Physical chemistry
Adsorption, heterogeneous catalysis
How molecules stick to surfaces and why that sticking makes catalysts work: physisorption versus chemisorption, the Langmuir and BET isotherms, and Langmuir–Hinshelwood kinetics.
IntuitionIntuition: a parking lot where cars change
Picture a solid surface as a huge parking lot: molecules from the gas or liquid land on empty spots and stick. Weakly held visitors soon leave — that is physisorption. Some stick so firmly that their bonds rearrange while parked — chemisorption. A catalyst is a lot designed so that a molecule parking next to the right neighbour reacts, and the product simply drives away. Reactants that never park cannot react; products that never leave block the lot. Good catalysts tune the parking strength between these two failures.
SchoolSchool: two ways to stick, and why sticking speeds things up
Definition: Adsorption, physisorption, chemisorption
Adsorption is the accumulation of molecules at an interface. Physisorption binds them through weak, nonspecific van der Waals forces (heats of only 5–50 kJ mol⁻¹), is easily reversed by lowering the pressure, and may form several molecular layers. Chemisorption forms true chemical bonds with surface atoms (50–500 kJ mol⁻¹), is limited to one layer, is selective about sites, and may be activated — the very process that creates and breaks bonds in catalysis.
| Property | Physisorption | Chemisorption |
|---|---|---|
| Binding force | van der Waals | Chemical bond |
| Heat of adsorption | 5–50 kJ mol⁻¹ | 50–500 kJ mol⁻¹ |
| Layers | Multilayer possible | Monolayer only |
| Selectivity | Low, any surface | High, specific sites |
| Reversibility | Fully reversible | Often needs heat to desorb |
A heterogeneous catalyst turns sticking into speed. In the Haber–Bosch synthesis, N₂ — whose triple bond resists attack in the gas phase — dissociates upon chemisorbing on iron, and the adsorbed atoms are progressively hydrogenated to NH₃ which desorbs. In a car’s catalytic converter, CO and NO chemisorb on Pt/Rh nanoparticles, react at the surface, and leave as CO₂ and N₂. In both cases the surface provides a new, lower-barrier pathway: the catalyst is consumed by adsorption and regenerated by desorption every turnover.
UndergraduateUniversity: isotherms and catalytic rate laws
The Langmuir isotherm follows from a balance of parking and leaving on a surface of N_S equivalent, independent sites, each holding at most one molecule. With adsorption rate ∝ p(1 − θ) and desorption ∝ θ, equilibrium gives a simple fraction. The model underpredicts multilayer adsorption, which the BET extension treats by stacking further layers like condensation; its linearised form turns a measured adsorption curve into the monolayer capacity V_m and hence the specific surface area — the routine way powders are characterised.
Surface reactions inherit their rate laws from coverage. In the Langmuir–Hinshelwood picture both reactants adsorb and the event is A(ads) + B(ads) → products, so the rate r = kθ_Aθ_B has the competition factor (1 + K_Ap_A + K_Bp_B)² in the denominator: raise one pressure too far and you crowd out the other reactant — the rate passes through a maximum. In the Eley–Rideal alternative, an adsorbed species reacts directly with a gas-phase molecule (r ∝ θ_Ap_B), a mechanism documented for some hydrogenation and oxidation steps.
Example: Extracting K from an adsorption isotherm
For CO adsorbed on a zeolite at 300 K, θ = 0.50 at p = 2.0 kPa. Assuming Langmuir behaviour, find K and the pressure needed for θ = 0.90.
Solution
Rearrange θ = Kp/(1 + Kp): at θ = 0.5 the denominator gives Kp = 1, so K = 1/p = 0.50 kPa⁻¹. For θ = 0.90, Kp = θ/(1 − θ) = 9, hence p = 9/K = 18 kPa. General rule: at half coverage the pressure equals 1/K — K is literally the reciprocal pressure needed to half-fill the surface.
AdvancedAdvanced: what the surface really looks like
A real catalyst surface is not a smooth checkerboard but a corrugated energy landscape — the simulation above — with terraces, steps, kinks and defects of very different adsorption energies. Taylor’s 1925 insight that reactions may occur only at a minority of such “active sites” explained why tiny amounts of poison kill catalysts. Modern surface science measures this directly: temperature-programmed desorption maps binding energies, sticking coefficients quantify the landing probability, and the Brønsted–Evans–Polanyi relation links adsorption energies to activation barriers, so that a single descriptor — the binding energy of a key intermediate — orders metals on a Sabatier “volcano”: bind too weakly and nothing adsorbs, too strongly and nothing leaves.
Kinetics at this level is a sequence of elementary surface steps — adsorption, diffusion between sites, reaction, desorption — each with its own rate constant and coverage dependence. The measured order in pressure is therefore rarely integer: r ∝ p_A^m changes from m ≈ 1 at low coverage to m ≈ 0 or negative at saturation, and apparent activation energies blend true barriers with adsorption enthalpies (they can even turn negative when heating ejects reactants faster than it accelerates the surface step). Reading a rate law on a surface means reading coverages, not just concentrations.
References
- The Adsorption of Gases on Plane Surfaces of Glass, Mica and Platinum · I. Langmuir, 1918
- Adsorption of Gases in Multimolecular Layers · S. Brunauer, P. H. Emmett, E. Teller, 1938
- Why gold is the noblest of all the metals · B. Hammer, J. K. Nørskov, 1995