In chloromethane-slurry butyl polymerization, the reactor temperature is not a freely adjustable parameter; it is pinned by the freezing point of chloromethane and by the kinetic requirement for high molecular weight. Chloromethane solidifies at −97.7°C at atmospheric pressure and boils at −23.8°C, while the polymerization is generally operated in the range of −90°C to −98°C. The useful control band at the low end is therefore less than 2 K wide. A feed mixture of isobutylene, isoprene, and chloromethane is chilled before entering a stirred reactor, and initiation is carried out by a water-activated aluminum chloride complex dissolved in chloromethane. Because cationic propagation proceeds through a tertiary carbenium ion, the reaction remains rapid even at −95°C; however, chain transfer to monomer, beta-proton elimination, and termination are more strongly retarded by low temperature than propagation. This differential temperature response produces a high-molecular-weight copolymer with commercial Mooney viscosity only if the reactor internal temperature is kept uniform. The heat of polymerization of isobutylene is approximately −48 kJ/mol, with published values spanning −43 kJ/mol to −53 kJ/mol depending on measurement technique, and this enthalpy must be removed by boiling ethylene at near-atmospheric pressure in the jacketed vessel. Because polyisobutylene is insoluble in chloromethane at −95°C, the copolymer precipitates as a fine slurry, which limits the viscosity of the continuous phase but creates fouling and heat-transfer boundaries that dominate reactor operation.

The diluent itself contributes to temperature control in two ways. First, chloromethane is partially vaporized from the reactor headspace and condensed in a reflux loop, providing direct evaporative cooling in addition to jacket heat removal. Second, because the polymerization is conducted as a slurry, the liquid chloromethane acts as a heat-sink fluid that surrounds precipitated polymer particles and reduces the localized temperature rise at the growing chain site. At 20 wt% to 30 wt% solids, the slurry can still be pumped and agitated, but the margin to settling and fouling is small. Reactor temperature is measured by redundant resistance temperature detectors or thermocouples inserted at multiple depths. The control strategy is typically cascade: catalyst feed is manipulated by a flow controller whose setpoint is trimmed by a temperature controller. If the temperature exceeds the setpoint by more than 2°C, the catalyst feed is reduced and the ethylene coolant valve opens further. If the temperature falls below −97.5°C, the coolant supply is throttled and a warm chloromethane bypass loop is activated to prevent localized solidification. The practical setpoint window is thus ±1°C to ±2°C around the target grade temperature, and this window is narrower for high-molecular-weight grades that operate at the lower end of the range.

What Cooling Configuration Prevents Local Solidification at −97.7°C?

The cooling train for a low-temperature butyl reactor is designed around boiling ethylene rather than a single-phase secondary refrigerant. Ethylene boils at −103.7°C at 101.325 kPa, so its saturation temperature is below the chloromethane freezing plateau but within 6 K of the typical reactor setpoint. The jacket is divided into multiple zones, each with a separate ethylene feed and vapor return line, and the reactor also includes internal cooling coils or a draft tube. This arrangement prevents a single refrigerant temperature from creating a cold spot at the bottom head where slurry velocity may be low. The coolant is supplied to the jacket as a boiling liquid; the liquid level in the jacket determines the fraction of wall area exposed to condensing vapor and thus the heat-flux profile. The process-side heat-transfer coefficient is controlled by agitator intensity and by the solids concentration of the slurry. At lower agitator speeds, a layer of high-solids slurry can stagnate near the wall and reduce the local heat-transfer coefficient by 30% or more, and the corresponding wall temperature may fall below the chloromethane solidification point. Published data for specific commercial agitator designs are limited, but the operating constraint is well established: the wall temperature must remain above −97.7°C at all points in contact with liquid chloromethane. To maintain this condition, the temperature difference between the coolant and the reactor contents is kept below 8 K, and the ethylene pressure is adjusted to change the saturation temperature. A lower pressure gives a lower saturation temperature and a higher driving force, but it also increases the risk of forming solid chloromethane on the wall. The control system contains an interlock that closes the ethylene feed if any wall-mounted temperature probe reads below −98.0°C for more than 5 s.

