Chemical Bonding: The Ionic and Lewis Picture

In 1916, Kössel and Lewis independently worked out the first genuinely satisfying explanation of chemical valence, grounded in one striking observation: highly electronegative halogens and highly electropositive alkali metals sit on opposite sides of the periodic table, separated by the chemically inert noble gases, and both groups form ions by gaining or losing exactly enough electrons to match a noble gas's own electron count. Lewis pictured an atom as a positively charged 'kernel', the nucleus plus inner electrons, surrounded by an outer shell that could hold at most eight electrons, imagined sitting at the eight corners of a cube. A single outer electron, as in sodium, occupies just one corner; a noble gas fills all eight. This complete octet represents a particularly stable arrangement, and atoms achieve it either by transferring electrons entirely, forming ions, or by sharing them, forming covalent bonds.

G.N. Lewis introduced a simple, powerful notation: Lewis symbols, where dots placed around an element's symbol represent its valence electrons, the electrons in the outermost shell that actually participate in bonding, since inner-shell electrons are well shielded and generally uninvolved. The number of dots directly gives an element's group valence, either that number itself or eight minus that number. Building on this, Lewis dot structures represent entire molecules: each bond is a shared electron pair, drawn as a pair of dots or a line between two atoms, and each combining atom contributes at least one electron to that shared pair. A specific set of rules governs constructing these structures correctly: total up every combining atom's valence electrons (adjusting for any overall ionic charge), sketch a reasonable skeletal arrangement with the least electronegative atom generally at the centre, then distribute electrons as shared bonding pairs first and remaining lone pairs second, checking that every atom, wherever possible, ends up with a complete octet.

In a polyatomic ion, the overall charge belongs to the whole species, not to any single atom, yet it's genuinely useful to assign each atom a formal charge, calculated as its valence electrons in the free atom, minus its non-bonding (lone pair) electrons, minus half its bonding (shared) electrons. Take the middle oxygen in ozone (O3): starting with 6 valence electrons, subtracting 2 lone-pair electrons, subtracting half of its 6 bonding electrons (3), gives a formal charge of +1. This isn't a claim about real, physically separated charge within the molecule; it's purely a bookkeeping convention for tracking valence electrons and, crucially, for choosing between multiple possible Lewis structures for the same species, since the structure with the smallest formal charges, closest to zero, is generally the lowest-energy, most realistic one.

The octet rule is useful, especially for second-period elements, but it's not universal, and three genuine categories of exception exist. Some central atoms simply can't reach eight electrons, since they don't have enough valence electrons or bonding partners to begin with; lithium, beryllium, and boron, with only 1, 2, and 3 valence electrons respectively, form compounds like LiCl, BeH2, and BCl3 with an incomplete octet. A small set of molecules have an odd total electron count, making a complete octet on every atom mathematically impossible, as in nitric oxide (NO) and nitrogen dioxide (NO2). And elements from the third period onward have accessible d orbitals in addition to s and p, letting them accommodate more than eight electrons around a central atom entirely, an expanded octet, seen in PF5, SF6, and H2SO4. Beyond these three categories, the octet rule also says nothing at all about a molecule's actual shape or its relative energetic stability, real gaps that later theories were built to close.

Ionic bond formation depends on two things: how readily positive and negative ions form from neutral atoms, and how those ions arrange themselves in the resulting solid lattice. Forming a cation always costs energy (ionization enthalpy is always endothermic), while forming an anion may release or cost energy (electron gain enthalpy). Surprisingly, for many real ionic compounds, the sum of these two steps is actually unfavourable overall: sodium's ionization enthalpy is +495.8 kJ/mol, while chlorine's electron gain enthalpy is only -348.7 kJ/mol, a net cost of +147.1 kJ/mol just to form the separated gaseous ions. What makes NaCl form anyway is the third, decisive step: packing those ions into a crystal lattice releases a huge amount of energy, NaCl's lattice enthalpy is -788 kJ/mol, more than enough to cover the earlier net cost. This is the real, complete explanation for ionic stability, not simply 'atoms want a full octet', but a genuine energy balance across ionization, electron gain, and lattice formation together.

