# Nitrogen from the Air

Laplace’s Mini World · Plate 54 · Matter, Light & Sound · Matter

> Air is nearly four-fifths nitrogen, and crops can’t use it. At 100⁠–⁠250 bar and 350⁠–⁠550 °C, nitrogen and hydrogen join on iron as ammonia, 70 % of it for fertiliser.

- Interactive model: https://world.lapla.si/en/haber/
- Embed: https://world.lapla.si/embed/en/haber/
- 中文: https://world.lapla.si/zh/haber/
- Updated: 2026-10-28

## 01 Nitrogen Out of Reach

Dry air is 78 % nitrogen. Proteins and DNA all need nitrogen, but in nitrogen gas two atoms are held by a triple bond that takes 945 kJ/mol to break, the strongest bond of any molecule made of two like atoms. Plants can’t use it directly.

In nature, some nitrogen-fixing microbes and lightning break it. By the estimate of Fowler and colleagues (2013), microbes on land fix about 58 million tonnes of nitrogen a year, in the sea about 140 million, and lightning about 5 million.

In 2010 industry fixed about 120 million tonnes of nitrogen from the air, 100 million of it as fertiliser: already twice the natural fixation on land. All of it starts with one reaction: nitrogen and hydrogen to ammonia.

*Nitrogen 78 % of dry air · the N≡N bond: 945 kJ/mol · industry fixed about 120 Mt N in 2010 (Fowler et al. 2013)*

## 02 A Reaction That Runs Both Ways

N₂ + 3 H₂ ⇌ 2 NH₃: one nitrogen molecule and three hydrogen molecules make two ammonia molecules. The reaction gives off heat: about 92 kJ for every 2 mol of ammonia at 25 °C.

It runs both ways: ammonia keeps breaking back into nitrogen and hydrogen. When the two directions run equally fast, the ammonia stops increasing: equilibrium.

Four molecules on the left, two on the right: the reaction shrinks the gas, so the higher the pressure, the further the equilibrium leans to ammonia; it gives off heat, so the lower the temperature, the further it leans too. The balls in the converter are a schematic: blue for nitrogen, white for hydrogen, not to scale.

*N₂ + 3 H₂ ⇌ 2 NH₃ · about 92 kJ released per 2 mol of ammonia at 25 °C · four molecules become two*

## 03 Too Slow When Cold, Too Little When Hot

The colder it is, the more ammonia at equilibrium: at 200 bar, about 68 % at 300 °C, only about 19 % at 500 °C (computed on this page from thermodynamic data). But when it is cold, the reaction all but stops.

Heat speeds it up. At 200 bar and for the same gas, by the rate expression of a model of an industrial loop, it runs about 3 times faster at 450 °C than at 400 °C, and about 10 times faster than at 350 °C.

Plants settle on a compromise: usually 100⁠–⁠250 bar and 350⁠–⁠550 °C, with a catalyst to speed things up. Even so, each pass through the converter turns only 20⁠–⁠30 % of the gas into ammonia.

*Ammonia at equilibrium, 200 bar: about 68 % at 300 °C, about 19 % at 500 °C (computed) · in industry: 100⁠–⁠250 bar, 350⁠–⁠550 °C · 20⁠–⁠30 % converted per pass*

## 04 On the Surface of Iron

What goes into the converter is magnetite (Fe₃O₄) with a few per cent of alumina and a little potassium and calcium oxide. At start-up the synthesis gas reduces it to porous iron, which is the catalyst.

A nitrogen molecule lands on the iron, its triple bond comes apart there, and each nitrogen atom takes on hydrogen until it leaves as ammonia. Gerhard Ertl’s surface experiments showed that the slowest step is the nitrogen molecule splitting on the iron; potassium oxide changes the iron’s electrons, alumina and calcium oxide keep its structure from collapsing.

BASF’s Alwin Mittasch found this catalyst. By 1910 his team had tried more than 2,500 recipes in 6,500 runs, each sample about 2 g; in November 1909 a magnetite from Sweden did exceptionally well.