In addition to wall cooling, a portion of the chloromethane may be allowed to vaporize from the reactor headspace, condensed in an external exchanger, and returned as subcooled liquid. This reflux loop removes heat at the vapor-liquid interface and helps maintain a uniform temperature in the headspace. Because the polymerization is exothermic and the reaction mass is viscous at high solids, relying on jacket cooling alone may be insufficient for the highest production rates. The reflux loop also strips a small amount of unconverted isobutylene and isoprene, which are recovered and recycled. The heat removal duty is split between the jacket, internal coils, and reflux condenser in proportions that depend on reactor scale and grade. For a reactor producing butyl rubber at a rate of several tonnes per hour, the total heat removal duty may exceed several megawatts; however, published data for specific reactor-scale duties are limited. The temperature control system is therefore not a simple proportional-integral loop; it includes feedforward from the catalyst flow and the slurry density, and it includes constraint logic based on wall temperature, agitator torque, and reflux accumulator level. Any imbalance among these constraints can cause the reactor to drift toward either a high-temperature condition that reduces molecular weight or a low-temperature condition that freezes chloromethane.

Moisture, Cocatalyst Stoichiometry, and the Molecular Weight Collapse Boundary

Within the same reactor envelope, the moisture content of the chloromethane and mixed olefin feed controls the concentration of the initiating species. Water participates in the formation of the proton-donating complex that activates aluminum chloride. A feed moisture specification of less than 10 mg/kg is typically applied to the mixed olefins, while the chloromethane stream is dried through fixed-bed molecular sieves to below 5 mg/kg. These values are not arbitrary; they reflect the stoichiometric demand for initiation without excessive chain transfer. If the moisture content exceeds the design ceiling, uncontrolled initiation creates too many growing chains, and the number-average molecular weight falls. The resulting rubber may pass the Mooney viscosity test but may exhibit a broadening of the molecular weight distribution and an increase in low-molecular-weight extractables. If the moisture content is too low, initiation becomes sluggish, monomer conversion falls, and the reactor residence time must be increased to achieve the same production rate. Online moisture analyzers using tunable diode laser absorption spectroscopy are placed on the feed lines, and the drying beds are arranged in parallel with automatic switchover based on breakthrough. The adsorbent is typically a 3A or 4A molecular sieve, which excludes chloromethane and hydrocarbons from the pores so that water is adsorbed selectively. The regeneration cycle uses hot nitrogen at 220°C to 260°C and a final cool-down with dry chloromethane or nitrogen. During high-humidity operation, pre-drying of regeneration nitrogen is required at ambient relative humidity above 60%, because the molecular sieve may not reach its design dew point after regeneration. This drying sequence is critical because chloromethane at low temperature has limited water solubility, and any free water can form ice or hydrate deposits in the feed piping and reactor nozzles.

Because chloromethane at −95°C is only a few degrees above its freezing point, any free water present in the feed system can form solid ice or chloromethane hydrate. The hydrate formation temperature depends on pressure and water content; in lines containing stagnant chloromethane, hydrate deposits can restrict flow and create pressure drops that fluctuate. The feed blending system therefore includes a chilled filter or coalescer to remove free water droplets before the stream reaches the final heat exchanger. Heat exchangers are arranged in a descending temperature cascade: the first stage cools the feed from ambient to −20°C, the second stage to −60°C, and the final stage to the reactor feed temperature using ethylene or a flashed chloromethane stream. This cascade prevents a single large temperature difference from causing water or chloromethane to freeze on the tube wall. The final stage is designed with high fluid velocity, typically above 1.5 m/s, to keep the wall shear stress high and to minimize the thickness of any boundary layer where solid might nucleate. If the pressure drop across the final exchanger rises by more than 10% from baseline, the control system initiates a defrost cycle or switches to the parallel exchanger. These feed-line constraints are part of the same low-temperature control problem as the reactor itself: the entire circuit from feed drying to reactor injection must remain above the chloromethane freezing point while still delivering a chilled, single-phase feed.

