Effects of liquid saturation on growth

Figure 11 shows the effect of the amount of aqueous binder solution on the mean granule size achieved in the liquid addition phase when granulating calcium hydrogen phosphate in a high shear mixer. The data demonstrate that the type of binder has an effect upon the granule growth. In particular, povidone solutions give rise to greater granule sizes than solutions of hydro-xypropylmethylcellulose (HPMC). Granulation with the povidone solutions produced denser granules than did the other binder solutions. Ritala etal. (1988) attributed the effect to a high surface tension, c.f. Table 1, which according to equation 5 produces a high liquid bonding strength. As a consequence, consolidation of the moist agglomerates becomes more pronounced.

Figure 12 shows the correlation between liquid saturation and mean granule size dgw obtained by the experiments shown in Fig. 11 and additional experiments with concentrated solutions of the binders having viscosities up to about lOOmPas. The viscosity had only a slight effect upon the granule growth.

The close correlation shown in Fig. 12 confirms equation 12 in predicting that the liquid saturation is a primary factor controlling agglomerate growth. The range of liquid saturations exceeds 100% due to a slight bias

Table 1 Viscosity and surface tension of the binder solutions shown in Fig. 11

Viscosity

Surface tension

Binder

(mPa s, 30°C)

(mNm-1, 25°C)

Kollidon 90, 3%

9

68

Kollidon 25, 3%

1

68

Kollidon VA64, 10%

4

50

Methocel E5, 3%

6

48

Methocel E15, 2%

11

50

0 20 40 60 80 100

Liquid saturation (%)

0 20 40 60 80 100

Liquid saturation (%)

fig. 12 The correlation between liquid saturation and the geometric mean weight diameter obtained in the same series of experiments as shown in Fig 11. Kollidon 90, 3%, 5%, 8%, •; Kollidon VA64, 10%, 20%, 30%, O; Kollidon 25, 3%, 20%, V; Methocel E5, 3%, 6%, 8%, A; Methocel E15, 2%, 3.5%, 4.5%, □. Reproduced with permission from Ritala etal. (1988), Drug Dev. Ind. Pharm. 14, 1041-1060. Marcel Dekker Inc., USA.

in the determination of the intragranular porosity by a mercury immersion method, Jaegerskou etal. (1984). During measurement of the granule density, mercury may penetrate the surface of the granules giving rise to a density value that is too high. An error of, say, 2% in the intragranular porosity calculated from the density determination produces a systematic error of about 10% in the liquid saturation value.

Figure 12 demonstrates that significant agglomerate growth of calcium hydrogen phosphate requires liquid saturations close to 100%. This agrees well with the strain behaviour of the material described in Fig. 6. Complete saturation is required to achieve plasticity when the porosity is in the range of 20-30%, as is the case in the described granulation experiments.

Figure 6 also shows that moistened lactose becomes highly deformable at liquid saturations well below saturation. Kristensen etal. (1984) have, accordingly, found that wet granulation of lactose proceeds at liquid saturations below about 60%. Figure 13 shows the effect of the liquid saturation upon the mean granule size obtained by granulating lactose (dgw = 56 /¿m) in a high shear mixer. Examination of the moist agglomerates by microscopy revealed that growth by coalescence was significant at saturations in the range 25-60%, above which the mass appeared overwetted.

Impeller rotation speed has an effect upon the granule growth of lactose, as shown in Fig. 13. It is to be expected that the agglomerate deformation produced by collision is dependent on the intensity of agitation, i.e. that the impeller rotation speed has an effect upon the agglomerate growth in addition to its effect upon the consolidation. Unpublished results have verified that the S-dgw correlation is affected by the intensity of agitation when the starting materials are free flowing. Cohesive powders produce agglomerates with high strength because of the small particle sizes. The

Fig. 13 Effect of liquid saturation in granulating lactose (d^ = 56 /jm) in a high shear mixer. Impeller speed 250 r.p.m. (O) and 500 r.p.m. (•). Reproduced with permission from Kristensen eta/. (1984), Pharm. Ind. 46, 763-767. Editio Cantor, Denmark.

strength resists the stress produced by the impact so that the resulting strain becomes dependent only on the plasticizing effect of the liquid.

The fact that the particular lactose used in the study showed plasticity and growth by coalescence at low liquid saturations is probably partly due to dissolution of lactose in the binder liquid. In a study on melt granulation of lactose (dgw = 68 /¿m) with polyethylene glycol (PEG) 3000 and 6000 in a 10-1 Baker Perkins high shear mixer, Schaefer etal. (1990b) found that liquid saturations between 80 and 90% were necessary to achieve growth by coalescence (see section on process and product variables). With calcium hydrogen phosphate these authors found that significant growth by coalescence proceeded within the same range of liquid saturations. This result indicates that the error in the determination of the liquid saturations shown in Fig. 12 accounts for 10-15%, corresponding to an error of 2-3% in the determination of the granule density. Elema and Kristensen (1992) have recently compared the mercury immersion method for determining the density and porosity of pellets with a gas-permeametric method. They found that the results obtained by mercury immersion were systematically 10-15% higher than the results obtained by the gas-permeametric method.

The correlation between liquid saturation and granule size is valid in the liquid addition phase as well as the wet massing phase of agglomeration. This is due to the effect of liquid saturation on the strain behaviour of the agglomerates. The liquid saturation is controlled by the consolidation of the agglomerates and, thus, is particularly dependent on material properties such as particle size distribution, particle shape and surface texture. The correlation is assumed to be characteristic of a particular starting material or formulation insofar as the starting material has a particle size distribution which renders it cohesive. This is the case with most pharmaceutical formulations. The correlation may, therefore, be applied in analysing the effects of scaling up and comparisons between mixer-granulators.

Was this article helpful?

0 0

Post a comment