Investigate inorganic filler settling by measuring the concentration gradient and testing whether the bottom deposit can be recovered within the available mixing procedure. Then compare particle structure and aged low-shear behavior to select a targeted corrective trial. Sediment height alone cannot establish stability: a thin, resistant bed may be harder to recover than a larger, softer deposit. A correction succeeds when the stored material returns to the required composition and processing performance without exceeding the permitted handling conditions.

This article addresses nonaqueous, uncured mineral-filled liquid resin components, using the resin-side component of a solvent-free epoxy system before hardener addition as the main example. Waterborne latex destabilization and settling during mixed-system cure require separate investigations.

Record the Separation Before Remixing

Photograph an undisturbed container and record clear-layer height, sediment height, fill depth, storage time and temperature history. Preserve separate containers for subsequent time points: repeated opening, probing and remixing change the sample.

Assess the bottom as well as the upper liquid. Describe whether the deposit yields to a defined probe, breaks into fragments, or resists penetration. In its guidance for energy-cured matte coatings, INX International Coatings & Adhesives distinguishes easily reincorporated sediment from a hardened particle deposit and recommends inspecting the container bottom. This supports a practical distinction; its coating-specific operating instructions should stay within that application.

For the resin under investigation, define hard settling operationally: a deposit that cannot be restored to acceptable homogeneity within the permitted mixing procedure.

Take top, middle and bottom samples from dedicated containers before remixing, documenting sampling depths and avoiding disturbance of the lower layers. Use a separate container for bottom probing and another for the recovery trial. This proposed sampling design preserves the distinctions between the original gradient, deposit resistance and recoverability.

A clear upper layer alone does not establish the mechanism. If the liquid and particle profiles do not support simple sedimentation, investigate resin phase separation or contraction of a structured network.

Establish a Filler-Concentration Measurement

Total nonvolatile content cannot distinguish filler from a solvent-free binder when both remain in the residue. NETZSCH describes thermogravimetric analysis (TGA) as a method for compositional and filler-content analysis; interpretation depends on decomposition and residue formation.

For an epoxy component whose only inorganic residue-forming ingredient is alumina, the following is a proposed method-development route:

1. Test the binder formulation without alumina and the supplied alumina separately under the selected TGA program. Establish their residue contributions, including any powder treatment.
2. Test homogeneous mixtures of known alumina mass fraction spanning the expected sample range. Use these to verify the residue-to-content relationship and check recovery and repeatability.
3. Analyze the top, middle and bottom samples using the same program, mass basis and validated sample-mass range. Compare concentration differences with the method’s demonstrated variability.

If matrix residue, filler transformation or poor recovery prevents reliable quantification, resolve the method before interpreting the profile. If fumed silica or other inorganic ingredients contribute residue, a single endpoint cannot identify alumina independently; constituent-specific measurement requires separate validation.

Repeat the validated assay after remixing. Agreement supports homogeneity only for the sampled locations and measured constituents.

Check the Particles Present in the Resin

A powder’s median size cannot describe everything that settles.

The Micromeritics explanation of gravity sedimentation relates settling velocity to equivalent particle diameter, particle–liquid density difference and liquid viscosity through Stokes’ law. For isolated spheres in a Newtonian liquid under its applicable flow conditions, larger diameter and greater density difference increase settling velocity.

Use this relationship to organize hypotheses. Concentrated, irregular, flocculated or yield-stress resin dispersions require additional evidence; a settling-time calculation from D50 and one viscosity value cannot predict their shelf life.

Compare the complete size distributions, coarse tails and particle images for the satisfactory and affected fillers. Record morphology explicitly: plate-like talc or mica should not be interpreted as spherical particles merely because a sizing method reports an equivalent diameter.

Then examine the dispersions themselves. A coarse object may be a primary grain, a poorly wetted agglomerate, or a cluster formed during storage. Check representative samples before storage and after aging, using preparation that preserves the structure being investigated.

A concentrated bed raises a further question: have particles simply accumulated, or has the deposit become progressively harder to mobilize? Treat compaction, strong particle contacts and binder aging within the deposit as competing explanations. Track bottom concentration and recovery effort over time; the feel of a hard bed cannot identify a chemical bond or prove that cure has occurred.

For blends, a similar total filler concentration can conceal a changed constituent ratio. Use a method demonstrated to distinguish those constituents before claiming that the original blend has been restored.

ChemicalCell’s Non-Metallic Minerals category provides a browsing route for materials including quartz, talc and mica; oxide fillers can be explored through the Metal Oxides category. These listings do not establish equivalent particle shape, surface treatment or suitability for a particular resin.

When comparing different-density fillers, state both mass loading and calculated volume loading. A fixed mass replacement changes the comparison if the occupied solid volume changes substantially. Record whether the trial evaluates a direct substitution or isolates a material property.

