Geochemistry
Limestone Formation: How the Ocean Turns CO₂ Into Stone
The White Cliffs of Dover are roughly 500 meters of dead plankton. Each cubic centimeter of that chalk holds the crushed skeletons of on the order of a billion coccolithophores — single-celled algae that, ~90 million years ago, pulled dissolved carbon out of a warm Cretaceous sea and locked it into calcium carbonate (CaCO₃). Multiply that across every reef, seafloor ooze, and ancient carbonate platform on Earth and you get the planet's largest carbon reservoir: roughly 60,000,000 gigatonnes of carbon stored as limestone and dolomite — about 1,500 times more than lives in the atmosphere, ocean, soils, and all life combined.
The chemistry that builds this stone is a chain of coupled equilibria that starts when CO₂ dissolves in seawater and ends when Ca²⁺ and CO₃²⁻ ions drop out of solution as a crystalline solid. Understanding it explains why the deep sea is littered with dissolving shells, why reefs are dissolving faster as we burn fossil fuels, and why — over millions of years — this single reaction acts as Earth's thermostat.
- Net reactionCa²⁺ + 2HCO₃⁻ ⇌ CaCO₃↓ + CO₂ + H₂O
- Main solidCaCO₃ (calcite / aragonite)
- Ksp (calcite, 25°C)≈ 4.5 × 10⁻⁹
- Surface Ω_aragonite≈ 3–4 (supersaturated)
- Seawater pH≈ 8.1 (−0.1 since 1750)
- Timescaledays (biology) → 10⁶–10⁸ yr (burial)
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
From gas to acid to ion: the carbonate ladder
Before any stone can form, gaseous CO₂ has to become a dissolved ion. When atmospheric CO₂ (≈ 420 ppm today) contacts the sea surface, a small fraction dissolves — its solubility set by Henry's law (K_H ≈ 3.3 × 10⁻² mol L⁻¹ bar⁻¹ at 25 °C). The dissolved gas then hydrates and ionizes down a four-species ladder:
- CO₂(g) ⇌ CO₂(aq) (dissolution)
- CO₂(aq) + H₂O ⇌ H₂CO₃ (hydration — slow, only ~0.3% converts)
- H₂CO₃ ⇌ H⁺ + HCO₃⁻ (pKₐ₁ ≈ 6.0 in seawater)
- HCO₃⁻ ⇌ H⁺ + CO₃²⁻ (pKₐ₂ ≈ 9.1 in seawater)
At the ocean's pH of ≈ 8.1, these constants dictate the mix: about 90% bicarbonate (HCO₃⁻), ~9% carbonate (CO₃²⁻), and only ~1% CO₂(aq). That partitioning matters enormously — the mineral needs the CO₃²⁻ end of the ladder, but the ocean keeps most of its carbon one rung down as HCO₃⁻. This is why the sea can hold ~50 times more carbon than the atmosphere: the carbonate buffer converts incoming CO₂ into charged ions that don't escape back to the air.
The precipitation reaction, balanced
Once carbonate ions are available, stone forms when they pair with calcium — the ocean's third-most-abundant cation (Ca²⁺ ≈ 10.3 mmol/kg, ~412 ppm). The clean way to write it is:
- Ca²⁺ + CO₃²⁻ ⇌ CaCO₃(s)
But because carbonate is scarce and bicarbonate is abundant, the reaction organisms and reefs actually run is the bicarbonate form, and it hides a crucial twist:
- Ca²⁺ + 2HCO₃⁻ ⇌ CaCO₃(s) + CO₂ + H₂O
Read that carefully: making one mole of solid carbonate releases one mole of CO₂. This is the great counterintuitive fact of marine geochemistry. Building a shell is not, in the short term, a carbon sink — calcification pushes CO₂ back toward the atmosphere. A coral reef is a net source of CO₂ to surface water on the timescale of the reaction. The permanent burial of carbon happens only because organic matter and mineral together sink and are entombed faster than they redissolve, and because on geological timescales the counterbalancing CO₂ is reconsumed by rock weathering (below).
Ksp and Ω: the on/off switch for stone
Whether CaCO₃ precipitates or dissolves is governed by one dimensionless number, the saturation state Ω:
- Ω = [Ca²⁺][CO₃²⁻] / K′_sp
Here K′_sp is the stoichiometric solubility product in seawater. The two common polymorphs differ: calcite (the stable form, trigonal) has K_sp ≈ 4.5 × 10⁻⁹ (25 °C), while aragonite (the metastable orthorhombic form corals and pteropods build) is more soluble at K_sp ≈ 6.5 × 10⁻⁹. When Ω > 1 the water is supersaturated and crystals can grow; when Ω < 1 they dissolve. Surface tropical water sits at Ω_aragonite ≈ 3–4 — thermodynamically primed to precipitate.
