PRODUCTION OF PELLETS 1 MIXING

According to the standard operation procedure prepared for each experiment, first the weighed powders on a sensitive balance (SAUTER RC 1631, August Sauter GmbH, Ebingen, Switherland) were blended in a planetary mixer (Model A200, Hobart LTD, London, UK) for 10 minutes. The speed of the impeller was at the lowest rate to reduce its effect on the surface properties of the powder. The liquid binders were also mixed separately (if more than

one liquids were included in a formulation) and were added into the mixed powder using a syringe. The rate of addition of liquid was slow approximately 50 ml/min. The mixing process was continued for further 15 minutes. The sides of the bowl and the stirrer were scraped every 5 minutes to detach the material adhering and to ensure a homogenous mixture. The wet mass was placed in a sealed container until it was extruded in few hours time.

2.2.1.2 EXTRUSION

The wet mass was extruded using a ram extruder mounted in a mechanical press (Lloyd Instruments, MX50, Warsash, Southampton, UK), which was fitted with a 50 kN load cell. About lOOg of the wet mass was first packed and manually consolidated in the stainless streel barrel of a 2.54 cm internal diameter and approximately 20 cm long fitted with a centrally mounted die having 1.0 mm diameter and 5.0 mm length (L/R=10) by inserting a stainless steel piston. The cross-head positioned above the piston-barrel-die assembly was driven down at a constant speed, which was 200 mm/min in all experiments unless stated, to extrude the wet mass. The extrudate was collected in a plastic bag before it was spheronized. In a specific experiment where the effect of the speed of extrusion was studied, a die having 1.5 mm diameter and 15.0 mm length (L/R = 20) was used. The force acting on the material during extrusion was recorded as a function of time, and a force-time profile was produced on an attached computer. The average of about 250 points in the steady state flow stage (Harrison, 1982) of the force/displacement curve was calculated and recorded as an average extrusion force of the run. The mean of at least 10 runs was used to describe the steady state extrusion force of a formulation in the given experimental system.

2.2.1.3 SPHERONIZATION

About 40g of extrudate, at a time, was spheronized on a 12 cm diameter spheroniser (Model- 120, G.B. Caleva Ltd. Sturminster Newton, Dorset, UK) fitted with a cross hatch grooved plate, for about 15 minutes at a speed of 1800 rpm. The load was adjusted after a trial and error experiments to be approximately 40g in each run, and the time was determined by visual observation of the roundness of the pellets formed, however, it was approximately 15 minutes for all formulations unless specifically stated. The spheroniser was partially covered to limit the condensation of the binding liquid. The variation related to the processing parameters of the spheronization have been specified in each experiments. For the last set of experiments (PART-HI) where drug was incorporated in the pellets, a spheronizer with

22.5cm diameter (G. B. Caleva, Sturminster Newton, Dorset, UK), fitted with a parallel grooved plate was used. About 250g of extrudate was spheronized at a speed of 1000 rpm. The time was 15 minutes again in all the formulations, unless differently stated.

2.2.1.4 DRYING

Unless stated, all pellets were dried in a laboratory fluid bed drier (Laboratory fluid bed dryer Model No. FBD/L70, PRL Engineering, Mostyn, Flintshire, UK). About 200 g of the pellets were placed in the vessel with a perforated base, which enables air to pass through. A high velocity of hot air (60°C) separated the pellets and enabled them to move freely. The gas was distributed evenly through small orifices supporting the pellets at a rate sufficient high to cause incipient fluidization but not so high as to give the appearance of a vigorously boiling bed. This technique was efficient to dry the pellets in 30 minutes.

In a specific experiment where the effect of different drying techniques were studied, MCC pellets, produced with water/ethanol mixture, were additionally dried in a hot air oven (Pickerston Instruments Ltd., Romfolk, England), freeze drier (Micro Modulyo, Edwards, Sussex, England), and were placed in a desiccator with dried silica-gel for different duration of time. Pellets were spread on a shallow tray and were placed in the compartments of the oven, set at 60°C. The pellets dried by convection due to constant flow of the heated turbulent air over their surface. This continued for 24 hours. In freeze drying system the wet pellets were first frozen by the addition of liquid nitrogen, and were then subjected to low pressure (below the triple point, 4.579 mm Hg and 0.0099 °C) by high vacuum. The heat supplied by conduction or radiation, or by both changed the ice directly to vapour leaving only the solid structures. This process took 14 hours to produce dried pellets. Some pellets were placed on a perforated plate in a sealed desiccator with a dried silica-gel. The silica-gel absorbed the water vapour from the atmosphere of the sealed desiccator. As the vapour pressure in the desiccator reduced, water vaporized from the pellets until they were dried. This technique took 5 days to dry the pellets. Finally, the moisture content of the pellets was measured using Thermogravimetric analyzer (TGA) (HI-RES TGA 2950, TA-instruments, Leatherhead, Surrey, UK). The dried pellets were placed in a sealed container until further analysis.

2.2.1.5 FILM COATING

The pellets containing the model drug were coated in a fluid bed coater (Aeromatic AG,

Germany) which had a 8.5 cm diameter perforated bottom plate and with a bottom spray pneumatic nozzle. Preliminary studies established that the optimum conditions, i.e. even coating with no pellet agglomeration was possible with 40 g of pellets per batch, inlet air temperature of 60^C, outlet air temperature of 40^C, atomized air pressure of 0.2 bar, and coating feed rate of 0.6 ml/min. The coating suspension was heated to 60°C and was applied for 45 minutes for each 5% weight gain. Each batch of the pellets was coated at three levels (i.e. 5%, 10%, and 20% w/w). To assess friability, some uncoated pellets were fluidized in the coater without the spray of the coating material for 60 minutes.

2.2.2 STRUCTURAL CHARACTERIZATION OF THE PELLETS