Geochemistry

How Dead Plankton Become Crude Oil

A barrel of crude oil is a cemetery. Roughly 23 tonnes of ancient marine phytoplankton and algae had to die, sink, and be buried to make a single 159-litre barrel — and it took them somewhere between 50 and 300 million years, buried 2 to 4 kilometres down where temperatures hold steady between about 60 and 120 °C. What comes out of the drill is a molecular fossil: the carbon skeletons of Emiliania huxleyi and its relatives, cooked slowly, stripped of oxygen and nitrogen, and re-arranged into a soup of hydrocarbons.

The chemistry that turns diatom lipids into gasoline is not combustion, not fermentation, and not any single reaction — it is thermal cracking of kerogen, a kinetically controlled maturation that runs so slowly that geologists model it with Arrhenius kinetics over geological time. Get the temperature 20 °C too high and you burn the oil to gas; too low and it never forms. This narrow band is called the oil window, and every field on Earth is a place where organic carbon happened to sit inside it long enough.

  • FeedstockMarine algae/plankton lipids → kerogen (C, H, O, N, S)
  • Oil window≈ 60–120 °C (2–4 km depth)
  • Key processThermal cracking of kerogen (catagenesis)
  • Timescale10⁶–10⁸ years
  • Maturity gaugeVitrinite reflectance %Rₒ ≈ 0.6–1.35
  • Buried fraction≈ 0.1% of marine primary production

Interactive visualization

Press play, or step through manually. The visualization is yours to drive — try it before reading on.

Open visualization fullscreen ↗

Watch the 60-second explainer

A condensed visual walkthrough — narrated, captioned, under a minute.

It starts with a rain of dead algae — and a lack of oxygen

The story begins in the sunlit top 100 m of the ocean, where phytoplankton fix inorganic carbon into biomass following the Redfield ratio of roughly 106 C : 16 N : 1 P. Photosynthesis writes the net reaction as:

  • 106 CO₂ + 16 NO₃⁻ + HPO₄²⁻ + 122 H₂O + 18 H⁺ → C₁₀₆H₂₆₃O₁₁₀N₁₆P + 138 O₂

When these cells die, most of that organic matter is recycled on the way down — respired back to CO₂ by bacteria and zooplankton in the biological pump. Roughly 1% of surface primary production reaches the seafloor intact, and only about 0.1% of primary production ultimately escapes remineralisation to be buried. The single most important control on whether carbon survives is the absence of oxygen. In a normal oxygenated basin, aerobic bacteria strip organic matter down completely (ΔG ≈ −479 kJ per mol CH₂O respired). Survival requires an anoxic bottom water — a stagnant, stratified basin like the modern Black Sea, or an ocean-wide anoxic event. Under anoxia, the electron-acceptor ladder shifts to weaker oxidants — NO₃⁻, then Mn⁴⁺/Fe³⁺, then SO₄²⁻ (sulfate reduction, the dominant marine anaerobic pathway), and finally methanogenesis. Each step yields less energy, decomposes organics more slowly, and lets more carbon slip into the sediment.

Diagenesis: building kerogen, the insoluble parent of all oil

Once buried in fine-grained mud, the organic debris enters diagenesis — the low-temperature (< ~50 °C), shallow reworking that happens in the first few hundred metres over the first few million years. Sulfate-reducing bacteria run reactions like:

  • 2 CH₂O + SO₄²⁻ → H₂S + 2 HCO₃⁻ (sulfate reduction; the H₂S then reacts with Fe²⁺ to precipitate pyrite, FeS₂)
  • CH₃COO⁻ + H⁺ → CH₄ + CO₂ (acetoclastic methanogenesis, producing shallow biogenic gas)

The labile molecules — sugars, most proteins — are consumed. What resists is the lipid fraction: algal cell-wall biopolymers, long-chain fatty acids, sterols, and pigments. Through condensation, defunctionalisation (loss of –OH, –COOH, –NH₂ groups as CO₂, H₂O, NH₃), and polymerisation, these fragments cross-link into kerogen — a huge, insoluble, three-dimensional macromolecule that is the true precursor of petroleum. Kerogen makes up over 90% of the organic carbon in sedimentary rocks and is the largest organic-carbon reservoir on the planet (~10¹⁶ tonnes C). Geochemists sort it by hydrogen richness on a van Krevelen diagram (H/C vs O/C atomic ratios):

  • Type I (H/C > 1.5): lacustrine algae — oil-prone, the richest source.
  • Type II (H/C ≈ 1.2–1.5): marine plankton — oil- and gas-prone; most petroleum source rocks.
  • Type III (H/C < 1.0): land-plant lignin — gas-prone (this is coal's cousin).

