A comparison between the extrusion forces and sphere quality of a gravity feed extruder and a ram extruder

A comparison of screen and ram extrusion-spheronisation of simple pharmaceutical pastes based on microcrystalline cellulose

The performance of two laboratory-scale extrusion apparatuses used to approximate the action of an industrial screen extruder, namely a multi-holed die ram extruder and a roller screen extruder, were compared. Both devices featured short dies (ram 2mm, screen 1mm) with die diameter 1mm and hole area fraction approaching 0.25. A series of water/microcrystalline cellulose (MCC) pastes with water contents varying from 45 to 60 wt% were extruded and pellets obtained from subsequent spheronisation of the extrudates characterised in terms of size and shape. Each device exhibited a different range of processing windows for acceptable spheronised products, with the ram apparatus being able to extrude a wider range of paste water contents than the screen device. The pellets obtained from extrusion-spheronisation (E-S) of the pastes via the screen device were in general smaller, with a wider size distribution, than those from ram E-S. These results are attributed to the different mechanical histories experienced by the pastes in the two types of extruder, which lead to different extrudate densities being achieved. MCC/water/calcium carbonate pastes were also tested, where the latter component represented a 'hard' (non-deformable) active pharmaceutical ingredient. Addition of calcium carbonate increased the stiffness of the paste, which could be countered by adjusting the water content of the deformable MCC/water matrix within the extrudability limits of the latter material.

The Evaluation of Formulations for the Preparation of Pellets with High Drug Loading by Extrusion/Spheronization

A capillary rheometer was used to evaluate rheological properties and the fluid mobility of mixtures with a high drug loading (80%) of three model drugs (ibuprofen, lactose, and ascorbic acid) when extruded. These drugs have a range of solubility in water, with 20% microcrystalline cellulose (MCC) as the spheronization aid, and water, pH 2.0, and pH 10.0 buffer as the binder liquid. The results were compared with the ability of the systems to form spherical pellets by the process of extrusion/spheronization. It was found possible to produce round pellets with a narrow size distribution by the process of extrusion/spheronization for formulations containing 80% of either lactose or ascorbic acid with MCC as the spheronization aid. It was not, however, possible to form pellets containing the same level of ibuprofen. This appears to be associated with the high level of fluid mobility observed when the wet masses were extruded in a ram extruder. A range of rheological characteristics in terms of shear stress, die entry pressure, angles of convergence, extensional flow, and elasticity were determined, but the variations in the values of these, which were observed, did not give an indication of the ability of the wet mass to form spherical pellets when subjected to the spheronization process. This could be associated with the fact that the selection of the conditions necessary to provide a valid quantification of the extrusion process did not truly represent the stability of the systems in terms of the mobility of the fluid when the wet mass was processed. The formulation of a wet mass with limited fluid mobility appears to be the first priority of formulations used in extrusion/spheronization.

The evaluation of modified microcrystalline cellulose for the preparation of pellets with high drug loading by extrusion/spheronization

The performance of microcrystalline cellulose (MCC) which had been modified by the inclusion of various levels of sodium carboxymethylcellulose (SCMC) in the wet cake prior to drying, in terms of their ability to form pellets by a standardised extrusion/spheronization process has been assessed. Initial screening of the ability of the modified MCCs to form pellets with an 80% level of lactose as a model drug identified two potential products containing 6 or 8% of SCMC (B 6 and B 8). These two products were compared with a standard grade of MCC (Avicel PH101) in terms of their ability to produce pellets with 80% of model drugs of low (ibuprofen), intermediate (lactose) and high (ascorbic acid) water solubility when subjected to a standardised extrusion/spheronization process. Also assessed was their ability to retain water with applied pressure using a pressure membrane technique and their ability to restrict water migration during extrusion with a ram extruder. The two new types of MCC (B 6 and B 8) were able to form good quality pellets with all three model drugs, whereas Avicel PH101 could not form pellets with this high level of ibuprofen. This improved performance was related to the ability of the new types of MCC to hold higher levels of water within their structure and restrict the migration of water in the wet mass when subjected to pressure applied during the process of preparing the pellets. There is evidence to show that the two new types of MCC can function over a wider range of water contents than Avicel PH101 and that they have an improved performance if the extrusion process is rapid and if, after incorporation of the water into the powder, the sample is stored for some time before extrusion.

