Prof. dr. Chris Vervaet

Prof. Dr. Chris Vervaet
Contact info

Laboratory of Pharmaceutical Technology
Harelbekestraat 72, 9000 Gent (Belgium)
Telephone nr. : ++ 32 9 264.80.69
Faxnr. : ++ 32 9 222.82.36
E-mail :

Research area : Extrusion, spheronisation, pelletisation

PhD. Thesis

Title: "Influence of process- and formulationparameters on the quality of pellets, produced by use of extrusion/spheronisation." (June 1997)


Extrusion/spheronisation is one of the most popular techniques for the production of pellets in the pharmaceutical industry. This technique, discussed in Chapter I, involves four steps: preparation of the wet mass (granulation), shaping the wet mass into cylinders (extrusion), breaking up the extrudate and rounding off the particles into spheres (spheronisation) and finally drying of the pellets.

The first aim of this thesis is to determine the influence of some process parameters during granulation, extrusion and spheronisation on the quality of pellets made by extrusion/spheronisation. The determination of the critical parameters will allow the optimization of formulations processed by extrusion/spheronisation and the validation of the production process. The instrumentation of a production-scale basket extruder by means of the measurement of the power consumption of the impeller is described in Chapter II. The data show that this kind of instrumentation allowed to differentiate between different formulations, the power consumption being dependent on the moisture content of the formulation and on the solubility of the drug. It was concluded that the measurement of the power consumption of the impeller could be used as an in-process control during extrusion. The second part of Chapter II deals with the modification of the Caleva Model 10 lab-scale basket extruder. Validation of the modified mini-extruder showed that pellet batches were produced in a reproducible way, for a low (100 g) as well as for a high (600 g) load of the extruder. The influence of granulation parameters on the particle size distribution and on the sphericity of pellets is discussed in Chapter III. The granulation was performed in a planetary mixer and in a high shear mixer. The process parameters were the granulation time, the granulation speed and the speed of addition of the granulation liquid. None of these process parameters had an influence on the quality of the pellets when the granulation was performed in a planetary mixer. Using the high shear mixer, the average particle size decreased when the energy input into the wet mass was increased (longer mixing time, higher rotational speed of the impeller). The speed of addition of the granulation liquid and the rotational speed of the chopper had no influence on the particle size distribution of the pellets. The sphericity of the pellets was not influenced by the granulation parameters. Storing of a mixture formulated with b-lactose in sealed containers before extrusion yielded a higher fraction of small particles due to the mutarotation of b-lactose in a-lactose monohydrate. The influence of the extrusion screen and of the impeller on the quality of pellets is determined in Chapter IV, outlining a zone in the phase diagrams where mixtures were situated which met all evaluation criteria. During the first part of this study it was shown that a more robust system was formed when an extrusion screen, characterized by the length-to-radius ratio (L/R-ratio) of the perforations, with a longer die length was used. This observation was due to the higher densification of granules during extrusion with a screen with a L/R-ratio of 4 compared to a screen with a L/R-ratio of 2. The second part of this chapter revealed that modifications to the design of the impeller improved the efficiency of the extrusion process. The influence of the perforation technique and of the geometry of the screen perforations is described in the last part of this study. Compared to extrusion cycles with a punched screen, extrusion with a drilled screen allowed to process mixtures with a higher drug concentration before they failed on one of the evaluation criteria. This observation was due to the more regular shape of the drilled screen perforations. Using a profile screen the maximal drug load was increased when b-lactose was incorporated as a model component, due to the extra densification inside the perforations. Extrusion of formulations with a high load of dicalciumphosphate dihydrate as a model component resulted in a complete block of the extrusion screen. The influence of a variable spheronisation speed on the particle size distribution and on the sphericity of pellets is discussed in Chapter V. Running a computer program the spheronisation speed during one cycle could be varied using the modified spheroniser. The complete spheronisation cycle can be divided into three parts: the initial speed (Phase I), the slope (Phase II) during which the spheronisation speed changes and the final speed (Phase III). Using a roll extruder no influence of a variable spheronisation speed on the particle size distribution was seen. Increasing the speed during spheronisation had an influence on the particle size when the wet mass was extruded with the Caleva Model 10 basket extruder. Prolonging Phase III resulted in a smaller amount of fines due to the agglomeration of small particles during spheronisation. The different behavior during spheronisation of extrudate produced with a roll and a basket extruder was be explained by the different structure of the extrudate. The sphericity of the pellets was determined by the highest speed during the spheronisation cycle.

