How it Works
Forming a spherical nucleus of radius r from a supersaturated solution or supercooled melt costs surface energy but releases volume energy: ΔG(r) = 4πr²γ − (4/3)πr³ΔG_v. The surface term (positive, proportional to r²γ) dominates at small r, while the volume term (negative, proportional to r³ΔG_v) dominates at large r. This tug-of-war produces a barrier: ΔG(r) rises to a maximum ΔG* at the critical radius r* = 2γ/ΔG_v, then falls for r > r*.
Nuclei smaller than r* are thermodynamically pushed to shrink back to nothing, because further growth would raise ΔG. Nuclei larger than r* are pushed to grow indefinitely, because further growth lowers ΔG. Because real nuclei form and fluctuate stochastically, whether a given cluster crosses r* is a matter of chance — which is exactly what the seeded nucleus in this simulation does, taking a biased random walk in size, drifting away from r* toward zero if it starts below, and accelerating growth if it starts above.
Critical radius: r* = 2γ / ΔG_v
Nucleation barrier: ΔG* = (16π/3)·γ³/ΔG_v² = (4/3)πr*²γ
Rate law: J = A·exp(−ΔG*/kT)
Frequently Asked Questions
What does classical nucleation theory describe?
Classical nucleation theory (CNT) describes the free-energy cost of forming a small nucleus of a new, more stable phase (e.g. a crystal) inside a metastable parent phase (e.g. a supersaturated solution or supercooled melt). Even though the new phase is ultimately lower in free energy, forming its surface costs energy, creating a barrier that must be overcome before the new phase can grow.
What does the critical radius r* represent?
r* = 2γ/ΔG_v is the nucleus size at which the free-energy curve ΔG(r) reaches its maximum. It marks the boundary between unstable sub-critical nuclei, which tend to shrink and redissolve, and stable super-critical nuclei, which grow spontaneously because further growth lowers their free energy.
Why do nuclei smaller than r* behave oppositely from those larger than r*?
ΔG(r) is a competition between a positive surface term (∝r²γ) and a negative volume term (∝r³ΔG_v). Below r*, the surface term dominates, so growing further increases ΔG — the nucleus is driven to shrink. Above r*, the volume term dominates, so growing further decreases ΔG — the nucleus is driven to grow. This sign flip is a direct mathematical consequence of r² versus r³ scaling.
Why does higher supersaturation or undercooling make nucleation easier?
Increasing supersaturation or undercooling raises the magnitude of ΔG_v, the volume free-energy difference driving the transformation. Since r* = 2γ/ΔG_v and ΔG* ∝ γ³/ΔG_v², a larger ΔG_v shrinks both the critical radius and the barrier height. Because nucleation rate depends exponentially on −ΔG*/kT, even modest increases in supersaturation can produce dramatic increases in nucleation rate.
Why do some substances resist crystallizing even though it's thermodynamically favorable?
A high surface tension γ raises both r* and ΔG*, since the barrier scales as γ³, making the nucleation barrier very difficult to overcome by thermal fluctuations alone. This is why some liquids can be deeply supercooled or vitrify into a glass without crystallizing, and why industrial crystallization often requires seed crystals or nucleating agents to bypass the barrier.
What is heterogeneous nucleation and why is it usually easier?
Heterogeneous nucleation occurs on an existing surface — a dust particle, container wall, or seed crystal — rather than spontaneously within the bulk parent phase (homogeneous nucleation, the idealized case this simulation models). A foreign surface lowers the effective surface energy penalty by providing a template, sharply reducing the barrier. This is why nearly all real-world nucleation happens heterogeneously.
What is the CNT nucleation-rate law?
The nucleation rate follows an Arrhenius-like law, J = A·exp(−ΔG*/kT), where A is a kinetic prefactor. Because ΔG* appears in the exponent, nucleation rate is extraordinarily sensitive to small changes in temperature, supersaturation, or surface tension — a barrier just a few kT higher can suppress nucleation by many orders of magnitude.
How does classical nucleation theory apply to rain and cloud formation?
Water droplets in clouds nucleate far more readily on aerosol particles (dust, salt, pollution) than they would from pure water vapor alone, because these particles provide heterogeneous nucleation sites that drastically lower the effective barrier. Cloud seeding deliberately introduces such particles, like silver iodide, to encourage precipitation.
Why is protein crystallization for structural biology so difficult?
Proteins have complex, irregular surfaces that make the effective surface tension γ for forming an ordered crystal lattice unusually high, so the nucleation barrier ΔG* is large and homogeneous nucleation is rare. This is why growing diffraction-quality protein crystals for X-ray crystallography often requires extensive screening of conditions and can take months.