Methyl Methacrylate Amidation Esterification Sequence Control with Acetone Cyanohydrin

What Determines the Optimal Acid‑to‑Cyanohydrin Ratio in Amidation?

In the acetone cyanohydrin (ACH) route to methyl methacrylate (MMA), the first critical sequence control point is the amidation of ACH with concentrated sulfuric acid. The reaction produces methacrylamide sulfate (MAAS) and ammonium bisulfate, with the stoichiometry demanding 1.0 mol H₂SO₄ per mol ACH to form the amide salt; however, industrial practice consistently operates at 1.2 to 2.5 mol H₂SO₄ per mol ACH. The excess acid serves three functions: it suppresses premature polymerization of methacrylamide intermediates, it controls the viscosity of the highly exothermic reaction mass, and it shifts the equilibrium of side reactions that regenerate acetone. At ratios below 1.15, yield loss to acetone and diacetone alcohol exceeds 8% of ACH feed, while ratios above 2.5 generate an ammonium bisulfate slurry that is difficult to pump and increases downstream neutralization costs. The operating window for acid concentration itself is equally narrow: 98–100 wt% H₂SO₄ is necessary to keep water below the threshold where hydrolysis of ACH to α‑hydroxyisobutyramide becomes competitive. Oleum (up to 20% free SO₃) is occasionally used to scavenge residual water but introduces a sulfonation pathway that leads to methyl methacrylate sulfonate impurities detectable by HPLC at ≥5 ppm. Temperature control in the amidation reactor is a key determinant of product colour and distillability. A cascade of two to four continuous stirred‑tank reactors (CSTRs) is standard, with the first vessel held at 80–95°C and subsequent stages allowed to rise to 110–130°C. Residence time per reactor typically falls in the range of 20–45 min. Published heat‑flow calorimetry data (RC1e, Mettler‑Toledo) from process safety evaluations indicate that the adiabatic temperature rise for the ACH‑sulfuric acid system can exceed 150°C if cooling is lost, necessitating a pressure‑relief system sized for two‑phase runaway per DIERS methodology (following ISO 4126‑10:2021). Agitator design shifts from pitched‑blade turbines in the first stage to anchor or helical‑ribbon impellers in later stages where the slurry density reaches 1.6–1.8 g/cm³ and apparent viscosity rises above 5,000 cP. The amidation sequence is completed once the free sulfuric acid concentration in the reaction mixture drops to 15–25 wt%, at which point the MAAS is routed to esterification without intermediate isolation.

Industrial production facilities that process 100,000 tonnes/year of MMA frequently report batch‑to‑batch variability in amidation yield within a ±1.5% band when the ACH feed contains more than 0.2 wt% water. Pre‑drying of ACH by azeotropic distillation with n‑heptane or by molecular sieve dehydration (3A, ≤0.05 wt% water achievable) is standard upstream. Failure to control the water content converts the amidation into an amide‑hydrolysis‑dominated sequence, where methacrylamide is cleaved to methacrylic acid and ammonium ion. This not only reduces MMA output but acidifies the subsequent esterification mixture, requiring additional sodium methylate neutralisation that raises the ash content of the final monomer above the 10 ppm limit required for optical‑grade polymerisation (ASTM D788‑24, Group 2). The amidation control logic also addresses the competition between N‑ and O‑sulfonation of the intermediate species: the desired N‑sulfonated methacrylamide is favoured at temperatures below 100°C and acid strengths above 95 wt%, whereas O‑sulfonation to methacrylic anhydride becomes kinetically significant at temperatures above 115°C and leads to a product stream with a titrated saponification number deviation of +4 mg KOH/g that interferes with precise copolymer composition targeting in downstream acrylic sheet casting.

