Process and product variables

Schaefer etal. (1990b) investigated melt granulation in a 10-1 Baker Perkins high shear mixer using polyethylene glycol (PEG) 3000 and 6000. The starting materials were calcium hydrogen phosphate (dgw = 23 /¿m) and lactose (dgw = 68 fim) of a quality identical to that used in a study on wet granulation in the same mixer (Schaefer etal., 1990a). Direct comparisons between wet and melt granulation can therefore be made. A conclusion of the work, which agrees well with the conclusions made by Kinget and Kernel (1985) in their study on melt granulation in a 10-1 Gral mixer, is that the main factors influencing agglomeration are the relative amount of binder and its viscosity, the impeller rotation speed and massing time. Except for an effect of liquid binder viscosity, the effects of the other factors mentioned agree well with results of wet granulation.

Schaefer etal. (1990b) showed that the agglomerate growth of calcium hydrogen phosphate can be correlated with the liquid saturation of the moist agglomerates. Rapid growth by the coalescence mechanism was seen at 80-85% saturation while a slightly higher saturation was observed by wet granulation. As discussed earlier, the difference is supposed to be caused by an error in the measurement of the density of wet granulated granules (Elema and Kristensen, 1992). Granules prepared by melt granulation contain the solidified binder and are, therefore, less porous than wet granulated granules. Smaller surface pores give rise to reduced penetration by mercury during the measurement. The optimum amount of binder to agglomerate the particular calcium hydrogen phosphate was in the range of 37-43% v/v at melt granulation and slightly higher by wet granulation, because loss of water by evaporation affects the liquid requirements during wet granulation in high shear mixers.

The results obtained with lactose were different. Agglomerates of lactose were consolidated to their minimum porosity after a short massing time. The ease of consolidation is affected by the particle size of the feed material. In wet granulation processes this means that the granule size remains unchanged during further massing because the liquid saturation is constant. In contrast, melt granulation of lactose showed a constant growth with massing time, despite constant saturation with the liquid binder: the higher the content of binder, the higher the growth rate. The amount of binder necessary for wet granulation (14.4% v/v) was considerably less than that required for melt granulation (21.2-25.4% v/v). The range of liquid saturations required to produce growth by coalescence by melt granulation was 80-90%, i.e. the same range as determined for calcium hydrogen phosphate. The different liquid requirements for agglomeration of lactose are undoubtedly caused by dissolution of lactose in the aqueous binder liquid.

Figure 22 shows the granule growth of lactose (a-monohydrate lactose) by melt granulation in a high shear mixer. Except for the effect of impeller rotation speed, which gives rise to different mean granule sizes, the graph demonstrates that the use of PEG 3000, independently of impeller speed, produces larger granules than does PEG 6000. Liquified polyethylene glycols wet lactose and have the same surface tension. The viscosity of PEG 3000 at 90°C is about 135 mPa s, and that of PEG 6000 about 500 mPa s -the less viscous binder liquid produces the largest granules. If the effect of

Lactose Mesh

9 13 17

Massing time (min)

Rg. 22 Effects of impeller rotation speed and polyethylene glycol (PEG) grade upon granule growth by melt granulation of 450 mesh lactose in a Pellmix 10 high shear mixer. Binder concentration 23% v/v; PEG 3000 (O, □); PEC 6000 (•, ■); impeller speed 500 r.p.m. (O, •) and 700 r.p.m. (□, ■). Reproduced from Schaefer et al. (1992a) with permission from the authors.

9 13 17

Massing time (min)

Rg. 22 Effects of impeller rotation speed and polyethylene glycol (PEG) grade upon granule growth by melt granulation of 450 mesh lactose in a Pellmix 10 high shear mixer. Binder concentration 23% v/v; PEG 3000 (O, □); PEC 6000 (•, ■); impeller speed 500 r.p.m. (O, •) and 700 r.p.m. (□, ■). Reproduced from Schaefer et al. (1992a) with permission from the authors.

the two PEGs on the growth results from the different viscosities, it conflicts with the claimed effect of liquid viscosity upon the probability of coalescence of agglomerates. Unpublished results have, however, shown that the product temperature, varied by applying a heating jacket, has an effect upon the growth. The higher the temperature and, consequently, the lower the binder viscosity, the smaller the granules produced by experiments using one of the two PEG grades. The difference between the effects of PEG 3000 and 6000 on the growth may be attributed to the very high viscosity of the molten PEG 6000, or it may be caused by interactions between the liquid binder and the lactose. The effect of binder concentration upon growth of two types of lactose shown in Fig. 23 may also be attributed to binder-substrate interactions.

Figure 23 shows the effect of lactose particle size upon the mean granule a

18 19 20 21 22 23 24

Binder concentration (%)

Fig. 23 Effect of binder concentration upon mean granule size by melt granulation of different lactose qualities in a Pellmix 10 high shear mixer. Impeller rotation speed 700 r.p.m.; 13 min massing time; a-monohydrate lactose 200 mesh (□), 350 mesh (A) and 450 mesh (O); anhydrous lactose (•). Reproduced from Schaefer eta/. (1992c) with permission from the authors.

18 19 20 21 22 23 24

Binder concentration (%)

Fig. 23 Effect of binder concentration upon mean granule size by melt granulation of different lactose qualities in a Pellmix 10 high shear mixer. Impeller rotation speed 700 r.p.m.; 13 min massing time; a-monohydrate lactose 200 mesh (□), 350 mesh (A) and 450 mesh (O); anhydrous lactose (•). Reproduced from Schaefer eta/. (1992c) with permission from the authors.

size by melt granulation of lactose with PEG 3000. The grade and particle size dgw of the lactose were mesh 200 (44 /¿m), mesh 350 (34 fim), mesh 450 (22 (im) and anhydrous lactose (13 pun). The effects of particle size, binder concentration and additional results on the effect of impeller rotation speed upon the growth of lactose agglomerates are in good agreement with the results of wet granulation of the material. They can be attributed to the effect of the agglomerate consolidation during the process. It is surprising, however, that anhydrous lactose, which has the smallest mean particle size, is agglomerated by less liquid than a lactose of similar particle size, and that larger granules than those shown in the graph could not be produced.

Although there are similarities between melt and wet granulation as to the effects of intensity of agitation, process time, particle size of the starting material and liquid requirements, there appear to be additional factors to be taken into account. The effect of the PEG grade upon granule growth (Fig. 22) and the effect of feed material properties (Fig. 23) are examples of factors that do not influence wet granulation processes to the same extent.

Was this article helpful?

0 0

Post a comment