*Magnetite + alumina, potassium and calcium oxides → reduced to iron · slowest step: N₂ splitting on iron · by 1910: 2,500+ recipes, 6,500 runs*

## 05 The Converter, Cut Open

The converter is a thick-walled pressure vessel with a thin inner basket, the catalyst held in it in several beds. The cold feed gas first flows through the annulus between the shell and the basket, keeping the pressure shell cool.

The gas passes down through a bed and heats up as it reacts: in one design it enters the first bed at about 400 °C and leaves at about 500 °C. Hotter means a worse equilibrium, so between beds the gas is cooled before it enters the next.

On the stand, a model of the principle with three beds and two interbed heat exchangers, not any maker’s design. Real converters are slimmer and taller, and the grains are only millimetres across: 6⁠–⁠10 mm in axial-flow converters, 1.5⁠–⁠3 mm in radial-flow ones.

*The shell cooled by the incoming gas · first bed: in at about 400 °C, out at about 500 °C (Appl 1997) · cooling between beds*

## 06 Round and Round

Each pass turns only a fifth to a third into ammonia, and the rest can’t be thrown away. The gas leaving the converter is cooled to around 0 °C; the ammonia condenses and runs to the bottom of the separator, where it is drawn off.

The unreacted nitrogen and hydrogen leave the top of the separator, join the fresh synthesis gas, and are compressed and sent back to the converter.

The feed carries a little argon and methane, which don’t react and would build up, so a small stream is bled off all the time to hold them at about 10⁠–⁠15 %.

*Cooled to about 0 °C, the ammonia condenses · the unreacted gas is compressed and returned · a small purge holds the inert gases at 10⁠–⁠15 %*

## 07 Where the Hydrogen Comes From

The nitrogen comes from the air; the hydrogen mostly from natural gas. Methane and steam turn into hydrogen and carbon monoxide at high temperature, and the carbon monoxide reacts with more steam to give carbon dioxide and more hydrogen. Air added in the secondary reformer brings the nitrogen for the synthesis, three parts hydrogen to one part nitrogen.

In 2020 the world made 185 million tonnes of ammonia, 72 % from natural gas and 26 % from coal. On average each tonne emitted about 2.4 t of carbon dioxide; the best gas-based plants about 1.8 t, the best coal-gasification plants about 3.2 t.

Making ammonia takes about 2 % of the world’s final energy use and gives about 1.3 % of the energy system’s CO₂ (IEA 2021). Low-emissions hydrogen, made for instance by splitting water with low-carbon electricity, was still less than 1 % of the world’s hydrogen in 2024.

*About 2.4 t of CO₂ per tonne of ammonia (average) · about 2 % of world final energy · low-emissions hydrogen under 1 %*

## 08 The Bench Apparatus of 1909

At the Technische Hochschule in Karlsruhe, Fritz Haber and Robert le Rossignol built a small apparatus: the gas was compressed to about 200 atmospheres and passed over osmium or uranium at 500⁠–⁠600 °C, the ammonia condensed out, and the unreacted gas went round again.

On 2 July 1909, with BASF’s Alwin Mittasch watching, it ran for about five hours without a stop, making liquid ammonia all the while (Haber’s letter the next day). That day’s ammonia, sealed in glass tubes, is in the Science Museum in London.

Haber’s yield was several grams of ammonia an hour for every cubic centimetre of the heated high-pressure chamber. BASF had bought the process in 1908 and gave Carl Bosch the job of scaling it up.

*About 200 atmospheres · 500⁠–⁠600 °C · osmium or uranium · 2 July 1909: ammonia for about five hours without a stop*

## 09 Steel Eaten by Hydrogen

BASF’s early contact tubes were carbon steel, their walls about 30 mm thick, heated from outside. Two tubes drawn by Mannesmann each lasted 80 hours, then burst.

