On a woodfree fine-paper wet end operating at a headbox pH of 7.0–8.5 with continuous cooked cationic starch addition and on-machine ASA emulsification, a pH excursion below 4.0 simultaneously disturbs fibre charge, starch molecular integrity, sizing agent reactivity, filler solubility, and deposit equilibria. The starch component is typically a cationic waxy maize or potato starch with a degree of substitution of 0.02–0.05, cooked in an automated jet cooker at 120–130 °C for 20–30 min, then diluted to 0.5–1.5% solids before addition at the thick-stock pump or thin-stock approach. The ASA is emulsified on-machine through a high-shear dispersion unit at an ASA-to-starch solids ratio of 1:1 to 1:2, generating an emulsion with median particle size of 1–3 µm and a target zeta potential above 20 mV. At pH 4.0 and below, the fibre surface carboxylic acid groups protonate, reducing the anionic site density that cationic starch requires for adsorption; the cooked starch undergoes acid-catalysed depolymerisation of α-1,4-glycosidic bonds; the ASA anhydride ring hydrolyses to a non-sizing diacid; calcium carbonate filler dissolves with release of calcium ions and carbon dioxide; and free aluminium species from alum shift towards Al³⁺. The result is a measurable loss of first-pass retention as determined by ISO 5267-1:2000 drainage tests, a sharp increase in Cobb values measured according to ISO 535:2023, and higher deposition on forming fabrics, press felts, and dryer can surfaces. The following sections examine the separate mechanistic pathways and the on-line sensor strategies that allow a wet end to be stabilised when the pH falls below 4.0.

When pH Falls Below 4.0 in an ASA-Starch Wet End, Which Degradation Pathway Accelerates First?

Between ASA hydrolysis and starch chain scission, the ASA hydrolysis reaction is the faster of the two at equivalent temperature and pH. The ASA molecule contains a five-membered cyclic anhydride that reacts with water to form an alkenyl succinic acid; the hydrolysis rate is acid-catalysed, so a drop from pH 7.0 to pH 4.0 can reduce the hydrolysis half-life from several hours to less than 30 min at 25 °C, and to 5–10 min at typical stock temperatures of 40–50 °C. The hydrolysis product R–CH=CH–CH₂–C(O)–O–C(O)–CH₂–CH=CH–R reacts with water to yield two carboxylic acid groups, increasing the proton concentration and accelerating further hydrolysis in an autocatalytic loop. Once the diacid is formed, it loses the hydrophobic orientation necessary for sizing and can form calcium or aluminium salts with high stickiness. The starch acid hydrolysis pathway is slower but still significant: the hydronium ion cleaves α-1,4-glycosidic bonds in the amylose and amylopectin chains, lowering intrinsic viscosity and reducing the dry-strength contribution. For a cationic potato starch with a degree of substitution of 0.035, laboratory viscosity data show a 20–40% drop in Brookfield viscosity at 50 °C after 60 min at pH 3.5, depending on the buffering salts present; published data for the exact whitewater matrix in high-speed paper manufacturing is limited. The practical consequence is that a pH excursion below 4.0 destroys the sizing response of ASA before the starch loses all of its dry-strength function, but the starch loss is rapid enough to affect strength and retention within one machine retention time.

