(a full version of this article is available at http://dx.doi.org/10.1016/j.ijpharm.2011.03.037)
There is a need for rapid pharmaceutical cocrystal screening. Many crystallization technique have the potential to be successful at generating cocrystals, including solvent evaporation, ultrasonication, supercritical fluid crystallization, slurry conversion, moisture sorption, melting/cooling, heating, co-grinding and even simply mixing components together. However, in pharmaceutical research, the selection of a suitable screening approach is often biased towards strategies that are both productive (proven high success rate) and time-efficient.
Slurry, solution-phase (e.g., solvent evaporation) or suspension techniques have developed as the preferred techniques, though they are less productive in terms of cocrystals produced. Now, by applying Automaxion's patented milling system to dry grinding, solvent-drop grinding (SDG) or liquid assisted grinding (LAG) which were previously impractical and laborious, cocrystallizations can be very rapidly executed.
The effectiveness of the system in producing cocrystals was tested using four known pharmaceutical cocrystals: carbamazepine/saccharin (1:1), caffeine/oxalic acid (2:1), caffeine/maleic acid (1:1), and caffeine/maleic acid (2:1). These model cocrystal cases were chosen due to their well-reported reproducibility, polymorphic and stoichiometric diversity, and availability of characterization data, which served as the basis for validating the outcomes of the study. The cocrystal components were weighed into 2-mL glass vials containing two 3-mm stainless steel beads. Where designated in the table following, 10-uL of solvent was added and the samples were subjected to neat grinding or SDG for 0.5, 2 or 4 hours using the Automaxion system.
Once milled, the samples were manually transferred onto an analytical plate and subjected to automated analyses via FT-Raman spectroscopy (Nicolet NXR 9650 equipped with a stainless steel 96-well sample stage). The unique screening hits were validated by Powder X-ray Diffraction (PXRD, PANalytical X'Pert Pro). The cocrystal formation was confirmed or refuted based on comparison of the PXRD peaks positions to those reported for the four model cocrystals. The sample composition, milling times and results are shown below:
|
Component 1 |
Component 2 |
Solvent (10µL) |
RESULTS |
||
|
Milling time 0.5hr |
Milling time 2hrs |
Milling time 4hrs |
|||
|
Saccharin 15.5+/-0.2mg (1equivalent) |
None |
UC |
UC |
UC + CC1 |
|
|
MeOH |
CC1 |
CC1 |
CC1 |
||
|
Toluene |
CC1 |
CC1 |
CC1 |
||
|
Caffeine 30+/-0.3mg (2equivalents) |
Oxalic acid 7+/-0.1mg (1equivalent) |
None |
UC + CC2 |
UC + CC2 |
UC + CC2 |
|
MeOH |
CC2 |
CC2 |
CC2 |
||
|
Toluene |
UC + CC2 |
UC + CC2 |
UC + CC2 |
||
|
Caffeine 25+/-0.2mg (1equivalent) |
Maleic Acid 14.9+/-0.1mg (1equivalent) |
None |
UC + CC1 |
UC + CC1 + CC2 |
UC + CC1+ CC2 |
|
MeOH |
CC3 |
CC3 |
CC3 |
||
|
Toluene |
UC + CC2 |
UC + CC2 |
UC + CC1 + CC2 |
||
|
Caffeine 30+/-0.3mg (2equivalents) |
Maleic Acid 9.0+/-0.1mg (1equivalent) |
None |
UC + CC2 |
UC + CC1 + CC2 |
UC + CC1 + CC2 |
|
MeOH |
UC |
UC + CC3 |
UC + CC3 |
||
|
Toluene |
UC + CC2 |
CC2 |
CC2 |
||
UC- unreacted components; CC1 – cocrystal having stoichiometry 1:1; CC2 – cocrystal having stoichiometry 2:1; CC3 – cocrystal of undetermined composition
Here we show that all unique mixtures underwent some degree of cocrystallization under at least one set of applied grinding conditions. Partial cocrystallization was observed within 30 min in the majority of experiments conducted. As expected, SDG was more effective as compared to neat grinding in that milling times of 0.5-2 hrs were sufficient to achieve complete cocrystallization in all four cases studied, whereas unreacted starting components were detected in all neat samples, even after 4 hrs of continuous grinding. While carbamazepine/saccharin and caffeine/oxalic acid cocrystal were obtained in a single solid-state form (polymorphic and/or stoichiometric), grinding caffeine and maleic acid resulted in a total of three solid forms. Two of these forms were identified as 1:1 and 2:1 cocrystals, (CC1 and CC2 respectively) and the third solid form (CC3) was new. Thermal analyses confirmed no appreciable solvent loss upon heating the solid, which suggests it is a non-solvated solid form. PXRD analysis of the form formed after 4 hours using 1:1 ingoing mixture indicated that only traces of unreacted caffeine and maleic acid were present, whereas large excess of caffeine and trace of maleic acid were present in the product obtained using 2:1 ingoing mixture. These data suggest that CC3 is likely a polymorphic form of 1:1 caffeine/maleic acid cocrystal. The milling experiments were performed without sample loss, as the ground material was contained within the vials during the entire course of the milling operation. Such design eliminates cleaning steps between experiments that are otherwise required for the currently known milling vessels and reduces the risk of sample cross-contamination. Although, in this study the samples were manually transferred onto an analytical plate for subsequent FT-Raman analysis, it should be noted that the milled samples could also be analyzed directly through the vial glass, thus preserving an intact sample. The breadth and diversity of the study could then be expanded further by applying additional portions of solvent to the intact sample in order to generate suspensions or solutions for subsequent crystallizations, without any additional demand on the solid components.
Raman spectra of Carbamazapine (Green) and Saccharine (Blue) starting materials and the Cocrystal formed by Automaxion milling (Red).
The patented Automaxion multisample milling system was successful at producing the expected cocrystals and led to the discovery of a previously unreported solid form in the presence of methanol. The device is capable of carrying out up to 48 experiments at a time, which delivers the highest capacity available for cocrystal screening via mechanochemical methods. Such throughput enables more experimental variables to be investigated in less time enabling faster and more extensive optimisation of conditions for cocrystal formation, e.g., cocrystal formers and stoichiometries, solvents type and amount, and cocrystallization time. In addition to the increased throughput, automated approaches provide greater control of experimental conditions, such as timing and grinding force exerted on a sample, and therefore improve reproducibility as compared to manual and low-throughput mechanised grinding. The Automaxion system was originally designed to be part of an automated syststem and incorporating it into fully automated cocrystal screening work-flows can be envisioned where automated powder, bead, and solvent dispensing would be followed by automated grinding and automated analysis. Such a system would not only be applicable to cocrystal screening but also salt screen allowing the pharmaceutical scientists to explore solid-form screening conditions faster.

Automaxion is grateful for the assistance Joanna A. Bis, David Igo, Beth Northon and Daniel Kinder of Catalent Pharma Solutions for help with the experiments, especially with the analyses and interpretation of the resullts.


Automaxion Limited. Registered in England No. 5396554. Automaxion SARL. Registered in France SIRET 48520026500018
UK Registered Office: Winnington House, Woodberry Grove, North Finchley, LONDON, N12 0DR
Automaxion does not claim rights to Logo’s or endorse services or products from other organisations that may appear on this site