Cut open, the inside of the tubes had lost its pearlite: at high temperature and pressure, hydrogen had got into the steel and taken its carbon (as methane), leaving it hard and brittle.

Bosch’s answer: a pressure-bearing steel jacket lined inside with thin soft steel. Hydrogen that gets through the lining escapes along the grooves left on the lining’s outside by turning and through many holes drilled in the jacket, so no pressure builds up; the holes came to be called “Bosch holes”. The specimen at the front shows the idea. The first lined converter dates from March 1911; Oppau began production in September 1913, built for 30 t a day.

*Carbon steel: burst after 80 hours · soft-steel lining + drilled jacket, the hydrogen escapes through the holes · Oppau: from September 1913, 30 t a day*

## 10 The Nitrogen in Our Food

By the U.S. Geological Survey’s estimate, the world made ammonia containing about 160 million tonnes of nitrogen in 2025, about 190 million tonnes of ammonia; about 70 % of it goes into fertiliser.

By the estimate of Erisman and colleagues (2008), about 48 % of the world’s people that year were fed thanks to nitrogen fertiliser from synthetic ammonia; Smil’s 1999 estimate was almost two-fifths. These are estimates of what would be missing without it, not counts.

Most of the fertiliser nitrogen never reaches our food: in 2005, farming used about 100 million tonnes of Haber–Bosch nitrogen, and people ate only about 17 million. The same plants also supplied the war: the Nobel Prize’s biography of Haber says that when Germany’s nitrates for explosives failed in 1914, Oppau and the Leuna works, opened in 1917, let it prolong the First World War.

*2025: about 160 Mt of nitrogen (USGS estimate) · about 70 % for fertiliser · about 48 % of people fed by it (Erisman et al. 2008, an estimate)*

## 11 Lifted Out, Layer by Layer

The converter comes apart into the cover, three catalyst beds, two interbed heat exchangers, the basket and the pressure shell.

The part under the pointer lifts out alone and is named; left alone, the parts lift out in turn.

*Cover · bed 1 · exchanger · bed 2 · exchanger · bed 3 · basket · shell*

## 12 Temperature and Pressure

Reading one: ammonia at equilibrium, for three parts hydrogen to one of nitrogen, computed from thermodynamic data with the Peng–Robinson equation of state. Reading two: the relative rate on the catalyst, by the rate expression of a model of an industrial loop, for a gas with 3 % ammonia, taken as 1 at 450 °C and 200 bar.

The pale band on the chart is the range plants usually run in. The balls in the converter are a schematic: the further the equilibrium leans to ammonia, the more of them turn into it; the faster the reaction, the sooner.

*Readings: ammonia at equilibrium · relative rate (1 at 450 °C, 200 bar) · the molecules are a schematic*

**Try this**: At 200 bar, move the temperature from 350 °C to 550 °C: how far does the ammonia at equilibrium fall, and how many times faster does it react?

## What the Model Leaves Out

The stand is a model of the principle: the look, size and place of the converter, compressor, cooler and separator are drawn, not taken from any plant; real converters are slimmer and taller, and their grains are only millimetres across. Real plants cool in several stages (water first, then evaporating ammonia) and often compress in several, shown here as one each. The molecules in the converter are a schematic, not to scale and not in real numbers. The lab’s equilibrium is for three parts hydrogen to one of nitrogen with no inert gas, corrected for real gases with the Peng–Robinson equation; as ideal gases it agrees with Haber’s 1920 table within 0.5 percentage points. The rate is shown only as a ratio, from the rate expression of a model of an industrial loop, without diffusion inside the grains or the temperature changing along the bed.

## Easily Misread

A catalyst doesn’t shift the equilibrium; it only gets there faster. Cooling moves the equilibrium towards ammonia but slows the reaction. Turning only 20⁠–⁠30 % per pass doesn’t waste the feed: the unreacted gas goes round again, and only the small purge that carries off the inert gases is lost. “Feeding about half the world” is an estimate that rests on assumptions about yields, diets and alternatives; Erisman and colleagues give about 48 % for 2008, Smil almost two-fifths in 1999. Making the hydrogen by splitting water leaves the synthesis itself unchanged, only the hydrogen’s source; such low-emissions hydrogen was still under 1 % of the world’s hydrogen in 2024.

