Given the improved nanoparticle entrapment seen with NIMslurry (Figs. 2C, 3B and D), it appears that the maintenance of the wet state/absence of the oven-drying stage in the preparation of Nslurry was important. This helped to impart surface characteristics that facilitated nanoparticle residency in [w1] and/or prevented drying-induced augmentation of the hydrophobicity associated with PCL. With respect to the former hypothesis,

maintaining the wet state of the nanoparticles BI 6727 datasheet and resuspending them immediately in PVA solution may have allowed a satisfactory PVA ‘corona’ to form around the nanoparticles. It has previously been suggested that PVA can strongly absorb on the surface of protein-loaded PLGA nanoparticles [18], while its hydroxyl groups have also been envisaged to fix to the acetyl group of PLGA and thus improving the rehydration-ability of freeze-dried nanoparticles [19]. In the present work, the vinyl acetate segment of the partially hydrolysed PVA could have interpenetrated with the PCL molecule when the solvent diffuses

towards the aqueous phase during the polymer solidification process [20]. The adsorption of PVA on polymeric particles surface during their preparations is common [21], [22] and [23]. INK1197 datasheet It could be suggested that subsequent drying has disrupted the interaction between the PVA and the PCL molecules resulting in a more hydrophobic product (i.e. Ndried). Fig. 4A shows that when fractured to reveal their interiors, NIMslurry particles are seen to have a hollow core with nanoparticles below embedded within the wall of the microparticles. A mechanism leading to nanoparticle residency in the wall is proposed in Fig. 4B. The hollow core may be advantageous if capacity for the encapsulation of other agents is desired. Alternatively, if disadvantageous (e.g. leading to mechanical weakness), decreasing the

volume of [w1] or reducing water droplet size could be employed to reduce the volume of the void, or redistribute it into a number of smaller, individual voids. To determine the drug loading of typical NIM systems, three separate batches of NIMdried and NIMslurry were prepared and three samples taken from each for analysis. Drug loadings were found to be 3.80 ± 0.82% and 6.46 ± 1.26% for NIMdried and NIMslurry, respectively. This difference is statistically significant (Mann–Whitney U-Test; α = 0.05), again suggesting improved nanoparticle entrapment for NIMslurry. The in vitro cumulative drug release profiles are shown in Fig. 5 and provide further evidence of the different entrapment profiles for NIMslurry and NIMdried. For the latter, the drug release profile was very similar to that seen for nanoparticles alone, supporting other evidence that the nanoparticles were largely surface associated ( Fig. 3A). For NIMslurry, an initial lag phase was observed (no release for ∼1 day; only ‘noise’ on HPLC chromatograms).