Heterogeneous Catalysis: How Solid Catalysts Work

Short answer

Heterogeneous catalysis is catalysis in which the catalyst and the reacting molecules are in different phases, most often a solid catalyst working on gas or liquid reactants. The reaction happens on the solid’s surface: molecules attach (chemisorb), react, and leave, so the catalyst can run continuously in a flow reactor without being consumed. It is the dominant mode of catalysis in technical chemistry.

Definition and scope

In his 2007 Nobel Lecture, Gerhard Ertl quotes Wilhelm Ostwald’s definition from around 1900: a catalyst is “a substance which affects the rate of a chemical reaction without being part of its end products.” The word “catalysis” itself goes back to Berzelius in 1835. An early observation is Döbereiner’s in 1823, that finely divided platinum powder makes hydrogen and oxygen react to water while the platinum stays unaltered.

Ertl separates three settings. If the reacting molecules and the catalyst are in the same gaseous or liquid phase, the effect is called homogeneous catalysis. In living systems, macromolecules (enzymes) play the role of catalysts. In technical reactions, mostly the interaction of molecules with the surface of a solid is decisive, and that is heterogeneous catalysis. A catalyst works by forming intermediate compounds with the reactants, which gives an alternative reaction path with a lower activation barrier; the fraction of molecules able to cross a barrier scales roughly as exp(−E*/kT), so even a modest barrier reduction changes the rate substantially.

This page sits in the protocols and lab techniques part of the lab operations cluster because the same ideas govern how solid catalysts are handled, run in flow reactors and diagnosed when they fail. For the neighbouring disciplines, see the CASRAI guides on physical chemistry, inorganic chemistry and chemical engineering.

How a solid catalyst works in practice

Adsorption, surface reaction, desorption

According to Ertl, atoms at a solid surface have fewer neighbours than atoms in the bulk, so they are chemically unsaturated and can form bonds with molecules arriving from the adjacent gas or liquid phase. That attachment is chemisorption. The bonds inside the adsorbed molecule can be modified or broken (dissociative chemisorption). The surface species then move between sites and react with one another, and the products leave the surface (desorption). Because the surface is restored after each cycle, a catalyst operated in a flow reactor can run continuously without being consumed.

Two practical consequences follow directly. Activity depends on which surface sites are exposed, not just on what the material is made of, and the reaction is limited by whichever step in the cycle is slowest. The two worked examples below show both points with data from the same lecture.

Worked example 1: ammonia over an iron catalyst

The reaction is N₂ + 3H₂ → 2NH₃. Fritz Haber realised it in the laboratory in 1909 with an osmium catalyst in a high-pressure flow apparatus (Nobel Prize 1919), and Carl Bosch and BASF scaled it up to the first industrial plant in 1913. Ertl notes that the world’s osmium supply was then only about 80 kg, so Alwin Mittasch tested thousands of materials and found an iron-ore-derived catalyst. He reports that this doubly promoted iron catalyst is in fact still in use in almost all industrial plants.

The industrial catalyst is not plain iron. Its specific surface area is about 20 m²/g, because the iron is present as nanometre-sized particles. Under reaction conditions the metallic iron is covered by a submonolayer of potassium plus oxygen, which Ertl calls an “electronic” promoter. An Al₂O₃ framework (with CaO) acts as a “structural” promoter that stabilises the particles against sintering.

Why is N₂ the bottleneck? Paul Emmett concluded in 1974 that the rate-limiting step over iron is the chemisorption of nitrogen. Ertl’s surface-science measurements are consistent with this: the sticking coefficient for dissociation of N₂ on iron is about 10⁻⁶, and the Fe(111) surface is the most active while the densely packed Fe(110) surface is the least active. Potassium raises the probability of dissociation. The same order of reactivity was found for NH₃ yield at 20 bar, so for this reaction Ertl reports no “pressure gap” between low-pressure surface studies and technical conditions.

Worked example 2: CO oxidation and why a car catalyst needs heat

The car-exhaust reaction 2CO + O₂ → 2CO₂ runs on platinum-group metal particles. O₂ is dissociatively chemisorbed, and the adsorbed oxygen reacts with chemisorbed CO in a Langmuir-Hinshelwood mechanism, with an activation barrier of about 100 kJ/mol (roughly half that at higher coverage; density-functional calculations agree). The surface tends to become covered with CO, which blocks adsorption of O₂. CO has to desorb to free sites, which requires a temperature of at least about 450 K. In Ertl’s words, “the catalyst of your car does not work in the cold.” The same system shows rate oscillations on platinum, first reported around 1970 by Wicke, and later on Pt(110).

The lesson is general: site competition between adsorbates, not only the intrinsic reaction step, can set the temperature window in which a solid catalyst operates.

