Lattice Energy and Thermodynamic Stability of s-Block Salts

Lattice-energy reasoning underpins battery electrolyte design: predicting whether a proposed lithium or sodium salt (for solid electrolytes or liquid electrolyte solvation) will dissolve favorably in a given solvent requires balancing lattice energy against solvation energy, exactly the Born–Haber-style bookkeeping used here — this is central to modern lithium-ion and sodium-ion battery electrolyte formulation. In cement and glass chemistry, thermal stability trends of alkaline-earth carbonates and silicates (traced directly to size-mismatch lattice arguments) determine calcination temperatures in industrial lime and cement production. Pharmaceutical salt selection for drug formulation routinely uses lattice-energy/solubility trade-off reasoning to choose counter-ions (chloride vs. mesylate vs. tartrate) that balance crystallinity (needs sufficient lattice energy for a stable solid form) against aqueous solubility (needs the lattice energy not to overwhelm hydration energy) — a direct descendant of the size-matching principle developed for simple s-block salts. In geochemistry and planetary science, Kapustinskii-type estimates predict the stability of exotic mineral phases (including hypothetical high-pressure ionic solids) where no experimental thermochemical data exist. Finally, the whole apparatus is the standard explanatory tool for the Anomalous Behaviour of Lithium and the even more extreme Anomalous Behaviour of Beryllium, both of which trace back to unusually small cationic radii producing disproportionately large lattice (and hydration/covalency) effects relative to their heavier congeners.

Ask the tutor: How would the Born–Haber cycle change if we used lattice energy (constant volume) instead of lattice enthalpy (constant pressure), and does the pV correction ever matter numerically?

An ionic solid is held together by the electrostatic attraction of a periodic array of point (or nearly point) charges. The starting point for quantifying this is the Madelung energy of an infinite lattice: sum the Coulomb interaction of one reference ion with every other ion in the crystal, positive contributions from unlike-charge neighbors and negative contributions (repulsive, same-sign) from like-charge ions further out. The infinite alternating sum converges to a structure-dependent constant $A$, the Madelung constant (1.7476 for rock-salt NaCl, 1.7627 for CsCl, 1.6381 for zinc blende). Multiplying by $N_A e^2/(4\pi\varepsilon_0)$ and dividing by the shortest cation–anion distance $r_0$ gives the purely electrostatic (attractive) part of the lattice energy. But point charges pushed arbitrarily close together would give infinite attractive energy — real ions have electron clouds that repel via Pauli exclusion at short range. The Born–Landé model adds a short-range repulsive term $B/r^n$, with $n$ (the Born exponent, typically 5–12, tabulated by ion type) representing the steepness of this repulsion. Minimizing the total energy with respect to $r$ (setting $dU/dr = 0$) eliminates the unknown constant $B$ and yields the compact Born–Landé formula quoted below. The key qualitative lesson survives every refinement: lattice energy magnitude increases with ionic charge and decreases with ionic size, because both the Madelung sum and the equilibrium separation $r_0$ are governed by these two variables. The Born–Haber cycle sidesteps the need for any structural model. It simply asserts that the enthalpy of formation of a salt from its elements can be reached by two different thermochemical paths — direct combination, or a stepwise route through gaseous atoms and ions — and since enthalpy is a state function, both paths give the same total. This is Hess's law dressed in ionic clothing. Because every step except the lattice enthalpy is independently measurable (sublimation/atomization enthalpies, bond dissociation energies, ionization energies, electron affinities), the cycle isolates $U$ as the one unknown, or conversely can isolate an otherwise hard-to-measure electron affinity if $U$ is estimated by a model like Kapustinskii's.

The Born–Landé equation gives lattice energy from purely electrostatic and repulsive contributions:

where: $N_A$ is Avogadro's number, $A$ is the Madelung constant (structure-dependent, dimensionless), $z_+,z_-$ are the (signed magnitude) ionic charges, $e$ is the elementary charge, $\varepsilon_0$ is the vacuum permittivity, $r_0 = r_+ + r_-$ is the equilibrium interionic distance, and $n$ is the Born exponent (≈5–12, from ion compressibility/electronic configuration).

