CMP Slurry Stability and Particle Agglomeration
A slurry that performs perfectly on day one is worthless if it drifts in the tank or on the shelf. This guide explains slurry stability, the science of dispersion, why particle agglomeration causes defects, and how to keep a polishing slurry consistent from delivery to drain.
What Slurry Stability Means
Stability is the slurry’s ability to keep its abrasive uniformly dispersed — without agglomerating, settling or changing chemically — from manufacture through storage, dilution and use. It is one of the most important and most underestimated properties of any polishing slurry, and it is built in through careful composition long before the slurry reaches a tool. For the broader context, see the pillar guide.
Crucially, stability is not a single number but a behaviour over time and under stress. A slurry can pass a snapshot measurement yet still drift under the shear, temperature and dilution it sees in a real distribution loop — which is why stability is assessed across the full handling chain, not just at the point of manufacture.
The Science of Dispersion
Abrasive particles are kept apart by a balance of forces. Electrostatic repulsion arises from the surface charge on each particle — quantified by the zeta potential — while steric repulsion comes from adsorbed dispersant molecules that physically block particles from touching. Working against these is the ever-present van der Waals attraction that pulls particles together. As long as repulsion dominates, the slurry stays dispersed; when it weakens, attraction wins and particles begin to clump together.
This balance is the heart of classical colloid stability (DLVO) theory. The practical lesson is simple: anything that lowers the energy barrier between particles — collapsing surface charge, stripping dispersant, or concentrating the system — pushes the slurry toward agglomeration.
How Agglomeration Causes Defects
When the repulsive forces weaken, even a small population of oversized agglomerates shifts the effective particle-size distribution and drives up the large-particle count — and large particles are exactly what produce scratches. In other words, an unstable slurry starts creating the very defects CMP exists to remove. Worse, agglomerates can be transient: they form, scratch a wafer, then partly break up, making them hard to catch after the fact.
Mean particle size can look unchanged on a routine measurement while the large-particle tail quietly grows. That tail, not the mean, governs scratch defectivity — which is why large-particle-count monitoring is a standard incoming-quality control.
Common Triggers
- pH excursions — surface charge collapses near a particle’s isoelectric point, removing the electrostatic barrier to clumping.
- Contamination — stray ions, especially multivalent ones, screen the charge and destabilise the dispersion; incompatible chemistries do the same.
- Shear and pumping — excessive shear can force collisions while stagnant zones let particles settle; both promote agglomeration.
- Temperature and time — heat and long or improper storage accelerate settling, chemical drift and oxidiser decomposition.
- Freezing — many slurries are irreversibly destabilised by freezing during transport or storage.
- Improper dilution or mixing — adding components in the wrong order or too quickly can locally shock the system.
Keeping Slurry Stable: Formulation, Manufacturing, Handling
Stability is protected at three stages. Formulation provides robust dispersant and pH systems with a wide process window and a zeta potential held well away from the isoelectric point. Manufacturing uses high-purity raw materials and tight filtration to control the large-particle tail from the start, plus milling and classification to set the size distribution. Handling covers correct storage temperature, shelf-life discipline, gentle recirculation, protection from freezing, and validated dilution procedures. Weakness at any stage undoes the others — which is one reason supplier process control features heavily in supplier qualification.
Shelf Life, Handling and Incoming QC
Every slurry carries a specified shelf life and storage window; exceeding either invites agglomeration and chemical drift. In the distribution system, gentle continuous recirculation prevents settling without inducing damaging shear, and reactive components such as oxidisers are often blended in only at point of use. On receipt, fabs run incoming quality control — large-particle count, particle-size distribution, pH, conductivity, zeta potential and solids content — to catch a marginal lot before it ever reaches a wafer.
Treat large-particle count as the headline stability metric on incoming lots. It is the most direct predictor of scratch defectivity and the earliest warning that a slurry has begun to destabilise.
The Cost Connection
Instability is expensive in ways that never appear on the slurry invoice: scrapped wafers, extra rework, shorter slurry shelf life and tighter, more frequent quality checks. A single agglomeration excursion can scratch a whole lot of high-value wafers. Treating stability as part of the total cost of ownership — and as a selection criterion in slurry selection — usually pays for itself quickly.
Questions fréquemment posées
Why does slurry stability matter so much?
What keeps abrasive particles dispersed?
What causes CMP slurry to agglomerate?
How is slurry stability measured?
How should CMP slurry be stored and handled?
Does CMP slurry expire?
Talk to the JEEZ slurry engineering team
From first slurry selection to defectivity optimisation and multi-source qualification, JEEZ — Jizhi Electronic Technology Co., Ltd. — helps you match the right polishing slurry to your material and process targets.
Contact JEEZ →Part of the JEEZ Polishing Slurry knowledge series. Reviewed and updated June 2026 by Jizhi Electronic Technology Co., Ltd.