Across a five-reactor cold SBR train operating at 5 °C to 8 °C with a butadiene/styrene feed ratio near 71:29 by mass, emulsifier selection determines not only latex stability but also particle number, conversion profile, and coagulum load. Disproportionated rosin acid, converted in situ to potassium soap, exhibits a critical micelle concentration that is substantially lower than that of straight-chain C14–C18 fatty acid soaps under alkaline pH conditions. This low CMC allows the emulsion to maintain a micellar population at concentrations where fatty acid-only systems would exist primarily as dissolved unimers. In continuous cold SBR lines, the emulsifier is metered into the aqueous phase ahead of the first reactor, blended with potassium hydroxide, electrolyte, ferrous sulfate, sodium formaldehyde sulfoxylate, and chelating agent. The resulting soap solution is then contacted with premixed butadiene and styrene under high-shear recirculation. Because disproportionated rosin soap provides a bulky hydrophenanthrene hydrophobic group, its aggregation behavior is less dependent on phase separation temperature than that of saturated alkyl carboxylates, which is advantageous at the low polymerization temperatures required for high trans-1,4 content and reduced branching. The technical justification for selecting disproportionated rosin is therefore tied to its micelle persistence at low temperature and low concentration, not merely to its acid number or rosin acid composition.

Disproportionated rosin is produced by heating gum rosin or tall oil rosin in the presence of a catalyst to convert abietic acid into dehydroabietic acid and dihydroabietic acid, reducing conjugated unsaturation and improving oxidative stability. The acid number is controlled by ASTM D465-15 and typically falls in the 160–175 mg KOH g⁻¹ range for disproportionated tall oil rosin. Gas chromatographic analysis after methyl esterification quantifies the residual abietic acid content, which is usually held below 5 wt% of total rosin acids. The soap is prepared by saponification with potassium hydroxide in demineralized water; potassium is preferred over sodium in cold SBR because potassium rosin soap gives a lower viscosity aqueous concentrate at 5 °C and reduces the risk of gel formation during storage. The concentration of the soap concentrate is adjusted to maintain an aqueous phase viscosity that can be metered by positive displacement pumps without cavitation. Because the CMC of the final emulsifier package depends on the distribution of rosin acids, the analytical profile of the rosin is treated as a critical raw material attribute rather than a routine lot-release formality.

How Does Disproportionated Rosin Affect Micellization at 5°C?

Published CMC values for potassium rosin acid soaps, aggregated from surface tension and conductometric titrations under alkaline conditions, typically fall between 10⁻³ mol L⁻¹ and 10⁻² mol L⁻¹, depending on the ratio of dehydroabietic, dihydroabietic, and tetrahydroabietic acids after disproportionation. The hydrophobic group consists of a fused three-ring hydrophenanthrene structure with a carboxylate headgroup; the rigidity of this structure reduces the entropy penalty associated with micelle assembly, which lowers the CMC relative to flexible alkyl chains of equivalent carbon number. At 5 °C, the Krafft point of potassium rosin soap is low enough to avoid crystallization, but the micelle dissociation rate is reduced compared with operation at 25 °C. This slow exchange between free emulsifier and micellized emulsifier has practical consequences for continuous reactors because emulsifier added to a cold aqueous stream requires longer residence time in the dissolution tank to reach equilibrium. In diluted latex serum samples drawn from the first reactor, surface tension measurements often show a broad transition rather than a sharp break, because the aqueous phase contains residual monomer, electrolyte, and water-soluble oligomers that alter the activity coefficient of the soap. Consequently, CMC values determined in pure water cannot be directly transferred to the reactor; the value must be corrected for ionic strength, pH, and the presence of butadiene and styrene saturation at the measurement temperature. The micellar size distribution is also broader than that of a pure anionic alkyl sulfate, which means that the number of micelles per unit volume is not simply proportional to the soap concentration above CMC. This distribution affects the rate of radical capture by micelles and, therefore, the rate of particle nucleation in the induction period.