## Common questions

### Why can’t plants use the nitrogen in the air?

Proteins and DNA all need nitrogen, but in nitrogen gas two atoms are held by a triple bond that takes 945 kJ/mol to break, the strongest bond of any molecule made of two like atoms.

See: [01 Nitrogen Out of Reach](https://world.lapla.si/en/haber/#air)

### At what temperature and pressure is ammonia made?

Plants settle on a compromise: usually 100⁠–⁠250 bar and 350⁠–⁠550 °C, with a catalyst to speed things up.

See: [03 Too Slow When Cold, Too Little When Hot](https://world.lapla.si/en/haber/#compromise)

### What is the slowest step in making ammonia?

A nitrogen molecule lands on the iron, its triple bond comes apart there, and each nitrogen atom takes on hydrogen until it leaves as ammonia.

See: [04 On the Surface of Iron](https://world.lapla.si/en/haber/#catalyst)

### What happens to the gas that does not react?

Each pass turns only a fifth to a third into ammonia, and the rest can’t be thrown away.

See: [06 Round and Round](https://world.lapla.si/en/haber/#loop)

### How did Bosch stop hydrogen from wrecking the steel?

Bosch’s answer: a pressure-bearing steel jacket lined inside with thin soft steel.

See: [09 Steel Eaten by Hydrogen](https://world.lapla.si/en/haber/#liner)

