From shared pairs to a map on paper
In the last guide you saw the three idealised bonding models — ionic, covalent and metallic — and learned that a covalent bond is just a pair of electrons shared between two nuclei. That is a lovely idea, but a real molecule has many electrons, and you need a way to keep track of where they all sit. A [[inorg-lewis-structure|Lewis structure]] is exactly that: a back-of-the-envelope map invented by G. N. Lewis around 1916, in which every shared pair is drawn as a line between two atoms and every leftover pair sits as two dots on a single atom.
Two kinds of electron pairs live on a Lewis structure. A bonding pair sits between two atoms and glues them together; a lone pair belongs to one atom alone and points off into empty space. The whole point of the picture is that those two types behave very differently — bonding pairs hold the skeleton together, while lone pairs shove the bonds around and, as you will see in the very next guide, decide the molecule's shape. For now, just remember the bookkeeping: count the pairs, draw the lines and dots, and you have captured a molecule's electron arrangement on a single line of a notebook.
Why eight? The octet rule
Lewis noticed that main-group atoms keep gaining, losing or sharing electrons until they are surrounded by eight valence electrons — the same count the noble gases enjoy. This is the [[inorg-octet-rule|octet rule]], and it is the engine behind almost every Lewis structure you will ever draw. Eight is special for a simple reason: in a given shell, one s orbital plus three p orbitals hold two plus six equals eight electrons, and a filled s-and-p set is unusually low in energy and unreactive.
An atom can reach an octet two ways. It can transfer electrons outright — sodium loses one to look like neon, chlorine grabs one to look like argon, which is ionic bonding — or it can share them, so two chlorines split a pair and each one counts eight. Hydrogen is the standard, honest exception: it is content with just two electrons, a filled 1s, a duet rather than an octet, because its valence shell has no p orbitals to fill. So when you draw H, never hang dots on it expecting eight; two is its goal.
The octet rule works beautifully for the second-period workhorses — carbon, nitrogen, oxygen and fluorine — which is why every textbook teaches it first. In CO2, for instance, carbon forms a double bond to each oxygen; count the shared electrons and carbon sees eight, each oxygen sees eight, and everybody is happy. The trouble, as you will discover below, is that inorganic chemistry roams far beyond the second period, and out there the rule is a strong default rather than a law of nature.
Drawing one, step by step
Here is the recipe that turns a chemical formula into a Lewis structure. It is mechanical, and that is the beauty of it: do the same steps every time and you rarely go wrong. Let us run it on the nitrate ion, NO3-, a classic inorganic anion that also happens to show off resonance.
- Count all valence electrons. Add one for each unit of negative charge, subtract one for each positive charge. NO3- has 5 (from N) + 3 times 6 (from three O) + 1 (for the minus charge) = 24 electrons, that is 12 pairs.
- Pick the central atom — usually the least electronegative one (and never hydrogen). Here nitrogen sits in the middle with the three oxygens around it.
- Join everything with single bonds first. Three N-O single bonds use 3 pairs, leaving 9 pairs to place.
- Hand out the remaining pairs as lone pairs to the outer atoms first, filling their octets. Three oxygens taking three lone pairs each uses all 9 remaining pairs.
- Check the centre. Nitrogen now has only 3 bonds, that is 6 electrons — short of an octet. Fix it by sliding one lone pair off an oxygen into a double bond. Now N has a double bond plus two single bonds: 8 electrons, octet satisfied.
Now a question stares back at you: which of the three oxygens gets the double bond? Nothing in the molecule singles one out — all three N-O distances are measured to be identical. That is your cue for [[inorg-resonance|resonance]]: the real ion is not any one of the three drawings but a blend of all three, with the double-bond character smeared evenly over every N-O link. Resonance is not the molecule flickering between forms; it is an admission that one Lewis structure is too crude, and the truth lies in the average. A double-headed arrow between the structures is the standard shorthand.
Formal charge: which drawing is best?