Next to the determination of the influence of the process parameters the second aim of this study was to develop a pellet formulation which exhibited a fast release of a poorly water soluble drug. Hydrochlorothaizide (HCT) was used as a model drug. Chapter VI deals with the influence of solubilisers on the in-vitro dissolution rate of HCT form microcrystalline cellulose pellets. The in-vitro dissolution profiles showed that the release of HCT was increased when polyethylene glycol or Cremophor was incorporated into the formulation. Pellets formulated with 20 % (w/w) PEG 400 released 70 % of the total dose within the first 5 min. of the dissolution test, whereas pellets without PEG 400 released only 39 % of the dose after 90 min. The X-ray diffraction pattern showed that HCT was completely solubilised in the amount of PEG 400 present in the formulation. Decreasing the PEG 400-concentration decreased the HCT release rate from the pellets. The release from Cremophor RH40® also depended on the concentration of the solubiliser. The increase of the HCT release rate from pellets formulated with 21 % (g/g) Cremophor RH40® after a thermal treatment or during the stability study was due to the increased solubilisation of the drug in Cremophor RH40®. The in-vitro release of HCT from PEG 400-pellets was not altered during a 24 month stability study. The tabletting properties of pellets formulated with PEG 400 are described in Chapter VII. The data showed that pellets formulated with only microcrystalline cellulose did not yield tablet of an acceptable quality, whereas the hardness of tablets formulated with PEG 400 depended on the compression force and on the PEG 400 concentration. Adding PEG 400 to the formulation allowed to produce tablets with a low friability and a fast desintegration time. The external or internal addition of a desintegrator had little influence on the quality of the tablets. The fast release of HCT from microcrystalline cellulose pellets formulated with PEG 400, discussed in Chapter VI, was not lost if these pellets were compressed into tablets. The validation of a HPLC method for the determination of HCT in human plasma is described in Chapter VIII. Hydroflumethiazide (IS) was chosen as the internal standard. The chromatographs showed that, following extractions with methyl tert.-butylether and toluene, no interferences between HCT, IS and endogenic component occurred. All calibration curves were linear (r2 > 0.999) in the concentration range tested (0 - 1000 ng HCT / ml). The coefficient of variation for the variation within days and between days were < 5.01 % and < 5.89 %, respectively; while this values were < 3.03 if the accuracy of the HCT-determination was tested. The recovery of HCT (0 - 1000 ng/ml) was between 87.5 and 91.5 %, while 93.5 % of IS (1.25 µg/ml) was recovered from the serum. The limit of detection and quantification were determined at 3.35 and 11.18 ng/ml, respectively. The bioavailability of HCT in human volunteers was determined in Chapter IX. An oral dose of 50 mg HCT was administered to healthy volunteers, once as a commercially available tablet and twice as pellets. Type I-pellets were formulated with microcrystalline cellulose and HCT, while 20 % (w/w) PEG 400 was added to Type II-pellets. Cmax-values after administration of Type I, Type II and the tablet formulation were 105.9, 254.2 and 180.2 ng/ml, respectively; while tmax was determined at 195, 83 and 165 min, respectively. The relative bioavailability (Frel) of Type II-pellets was 117.3 % compared to the tablet, whereas Type I-pellets yielded a Frel of only 70.4 %. The lower bioavailability of Type I-pellets was correlated with a low in-vitro release from these pellets, while the better pharmaco-kinetic parameters of Type II-pellets were due to the faster dissolution rate from these pellets as the drug was present in a solubilised form.