Continuous production of butyl rubber by low-temperature cationic copolymerization of isobutylene and isoprene in methyl chloride diluent operates as a precipitated slurry process in which the polymer forms discrete particles suspended in a chlorinated hydrocarbon continuous phase. The reactor setpoint is typically -92 °C to -95 °C, and commercial continuous stirred tank reactors are equipped with internal cooling coils, an overhead condenser for boiling methyl chloride heat removal, and a low-shear axial-flow hydrofoil impeller. Slurry stability in this system is not governed by a single variable but by the coupled response of solids concentration, molecular weight, particle size distribution, residual catalyst activity, methyl chloride purity, and local hydrodynamic shear. Industrial lines are commonly operated at slurry solids between 25 wt% and 32 wt%; at solids below 20 wt% the downstream coagulation and finishing stages become energy-intensive, while at solids above 33 wt% the impeller power draw increases disproportionately because the relative slurry viscosity follows a Krieger-Dougherty divergence with a maximum packing fraction φ_m in the range 0.58–0.64 for polydisperse rubber particles. Mooney viscosity ML 1+8 at 125 °C is maintained at 46–55 MU according to ASTM D1646-19a, corresponding to a weight-average molecular weight of 350,000–450,000 g/mol by gel permeation chromatography calibrated against narrow polyisobutylene standards. Temperature excursions greater than ±5 °C around the setpoint increase rubber solubility in methyl chloride, soften the particle surfaces, and initiate irreversible particle coalescence that cannot be reversed by subsequent cooling. Downstream equipment, including steam-stripping vessels and dewatering extruders, further constrains the acceptable particle size; excessively fine slurry particles below 0.2 mm cause carry-over into the solvent recovery system, while particles above 2.0 mm settle in the reactor bottom and reduce heat transfer uniformity.

Slurry Viscosity Escalation and Reactor Fouling Thresholds in MeCl Diluents

In methyl chloride-diluted slurry, viscosity escalation occurs in two regimes. At solids loadings from 20 wt% to 27 wt%, the slurry behaves as a weakly shear-thinning fluid with apparent viscosity controlled largely by the continuous phase and particle translation; impeller power number remains within ±10% of the single-phase value. Between 28 wt% and 32 wt%, particle-particle interactions become significant, and the apparent viscosity at a shear rate of 10 s⁻¹ can increase by a factor of 2–3 relative to the continuous phase. This shear rate corresponds to the average shear in a 2.0 m diameter reactor with an impeller tip speed of 2.2–2.8 m/s. The transition is aggravated by residual aluminum chloride-based catalyst because the catalyst generates local cationic sites on the particle surface and promotes interparticle bridging when fresh monomer diffuses into partially coalesced particles. Reactor fouling initiates on cooling coil surfaces where the local wall temperature is 3–8 °C below the bulk slurry temperature; methyl chloride-swollen rubber particles adhere to the cold metal, form a stagnant film, and then densify over successive residence times. The fouling layer reduces the overall heat transfer coefficient from a clean condition of approximately 350–500 W/(m²·K) to below 150 W/(m²·K) within 72–120 h if the solids level is not reduced or the coagulant injection is not adjusted. Plant experience on continuous lines indicates that the onset of rapid fouling occurs when the volume fraction of particles exceeds 0.40 because the hydrodynamic compression of the adsorbed polymer layer exceeds the steric barrier provided by the chlorinated solvent. The corresponding mass-based solids limit is generally 31–33 wt% depending on particle density and occluded solvent content.

Because the initiation reaction between aluminum chloride and trace water yields a fast carbocationic center, particle formation in methyl chloride proceeds through rapid initiation, propagation, and precipitation after the growing chain exceeds the critical solvent-polymer miscibility limit. Propagation rates at cryogenic temperature are high, although published kinetic constants for this specific system vary with catalyst and impurity levels; localized reaction zones can reach cooling-limited temperatures in less than 1 s. Because the apparent overall activation energy is negative over the range -100 °C to -80 °C, localized temperature increases reduce reaction rate and simultaneously increase polymer solubility, creating a destabilizing feedback loop: soluble polymer raises continuous-phase viscosity, reduces monomer diffusion, and promotes gel-like deposits in the recirculation loop. The precipitated primary particles are composed of loose chains with a high occluded solvent content; occluded methyl chloride can be 30–60 wt% of the swollen particle mass before degassing. The particle size distribution is set by the balance between aggregation of primary particles and breakage by the impeller shear field. A bimodal distribution with a fine mode at 0.1–0.4 mm and a coarse mode at 0.6–1.2 mm improves packing and reduces viscosity at a given solids level, but bimodality is difficult to maintain because coarse particles are selectively retained in the reactor and eventually dominate if the slurry withdrawal point is not placed in the well-mixed zone. Laser diffraction particle sizing of slurry samples from production reactors generally shows a volume median diameter between 0.4 mm and 0.9 mm, with the upper end of the distribution governed by the clearance between the impeller blades and the vessel wall.

