Influence of excipients on mixtures containing high amount of dry extract from Aesculus hippocastanum L .

The aim of this study was to evaluate the influence of the qualitative composition of excipients on formulation technology of hard gelatin capsules containing high amount of dry extract from Aesculus hippocastanum L. (ESAH). The rheological and technological characteristics of ESAH were determined and the influence of pharmaceutical excipients on the properties of mixtures was studied by a 2 factorial design. The dissolution profiles of the capsules containing the best mixtures were obtained using sodium lauryl sulfate aqueous solution 1% (w/v) as medium. The ESAH showed high cohesivity and therefore low flow. Moreover, the ESAH showed lower level of residual moisture and was stable under controlled atmosphere. When considering pre-formulation, the filler/binder materials (microcrystalline cellulose or lactose) and glidant (colloidal silicon dioxide) increased the technological properties of the mixtures in comparison to pure ESAH. In conclusion, the dissolution profiles showed that there was no significant difference between the formulations and the ESAH, which demonstrated totally release after 30 minutes.


Preparation of mixtures
The mixtures were obtained by homogenization of ESAH with the filler/binder at 100 rpm for 10 minutes.
Then, the lubricant/glidant was added and the final mixing was performed for 5 minutes (Blender MixPlus ® , Tepron, Brazil). Six formulations were prepared (TABLE 1).

Determination of bulk and tapped densities
Samples of 5.0 g from ESAH and mixtures I, II, III, IV, V and VI were weighed independently and carefully transferred to a 25 mL graduated cylinder and the volume of powder column was measured (bulk volume).
The bulk density (bd) was established from the ratio between the mass of the sample and the bulk volume.
To determine the tappet density (td), the powder column was submitted to 1250 falls using a tapping device (n = 3) (PODCZECK, 1999).

Determination of Hausner Factor
The Hausner Factor was established from the ratio between the compaction density and gross density of dry extract and mixtures produced according to equation 1 (HAUSNER, 1967). where: HF = Hausner Factor, bd = bulk density (g/mL), td = tapped density (g/mL).

Determination of the Carr Index or Compressibility
The compressibility was established from the gross density and compaction, according to equation 2 (CARR, 1965 where: CI = Carr Index; bd = bulk density (g/mL), td = compaction density (g/mL).

Flowability
The flow properties of the sample were estimated by the angle of repose generated by samples of 30 g using a funnel with an outflow orifice of 1.4 cm at 7 cm from the surface. The analysis was performed in triplicate and the angle was calculated from the cone formed according to equation 3 (BAUER, FRÖMMING and FÜHRER, 2012;FAHR and VOIGT, 2015).
Calibration curve for the dry extract 250.0 mg of ESAH were transferred to the dissolution cube containing 900 mL of water (1.0% of SLS; w/v).
After 60 minutes under stirring at 100 rpm and 37.5 ± 0.5 °C, the solution was filtered and diluted to concentrations ranging from 0.055 to 0.300 mg/mL. The calibration curve was fitted by linear regression using PLS method and represents the average of three curves obtained from three determinations of each concentration at 270nm.