The oil window: thermal cracking under the Arrhenius clock

As burial continues and the geothermal gradient (typically 25–30 °C/km) raises the temperature past ~60 °C, kerogen begins to break apart thermally. This is catagenesis, and the temperature band from about 60 to 120 °C is the oil window. Chemically, it is homolytic cracking of C–C and C–heteroatom bonds via a free-radical chain, closely analogous to industrial thermal cracking but running at a thousandth of the temperature over a billion times the duration:

  • Initiation: R–CH₂–CH₂–R′ → R–CH₂• + •CH₂–R′ (a weak C–C bond, ~350 kJ/mol, breaks homolytically)
  • Propagation (β-scission): R–CH₂–CH₂–CH₂• → R–CH₂• + CH₂=CH₂; and H-abstraction R• + R″H → RH + R″•
  • Termination: R• + R′• → R–R′ (radical recombination)

The kerogen loses its heteroatoms and its cross-links, shedding smaller, hydrogen-rich, mobile molecules — first heavy asphaltenes and resins, then progressively lighter alkanes, cycloalkanes (naphthenes), and aromatics that make up crude oil. Because this is a first-order-ish thermal process, geochemists model it with the Arrhenius equation, k = A·e^(−Eₐ/RT), using activation energies Eₐ of roughly 200–250 kJ/mol distributed across many parallel bonds. The critical insight is the time–temperature trade-off: rock held at 100 °C for 100 million years generates the same oil as rock flashed briefly to a much higher temperature. Maturity is tracked by vitrinite reflectance (%Rₒ): oil generation runs from Rₒ ≈ 0.55 to 1.35%, peaking near 0.9%.

Cook it too long and you get natural gas

The oil window has a hard upper edge. Push past ~120–150 °C (Rₒ > ~1.35%) and you enter metagenesis, the gas window. Here the liquid hydrocarbons already generated are themselves cracked — every remaining C–C bond is a target, and cracking always drives toward the most thermodynamically stable small molecule, methane:

  • C₁₀H₂₂ (decane) → C₇H₁₆ + C₃H₆, and ultimately long chains → CₙH₂ₙ₊₂ → (n) CH₄ + carbon-rich residue

Cracking a big alkane into methane is entropically favoured (many small molecules from one large one, ΔS > 0) and becomes spontaneous (ΔG < 0) at high T. The carbon that methane can't carry off — because CH₄ is the most hydrogen-rich hydrocarbon per carbon — is left behind as pyrobitumen, a solid, insoluble, increasingly graphitic residue. Above ~200 °C the system is overmature: even methane is gone or destroyed, and the kerogen has been reduced to graphitic carbon. This is why very deep or very old, deeply buried basins hold only dry gas or nothing at all — the oil has been thermally destroyed. It also explains the depth zonation of a producing basin: shallow biogenic gas, then oil, then wet gas, then dry gas, then barren, top to bottom.

Migration and the trap: why oil pools where it does

Generating oil is only half the problem; oil forms inside a low-permeability source rock (an organic-rich shale or marl) where it can't be extracted. It has to migrate. Newly formed hydrocarbons are less dense than the pore water (oil ≈ 0.8–0.9 g/cm³ vs brine ≈ 1.0–1.2 g/cm³), so buoyancy drives them upward and sideways through connected pore space and micro-fractures — primary migration out of the source, then secondary migration through a permeable carrier bed like sandstone or fractured limestone. The journey ends at a trap: a geometry where an impermeable seal or cap rock (evaporite salt, dense shale, anhydrite) blocks further ascent. Classic traps include:

  • Structural traps — anticlines (upfolds) and fault-sealed blocks, where oil collects at the crest.
  • Stratigraphic traps — pinch-outs, reefs, and unconformities where reservoir rock is sealed by a lithology change.
  • Salt-dome traps — buoyant salt diapirs that pierce and drape strata (the Gulf of Mexico's signature).

In the reservoir, oil, gas, and water separate by density: a free gas cap on top, oil in the middle, and formation brine below, filling only the pore space of the rock — there is no underground 'lake' of oil. Without the coincidence of source, maturity, migration path, reservoir, seal, and trap all in the right place at the right time, no field forms. That is why petroleum is geographically rare and economically concentrated.

