Dispersing Inorganic Pigments in Solvent-Based and UV/EB Systems: Titanium Dioxide, Carbon Black and Iron Oxide
Inorganic pigment dispersion must be treated separately in solvent-based and UV/EB systems.
Inorganic pigment dispersion must be treated separately in solvent-based and UV/EB systems. Inorganic pigments — titanium dioxide, carbon black and iron oxides — present hydroxyl groups, lattice acid sites and residual oxide groups on their surfaces. Their polarity and surface energy are high, and their oil absorption and specific surface area differ widely. In a non-aqueous medium, electrostatic repulsion is largely suppressed by the high dielectric or high viscosity environment of the organic solvent or reactive monomer. Stabilisation must therefore rely primarily on steric stabilisation, with electrostatic contribution secondary. Unlike water-borne systems that depend on the electrical double layer, dispersants in solvent-based and UV/EB systems must adsorb firmly onto the inorganic surface through anchoring groups and then extend solvated chain segments into the resin or monomer phase to build an adsorption layer thick enough to resist competitive adsorption between pigment, resin and dispersant.
1. Surface characteristics of the three pigment classes
Titanium dioxide (rutile or anatase) carries a high density of Ti–OH hydroxyl groups. Inorganic surface treatment with Al2O3, SiO2 or ZrO2 changes the hydroxyl type and acidity or basicity further. Carbon black carries carboxyl, phenolic hydroxyl, quinone and oxygen-containing functional groups. Its surface energy is high, its primary agglomerates are difficult to open, and its specific surface area is usually far above that of titanium dioxide. Iron oxide red, yellow and black (Fe2O3, FeOOH, Fe3O4) show a moderate hydroxyl density with amphiphilic lattice sites; they are light-fast and weather-resistant but have relatively high oil absorption. The practical difficulty therefore differs by pigment. Titanium dioxide is mainly about viscosity reduction and anti-flooding at high volume fraction. Carbon black is mainly about opening primary agglomerates and controlling paste viscosity. Iron oxide is mainly about matching dosage to high oil absorption and preventing re-agglomeration.
2. Steric stabilisation in non-aqueous systems and anchoring selection
In organic solvents or reactive monomers the dielectric constant is low and the double layer is compressed, so the DLVO electrostatic contribution is limited. Stabilisation comes mainly from steric and osmotic repulsion generated by adsorbed polymer chains between particles. Non-aqueous dispersants are therefore typically two-segment hyperdispersants. The anchoring end must be strongly affinitive to the polar surface: acidic groups such as carboxylic, phosphoric or phosphonic acid for titanium dioxide and other metal oxide surfaces, and alkaline, amine or polyamine anchoring groups for acidic carbon black surfaces. The solvated chain must be highly compatible with the resin or monomer phase. For solvent-based systems, choose acrylic or polyurethane segments compatible with the main resin — acrylic, alkyd, polyester or nitrocellulose. For UV/EB systems, choose segments compatible with oligomers such as epoxy acrylate, polyurethane acrylate and polyester acrylate, and with mono-, di- or polyfunctional acrylate monomers. A collapsing chain thins the steric layer and allows the pigment to flocculate again.
3. Compatibility in solvent-based and UV/EB systems
True solvent-free UV/EB systems (FLUV; the EB route uses no photoinitiator) consist almost entirely of oligomer and reactive monomer, with no organic solvent as buffer. The dispersant solvated chain must be soluble in the acrylate or methacrylate monomer phase and must not interfere with curing. Excess dispersant or poor monomer compatibility raises viscosity during storage, reduces cured film adhesion or causes yellowing. RFNC is a resin-free carrier paste but still contains solvent; it is not a solvent-free system, and dispersant selection follows general solvent-based logic. FLTY contains approximately 20 percent PMA and must never be described as solvent-free in export documents or technical data. In solvent-based systems using PMA, DBE or acetate solvents, attention must be paid to the co-solvent window between the dispersant solvated chain and the main resin during solvent evaporation. As solvent evaporates the chain moves from an extended to a collapsed state; if the resin competitively adsorbs and occupies the surface, flooding and floating are likely during the drying stage.
4. Pigment–dispersant–resin competitive adsorption
This is the most common failure mode in non-aqueous inorganic pigment dispersion. Resin molecules are large and present at high concentration, and their polar segments can also adsorb onto the hydroxyl sites of titanium dioxide or iron oxide. When resin adsorption is stronger than the dispersant anchor, the dispersant is displaced, particles lose the steric layer, and the paste re-agglomerates in storage, viscosity climbs and the film floods. Countermeasures: match the anchoring group to the pigment surface — phosphoric or carboxylic anchors for metal oxides, polyamine anchors for carbon black; add the dispersant during the pre-dispersion stage before the main resin so that it wets and occupies the surface first; keep dosage close to just covering the surface, since excess dispersant remains free in the resin phase and raises viscosity; and for high surface energy carbon black, grind in two stages, pre-dispersion coarse then fine, to avoid resin competition in a single charge.
5. Dosage from oil absorption or BET plus gradient trial
There is no cross-system fixed ratio. Engineering practice starts from the pigment oil absorption (GB/T 5211 series) or BET specific surface area to estimate a starting anchor quantity, then runs five gradients — 50, 75, 100, 125 and 150 percent of the estimate — on the same mill, media and tip speed. Fineness is measured to GB/T 6753.1, viscosity to GB/T 9751, and storage stability by seven days at 50 °C observing re-agglomeration and settling. The best point is the lowest viscosity with stable fineness and the smallest heat storage change. Recommended values from a pigment supplier must not be copied directly. Optimum dosage can differ by several times across surface treatment batches, resins or monomers, and equipment.