Results and Discussion
Seeds of Aesculus hippocastanum L. are used in several countries for the treatment of varicose veins and hemorrhoids. In Brazil, numerous industrial products are sold containing its extract. It is mainly marketed as seed powders, dried extracts or tablets. The technological characterization of powder beds is a crucial step in the selection of excipients that are more appropriate for the development of formulations and/or for the choice of operations used during the manufacturing process to obtain a technologically feasible final product.
In this context, analysis and determination of the size of the particles have a particular importance in the development of solid dosage forms. This is due to the important technological implications of packaging properties such as stability and flowability, which can lead to improper filling of dies from capsule or tablet machines causing a high variability in dose uniformity (WELLS, 1988;RITCHEL and BAUER-BRANDL, 2002;FAHR and VOIGT, 2015). The mean particle size can be estimated by several methods. The graphical and the mathematical methods are the most widely used (RITCHEL and BAUER-BRANDL, 2011;FAHR and VOIGT, 2015). Some discrepancies between procedure results have been related for the analysis of dried herbal extracts, but some correlation can be observed (DE SOUZA et al., 2000). Regarding the Revista Fitos, Rio de Janeiro, Vol, 10(4), 375-547, Out-Dez 2016 | e-ISSN: 2446-4775 | www.revistafitos.far.fiocruz.br Influence of excipients on mixtures containing high amount of dry extract from Aesculus hippocastanum L.
Luana Costa et al. analysis of the ESAH, the estimated mean particle sizes were 9.8 μm and 12.33 μm for graphical analysis and for the mathematical method, respectively. The data suggesting that the mean particle size of ESAH stays around 10 μm. At this range of particle size, it is classified as a very fine powder. Several rheological difficulties could be expected, mainly due to interparticle interaction and lower density. Additionally, the complex and amorphous composition of dried extracts provide a higher affinity with the surrounding water vapor, improving the hygroscopic tendency (TONG et al., 2008). In this way, the residual moisture and the stability under controlled atmospheres play an important role in the formulation, development and storage of phytopharmaceuticals (SOUZA et al., 2008a;COUTO et al., 2012).
Concerning the moisture requirements for powder systems used in pharmaceutical production, the residual humidity should be lower than 8% (FAHR and VOIGT, 2015).
However, this requirement becomes more critical for products from a herbal source. Thus, due to the hygroscopic nature of their compounds, as well as the microbiological risk, the residual humidity of dried herbal products should be lower than 4% (w/w) (FB5, 2010).
Concerning our results, the dried extract of A. hippocastanum showed residual humidity of 9.5 ± 0.0075% (w/w), suggesting that care should be taken with packaging and storage conditions. In this way, the behavior of products under a controlled atmosphere provides a basis for the prediction of physical stability during storage, by observing physical and/or chemical changes. Regarding the preliminary studies for the ESAH performing under a controlled atmosphere of 65%, the product showed mass increases in the first 24 hours, stabilizing from the fifth day at about 4.0% (FIGURE 1). The results show an interesting behavior for dried extract. Although the EASH presents high residual moisture, it was less sensitive to the atmospheric humidity. This data suggests an important stability of the ESAH during storage.
The determination of bulk and tapped densities is especially interesting in regards to the determination of the volume of powder column. Thus, the process performance of solid dosage forms as well as the packaging requirement can be influenced (PODCZECK, 1999).
The bulk density is a property of the powder bed and not of the individual particles. This depends on how the particles are packed, and the porosity of the powder bed. The determination of the tapped density is achieved when the powder column achieves a state of stability and the volume stays invariable. These assays are important when the powder is intended for the preparation of medicinal capsules, because the choice of appropriate capsule size will depend on the value of the apparent volume of the powder to be distributed (BAUER, FRÖMMING and FÜHRER, 2012;FAHR and VOIGT, 2015).
The ratio td/bd is related to the friction between the particles and therefore can be used to predict the flow properties of the particles. Powders of low friction between the particles show a Hausner Factor (HF) lower than 1.2, while the powders with higher cohesivity and restricted flow show HF exceeding 1.6 (HAUSNER, 1967). Regarding the compressibility index (CI), low values (< 15%) indicate materials with good compressional properties. On the other hand, materials with higher values of CI (> 25%) show poor compressibility. In this study, the HF and CI for both ESAH or mixtures were typical of powdered systems with poor flowability (SOARES et al., 2005a;RITCHEL and BAUER-BRANDL, 2011;FAHR and VOIGT, 2015).
Although the determinations of the CI and HF have been widely used for preliminary assessment of the flowability of powdered systems, the use of other methods is often report in the literature. The use of methods for measurement of flow known as dynamic procedures presents close correlations between the experimental data and the real technological characteristics of the powder. In this way, the angle of repose is a method that allows a dynamic measurement of flow by assessing the difficulty presented by the particles to flow freely through an orifice to a flat surface (PODCZECK, 1999;RITCHEL and BAUER-BRANDL, 2011;BAUER, FRÖMMING and FÜHRER, 2012). Thus, powder or granular material flows through the orifice and generates a cone at the surface. The angle of the base of the cone is known as the angle of repose. This definition was introduced by Dallavalle (1948), and systems which show angles of repose higher than 30° are typical of those with resistance to flow. The results observed for the materials showed a system with high difficulty to flow (TABLE 1).