Yet, curiously, the open surface ocean does not spontaneously turn to chalk despite being 3–6× supersaturated. The kinetic barrier is real: Mg²⁺ ions (≈ 53 mmol/kg in seawater) poison calcite nucleation sites, and homogeneous nucleation needs Ω well above 10. This is why biology dominates: organisms use energy and specialized proteins to force precipitation against this barrier, templating crystals inside vesicles where they can control ion supply and exclude Mg²⁺.
Life does the hard part: biomineralization
The overwhelming majority of the ~1 gigatonne of CaCO₃ produced in the ocean each year is made by organisms, not by inorganic precipitation. The chief builders:
- Coccolithophores — algae that plate themselves in calcite scales; the source of chalk and much deep-sea ooze.
- Foraminifera — amoeboid protists with calcite tests; their shells drive most pelagic carbonate rain.
- Corals and coralline algae — aragonite reef framework built by symbiotic cnidarians.
- Pteropods, mollusks, echinoderms — aragonite and high-Mg calcite shells and plates.
Inside these organisms, calcifying fluid is actively pumped to raise its pH and CO₃²⁻ concentration — corals elevate the internal pH to ~8.5–9 (versus 8.1 seawater), driving Ω at the crystal site far above ambient. The organic carbon fixed by the same plankton (Redfield ratio ≈ 106 C : 16 N : 1 P) sinks alongside the mineral in the biological pump. When these particles fall below the depth where Ω < 1, the carbonate begins to redissolve — but a fraction reaches the seafloor and, over millions of years, compacts and lithifies into limestone.
The saturation horizon and the CCD
Because CO₃²⁻ falls with depth (cold water holds more CO₂, and respiration of sinking organic matter adds CO₂, both lowering pH and carbonate ion), Ω drops below 1 at a depth called the saturation horizon — around 500 m for aragonite in parts of the North Pacific, deeper for calcite. Below it, shells actively corrode. Deeper still lies the carbonate compensation depth (CCD), roughly 4,000–5,000 m, where the rate of CaCO₃ supply from above exactly matches its dissolution rate. The signature is stark on the seafloor: above the CCD, pale carbonate ooze blankets the bottom; below it, only red clay remains, every calcite shell dissolved away.
The pressure and temperature dependence is why the deep sea and the surface behave so differently. Higher pressure and colder temperature both increase CaCO₃ solubility (raise K_sp), pushing Ω down. The CCD is therefore a moving boundary — it shoaled dramatically during past ocean-acidification events, dissolving carbonate that had already reached the seafloor. Geologists read those dissolution horizons in the rock record as fingerprints of ancient carbon-cycle upheavals like the Paleocene–Eocene Thermal Maximum (~56 Ma).
Weathering closes the loop: Earth's thermostat
The calcium and alkalinity feeding marine carbonate come from land. Rain, mildly acidic from dissolved CO₂ (pH ≈ 5.6), attacks silicate rock in the Urey weathering reaction:
- CaSiO₃ + 2CO₂ + 3H₂O → Ca²⁺ + 2HCO₃⁻ + H₄SiO₄ (silicate weathering — consumes 2 CO₂)
- Ca²⁺ + 2HCO₃⁻ → CaCO₃ + CO₂ + H₂O (marine precipitation — releases 1 CO₂)
Sum the two and one net CO₂ molecule is removed from the atmosphere for every CaCO₃ buried. This coupled system is the silicate weathering thermostat: when CO₂ and temperature rise, weathering speeds up (warmer, wetter, faster reaction kinetics), pulling CO₂ down; when the planet cools, weathering slows and volcanic CO₂ rebuilds. Over ~10⁵–10⁶ year timescales this negative feedback has kept Earth's surface liquid-water-friendly for billions of years. Limestone is the ledger where the withdrawn carbon is recorded — and volcanoes and subduction-zone metamorphism (CaCO₃ + SiO₂ → CaSiO₃ + CO₂) eventually return it to the sky, completing the geological carbon cycle over ~200 million years.