Injection molding grades of MMA monomer intended for high‑transparency polymethyl methacrylate (PMMA) require that the dimer and oligomer content arising from premature amidation‑stage polymerisation be held below 200 ppm. This is enforced by inline UV‑Vis monitoring at 254 nm on the MAAS slurry transfer line. When absorbance exceeds a set point equivalent to 0.15 AU, a proprietary nitroxyl‑based inhibitor (typically 50–200 ppm of 4‑hydroxy‑TEMPO) is added to the first CSTR. The cost of such intervention is justified by the specification penalty for PMMA yellowing index shift beyond ΔYI +0.8 under ASTM D1925.

When Esterification Precedes Amidation: Anomalous Sequences in By‑product Formation

The canonical ACH‑to‑MMA sequence is thermally and chemically unidirectional under standard conditions, yet deviations in feedstock purity or heat‑integration schemes can inadvertently create a local environment where esterification precedes complete amidation. This occurs most commonly when methanol is introduced into the amidation‑zone flash vapours through poorly designed condensate return loops. Methanol, present at concentrations as low as 2 vol% in the vapour phase, reacts with free SO₃ or unreacted ACH to produce methyl sulfate esters and methyl methacrylate directly, bypassing the methacrylamide intermediate. While this might appear beneficial, the uncontrolled early esterification consumes methanol that is stoichiometrically required in the subsequent esterification reactor and generates dimethyl sulfate, a potent methylating agent that is subject to strict workplace exposure limits under OSHA 1910.1020 and REACH Annex XVII entry 28. Analytical data from process gas chromatography (ASTM D4773‑22) show that the resulting crude MMA can contain 50–300 ppm dimethyl sulfate, requiring an alkaline hydrolysis scrubber downstream that reduces overall yield by 0.3–0.7% of theoretical MMA.

The reversion of ACH to acetone and hydrogen cyanide under the acidic conditions of amidation is a competing reaction that cannot be fully suppressed but is kept below 4% of ACH feed by rapid injection of ACH into the acid medium at −5 to +5°C sub‑cooling. The sequence reversal risk becomes acute during turn‑down operation below 60% of nameplate capacity, when heat‑integration exchangers suffer from reduced velocity and localised hot spots can reach 140–150°C. In such hotspots, the esterification of methacrylic acid (formed by hydrolysis) competes with the intended amidation, producing heavy esters that foul reboiler tubes within 48–72 hours. The deposition layer, analysed by SEM‑EDX on pulled tubes, consists predominantly of polymethacrylate oligomers with a manganese content of 120–400 ppm—manganese originating from stainless steel corrosion at high‑temperature acid service. Cleaning cycles using 10% nitric acid at 60°C (per NACE SP0292) restore heat transfer coefficients to within 85% of design but impose a 3‑day outage per annual campaign. Process licensors that embed esterification preceding amidation—even inadvertently—thus face a mandatory requirement for continuous online TOC monitoring of the condensate return with a shutdown interlock set at 25 ppm TOC, as specified in IEC 61511 safety instrumented system guidelines.

The amidation‑first, esterification‑second sequence ensures that any residual methacrylamide sulfate is converted to MMA in the final vessel, whereas a reversed sequence leaves unreactive ammonium sulfate‑rich solids that blind the distillation column trays. To enforce the correct sequence, the primary control loop is the interlock between the ACH feed pump and the methanol feed pump: methanol introduction is permitted only when the MAAS slurry temperature is ≥90°C and the free H₂SO₄ concentration, measured by online conductivity, has fallen below 30 wt%. This interlock, validated by a SIL‑2 safety function, prevents the onset of competitive esterification during the induction period of amidation and is common across multiple ACH‑based MMA process technologies currently in operation.

