Colloidal Forces and DLVO Theory

Two Particles, One Pull: van der Waals Attraction

Guide 1 handed us a puzzle. Stir a fine oxide powder into water and, left to itself, it collapses into clumps and settles out. Gravity is not the villain — a sub-micron particle is so light that random thermal kicks from water molecules (Brownian motion) keep it aloft for days. The real culprit is that the particles want to touch. To see why, we have to name the force that pulls any two grains together no matter what, and it is the first of our colloidal cast: the van der Waals attraction.

Where does it come from? Even a neutral atom has electrons that flicker from side to side, and for an instant it becomes a tiny dipole — plus on one face, minus on the other. That flicker induces a matching dipole in a neighbour, and the two lean toward each other. Sum this over the trillions of atoms in one particle pulling on the trillions in another and you get a force that, unlike the fickle single-atom version, is always attractive between two particles of the same material across a liquid. For two spheres of radius R almost touching, the van der Waals energy is roughly V ~ -A times R / (12 times D), where D is the tiny gap between their surfaces and A is the Hamaker constant — a material number, about 1 x 10^-20 J for common oxides across water.

That formula hides a warning. Because the energy grows as -1/D, the pull becomes brutal as the gap closes to a nanometre. Put numbers in: two half-micron alumina spheres (R = 0.25 micron = 2.5 x 10^-7 m) at a gap of D = 1 nm, with A = 4 x 10^-20 J, give V = -(4 x 10^-20 times 2.5 x 10^-7)/(12 times 1 x 10^-9), about -8 x 10^-19 J. Compare that to the thermal energy a particle carries, kT = 1.38 x 10^-23 times 300, about 4 x 10^-21 J at room temperature. The attractive well is nearly 200 times kT deep. A particle that stumbles that close is trapped in a pit far too deep for thermal jiggling to climb back out. This is exactly why guide 1's fine powder — with its enormous specific surface area — clumps: left alone, every collision is a permanent marriage.

The Electrical Double Layer: a Cloud of Counter-Ions

If attraction never rests, the only way out is repulsion — and nature offers a beautiful one for particles in water. Drop an oxide into water and its surface is not inert: the surface metal-OH groups react with the water. In acid they grab a proton and go positive (M-OH + H+ gives M-OH2+); in base they shed one and go negative (M-OH gives M-O- + H+). Either way the particle surface ends up charged, and how charged depends on the pH. That surface charge is the seed of the repulsion we need.

A charged surface in water does not sit alone. It pulls a swarm of oppositely charged ions (counter-ions) out of the liquid to screen it. The nearest of these stick fast in a thin, tightly bound sheet — the Stern layer — while the rest form a looser, jostling cloud that thins out with distance, the diffuse layer. Surface charge plus its ionic cloud together make the electrical double layer: think of it as a charged particle wearing a fuzzy coat of the opposite sign. Two particles feel no repulsion until these coats begin to overlap.

How thick is this coat? Its reach is set by the Debye length, written 1/kappa — the distance over which the counter-ion cloud screens the surface charge to nothing. In water at room temperature it runs about 1/kappa (nm) = 0.3 / sqrt(I), with I the ionic strength in mol/L. So in fairly pure water (I = 0.001 mol/L) the coat is a fat ~10 nm thick; add salt to I = 0.1 mol/L and it collapses to a thin ~1 nm. Hold onto that: more salt means a thinner coat, and we will see in a moment why that quietly wrecks a suspension. The measurable potential right at the edge of the tightly-bound part of this cloud is the zeta potential — the leftover charge that keeps particles apart and the slip pourable. It is the handle you actually turn in the lab, and guide 3 is devoted to it; here we just need to know it exists and rides on this double layer.

DLVO: Adding Attraction and Repulsion Together

Now the whole story falls out of one addition. Two Russian pairs — Derjaguin and Landau, Verwey and Overbeek — realised in the 1940s that you can just add the two energies at every separation D and read off the fate of the pair. This is DLVO theory. The van der Waals term is always negative and bites hardest up close (~ -1/D). The double-layer repulsion is positive and dies away over the Debye length (~ exp(-kappa times D)). Their sum, V_total(D), is a curve with a very particular shape — and that shape is everything.

DLVO interaction energy   V(D) = attraction + repulsion

  V |
    |          _
  + |        .' '.  <- ENERGY BARRIER (its height decides stability)
    |       .     '.
  0 |____.__'________'.__________.__________  D (separation) ->
    |   .              '.     .'   secondary minimum
  - |  .                 '---'     (shallow: loose, reversible flocs)
    | .
    | .   PRIMARY MINIMUM
    |.    (deep well near contact:
    ||    van der Waals wins -> hard, irreversible sticking)

  attraction  V_A ~ -A R / (12 D)     always pulls  (Hamaker A ~ 10^-20 J)
  repulsion   V_R ~ exp(-kappa D)     range ~ Debye length 1/kappa

  barrier > ~15-25 kT  -> collisions bounce off  -> suspension STABLE
  barrier ~ 0 (add salt, or sit at IEP) -> collisions stick -> FLOCCULATES

Read the curve left to right. Very close in, attraction wins and the sum plunges into a deep primary minimum — the permanent trap from the last section. But if the surface charge is high and salt is low, repulsion pokes above zero in the middle distance to raise an energy barrier, a hill two colliding particles must climb before they can reach that pit. Farther out, a whisper of leftover attraction can dig a shallow secondary minimum, where particles loiter loosely and reversibly. Here is the key idea: stability is not about which minimum is deeper — the primary always wins that — it is kinetic, about whether particles can get over the barrier. A rough rule of thumb: a barrier taller than roughly 15 to 25 kT stops nearly every collision, since a particle rarely borrows that much thermal energy at once, and the suspension stays dispersed for practical times. Flatten the barrier toward zero and every collision sticks — the suspension flocculates fast.