Why it matters now: acidification and carbon storage
Human CO₂ emissions have added carbon to the surface ocean faster than weathering can supply new alkalinity. The result: surface pH has fallen from ≈ 8.2 to ≈ 8.1 since 1750 — a ~30% increase in H⁺ concentration — and CO₃²⁻ has dropped by roughly 10–20%. That directly lowers Ω. Under high-emission scenarios, large regions of polar and deep water tip to Ω < 1 by 2100, and the aragonite saturation horizon is already shoaling by meters per year in parts of the North Pacific and Southern Ocean, corroding pteropod shells that have been photographed pitting and dissolving in the wild.
The same chemistry is being harnessed deliberately. Enhanced weathering spreads crushed silicate rock (e.g., olivine, basalt) on land or coasts to accelerate the Urey reaction and boost ocean alkalinity, permanently storing CO₂ as bicarbonate and carbonate. Ocean alkalinity enhancement adds base (e.g., olivine or lime) directly to seawater to shift the carbonate ladder toward HCO₃⁻/CO₃²⁻ and re-supersaturate the water. Both are attempts to run, on a human timescale, the reaction the ocean has used for four billion years to turn CO₂ into stone.
| Property | Warm surface ocean | Cold deep ocean (>~4 km) |
|---|---|---|
| Saturation state Ω | > 1 (Ω ≈ 3–6, supersaturated) | < 1 (undersaturated) |
| CO₃²⁻ concentration | ~200–300 µmol/kg | ~50–80 µmol/kg |
| Pressure / temperature | ~1 bar, 20–30 °C | ~400 bar, 1–4 °C |
| Net CaCO₃ behavior | Precipitates (reefs, shells, ooze) | Dissolves back to Ca²⁺ + CO₃²⁻ |
| Result | Carbonate platforms, chalk, reefs | Clay-rich red seafloor, no carbonate |
Frequently asked questions
Is making a shell a carbon sink or a carbon source?
Counterintuitively, the precipitation reaction Ca²⁺ + 2HCO₃⁻ → CaCO₃ + CO₂ + H₂O releases CO₂. So calcification itself is a short-term CO₂ source to surface water. Carbon is only truly sequestered when the mineral is buried and, on geological timescales, when silicate weathering consumes the released CO₂ — netting one CO₂ removed per CaCO₃ permanently buried.
Why doesn't the whole supersaturated ocean just turn to limestone?
Surface water is 3–6× supersaturated (Ω_aragonite ≈ 3–4), but a kinetic barrier stops spontaneous precipitation. Magnesium ions (≈ 53 mmol/kg) poison calcite crystal nuclei, so homogeneous nucleation needs Ω well above 10. Organisms overcome this by using energy and proteins to template crystals internally — which is why biology, not chemistry alone, builds most marine carbonate.
What's the difference between calcite and aragonite?
They are polymorphs — same formula CaCO₃, different crystal structure. Calcite is trigonal and thermodynamically stable (K_sp ≈ 4.5 × 10⁻⁹); aragonite is orthorhombic and more soluble (K_sp ≈ 6.5 × 10⁻⁹). Corals and pteropods build aragonite, making them more vulnerable to acidification; foraminifera and coccolithophores build calcite. Over geologic time aragonite recrystallizes to calcite.
What is the carbonate compensation depth (CCD)?
It's the depth (~4–5 km) where the rate CaCO₃ falls from above exactly balances its dissolution, because deeper water is colder, higher-pressure, and CO₂-rich (Ω < 1). Above the CCD the seafloor is pale carbonate ooze; below it, only red clay survives because every shell dissolves. The CCD rises and falls with ocean chemistry, recording past carbon events in the rock.
How fast does limestone form, and is it reversible?
Biological precipitation takes days; a coral reef accretes millimeters to centimeters per year. Compaction and lithification into rock takes 10⁴–10⁸ years. The precipitation step is fully reversible — carbonate dissolves wherever Ω < 1, which is why the deep sea and acidifying surface waters actively corrode shells. Only deep burial makes the storage effectively permanent.
Can we use this chemistry to fight climate change?
Yes — that's the basis of enhanced weathering and ocean alkalinity enhancement. Spreading crushed silicate rock (olivine, basalt) accelerates the Urey reaction, consuming CO₂ and adding alkalinity that keeps carbon locked as HCO₃⁻ and CaCO₃. It works with, rather than against, the four-billion-year-old reaction, but faces cost, scale, and ecological hurdles before it can offset meaningful emissions.