Microscopic image analysis techniques for the morphological characterization of pharmaceutical particles: Influence of the software, and the factor algorithms used in the shape factor estimation

The present report highlights the difficulties of particle shape characterizations of multiparticulate systems obtained using different image analysis techniques. The report describes and discusses a number of shape factors that are widely used in pharmaceutical research. Using photographs of 16 pellets of different shapes, obtained by extrusion-spheronization, we investigated how shape factor estimates vary depending on method of calculation, and among different software packages. The results obtained indicate that the algorithms used (both for estimation of basic dimensions such as perimeter and maximum diameter, and for estimation of shape factors on the basis of these basic dimensions) have marked influences on the shape factor values obtained. These findings suggest that care is required when comparing results obtained using different image analysis programs.

The preparation by extrusion/spheronization and the properties of pellets containing drugs, microcrystalline cellulose and glyceryl monostearate

Pellets have been prepared by extrusion and spheronization containing microcrystalline cellulose (MCC) and four model drugs with decreasing order of solubility, paracetamol (P), diclofenac sodium (D), ibuprofen (IB) and indomethacin (IN) at a 10% level with and without the addition of a range of levels of glyceryl monostearate (GMS). The drugs differed in their response to extrusion in that all formulations containing the drug D had a 'steady state' extrusion profile whereas the other three drugs exhibited 'forced flow' indicating the possibility of water migration during the process of ram extrusion. The presence of GMS did not influence this effect. The drug D also required consistently less water to function than the other three drugs. In spite of these differences in extrusion performance, it was possible to prepare satisfactory pellets from formulations of all the drugs with 0, 30 and 60% GMS combined with 90, 60 or 30% of MCC at a range of water levels. It was also possible to prepare pellets containing the drug D with 70, 80 and 90% GMS, with corresponding quantities of 20, 10 and 0% of MCC. It was also possible to prepare the pellet formulations by dispersing the drugs in molten GMS, grinding and processing this with MCC and water. Such systems retained the processing characteristics of the composition made by the blending of the powder. The presence of GMS in all cases reduced the quantity of water required for the process to function. The steady state or the mean of the range of the forces observed during forced flow, were dependent on the composition and the quantity of water added. The surface of the extrudate appeared smooth and measurements of surface roughness established that the value of the rugosity R(a) for any of the extrudates did not exceed 6 microm. The extrudate diameter was found to increase with the quantity of GMS in the formulation. The pellets produced were all within a relatively narrow size range (three sieve fractions of a root two progression), the median value of which increased with the level of GMS. For the drug D, there was a linear increase of pellet diameter with increase in the extrudate diameter. For the three other drugs this relationship was less certain but nevertheless there was a similar trend for the pellet diameter to increase as the extrudate diameter increased, suggesting the mechanism of the process is the same irrespective of the composition. Considering the value of the shape factor e(R), all the pellets produced from the various formulations were well within acceptable levels for further processing and the only observable trend in the values was that the formulations with the lower water contents were the least round. The porosity of the pellets of the different formulations generally decreased with the increase in water used to prepare the pellets, the extent of this decrease being dependent on the drug and the level of GMS. The in vitro drug release from the pellets was controlled by the solubility of the drug, the lower the value of the solubility, the longer the mean dissolution time (MDT). This was not influenced by the presence of GMS or the method of incorporation of the drug into the formulation.

Image Analysis of the Shape of Granulated Powder Grains

This study presents and evaluates two new form factors for the characterization of pharmaceutical microparticles using image analysis techniques. The first factor, denoted Vr, is mean percentage variation in radial chord length (for a large number of radial chords drawn at small angular intervals) with respect to mean radial chord length. The second factor, denoted Vp, is percentage deviation of measured perimeter from the perimeter of a circle with radius equal to the mean radial chord length of the particle. Considering both ideal shapes and real pharmaceutical particle populations, these factors are compared with other form factors widely used in pharmaceutical technology. Our results indicate that Vr and Vp allow effective assessment of whether the particles of a given population show pharmaceutically significant deviations from sphericity. The two factors additionally facilitate identification of the basic shapes of particle outlines (notably ellipsoid, rectangular, and irregular). These factors may thus be of value for the characterization and monitoring of pharmaceutical pelleting processes.

