Adaptation and optimization of the emulsification-diffusion technique to prepare lipidic nanospheres

Abstract

In this study, the emulsification-diffusion method traditionally used to prepare polymeric nanoparticles was adapted to obtain lipidic nanospheres (LN) using four model lipids. The method consists of dissolving the lipid in a partially water-miscible solvent (previously saturated with water) at room temperature or at controlled temperature depending on lipid solubility. This organic phase is emulsified in an aqueous solution of a stabilizing agent (saturated with solvent) by conventional stirring at the same temperature used to dissolve the lipid. This oil-in-water emulsion is then diluted with an excess of water at controlled temperature in order to provoke the diffusion from the internal phase into the external phase thereby causing lipid aggregation in the form of LN. This new approach for the preparation of LN has clear advantages over the existing methods, namely: (i) it is efficient and versatile; (ii) easy implementation and scaling up (with no need of high energy sources); (iii) high reproducibility and narrow size distribution; (iv) less physical stress (i.e., long exposure to high temperatures and to mechanical dispersion); (v) it is not necessary to dissolve the drug in the melted lipid. The selection of the water-miscible solvent and the stabilizers are critical parameters to obtain lipidic particles in the nanometric range. In general, solvents with high water miscibility and stabilizers able to form stable emulsions are preferred. The results demonstrated that it was possible to reduce the particle size by increasing the process temperature, the stirring rate, the amount of stabilizer, and by lowering the amount of lipid. Control of the preparative variables allowed to obtain LN with diameters under 100 nm. It was found that the influence of preparative parameters was associated with a mechanism based on a physicochemical instability. In this sense, it is suggested that the rapid solvent diffusion produces regions of local supersaturation near the interface, and LN are formed due to the ensuing interfacial phase transformations and lipid aggregation that occur in these interfacial domains. In terms of stability, only poly(vinyl alcohol) (PVAL) was able to preserve the physical stability of the dispersion for long periods after preparation. This effect was attributed to the ability of PVAL chains to form a strongly attached layer on the nanoparticle surface with an excellent repulsion effect.

Introduction

At the beginning of the 90s, submicronic particles composed of solid lipids were proposed as a novel colloidal drug carrier. They can be considered as a new generation of submicron-sized lipid emulsions where the liquid lipid (oil) has been substituted by a solid lipid (Müller et al., 2000). Unlike their predecessors, these suspensions show a very different behavior mainly concerning their drug delivery and physical stability (Westensen, 2000). In this respect, terms such as lipospheres (LS) (Cortesi et al., 2002, Bekerman et al., 2003), lipidic nanospheres (LN) (Cavalli et al., 2000), solid lipid nanoparticles (SLN) (Mehnert and Mäder, 2001), and more recently, nanostructured lipid carriers (NLC) (Müller et al., 2002a, Müller et al., 2002b) and lipid drug conjugates (LDC) (Olbrich et al., 2002) have been used to describe these systems. In all cases, we have a lipidic matricial structure with an average diameter in the nanometer range. SLN is the name generally accepted to describe these types of systems, although it is worth pointing out that this term is not necessarily descriptive of their composition and arrangement. In this paper, LN is the term preferred to generalize these types of systems. LN show several advantages over the traditional carriers (i.e., macro-, micro- and nanoemulsions, liposomes, polymeric micro- and nanoparticles, etc.), such as an excellent tolerability, good physical stability, protection of labile drugs, high drug payload, controlled release (fast or sustained) and likely drug targeting (Müller et al., 2000, Wissing et al., 2004).

Technologically, there are four main methods to prepare LN: (i) high shear homogenization (HSH) and/or ultrasound; (ii) high-pressure homogenization (HPH) including two approaches, the hot (H-HPH) and cold (C-HPH) homogenization techniques; (iii) solvent emulsification/evaporation (SEE) and (iv) via microemulsion (Gasco, 1993, Shahgaldian et al., 2003). The predominant preparation method is HPH, which involves the use of high-pressure homogenizers to disrupt the particles down to the submicron range. In both approaches (H-HPH and C-HPH), it is necessary to melt the lipid in order to incorporate the drug into the bulk lipid (Mehnert and Mäder, 2001). Under these conditions, the drug, the carrier, and other materials (e.g., stabilizers) may suffer degradation. Furthermore, the homogenization step causes an increase in the sample temperature, and therefore, an effective temperature control system is required. The HSH and/or ultrasound method share the problem that the SLN obtained generally have a broad particle size distribution ranging within the micrometer size. This can lead to physical instabilities such as particle growth upon storage. This could be improved by higher surfactant concentrations but with the correspondent toxicological implications. Further disadvantages of this method are potential metal (titanium) contamination, induction of chemical reactions or degradation, and an increase in temperature when ultrasonication is used. The SEE requires organic solvents such as cyclohexane, toluene and chloroform, which are classified as class 1 (highly toxic solvents) or class 2 (toxic solvents) by the ICH guidelines for residual solvent limit settings (Witschi and Doelker, 1997). The use of these solvents could result in severe acceptability and toxicity problems mainly for parenteral administration. Production of LN via microemulsions shows two important drawbacks: the lipid must be melted during microemulsion formation, and high concentrations of surfactants (≈15%) and cosurfactants (up 10%) are required (Mehnert and Mäder, 2000; Cortesi et al., 2002, Wissing et al., 2004).

Considering the above-mentioned aspects, it is clear that the development of other technological approaches to prepare SLN is an interesting challenge. In the middle of the nineties, our group developed and patented (Quintanar-Guerrero et al., 1996, Quintanar-Guerrero et al., 1999a) a method called emulsification-diffusion to prepare nanoparticles from preformed polymers. In this method, an oil/water emulsion is formed between a well-accepted partially water-miscible solvent containing the drug and polymer and an aqueous solution containing a stabilizer. The subsequent addition of water causes the solvent to diffuse into the external phase, resulting in the formation of nanoparticles. Depending on the boiling point of the solvent, this can be eliminated under reduced pressure or by cross-flow filtration. The technique has been reported to be efficient, versatile, of easy implementation and allows a high efficiency entrapment of lipophilic drugs in polymeric matrices (Kwon et al., 2001, Colombo et al., 2001). In the present work, the feasibility of using the emulsification-diffusion method to prepare LN was evaluated. The optimal conditions to obtain lipidic particles with submicronic size were investigated with four model lipids: glyceryl behenate (Compritol^® ATO 888), glyceryl monostearate (Geleol^®), lauroyl polyoxylglycerides (Gelucire^® 44/14) and stearoyl polyoxylglycerides (Gelucire^® 50/13). The influence of preparation parameters such as: (a) type of solvent, (b) process temperature, (c) amount of lipid, (d) type and amount of stabilizer and (e) stirring rate, are also investigated.