Reading: the negative sign signals energy released on lattice formation (using the "formation from gaseous ions" sign convention — lattice energy is exothermic). The factor $(1-1/n)$ is a modest correction (typically 0.7–0.9) capturing how short-range repulsion softens the pure Coulomb attraction; it never changes the qualitative charge/size scaling. The Kapustinskii equation replaces the need for a known crystal structure (hence no Madelung constant lookup) with an empirical observation: the Madelung constant per ion pair is roughly proportional to the number of ions in the formula unit, $\nu$. This lets you predict $U$ for any ionic formula, real or hypothetical:

where: $\nu$ is the number of ions per formula unit (e.g. $\nu=2$ for NaCl, $\nu=3$ for CaCl₂ or Na₂O), $z_+,z_-$ are charge magnitudes, and $r_+ + r_-$ is in picometers (this specific numeric-constant form assumes radii in pm and gives $U$ in kJ mol⁻¹; an equivalent form exists for radii in Å giving eV).

Reading: because $\nu$ absorbs the structure-dependence that the Madelung constant would otherwise carry, this formula needs no crystallographic information at all — only ionic charges and a tabulated (Shannon–Prewitt-type) radius. It is therefore predictive even for compounds that have never been synthesized, and it is the standard tool for estimating lattice energies of "thermochemically forbidden" or hypothetical species (like NeCl or a fictitious Ar⁺ salt) used to explain why noble-gas halides don't form, or why Mg²⁺ + O²⁻ can't easily be replaced by a hypothetical MgCl in a stable lattice. The Born–Haber cycle, meanwhile, is not a new energy law but a bookkeeping identity from Hess's law:

where: $\Delta H_f^\circ$ is the standard enthalpy of formation of the solid salt MX from elements in their standard states, $\Delta H_{\text{atom}}$ are atomization (sublimation/dissociation) enthalpies of the elements to gaseous atoms, $IE(M)$ is the (sum of, if multiply charged) ionization enthalpy of the metal, $EA(X)$ is the (sum of, if multiply charged) electron-attachment enthalpy of the nonmetal, and $U$ is the lattice enthalpy (formation-from-ions convention, so it is the most negative term).

Reading: every term on the right except $U$ is measured by an independent spectroscopic or calorimetric technique; solving for $U$ turns a five-step cycle into a lattice-energy measurement — or, if $U$ is instead estimated via Kapustinskii, the same equation solves for an unmeasured $EA$.

Example 1 (foundational): Which is larger in magnitude, $U(\mathrm{LiF})$ or $U(\mathrm{LiI})$, and by how much (Kapustinskii)? Radii: $r(\mathrm{Li^+}) = 76$ pm, $r(\mathrm{F^-}) = 133$ pm, $r(\mathrm{I^-}) = 220$ pm. Both are 1:1 salts, $\nu=2$, $z_+z_-=1$. For LiF: $r_++r_- = 76+133 = 209$ pm.

For LiI: $r_++r_- = 76+220 = 296$ pm.

Difference: $U(\mathrm{LiF}) - U(\mathrm{LiI}) = -962-(-718) = -244$ kJ mol⁻¹ — LiF's lattice energy is 244 kJ mol⁻¹ more exothermic, driven entirely by F⁻ being far smaller than I⁻ (133 pm vs 220 pm) while paired with the same small Li⁺. This is the microscopic reason LiF is a high-melting, poorly soluble, hard refractory salt while LiI is comparatively soft and highly soluble. Limiting case: if F⁻ and I⁻ had the same radius (an artificial equal-anion-size limit), the two lattice energies would be numerically identical since $\nu$, $z_+z_-$, and $r_+$ are otherwise unchanged — confirming the entire 244 kJ mol⁻¹ gap traces to anion radius alone.

