Free Alkali Drift Effects in Full Boiled Soap Saponification

In full-boiled saponification, the nominal completion of triglyceride splitting is not a static chemical end state but a distribution boundary in a multicomponent electrolyte–fatty acid salt system. The kettle content after neat soap fitting comprises at least three separable zones: a top neat soap phase containing hydrated sodium fatty acid carboxylates, a lower nigre phase enriched in glycerol and electrolyte, and a discardable lye layer with concentrated salt and caustic. Free alkali, measured as titratable sodium hydroxide equivalent under ISO 456:2001 and ISO 684:2001, is routinely controlled below 0.10 wt% in toilet soap base but can drift upward or downward within hours when kettle temperature, surface exposure, salt concentration, and glycerol distribution change. The term drift encompasses measurable losses of free caustic through carbonation, migration of caustic from the neat soap into adjacent phases, and apparent increases from hydrolysis of sodium palmitate or sodium stearate at low electrolyte concentration. A full-boiled kettle operating at atmospheric pressure with saturated steam coils at 3 bar to 5 bar absolute and working volumes between 40 m³ and 120 m³ exhibits spatial nonuniformity; the upper neat soap layer may be 2 °C to 8 °C cooler than the bottom lye zone, and caustic migration across the phase boundary follows temperature-dependent partition coefficients. Because the analytical result is method-defined, direct comparison between free alkali as NaOH by ethanol extraction and by aqueous titration is meaningful only when the same solvent system and indicator endpoint are specified. In addition, the measured value does not distinguish sodium hydroxide from sodium carbonate or from carboxylate hydrolysis unless the method includes a barium chloride precipitation step for carbonate and a fixed end-point pH. The full-boiled process is therefore understood as a series of phase-titration states in which free alkali drift is a leading indicator of whether the neat soap will fit, grain, or re-emulsify during subsequent washing.

How Does Carbon Dioxide Ingestion Alter Titratable Caustic in Open Kettles?

Atmospheric carbon dioxide in process air, typically 0.04 vol% to 0.06 vol% depending on ventilation and combustion sources, is absorbed at the alkaline surface of an open kettle and neutralises free sodium hydroxide according to 2 NaOH + CO₂ → Na₂CO₃ + H₂O. The resulting sodium carbonate is less effective as a saponifying agent and alters the titration end-point: under ISO 456:2001, free alkalinity is determined with a pH endpoint near 8.3, and carbonate contributes only partially, so the recorded free caustic drops even though total alkalinity may remain unchanged. This conversion is particularly pronounced when a crust of neat soap remains stationary on the kettle surface for several hours; the high-pH surface film has a large specific area for gas absorption and can convert free caustic into sodium carbonate seam inclusions that later appear as white specks in milled soap. In open kettles with surface-to-volume ratios of 0.1 m²/m³ to 0.3 m²/m³, free alkali drift due to carbonation is commonly 0.005 wt% to 0.02 wt% per hour during extended hold periods, although published data for specific kettle geometries is limited because most plants do not routinely monitor headspace carbon dioxide. At relative humidity above 60%, condensation on kettle walls and surface films accelerates CO₂ uptake; additional soda ash formation can also elevate the ionic strength in the upper neat soap layer, shift the phase boundary downward, and reduce the volume of clean neat soap prior to finishing. Process control therefore includes minimising open-kettle hold time after fitting, maintaining a closed or inert purge where possible, and sampling the top 5 cm of neat soap separately from the bulk phase when free caustic is used as a release specification.

