Evaporites: How a Dried Sea Leaves Its Salt Behind

The chemistry of drawing salt out of solution

Seawater is a dilute ionic soup. By mass its major ions are chloride (Cl⁻ ≈ 19.4 g/kg), sodium (Na⁺ ≈ 10.8 g/kg), sulfate (SO₄²⁻ ≈ 2.7 g/kg), magnesium (Mg²⁺ ≈ 1.28 g/kg), calcium (Ca²⁺ ≈ 0.41 g/kg), potassium (K⁺ ≈ 0.40 g/kg), and bicarbonate (HCO₃⁻ ≈ 0.14 g/kg). None of these is anywhere near saturation in open ocean. Evaporation changes that by removing the solvent while conserving the solutes, so the concentration of every ion climbs in lockstep.

A salt precipitates the instant its ion activity product exceeds its solubility product Ksp. For a generic salt, that condition is written:

• Dissolution equilibrium: MX(s) ⇌ M⁺(aq) + X⁻(aq)
• Precipitation begins when the reaction quotient Q = [M⁺][X⁻] > Ksp

Because each mineral has its own Ksp, the salts fall out in a strict order — the least soluble first. This is fractional crystallization on a basin scale. Crucially, once a mineral starts precipitating it draws down its own ions and holds Q pinned near Ksp, so the brine composition marches down a well-defined path rather than jumping.

First out: the carbonate cap

The first solid to appear, at only ~1.8× concentration, is calcium carbonate. The relevant equilibria are:

• CO₂ + H₂O ⇌ H₂CO₃ ⇌ H⁺ + HCO₃⁻ ⇌ 2 H⁺ + CO₃²⁻
• Ca²⁺ + CO₃²⁻ ⇌ CaCO₃(s) (Ksp ≈ 3.3 × 10⁻⁹ for calcite)

As water leaves, both Ca²⁺ and dissolved inorganic carbon rise. Warming and CO₂ degassing push the carbonate equilibrium right (raising CO₃²⁻ and pH toward ~8.3–8.5), so calcite and its polymorph aragonite precipitate as a thin basal layer. Because calcium is scarce relative to the other ions (only 0.41 g/kg), this carbonate cap is quickly exhausted of new material — but it consumes some of the calcium budget, which matters for the next step.

Gypsum: the sulfate handshake

By ~3.5× concentration the brine saturates in gypsum, hydrated calcium sulfate:

• Ca²⁺ + SO₄²⁻ + 2 H₂O ⇌ CaSO₄·2H₂O(s) (Ksp ≈ 3.1 × 10⁻⁵ at 25 °C)

Gypsum is the great volume-maker of many evaporite sequences, laying down beds of pale, layered rock. Which calcium-sulfate phase forms depends on temperature and brine activity: below ~40 °C the dihydrate gypsum is stable, while hotter or more concentrated brines favor the anhydrous form anhydrite (CaSO₄), and the metastable hemihydrate (bassanite, CaSO₄·½H₂O — the same phase as plaster of Paris) can appear transiently. Because calcium (0.41 g/kg) is far scarcer than sulfate (2.7 g/kg), gypsum precipitation runs out of Ca²⁺ first, leaving the residual brine sulfate-rich — which is exactly why the last salts to form are magnesium and potassium sulfates, not just chlorides.

Halite: when the sea finally turns to rock salt

Sodium chloride is enormously soluble (~360 g/L, roughly 6 mol/L), so it holds off until the brine has been evaporated to about 10% of its original volume — a tenfold concentration:

• Na⁺ + Cl⁻ ⇌ NaCl(s)

At that point halite crashes out fast and abundant, because Na⁺ and Cl⁻ together make up more than 85% of seawater's dissolved solids. This is the thick, clear-to-milky rock salt that dominates commercial salt deposits and the great buried salt sheets. Halite is also mechanically weak and buoyant (density ≈ 2.16 g/cm³, less than overlying sediment), so once buried under kilometers of rock it flows plastically over geologic time and rises as salt domes (diapirs) — the very structures that trap much of the world's oil and gas.

