The Design and Use of Supramolecular Gels as Crystallisation Media

Abstract

The solid-state of pharmaceuticals is a critical area of research for pharmaceuticals companies, primarily due to the impact the solid form can have on the physicochemical properties of a drug. Some of these forms can be challenging to isolate experimentally, thus new techniques to screen for unidentified forms and to control crystallisation outcomes are required to reduce the risk of forms being identified late in product development. This work focuses on the design and use of supramolecular gels to crystallise pharmaceutical compounds and attempt to control the crystallisation outcomes. The work builds on previous studies by the Steed group.

Chapter One focuses on the design of bis(urea) supramolecular gels using a 4,4-methylenebis(2,6-diethylphenyl) spacer group, which has previously been the core structure of numerous versatile molecular gelators, with the aim to build on our understanding of the structural factors that may promote or prevent gelation for this system. A methodical stepwise approach to crystallise these gelators successfully yielded the first reported crystal structure of a bis(urea) derived from the 4,4-methylenebis(2,6-diethylphenyl) spacer, providing valuable insight to the molecular conformation and intermolecular interactions formed. The findings corroborate the results of previous computational crystal structure prediction studies on gelators derived from the 4,4-methylenebis(2,6-diethylphenyl) spacer.

A further study investigates the development and use of API-mimetic supramolecular gelators as crystallisation media for two polymorphic AstraZeneca active compounds (AZD-6140 and AZD-2281) to influence the crystallisation outcomes. Whilst a new nitrobenzene solvated of AZD-2281 is reported, the gels failed to influence crystal factors, such as the polymorphic form, crystal habit or size. This is despite an extensive screening procedure using a wide range of solvents and gelator/substrate concentration ratios. An initial proof of concept is also reported to demonstrate the gelation capability of free and pinacol protected boronic acid gelators. This provides the foundations for further work to investigate whole molecule API-mimicking gelators by exploiting reversible covalent bond formation between diol groups and boronic acid functionalities.

The final two studies investigate the self-assembly behaviour of novel bis(acyl-semicarbazide) compounds. A series of aliphatic and aromatically spaced bis(acyl-semicarbazides) with chiral 1-R-phenylethyl/1-S-phenyl ethyl and achiral benzyl end groups are synthesised and characterised. A crystallisation screening of these compounds, including racemic mixtures of the pure R,R and S,S enantiomers, resulted in thirteen single crystal structures of these compounds being determined. The resulting structures provide insight to their self-assembly behaviour. The aromatic-spacer compounds exhibit dimer assembly without infinite unidirectional hydrogen bond synthons that may promote anisotropic assembly. This lack explains their poor gelation profile. Two distinct conformations (linear and folded conformations) were observed in the structures for malonyl- and succinyl-spaced compounds. The linear conformations assemble via urea α-tape synthons, which are thought to give rise to gelation behaviour. The folded conformers, which are stabilised by an intramolecular hydrogen bond, assemble into bilayer arrangements with hydrophilic cores and hydrophobic surfaces. Ultrasound is required in many solvents to induced gelation of the malonyl and succinyl-spaced compounds and this is attributed to the requirement to disrupt the intramolecular hydrogen bond and induce a conformational change. Only linear conformers and urea α-tape assembly were identified for the oxalyl- and adipic-spaced compounds with both spacers yielding highly versatile gelators. The impact of chirality on gelation behaviour is also described, with chiral compounds exhibiting superior gelation behaviour to their achiral counterparts. Racemic mixtures of R,R and S,S enantiomers also exhibit significantly reduced gelation behaviour compared to the enantiomerically pure components, likely due to enabling crystallisation in centrosymmetric space groups. The sensitivity of gels to anions were explored, with most gels undergoing a gel-sol upon the addition of trace amounts of anionic salts. However, the chiral oxalyl-spaced gelators are highly resistant to anion-disruption, maintaining gelation upon addition of up to 10 molar equivalents of anions. This was shown to be useful in the crystallisation of the salt procaine hydrochloride, with the oxalyl-based gelator the only compound able to form stable gels in the presence of the API salt and facilitate API crystallisation. The bis(acyl-semicarbazide) gels were also used to crystallise a range of pharmaceuticals to influence the crystallisation outcomes, including via an API-mimetic approach and attempts to achieve chiral resolution of chiral APIs. Whilst in most cases the crystallisation outcomes were not influenced by the presence of gel fibres, one new form of polymorphic RMOY was successfully isolated from chlorobenzene gels of chiral adipic-spaced bis(acyl-semicarbazide) gelators.