At the conclusion of the final fitting operation, free alkali is not uniformly distributed through the neat soap mass. The continuous aqueous phase within neat soap is a concentrated electrolyte solution in which sodium ion activity is set by the balance of sodium chloride, glycerol, residual caustic, and the carboxylate anion itself. When the salt level in the nigre exceeds 8 wt%, the salting-out pressure on the hydrated soap lamellae rises, and free caustic partitions preferentially into the aqueous nigre because of its higher dielectric constant and greater solvation capacity for hydroxide ions. As a result, the ratio of free NaOH in neat soap to free NaOH in nigre typically remains below 0.5 under equilibrium conditions at 90 °C to 100 °C. A drop in salt concentration below 6 wt% reduces that partition, allows more water into the neat soap, and tends to increase its titratable free alkali by drawing caustic back into the neat phase. Conversely, a rise in glycerol concentration in the nigre above 15 wt% can reduce the caustic activity coefficient and hold free alkali in the lower phase. The phase behaviour is thus a coupled electrolyte–polyol–water system, and free alkali drift cannot be controlled by caustic addition rate alone; the salt-to-glycerol ratio in the recycle stream must also be kept within a narrow band. During the resting phase that follows washing, the free alkali measurement of the neat soap can increase slightly even when no caustic is added. This increase arises because sodium salts of long-chain fatty acids undergo hydrolysis at the organic–water interface: RCOO^- + H₂O ⇌ RCOOH + OH^-. The equilibrium population of hydroxide ions is suppressed by excess sodium chloride and by high soap concentration, but at water contents above 30 wt% and salt contents below 0.5 wt% the hydrolysis can add 0.01 wt% to 0.04 wt% to the titratable free alkali over several hours. This is partly an analytical artefact of the solvent extraction method, because the alcohol-water mixture shifts the hydrolysis equilibrium; it also reflects genuine pH development at the soap lamella surface. Production kettles therefore do not attempt to hit an absolute zero free alkali; instead, the final fit is adjusted to a method-defined plateau that will remain within specification after the hydrolysis and carbonation drifts experienced during hold and transfer. Table 1 lists analytical windows observed in full-boiled soap phases after equilibrium at 90 °C to 100 °C, based on method-defined values rather than universal thermodynamic limits.

When Free Alkali Exceeds 0.15 wt% After the Fitting Stage

If the neat soap leaving the kettle carries a titratable free alkali above 0.15 wt% as NaOH, the finishing line does not simply experience a higher pH; the hydrated soap phase begins to shift its rheological and thermal responses. In a vacuum spray dryer, the excess free caustic can reduce the surface tension of the soap film, produce uneven moisture distribution, and increase the tendency for nozzle fouling if the feed stock is not filtered. In a triple-roll mill with nip gaps of 0.2 mm to 0.6 mm, free alkali above 0.15 wt% is often associated with a shorter plastic working range and higher motor current at fixed roll pressure, because the soap mass becomes stiffer when sodium chloride and caustic lower the water activity of the continuous phase. A plodder operating with a conical barrel wall temperature of 35 °C to 50 °C and die pressures between 40 bar and 90 bar may exhibit surface striations and lower shine if the free alkali is not reduced before final extrusion. The effect is not solely mechanical. Free alkali at 0.15 wt% to 0.40 wt% can intensify the autoxidation of unsaturated fatty acid residues in tallow, palm, or palm stearin soap, producing off-odour development during storage at 35 °C and 65% relative humidity. In addition, high free alkali in the presence of residual glycerol can lead to alkali-catalysed colour bodies and a yellow-to-brown shift in the finished bar. For toilet soap base, a final free alkali specification of ≤0.05 wt% by ASTM D460-91(2014) is often applied at the milled bar stage, while laundry-grade soap may be released with free alkali up to 0.40 wt% because the cleaning function tolerates higher electrolyte. Published data for the exact relationship between free alkali and plodder barrel pressure in full-boiled soap is limited; most production lines rely on trial runs at a given feedstock titre and moisture to define the acceptable upper limit. The controlling principle is that free alkali is not merely a residual saponification variable but a phase modifier that alters how the soap lamellae pack under shear.

