Kossel-Lewis approach
Why does sodium never sit alone next to chlorine without reacting, and why does one nitrogen atom hold hands with exactly three hydrogens in ammonia, not two or four? In 1916, working independently, Walther Kössel and Gilbert Lewis answered both questions with the same idea: atoms react because they are chasing the electronic configuration of the nearest noble gas — usually eight electrons in the outer shell. Kössel pictured atoms transferring electrons outright to become ions; Lewis pictured atoms sharing electron pairs instead. Together their reasoning, the Kössel–Lewis approach, is the doorway into the whole unit on chemical bonding: before VSEPR shapes or hybridisation can make sense, you need this octet-driven logic and the dot-and-line bookkeeping — Lewis structures and formal charge — that lets you judge whether a proposed structure is reasonable.
Atoms bond — by fully transferring electrons (Kössel's ionic route) or by sharing electron pairs (Lewis's covalent route) — because both routes let each atom attain a stable, noble-gas-like octet (or duplet for H).
The core idea
Why noble gases are the benchmark. Helium, neon, argon, krypton and xenon are chemically inert, and each (except He, with 2) carries eight electrons in its outermost shell. Kössel and Lewis both reasoned backwards from this fact: if eight outer electrons make an atom unreactive, then any atom short of eight "wants" to get there, and chemical bonding is simply atoms rearranging their electrons to reach that stable count. This became the octet rule — the organising idea of the entire unit, even though (as the traps below note) it has well-known exceptions.
Kössel's route — electron transfer. Metals on the left of the periodic table have low ionisation enthalpy, so they lose their few valence electrons easily; non-metals on the right have highly negative electron gain enthalpy, so they accept electrons readily. When a metal atom hands its electrons to a non-metal atom, both end up with noble-gas configurations — the metal as a cation, the non-metal as an anion — and the oppositely charged ions attract electrostatically. This electrostatic attraction between ions of opposite charge is the ionic (electrovalent) bond.
Lewis's route — electron sharing. When the electronegativity difference between two atoms is small, neither can strip an electron cleanly from the other. Instead, the two atoms share one or more electron pairs, and that shared pair counts toward both atoms' octets simultaneously. This is the covalent bond. Lewis represented an atom's valence electrons as dots placed around its symbol (a Lewis dot symbol), with a shared pair shown either as two dots between the atoms or as a single line.
Building and checking a Lewis structure. Once bonding pairs are placed so that every atom (except H, which needs only a duplet) has eight electrons around it, any electrons left over are assigned as lone pairs on the most electronegative atoms. When more than one arrangement seems plausible, chemists use formal charge to pick the better one:
where: is the number of valence electrons the free (isolated) atom would have, is the number of non-bonding (lone-pair) electrons it carries in the structure, and is the number of bonding electrons around it, so is the count of pairs it "owns" by equal sharing. The structure with formal charges closest to zero, and any negative formal charge sitting on the more electronegative atom, is judged the more acceptable one — this is a bookkeeping rule of thumb, not a claim about the true charge distribution.
The goal is to show, using nothing beyond electron counting, why a Lewis structure with roughly zero formal charge on every atom is preferred over an alternative structure that piles up charge, so that when two skeletons both satisfy the octet rule you have a way to choose between them rather than guessing.
Step 1. Fix the total electron budget. Add up the valence electrons contributed by every atom in the formula, then add one electron for each unit of negative charge on the species or subtract one electron for each unit of positive charge. This total, divided by two, is the number of electron pairs you have to place as bonds and lone pairs.
Step 2. Place bonding pairs to complete every atom's octet (or duplet for hydrogen), and pattern-check that the total pairs used exactly matches the budget from Step 1 — this is a conservation check, not a free choice.
Step 3. Compute the formal charge on each atom individually with , where is that atom's free-atom valence electron count, its lone-pair electrons in the structure, and its bonding electrons. Do this for every atom in the structure.
Step 4. Sum the formal charges over all atoms; the sum must equal the overall charge on the species — this is a mandatory check, and a mismatch means an arithmetic slip in Step 2 or Step 3, not a real alternative structure. Among structures that pass this check, prefer the one whose individual formal charges sit closest to zero, with any unavoidable negative formal charge placed on the most electronegative atom present.
