(98e) Inverse Gas Chromatography (IGC) of Pharmaceutical Powders: Harnessing the Power of Adhesion-Cohesion Forces to Improve Pharmaceutical Development

The surface energy of pharmaceutical powders can affect their processing behavior including flowability and miscibility with other powders. Traditionally, surface energies are measured by classical contact angle measurements. However, the surface roughness, the presence of pores, and surface energy gradients of some materials make the contact angle measurements less appropriate for surface energetics determinations [1]. Surface energies has also been reported as the property of pharmaceutical powders that has the most impact on aerosol formation and powder aerosol performance, in inhalation formulations. Therefore, the measurement of these physicochemical properties is considered critical to describe and predict the aerodynamic performance of inhalation products [2].

An alternative technique that can be used for surface energy measurements is inverse gas chromatography (IGC). This technique has been previously applied to the measurement of important physicochemical properties of pharmaceutical materials such as powder surface energies, acid/base/polar functionality of surfaces, diffusion kinetics, solubility parameters and surface heterogeneity. In IGC the roles of the stationary (solid) and mobile (gas or vapor) phases are inverted with respect to traditional analytical gas chromatography (GC).

This work aims at describing the development and application of a home-made inverse gas chromatograph for the determination of surface properties in different pharmaceutical powders at infinite dilution regimen. In this study, IGC was used to explore different properties of formulations used in dry powder inhalers (DPI) and oral dosage formulations and the results obtained were compared to those obtained using next generation impactor (NGI) and contact angle measurements.

Two different case studies where devised to demonstrate the applicability of IGC as a tool for the characterization of pharmaceutical powders during the development stages of formulation design and development. This study was conducted not only to demonstrate the suitability of IGC for surface energetics measurement but also to benchmark the results obtained using the home-made equipment and the results obtained by the well-established classical methodology.

For the first case study DPI formulations were used and a correlation between IGC results and emitted dose (ED) and fine particle fraction (FPF) obtained by NGI was evaluated. Lactose carrier-based formulations were prepared with different active pharmaceutical ingredient (API) content (2% 3.5% and 5%) and different proportions of fine lactose (5%, 7.5% and 10%). The aerodynamic performance of each formulation was determined using NGI performed in duplicate for each formulation. The dispersive (apolar) component of the surface energy, as well as the specific (polar) component were measured for each formulation in triplicate. By calculating the specific work of adhesion (W_a^S) of tetrahydrofuran (slightly basic) and chloroform (slightly acid) to the surface particles it was possible to chemically characterize the surface of the particles for all formulations, regarding their acid/base properties. The ratio between the works of adhesion calculated is a good indicator of the chemical behavior of the surface. The results obtained for the ratio of the work of adhesion varied between 1.63 and 1.71 for all formulations tested, demonstrating that the surface chemistry does not change significantly with changes in the formulation variables. This observation is consistent to the fact that the API content is always much lesser than the Lactose content. The average surface chemistry is not significantly changed by varying the API content between 5% (w/w) and 10%(w/w).

The results obtained for the dispersive component of the surface free energy (g^d) did not correlate well with the % Emitted Dose (ED) (r²=0.199) neither with the Fine Particle Fraction (FPF) (r²=0.548) measured by NGI. Only when the contribution of the specific interactions (Acid/Base) was accounted for, the determination of the surface interaction energy (SIE) correlation with FPF was observed with good agreement (r²=0.808; Figure 1 a) confirming that the polar contributions of the surface free energy measurements of particles play a crucial role in understanding and determine particle-particle interactions as previously described [3]. It was also observed that the increase in the % of fine lactose present in the formulation increases the SIE (fic. 1 c), which is in agreement with previous observations [2] and counter intuitive to observations made for other materials since higher SIE are typically associated to more particle to particle interaction. However, it is proposed that for DPI formulations it is required some additional energy to ensure that micronized API which is very cohesive to be easily deagglomerated to originate better aerodynamic performance. The fine lactose increases the total surface area of the powder and competes with the API fort the high energy spots present in the coarse lactose therefore the increase in FPF observed when fine lactose is added to the formulation.

In addition to the IGC experiments, the traditional liquid wetting angle approach, for surface energy determination was also applied to surface characterization of four excipients commonly used in oral dosage forms, namely two diluents (lactose monohydrate, LM, and microcrystalline cellulose, MCC), one glidant (silicon dioxide, SD) and one lubricant (magnesium stearate, MgSt). Throughout these wetting angle studies, the surface energy, work of cohesion, work of adhesion and spreading coefficient between different excipients were determined.

Results showed that both SD and MgSt present a lower work of cohesion between their particles (28 mJ/m² for MgSt and 31 mJ/m² for SD) compared to the diluents (120 mJ/m² for LM and 105 mJ/m² for MCC). Since MgSt and SD present lower work of cohesion than adhesion to the different excipients (45 mJ/m² for LC-MgSt, 43 mJ/m² for LC-SD; 45 mJ/m² MCC-MgSt and 37 mJ/m² MCC-SD) they are likely to adhere and spread over both diluents surfaces. Raman imaging confirmed the surface coverage of diluents by both MgSt and SD when physical mixtures between the diluents and glidant/lubricant were prepared. As a result of the ability of MgSt or SD to spread over diluent particles during blending, these additives showed to be able to reduce the angle of repose, cohesion, compressibility and specific energy of the individual diluents.

IGC measurements were conducted in the same four excipients to determine the surface energy as well as the work of adhesion and cohesion and correlations between the IGC and liquid wetting angle results were explored, allowing to understand the benefits of each technique and major inconvenient.

The work conducted demonstrated that IGC can be used as an alternative methodology to support both oral dosage forms and DPI formulation development. The proposed methodology was successfully used to estimate the aerodynamic performance of a DPI formulation as demonstrated by the good correlation between NGI data and surface energy obtained using IGC. Preliminary work with excipients used in oral dose formulations also allowed identifying that the IGC data can be translated to this field of science, allowing insights on blending efficiency, predicting the impact of using excipients in formulations regarding final blend flowability and performance.

References:

[1] B. Riedl and P. D. Kamdem, “Estimation of the dispersive component of surface energy of polymer-grafted lignocellulosic fibers with inverse gas chromatography,” J. Adhes. Sci. Technol., vol. 6, no. 9, pp. 1053–1067, 1992.

[2] D. Cline and R. Dalby, “Predicting the quality of powders for inhalation from surface energy and area,” Pharm. Res., vol. 19, no. 9, pp. 1274–1277, 2002.

[3] D. Traini, P. Rogueda, P. Young, and R. Price, “Surface energy and interparticle forces correlations in model pMDI formulations,” Pharm. Res., vol. 22, no. 5, pp. 816–825, 2005.