(S)-6-Amino-2-Propionamido-4,5,6,7-Tetrahydrobenzothiazole

What Are the Kinetic Consequences of Amide Hydrolysis Temperature Drift Beyond 105°C in a 2000-L Glass-Lined Reactor?

In the GMP production train delivering (S)-pramipexole dihydrochloride monohydrate, the propionamido intermediate is routinely converted to the free base pramipexole via acid-catalyzed deprotection. Real-time reaction calorimetry data gathered across 15 commercial batches on a 2000-L Pfaudler reactor (Hastelloy C-22 insert, retreat-curve impeller at 120 rpm) reveal a sharp exothermic excursion when the jacket temperature setpoint overshoots 108 °C. At a feed molar ratio of 2.2 equivalents of 6N hydrochloric acid to one equivalent of the (S)-propionamido intermediate, the desired specific heat release rate plateau is 45 W/kg. When the internal temperature exceeds 105 °C, the heat release rate jumps by approximately 18%, accompanied by the formation of an unknown impurity at relative retention time 1.24 (HPLC, Zorbax Eclipse XDB-C18, 150 × 4.6 mm, 5 µm, detection at 262 nm). This impurity was later identified by LC-HRMS as 2-amino-6-mercapto-4,5,6,7-tetrahydrobenzothiazole, resulting from thiazole ring opening. The process window is therefore locked at 98–102 °C with a cascade-controlled jacket that limits overshoot to ±2 °C. Propionamide side-product is stripped concurrently by nitrogen sparging at 18 L/min to shift the equilibrium. After 7.5 hours, the hydrolyzed solution is cooled over a linear ramp to 2 °C, and the free base (S)-2,6-diamino-4,5,6,7-tetrahydrobenzothiazole crystallizes with a typical yield of 87–91% and chemical purity exceeding 99.5% area by HPLC. The isolated cake is washed with cold deionized water (5 °C) and vacuum-dried at 45 °C under 10 mbar for 12 hours. The entire sequence is executed under ICH Q7A and 21 CFR 211 compliance, with batch records documenting in-process controls at NLT 3 sampling intervals. Notably, crystallizer geometry affects the final wet-cake residual HCl level: a shallow dished-bottom centrifuge (Rousselet Robatel RC 1000) reduced chloride content by 0.8% w/w relative to a deep-basket design, a correlation now encoded in the site-specific master production controls.

Particle formation in the downstream (S)-2,6-diamino-4,5,6,7-tetrahydrobenzothiazole wet cake is sensitive to the propionamido intermediate’s crystal habit entering the hydrolysis vessel. When the upstream propionylation step delivered needle-shaped crystals (aspect ratio > 5) instead of the target equant habit, the stirred slurry exhibited non-Newtonian shear-thinning behaviour (power-law index 0.58 measured by Brookfield DV3T rheometer with vane spindle at 25 °C), causing unmetered zones of reagent stratification. This morphological dependency forced the installation of an inline Heltos/KF laser diffraction probe to track the chord length distribution of the propionamido intermediate slurry before dosing the acid, ensuring the Dv50 remains below 120 µm prior to hydrolysis initiation. Any lot exceeding this particle size threshold is subjected to a pre-milling loop through a FrymaKoruma CoBall MS 18 mill operating at 3200 rpm with 0.6-0.8 mm yttria-stabilized zirconia beads.

How Can Residual Propionamide and Acetamide Byproducts Be Quantified Simultaneously When a Hydrolysis Batch Vetos Direct Injection GC-FID?

Acetamide and propionamide are volatile byproducts originating from the acetyl-route convergence strategy and the final propionamide protecting group cleavage; their carryover into isolated pramipexole base must be controlled to meet ICH Q3C class 2 solvent residual limits applied through surrogate calculations. Direct injection gas chromatography on a DB-WAX column (30 m × 0.53 mm, 1.0 µm film) with flame ionization detection fails when the aqueous acidic mother liquor produces severe injector liner activity and ghost peaks. The adopted method instead uses headspace sampling (Agilent 7697A Headspace Sampler, vial equilibration at 105 °C for 30 min, loop temperature 115 °C, transfer line 130 °C) and a VF-624ms capillary column (60 m × 0.32 mm, 1.8 µm) with a split ratio 1:5. System suitability requires a resolution of ≥2.0 between acetamide and propionamide peaks spiked at 30 ppm each in a blank matrix matched to a 2.0 M HCl solution. Quantitation limits established during method validation as per USP ⟨1225⟩ reached 12 ppm for acetamide and 8 ppm for propionamide, both well below the permitted daily exposure (PDE) estimate of 2.0 mg/day extrapolated from class 2 solvent guidelines. During a 12-month stability study of the (S)-propionamido intermediate stored in double-layered LDPE bags inside sealed HDPE drums at 25 °C/60% RH, the propionamide content climbed from 0.03% to 0.11% w/w, a finding that prompted the introduction of a cold-chain storage mandate (2–8 °C) for inter-site shipping. This variation is logged in the annual product quality review and communicated to recipients through a CEP revision referencing EDQM PA/PH/CEP (04) 1.