Worked example
Draw the Lewis structure of the ammonium ion, , and use formal charge to confirm it is reasonable.
Step 1. Count valence electrons. Nitrogen (Z = 7) contributes 5; each of the four hydrogens contributes 1, giving electrons from the atoms. The ion carries a charge, meaning one electron has effectively been removed, so the usable total is electrons, i.e. 4 electron pairs.
Step 2. Place nitrogen at the centre and use all 4 pairs to form four N–H single bonds. Nitrogen now has no lone pair left, and each hydrogen has a duplet (2 electrons) through its one bond — every atom's octet or duplet requirement is satisfied, and all 4 pairs from Step 1 are accounted for.
Step 3. Formal charge on N: (free-atom valence electrons of nitrogen), (no lone pair), (four bonds, 2 electrons each). Formal charge . Formal charge on each H: , , , so formal charge .
Step 4. Sum: , which matches the actual charge on exactly, so the structure passes the check. The Lewis structure places the entire formal positive charge on nitrogen, with all four N–H bonds identical once formed — even though one of them originates as a coordinate bond, where the lone pair of is donated to , all four are indistinguishable in the finished ion.
Competitive stretch
For the carbonate ion , one skeleton places one C=O double bond and two C–O single bonds. Formal charge on the double-bonded O is ; on each singly-bonded O it is ; on C it is . Sum , matching the ion's charge, with the negative formal charges correctly on the electronegative oxygens.
Common traps
- Treating the octet rule as an absolute law rather than a first approximation — species such as (an incomplete octet on B) and (an expanded octet on P) are stable exceptions, studied separately once bonding basics are secure.
- Confusing formal charge with the actual (partial) charge on an atom — formal charge assumes bonding electrons are shared exactly equally, which is only true when both atoms are identical; real charge distribution depends on electronegativity difference.
- Forgetting hydrogen and helium need only a duplet (2 electrons), so don't force an octet onto every atom when counting stable configurations or lone pairs.
- Miscounting the electron budget for an ion — add electrons for a negative charge and subtract for a positive charge before dividing into pairs, and always re-check with the Step 4 sum rule above.
Ask the tutor: Want to see how the Kössel–Lewis dot structures you just drew connect to VSEPR theory and the actual 3-D shapes molecules take?
Why did Kossel and Lewis both pick eight electrons as the target, and not some other number?
They were reading the evidence, not choosing arbitrarily. The noble gases — helium aside — all happen to have eight electrons in their outermost shell, and they are strikingly unreactive. Since chemical activity elsewhere in the periodic table clearly aims toward that same inertness, eight became the natural target for the outer shell of other atoms too. Helium's exception (a duplet, two electrons) simply reflects that its very first shell is full at two, not eight, so hydrogen mimics helium's two-electron stability rather than a fictitious eight.
If a bond forms by sharing electrons, why does formal charge sometimes come out non-zero even when the octet is satisfied?
Formal charge is a bookkeeping convention, not a description of where electrons actually sit. It is calculated by pretending every bonding pair is split exactly in half between the two atoms, regardless of which atom is more electronegative. In species like or , the overall ionic charge has to live somewhere once you divide electrons this way, so one or more atoms are assigned a non-zero formal charge even though every atom's octet is complete. The formula does not claim that atom truly holds that much charge — only that the bookkeeping puts it there.
How do I decide whether to use the Kossel (ionic) picture or the Lewis (covalent) picture for a given pair of atoms?
Look at the electronegativity difference. When one atom is a metal with low ionisation enthalpy and the other a non-metal with strongly negative electron gain enthalpy — a large gap, as between sodium and chlorine — complete electron transfer is favourable and the Kossel ionic picture applies. When the two atoms have close electronegativities, as between the chlorines in or carbon and hydrogen in most organic molecules, neither can pull an electron away from the other, so they share pairs instead and the Lewis covalent picture applies. Real bonds often sit between the two pure pictures, but this difference decides which one you draw first.