Catalyst is introduced as a dilute solution of aluminum chloride in chloromethane, typically below 1 wt% active metal halide, to allow precise flow control and to distribute the initiation sites evenly. The catalyst solution line is cooled to the reactor temperature before entering through a dip pipe or sparger. A small amount of water is often deliberately added to the catalyst stream to adjust the cocatalyst ratio. The ratio of water to aluminum chloride influences both initiation efficiency and molecular weight distribution. If the water-to-aluminum ratio is too high, the catalyst complex precipitates as an inactive hydroxide and the reactor fills with unreacted monomer. If the ratio is too low, the initiation rate is low and the catalyst may be consumed by side reactions. The narrow control band for the cocatalyst ratio is a second low-temperature reaction-control problem: at −95°C, the solubility of water in chloromethane is so low that local water droplets can persist and create zones of over-initiation. Some producers pre-dissolve water in a polar modifier or use an alkylaluminum chloride instead of aluminum chloride to improve solubility and control. Published data for specific cocatalyst formulations are limited, but the operational boundary is clear: the moisture and aluminum chloride feeds must be matched to the heat-removal capacity of the reactor, because an initiation burst also produces a thermal burst.

When Isoprene Feed Ratio Deviates Below 1.0 mol%

If the isoprene concentration in the mixed olefin feed falls below 1.0 mol%, the copolymer still forms, but the average unsaturation in the final butyl rubber may approach the lower end of the commercial specification, near 0.8 mol%. This reduction changes vulcanization kinetics because sulfur crosslinking in butyl rubber proceeds through the residual double bonds. A grade with unsaturation below 1.0 mol% will require longer cure times under a standard sulfur-accelerator system and may not achieve the same modulus or compression set resistance. Conversely, if the isoprene concentration rises above 3.0 mol%, the molecular weight tends to fall because isoprene acts as a chain-transfer agent relative to isobutylene under the same conditions; the final polymer may also show increased branching and gel formation during finishing and drying. At temperatures below −95°C, the copolymerization may not be strictly random, and the measured unsaturation by ASTM D5902 or iodine titration can deviate from the feed ratio because of compositional drift. Producers therefore maintain the isoprene-to-isobutylene ratio through inline gas chromatography or Raman spectroscopy, with a control deadband of ±0.05 mol% around the grade-specific setpoint. The low-temperature constraint intersects with comonomer control: compensating for low isoprene by increasing catalyst addition can increase local heat release and push the reactor toward the freezing margin, while compensating for high isoprene by reducing temperature may increase molecular weight beyond the extruder torque limit downstream.

The cure behavior of butyl rubber is not only a function of total unsaturation; the distribution of isoprene units along the chain also matters. At low temperature, the copolymerization is largely random but small sequence-length variations can occur. Isoprene enchained in the butyl backbone is predominantly 1,4-trans, which provides the double bonds required for sulfur cure and for bromination in halobutyl grades. If the polymerization temperature is allowed to fluctuate by more than 2°C, the composition drift can be amplified because the reactivity ratios of isobutylene and isoprene are temperature-dependent. The result may be a batch that meets Mooney viscosity but fails the unsaturation specification. In a continuous reactor, such a drift may persist for several residence times, requiring the diversion of off-specification rubber and the loss of production capacity. This is why comonomer ratio, monomer conversion, and slurry density are continuously monitored and integrated into the temperature control strategy.

Across continuous slurry reactors, molecular weight is not controlled by temperature alone; the ratio of chain propagation to chain transfer depends on catalyst concentration, monomer concentration, and the concentration of adventitious proton sources. In commercial butyl plants, Mooney viscosity ML 1+8 at 125°C is the principal release parameter and is typically maintained between 45 MU and 60 MU for general-purpose butyl grades, while specialty grades may fall outside that range. The corresponding weight-average molecular weight is generally in the region of 300,000 g/mol to 600,000 g/mol, although gel-free measurement by size-exclusion chromatography is difficult because of high molar mass and long-chain branching. The relationship between temperature and Mooney viscosity is steep: a shift of 2°C can alter torque by several Mooney units depending on the grade. Therefore, even a well-tuned reactor cannot compensate for poor temperature control by adjusting residence time alone, because residence-time changes also affect isoprene conversion and slurry concentration. A high-conversion grade may reach 85% to 95% monomer conversion, but pushing beyond that envelope risks increasing the slurry solids to a level where agitator power draw exceeds design limits and the heat-transfer coefficient collapses.

