(T1030-09-60) Investigation into How Solvent Selection Affects the Metastable Zone of Paracetamol During a Thermally Controlled Recrystallisation

Purpose: Understanding the solubility profile of an API (Active Pharmaceutical Ingredient) is crucial when it comes to designing a reproducible recrystallisation procedure in which a final product which has the desired characteristics. With paracetamol, the solubility in a range of organic solvents has been published (Granberg & Rasmuson, 1999), but the temperature-solubility profile has not been thoroughly elucidated. Using a set of controlled thermal experiments, a dataset was created to understand the solubility profile and when the solution was supersaturated. This allowed us to plot the MSZW (metastable zone width) for paracetamol in various solvents and then identify an optimum solvent for use in future crystallization development studies.

Methods: The solubility profile of paracetamol in MEK (methyl ethyl ketone), IPA (Isopropyl alcohol), MeCN (acetonitrile) and 1-BuOH (1-Butanol) was investigated using a set of controlled thermal cycling experiments, using a Technobis Crystalline. For every solvent, four aliquots of paracetamol were weighed into glass vials and solvent was added to form suspensions at known concentrations. Each suspension was heated to 5°C below the boiling point of the solvent at a rate of 0.5°C minute^-1. Samples were then cooled to 5°C at a rate of 0.3°C minute^-1 and held at that temperature for 30 minutes. The temperature cycle was repeated a further two times. For the entire duration of the experiment turbidity was measured every 1 second, with an image being taken every 5 seconds. A bottom magnetic stirrer was used spinning at a rate of 800 rpm. After the controlled thermal cycling phase had been completed, samples were filtered on a frit filter and dried under suction for several hours at RT. Samples were then characterized to assess for crystal form, degradation, evidence of residual solvent, and morphology. MSZW determination was calculated using CrystalClear software from turbidity measurements.

Results: MSZW analysis showed a range of widths in the solvent systems used. The two alcohols (IPA & 1-BuOH) showed their capability of dissolving a larger amount of paracetamol when compared to MEK or MeCN. The advantage that a higher solubility brings is that a smaller volume of solvent would be required upon scale up. 4 volumes of IPA (250 mg/mL) would be sufficient to induce a solution from ~65°C. 5 volumes of 1-BuOH (200 mg/mL) would be sufficient to induce a solution from ~70°C. The cloud point measurements from the two alcohols range from 30°C to 40°C lower than the clear points. This in turn shows that the metastable zone width is large. This would allow more leeway when it comes to implementing a seeding method for optimization using these two solvent systems. The key disadvantage from the assessment of the MSZW for the two alcohols is that at 0°C both solvents were still estimated have a solution concentration of paracetamol > 100 mg/mL. This in turn would result in a lower yield. Further iterations of using the two alcohols as a solvent may need to involve an antisolvent to improve the final yield. The experiments performed in MeCN show the opposite advantages and disadvantages to the two alcohols, with a lower solubility, smaller MSZW, but with a smaller amount of compound in solution at low temperatures. The MSZW data for MEK seems to fit between the alcohols and MeCN, with the capability of forming a solution at 6.7 volumes (150 mg/mL) at ~70 °C, whilst only having 20 mg/mL in solution at 0°C. The MSZW in MEK is still at a size where there is an opportunity for seeding experiments without the cloud point line being crossed and thus reducing the concentration to below a supersaturated state on the cooling ramp (Figure 1). All samples were shown to be Form I from XRPD (X-Ray Powder Diffraction) analysis, with the same melting point as adjudged by DSC (Differential Scanning Calorimetry), with no evidence of the presence of Form II or III. Microscopy analysis from the samples reveal how the degree of supersaturation can impact the particle size, aspect ratio and morphology of the isolated material. This is another factor to be mindful of when developing a crystallization development method, as this may impact performance of the API further along in the drug development process.

Conclusion: MEK was identified as the most suitable solvent based on the controlled temperature experiments. The solubility was sufficient enough to ensure low solvent use on scale-up, whilst providing a wide MSZW to allow for further optimization. The solubility at 0°C was sufficiently low to ensure a good yield would be achievable.

References: Roger A. Granberg and Åke C. Rasmuson, Journal of Chemical & Engineering Data 1999 44 (6), 1391-1395

Side by side comparison of the MSZW of the four solvents evaluated

Microscopy images of paracetamol crystals isolated from MEK