The Coffee-Ring Effect: How Evaporation Sorts Particles to the Edge

What it is and why it matters

When a drop of a colloidal suspension — coffee, ink, blood serum, a DNA solution — dries on a solid surface, the solute does not spread evenly. It piles into a dense ring tracing the drop's original perimeter, with the interior left nearly bare. Deegan and coworkers named and explained this coffee-ring effect in 1997, and the term now spans thousands of papers in soft-matter physics, colloid science, and printing technology.

The effect matters because it is both a nuisance and a tool. It ruins the uniformity of inkjet-printed electronics, OLED pixels, and spotted DNA/protein microarrays, where you want a flat, even film. Conversely, it concentrates analyte for detection — a passive, self-assembling preconcentration step exploited in trace chemical sensing and forensic bloodstain interpretation. Understanding it is essentially understanding how a moving free surface, a pinned boundary, and mass conservation together sculpt where micron-scale matter ends up. The physics is deterministic and geometric, not chemical: it happens for almost any suspended species whenever the contact line stays pinned.

The mechanism, step by step

Three ingredients conspire. First, contact-line pinning: surface roughness and the deposited solute itself anchor the drop's edge so its radius R stays fixed as liquid is lost. Second, diffusion-limited evaporation: the evaporation rate is set by how fast vapor diffuses away into the surrounding air, so the vapor concentration obeys Laplace's equation ∇²c = 0 above the drop. The drop's surface acts like a charged conductor of the same shape — and just as electric field lines crowd at a sharp edge, the evaporative flux diverges at the thin wedge where liquid meets substrate.

Third, mass conservation. Because evaporation is fastest at the rim but the rim cannot recede (it is pinned), liquid must flow radially outward from the interior to replenish what leaves at the edge. This is a genuine hydrodynamic current, not diffusion. Suspended particles are advected along with it. They accumulate at the pinned contact line, and as drying nears completion the surviving fluid rushes ever faster to the edge — a 'rush-hour' that dumps the remaining solute into a sharp ring.

The key equation, exponents, and scales

Solving Laplace's equation for the vapor field around a thin spherical-cap drop with contact angle θc gives the surface flux near the pinned edge:

J(r) ∝ (R − r)^−λ,   with  λ = (π/2 − θc) / (π − θc).

For a nearly flat drop (θc → 0) this gives λ → 1/2, an inverse-square-root divergence exactly analogous to the field at the edge of a charged conducting disk. Integrating the resulting depth-averaged outward velocity, Deegan (PRE 2000) showed the mass accreted at the ring grows and saturates as an increasing power law, m(t) ∝ t^(2/(1+λ)) (≈ t^(4/3) for a flat drop, λ = 1/2); what actually diverges as t → tf is the radial flow speed and evaporative flux — the 'rush-hour' acceleration — not the finite deposited mass.

Whether particles actually reach the rim before the drop dries is set by the Péclet number Pe = v_r R / D, comparing outward advection to Brownian back-diffusion. For a 1 mm drop with radial speed v ~ μm/s and colloids of D ~ 10⁻¹² m²/s, Pe ≫ 1, so advection wins and a ring forms. Typical rings are ~10 μm wide for 100 nm particles; total drying time is seconds to minutes.

How it is realized and measured

The canonical experiment is disarmingly simple: deposit a microliter drop of dilute latex-sphere suspension (fluorescent polystyrene, ~1 μm) on cleaned glass and watch it dry under a microscope. Confocal or bright-field imaging tracks individual particles; particle-image velocimetry (PIV) reconstructs the internal flow field, directly revealing the outward radial current and, if present, any counter-rotating Marangoni recirculation. The final deposit is imaged by optical or scanning-electron microscopy and profiled to measure ring width and height.

Deegan's group verified the predicted flux singularity by measuring how deposit mass and ring growth scaled with time and drop size, confirming the increasing ring-mass growth law m(t) ∝ t^(2/(1+λ)) (≈ t^(4/3) for a flat drop) together with the diverging near-end flow velocity and evaporative flux. Later work varied the atmosphere: raising humidity slows evaporation and thins the ring; a directed external vapor source can steer where the flux — and hence the deposit — concentrates. Interferometry and reflection microscopy measure the receding drop-height profile, and controlling substrate temperature lets experimenters switch on thermal Marangoni flow on demand, providing a clean knob to test each ingredient of the theory independently.

Where it operates and how it differs from related effects

The coffee-ring effect is universal wherever a drop has a pinned contact line and evaporates by diffusion into a passive gas. It appears in coffee, wine, salt water, blood, printer ink, and nanoparticle inks alike. It weakens or vanishes when any pillar is removed: on slippery or superhydrophobic substrates the contact line recedes rather than pins, giving a central spot; at low Péclet number (small drops, large slow-drying particles) diffusion homogenizes the deposit into a disk.

The most important competing flow is thermal Marangoni convection. Evaporative cooling makes the drop apex slightly colder than its edge; since surface tension σ rises as temperature falls, a σ-gradient drives surface flow toward the apex, setting up a recirculation that can carry particles back to the center. Whether this reverses the ring is governed by the Marangoni number Ma = (dσ/dT)(ΔT·R)/(ηα). Pure water has weak, surfactant-suppressed Marangoni flow — which is exactly why coffee (aqueous) rings so reliably, while many organic solvents do not.

Applications, suppression, and open questions

Because uniform films are prized in printed electronics, much effort targets suppressing the ring. Yunker, Still, Lohr and Yodh (Nature 2011) showed a purely mechanical route: ellipsoidal particles, unlike spheres, strongly deform the air–water interface and attract one another via capillary forces; once carried to the surface they jam into an arrested network that resists edge accumulation, yielding a uniform coat. Other routes add surfactants (to tune Marangoni flow), mix particle sizes, control humidity and temperature, or use electrowetting to stir the interior.

Applications cut both ways. The ring is deliberately exploited for self-assembly of lines and gratings, for concentrating analyte in low-cost diagnostics and SERS substrates, and forensically — the width and morphology of dried-bloodstain rings encode droplet age and impact. Open questions remain quantitative: predicting deposit morphology across the full parameter space of Pe, Ma, particle shape, DLVO interactions, contact-line dynamics and concentration-dependent viscosity is still unsolved, and multi-component or biologically active drops (with living cells, or surfactant self-assembly) show patterns no single-mechanism theory yet captures.

Deposition regimes for an evaporating sessile drop: what balance decides the pattern

Mechanism / regime	Governing quantity	Resulting deposit
Pinned contact line + diffusive evaporation	Flux singularity J ∝ (R−r)^−λ; high Péclet Pe = vR/D ≫ 1	Sharp coffee ring at rim
Diffusion dominates over advection	Low Péclet Pe = vR/D ≲ 1	Uniform / disk-like deposit
Thermal Marangoni flow	Marangoni number Ma = (dσ/dT)(ΔT·R)/(ηα) large	Recirculation returns particles to center
Anisotropic (ellipsoidal) particles	Capillary interface deformation, strong interparticle attraction	Uniform monolayer; ring suppressed
Depinned / receding contact line	Low pinning; slippery/superhydrophobic substrate	Central spot, no ring
Electrowetting / AC forcing	Internal mixing overrides outward flow	Tunable, often uniform