How Does Isoprene Placement Alter Chain Swelling and Particle Coalescence?

When isoprene is incorporated into the polyisobutylene backbone, the unsaturation changes not only vulcanization response but also the low-temperature solubility of the polymer in methyl chloride. Commercial butyl rubber grades contain 0.7–3.0 mol% isoprene, with the higher levels used for halogenated butyl grades where allylic bromine or chlorine is introduced later. The random placement of isoprene units disrupts crystallinity and stiffens the chain locally, reducing the potential for crystallite formation at cryogenic temperatures. This disruption increases the methyl chloride uptake of the slurry particles by altering the polymer-solvent interaction parameter; the swollen particle diameter at -90 °C can increase by 8–15% when isoprene content is raised from 1.5 mol% to 2.5 mol%, as measured by ISO 21561 methods for microstructure. A larger swollen diameter reduces the critical solids mass at which particle contacts become permanent. In production terms, a reactor running at 30 wt% solids with a 2.2 mol% isoprene grade may have the same slurry viscosity as a reactor running at 32 wt% solids with a 1.5 mol% isoprene grade. The presence of 1,4-isoprene units also introduces cis and trans configurations, and the trans-1,4 units increase chain flexibility relative to the rigid isobutylene sequences, further modifying particle deformability. Mechanical spectroscopy of concentrated slurries indicates that the transition from viscous to elastic-dominated response occurs at lower solids when isoprene content is higher because the more swollen, deformable particles exhibit greater contact area under compression. This is why high-unsaturation butyl grades are typically produced at the lower end of the commercial solids range and with a narrower temperature control band.

Across the methyl chloride feed circuit, water has a disproportionate influence on slurry stability because it acts as a cocatalyst for the aluminum chloride initiator and also as a poison when present in excess. Methyl chloride as received from chloromethane plants can contain 5–50 ppm water unless dried over molecular sieves or alumina; the target for polymerization is typically below 5 ppm water by Karl Fischer titration following ASTM D6304. Above 15 ppm, initiation becomes spatially heterogeneous around the catalyst feed nozzle, producing a population of very high molecular weight particles that resist flow-induced breakage and create a sticky coarse fraction. The high molecular weight fraction has a Mooney viscosity above 60 MU and can survive steam stripping, appearing as gel in downstream finishing. Conversely, water below 1 ppm can reduce catalyst activity to the point where monomer conversion falls and the slurry density drops; under-converted reactors contain more liquid-phase polymer, which increases continuous-phase viscosity and can cause pump cavitation. Methanol and dimethyl ether are common oxygenated impurities in methyl chloride and act as chain transfer agents, lowering molecular weight and softening particles when present above 10 ppm. The combined effect of water and oxygenates is usually monitored by gas chromatography with an electrolytic conductivity detector on a 30 m capillary column; however, published tolerance limits for each impurity in this specific polymerization configuration are limited and often plant-specific.

When Methyl Chloride Purity Drops Below 99.5 wt%, Slurry Stability Requires Supplemental Drying

Methyl chloride purity below 99.5 wt% shifts the precipitation boundary and broadens the particle size distribution. The principal contaminants—moisture, methanol, dimethyl ether, and higher chlorinated methanes—each alter polymer-solvent interactions differently. Methylene chloride, for example, is a stronger solvent for butyl rubber than methyl chloride, and its presence at 0.5–1.0 wt% increases particle swelling sufficiently to lower the maximum stable solids by 1–2 wt%. Chloroform and carbon tetrachloride have even greater solvent power but are usually present at trace levels below 50 ppm. The commercial response to off-spec methyl chloride is to reduce reactor solids from 31 wt% to 27 wt% and to increase the coagulant dosing in the stripping vessel while the dryer system is regenerated. Molecular sieve dryers with a minimum adsorption capacity of 12 wt% water relative to sieve mass and a bed superficial velocity of 0.5–1.0 m/s are standard; regeneration is conducted at 230–250 °C under dry nitrogen. Without supplemental drying, water excursions above 20 ppm in the recycle methyl chloride stream can initiate cyclic fouling because the destabilized particles that deposit on the reactor walls release water upon decomposition of the catalyst residue, creating localized high-activity zones on the fouled layer. The resulting hot spots on the wall are typically only 2–5 °C above bulk conditions but are sufficient to accelerate particle coalescence in the immediate vicinity.