Controlling Bulk Slurry Viscosity and Agitator Power Draw in Methyl Chloride

The continuous stirred-tank reactor is not a simple kettle; it is a baffled vessel with a top-entering agitator, internal cooling coils, and an overhead slurry take-off to the finishing line. At 20 wt% to 30 wt% solids, the slurry exhibits shear-thinning behavior because the precipitated copolymer particles are irregular and partly swollen with chloromethane. Agitator power draw scales with slurry apparent viscosity and impeller pitch. If the particle-size distribution shifts toward fines, the apparent viscosity increases at equivalent solids, and agitator amperage rises. A rise of more than 10% in agitator power from baseline usually indicates either an increase in solids, a change in particle morphology, or localized solidification of chloromethane on internal surfaces. The control room alarm is set at 90% of the motor nameplate torque, with an automatic catalyst cutback at 95%. Continuous removal of the slurry through a letdown valve connected to a degassing extruder or hot-water stripping system maintains a constant reactor level. The letdown line is jacketed and tempered with liquid chloromethane at −80°C to avoid plugging; if the line temperature drops below −92°C, polymer particles can agglomerate at the valve seat and cause a pressure excursion. Some designs use a positive-displacement rotary pump with hardened clearances to transfer the slurry to the finishing section. Published data for specific vendor configurations are limited, but the operational boundary is defined by the freezing point of the diluent and the softening point of the polymer.

Agitator tip speed is usually set between 2.0 m/s and 4.0 m/s for slurry suspension, although the exact value depends on impeller diameter and blade angle. Axial-flow impellers are preferred because they provide top-to-bottom circulation and re-suspend settled solids. Radial-flow impellers may create high shear near the wall and can fracture polymer particles, increasing the fines population and raising slurry viscosity. The blades and shaft are constructed of alloys resistant to chloride stress-corrosion cracking, and the mechanical seal is purged with dry nitrogen or chloromethane to prevent moisture ingress. A loss of agitation is a critical event: within seconds, polymer particles begin to settle, and within minutes the bottom cooling surface can become insulated by a compacted polymer layer. When agitation is restored after such an event, the power draw may spike above the motor limit if the settled layer has consolidated. For this reason, agitator start-up is interlocked with a minimum reactor temperature and a maximum slurry density. The polymerization reactor may also be equipped with a variable-frequency drive so that agitator speed can be reduced during grade transitions without losing suspension. These mechanical boundaries, rather than the polymerization chemistry alone, define the maximum solids concentration and the maximum heat-removal rate.

Start-up of a low-temperature butyl reactor requires pre-cooling the chloromethane inventory without exceeding the freezing plateau. The vessel is first dried with warm nitrogen, then charged with chloromethane vapor, and finally cooled slowly while agitation is maintained. If liquid chloromethane is introduced too quickly into a cold vessel, the metal surfaces may be below −97.7°C and the first liquid to contact them can freeze. The cool-down procedure therefore brings the vessel wall to at least −90°C before liquid chloromethane is admitted. During shutdown, the monomer feed is stopped first, followed by the catalyst feed, and the reactor contents are transferred to the finishing line while the temperature is kept above −97.7°C. A shutdown below the freezing point would require reheating the entire vessel and could damage the agitator seals and internal coils. Published data for specific start-up procedures are limited, but the general sequence is dictated by the physical properties of chloromethane and by the need to avoid static layers in the vessel.