From plane surfaces to working catalysts

Ertl describes the surface-science approach as Langmuir’s 1922 idea of studying plane, well-defined surfaces first. It became experimentally accessible in the 1960s with ultra-high vacuum and surface-sensitive techniques; the lecture names low-energy electron diffraction (LEED) and scanning tunnelling microscopy (STM). In the ammonia case, the reactivity order found on iron surfaces matched the order seen at technical pressure.

Types and related concepts

The table compares the three catalytic settings Ertl distinguishes, using his wording.

Setting	Catalyst and reactants	Where the interaction happens	Example in this page
Homogeneous catalysis	Same gaseous or liquid phase	Within that single phase	None
Enzyme catalysis	Macromolecules in living systems	In the biological system	None
Heterogeneous catalysis	Solid surface with gas or liquid reactants	At the solid surface	Fe for NH3; Pt-group metals for CO oxidation

For a research centre working on catalysis at surfaces, see the CASRAI guide to SUNCAT.

Why catalysts stop working

Argyle and Bartholomew (Catalysts, 2015) open their review of deactivation and regeneration by stating that deactivation of heterogeneous catalysts is “a ubiquitous problem that causes loss of catalytic rate with time.” Their abstract classifies deactivation along two axes, summarised below.

Classification axis	Categories named in the review
Type	Chemical, thermal, mechanical
Mechanism	Poisoning; fouling; thermal degradation; vapor formation; vapor-solid and solid-solid reactions; attrition and crushing

The review discusses regeneration separately for each mechanism and works through case studies on cobalt Fischer-Tropsch catalysts and on selective catalytic reduction, covering sintering, coking, poisoning and fouling. Connecting this back to the ammonia example: the Al₂O₃ structural promoter described above is there to protect against sintering, one of the thermal routes by which active surface area is lost.

Where it matters for labs and process decisions

• Surface area and structure are part of the specification. The ammonia catalyst’s roughly 20 m²/g area and its promoter layers are what make the material work, so two catalysts of the same nominal metal can behave differently.
• Operating temperature has a lower bound as well as an upper one. The CO oxidation case needs about 450 K or more to clear adsorbed CO, while thermal degradation is a deactivation mechanism at the other end.
• Loss of activity over time is expected. A rate that drifts down during a run is the normal subject of the deactivation literature, and the review classifies it by mechanism so the cause can be named before regeneration is considered.
• Flow operation is the natural mode. Per Ertl, continuous operation without consuming the catalyst is what flow reactors exploit.

For environmentally oriented process design, the CASRAI green chemistry lab guide covers the wider context in which catalyst choices are made.

Common misunderstandings

“A catalyst takes no part in the reaction”

It participates: it forms intermediate compounds with the reactants. Ostwald’s definition says only that it is not part of the end products.

“A catalyst lasts forever”

Ertl’s statement is that a catalyst in a flow reactor operates continuously without being consumed by the reaction itself. Argyle and Bartholomew’s point is that rate nevertheless falls over time through poisoning, fouling, thermal and mechanical processes.

“The surface is just a passive support”

Different crystal planes of the same metal can differ strongly, as the Fe(111) versus Fe(110) comparison shows, and promoters such as potassium change the dissociation probability.

“A catalyst that works in a vacuum experiment will work at plant pressure”

That question is called the pressure gap. For ammonia over iron, Ertl found the same reactivity order at 20 bar as in the low-pressure experiments, but he presents this as a result for that case, not a rule.

Frequently asked questions

What is heterogeneous catalysis in simple terms?

It is catalysis where the catalyst is in a different phase from the reactants, typically a solid surface acting on gas or liquid molecules.

What is the difference between homogeneous and heterogeneous catalysis?

In homogeneous catalysis the catalyst and reactants share one phase; in heterogeneous catalysis the interaction occurs at the surface of a solid.

What steps occur on a catalyst surface?

Chemisorption (possibly dissociative), movement of surface species between sites and reaction, then desorption of products.

Which catalyst is used for ammonia synthesis?

Haber’s 1909 laboratory work used osmium; Mittasch’s iron-based catalyst with potassium and Al₂O₃ promoters is, per Ertl, still used in almost all industrial plants.

Why does a car’s catalytic converter not work when cold?

Per Ertl, CO covers the platinum-group surface and blocks O₂ adsorption until the temperature reaches about 450 K or higher, when CO desorbs.

Why do heterogeneous catalysts deactivate?

The main routes named in the literature are poisoning, fouling, thermal degradation, vapor formation, solid-state reactions and attrition or crushing.

Sources

• Ertl, G. “Reactions at Surfaces: From Atoms to Complexity.” Nobel Lecture, 8 December 2007. nobelprize.org. Also in Angew. Chem. Int. Ed. 2008, 47, 3524-3535, doi 10.1002/anie.200800480. Read 2026-11-12.
• Argyle, M.D.; Bartholomew, C.H. “Heterogeneous Catalyst Deactivation and Regeneration: A Review.” Catalysts 2015, 5(1), 145. doi 10.3390/catal5010145. Abstract read via Crossref, 2026-11-12.