The solubilization capacity of disproportionated rosin micelles for butadiene and styrene is finite and depends on the aggregation number and the volume of the hydrophobic core. Micelles formed by rosin acid soaps have a more rigid core than those formed by linear alkyl sulfates, which reduces the maximum molar solubilizate-to-soap ratio but also lowers the equilibrium monomer concentration in the aqueous phase. This shifts the partition coefficient for styrene toward the micelle phase and can alter the copolymer composition in the early stages of polymerization if the styrene mass transfer rate from monomer droplets is not sufficient. In practice, the monomer droplet phase acts as the primary reservoir, and the micelle core concentration during the early interval is maintained near saturation. Because the polymerization temperature is low, the diffusion coefficient of styrene in the aqueous phase is reduced; the presence of a lower CMC emulsifier does not eliminate this diffusion limitation but does provide a larger number of micelle interfaces through which monomer can diffuse. Reactor sampling during the first 10% conversion often shows bound styrene slightly below the feed ratio if agitation is insufficient, independent of the emulsifier CMC. This behavior is a mass transfer phenomenon and should not be mistaken for a reactivity ratio effect.

Direct measurement of CMC on the production aqueous phase is carried out before monomer addition. Surface tension is measured by the Wilhelmy plate method according to ASTM D1331-20, using a jacketed glass cell maintained at 5 °C ± 0.5 °C and equilibrated for 30 min before each reading. Conductometric titration may be used in parallel, but the presence of multivalent ions from redox activator solutions depresses the conductivity inflection and broadens the CMC transition. For this reason, surface tension remains the preferred primary method in production laboratories. Table 1 compares the analytical procedures that are applicable to CMC verification in cold SBR emulsifier feedstocks.

Electrolyte Tolerance and Nucleation Kinetics in Redox-Initiated Systems

The low CMC has direct kinetic consequences in redox-initiated cold SBR because particle nucleation is governed by micellar entry of free radicals generated in the aqueous phase by the reaction of organic hydroperoxide with ferrous iron. Under Smith-Ewart Case II kinetics, the number of polymer particles scales with the micelle concentration and the radical generation rate. A low-CMC emulsifier maintains a higher micelle number at equivalent mass loading, which increases particle number and, in a fixed residence time system, raises conversion. However, this effect is constrained by electrolyte concentration. The salt generated by sodium formaldehyde sulfoxylate and ferrous sulfate compresses the electrical double layer around carboxylate micelles, lowering the CMC further but also reducing electrostatic stabilization of growing latex particles. Production experience on continuous stirred-tank reactors equipped with low-shear radial impellers shows that an increase in particle number without a corresponding increase in electrolyte concentration can produce a latex with high viscosity and poor heat transfer in the first reactor. Therefore, the emulsifier-to-electrolyte ratio is treated as a paired control variable rather than an independent formulation parameter. The induction period in cold SBR is particularly sensitive to micelle concentration because the radical generation rate at 5 °C is lower than in hot SBR at 50 °C, and the propagation rate constant is also reduced. A low CMC compensates by providing more sites for radical entry, but it can also lead to runaway nucleation if the soap solution is overdosed after a line stoppage or during restart. Published data for specific continuous cold SBR configurations is limited; therefore, plant validation is required for any change in rosin source or disproportionation degree.

In a continuous emulsion polymerization train, the aqueous soap solution is chilled to 5 °C before monomer addition because the heat of polymerization in the first reactor can raise the reaction temperature by 8 °C to 12 °C if heat removal through the jacket and internal coils is not adequate. The low CMC of disproportionated rosin soap affects this energy balance indirectly through particle number: higher particle number increases the radical flux into particles, raises conversion in the first vessel, and therefore increases the exotherm that must be removed. Reactors are typically equipped with shell-and-tube precoolers on the aqueous and monomer feed lines, as well as internal coil circuits with chilled brine. The temperature control system usually holds the first reactor at 5 °C ± 0.5 °C using a cascade loop that adjusts brine flow based on the rate of temperature rise and the coolant return temperature. Batch-to-batch variance in emulsifier CMC, caused by variation in disproportionated rosin acid distribution, can appear as a shift in the first reactor temperature profile even when the soap metering weight is unchanged. For this reason, the degree of disproportionation is monitored by gas chromatography after methyl esterification, and the acid number is checked by ASTM D465-15. In agitated reactors, the low CMC also changes the apparent viscosity of the latex at a given conversion because smaller particle size increases the particle-particle interaction contribution to rheology. The agitator power draw is therefore recorded alongside reaction temperature and latex total solids to detect nucleation upsets that are not visible from the monomer feed ratio alone.