The esterification stage itself operates at 100–125°C and 0.2–0.5 MPa(g) in a reactive distillation column. A molar ratio of methanol to methacrylamide moiety of 1.5–2.2 is maintained; ratios below 1.3 reduce single‑pass conversion to <75%, while ratios above 2.5 increase the methanol recovery energy demand beyond 3.2 GJ/tonne MMA and elevate the methyl formate by‑product from formic acid contamination of the methanol recycle. Published data for this specific configuration is limited regarding proprietary catalyst packages, but un‑catalysed systems at 120°C achieve an equilibrium conversion of 88–92% after 4–6 h residence time. Heterogeneous acid catalysts such as sulfonated polystyrene‑divinylbenzene resins (typically 4–8% crosslinking) have been evaluated in pilot‑scale tests at 100 kg/day throughput, where they reduced free sulfuric acid carryover to <50 ppm in the crude MMA but suffered deactivation from ammonium bisulfate deposition within 200 hours on stream, requiring regeneration with dilute ammonia.

Sulfuric Acid Concentration Windows and By‑product Suppression

Maintaining a precise sulfuric acid concentration profile across the two‑step sequence is the most operationally demanding aspect of the ACH process. In the amidation reaction, the H₂SO₄ strength must remain above 96 wt% to avoid hydrolysis of ACH to α‑hydroxyisobutyramide, a compound that decomposes to acetone and carbon monoxide at temperatures above 90°C, creating pressure hazards and gas‑phase fouling in overhead condensers. The water that enters the system—via ACH (typically 0.1–0.3 wt% moisture), methanol (<0.1 wt% water specification), and spent sulfuric acid recycle—is the primary variable that shifts the concentration window. A mass balance over the amidation CSTR series shows that every 0.5% increase in total feed water reduces the effective H₂SO₄ strength in the first reactor by 1.8–2.2%, depending on the oleum make‑up rate. Once the acid strength drops below 90 wt%, the rate of methacrylamide sulfate precipitation slows and the slurry viscosity drops below 500 cP, yet the N‑selectivity of sulfonation falls so sharply that the yield of distillable MMA after esterification declines by 10–15% absolute. Plant data from a 50 kt/a line published in a 2017 process optimisation study revealed that restoring acid strength from 89 wt% to 97 wt% through oleum injection increased the overall MMA yield from 82% to 93% of theoretical, with a payback period of under 4 months for the additional oleum consumption of 12 kg/t MMA.

The esterification step, by contrast, benefits from a controlled amount of water—typically 5–15 wt% of the total liquid mass in the reactive distillation column—because the hydrolysis of MAAS to free methacrylamide and ammonium bisulfate is necessary before esterification can proceed. Too little water (<3 wt%) stalls the hydrolysis and accumulation of solid MAAS in the column sump raises the differential pressure beyond the design limit of 20 kPa; too much water (>18 wt%) reverses the esterification equilibrium and favours methacrylic acid formation, reducing the once‑through MMA yield and increasing the downstream acid‑removal load on the anion‑exchange resin beds (typically Lewatit MP 62 WS) to a breakthrough capacity of 0.6 eq/L‑resin. The column water balance is managed by a side‑draw decanter that separates an aqueous phase containing ammonium bisulfate (35–45 wt% concentration) from the organic phase; this aqueous phase is sent to an ammonium sulfate crystalliser, where the ammonium bisulfate is neutralised with ammonia to yield agricultural‑grade ammonium sulfate crystals conforming to EU Fertiliser Regulation 2019/1009. The decanter temperature is held at 55–65°C to prevent methacrylamide crystallisation, which occurs at <52°C for the typical concentration profile observed in the recycle loop.

Sulfuric acid regeneration from the ammonium bisulfate stream via thermal decomposition at 900–1,100°C in a spent acid recovery furnace is practiced at integrated production sites and can recover up to 88% of the initial H₂SO₄ inventory. The furnace off‑gas, containing 9–12 vol% SO₂, is processed through a contact process plant (V₂O₅ catalyst, 420–600°C) to produce fresh 98.5 wt% acid. However, the energy intensity of this regeneration—approximately 4.5–5.0 GJ/tonne of recovered H₂SO₄—makes it economically viable only when the integrated site has access to low‑cost fuel and can utilise the co‑produced high‑pressure steam. For stand‑alone MMA plants, the ammonium sulfate is sold as a co‑product, and fresh sulfuric acid is procured to maintain the required window. This commercial boundary condition directly influences the acid‑to‑cyanohydrin setpoint choice: sites with spent acid regeneration tend to operate at the higher end of the 1.8–2.5 ratio to maximise recovery efficiency per tonne of MMA, while merchant‑acid dependent plants cluster at 1.2–1.6 to minimise fresh acid consumption, accepting a tighter amidation control tolerance.

