1$ lone pair $\times 2 = 6+2=8$ , matching $N_{\text{total}}$ exactly. No multiple bonds are needed because the shortfall calculation (Step 4 of the second derivation, adapted for H which needs no extra lone-pair electrons) gives zero deficit at nitrogen. This lone pair is exactly the one responsible for ammonia's Brønsted basicity and its pyramidal shape under VSEPR theory.

Example 2 (applied) — the nitrate ion, NO₃⁻. Count valence electrons: N contributes 5, three O contribute $3\times6=18$ , plus 1 electron for the single negative charge:

N_{\text{total}} = 5 + 18 + 1 = 24.

Skeleton: N central, three N–O single bonds, using $3\times2=6$ electrons, leaving $24-6=18$ electrons for lone pairs. Distribute to give each terminal O an octet: each O already has 2 electrons from its bond, needs 6 more as three lone pairs, consuming $3\times6=18$ electrons — exactly the amount remaining, so N receives zero additional lone pairs. Check N's total: three bonding pairs only, $3\times2=6$ electrons — a shortfall of 2 relative to an octet. Apply the promotion rule: convert one O's lone pair into a second N–O bond, forming one N=O double bond. Now N has 2 single bonds + 1 double bond = 4 bonding pairs = 8 electrons (octet satisfied); the doubly bonded O has 2 bonding pairs + 2 lone pairs = 8 electrons; the two singly bonded O⁻ centers each have 1 bonding pair + 3 lone pairs = 8 electrons. Total electron count: bonds contribute $4\times2=8$ electrons-worth of pairs... more carefully, count pairs: 4 bonding pairs (one double = 2 pairs, two singles = 2 pairs) $\to 4\times2=8$ electrons, plus lone pairs: $2+3+3=8$ pairs $\to 16$ electrons; grand total $8+16=24$ , matching $N_{\text{total}}$ . Because any of the three oxygens could equally have received the double bond, nitrate is properly drawn as three resonance structures — a preview of the reasoning developed fully under Formal Charge and Resonance.

Example 3 (exam/competition level) — xenon tetrafluoride, XeF₄, an expanded-octet case. Count valence electrons: Xe (group 18, but retains 8 valence electrons as a reactive noble gas here) contributes 8, four F contribute $4\times7=28$ :

N_{\text{total}} = 8 + 4(7) = 36.

Skeleton: Xe central, four Xe–F single bonds, using $4\times2=8$ electrons, leaving $36-8=28$ electrons. Give each F an octet: each F has 2 electrons from its bond, needs 6 more as three lone pairs, consuming $4\times6=24$ electrons. Remaining electrons for Xe: $28-24=4$ , i.e., two lone pairs placed on Xe. Check Xe's total: 4 bonding pairs (8 electrons) + 2 lone pairs (4 electrons) = 12 electrons, not 8. This is a genuine expanded octet — Xe, a period-5 element, is not restricted by the strict $n=2$ orbital-counting argument (only four valence orbitals) that makes 8 special for C–F. With accessible low-lying 5d-character-mixed or, in a more modern rationale, three-center-four-electron bonding, period 3+ (and heavier) central atoms can comfortably exceed 8 electrons. Total electron count check: 4 bonding pairs ( $8$ e⁻) + 2 Xe lone pairs ( $4$ e⁻) + 12 F lone pairs ( $24$ e⁻) $=8+4+24=36$ , matching $N_{\text{total}}$ . The two lone pairs on Xe, arranged opposite each other in an octahedral electron-pair geometry, are exactly why XeF₄ is square planar rather than tetrahedral — a direct handoff to VSEPR Theory Fundamentals.

Limiting cases, scope & common traps

Situation	What happens to the octet rule	Representative example
Central atom is period 2 (Li–F)	Strict octet rule holds almost without exception	CH₄, NH₃, H₂O, HF
Central atom has fewer than 8 valence electrons available even after bonding	Electron-deficient center, stable "incomplete octet"	BF₃, BCl₃ (6 electrons on B)
Central atom is period 3 or below with accessible extra bonding	Expanded octet, 10 or 12 electrons	PCl₅ (10 e⁻ on P), SF₆ (12 e⁻ on S)
Odd total valence electron count	No structure can pair all electrons; one atom is left with an unpaired electron (radical)	NO (11 e⁻ total), NO₂, ClO₂
Hydrogen or helium as terminal atom	Duet rule (2 electrons), never an octet	H₂O, CH₄, all hydrides
Ionic charge present	Add/subtract electrons before starting the algorithm	NO₃⁻, NH₄⁺, SO₄²⁻

The chemist behind it

Origins · 1 minGilbert Newton Lewis and how this was discovered

Gilbert Newton Lewis

1875–1946 · American physical chemist, University of California, Berkeley

Lewis trained at Harvard and spent most of his career transforming Berkeley's chemistry department into a world center for physical chemistry, contributing foundational work on thermodynamics (chemical activity, free energy), acid–base theory (the Lewis acid/base definition, generalizing Brønsted–Lowry to electron-pair donors and acceptors), and the theory of the chemical bond. His 1916 paper "The Atom and the Molecule" introduced the shared electron-pair bond and the cubical/dot notation for valence electrons that still bears his name, more than a decade before quantum mechanics gave that sharing a rigorous orbital foundation. Lewis was nominated for the Nobel Prize numerous times but never received it, a fact regarded by many historians of chemistry as one of the award's notable omissions; he died in his Berkeley laboratory in 1946, apparently during an experiment with hydrogen cyanide.