## Sources

1. [European Commission, Best Available Techniques reference document: Ammonia, Acids and Fertilisers (2007): 100⁠–⁠250 bar, 350⁠–⁠550 °C, 20⁠–⁠30 % converted per pass, cooling and separation, the purge, air added in the secondary reformer](https://bureau-industrial-transformation.jrc.ec.europa.eu/sites/default/files/2019-11/lvic_aaf.pdf)
2. [M. Appl (1997): The Haber-Bosch Heritage, IFA Technical Conference (400⁠–⁠500 °C, beds and cooling between them, grain sizes, the catalyst search and make-up, Oppau’s 30 t a day)](https://www.fertilizer.org/wp-content/uploads/2023/01/HABER.pdf)
3. [F. Haber (1920): The synthesis of ammonia from its elements, Nobel lecture (200 atmospheres, osmium and uranium catalysts, his equilibrium table)](https://www.nobelprize.org/uploads/2018/06/haber-lecture.pdf)
4. [C. Bosch (1932): The development of the chemical high pressure method during the establishment of the new ammonia industry, Nobel lecture (the tubes that burst after 80 hours, decarburisation, the soft-steel lining and the drilled holes)](https://www.nobelprize.org/uploads/2018/06/bosch-lecture.pdf)
5. [G. Ertl (2007): Reactions at surfaces: from atoms to complexity, Nobel lecture (nitrogen splitting on iron is the slowest step; the promoters)](https://www.nobelprize.org/uploads/2018/06/ertl_lecture.pdf)
6. [V. Smil (1999): Detonator of the population explosion, Nature 400, 415 (the run of 2 July 1909, Oppau’s start, the two-fifths estimate)](https://doi.org/10.1038/22672)
7. [Science Museum Group, object 1952-391/2: the first liquid ammonia made by the Haber process, 2 July 1909](https://collection.sciencemuseumgroup.org.uk/objects/co525226/)
8. [D. Fowler et al. (2013): The global nitrogen cycle in the twenty-first century, Phil. Trans. R. Soc. B 368: 20130164 (nitrogen fixed each year)](https://doi.org/10.1098/rstb.2013.0164)
9. [J. W. Erisman et al. (2008): How a century of ammonia synthesis changed the world, Nature Geoscience 1, 636 (the 48 % estimate, nitrogen use)](https://doi.org/10.1038/ngeo325)
10. [IEA (2021): Ammonia Technology Roadmap (185 Mt produced, shares of energy and CO₂, CO₂ per tonne, about 70 % for fertilisers)](https://www.iea.org/reports/ammonia-technology-roadmap)
11. [IEA (2025): Global Hydrogen Review 2025 (low-emissions hydrogen under 1 %)](https://www.iea.org/reports/global-hydrogen-review-2025/executive-summary)
12. [U.S. Geological Survey (2026): Mineral Commodity Summaries, Nitrogen (Fixed)—Ammonia (world production)](https://pubs.usgs.gov/periodicals/mcs2026/mcs2026-nitrogen.pdf)
13. [US EPA AP-⁠42 §8.1 (1993): Synthetic Ammonia (cooling to 0 °C to separate the ammonia, the purge)](https://www.epa.gov/sites/default/files/2020-09/documents/8.1_synthetic_ammonia.pdf)
14. [NIST Chemistry WebBook: thermochemical data (Shomate equations) and critical constants for ammonia, nitrogen and hydrogen](https://webbook.nist.gov/cgi/cbook.cgi?ID=C7664417&Units=SI&Mask=1)
15. [CRC Handbook of Chemistry and Physics, 81st ed. (via WebElements): the N≡N bond, 945 kJ/mol](https://www.webelements.com/nitrogen/compound_properties.html)
16. [NOAA: The Atmosphere (nitrogen 78.084 % of dry air)](https://www.noaa.gov/jetstream/atmosphere)
17. [A. Araújo, S. Skogestad (2008): Control structure design for the ammonia synthesis process, Computers & Chemical Engineering 32, 2920 (this page’s rate expression)](https://doi.org/10.1016/j.compchemeng.2008.03.001)
18. [US patent 7,867,465 B2 (2011): feed gas through the annulus between basket and shell cools the pressure shell](https://patents.google.com/patent/US7867465B2/en)
19. [nobelprize.org: biographies of Carl Bosch and Fritz Haber (BASF’s purchase of the process in 1908, Oppau and Leuna, the First World War)](https://www.nobelprize.org/prizes/chemistry/1931/bosch/biographical/)

## Data

The equilibrium is computed from the NIST Chemistry WebBook’s thermochemical data (Shomate equations from the NIST-JANAF tables, M. W. Chase, 1998) and critical constants, with the Peng–Robinson equation of state; the relative rate uses the Temkin–Pyzhev expression with the constants of A. Araújo and S. Skogestad (Computers & Chemical Engineering 32, 2920, 2008). Haber’s own equilibrium table (Nobel lecture, 1920) is the check. Plant conditions are from the EU’s Best Available Techniques reference document for ammonia, acids and fertilisers (2007) and M. Appl, “The Haber-Bosch Heritage” (IFA, 1997); the history from the Nobel lectures of F. Haber (1920), C. Bosch (1932) and G. Ertl (2007) and V. Smil (Nature 400, 415, 1999); the flows and totals from D. Fowler et al. (2013), J. W. Erisman et al. (2008), the IEA (2021, 2025) and the USGS (2026). The converter, the loop and the specimen are a model of the principle, not any one plant or maker’s design. Material presets, parts and the softbox studio of the site’s model kit are ported from AirsupHQ/machine-skill (MIT, © 2026 Air Supply Technologies Inc.).

## Licence CC BY-NC 4.0

The text, pictures and models on this page are licensed CC BY-NC 4.0: free to use and adapt for teaching and other non-commercial purposes, with credit to world.lapla.si. Third-party material listed above keeps its own licence.

https://creativecommons.org/licenses/by-nc/4.0/deed.en

## Cite this plate

Laplace’s Mini World, Plate 54, “Nitrogen from the Air”, https://world.lapla.si/en/haber/, CC BY-NC 4.0, updated 2026-10-28.