Assessing Catalyst Jet Penetration and Gel Agglomerate Formation

The method of catalyst introduction determines whether the reactor operates as a stable slurry or as a partially gelled fluidized bed. Aluminum chloride catalyst is typically diluted in methyl chloride and injected through multiple symmetrically spaced nozzles into the reactor at a superficial velocity of 1–3 m/s to prevent backflow and local high catalyst concentration. Each injection point creates a reaction zone in which the conversion can exceed 20–40% within the first 0.5 s, generating high local particle concentrations and heat. If the nozzles are unequally fouled or if the dilution flow is maldistributed, the local solids concentration can reach 35–40 wt% at the injection point, leading to gel-like agglomerates that persist through the remainder of the reactor. These agglomerates have a characteristic size of 5–20 mm, are rich in catalyst residues, and tend to adhere to the impeller shaft and baffles. Removing them requires a reactor wash with methanol or water, which disrupts production and lengthens the turnaround. The spacing and number of catalyst nozzles are designed so that the jet momentum ratio, defined as the catalyst stream mass flux divided by the bulk slurry mass flux, remains below 0.2. At higher momentum ratios, the jet penetrates too far into the slurry and impinges on the vessel wall; at lower ratios, the catalyst is consumed near the nozzle and produces fine particles that circulate but do not aggregate sufficiently. The stable operating window for a 10 m³ reactor is typically 8–16 catalyst injection points, depending on circulation rate and impeller pumping capacity.

Heat transfer and slurry pump performance are coupled through viscosity and particle size distribution, not through temperature alone. Centrifugal slurry pumps in methyl chloride service are typically specified with open impellers, hardened wear plates, and a minimum suction pressure corresponding to a net positive suction head margin of 1.0–1.5 m above the boiling point of methyl chloride at the pump inlet. When reactor solids exceed 31 wt% or when the coarse particle fraction above 1.0 mm exceeds 10 wt%, pump vibration increases and discharge pressure oscillations become measurable. In some continuous plants, the reactor circulation rate is reduced automatically when pump motor current exceeds a preset threshold by 5–8%, which itself reduces heat transfer and accelerates fouling. Plate-and-frame heat exchangers in the external slurry loop are generally avoided because the narrow gaps plug rapidly with swollen rubber particles; shell-and-tube exchangers with tube inner diameters of at least 25 mm and slurry velocities above 2 m/s are preferred. The relationship between shear rate, particle deformation, and heat transfer is often evaluated at pilot scale using a 50 L jacketed CSTR before full-scale implementation; however, scale-up from pilot to production is nonlinear because the impeller Reynolds number changes by more than an order of magnitude between scales and because the pilot vessel typically operates with a lower solids concentration due to greater wall cooling per unit volume.

Post-polymerization handling extends the stability boundary into steam stripping

Once the slurry leaves the reactor, the stability limit shifts from polymerization-driven aggregation to coagulation and solvent removal. The reactor slurry is typically transferred to a steam stripping vessel where hot water and steam contact the methyl chloride-swollen particles, flash the solvent, and precipitate the polymer as crumb. The transfer line must maintain a velocity of 1.5–3.0 m/s to prevent settling without imposing excessive shear that would generate fines; line sizes are often selected to limit pressure drop to 50–100 kPa over the transfer distance. If the reactor slurry has already begun to agglomerate, the transfer line can become blocked at elbows and reducers where the local velocity drops below 0.5 m/s. Steam stripping efficiency depends on particle size, with particles above 2.0 mm requiring longer residence times or higher steam rates to reduce residual methyl chloride below 10 ppm in the dried product. Residual solvent content is measured by headspace gas chromatography following ISO 13302 or equivalent plant methods. The steam stripping vessel is typically operated at 60–80 °C and 20–50 kPa gauge to balance solvent removal against crumb agglomeration; published data for the exact slurry stability limit at the stripper inlet is limited because the transition depends heavily on the upstream particle size distribution and residual catalyst deactivation. Premature addition of coagulant or pH control agents before adequate deactivation of aluminum chloride can create a sticky crumb that adheres to the stripper walls and downstream dewatering extruder screens.

Residual catalyst deactivation downstream of the reactor controls whether the slurry remains stable during transfer and stripping. Aluminum chloride residues are typically deactivated by addition of a stoichiometric excess of water, methanol, or an aqueous base such as sodium hydroxide or calcium stearate, depending on the intended product grade and downstream halogenated butyl processing. If deactivation is delayed until the stripper, the active catalyst continues to oligomerize residual isobutylene and isoprene inside the particles, raising local molecular weight and releasing heat. This post-reactor reaction can raise the temperature of the transfer line by 2–4 °C, enough to soften the slurry particles and cause line coating. The deactivator dosing rate is often set to maintain a molar ratio of deactivator to aluminum of 10:1 to 20:1, with the residual aluminum in the crumb measured by inductively coupled plasma optical emission spectroscopy and controlled to below 50 ppm in finished butyl rubber. Antioxidants such as butylated hydroxytoluene or octadecyl-3-(3,5-di-tert-butyl-4-hydroxyphenyl)propionate are added to the slurry or crumb at 0.05–0.30 wt% to prevent polymer degradation during finishing; these additives do not generally alter slurry stability but can reduce the effectiveness of certain coagulants if added too early.