The temperature‑concentration coupled control problem in the amidation CSTR train is often addressed by model predictive control (MPC) that manipulates oleum flow, ACH feed temperature, and jacket cooling water flow simultaneously. A benchmark MPC implementation on a 90 kt/a plant, described in publicly available control literature, maintained the first‑stage H₂SO₄ strength within ±0.5 wt% of the 97.5% target and reduced the standard deviation of amidation yield from 1.8% to 0.6%, achieving a product colour improvement from 15 APHA to 8 APHA. The controller constraints included a maximum cooling water outlet temperature of 45°C to prevent silica scaling on the exchanger surface and a minimum agitator motor current that served as a soft sensor for slurry viscosity—below 45% of full load amperes, the slurry was deemed too dilute in solids for efficient subsequent esterification.

When Methanol‑to‑Amide Molar Ratios Shift from Optimal Values

The reactive distillation column that carries out the esterification step is highly sensitive to the methanol‑to‑methacrylamide moiety ratio. At ratios below 1.3, the vapour‑phase methanol concentration in the column drops sufficiently that the temperature on the uppermost trays can exceed 90°C, triggering thermal polymerisation of MMA in the partial condenser. The polymerisation deposits polymethyl methacrylate on the condenser tubes, reducing the overall heat transfer coefficient from a design value of 800–1,000 W/m²·K to below 300 W/m²·K within 72 hours of continuous operation. To prevent this, an inhibitor injection system distributes a solution of MEHQ and air (as an oxygen co‑catalyst) to each tray above the feed point, with the MEHQ concentration in the reflux maintained at 30–60 ppm. The air injection rate—typically 0.1–0.3 Nm³/h per tray—is regulated to keep the dissolved oxygen in the reflux at 5–8 ppm, measured by an in‑line optical DO probe. Below 3 ppm DO, inhibition becomes ineffective even at high MEHQ loads, a finding attributed to the regeneration cycle of the phenoxyl radical requiring molecular oxygen (as per the classical inhibition mechanism documented in polymer science literature). When the ratio rises above 2.5, the excess methanol breaks the azeotrope with MMA (64.5°C at 101.3 kPa, 80.5 wt% methanol) and carries MMA into the overhead distillate, reducing the bottom product purity to below 98 wt% and increasing the recycle load on the methanol recovery column, which must then separate a methanol‑MMA‑water mixture that forms three binary azeotropes. The increased reboiler duty can exceed the utility margin by 15–20%, forcing a reduction in throughput.

The methanol feedstock specification is itself a determinant of sequence fidelity. Methanol containing more than 50 ppm ethanol undergoes transesterification to ethyl methacrylate, a monomer that copolymerises with MMA at a different reactivity ratio (r₁ = 0.68, r₂ = 1.45 for MMA/EMA in bulk polymerisation at 60°C, as reported in the literature) and shifts the glass transition temperature of the resulting copolymer away from the target of 105°C for standard PMMA. To avoid this, methanol purchased for MMA esterification is certified against ASTM D1152‑24 with an ethanol concentration limit of <10 ppm and a total carbonyls limit of <20 ppm. The methanol‑to‑amide ratio is therefore not just a productivity variable but a product‑quality lever that directly influences the chain architecture of downstream polymer. An online near‑infrared analyser (NIR) scanning the 1,450–1,650 nm region provides a closed‑loop feedback signal for the methanol feed ratio controller, with a response time of <30 seconds and an accuracy of ±0.02 in molar ratio units, which is essential given the 6–8‑hour mean residence time of the esterification system.