Under the same low-pH condition, the charge balance between cationic starch and bleached kraft fibre changes markedly. Bleached hardwood and softwood kraft fibres exhibit zeta potentials of roughly -25 to -35 mV at pH 7.0, but at pH 4.0 the surface carboxyl groups are substantially protonated and the zeta potential commonly falls to -5 to -10 mV. This reduction in anionic surface charge weakens the electrostatic attraction that binds cationic starch and cationic ASA emulsion droplets to the fibre. Simultaneously, anionic trash from broke, mechanical pulp, and dissolved calcium carbonate filler begins to compete for the remaining cationic charges. Calcium carbonate dissolves according to CaCO₃ + 2H₃O⁺ → Ca²⁺ + CO₂ + 3H₂O; at pH 4.0, the equilibrium solubility of calcium carbonate rises sharply, releasing calcium ions that increase conductivity and form deposits with alkenyl succinic acid and fatty acid soaps. The dissolved carbon dioxide can produce foam in open headboxes and whitewater trays. Free Al³⁺ from alum at pH below 4.5 becomes the dominant aluminium species, and its high charge density can over-cationise the furnish, causing local charge reversal and sticky aluminium–anionic trash complexes. The resulting retention loss is observable as lower first-pass retention and higher turbidity in the forming section tray water. Drainage testing under the conditions of ISO 5267-1:2000 may show an apparent improvement because the sheet becomes less swollen and fines are lost, but the improvement is misleading because the retained fines and fibre fines content decline.

Charge Reversal, Starch Retrogradation, and Hydrolytic Cleavage Reactions

Cationic starch behaviour at pH below 4.0 depends strongly on the type of cationic substituent. Quaternary ammonium starch retains its positive charge across the entire pH range because the quaternary ammonium group is ionised independently of pH; however, the loss of fibre anionic sites means that retention of the starch is still reduced. Tertiary aminoalkyl starch becomes more protonated as pH falls, but its charge density may not be sufficient to overcome the fibre surface protonation, and the acid environment simultaneously hydrolyses the polymer backbone. Amphoteric starches containing both quaternary ammonium and phosphate or carboxyl groups can exhibit an isoelectric point in the range of 4.0–5.5; below the isoelectric point the molecule carries a net cationic charge, but the hydrolytic depolymerisation dominates. Acid-catalysed cleavage of α-1,4-glycosidic bonds produces shorter amylose and amylopectin fragments, which have lower hydrodynamic volume and lower strength-transfer capability. The viscosity of the cooked starch after pH adjustment to 3.5 may drop by 30–60% within 60–90 min at 50 °C, which in turn reduces the starch’s ability to form a continuous film at fibre–fibre joints. Starch retrogradation, the reassociation of amylose helices, is influenced by chain length and concentration; acid-hydrolysed starch with reduced chain length can form gels of different texture, but the dominant wet end effect is the loss of adsorption onto fibre surfaces rather than retrogradation. The starch content in the finished sheet can be measured by TAPPI T 419 om-15, and the dry tensile strength can be measured according to ISO 1924-2:2008; in wet end investigations, a decrease in starch retention of 15–25% is typically accompanied by a measurable loss in tensile index when pH falls below 4.0 without corrective action.

The following compliance checklist summarises the standard test methods and typical control parameters used when auditing a wet end after a pH excursion below 4.0.

If pH Falls Below 4.0, Why Do Calcium Carbonate Filler and Alum Chemistry Dominate Deposit Formation?

Calcium carbonate and alum are the two wet end additives most likely to create deposit problems when the pH drops below 4.0. Calcium carbonate filler dissolves rapidly in acid, releasing calcium ions; the solubility increases by several orders of magnitude between pH 7.0 and pH 4.0. The liberated calcium ion reacts with the hydrolysed ASA diacid to form an insoluble calcium alkenyl succinate, which is tacky and deposits on forming fabrics, couch rolls, and press felts. Aluminium sulfate at pH below 4.5 exists predominantly as Al³⁺, which hydrolyses to a range of polynuclear hydroxyaluminium species with high positive charge. These species can complex with dissolved and colloidal anionic substances, such as oxidised starch fragments, fatty acid soaps, and hydrolysed ASA diacid, forming sticky precipitates that are difficult to remove. The combination of free Al³⁺ and calcium ions in the presence of alkenyl succinic acid is particularly troublesome because the mixed metal carboxylate deposits have both calcium and aluminium cross-links. Deposit removal requires acidic or alkaline boil-out procedures, and the downtime associated with deposit-related felt plugging or fabric filling is typically higher than the cost of preventing the pH excursion. Monitoring of conductivity and calcium ion activity in the whitewater, coupled with ISO 5267-1:2000 drainage measurements, provides an indirect indication of filler dissolution. Published data for deposit composition under acidic ASA wet end conditions is limited, but the underlying chemistry of calcium and aluminium carboxylate precipitation is well established.