When Fatty Acid Soaps Are Partially Replaced by Disproportionated Rosin

Partial replacement of fatty acid soap with disproportionated rosin alters the emulsifier hydrophobic group distribution, broadens the micelle size distribution, and reduces temperature sensitivity of the soap solution during storage. The outcome for cold SBR is typically a finer primary particle size and higher latex surface area at equal solids. That increase in surface area raises the demand for post-polymerization antioxidant emulsion because residual rosin soap at the particle surface competes with antioxidant adsorption. Published data for specific rosin-to-fatty acid ratios under continuous cold SBR conditions is limited; therefore, exact Mooney and bound styrene responses must be generated on the target line rather than extrapolated from batch autoclave studies. The selection of replacement ratio is influenced by the electrolyte package, because fatty acid soaps and rosin acid soaps have different sensitivities to calcium and ferrous ions. A mixed emulsifier system may exhibit a CMC below that predicted by ideal mixing of the two soap components, which means that the total emulsifier concentration required to achieve a target particle number can be reduced. However, this synergy is sensitive to pH: fatty acid soaps require pH above 9.0 to remain fully ionized, while disproportionated rosin soap is less pH-sensitive due to the lower pKa of the rosin acid carboxyl group. Table 2 summarizes the production control matrix used to hold the emulsifier system within a narrow operating window during continuous cold SBR manufacture.

Following polymerization, the latex is transferred to blowdown tanks where residual butadiene and styrene are stripped under vacuum, and the dispersion is acidified below the pKa of disproportionated rosin acid to precipitate the polymer. The low CMC of the emulsifier influences coagulation because the concentration of free soap in the serum phase is relatively low, while the adsorbed soap at the particle interface controls the acid demand. A narrow pH window is required to avoid incomplete serum separation on one side and sticky crumb formation on the other. Ash content of the finished dry rubber is checked according to ISO 247-1:2018, and Mooney viscosity is checked according to ASTM D1646-19; both are sensitive to residual rosin soap and coagulant carryover. When the rosin soap is overdosed relative to CMC, the additional electrolyte introduced with the soap increases the ash content and can shift vulcanization kinetics in downstream compounding. Conversely, underdosing below 1.2× CMC can produce latex instability during monomer stripping, causing foam management problems and heat exchanger fouling. These effects are most pronounced in recipes that use low emulsifier levels to reduce raw material cost; the low CMC of disproportionated rosin is then both an advantage and a process risk, because the micelle population cannot be inferred from total solids alone.

Process Water Hardness, pH Drift, and Coagulum Formation

Process water hardness is critical because rosin acid carboxylates form calcium and magnesium salts with lower solubility than sodium or potassium rosinates. At 5 °C, the solubility product of calcium rosinate is low enough that even 10 mg L⁻¹ hardness as calcium carbonate can produce filter plugging in static mixers and heat exchanger preheat sections. Demineralized water with conductivity below 5 µS cm⁻¹ is typically specified. pH drift caused by carbon dioxide absorption into alkaline soap solutions reduces effective alkalinity and can locally drop pH below the pKa of rosin acid, leading to free acid precipitation and loss of micelles. The use of closed soap-dissolving vessels with nitrogen blanketing prevents carbon dioxide ingress. In addition, ferrous sulfate and sodium formaldehyde sulfoxylate contribute electrolyte that interacts with the carboxylate headgroups; at low CMC, micelles persist but the diffuse double layer is compressed, lowering the zeta potential of nucleated particles. This can be beneficial in controlling nucleation rate if the electrolyte mass is held within a narrow window, but it becomes detrimental when multivalent cationic impurities are present. Coagulum formation in the first reactor is a sensitive indicator of this balance; an increase in coagulum under constant agitation and temperature frequently corresponds to a CMC shift caused by hardness or pH drift rather than a change in rosin quality.

Because the CMC of disproportionated rosin soaps is not a single fixed value but a distribution-dependent property, plant qualification requires measurement on the actual aqueous phase and under representative ionic strength. Published data for this specific configuration is limited, particularly with respect to the effects of monomer partitioning and water-soluble oligomers on CMC in the first reactor. The use of ASTM D1331-20 for surface tension and ISO 4311:1979 for conductometric CMC provides a common basis, but neither method fully captures the dynamic micelle exchange that occurs during radical entry and particle nucleation. In practice, the emulsifier concentration is therefore controlled by a ratio to CMC determined in the plant laboratory, and that ratio is stored in the distributed control system as a feed-forward compensation factor for rosin acid distribution and water hardness variation. Finished rubber properties remain the definitive acceptance criterion; they are evaluated by ASTM D1646-19 for Mooney viscosity, ISO 247-1:2018 for ash content, and ASTM D1417-16 for latex total solids and coagulum content. Without these paired measurements, a low CMC alone does not guarantee stable cold SBR production, because particle size distribution, electrolyte tolerance, and coagulation behavior are equally governed by the interaction of the soap with the emulsion and finishing line conditions.