The Making of Ammonia
From air and gas to the world's food

The Making of
Ammonia

The air above you is four-fifths nitrogen — but the atoms are bonded so tightly that turning them into fertiliser took a century of chemistry, and still accounts for roughly 2% of global energy use.

Chemical Agriculture Chemical Feedstock ≈ half the world's food depends on it

You almost certainly ate some of its work today.

Food Food
Refrigeration Refrigeration
Cleaners Cleaners
Explosives Explosives
Textiles Textiles
Clean fuel Clean fuel

The journey

Two atoms of air, split apart and rebuilt

Ammonia is made by tearing nitrogen out of the air and bolting it onto hydrogen, at pressures many times greater than a car tyre and temperatures hot enough to make steel glow red. The nitrogen is free and everywhere; the hard part is that its two atoms cling together so fiercely that only extreme conditions and a metal catalyst will persuade them to let go and grab hydrogen instead.

The five stages of making ammonia, from feedstocks to liquid ammonia 1 · Air + gas 2 · Reforming 3 · Cleaning up 4 · Haber–Bosch 5 · Ammonia out
Stage 1 of 5

Where it starts: methane and air

Two cheap feedstocks: natural gas (methane) for the hydrogen, and ordinary air for the nitrogen — air is 78% nitrogen (Royal Society).

Air ≈ 78% N₂
What it's made of

Three ingredients, and only one is scarce

The idea

Ammonia is built from just nitrogen and hydrogen — but where those come from decides everything about its cost and its carbon.

Nitrogen and hydrogen are the only two elements in ammonia (NH₃) — by weight it's 82% nitrogen, 18% hydrogen — yet almost the entire economic and environmental story lives in the third, invisible ingredient: the energy to get the hydrogen.

Nitrogen from air N₂

Nitrogen — from the air

Free, inexhaustible, 78% of every breath. In the usual gas route it's carried in with a blast of air during reforming — the oxygen burns off, the nitrogen stays; coal plants use a dedicated air-separation unit instead. Either way, separating nitrogen from air is relatively easy (Royal Society). The easy ingredient.

Hydrogen from natural gas H₂

Hydrogen — from natural gas (today)

Almost all of it stripped from methane by steam reforming — actually several reactors and purification stages in series — where ~90% of the carbon and most of the energy go (Royal Society). Ammonia is the single largest use of hydrogen on earth — about 33 of the world's ~94 Mt of hydrogen a year (Saygin). The hard, dirty ingredient — and the one everyone is trying to change. See the challenge →

Energy

Energy — the real feedstock

Making a tonne at best practice takes about 8 MWh (28 GJ; Royal Society), but the global-average plant uses about 11 MWh (41 GJ; IEA) — roughly a typical UK household's total energy for a year.

How it's made

The reaction that feeds the world — and why it's so hard to run

The idea

Nitrogen from the air is almost unreactive: its two atoms are triple-bonded, one of the strongest chemical bonds found in nature. Haber–Bosch beats that bond into submission with pressure and heat — a triumph of engineering, not just chemistry.

The problem. Air isn't already fertiliser because N₂ is held by a triple bond that plants and most chemistry can't break. For centuries, farming depended on the nitrogen that lightning and soil bacteria fixed naturally.

Nitrogen's triple bond, the energy barrier, broken and rebuilt into ammonia's three nitrogen–hydrogen bonds N N triple bond + 3H₂ NH₃ NH₃
Breaking one nitrogen triple bond (gold) is the whole difficulty; once open, each nitrogen atom grabs three hydrogens to become ammonia. Bond depiction after Royal Society, 2020

The solution. To break that bond at scale you need three things together — high pressure, high temperature, and a catalyst. The reactor runs at 150–300 bar and 350–500 °C over an iron catalyst (Royal Society). Only then does nitrogen let go and pair with hydrogen.

The reactions, in order
CH₄ + H₂O CO + 3H₂ Steam-methane reforming — strips hydrogen from gas (endothermic, ΔH = 206 kJ/mol)
CO + H₂O CO₂ + H₂ Water-gas shift — squeezes out more hydrogen, makes CO₂ to remove (exothermic)
N₂ + 3H₂ 2NH₃ Haber–Bosch — the famous, reversible step; only a fraction converts each pass

The consequence. Even under that assault the reaction is stubborn and reversible — only a fraction converts each pass — so the unreacted gas is recycled round the loop again and again. That recycling, and the pressure, are why ammonia plants are among the most energy-hungry factories on earth: about 8 MWh per tonne at best practice, about 11 MWh on the global average (Royal Society; IEA).

Fritz Haber found the reaction in 1909 and Carl Bosch turned it into a factory; between them the work drew two Nobel prizes. It is often called the invention that lets the planet feed itself.
Making the hydrogen — ~90% of the CO₂
synthesis
Steam reforming & gas cleanup (>80% of energy, ~90% of carbon) Haber–Bosch synthesis & the rest

The famous reaction isn't the dirty part. Cleaning up ammonia means cleaning up hydrogen. Source: Royal Society, 2020.

Where it goes

Mostly food — and then almost everything else

The idea

Most ammonia becomes fertiliser. The rest quietly runs through explosives, plastics, refrigeration and cleaning — most people meet ammonia's products daily without ever seeing ammonia.