Properties of microcrystalline cellulose and powder cellulose after extrusion/spheronization as studied by fourier transform Raman spectroscopy and environmental scanning electron microscopy

In this study, the effect of powder cellulose (PC) and 2 types of microcrystalline cellulose (MCC 101 and MCC 301) on pellet properties produced by an extrusion/spheronization process was investigated. The different investigated types of cellulose displayed different behavior during the extrusion/spheronization process. Pure PC was unsuitable for extrusion, because too much water was required and the added water was partly squeezed during the extrusion process. In contrast, MCC 101 and MCC 301 were extrudable at a wide range of water content, but the quality of the resulting products varied. In the extrusion/spheronization process, MCC 101 was the best substance, with easy handling and acceptable product properties. The properties of the extrudates and pellets were determined by Fourier transform (FT) Raman spectroscopy and environmental scanning electron microscopy (ESEM). FT-Raman spectroscopy was able to distinguish between the original substances and also between the wet and dried extrudates. The particle sizes of the raw material and of the extrudates were determined by ESEM without additional preparation. For MCC, the size of the resulting particles within the extrudate or pellet was smaller. However, in the extrudates of PC, changes in particle size could not be observed.

The evaluation of the rheological properties of lactose/microcrystalline cellulose and water mixtures by controlled stress rheometry and the relationship to the production of spherical pellets by extrusion/spheronization

The consistency of wet powder masses produced from two ratios (7:3 and 8:2) of alpha-lactose monohydrate (L) and microcrystalline cellulose (MCC) mixed with a range of water contents has been assessed with a parallel plate controlled stress rheometer. The range of water contents, which could be studied, was restricted to those, which could be extruded uniformly by a ram extruder. In the creep mode, the instantaneous compliance increased as the water content increased for both L:MCC ratios illustrating the increasing deformability of the mixtures with increasing water content. The derived apparent viscosity of the mixtures as a function of shear rate, increased as the water content decreased and the values for all the systems fell on a common line. This indicates that the measurements are providing a reliable assessment of the mixtures and that the change in water content and L:MCC ratio provides systems, whose change of viscosity with rate of shear is consistent at low rates of shear. The values of the storage and loss moduli obtained from oscillatory measurements, increased with a decrease in water content but this time the two ratios of L:MCC were not on a common line when related to the water content of the mixtures. There was a range of water levels over which both the values of the storage and loss moduli were approximately constant. This corresponded to the level of water, which produced the pellets of the smallest diameter and range of diameters and were of the most spherical shape when produced by a ram extruder and spheronization. For 8:2 L:MCC ratio, there appeared to be a value for both the storage and the loss moduli above which the wet mass could not produce good pellets. For the 7:3 L:MCC these limiting levels were not achieved before extrusion with steady state conditions could be maintained without the mass being too wet or too dry. Instead, there appeared to be minimum levels of the moduli required to ensure that the mixtures were able to produce good pellets.

The influence of liquid binder on the liquid mobility and preparation of spherical granules by the process of extrusion/spheronization

The influence of the type of liquid on the movement of water and the performance of the preparation of pellets by the process of extrusion/spheronization has been studied. Liquid movement was assessed by a pressure membrane technique and by extrusion, while the pellet properties were assessed in the terms of their median size, size range (interquartile range), roundness (by a two-dimensional shape factor) and porosity. The model formulations studied were microcrystalline cellulose (MCC) and mixtures of MCC and barium sulphate at 20, 50 and 80% levels. The liquids were water, a 25% solution of glycerol in water and an anionic surfactant (sodium lauryl sulphate) below its c.m.c. and two concentrations (0.01 and 0.0001%) of a non-ionic surfactant (Pluronic PF68). The presence of the different liquids changed the ease and extent with which the liquid could be removed (drying) and reabsorbed (wetting), resulting in lower levels of saturation with the glycerol solution and considerably increased levels of saturation with the surfactants. Changes in liquid movement during extrusion, were influenced more by the level of liquid and the rate of extrusion, than by its composition. The level of liquid was also an important factor in terms of the force necessary to extrude the different formulations. For a given level of liquid, the glycerol solution tended to increase extrusion force, while the surfactant solutions tended to decrease the extrusion force. The liquid levels, particulate composition and rate of extrusion were important in terms of pellet size, size range, roundness and porosity. The low level of liquid involved produced elongated pellets. The wet formulations produced larger, agglomerated pellets with a wide particle size range and a higher porosity. The lowest porosity pellets were prepared from mixtures with no or a low barium sulphate content while the highest levels of porosity were found with equal parts MCC and barium sulphate. In general, for equivalent liquid contents, pellets made with the glycerol solution were more porous than those prepared with water while the opposite was true for pellets made with surfactants. Although the different liquids influenced water movements, they did not prevent the formation of high quality pellets by the process of extrusion/spheronization.

