If dry sieve residue is within the reference range before transfer and rises in the receiving tank, investigate the transfer interval first. Check for new aggregates, detached deposits, or differences in how the two samples were collected. When residue rises earlier, align batches by the operation and cumulative monomer feed, then check whether reaction progress differs. The first reproducible increase narrows where to investigate. Identifying the formation mechanism requires evidence from that interval. This workflow covers conventional semi-batch styrene–acrylic and all-acrylic latexes stabilized with anionic surfactants, with or without a nonionic co-stabilizer.

Choose the next check from the first abnormal stage

Use approved sampling points and compare equivalent process states against a representative reference batch. If one feed paused, equal clock times can represent different material additions. Record elapsed time, cumulative feed, and relevant process events together.

Pre-emulsion differs at the feed outlet — Next check: Compare fresh and end-of-hold feed and samples representative of actual withdrawal. · What the comparison resolves: Whether the reactor received the intended feed composition.

Latex residue rises at seed formation or early growth — Next check: Compare seed residue, solids, and the particle-size distribution before the main feed. · What the comparison resolves: Whether the difference was present before the main feed entered.

Residue rises after an aqueous addition — Next check: Reconstruct the stream’s composition, delivered quantity, timing, and entry conditions. Compare stage pH for acid-functional latexes. · What the comparison resolves: Whether the first controlled comparison should focus on stream composition or delivery conditions.

Residue rises later in the monomer feed — Next check: Align samples by cumulative monomer feed and compare conversion or residual monomer, temperature history, and mixing records. · What the comparison resolves: Whether comparable feed progress also represents comparable reaction progress.

Residue rises after neutralization, finishing, or transfer — Next check: Compare matched samples immediately before and after the operation, accounting for line hold-up and sample delay. Collect deposits separately. · What the comparison resolves: Which operation brackets the change and whether deposit release warrants testing.

Feed and latex samples answer different questions. In conventional emulsion polymerization, monomer transfers from droplets through the aqueous phase to growing polymer particles. A uniform pre-emulsion supports the feed-uniformity check. Latex samples are needed to assess particle stability during reaction. See Schork’s review of monomer transport (https://pubs.acs.org/doi/10.1021/acs.biomac.4c00412).

Make sieve residue comparable across batches

Fix the sieve aperture, sample mass, time from sampling to analysis, and the dilution, washing, and drying procedure. Report the result on a defined basis, such as mg of dry retained material per kg of sampled latex, and record latex solids. Use dry sieve residue until the material is identified. Retained material can include polymer aggregates, detached deposits, and foreign matter. Use microscopy or composition analysis when distinguishing those sources would change the next trial.

Keep reactor-wall, line, and filter deposits separate from sampled latex. If particle sizing is performed on filtered samples, record the pretreatment. The size distribution describes the measured filtrate. Use the sieve measurement to track the retained coarse fraction.

Reconstruct what reached the reactor

For an emulsifier change, calculate the surfactant active mass from the delivered product mass and its applicable active-content result. Track cumulative active mass at the first affected stage, including the split between the initial charge and later feeds. Confirm water and other carrier contributions separately. A nonvolatile-solids result can include components outside the surfactant active measurement. If the supplier’s COA gives only solids, ask which method measures active matter and how it is defined. See the guide to active matter and solids in surfactant quotations (https://www.chemicalcell.com/news/1285).

If the difference follows an aqueous addition, compare stream preparation and entry conditions at the actual fill level. For acid-functional latexes, check pH at the affected stage because final pH after neutralization describes a later condition. Reynhout and colleagues (https://doi.org/10.1002/pola.20547) observed increased macroscopic coagulation and fouling under insufficient mixing and concentrated electrolyte delivery in their studied system. They also found that the stability benefit of surface carboxylic groups depended on their ionization. Use those results to choose a testable hypothesis for the affected formulation.

When a pilot trial changes feed concentration, document which quantities remain matched, including solute delivery over time and total formulation water. Changing the solution concentration while matching the solute delivery rate changes solution flow. If flow or the timing and location of compensating water also changes, interpret the outcome as an effect of the tested delivery arrangement.

Cross one suspect lot with one defined process condition

If the reference lot ran in the lab and the current lot ran in production, both the material and the process changed. After locating the first divergence, select one suspect stream and one condition that represents the affected operation, such as pre-emulsion hold time. Test both lots at both condition levels in a suitable development or pilot setup. The four combinations follow NIST’s two-factor factorial design (https://itl.nist.gov/div898/handbook/prc/section4/prc437.htm).

Reference lot — Reference condition: Baseline · Selected scale-relevant condition: Process comparison with the reference lot

Current lot — Reference condition: Lot comparison at reference conditions · Selected scale-relevant condition: Combined lot and process comparison

Match the other lots, seed history, intended active dose, total water, and starting equipment condition. Use independent batches to estimate run-to-run variability. Repeated measurements within one batch estimate measurement variability. Randomize run order where practical or balance both lots across the relevant day or equipment blocks.

A consistent difference between lots across both conditions supports a lot-associated effect in the tested range. A consistent increase for both lots under the selected condition supports a process effect. If the lot difference grows or reverses under that condition beyond the relevant variability, investigate an interaction. A lot-associated result warrants identifying the responsible property before assigning a specification failure. If none of the combinations reproduces the affected stage, review which plant exposure the trial omitted.

Verify the correction at the stage that first changed

Before the confirmation run, define the residue endpoint at the affected stage, how deposits will be assessed, and which product properties must remain within acceptance limits. Confirm the correction in independent batches under production-relevant conditions. If the final filter collects less residue while reactor or line deposits increase, the filter result alone cannot show that the fault was resolved.

Schork’s review of monomer transport

guide to active matter and solids in surfactant quotations

Reynhout and colleagues

NIST’s two-factor factorial design