Roughly 70–80% of ammonia becomes nitrogen fertiliser — sources draw the line differently, from about 70% (IEA) up to about 80% (Royal Society; IRENA; IFA). Of that fertiliser, about 55% becomes urea (Hydrogen Europe, citing IEA). The remainder spreads across industry: explosives (ammonium nitrate), nitric acid via the Ostwald process, plastics and synthetic fibres, refrigeration, and the ammonium sulphate that ties it to the sulphuric-acid story.

The gold-outlined slice marks the urea share within fertiliser — not a separate quantity. Shown as a band because sources disagree: IEA ~70/30, Royal Society / IRENA ~80/20 (in the US, USGS puts fertiliser nearer 88%). Share of ammonia demand by use — IEA / Royal Society, 2020–2023.

By the Numbers

The scale of the invisible industry

The idea

Ammonia is one of the most-produced chemicals on earth — second only to sulphuric acid — and one of the dirtiest to make: a paradox at the centre of the food system.

>190 Mt
Ammonia produced worldwide each year derived
USGS 2026 (160 Mt N × 1.216); IEA/IFA 185 Mt (2020)
~30%
Of it made in China alone — mostly from coal, the dirtiest route (IEA 2020)
−33 °C
Temperature at which ammonia stores as a liquid, so it ships like fuel (Royal Society)
~90%
Of the process's CO₂ comes from making the hydrogen, not the famous reaction (Royal Society)
1.3–1.8%
Share of global emissions from ammonia basis differs
IEA ~1.3% energy-CO₂ · RS ~1.8% global CO₂
~2.4 t
CO₂ per tonne of ammonia — about 2× steel, 4× cement, pound for pound (IEA)

Who makes it

The carbon twist: China makes ~30% of the world's ammonia but roughly 45% of its CO₂ — because most Chinese ammonia is made from coal, the dirtiest route. The country that feeds a third of the world's crops also runs the dirtiest ammonia. Bars scaled to the largest share; figures give the true percentages. Share of world production — USGS MCS 2026; China share corroborated by IEA (30%, 2020).

Cleaning up the hydrogen

The colours of ammonia

The idea

Ammonia's problem isn't the famous reaction — it's the hydrogen feeding it. Change where the hydrogen comes from and you can decarbonise nearly the whole industry; but the clean routes still cost more than natural gas, and that gap is the whole fight.

The "colours" of ammonia are really a story about the hydrogen (the Royal Society's framing). Same molecule, same Haber–Bosch reactor — what changes is where the hydrogen is torn from, and what that costs in carbon.

Brown / grey
today's default
~1.6 t
CO₂ per tonne (gas, direct); coal is worse, 2.5–3.8 t. About 90% of that carbon is in the hydrogen step. (Royal Society)
Blue
capture bolted on
−16–82%
The same process, with much of the CO₂ captured. How much climate benefit that delivers is genuinely debated — studies make different assumptions about methane leaks and capture efficiency (see methodology). A transition, not a destination.
Green
hydrogen from water
~99.7%
Split from water with renewable power; near-zero carbon. But ~85% of its cost is electricity, so it needs very cheap renewables (~$730/t vs grey ~$470/t). (RS; Saygin)
Carbon-negative
the surprising one
−1.8 kg
CO₂ per kg, if biomass (which absorbs CO₂ as it grows) is gasified and the carbon captured. Early and costly — but a heavy industry that could be a carbon sink. (Zhang et al., 2025)

One way to state the whole tradeoff

Blue Cuts greenhouse emissions ~71% for about a 23% cost rise — carbon avoided at roughly $68 a tonne.
Green Cuts emissions ~99.7% for about a 46% cost rise — carbon avoided at roughly $97 a tonne.

Carbon falls a long way either route; the price is the fight. And it's still a small fight: as of 2025 only about 6 Mt of low-emissions ammonia is operating or firmly committed (plus ~13 Mt earlier in the pipeline) against more than 190 Mt made in total — clean ammonia is real, but barely started. Figures: Shin, Zang et al. (2026), across 63 countries; pipeline from IEA Breakthrough Agenda (2025).

Why anyone cares beyond fertiliser. Because ammonia is already shipped and stored worldwide, and packs energy at ~3 kWh/litre without hydrogen's extreme cold or pressure, it's eyed as a way to move clean energy itself — a carbon-free shipping fuel, and a carrier that lets sunny, windy regions export energy as liquid (Royal Society; Saygin). The infrastructure already exists: the US alone has about 4,830 km of ammonia pipelines and more than 10,000 storage sites (Royal Society), and ~10% of all ammonia is already traded internationally (IEA). The IEA and IRENA expect demand from new energy uses to grow ammonia roughly 3–4×, to 560–665 Mt by 2050 (Saygin/IRENA). The catch that keeps the story honest: burnt or leaked carelessly, ammonia forms nitrogen oxides and harms ecosystems — so any clean-energy future depends on handling that releases only harmless nitrogen gas (Royal Society).

The same molecule that lets the planet feed itself may help it decarbonise — if the clean hydrogen it needs can ever be made as cheaply as the dirty kind.