Factors influencing the physical characteristics of pellets obtained by extrusion-spheronization

The objective of this work was to analyse the influence of the solubility of the drug and the filler on the physical characteristics of pellets prepared by extrusion/spheronization. Different formulations were prepared according to a statistical plan, using five different drugs and five different fillers selected according to their water solubility. The pellets were then obtained by a standardized extrusion/spheronization process and evaluated in terms of their physical characteristics by measuring the pellet size, density, porosity, mechanical strength, residual moisture after drying and shape. The results were first analysed by the analysis of variance to identify the main factors involved. The results were further assessed by canonical analysis and the significant influence factors were quantified in terms of regression equations. It can be concluded that the solubility of materials used (both drugs and fillers) plays an important role in the quantity of water required to form satisfactory pellets and on the physical characteristics of pellets. Quantitative relationships were identified between (a) the extrusion force required to provide extrudate, which would form pellets and the natural log of the filler solubility; (b) the quantity of the pellets in the size range 1-1.4 and the solubility of both the filler and the drug; (c) the apparent pellet density and both the level of drug and filler plus the solubility of the filler; (d) the pellet porosity and the quantity of drug and the inverse function of the filler solubility; and (e) the mechanical strength of the pellets and the square root of the quantity of drug.

Characterization of wet massing behavior of silicified microcrystalline cellulose and alpha-lactose monohydrate using near-infrared spectroscopy

The purpose of this study was to investigate the energetic state of water in silicified microcrystalline cellulose (SMCC) and alpha-lactose monohydrate wet masses using near-infrared (NIR) spectroscopy. The applicability of NIR spectroscopy to studying pharmaceutical wet masses at a wide moisture range was evaluated in comparison with mixer torque rheometry (MTR). With increasing moisture content changes in the physical properties of the samples resulted in an apparent increase in log (1/R) throughout the whole spectrum. The upward displacement of baseline and the relative height of water bands were greatest with materials that had a poor liquid-retention capacity. In the case of SMCC and 1:1 mixture of SMCC and alpha-lactose monohydrate, the height of the baseline-corrected water bands increased linearly at low moisture contents, thereafter achieving a plateau stage. According to the MTR results, the plateau stage of the band heights indicated a capillary state of liquid saturation. The second derivative spectrum was capable of distinguishing monohydrate, absorbed, and adsorbed water, which overlapped in the absorbance spectrum. When water was absorbed to the internal structure of the material (SMCC), the water bands were first seen at higher wavelengths, then followed by a shift to lower wavelengths. When water was only adsorbed onto the surface of the particles (glass ballotini), the water bands were seen directly in the region of bulk water.

The influence of type and quantity of model drug on the extrusion/spheronization of mixtures with microcrystalline cellulose. I. Extrusion parameters

Five model drugs, (methyl, propyl and butyl paraben, 4-hydroxybenzoic acid and propyl gallate), similar in their chemical structure were mixed with microcrystalline cellulose and water in different proportions and forced through a ram extruder. The overall water movement was measured by the difference between the initial water in the formulation and the water content in the plug remaining after extrusion was completed. The differences in theoretical and practical volume occupancy of the materials inside the barrel were calculated to look for trapped air inside the barrel. The steady-state extrusion force for each formulation was recorded. All five materials demonstrated differences in behaviour during extrusion. The relationship between each of the three properties measured and both the drug-load and initial water content was examined, to establish the potential relationship that existed between the differences due to the drug models. The five drug models were divided into two sub-groups, when examining the way that they underwent extrusion. Methyl, propyl and butyl paraben formed one group while 4-hydroxybenzoic acid and propyl gallate formed the other group. Within the former group the relationship between steady-state extrusion force and the percentage of drug and water present tended to be lower than those in the latter group. For the former group these relationships were non-linear.