Example 2 (applied): Why does Li₂CO₃ decompose thermally far more readily than Na₂CO₃, using a lattice-energy argument? Thermal decomposition is $\mathrm{M_2CO_3(s) \to M_2O(s) + CO_2(g)}$. Write a thermodynamic cycle around this reaction using lattice energies of reactant and product:

The bracketed lattice-energy difference is the size-mismatch term: both $\mathrm{CO_3^{2-}}$ (large, ~178 pm effective radius) and $\mathrm{O^{2-}}$ (small, ~140 pm) are compared against the same cation. For a small cation like Li⁺ (76 pm), pairing with the small O²⁻ anion is disproportionately more stabilizing (large $|U(\mathrm{Li_2O})|$) relative to pairing with bulky carbonate, so $|U(\mathrm{Li_2O})| - |U(\mathrm{Li_2CO_3})|$ is a large positive number — meaning decomposition (forming the oxide) is strongly lattice-energy-favored. Using approximate Kapustinskii-type values: $U(\mathrm{Li_2O}) \approx -2799$ kJ mol⁻¹, $U(\mathrm{Li_2CO_3}) \approx -2523$ kJ mol⁻¹, giving a favorable gap of

For Na⁺ (102 pm, larger), the same comparison shrinks because a larger cation cannot exploit the small oxide radius as effectively — both lattice energies are smaller in magnitude and closer together, so the driving force for decomposition is weaker. This is exactly why lithium carbonate decomposes at a much lower temperature (~1300 K range onset, and in practice partial decomposition well below) than sodium carbonate, which is thermally very robust and melts largely intact. Limiting case: as the cation radius $r_+ \to \infty$ (very large cation, e.g. Cs⁺), the relative size mismatch between O²⁻ and CO₃²⁻ becomes negligible on the scale of $r_++r_-$, so $|U(\mathrm{M_2O})|-|U(\mathrm{M_2CO_3})| \to$ small, predicting (correctly) that Cs₂CO₃ is thermally far more stable than Li₂CO₃.

Example 3 (exam/competition level): Estimate the electron affinity of oxygen's second electron attachment, $EA_2(\mathrm{O})$, from a Born–Haber cycle on MgO, given $\Delta H_f^\circ(\mathrm{MgO}) = -602$ kJ mol⁻¹, $\Delta H_{\text{atom}}(\mathrm{Mg}) = +148$ kJ mol⁻¹, $IE_1(\mathrm{Mg})=+738$, $IE_2(\mathrm{Mg})=+1451$, $\tfrac12 D(\mathrm{O=O})=+249$, $EA_1(\mathrm{O})=-141$ kJ mol⁻¹, and a Kapustinskii-estimated $U(\mathrm{MgO}) = -3795$ kJ mol⁻¹ (using $r(\mathrm{Mg^{2+}})=72$ pm, $r(\mathrm{O^{2-}})=140$ pm, $z_+z_-=4$, $\nu=2$). First verify the Kapustinskii number to make sure it's usable: $r_++r_- = 72+140=212$ pm,

consistent with the given $-3795$ kJ mol⁻¹ within rounding — good, use it. Write the full cycle:

Insert numbers:

Sum the known terms on the right (excluding $EA_2$):

So:

This large positive value (endothermic!) is the classic and important result: adding a second electron to already-negative O⁻ costs a great deal of energy because the electron must approach a species that already repels it electrostatically. The reason MgO nonetheless forms is that the enormous lattice energy (−3795 kJ mol⁻¹, driven by $z_+z_-=4$ and small ionic radii) more than compensates for this unfavorable step — a textbook illustration that no isolated gas-phase O²⁻ ion is thermodynamically stable; O²⁻ only exists embedded in a strongly stabilizing lattice. Limiting case: if the lattice energy were instead only as large as a 1:1 salt's (say −800 kJ mol⁻¹, comparable to NaCl), the cycle would require $\Delta H_f^\circ$ to be strongly positive (unfavorable), correctly predicting that no stable "MgO-like" 2:2-charge oxide could exist with such a weak lattice — consistent with why high lattice energy is a precondition for any O²⁻-containing ionic solid to be viable at all.

Anatoli Kapustinskii, working in the Soviet Union in the 1930s–1950s, recognized the practical limitation of the Born–Landé approach: it requires a known crystal structure (to get the Madelung constant) before any prediction can be made — useless for hypothetical or newly proposed compounds. His 1943 empirical insight, that the Madelung constant per formula unit tracks closely with simple ion count regardless of detailed structure type, converted lattice energy estimation into an algebraic exercise needing only tabulated ionic radii. This made it possible to estimate, for example, whether a proposed but unsynthesized salt should be thermodynamically viable at all — a question central to explaining why certain oxidation states or stoichiometries are simply never observed in nature.