Colloidal Silica: Particle Size; Stabiliser Balance and the Frozen Drum That Gels a Mould

A drum of colloidal silica arrives frozen in winter and thaws without anyone noticing. The particles have already agglomerated, and the batch that looked normal in the tank gels in the mould.

This note is for buyers and process engineers specifying colloidal silica. It covers what the concentration figure leaves out, how storage and pH decide stability, and the checks that catch a batch before it reaches a process.

Colloidal silica is specified by particle size and stability, not by concentration alone

A thirty percent dispersion and a fifty percent dispersion of the same particle size behave differently in every application, because the solids loading changes viscosity, drying behaviour and gel strength.

Particle size decides surface area, and surface area decides how much stabiliser the manufacturer must use to keep the particles apart.

Two products at the same concentration can therefore have different stabiliser levels, different pH and different behaviour in your process.

Freezing is the failure mode that leaves no visible trace

When the water between particles crystallises, the growing ice pushes the silica particles together and they bond irreversibly.

The drum thaws, looks uniform after agitation, and the only sign is a slightly higher viscosity that nobody records.

Shipping in winter needs heated transport or an insulated container with a declared minimum temperature, and that requirement belongs in the purchase order.

pH is a stability control, not a quality parameter

Colloidal silica is normally stabilised at an alkaline pH, where the particle surfaces carry a charge that keeps them apart.

Contamination with an acid, or dilution with hard water, collapses that charge and the dispersion gels.

Cleaning the tank with the wrong detergent produces the same result, which is why cleaning procedure belongs in the handling instructions.

Investment casting uses the gelling behaviour, and it is sensitive to both size and pH

The slurry has to stay fluid while the pattern is invested, then set hard enough to hold its shape as the mould is built up.

That transition is controlled by the stabiliser and by an additive that triggers gelation, and the two have to match the particle size being used.

Substituting a supplier's product without re-qualifying the slurry recipe is one of the most common causes of shell cracking.

CMP and coatings use the same material for entirely different reasons

In polishing, the particle size distribution determines removal rate and surface finish, and oversize particles cause scratches that scrap a wafer.

In coatings, the particles act as a binder that improves hardness and scratch resistance without adding organic volatiles.

Each application needs its own measurement, and a certificate written for the casting market tells a coatings buyer very little.

Shelf life is real, and it is shorter than the drum suggests

Colloidal dispersions age. Particles slowly aggregate even under good conditions, and the viscosity creeps up over months.

Using material near the end of its shelf life usually works, but it changes the process window, and a plant that does not track batch age is running a variable it cannot see.

First in, first out sounds trivial until a plant discovers it has been drawing from the back of the store for two years.

Incoming checks that catch a bad batch cheaply

pH, specific gravity and a visual check for sediment catch most transport damage and contamination events.

A viscosity or gel-time check against a retained reference sample catches the subtler cases, where the dispersion has partly aggregated without separating.

Keeping a retained sample from each delivery costs a shelf and turns an argument with a supplier into a measurement.

Silica chemistry feeds a wide range of downstream processes, and the abrasives market provides a useful reference point for how demand shifts between foundry, polishing and coatings applications.

Handling and dilution follow rules that are easy to get wrong

Dilution must be done with deionised water, slowly and with agitation, because a sudden change in ionic strength destabilises the dispersion.

Pumps and lines need to be compatible with an alkaline abrasive dispersion, and standard carbon steel is not.

Training operators on the dilution procedure is worth more than any specification clause, because the failure happens in the drum, not in the certificate.

Where colloidal silica sits against the other binders

Against sodium silicate, it offers a longer working life and better dimensional control after firing, at a higher cost per unit of solids.

Against organic binders, it leaves an inorganic residue that survives high temperature, which is why it dominates investment casting shells.

The colloidal silica market is therefore pulled by foundry output and by semiconductor polishing demand at the same time, and the two cycles do not move together.

Transport and storage temperature need to be specified, not assumed

A standard insulated container is adequate in mild weather and inadequate in a hard frost, and the difference is invisible until the drum is opened.

Declaring a minimum transport temperature, and requiring the carrier to record it, converts a silent failure into a documented one.

On arrival, the drum temperature should be checked before acceptance, not after the batch has been charged to a tank.

Traceability matters more than usual because the failure is delayed

A batch that has partly aggregated behaves normally for weeks and then fails during a production run, by which time the drum is empty.

Recording the batch number against the orders produced, not just against the delivery, is what makes the investigation possible.

Without that link, every failure is attributed to the process, and the same supplier problem repeats on the next delivery.

Cleaning and changeover carry their own risk

Residue from a previous product in a shared tank can destabilise a fresh dispersion within minutes of transfer.

An alkaline detergent rinse followed by deionised water is the usual answer, and a conductivity check confirms the rinse has actually worked.

Skipping the check is common, and the resulting gel is then blamed on the incoming material rather than on the tank.

What a specification should contain

Particle size and distribution, solids content, pH, stabiliser type if the supplier will declare it, and viscosity at a stated shear rate.

Add the minimum transport temperature, the shelf life from the date of manufacture, and the tests that accompany each batch.

A specification written this way can be checked on arrival, which is the only point where a problem is still cheap to fix.

The economics are driven by water, so freight dominates the delivered price

A dispersion is mostly water. Moving it costs far more per tonne of useful solids than moving a dry powder would.

That is why producers serve a regional radius and why a distant quote that looks competitive rarely stays competitive once freight is added.

For a plant consuming steadily, a local supplier with a slightly higher ex-works price is usually cheaper delivered, and far more responsive when a batch needs replacing quickly.

Replacement supply is harder than it looks

Qualifying a second source means re-running the slurry recipe or the polishing trial, because particle size and stabiliser level differ between producers.

That work takes weeks, and it is the reason plants stay with a supplier they are unhappy with.

Starting the qualification before the primary source fails is the only way to keep the option, and it is rarely planned for.

FAQ: questions buyers ask

Can colloidal silica be frozen and reused?

No. Freezing causes irreversible aggregation, and the material will not return to its original dispersion whatever is done afterwards.

What is the difference between colloidal silica and fumed silica?

Colloidal silica is a dispersion of discrete particles in liquid. Fumed silica is a dry powder made by flame hydrolysis, used mainly as a thickener.

Why does my slurry gel too quickly?

Usually because the stabiliser level or the particle size changed between batches, or because the gelling additive was adjusted without re-qualifying the recipe.

Does particle size affect surface finish?

Directly. Larger particles cut faster and leave a rougher surface, and a few oversize particles can scratch a polished component.

How should it be stored?

Above freezing, out of direct sunlight, in the original container, and with the drum rotated rather than left undisturbed for long periods.