Grade changes introduce a different temperature control problem. Because different butyl grades have different Mooney viscosity targets and different isoprene levels, the reactor temperature setpoint may change by 1°C to 3°C during a transition. If the setpoint is lowered too quickly, the wall temperature may cross the freezing threshold before the slurry temperature has equilibrated. The accepted practice is to change the reactor temperature at a rate no greater than 0.5°C/min while holding the ethylene coolant temperature above −101°C. During the transition, the off-specification rubber is diverted to a separate holding tank or reprocessing line. The duration of the transition is typically several reactor residence times, and the low-temperature control loop must maintain stable operation while the feed composition and catalyst flow are changed. If the grade change also raises the isoprene content, the molecular weight may fall and the temperature may need to be reduced, but this reduction must not violate the freezing margin. The operator therefore works within a narrow three-variable matrix: temperature, isoprene feed ratio, and catalyst flow.

Chloromethane recovery from the slurry and from the finishing extruder is another temperature-sensitive operation. The unreacted monomer and chloromethane flashed from the slurry are compressed and distilled, but the column reboiler temperature must stay below the decomposition threshold for any residual initiator or inhibitor. The recycled chloromethane is dried and returned to the feed blender. If recycle chloromethane contains residual water or acidic species, it can shift the initiator balance in the reactor. Therefore, the recycle stream is monitored by online pH and moisture analyzers, and a portion is purged to prevent the accumulation of non-condensables and low-molecular-weight hydrocarbons. The purge rate is set by gas chromatography and is adjusted so that ethane and ethylene impurities remain below 0.1 mol%. These impurities can influence the bubble-point and heat-transfer performance of the cooling loop; they also alter the vapor-phase composition in the reactor headspace. The entire plant operates as an integrated low-temperature distillation and polymerization network, where the same chloromethane inventory is used as polymerization diluent, catalyst carrier, and heat-transfer fluid.

Reactor pressure is maintained above the vapor pressure of chloromethane at the setpoint temperature, and non-condensables are vented through a pressure-control valve to the recovery compressor. The vent stream is analyzed for oxygen before entering the compressor because oxygen can react with aluminum chloride residues and form corrosive species. The pressure controller also protects the mechanical seals: a sudden drop in reactor pressure can vaporize chloromethane in the seal cavity and cause dry running. The overhead condenser is designed for partial condensation so that non-condensables are rejected while chloromethane and unreacted monomers are returned. If the overhead system is over-cooled, chloromethane can condense and return as a subcooled liquid that locally lowers the reactor temperature. The condenser coolant is therefore controlled to maintain a return liquid temperature at least 2 K above the chloromethane freezing point. Published data for specific condenser designs are limited, but the pressure boundary is fixed by the need to avoid boiling in the reactor and freezing in the return line.

For halobutyl derivatives, the low-temperature reactor product is transferred to a separate halogenation step, but the molecular weight, unsaturation, and gel content fixed in the polymerization reactor determine the outcome of bromination or chlorination. A temperature excursion during polymerization that produces low-molecular-weight tail components can create sticky fines that accumulate in the extruder vent and degrade during subsequent halogenation. The extrusion finishing line typically includes a twin-screw extruder with a length-to-diameter ratio of 40:1 to 50:1, devolatilization ports, and a hot-face cutter; water injection and vacuum stripping reduce residual chloromethane to below 0.1 wt% in the finished bale. In this equipment, the Mooney viscosity of the feed is the dominant variable controlling screw torque, melt temperature, and die pressure. If the reactor temperature is allowed to drift upward by 3°C, the resulting lower-Mooney rubber may process with lower head pressure but may fail the compression set requirement after curing when tested according to ISO 815-1. Conversely, a 3°C downward drift raises molecular weight and may exceed the extruder torque limit, forcing a reduction in feed rate and lowering line capacity. The halogenation step is also sensitive to residual moisture and to amine-based stabilizers; amine-based antidegradants are avoided in halobutyl finishing because they can cause premature crosslinking and discoloration. For pharmaceutical stopper grades, the finished compound is subject to USP <381> extraction testing and 21 CFR 177.2600 compliance; these tests are sensitive to low-molecular-weight extractables generated by a high-temperature excursion in the polymerization reactor. For these reasons, the low-temperature reaction control loop is directly linked to the finishing extruder load control and to the downstream halogenation line, not as a cascaded setpoint but as a constraint gate that prevents the introduction of off-specification molecular weight into the melt train.