Molecular weight distribution, rather than average molecular weight alone, determines the tendency of particles to cold-flow under impeller shear. A high polydispersity index above 4.0, as measured by gel permeation chromatography, produces a low-molecular-weight fraction that plasticizes the particle surface and a high-molecular-weight fraction that resists breakage. This combination yields a sticky coarse fraction with poor packing. Control of molecular weight distribution is achieved by adjusting chain transfer to monomer, solvent, and added chain transfer agents such as diisobutylene, while maintaining reactor temperature and monomer conversion within narrow bands. The ratio of weight-average to number-average molecular weight is typically 2.8–4.5 for commercial butyl rubber; outside this range, slurry handling problems become more frequent. In continuous reactors, broadening of residence time distribution caused by poor mixing or stagnant zones further widens the molecular weight distribution and destabilizes the slurry. Reactor campaigns that fail to maintain adequate impeller pumping often show a characteristic shift in the coarse particle fraction at the discharge from 5 wt% to 18 wt% within 24 h, requiring immediate correction of agitator speed or slurry withdrawal points.

Residual isobutylene concentration in the slurry leaving the reactor is a direct indicator of conversion and a predictor of downstream stability. In commercial units, monomer conversion in the reactor is typically 80–95%; lower conversion increases the concentration of unreacted monomer in the recycle methyl chloride and swells the particles further. The flashed monomer from the stripper is compressed, condensed, and returned to the reactor, but if the recycle stream contains heavier oligomers, the maximum stable solids concentration is reduced by 1–3 wt% because the oligomers plasticize the particles and lower the glass transition temperature of the swollen polymer phase. On-line gas chromatography of the reactor overhead can provide a conversion estimate with an uncertainty of ±2%, and this value is used by process control systems to trim catalyst flow. A rapid drop in conversion without a corresponding change in catalyst flow often indicates an impurity excursion in the methyl chloride feed or a fouled catalyst injection nozzle, both of which lead to slurry instability before the operator observes a change in agitator power.

Materials compatibility with methyl chloride at cryogenic temperatures and with aluminum chloride residues restricts the selection of elastomers, gaskets, and pump seals. Chlorinated solvents swell natural rubber, nitrile rubber, and ethylene-propylene-diene monomer seals to different extents; perfluoroelastomer and polytetrafluoroethylene are preferred for dynamic seals in slurry service. Gasket materials should have a volume swell below 10% after immersion in methyl chloride at -20 °C for 70 h when tested according to ASTM D471. Stainless steel grades with molybdenum, such as 316L, are commonly used for wetted parts because aluminum chloride can generate hydrochloric acid in the presence of moisture, and pitting corrosion at low spots can introduce iron into the slurry, which affects color and stability. Published data for corrosion rates in the specific methyl chloride-water-aluminum chloride system is limited, but plant inspections typically show pitting rates below 0.1 mm/year when the pH of condensed water remains above 4.0.

Mechanical reliability of the slurry handling system imposes additional boundaries on the polymerization operating window. Reactor agitator seals in methyl chloride service must withstand low temperatures and the swelling effect of the solvent on elastomeric O-rings; perfluoroelastomer or polytetrafluoroethylene-based seal materials are typically specified. Seal leak rates above 5 mL/h can introduce moisture and air, both of which destabilize the catalyst and the slurry. Agitator shaft vibration is monitored continuously, and a trip is commonly set at 4.5 mm/s root mean square velocity in the horizontal plane. When vibration exceeds this limit, the immediate operational response is to reduce solids and increase methyl chloride dilution, even though this reduces reactor throughput. Slurry density meters based on Coriolis or nuclear attenuation principles provide feedback for solids control; the nuclear density meter reading is calibrated against gravimetric solids every 8 h to correct for occluded solvent changes. The acceptable deviation between the online meter and laboratory solids is typically ±1.5 wt%. Operators also monitor the power draw of the slurry transfer pump because an increase of 15–20% at constant flow often precedes a line blockage. These mechanical constraints, rather than reaction kinetics alone, frequently define the maximum stable solids concentration in methyl chloride-diluted butyl rubber polymerization.