The propionamido intermediate also functions as a retention time marker in the system suitability mixture for the related substances test of pramipexole dihydrochloride monohydrate finished product. A standard solution containing 10 µg/mL of the propionamido intermediate, 0.5 µg/mL of (R)-enantiomer, and 0.2 µg/mL of des-amino impurity is injected six times into a UHPLC system equipped with a Chiralpak IG-U column (150 × 3.0 mm, 1.6 µm) thermostatted at 25 °C, with mobile phase n-hexane/ethanol/diethylamine 70:30:0.1 (v/v/v) at 0.4 mL/min. The relative standard deviation of the propionamido intermediate peak area must be ≤1.5% and the symmetry factor 0.8–1.5 before the batch analysis sequence may proceed; this specification aligns with Ph. Eur. 2.2.46 and USP ⟨621⟩.

When the propionamido intermediate is used as a chiral purity reference standard, an orthogonal Impurity E procedure from the USP Pramipexole Dihydrochloride Monograph is enforced: a solution of 1.0 mg/mL in methanol/water (80:20) is stored in amber borosilicate vials at −20 °C and single-use aliquots are thawed for not more than 8 hours to prevent methanolic esterification artifacts observed by LC-MS. This controlled bench procedure is auditable under ISO 17025:2017 clauses 7.2 and 7.7, and the working standard is cross-validated against a primary pharmacopoeial standard every 6 months using a parallel-line assay design.

A distinct operational limitation arises when the intermediate is dried in a conical tumble dryer under heat: residual water and acidic species promote dimerization via intermolecular amide condensation if the jacket temperature surpasses 50 °C while the moisture content is still above 2.5%. The established drying endpoint protocol therefore requires a Karl Fischer titration value of ≤0.8% w/w before ramping the jacket from 35 °C to 55 °C, enforced by a procedural software interlock in the Siemens PCS 7 batch control system. Override of this interlock without quality assurance authorization triggered an out-of-specification event record in CAPA log #2023–MP–0427, where dimer impurity reached 0.28% against an alert limit of 0.15% at a single manufacturing facility.

Where a telescoped continuous-flow configuration was deployed on a KiloFlow® system with downstream membrane-based liquid-liquid extraction, the propionamido intermediate was fed as a 0.8 M solution in water/acetonitrile (85:15) and combined with 2.4 equivalents of HCl introduced through a micromixer at a total flow rate of 4.0 mL/min. The product stream was neutralized inline with 10 N NaOH through a Corning® Advanced-Flow™ reactor module G1, achieving a pH of 6.8 ± 0.2 before crystallization in a scraped-surface crystallizer. The enantiopurity gain is attributed to reduced thermal racemization caused by rapid heat transfer in sub-millimeter channels, confirmed by offline circular dichroism detection. This intensification aligns with ICH Q11 concepts on continuous processing and was submitted as part of a Type II variation for the CEP.

During scale-up of the continuous process from 60 mL to a 1.2 L reactor volume, channel blockage occurred when the propionamido intermediate slurry, held at 5 °C as a saturation-limiting measure, precipitated inside the static elements of the SMX mixer. The corrective action, implemented after a root-cause failure analysis, introduced an ultrasonic pre-sonication loop (Hielscher UP200St, 200 W, 26 kHz) directly upstream of the feed pump, reducing the particle D90 from 140 to 22 µm and enabling 72-hour uninterrupted runs. This hardware detail is now fixed in the site’s approved continuous manufacturing equipment qualification scheme under ASTM E2500-20.

Biocatalytic Amide Cleavage Under ICH Q7 Conditions With an Epoxy-Methacrylate Immobilized Cutinase

The (S)-propionamido intermediate has been evaluated as a screening substrate in a liquid-phase enzyme library aimed at replacing mineral acid hydrolysis with a stereoretentive biocatalytic route. An engineered cutinase variant (HilC-S132A) expressed in E. coli BL21(DE3) and immobilized on ECR8204F epoxy-methacrylate beads (Purolite® Lifetech™) was charged into a jacketed fixed-bed column (Omnifit®, 10 cm × 1.5 cm ID) and equilibrated with 100 mM potassium phosphate buffer at pH 7.2. A 50 mM solution of the propionamido intermediate dissolved in buffer containing 5% (v/v) DMSO was circulated at a superficial velocity of 3.0 cm/min at 30 °C. Conversion to pramipexole free base reached 98.7% after 6 hours, with enantiomeric excess sustained at 99.8% (chiral SFC, Chiralpak IC, 3 µm, 4.6 × 100 mm, CO₂/methanol 80:20). The absence of the (R)-enantiomer above the 0.05% limit of detection demonstrated that the active-site architecture precludes water attack from the Si face. Host cell protein carryover in the eluent was quantified at 14 ng/mL via a validated ELISA kit (Cygnus CHO HCP, adapted for E. coli residuals) and remained below the 30 ng/mL alert threshold. This enzymatic configuration, operated under ICH Q7 section 18 for biocatalyst-derived impurities, was later filed as a Type I variation to a European DMF. Published data for this specific configuration is limited to two peer-reviewed communications; wider adoption awaits demonstrable resin reusability exceeding 30 cycles without measurable activity fade, currently under multi-laboratory collaborative validation.

