Ostwald Ripening: Why Big Grains Eat Small Ones

Curved surfaces are always slightly less stable

Once nucleation has produced a population of small droplets, grains, or bubbles dispersed in a matrix, the system is not done evolving even at constant temperature and constant total amount of material. A small particle has a highly curved surface, and surface tension pulls that curved surface inward with extra pressure — the smaller the radius, the higher the pressure, following the Young-Laplace relation ΔP = 2γ/r. That extra internal pressure raises the local chemical potential of the material just inside the surface, which in turn raises the equilibrium solubility (or vapour pressure) of that material in the surrounding matrix, right next to a small particle, compared to next to a large one.

This is the Gibbs-Thomson effect, and it sets up a concentration gradient in the matrix between small and large particles, even though there is no temperature difference anywhere. Material diffuses down that gradient — away from small, highly soluble particles and toward large, less soluble ones. Small particles shrink and eventually vanish; large particles grow. Wilhelm Ostwald described the phenomenon in 1896, and it now carries his name.

LSW theory and the cube-root law

Lifshitz and Slyozov, and independently Wagner, worked out in 1961 what this diffusion-driven coarsening does to the whole size distribution over time, not just to one pair of particles. Each particle's growth rate depends on the local diffusion field set up by every other particle, which makes the full problem genuinely many-body — but averaged over a large population, the result collapses to a strikingly simple asymptotic law:

Gibbs-Thomson:   c_eq(r) = c_inf * exp(2*gamma*V_m / (r*R*T))  ~  c_inf * (1 + 2*gamma*V_m/(r*R*T))

diffusion flux to/from a particle of radius r  ~  D * (c_inf - c_eq(r)) / r

LSW asymptotic (late-time) result:

  <r(t)>^3  -  <r(0)>^3   =   K * t          (K depends on D, gamma, solubility, temperature)

  =>   <r(t)>  ~  t^(1/3)      the mean radius grows as the cube root of time

LSW theory also predicts a specific, universal shape for the particle size distribution once the system reaches this self-similar "late-stage" regime — a distribution with a sharp cutoff at roughly 1.5 times the mean radius, since any particle that would grow larger than that consumes its local supply too fast and slows itself down. Real systems usually broaden that distribution somewhat because of particle-particle interactions and non-dilute effects that the original theory neglects, but the core t^(1/3) growth law for the mean size is remarkably robust and shows up across an enormous range of materials.

Where it matters in practice

Ostwald ripening governs the long-term stability of almost anything dispersed as small droplets or particles in a continuous phase. In metallurgy it coarsens the strengthening precipitates in age-hardened alloys, which is why those alloys have a finite useful service temperature — run them too hot for too long and the precipitates coarsen until they stop blocking dislocation motion effectively. In food science it is the reason fine emulsions like mayonnaise, ice cream, or foamed products slowly become coarser and grainier over their shelf life, and why formulators add surfactants and stabilisers specifically to suppress it. In geology it drives the coarsening of mineral grains during metamorphism over geological timescales. And it is a serious nuisance in nanoparticle synthesis, where researchers actively fight it (often by using very insoluble capping materials) to keep expensively engineered nanoparticles from simply dissolving into each other.

Distinguishing ripening from ordinary grain growth

It is worth separating Ostwald ripening from plain grain growth in a single-phase polycrystal, where grain boundaries themselves move to reduce total boundary area and larger grains consume smaller neighbours directly, without any diffusion through a separate matrix phase. Both processes produce the same qualitative outcome — big features eating small ones, driven by surface or interfacial energy — and both often follow power-law coarsening in time, but the transport mechanism, the exponent, and the microscopic driving force differ. Ostwald ripening specifically requires a second phase (a matrix or continuous phase) through which the coarsening species can diffuse from small particles to large ones.

Frequently asked questions

Try it live

Everything above runs in your browser — open Ostwald Ripening and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.