Regarding the formulations employed to perform the factorial design, it could be observed that the combination of ESAH with filler/binder materials and lubricants/glidants changed the technological properties of the systems. Thus, the rheological properties of the mixtures were improved and the statistical analyses denote the improvement of the rheological parameters (HF, CI and angle of repose). Therefore, significant influence could be attributed to the following conditions: interaction between factors for the HF response, as well as the type of lubricant/glidant for the angle of repose (TABLE 2)   Concerning the complex and multifactorial nature of the rheological properties of powder bulk, such as interparticle interaction, the powder flow can be disturbed as a result of incomplete or irregular flow patterns.
Due to the typical intrinsic variability of such experiments, the results for the statistical analysis often cannot completely reflect the phenomenon that is happening during the determinations. In other words, it is not totally able to detect and clarify the majority of the phenomena. In fact, the statistical approaches by hypothesis tests allow us to identify and quantify the major factors or level set-ups crucial to a determined response. However, in cases like pre-formulation studies, where a continued nature should be taken into account, a graphic approach plays an important role in clarifying and understanding the physical and technological behavior during the formulation processes, such as the contribution of each factor, their combinations and levels, as well as the mathematical order followed.
In this context, the use of response surfaces can measure the tendency of the behavior of certain input variables on certain answers. The areas corresponding to the Hausner Factor (HF), Compressibility Index (CI%) and angle of repose (α), are represented in (FIGURE 2). from Aesculus hippocastanum L.
Luana Costa et al.  (TABLES 1 and 3). The rheological evaluation by Compressibility Index and Hausner Factor suggested that the mixtures containing colloidal silicon dioxide (V and VI), which showed lower angles of repose, also exhibited better flow properties. On the other hand, the mixtures containing magnesium stearate (II and IV) showed the largest angle of repose and probably the worst behavior of flow.
Regarding the contribution of each excipient, the influence of the CSD as a glidant enabled a decrease in the Hausner Factor and the angle of repose for both filling materials. In fact, the small spherical CSD agglomerates adhere to the surface of filler/binder material and reduce the attraction forces between particles, increasing the roller friction in comparison to sliding friction. This behavior can explains the better flow performance to mixtures containing CSD than other mixtures containing magnesium stearate. Additionally, hydrophobic lubricants such as magnesium stearate are able to forming capable of forming films on other excipients during mixing, decrease the solubility and prolonging the active liberation (PODCZECK, 1999;BAUER, FRÖMMING and FÜHRER, 2012;HURTADO et al., 2012;FAHR and VOIGT, 2015).
Orally administered drugs must be dissolved or in solution before they can be absorbed. Lower soluble molecules are solubilized in the gastro-intestinal tract by endogenous surfactants such as bile acids, bile salts, lecithins and so on, before they are absorbed (JONAT et al., 2004). In vitro tests using acids/bile salts showed an increase in the dissolution of poorly soluble drugs as the concentration of surfactant increases (KWAN et al., 1977;SHAH et al., 1989). However, due to the high cost of these substances, their use in routine testing of dissolution is impractical. Thus, the use of synthetic surfactants as an alternative has been frequently used in the routine. In this way, the surfactants Tween 80 and SLS have been widely used in order to promote the dissolution of poorly soluble molecules (SHAH et al., 1995;JAMZAD and FASSIHI, 2006).
However, as is well known, the dissolution of drugs is the result of the release of the molecule in the medium of dissolution. In this context, the diffusivity of dissolved specimens (drug molecules and complex moleculemicelle) plays an important role (JAMZAD and FASSIHI, 2006). The diffusivity of the complex drug-micelle is smaller than the molecule itself and change in the rate of dissolution is a result of an increase in solubility plus the decrease in diffusivity (CRISON et al., 1996;BALAKRISHNAN et al., 2004;JAMZAD and FASSIHI, 2006). Due to the high molecular weight of Tween 80 and the great mass of its aggregation micelles in comparison with SLS (TURRO and YEKTA, 1978;NERURKAR et al., 1997;JAMZAD and FASSIHI, 2006), the polysorbate provides low diffusivity of the complex drug-micelle and consequently reduces the rate of dissolution (BALAKRISHNAN et al., 2004;JAMZAD and FASSIHI, 2006). Therefore, the use of SLS is recommended by several official codes to study dissolution behavior of poor soluble molecules.
Thus, due to the low solubility of Escins (saponins from A. hippocastanum), the SLS was used as surfactant to perform the in vitro study. However, in order to quantify the release, the total ESAH was considered as the active principle. Therefore, it was assumed that its release represents the total release of all constituents of interest (DE SOUZA et al., 2001).
The data for the preliminary assessment of the release profile in different mediums from hard capsules containing the ESAH (250 mg) by using the paddle method at 100 rpm are shown in ( maximum performances were recorded at concentrations of SLS between 0.5 and 1.5 % (w/v). In this way, the medium containing SLS 1.0 % (w/v) was adopted as a release medium for capsules containing ESAH. The total release of ESAH in the medium was calculated by a calibration curve evaluated at concentrations ranging in accordance with experimental dissolution conditions. The curve was described by the equation y = 0.8142 x + 0.0071 and the determination coefficient (R 2 ) was 0.9997, indicating that the equation describes satisfactorily the experimental variance in the laboratorial conditions (FIGURE 3). The dissolution data showed similar profiles for the three batches tested: pure extract, formulations V and VI. Overall, maximum release of ESAH was achieved after 30 minutes (FIGURE 4). The capsules containing pure ESAH showed a slight increase in the released concentrations. However, no technological difference could be attributed to this behavior. Both formulations also showed similar profiles, but due to the solubility of the filler material, formulation containing lactose as filler presented faster release in the first minutes. The