API solid state: The challenges

In this speaker series, John Mykytiuk, Solid State Manager at our Dudley site, discusses API solid state chemistry challenges.

As a well-established PDMO^® with specialist expertise in handling complex chemistries, we bring over 50 years’ experience in contract development and manufacturing to the global pharmaceutical industry. Over the years, additional facilities have been added to our portfolio and we have continued to invest in our capabilities, dedicated to offering our clients first class API services and cGMP contract development manufacturing capabilities. As a part of Sterling’s ongoing expansion, the Material Science Centre at our Dudley headquarters is home to the dedicated solid state team to complement API development, allowing us to provide solid form development as a standalone activity or as part of a comprehensive API development program.

I would like to introduce you to the elements that make up solid state science at Sterling, and some of the challenges we have experienced.

Solid state science at Sterling is composed of the following elements which assist in affording an API with desired particle attributes of chemical purity, form or version, and particle size and habit. We can see the interconnectivity of these elements which are salt/co-crystal investigation in order to improve solubility, efficacy and physical properties, propensity to polymorphism which will identify the preferred form or version to progress into development, pre-formulation evaluation to aid in the selection of a form, salt or co-crystal of an API from versions which have similar characteristics, crystallisation development in order to produce the desired form or version which can include particle characteristics such as size, and if amenable, habit. Crystallisation is a bottom-up process where control is achieved from the crystallisation solution. We also conduct bulk particle manipulation to modify the particle size by milling, which can be an energetic process and small trial scale jet micronisation, which are both top-down processes.

API and substance manufacture can be fraught with challenges at all stages of development and the impact of solid state on a compound is no different. To help illustrate the various solid state elements, we have the following challenges. The first two examples, producing an API with a defined particle size and the solid form developed to an API deal with APIs in early development, while the change of process equipment concerns an API in production.

In the first example, the challenge was to produce an API salt with a defined particle size. During chemical development of the API salt, only one form had been isolated. Since the counter ion of the API had already been defined, the first element of solid state investigation that was pertinent was to establish the propensity of the API salt to polymorphism, which is form or version change. In this case, we found that the form originally produced during chemical development was a thermodynamically stable version and the preferred form. We also determined that the propensity of the API salt to polymorphism was low, and that a certain class of solvent should be avoided in the crystallisation process.

Subsequent production of additional API salt afforded another unique version, which was not a solvate and attributed to a change to the original process. Although the applied process change was considered advantageous in reducing the process time, it was deleterious for form control. The material isolated from the existing uncontrolled process was found to have a broad particle size range composed of thin plates which were prone to breakage and formed agglomerates and accretions. It was considered that the existing crystallisation process did not lend itself to further development.

Consequently, the controlled crystallisation required not only polymorph control producing the preferred version of the salt, but also particle size control which would improve the habit. Which is the shape of the particles and eliminate clustering and agglomeration of the primary particles. In order to address these challenges, it was necessary to identify an appropriate solvent system for the crystallisation and parameters, such as cooling rate, hold temperature, and seeding regime.

The seed regime was considered to be the main parameter for control since this would provide version control, but also be the most effective approach to particle size control. In order to identify an ideal solvent for the crystallisation of the API salt, concentration temperature profiles of the API salt in various solvents were measured in conjunction within silico evaluation. This revealed that there was a steep temperature response at elevated crystallisation solution temperature and afforded a short list of preferred crystallisation solvent systems.

Further in silico evaluation identified the particle size distribution of the API salt seed together with the required charge to achieve the desired particle size of the API salt from the crystallisation. As the seed was considered to be the main parameter for particle size control of the API salt, several iterations of seed formation were necessary. The simplest and most often employed approach is dry API particle size reduction by mechanical means. Several attempts with size fractionation afforded seed material which when employed in the crystallisation process resulted in flocculate and non-controlled particle growth, as the seed particles did not effectively disperse as individual particles in the crystallisation solution. This was of concern and suggested an inherent behaviour of the API salt to flocculate and could be a consequence of several factors.

The next level of seed formation employed solvent mediated ball milling. The conditions were readily established and produced a seed material of the required size range that did not flock in suspension. With the introduction of a defined temperature hold period following the addition of the seed material as a suspension, established from the concentration temperature profile of the API salt solution, and an appropriate cooling profile the crystallisation reproducing afforded the API salt with the required particle attributes of chemical purity form and particle size with no evidence of agglomerate or accretion and a uniform particle habit.