Often a formula admits more than one valid octet structure, and you need a tie-breaker. That tool is [[inorg-formal-charge|formal charge]]: for each atom, compare how many valence electrons it owns in the structure with how many it has as a free atom. You assign every lone-pair electron to its atom and split each bonding pair evenly, then subtract that count from the free-atom valence number. The best structure is usually the one with formal charges closest to zero, with any negative charge sitting on the most electronegative atom.
formal charge = (valence e- of free atom) - (lone-pair e-) - (1/2 * bonding e-) NO3- best structure: N (double + 2 single bonds): 5 - 0 - 1/2(8) = +1 O (double-bonded): 6 - 4 - 1/2(4) = 0 O (single-bonded), x2: 6 - 6 - 1/2(2) = -1 each sum of formal charges = +1 + 0 + (-1) + (-1) = -1 (matches the ion's charge)
The honest exceptions
Now for the misfits, which are the most instructive part of the whole story. The first family is the [[electron-deficient-compounds|electron-deficient compounds]], and they live near the left of the p-block. Boron trifluoride, BF3, is the poster child: boron brings only three valence electrons, so even after three B-F bonds it is surrounded by just six — two short of an octet. There is no extra pair to scrape together, so boron simply does without. That empty space makes BF3 a hungry Lewis acid, eager to accept a lone pair from a partner like ammonia to form the adduct F3B-NH3 — the kind of boron-trihalide Lewis acid you will study in the p-block rung.
Boron pushes things even further in the boron hydrides, like diborane B2H6, where there are not even enough electrons to draw ordinary bonds. The molecule's trick is a three-centre two-electron bond: a single pair of electrons smeared over a B-H-B triangle, holding three atoms together with what would normally bond only two. No Lewis structure with neat lines can capture that — your pencil simply runs out of electrons — which is the first loud hint that lines-and-dots is an approximation hiding something richer underneath.
At the opposite extreme sit the [[hypervalency|hypervalent]] or expanded-octet species, common from the third period downward — PCl5 with five bonds, SF6 with six, XeF4 with four bonds plus two lone pairs. Sulfur in SF6 appears to carry twelve electrons, far past an octet. The old textbook excuse was that sulfur borrows its empty d orbitals (sp3d2 hybrids) to make room — but this is now seen as largely wrong: careful calculations show those d orbitals lie too high in energy to contribute much. The honest modern picture is delocalised bonding (three-centre four-electron bonds), where the central atom binds several very electronegative neighbours using mainly its s and p orbitals, with bonding pairs spread over more than two atoms. Note too that hypervalency never happens for the tiny second-period atoms — nitrogen can never make NF5 — simply because they are too small to fit five or six neighbours, d orbitals or not.
The third exception is the simplest to state and the hardest to draw: [[odd-electron-molecules|odd-electron molecules]], or radicals. Nitric oxide, NO, has 11 valence electrons — an odd number — so no matter how you arrange them, one electron must stay unpaired and at least one atom cannot reach a tidy octet. NO is a perfectly real, stable, biologically vital molecule (your blood vessels use it as a signal), yet the octet rule cannot even begin to describe it honestly. Wherever you see an odd electron count, a single Lewis structure has already lost.
Why the exceptions matter
Step back and notice the pattern. The octet rule is not failing because nature is messy; it is failing because lines and dots are too rigid. Each exception breaks the picture in a different way — too few electrons (boron), too many around one centre (sulfur), an unavoidable odd one out (NO) — but they all point to the same fix: bonding electrons are not always neatly localised in pairs between two atoms. Sometimes a pair is shared over three atoms, sometimes density is delocalised over the whole molecule.
That is precisely the door into the next track. Drawing-board Lewis structures and the VSEPR shapes you will build from them in the coming guides are wonderfully practical, but they cannot explain a radical's unpaired electron, or why molecular oxygen is magnetic. The cure is molecular orbital theory, which lets electrons spread over the whole molecule instead of pinning each pair between two atoms. The exceptions you just met are not annoyances to memorise — they are the cracks through which a deeper, quantum-mechanical model of bonding shines through.