Measuring the water retention capacities (MRC) of different microcrystalline cellulose grades

The ability of various types of microcrystalline cellulose (MCC) to hold water when subjected to an applied force has been assessed by a centrifuge technique. By considering the final percent of water retained after a standard centrifugation, as a function of the initial water content, a moisture retention capacity (MRC) can be determined. Statistical analysis of the results identified that it was possible to divide the nine samples of MCC into six sub-sets in terms of final moisture content retained. Those types which contained added polymer had by far the highest water level remaining. In terms of the MRC value, statistical analysis again sub-divided the celluloses into six sub-sets, although different from those for the final water content. Again those types containing added polymer gave much higher MRC values. As experiments were carried out with initial water/MCC ratios of different levels, statistical analysis of the influence of initial water content on the MRC values was undertaken with each type of MCC. The results showed a varying dependence on initial water level with the different types of cellulose. To provide a value of MRC which characterised the cellulose, the maximum value of MRC at the lowest initial water level was identified. Samples of MCC with low values of MRC have been shown previously to require less water for processing by extrusion/spheronization, while celluloses with a high MRC value appear better for the limitation of water movement during the process of extrusion/spheronization. The water retention must, however, also be considered in association with the rheological properties of the wet powder mass. Thus, while the Avicel RC591 had the highest MRC value, its rheological properties are so that the production of pellets with such a type can be less effective than with other types of MCC.

Emulsion/solvent evaporation as an alternative technique in pellet preparation

Paracetamol/Eudragit RS, paracetamol/ethylcellulose, and paracetamol/cellulose acetate pellets of different drug/polymer ratios (w/w) were prepared by the dissolution/solvent evaporation technique. These pellets were then characterized by particle size distribution analysis, ultraviolet (UV) spectroscopy, differential thermal analysis, and scanning electron microscopy (SEM). Hard gelatin capsules were filled with each particle size fraction of these pellets, and in vitro dissolution studies were performed to verify the capability of each series of pellets to control drug release. Pellets were spherical, presented a polynucleated microcapsule structure, and under certain experimental conditions, the yield of the preparation process reached very high values. The dissolution studies pointed out the slow paracetamol release from these pellets.

Formulation of ranitidine pellets by extrusion-spheronization with little or no microcrystalline cellulose

The present study was concerned with the feasibility of formulating ranitidine into pellets with a range of alternative excipients in place of microcrystalline cellulose (MCC). Eight ranitidine formulations employing two or more of the excipients lactose, barium sulfate, glyceryl monostearate, and MCC were processed by extrusion-spheronization, and characterized according to a series of physico-mechanical and dissolution criteria. Formulations containing lactose produced unsatisfactory pellets of wide size distribution and irregular shape, whereas formulations incorporating barium sulfate and glyceryl monostearate with or without MCC resulted in relatively spherical pellets of narrow size distribution and good mechanical properties. Ranitidine release was found to be rapid and virtually complete within 15 min, regardless of the pellet formulation. A direct relationship was observed between the concentration of MCC in the formulation and the properties of the pellets. In general, the higher the concentration of MCC, the rounder, stronger, and less friable the pellets. However, even pellets without MCC were also successfully prepared with a superior size distribution and shape over those with MCC. Overall, these results confirm that ranitidine can be formulated into pellet dosage forms with little or no MCC by the extrusion-spheronization process.

Measuring water distribution in extrudates using magnetic resonance imaging (MRI)

Magnetic resonance imaging (MRI) has been applied to the evaluation of the distribution of water in extrudates produced by extruding pastes. Two model drugs similar in chemical structure were mixed with microcrystalline cellulose (MCC) and with two different amounts of water and extruded at two different extrusion speeds using a ram extruder. Extrudates were collected during the steady-state stage of the extrusion profile and were analysed for the water distribution using MRI. The percolation threshold for each sample was calculated to evaluate the degree of water structure within the sample. The water distribution inside the extrudates was surprisingly uniform. The extrudates produced using the faster extrusion speed had a significant lower percolation threshold, which suggests the existence of a greater water structure in the extrudates. A significant correlation was found between the percolation threshold and the extrusion force, which had been used to provide the extrudates.

A centrifuge technique for the evaluation of the extent of water movement in wet powder masses

A centrifuge method has been applied to the assessment of water retention in pharmaceutical powders. Five drug models and microcrystalline cellulose (MCC) were each mixed with different amounts of water and centrifuged at different speeds. The amount of water retained by the wet mass was evaluated by drying the powders to constant weight. Binary mixtures of each of the five model drugs, MCC and water were also processed in the same way. From the amount of water extracted the moisture retention capacity (MRC) was calculated. The MCC retained water more strongly than the different drug models over a wider range of initial water contents. The five drug models, although similar in their chemical structure, were divided into two groups, in terms of their MRC values. 4-HBA and propyl gallate recorded higher MRC values than methyl, propyl and butyl paraben. For the drug models mixed with MCC, the MRC values recorded were similar, though it was still possible to divide the drugs into the two subgroups. A correlation between the MRC value recorded for the different systems and the hydrogen bonding solubility component was found. The application of different centrifuge speeds indicated that within the same material there were different mechanisms of water retention.