After leaving the kettle, the neat soap is often transferred to agitated hold tanks at 70 °C to 85 °C before vacuum drying. During this transfer the free alkali may continue to decline if the holding tank is vented to plant air, because carbon dioxide absorption continues and surface crust can form. At the same time, free alkali can migrate downward from the upper neat soap layer into any separated aqueous heel, causing the upper layer to become slightly more acidic by hydrolysis and the lower heel to become caustic-rich. In an agitated tank this separation is partially masked, but if the agitator is stopped for more than 30 min, the upper layer may differ from the bulk by 0.02 wt% to 0.05 wt% free alkali. For soap bases containing volatile fatty acids or lauric oils, the free alkali drift during holding is coupled to moisture loss; a 1 wt% moisture loss from the top surface can increase the measured free alkali by concentrating the residual electrolyte in the remaining water phase. Operators monitor the free alkali not only at the kettle but also at the vacuum dryer feed tank, because refining and bleaching additions can be ineffective if the soap enters the dryer with a free alkali swing greater than 0.05 wt% against the target. A closed pipe loop or nitrogen blanketing is typically used where carbonation-sensitive formulations are processed. If the free alkali is too low after holding, small additions of dilute sodium hydroxide solution cannot be reliably mixed into a non-Newtonian soap mass without creating localized high-alkali zones; therefore the correction is preferably made at the fitting stage rather than downstream.

Multi-Kettle Countercurrent Wash Dynamics and Caustic Retention Time

In a multi-kettle countercurrent washing arrangement, the free alkali measured in the neat soap after the final wash is the delayed output of several interacting hold times, recycle ratios, and salt cuts. A typical train of four to six kettles operates with the nigre stream returned backward to the preceding wash, while the spent lye stream is segregated for glycerol recovery. The free alkali of the neat soap in kettle N depends on the free alkali of the nigre from kettle N+1, the salt concentration used to grain the soap, and the equilibrium partition coefficient between neat and nigre phases. When the nigre recycle stream contains 0.6 wt% to 0.8 wt% free NaOH, it can carry enough caustic into the previous wash to raise the neat soap free alkali only slightly, but if the recycle stream is not settled properly and contains entrained neat soap, the free alkali transfer becomes nonlinear. The retention time of free caustic in such a train is not equal to the liquid hydraulic retention time because hydroxide partitions into the neat soap during mixing and then redistributes as the phases separate. Experimental data from production kettles has shown that a step change in fitting caustic can require 2 to 4 batch cycles to appear fully in the final neat soap, depending on the recycle ratio and the degree of mixing in the wash tank. This lag is a practical source of drift: an operator correcting the free alkali too quickly can overfeed caustic and produce a later upward swing in the neat soap that is only visible after the batch has been transferred to finishing. Effective control therefore uses a feedforward adjustment based on nigre alkalinity and recycled salt, not simply a feedback loop on the last kettle.

Sampling frequency for free alkali drift is determined by the sensitivity of the downstream formulation to electrolyte and by the carbonation potential of the plant air. For a standard toilet soap base, a composite sample is taken at the kettle discharge, a separate sample from the top 5 cm of stagnant neat soap, and a final sample after the vacuum dryer. The kettle discharge sample is analysed by ISO 456:2001 for free caustic alkalinity and by ISO 684:2001 for total free alkali, while the dryer feed sample may be checked for moisture and chloride in addition. The same drift is also tracked by AOCS Da 4a-48 for free caustic alkali in soap. If the free alkali difference between the top surface and bulk neat soap exceeds 0.03 wt%, the hold time or nitrogen blanketing is adjusted because the surface has undergone carbonation or evaporation. In spent lye, free alkali is commonly 1.0 wt% to 3.5 wt% and is recovered during glycerol evaporation, but the presence of sodium carbonate from carbonation can complicate evaporator scaling and should be monitored as soda ash or carbonate alkalinity. The analytical control matrix for free alkali drift management is summarised in Table 2. These are method-defined control points, not thermodynamic constants, and should be recalibrated when the fat blend titre or the soap grade changes.