The separation of ammonium bisulfate from the esterification effluent represents the final stage where sequence control can be compromised. If the aqueous phase recycle from the decanter still contains traces of methanol—as happens when the decanter interface level control drifts by >5%—the sulfide‑rich aqueous phase can react with methanol in the ammonium sulfate crystalliser to regenerate methyl bisulfate, a corrosive agent that, in one documented failure incident, reduced the wall thickness of a 316L stainless steel crystalliser from 6 mm to 1.9 mm in 18 months, well below the corrosion allowance of 3 mm specified in the original engineering design. The corrective action required installation of a redox‑potential‑controlled purifier column upstream of the crystalliser, held at pH 4.0–4.5 and 70°C, to hydrolyse any remaining methyl bisulfate to methanol and bisulfate.

The selection of esterification vessel materials is driven by the need to resist chloride‑induced stress corrosion cracking, which can occur in sensitised austenitic stainless steels at temperatures above 60°C in the presence of residual chlorides (often introduced via the water used for methanol dilution). Standard practice for the main esterification column shell is to specify 904L or glass‑lined carbon steel with a thickness of ≥12 mm to allow for a 0.1 mm/year uniform corrosion rate over a 20‑year design life as per NORSOK M‑001. The reboiler tubes in the methanol recovery column are routinely specified in Hastelloy C‑276 after multiple documented pitting failures of 316L tubes within 2 years of service; the pitting was traced to the combination of 2–5 ppm chlorides and 15–20 ppm ferric ions originating from upstream corrosion. The incremental capital cost of C‑276 over 316L—approximately 3.5‑fold—was recovered through a reduction in unplanned downtime from 12 days/year to 0.5 days/year and an extension of tube bundle replacement intervals from 24 months to 96 months.

In‑line acid concentration measurement in the amidation trains is performed by conductivity cells immersed in a sidestream that is continuously diluted at a precise 100:1 ratio with demineralised water using a Coriolis‑based mass‑flow blending skid that maintains dilution accuracy to ±0.2%. The diluted sample is passed through a temperature‑compensated toroidal conductivity sensor calibrated against NIST SRM 3191 for sulfuric acid. This setup avoids the rapid fouling that plagues direct insertion probes exposed to the MAAS slurry. Each CSTR is equipped with one such measurement loop, and the signals are voted 2oo3 in the safety logic solver to trigger an emergency quench with chilled 30% aqueous ammonia in the event that the free acid concentration in the final amidation reactor exceeds 35 wt%, a condition that indicates incomplete ACH conversion and a risk of runaway during subsequent heating. Such quench systems, however, have a lifetime actuation limit of 3–5 events before the neutralised salts accumulate in the bottom of the CSTR and require a manual cleanout, a fact that compels operations teams to maintain sequence integrity through preventive measures rather than reliance on safety‑related trips.

The amidation‑esterification sequence is further safeguarded by installation of rupture discs between the amidation CSTRs and the esterification column that burst at a set pressure of 0.15 MPa(g), well below the 0.3 MPa(g) design pressure of the vessels, to redirect blowdown to a dedicated quench pool in the event of a methanol backflow. The quench pool, sized at 150% of the combined liquid inventories of all connected equipment, contains 5 wt% sodium hydroxide solution maintained at <30°C by cooling coils. These safety provisions, detailed in the HAZOP reports of multiple operating plants, underscore that the sequence control is not merely a yield optimisation parameter but a fundamental process safety requirement that dominates the HSE case for the ACH technology. Published data for this specific configuration is limited in the open literature, but industry safety guidelines such as the European Acrylate Sector Group’s BREF note on MMA production identify the amidation‑esterification boundary as a “critical sequence point” requiring independent protection layers with a risk reduction factor of ≥100.