Bond length is the equilibrium distance between two bonded nuclei, measured experimentally by spectroscopic, X-ray, or electron-diffraction methods, and each atom contributes its own share, called its covalent radius, half the distance between two identical atoms joined by a single bond; chlorine's Cl-Cl bond of 198 pm gives each chlorine a covalent radius of 99 pm. A separate radius, the van der Waals radius, describes an atom's overall non-bonded size instead, always larger than its covalent radius, since it reflects the atom's full electron cloud rather than the tighter overlap region of an actual bond. Bond angle, meanwhile, is the angle between two bonding orbitals around a central atom, measured spectroscopically, and it's genuinely informative: water's H-O-H bond angle of 104.5 degrees, rather than a perfect 109.5 or 180 degrees, is itself a direct clue to how oxygen's orbitals and lone pairs are actually arranged in space.

Bond enthalpy measures the energy needed to break one mole of a specific bond in the gas phase; hydrogen's H-H bond enthalpy is 435.8 kJ/mol, while a bond between fluorine atoms takes only 155 kJ/mol, despite both being simple single covalent bonds, since the actual strength depends on far more than just bond type. Bond order, the number of shared electron pairs between two atoms, correlates directly with both bond length and bond enthalpy: a higher bond order means a shorter, stronger bond, and a lower bond order means a longer, weaker one. Nitrogen's N#N triple bond (bond order 3) has one of the highest bond enthalpies known for any diatomic molecule, 946 kJ/mol, precisely because three shared pairs pull the two nuclei together far more strongly than one pair could. Isoelectronic species, sharing an identical total electron count, also share an identical bond order: F2 and O2^2- both carry bond order 1, while N2, CO, and NO+ all carry bond order 3.

Ozone's two oxygen-oxygen bonds are experimentally identical, both measuring 128 pm, yet any single Lewis structure for O3 has to draw one O-O bond as a single (148 pm) and the other as a double (121 pm), a genuine mismatch with reality. The concept of resonance resolves this: when no single Lewis structure captures a molecule accurately, several structures of similar energy, called canonical structures, are considered together, and the real molecule is understood as a resonance hybrid, a genuine average of all of them, not a molecule flickering between different forms over time. Ozone's two canonical structures average out to two identical, intermediate O-O bonds, exactly matching experiment. The carbonate ion (CO3^2-) shows the same idea even more clearly: its three canonical structures, each with two single C-O bonds and one double, average to three genuinely identical carbon-oxygen bonds, and this resonance hybrid always sits at lower energy than any single canonical structure alone, providing real extra stability.

No real bond is ever perfectly, purely covalent or purely ionic; even hydrogen's own H-H bond carries a trace of ionic character. When two identical atoms bond, as in H2 or Cl2, the shared electron pair sits exactly between them, a nonpolar covalent bond. When two different atoms bond, the pair shifts toward whichever atom is more electronegative, as with fluorine in HF, creating a polar covalent bond. This charge separation gives the molecule a dipole moment, the product of the separated charge's magnitude and the distance between its positive and negative centres, measured in Debye units and treated as a vector, pointing from positive toward negative. For molecules with more than two atoms, the overall dipole moment is the vector sum of every individual bond dipole, which is why water, bent at 104.5 degrees, has a substantial net dipole moment (1.85 D), while linear, symmetric CO2 has a net dipole moment of exactly zero, since its two equal, opposite bond dipoles cancel completely.

Just as covalent bonds carry some ionic character, ionic bonds carry some covalent character too, and Fajans worked out the rules governing how much. A small, highly charged cation, and a large, easily distorted anion, push an ionic bond toward greater covalent character, since the cation's concentrated positive charge pulls and distorts, or polarises, the anion's more diffuse electron cloud, building up genuine electron density between the two nuclei, exactly what happens in an ordinary covalent bond. Cations with a transition-metal-style (n-1)d^n ns^0 configuration polarise an anion more strongly than cations with a simple noble-gas-style ns^2 np^6 configuration of the same size and charge, since d electrons shield the nucleus less effectively. This is a genuinely useful predictive tool: it explains why, for instance, lithium's compounds, built from a small, highly charged Li+ cation, show noticeably more covalent character than the corresponding compounds of the larger alkali metals below it.