In this example the client’s API was in preclinical development and the solid state challenge was to improve the aqueous solubility of the API. An investigation into the polymorphic nature of the API was the first solid state element undertaken. This revealed that the supplied solid was the thermodynamically preferred form of the API while 10 other versions were identified which were partial or stoichiometric solvates. The propensity of the API to polymorphism was considered low. The solubility of the API was also found to be poor in deionised water. The low solubility of the preferred form of the API in water was attributed to interaction motifs which were revealed by structural determination using single crystal x-ray crystallography. At this point the API was considered to be either biopharmaceutical class 2, a compound that demonstrates low solubility but high permeability and well absorbed, or class 4, a compound that demonstrates low solubility, low permeability and is poorly absorbed.

Consequently, the second solid state element employed in order to improve the aqueous solubility of the API was a salt screen investigation. Various counter ion salt forming candidates were identified, some affording multiple versions. The two most promising API salt candidates were obtained with an alkali metal and an organic counter ion. Although both salts demonstrated desirable water solubility of the API, the alkali metal formation was not reproducible with different versions of the salt obtained with poor water solubility. The di-organic counter ion salt was reproducibly isolated, but the molecular weight of the salt was high and deliquesed above 75 percent relative humidity. The other issue with the salt candidate was revealed by pre-formulation investigations, which indicated that disproportionation of the API salt was a concern.

As salt investigations of the API did not afford a suitable candidate to progress and the preferred form of the API had a poor aqueous solubility, the next solid state elements that were considered appropriate required a controlled crystallisation of the preferred form in order to produce material of a uniform particle habit for micronisation to produce material with a DV90 of not more than 10 microns. Particle size reduction by jet micronisation was undertaken since the API was now considered to be biopharmaceutical class 2, and particle size reduction might enhance both solubility and permeability.

The evaluation of the API concentration/temperature profile in various solvents supported by in silico modelling, identified ideal solvent systems for crystallisation. The inclusion of an API seed charge to control form and a defined cooling profile of the crystallisation mixture afforded an effective crystallisation process. A preliminary crystallisation of the API that did not employ a seed charge without a hold temperature and a fast cooling profile, affording another solid form version, which increased the number of solvates to 11. A trial micronisation of atypical API in terms of form converted the material to the preferred form of the API, further reinforcing that the preferred form was a suitable development candidate.

Although the challenge here was to improve the aqueous solubility of the API, the polymorphism investigation revealed that the API was susceptible to solvate formation and that there was a structural explanation for the poor solubility of the preferred form. The investigation of counter ions as salt formers with the API afforded one potential candidate, this was considered to have drawbacks which did not make it suitable for further development. This directed the program to crystallisation development and particle size reduction of the API which reproducibly afforded the expected attribute and subjected to micronisation to produce material with the desired particle size for clinical development. This challenge demonstrates that certain characteristics of the API might limit the development opportunities for the API. In this case, the next option would be an extensive co-crystal investigation but crystallisation of the API from aqueous media might be inevitable.

The final example concerns an API which has been in production for a number of years. The API is produced by a seeded reactive crystallisation and then isolated on a filter dryer. The final operation is milling in order to achieve a critical particle size distribution. In order to reduce process time and increase throughput, an alternative filter dryer was considered. The alternative process equipment was expected to reduce the filtration time and accommodate twice the batch size and considered to be a low risk change to the API. The process time was reduced with the alternative process equipment, but the milled material produced did not pass the particle size distribution criteria. This was attributed to a subtle difference in the API solid isolated with the alternate filter dryer, which the existing milling parameters could not process satisfactorily.

The milling parameters were modified and successfully produced milled API with the required particle size distribution. The effectiveness of a particle size reduction operation can be dependent upon input material, something that was considered a low risk improvement in a manufacturing process had an unexpected consequence.

These examples reveal how solid state challenges for an API can be investigated and resolved with the solid state elements of salt cocrystal investigation, propensity to polymorphism investigation, pre-formulation evaluation, crystallisation development, and bulk particle manipulation here at Sterling.

Thank you for your time and if you would like to find out more about solid state science at Sterling, please visit our website at sterlingpharmasolutions.com, and if you have any questions please get in contact.