The planetary bookkeeping: a slow carbon leak we're reversing in a century

Petroleum formation is one arm of the long-term (geological) carbon cycle — the slow burial of reduced organic carbon that, over hundreds of millions of years, pulled CO₂ out of the atmosphere and left free O₂ behind (every carbon buried as kerogen instead of respired leaves one O₂ un-consumed). This organic-carbon burial, together with silicate weathering, is part of what has kept Earth's surface habitable. The scale is staggering but the rate is glacial: the sedimentary organic-carbon reservoir holds ~10¹⁶ t C, but it accumulated at a net rate equivalent to a tiny fraction of a percent of photosynthesis. Humans now burn conventional oil at ~100 million barrels per day (~4.5 billion tonnes/year), returning in a single year the carbon that took hundreds of thousands to millions of years to bury. Burning it reverses the original chemistry:

  • 2 C₈H₁₈ (octane, a proxy) + 25 O₂ → 16 CO₂ + 18 H₂O, ΔH ≈ −5,470 kJ/mol octane

That is the crux of climate change: fossil-fuel combustion is the reversal of ~10⁸ years of biological pump and burial chemistry, run in ~200 years. Atmospheric CO₂ has climbed from ~280 ppm preindustrial to over 420 ppm, and roughly a third of that has dissolved into the ocean, driving surface pH down ~0.1 units (from ≈8.2 to ≈8.1) via ocean acidification. Petroleum is, quite literally, buried sunlight and buried carbon; using it un-buries both. Because kerogen maturation takes millions of years, oil is not renewable on any human timescale — the reservoirs we drain do not refill.

The three maturation stages: what happens to buried organic carbon as depth and temperature rise
StageTemperatureDominant chemistryMain product
Diagenesis< ~50 °CMicrobial decay, dewatering, loss of O & N; polymerisation to kerogenKerogen + biogenic CH₄
Catagenesis (oil window)≈ 60–120 °CThermal cracking of C–C bonds; H redistributionLiquid crude oil + wet gas
Metagenesis (gas window)≈ 120–200 °CFurther cracking; aromatisation; graphitisation of residueDry gas (CH₄) + pyrobitumen
Overmature> ~200 °CTotal cracking; kerogen → graphitic carbonOnly CH₄, then barren

Frequently asked questions

Does oil really come from dinosaurs?

No — that's a persistent myth. The overwhelming majority of crude oil derives from marine microorganisms: phytoplankton, algae, and bacteria, not large animals. Dinosaurs and land plants contributed almost nothing to oil (land-plant matter mostly makes coal and gas). The 'fossil' in fossil fuel refers to microscopic marine biomass buried in ocean and lake sediments.

How long does it actually take to form oil?

Tens to hundreds of millions of years. After burial, organic matter must sit within the oil window (roughly 60–120 °C) for long enough that Arrhenius-governed thermal cracking runs to completion — typically 10 to 100+ million years, because the reaction is extraordinarily slow at those temperatures. This is why oil is not renewable: no reservoir refills within human timescales.

Is there an abiotic (non-biological) origin of oil?

For commercial petroleum, no. The abiogenic hypothesis proposes hydrocarbons form from deep mantle carbon, and Fischer–Tropsch-type reactions during serpentinization can indeed make trace methane and light hydrocarbons abiotically. But essentially all economically produced oil carries unambiguous biological fingerprints — biomarkers like hopanes and steranes (derived from bacterial and algal sterols) and characteristic ¹³C/¹²C isotope ratios that only biology produces.

Why is the temperature window so narrow?

It's a Goldilocks problem set by kinetics. Below ~60 °C, kerogen cracking is too slow to generate oil even over geological time. Above ~150 °C, cracking runs so far that it destroys the liquid oil, converting it to methane and solid pyrobitumen. Liquid crude only survives in the band where cracking is fast enough to happen but not so fast that it overshoots — the oil window.

What made a rock a good source of oil versus not?

Three things: enough organic carbon (source rocks typically have >1–2% total organic carbon, the best over 10%), the right kerogen type (hydrogen-rich Type I/II from algae and plankton), and the right thermal history (burial into the oil window, tracked by vitrinite reflectance %Rₒ ≈ 0.6–1.35). Miss any one — too little carbon, land-plant kerogen, or over-cooking — and you get gas or nothing.

How is burning oil connected to ocean acidification?

Directly and by mirror image. Oil formed by removing CO₂ from the ancient atmosphere-ocean and burying its carbon. Combustion reverses that, returning CO₂ far faster than it was buried. About a quarter to a third of that CO₂ dissolves into seawater, where CO₂ + H₂O ⇌ H₂CO₃ ⇌ H⁺ + HCO₃⁻ releases protons, lowering ocean surface pH by ~0.1 units since preindustrial times.