Separate Dispersion Quality From Suspension Support

Dispersants and rheology modifiers address different parts of the problem.

BYK’s explanation of dispersing additives describes stabilization against flocculation and the viscosity reduction that can accompany deflocculation. A formulation that becomes easier to mix after a dispersant change still needs storage testing.

BYK also describes controlled flocculation: additive-mediated networks can provide anti-settling behavior while remaining breakable under relatively low shear. This is a deliberately formulated state. Uncontrolled aggregation should not be treated as an acceptable substitute.

Evaluate dispersion quality, sediment formation and redispersibility separately. If a dispersant change reduces initial viscosity but produces a deposit that is harder to recover, the processing gain has introduced a storage tradeoff. That observation identifies a formulation response; microscopy and rheology are still needed to explain it.

Fumed silica introduces another distinction. In Evonik’s technical bulletin on solvent-free epoxy resins, Sections 2.4.2–2.4.3, suspension support is associated with development of a yield value. The bulletin also shows that silica grade, silane addition and aging can change rheology and filler suspension in the tested epoxy formulations. The results support compatibility testing of the complete additive package; they do not establish a universal silica grade or dose.

If the suspected change involves silane treatment, ChemicalCell’s silane selection for filled resins addresses reaction route and component placement. Confirm storage behavior separately from any intended adhesion benefit.

Measure What Happens After Shear Stops

A viscosity result at a production-relevant speed describes only part of the storage problem.

Compare fresh and aged dispersions at controlled temperature using the same loading procedure, preshear and rest history. Include low-shear behavior and, where suitable, a yield-stress assessment with its method stated. Check that the geometry accommodates the particles and monitor separation during measurement.

Check for wall slip before interpreting an apparent loss of low-shear structure. TA Instruments explains how slip affects rheological measurements and describes gap comparisons and roughened geometries as diagnostic approaches. Keep the measurement temperature and conditioning controlled when making these comparisons.

Use a recovery test when the material experiences pumping, mixing or filling before storage. Anton Paar’s three-interval method applies low shear, then higher shear, then low shear again to follow structural recovery. Define the shear exposure and recovery times around the relevant operation.

Keep two questions separate: does the formulation retain its low-shear structure during aging, and does that structure recover after handling? A favorable recovery curve supports a rheological explanation. Storage and redispersion tests must establish whether that behavior is sufficient.

Choose the Next Trial From the Evidence

The following is a proposed diagnostic framework, rather than a reported ChemicalCell experiment. Each observation narrows the investigation without uniquely proving its cause.

Start with a retained satisfactory formulation and the affected formulation. Where possible, compare the old and new filler lots in the same resin batch. Keep mixing sequence, temperature, additive composition and storage conditions controlled. Record both initial low-shear behavior and its change during storage so that a higher starting value is not mistaken for better retention.

Change one suspected factor in the first diagnostic comparison. If several changes are needed to produce a usable candidate, record that candidate as a complete reformulation. Use independent preparations to distinguish a repeatable response from mixing variation.

Make Redispersion a Measurable Requirement

Define the permitted mixer, impeller, vessel, fill level, mixing duration and temperature rise before evaluating recovery. Keep these conditions consistent within the comparison.

After remixing, inspect the bottom and repeat the validated composition measurements at representative locations. A uniform-looking upper region can coexist with a residual bed outside the impeller’s effective circulation.

Specify acceptance against the intended operation: allowable filler-concentration variation, residual deposit, delivery consistency and processing behavior. Include constituent ratios when the method can resolve them; otherwise, blend recovery remains unconfirmed. A recovery that requires scraping, additional solvent or mixing beyond the available production window does not validate the original handling procedure.

Check the remixed component through its expected holding and withdrawal sequence. For an epoxy system, combine it with the specified hardener and verify the processing and cured properties that the filler is intended to control. A depleted or enriched sample has a different binder-to-filler ratio; account for that composition before attributing a cure or performance change to filler chemistry.

Confirm the Change Across Storage and Use

LUM describes real-time and physically accelerated analysis of dispersions, while emphasizing that shelf life has no all-purpose prediction method. Use accelerated separation results as a defined screening comparison. Retain real-time storage at the intended conditions and establish relevance to the actual failure mode before making a shelf-life claim.

Confirm the shortlisted change at representative container scale. Include the relevant transport or handling exposure and the longest permitted interval between remixing and use.

The next step is a matched storage-and-remix comparison using a validated composition assay and a bottom-deposit check. Retain a change when it repeatedly restores acceptable composition, delivery and application performance within the available handling procedure. If the composition cannot be verified or the deposit remains unrecoverable, hold the material for investigation under the applicable quality procedure.