When the Barrier Falls: Salt and pH

DLVO does more than draw a pretty curve — it tells you exactly which knobs move the barrier, and both are things you can dial in a beaker. The first is salt. Recall that added electrolyte thins the counter-ion coat (a smaller Debye length), so the repulsive term dies away over a shorter distance. The barrier shrinks, and past a threshold called the critical coagulation concentration it vanishes altogether and the slurry flocculates. There is a sharp twist here worth knowing: multivalent counter-ions are savagely more effective. The Schulze-Hardy rule says the concentration needed scales as roughly 1/z^6 in the ion's charge z, so a trace of a divalent ion like Ca2+ can flocculate a suspension that would shrug off far more Na+. This is why a slurry that behaves in deionised water can curdle the moment it meets hard tap water.

1. Start well-dispersed: high surface charge, thick double layers, a tall DLVO barrier — every collision bounces off.
2. Add salt (or shift pH toward the isoelectric point): the double layer thins, or the surface charge falls.
3. The repulsion decays over a shorter range, so the barrier drops lower and lower.
4. Past the critical coagulation concentration the barrier reaches zero — nothing now stops the van der Waals pull.
5. Colliding particles drop straight into the deep primary minimum and lock together: the slurry flocculates.

The second knob is pH, and it works differently — it changes the charge itself rather than the coat's thickness. As you swing the pH, the surface charge (and with it the zeta potential) climbs, passes through zero, and flips sign. The pH where it crosses zero is the isoelectric point (IEP): there the particle carries no net charge, the repulsive term collapses, and only van der Waals is left, so the suspension flocculates hardest right there. Steer the pH well away from the IEP — for alumina the IEP sits near pH 9, so an acidic slip around pH 4 carries a strong positive charge and disperses beautifully. The IEP is such a central idea, and the zeta potential such a central measurement, that guide 3 gives them a whole chapter.

Steric Stabilisation: a Polymer Cushion

Charge-based (electrostatic) stabilisation is elegant but fragile: it lives and dies on pH and salt, and it barely works at all in non-aqueous liquids, where there is no ionic cloud to build. So potters and engineers reach for a second, sturdier trick — steric stabilisation. Coat each particle with polymer chains that dangle out into the liquid like the bristles of a brush, anchored at one end to the surface. When two coated particles approach, their brushes cannot interpenetrate freely: squeezing the chains together crowds them (an unfavourable osmotic pressure sucks solvent in to push the particles apart) and robs them of wiggle room (an entropic, springy resistance). The result is a soft but firm cushion that halts the particles a brush-thickness apart, well outside the reach of that -1/D pit.

Steric repulsion shrugs off the two things that break electrostatic stabilisation: it is largely deaf to salt, and it works in oil or solvent just as well as in water. Its price is that it must be done right. The polymer needs a good solvent to stay swollen (in a poor solvent the chains would rather clump than mix, and stabilisation fails); the surface must be fully covered; and the chains must be long enough to hold particles beyond the attraction's reach. Add too little polymer and you can trigger the opposite of what you want — a single long chain grabs two particles at once and staples them together, a mishap called bridging flocculation. Many real deflocculants hedge their bets by being charged polymers, stabilising by charge and bulk at once (electrosteric); guide 4 gets specific about which molecules do the job.

Stable versus Flocculated: Reading the Sediment

So the whole art comes down to which side of the DLVO ledger you keep the particles on — and you can see the answer with a jar and an hour of patience. A well-dispersed suspension, with its tall barrier intact, holds each particle apart; the particles drift down slowly one by one and pack into a thin, dense, and often stubbornly hard sediment. A flocculated one is the opposite: particles clump into open, fluffy flocs that fall fast and pile into a bulky, loose, low-density sediment that stirs back up with a nudge. That contrast is the cheapest diagnostic in the lab — watch the sedimentation and you are reading the DLVO curve with your eyes.

Here is the honest twist that makes this an engineering choice rather than a chase after one 'right' answer. For most work you do want the well-dispersed state: it lets you cram in a high solids loading at a low, pourable viscosity, and it packs particles densely, so a cast piece dries to a high green density that sinters uniformly. But a perfectly dispersed slurry has a vice — the particles settle out and the coarse ones outrun the fine, so the batch segregates before it can be shaped. That is why ceramists often court a mild, deliberate flocculation: a gentle floc network gives the slurry a small yield stress that suspends everything and holds a shape, without letting it curdle into a solid. Dispersed slips for slip casting, slightly flocculated pastes for others — the dial is set per forming method.

Step back and the payoff of DLVO is a single mental picture: two particles connected by an energy-versus-distance curve with a deep pit, a hill, and maybe a shallow dip, and four knobs — surface charge, pH, salt, and adsorbed polymer — that raise or flatten that hill at will. Master this and you no longer stir slurries by superstition; you decide in advance whether a batch pours like cream or sets like pudding. What the curve does not yet tell you is how the resulting slurry flows under a real push — its viscosity, its yield stress, its thixotropy. That flow behaviour, and how it is tuned for pressing, casting, and extrusion, is the subject of guide 5.