Importance of the fraction of microcrystalline cellulose and spheronization speed on the properties of extruded pellets made from binary mixtures

The aim of the study was to prove the importance of the binary mixture composition and spheronization speed on pellet properties. Extrudates from different binary mixtures of microcrystalline cellulose (MCC) and dicalcium phosphate dihydrate were prepared with a power-consumption-controlled extruder and spheronized at different speeds. The water content of the extrudate for the production of spherical pellets was evaluated. The pellets were characterized in terms of size, shape, porosity, mechanical properties, and disintegration. The fraction of MCC in the binary mixtures had the highest impact on the pellet properties. With an increasing fraction of MCC more water was required for successful pelletization, size and porosity of the pellets decreased, and the surface tensile stress increased. These observations were evaluated using the "sponge" and the "crystallite--gel" models for MCC. The latter led to the conception that an extrudate consists of two phases: a percolating crystallite--gel phase formed by MCC and water during extrusion and a filler phase formed by the second component of the binary mixture. This two-phase concept provides explanations for the extent of shrinking during drying and for the disintegration behavior. The spheronization speed had an influence on the size but not on porosity or surface tensile stress of the pellets. The best results were obtained at intermediate spheronization velocities of 10 and 13.4 m/sec. Fundamental properties of extrudates and pellets can be described by applying a two-phase concept of the crystallite--gel model.

Extrusion and spheronization in the development of oral controlled-release dosage forms

The concept of multiparticulate dosage forms was introduced in the 1950s. With the increasing use of multiparticulate controlled release (CR) oral dosage forms, in recent times there has been a rise in interest in the methods of preparing these dosage forms. A method that has gained increased usage over the past few years is that of extrusion and spheronization. It has been extensively explored as a potential technique and also as a future method of choice for preparation of multiparticulate CR dosage forms. In this review an attempt is made to outline the general process of extrusion and spheronization and to assess its importance in the development of multiparticulate CR oral dosage forms.

Extrusion/spheronization--effect of moisture content and spheronization time on pellet characteristics

As part of a larger effort aimed at optimizing the properties of pellets produced by spheronization of extruded masses, the effect of the moisture content of wet masses on extrusion force and torque was studied. The wet masses were composed of either microcrystalline cellulose (MCC) or mixtures of MCC with lactose or dicalcium phosphate. Based on the force and torque data, a moisture content "window" was defined for consistent extrusion. Moisture exerts a lubricant effect, and a moisture level of 100-120% w/w dry solid seemed necessary for the extrusion of MCC into rod-shaped, discrete pieces. Screen force clearly depended on the moisture content but was relatively insensitive to extruder speed, especially at 80% and 100% moisture content. The physical properties of pellets as a function of spheronization time were studied by sampling the material at known intervals. The percent yield, tapped density, and a two-dimensional sphericity index of an 18/20 mesh fraction of pellets were measured. Maximum yield, tapped density, and sphericity were achieved within 60 sec in the spheronizer. With increasing residence time, the shape and density were unchanged while the yield was severely reduced. Among the formulations studied, pellets with equal amounts of lactose and MCC were superior to those of pure MCC in yield, density, and sphericity. Based on these results, an outline to optimize the endpoint of the spheronization process for formulations containing MCC is suggested.

The power-consumption-controlled extruder: a tool for pellet production

Based on the assumption that there is a link between power consumption of an extruder and pellet properties, a control circuit for power consumption was developed. Powder and granulation liquid are fed separately into a twin-screw extruder. The power consumption is controlled by varying the pump rate at a given powder-feed rate; consequently each level of power consumption results in a specific water content of the extrudate for a particular formulation. The shape of pellets depends almost entirely on the level of power consumption irrespective of formulation. The size of dry pellets is additionally affected by a shrinking factor which depends on the water content. The power-consumption-controlled extruder is an appropriate tool for the production of pellets. The system is able to adapt the water content for a formulation automatically.