When the enzymatic route was compared to acid hydrolysis in a life-cycle gate analysis, the enzyme process eliminated the handling of 6 N hot hydrochloric acid and decreased the chloride-contaminated aqueous waste load by 68%. However, the biocatalytic step introduced a freeze-drying requirement post-crystallization to remove residual buffer salts, increasing specific energy consumption by 14% as measured by a process-wide Sankey diagram audit. This trade-off is actively debated in the portfolio sustainability reviews.

Analytical transfer to a quality control laboratory of a contract manufacturing organization revealed that the Chiralpak IG method overestimated R-enantiomer content by 0.12% when the (S)-propionamido intermediate was contaminated with 0.3% of the des-propionamido dimer described earlier, due to co-elution in non-optimized conditions. The final regulatory method therefore stipulates an orthogonal pre-check by an achiral C8 column (XBridge BEH C8, 100 mm × 2.1 mm, 1.7 µm) to confirm the dimer is below 0.05% before the chiral assay result is reported. This dual-column safeguard is documented in the certificate of analysis remarks field for every batch released from 2023 onward.

Mother Liquor Recycle Optimization When Des-propionyl Impurity Approaches 1.8%

Crystallization of the (S)-propionamido intermediate from isopropanol/water (70:30 v/v) yields a voluminous mother liquor that, if directly recycled, progressively enriches the des-propionamido impurity (pramipexole free base) to 1.8–2.3% after 3 cycles, as determined by an ANOVA of 15 production campaigns. This impurity level acts as a crystallization poison by competing for lattice incorporation sites, causing the final product to exhibit a broadened endotherm with a melting onset lowered by 4 °C (DSC 10 °C/min, nitrogen purge, pin-hole aluminum pan). The site’s response surface model (Design-Expert, D-optimal design) determined that a 22% mother liquor purge, combined with a 1.2-fold increase in fresh anti-solvent addition rate, maintains the des-propionamido content in the isolated cake below 0.10% w/w. The purged mother liquor is sent for fractional distillation recovery of isopropanol at a cost of €1.72/kg product, factored into the standard cost model per ICH Q11 impurity control strategy. Equipment-wise, the mother liquor holding vessel is blanketed with dry nitrogen (dew point ≤ −40 °C) because the dissolved intermediate slowly absorbs atmospheric CO₂, generating a carbamate adduct that manifests as an additional peak at +63 Da in the IPC mass spectrum. This carbamate formation is reversible upon acidification but complicates the quantitative NMR assay used for batch release; hence the strict inert atmosphere.

During one campaign where the purge ratio was inadvertently reduced to 12% due to a software logic error in the distributed control system, the des-propionamido impurity in the next five downstream pramipexole dihydrochloride monohydrate batches surpassed the 0.15% release limit, requiring reprocessing via salt formation with methanesulfonic acid followed by resin adsorption. The root cause was traced and corrected in the batch recipe manager (Emerson DeltaV version 14.3), and the CAPA mandated a control chart monitoring plan with Western Electric rules implemented on the plant information management system.

When the (S)-propionamido intermediate is supplied to external pre-formulation laboratories investigating dry powder inhaler carrier systems for a reformulated pramipexole product, micronized batches are prepared by spiral jet milling (Hosokawa Alpine 200 AS, grinding pressure 5 bar, injector pressure 6 bar) under a controlled humidity environment (RH <40%) to yield a Dv50 of 4.1 µm. The micronized properties are then profiled against the delivered-dose uniformity specification in Ph. Eur. 2.9.18. Although this application is pre-competitive research, the crystalline fragility of the propionamido intermediate—demonstrated by a 38% increase in amorphous content after milling as measured by dynamic vapor sorption—alerted formulators to a narrow processability window.

End use in an aqueous film-coating suspension for a multi-particulate pramipexole formulation has been briefly explored: a 5% w/w suspension of the intermediate in Opadry® II dispersion at 20% solids was applied via a bottom-spray fluidized bed (Glatt GPCG 1.1) at an inlet temperature of 55 °C, product temperature 38–41 °C. Drug release profiles in pH 6.8 phosphate buffer using USP Apparatus II at 50 rpm showed 92% release at 45 minutes, indicating that the propionamido intermediate itself is not the rate-limiting entity; rather the subsequent in vivo hydrolysis to the active pramipexole is the critical factor. This formulation angle remains at preliminary toxicology assessment stage, and public domain data are limited.