6. Countermeasure table
Table 1 — Dispersion strategy for three inorganic pigment classes
| Pigment | Surface characteristics | Typical specific surface area and oil absorption | Recommended anchoring group | Solvated chain compatibility | Main failure mode | Dosage and testing |
|---|---|---|---|---|---|---|
| Titanium dioxide (rutile / anatase) | High density Ti–OH; Al/Si/Zr treatment changes hydroxyl and acidity | Moderate specific surface area, medium to high oil absorption, high viscosity at high volume fraction | Carboxylic, phosphoric or phosphonic acid for strong anchoring to metal oxide surfaces | Acrylic or polyurethane chain compatible with main resin for solvent-based; acrylate-monomer-soluble chain for UV/EB | Resin competitive adsorption causing flooding; re-agglomeration during solvent evaporation; viscosity runaway at high loading | Estimate from oil absorption or BET, five-gradient trial; fineness GB/T 6753.1, viscosity GB/T 9751, seven-day heat storage |
| Carbon black | Carboxyl, phenolic hydroxyl, quinone groups; high surface energy, hard primary agglomerates | Specific surface area far above titanium dioxide, high oil absorption, difficult to wet and to reduce in viscosity | Polyamine or alkaline anchoring for acidic oxygen-containing surfaces | Chain must extend fully in resin or monomer; avoid free amine interfering with UV cure and yellowing | Agglomerates not opened, re-agglomeration; overdose raises viscosity; yellowing or adhesion loss in UV systems | Two-stage grinding recommended; start conservatively by BET and increase by gradient; D50/D90 to ISO 13320 |
| Iron oxide red / yellow / black | Moderate hydroxyl density, amphiphilic lattice sites, light-fast and weather-resistant | High oil absorption, high density, prone to settling | Carboxylic or phosphoric anchor for the amphiphilic surface | Compatible with solvent-based resins and UV oligomers; chain long enough to resist settling | Settling and re-agglomeration; high oil absorption limits solids; colour specks and undispersed particles | Match by oil absorption, balance solids; test fineness, particle size and heat storage together |
All dosages in the table are starting directions estimated from oil absorption or BET, not fixed ratios. Final dosage must be confirmed by gradient trial and heat storage on the selected equipment, media and tip speed.
7. Application case summary
A solvent-based industrial anti-corrosion black paste illustrates the method. Carbon black has high specific surface area and oxygen-containing surface groups, so a polyamine-anchored hyperdispersant is selected and added during pre-dispersion before the epoxy or polyamide resin, followed by two-stage bead milling. After the starting quantity is estimated from oil absorption, a five-point gradient is run. The point is fixed where fineness to GB/T 6753.1 is 25 micrometres or below, viscosity to GB/T 9751 is stable, and no re-agglomeration occurs after seven days at 50 °C. A UV white paste (FLUV, true solvent-free) is based on titanium dioxide. The dispersant solvated chain must be soluble in the acrylate or polyurethane acrylate monomer phase, and the cured film must be verified for absence of yellowing and for adhesion. Different pigments and different systems cannot share a single formulation table.
8. Frequently asked questions
Q1. Can inorganic pigments in non-aqueous systems rely mainly on electrostatic dispersion?
No. The dielectric constant of organic solvents and reactive monomers is low and the double layer is compressed, so the DLVO electrostatic contribution is limited. Stabilisation is primarily steric, achieved through anchoring segments plus compatible solvated chains.
Q2. Is more dispersant always more stable?
No. Excess dispersant remains free in the resin or monomer phase and raises viscosity, interferes with UV cure, causes yellowing or reduces adhesion. Dosage should be estimated from oil absorption or BET and optimised by gradient trial at the point of lowest viscosity with stable heat storage.
Q3. Can RFNC paste be used as a solvent-free UV system?
No. RFNC is a resin-free carrier paste but still contains solvent, which is a different class from FLUV true solvent-free UV/EB where the EB route uses no photoinitiator. FLTY contains approximately 20 percent PMA and must not be labelled solvent-free either.
Q4. Which standards apply to fineness and particle size distribution?
Scraper fineness to GB/T 6753.1, paste viscosity to GB/T 9751, and laser diffraction particle size distribution to ISO 13320. The three together determine whether grinding is complete; fineness alone is not sufficient.
Q5. Why does the same paste work in a water-borne system but flocculate in a solvent-based one?
Water-borne systems stabilise through the double layer and electrostatics; solvent-based systems stabilise through steric hindrance. Anchoring and solvated chain design differ, so the system change requires re-selection and a new trial rather than direct transfer.
Key conclusions
- Inorganic pigments — titanium dioxide, carbon black and iron oxide — are stabilised mainly by steric hindrance in solvent-based and UV/EB systems. The anchoring group must match the pigment surface acidity or basicity, and the solvated chain must be compatible with the resin or monomer phase.
- Dosage is estimated from pigment oil absorption or BET specific surface area and confirmed by gradient trial and heat storage; there is no fixed ratio. Competitive adsorption between pigment, dispersant and resin is the main cause of re-agglomeration and flooding.
- System classification must remain accurate: FLUV is true solvent-free UV/EB with no photoinitiator on the EB route; RFNC is resin-free but not solvent-free; FLTY contains approximately 20 percent PMA and must never be described as solvent-free. Fineness, viscosity and particle size are assessed to GB/T 6753.1, GB/T 9751 and ISO 13320.