Calibrating On-Line pH, Conductivity, and Zeta Potential Sensors Before ASA Addition

pH measurement in an acidic ASA-starch wet end is compromised by starch adsorption, fines deposition, and ASA hydrolysis products on the electrode surface. Differential pH electrodes with a flat or guarded junction and automatic temperature compensation are preferred over standard glass electrodes for this service because the reference junction remains more stable when coated with starch film. Calibration must be performed with traceable buffer solutions at pH 4.01 and 7.00 at 25 °C; the electrode slope should remain within 95–102% of Nernstian response, and the offset should not exceed ±15 mV. When the pH falls below 4.0, the electrode response often becomes sluggish because the active glass membrane is coated with alkenyl succinic acid and starch fragments; on high-speed gap formers producing woodfree printing papers, maintenance intervals for pH probes can drop from 7 days to 8–24 h. Conductivity sensors should be located after the thick-stock pump and in the whitewater tray to detect the rise in dissolved calcium and sulphate ions that accompanies calcium carbonate dissolution and alum dissociation. Streaming potential or zeta potential sensors on the headbox furnish provide a continuous charge signal; at pH 4.0 the furnish zeta potential often shifts from a normal range of -15 to -25 mV to -5 to 0 mV, indicating loss of anionic fibre charge and impending retention failure. Automatic acid and base dosing lines must be interlocked with the pH sensor and configured with dead-band control of ±0.2 pH units to avoid overshooting. If the pH drops below 4.0, the preferred response is to reduce or stop the Al³⁺-forming alum source, add a buffering agent such as sodium hydroxide or sodium bicarbonate in a controlled thin-stock loop, and reduce ASA addition until the headbox pH is restored to 6.0–7.5. The instrumentation should be verified against grab samples using a portable pH meter calibrated with the same buffer set and a charge demand analyser using a polyelectrolyte titrant standardised according to the supplier’s procedure. Published data for on-line zeta potential response during acid excursions in papermaking is limited, but the relationship between pH, fibre charge, and starch adsorption is well documented in colloid science literature.

Operating boundaries for this wet end chemistry are defined by the combined tolerances of the cooked cationic starch, the ASA emulsion, the calcium carbonate filler, and the retention system. The cooked starch should not be intentionally acidified below pH 4.0 for viscosity control because acid hydrolysis reduces the emulsifying and strength-building functions and because the resultant low-viscosity starch may fail to protect the ASA droplets from coalescence. Direct acid addition to an ASA emulsion is incompatible with sizing because it accelerates anhydride hydrolysis and produces tacky diacid deposits. Alum should not be added directly to the starch slurry before the emulsification step, because the resulting aluminium–starch complexes can cause gelation and clog the high-shear unit. When the pH of the headbox furnish falls below 4.0 due to broke return, acid cleaning residues, or excessive alum, the ASA addition rate must be reduced or stopped until the pH is restored to the 6.0–7.5 range; otherwise the hydrolysed ASA will form deposits on forming fabrics and press felts and the size specification measured by TAPPI T 441 om-13 or ISO 535:2023 will fail. For closed-loop systems with high anionic trash and acidic broke return, maintaining pH above 4.5 is required to preserve starch molecular weight, protect fibre anionic charge, and keep ASA hydrolysis half-life above the machine retention time. In furnishes where pH below 4.0 cannot be avoided for process reasons, the use of acid-resistant fillers such as calcined clay or titanium dioxide instead of calcium carbonate may be necessary, and the cationic starch dosage must be adjusted upward only after the fibre charge has been restored to a zeta potential below -10 mV.