History

- 1875 — Gilbert N. Lewis born in Weymouth, Massachusetts.

- 1916 — publishes "The Atom and the Molecule," introducing the shared electron-pair bond and dot structures.

- 1919 — Irving Langmuir extends and popularizes Lewis's ideas, coining "octet rule" and "covalent bond" terminology in wide use today.

- 1923 — Lewis publishes Valence and the Structure of Atoms and Molecules, and separately proposes the general Lewis acid–base definition.

- 1946 — Lewis dies in his laboratory at Berkeley.

Why do some atoms "want" eight electrons instead of some other number?

Because second-period atoms (C, N, O, F, and their neighbors) have exactly four valence orbitals available — one 2s and three 2p — and each orbital holds at most two electrons with opposite spin. Four orbitals times two electrons gives eight, the maximum that can be accommodated without reaching into the next principal shell (3s, which lies much higher in energy). Filling all four orbitals, whether with bonding pairs or lone pairs, leaves no partly occupied, energetically low-lying orbital exposed for further reaction, which is what makes the octet configuration both a good bonding target and a good stability criterion. This orbital-counting argument is why the rule is essentially exact for period 2 but must be relaxed for period 3 and below, where more valence orbitals (including d-like or multicenter bonding pathways) are accessible.

How do you decide which atom goes in the center of the structure?

Pick the atom that is not hydrogen (hydrogen only ever forms one bond, so it must be terminal) and that has the lowest electronegativity among the remaining candidates. Less electronegative atoms are more willing to be surrounded by several bonds and to share electron density outward toward more electronegative neighbors, which is exactly what a central atom must do. In practice this usually just means: the atom written first in the formula, or the one closer to the left/bottom of the periodic table, is central — e.g., C in CO₂, N in NH₃, S in SO₄²⁻. When a formula has multiple non-hydrogen atoms of similar electronegativity (as in acetic acid, C₂H₄O₂), you build a multi-center skeleton and apply the same counting rules atom-by-atom.

What is the difference between a bonding pair and a lone pair, physically?

A bonding pair is a pair of electrons whose probability density is concentrated in the region between two nuclei, lowering the total energy of the system by simultaneously attracting both nuclei — this is the shared pair Lewis proposed as the essence of the covalent bond. A lone pair is a pair of electrons localized almost entirely around a single nucleus, contributing to that atom's completed valence shell without directly holding two atoms together. Lone pairs still matter enormously for structure and reactivity: they occupy space (driving bond-angle distortions under VSEPR theory), and they are the electron source in Lewis acid–base chemistry, hydrogen bonding, and much of transition-metal coordination chemistry.

Can a Lewis structure ever be "wrong" even if the total electron count is correct?

Yes. Getting the total electron count right is necessary but not sufficient — you must also place those electrons to satisfy both the connectivity (correct skeleton) and, as far as possible, the octet/duet criterion, and among multiple electron-count-correct arrangements you should prefer the one(s) with the most favorable formal charges (developed fully under Formal Charge and Resonance). For example, drawing NH₃ with the lone pair split as two "half-pairs" on two different hydrogens, instead of one full lone pair on nitrogen, would use the correct total of 8 electrons but violate both the duet rule for hydrogen and the octet rule for nitrogen, and is not a valid Lewis structure even though the raw count matches.

Why does the octet rule fail for BF₃ but not for CF₄?

Boron has only 3 valence electrons (group 13) and forms only 3 bonds to fluorine, contributing just 6 electrons around itself — there simply are not enough electrons in the B + 3F system to also give boron a lone pair without leaving one fluorine electron-deficient instead, and moving a fluorine lone pair in to form a B=F double bond would place a very unfavorable positive formal charge on the highly electronegative fluorine. Carbon, in contrast, has 4 valence electrons and forms 4 bonds to fluorine, exactly filling its own octet with four bonding pairs and no charge-separation penalty. BF₃'s incomplete octet at boron is precisely why the molecule is a strong Lewis acid, avidly accepting an incoming lone pair (e.g., from NH₃ or F⁻) to complete boron's octet in the product adduct.

Do expanded octets in SF₆ or PCl₅ really mean sulfur or phosphorus use d orbitals?

The traditional teaching explanation invoked empty 3d orbitals on sulfur or phosphorus to "host" the extra electron pairs, but modern computational studies show 3d orbital participation is minimal; the energetically dominant picture instead uses three-center, four-electron bonding (delocalized σ bonds spanning three atoms with four electrons, none of them purely a simple 2-electron 2-center bond) or substantial ionic/charge-transfer character in the bonding description. Regardless of the precise orbital picture, the empirical, Lewis-structure-level bookkeeping — more than 8 electrons formally "around" the central atom — remains an accurate and useful description of connectivity and reactivity, which is why it is retained at the introductory level even as the deeper bonding picture has been refined.

How does the octet rule connect forward to resonance and VSEPR?

Whenever a correct electron-counting procedure allows a multiple bond to be placed on more than one equivalent terminal atom (as with the three oxygens of NO₃⁻), you get several equally valid Lewis structures called resonance contributors, whose true electronic structure is a blend of all of them — this refinement is the subject of Formal Charge and Resonance. Separately, once you know how many bonding pairs and lone pairs surround a central atom from Lewis-structure electron counting, that same total number of electron domains is the direct input to VSEPR theory, which predicts three-dimensional molecular shape from electron-pair repulsion. In both directions, the Lewis structure is the load-bearing first step that everything else in structural inorganic chemistry builds on.