Why Suspensions Keep Separating and What to Do About It
A suspension is just a heterogeneous mixture where solid particles are dispersed in a liquid medium. The particles are big enough to eventually settle out under gravity. That's the whole definition. In a true solution, solute molecules are nanometer-scale and stay dissolved indefinitely. In a suspension, you're looking at particles ranging from roughly 1 micrometer up to several hundred micrometers. They'll settle. The question is always how fast and whether you can control it. The definition is simpler than most textbooks make it sound. A suspension is a two-phase system where the dispersed phase (the solid) is insoluble in the continuous phase (the liquid). The particles are large enough to scatter light—that's why suspensions look cloudy or opaque—and large enough to be captured by standard filter paper. If it passes through a 10-micrometer filter, it's probably not a suspension anymore. That's a practical rule I use in the lab when I need to quickly classify something. The science behind making them stable has three main components. First is sedimentation kinetics, governed by Stokes' Law. Second is the interparticle forces that determine whether particles aggregate or stay dispersed. Third is the viscosity of the medium, which slows everything down.
Stokes' Law is straightforward: v = (2r²g(p - m)) / (9) Velocity of settling depends on particle radius squared, the density difference between particle and medium, gravitational acceleration, and the viscosity of the medium. Most beginners fixate on particle size because it's squared in that equation. Halving the diameter cuts settling velocity by four. That's the single most impactful variable. But it's not the only one you should be tuning.
Practical Stabilization Strategies
There are two main approaches to stabilizing a suspension, and they're often used together. Electrostatic stabilization relies on surface charge. Particles in water develop a charge at their interface with the liquid, creating what's called the electrical double layer. When two particles approach, their double layers repel each other. This is classic DLVO theory—Derjaguin, Landau, Verwey, and Overbeek. The balance between van der Waals attraction and electrostatic repulsion determines whether the suspension stays dispersed or flocculates. Steric stabilization is different. You add large molecules—polymers, surfactants, proteins—that adsorb onto the particle surface. When two particles come close, the polymer layers physically prevent them from getting near enough for van der Waals forces to dominate. This approach works in both aqueous and non-aqueous media, which is a significant advantage over electrostatic methods that require an ionizable surface and sufficient dielectric constant. I've found that combining both strategies usually gives better long-term results. A system with moderate zeta potential and a steric barrier from a small amount of polysorbate or PEGylated surfactant tends to survive temperature cycling and dilution better than either approach alone. Zeta potential above plus or minus 30 millivolts is the usual target, but that number alone tells you very little about actual stability. I learned this the hard way.
Get the Full Details
A Real Problem I Ran Into
Last year I was formulating a pigment dispersion for an industrial coating. The zeta potential measurements looked fine—all readings were in the minus 35 to minus 42 millivolt range. Based on that data alone, I would have said the suspension was stable. Three weeks later, the batch had developed a hard cake at the bottom of the container that wouldn't redisperse with normal agitation. The particles weren't just settling. They were aggregating into a dense, consolidated mass. The problem was secondary minimmum in the DLVO potential energy curve. At those high ionic strengths in the formulation, the energy barrier was tall enough to prevent rapid aggregation but low enough that over weeks, particles could occasionally tunnel through it and stick. Once they formed a compact sediment, breaking those bonds required far more energy than simple shaking could provide. The workaround was adding a small amount of a nonionic surfactant—polysorbate 80 at 0.5 weight percent. This created a steric barrier that raised the effective energy barrier well above the thermal energy available to the particles. The suspension stayed dispersed for months after that. Zeta potential alone had been misleading me because it only measures the electrostatic component. It says nothing about whether steric forces are present or absent.
Common Pitfalls
People often assume that smaller particles automatically mean better stability. That's true for settling velocity, but smaller particles also have higher surface energy, which drives aggregation. If you grind particles down to sub-micron sizes without adequate surface modification, they may settle slowly but then form a gel network that's nearly impossible to break up. There's a tradeoff here between sedimentation rate and redispersibility. Another mistake is ignoring the density match between particle and medium. If the density difference is small, sedimentation is slow regardless of particle size. I've seen formulators spend weeks optimizing particle size distribution when a simple density adjustment—adding a small amount of glycerol or sugar to the aqueous phase—would have achieved the same result in hours. The trick is that matching densities exactly is rarely practical and sometimes impossible if the particle density is very high, like with metal oxides or heavy pigments. But getting within 0.1 grams per cubic centimeter makes a noticeable difference. Flocculated versus deflocculated systems is another area where people get confused. A flocculated suspension forms loose, open aggregates that settle quickly but redisperse easily. A deflocculated suspension has individual particles that settle slowly but form a dense, hard cake. For most practical applications, a controlled flocculated system is actually preferable because the product remains usable after standing. You want the suspension to look like it's settling but actually be forming weak flocs rather than compacting into a solid layer.
Measuring What Matters
Zeta potential is useful but insufficient on its own. Particle size distribution by dynamic light scattering tells you about the size range and polydispersity. Visual observation of sedimentation over time under controlled conditions—some people call this a centrifuge stability test when you spin the samples, though regular storage at elevated temperature works too—is the most honest test you can run. No instrument reading predicts actual shelf-life better than sitting with the product and watching it. For rheology, you want to measure viscosity and yield stress. A suspension with a measurable yield stress won't settle at all until the applied stress exceeds that threshold. That's why thickeners and rheology modifiers are standard additives. Carrageenan, xanthan gum, and fumed silica are common choices. The amount needed depends on particle loading and size distribution. In my experience, getting the rheology right usually takes three to five iterations because the interactions between the thickener and the particles aren't always obvious from the component properties alone.
Where Suspensions Fail Completely
No stabilization strategy works forever. Temperature extremes can destabilize both electrostatic and steric systems. High temperatures reduce viscosity, which speeds up settling, and they can also desorb stabilizing polymers from the particle surface. Freezing and thawing cycles are particularly destructive because ice crystal formation physically disrupts the dispersion structure. Once the ice melts, the particles are often irreversibly aggregated. Very high solid loading is another failure mode. Above about 50 percent by volume, particle-particle interactions dominate over particle-medium interactions regardless of what stabilizer you add. The suspension behaves more like a paste than a fluid. At that point you're dealing with rheological control, not colloidal stability, and the approach changes entirely. Chemical incompatibility between the particle surface and the medium is the third common failure. If the suspension medium reacts with the particle surface—acidic conditions dissolving a metal oxide, for example—no amount of stabilization will help. The particles are being consumed. You need to change the medium or coat the particles first.
There's no universal solution here. Each suspension system requires individual attention to particle properties, medium composition, and intended use conditions. The ones that look good on paper often fail in practice because the real world introduces variables that ideal calculations don't account for. Temperature fluctuations, shear history during processing, container material interactions, and the age of the raw materials all matter. The lab bench is a controlled environment. The shelf is not.