The bitter end: potash and the K–Mg salts

After halite, only a tiny, dense, and chemically weird residue remains — the bittern brine, so named because magnesium salts taste bitter. Concentrated by roughly 60–70×, it finally saturates in the most soluble minerals, the ones that require nearly total desiccation:

• Epsomite: Mg²⁺ + SO₄²⁻ + 7 H₂O ⇌ MgSO₄·7H₂O(s)
• Sylvite: K⁺ + Cl⁻ ⇌ KCl(s)
• Carnallite: K⁺ + Mg²⁺ + 3 Cl⁻ + 6 H₂O ⇌ KMgCl₃·6H₂O(s)
• Kainite, polyhalite, bischofite (MgCl₂·6H₂O) and other complex phases

These are the potash ores mined for potassium fertilizer (the 'K' in N–P–K), so a fully desiccated sea is literally the source of agricultural potassium. Because carnallite and bischofite are deliquescent — they absorb atmospheric moisture and redissolve — they only survive geologically if buried and sealed quickly. Their presence in a core proves the basin was evaporated to near-dryness.

Why a whole sea has to dry: the basin machinery

Evaporating a beaker is easy; evaporating an ocean requires a restricted basin where evaporation outpaces inflow but a thread of seawater keeps topping the brine up. The two classic settings are:

• Sabkhas — coastal salt flats (like the Persian Gulf's Trucial Coast) where seawater soaks into flat sediments and evaporates in place, growing gypsum and anhydrite crystals within the mud.
• Barred marine basins — seas nearly cut off from the ocean by a sill, so fresh seawater seeps in over the barrier and replenishes the salt supply, allowing hundreds of meters of evaporite to stack up.

The scale can be staggering. During the Messinian Salinity Crisis (≈5.96–5.33 million years ago), the Mediterranean was cut off at Gibraltar and repeatedly desiccated, depositing roughly 1 million cubic kilometers of salt — locally over 2 km thick. Precipitating that much halite removed enough sodium and chloride to lower global ocean salinity by around 2‰, briefly reshaping seawater chemistry worldwide. To make just 15 m of pure halite you must evaporate a full 1000 m column of seawater, which is why thick beds demand continual replenishment.

Reading and reversing the record

Evaporite chemistry is a two-way street. Because the sequence is set by solubility, it is fully reversible: flood a salt flat and the minerals redissolve in reverse order — carnallite and halite go first, gypsum lingers, calcite is most stubborn. Geologists exploit this in reverse to read paleoclimate: the fluid inclusions trapped inside halite crystals are literally droplets of ancient seawater, and their chemistry records past ocean composition, temperature, even atmospheric CO₂ proxies going back hundreds of millions of years.

The sequence also shifts with the parent water. Modern seawater gives the carbonate→gypsum→halite→K-Mg path above, but ancient 'aragonite seas' versus 'calcite seas' (governed by the seawater Mg/Ca ratio, today ≈5.2) produced different early carbonates. And non-marine brines write different scripts entirely: soda lakes rich in Na⁺ and CO₃²⁻ precipitate trona (Na₃(CO₃)(HCO₃)·2H₂O) and natron, while boron- and lithium-rich closed basins like Chile's Salar de Atacama concentrate the brines now mined for lithium carbonate. In every case the rule is the same: evaporation drives Q past Ksp, and the least soluble salt always goes first.

The seawater evaporation sequence: what precipitates, when, and why

Mineral	Formula	Brine at onset (× seawater)	Approx. Ksp / solubility
Calcite / aragonite	CaCO₃	~1.8×	Ksp ≈ 3.3 × 10⁻⁹
Gypsum	CaSO₄·2H₂O	~3.5×	Ksp ≈ 3.1 × 10⁻⁵
Halite	NaCl	~10.6×	≈ 360 g/L (very soluble)
Epsomite / sylvite	MgSO₄·7H₂O / KCl	~60×	≈ 340–710 g/L
Carnallite (bittern)	KMgCl₃·6H₂O	~70×	highly soluble, deliquescent