Investigation of Pellet Agglomeration in Fluidised Bed Coating Process: Influences of Size, Material, and Dispersion

This poster has been presented at the 2026 World Meeting on Pharmaceutics, Biopharmaceutics and Pharmaceutical (PBP) Technology.

Introduction

The production of coated pellets is increasingly important in the pharmaceutical industry because these multiparticulate systems offer advantages such as improved drug release and flexible dosing. A central challenge in this process is avoiding agglomeration, which can significantly impair the homogeneity and quality of the coatings. For enteric film coatings, achieving a uniform layer thickness without clumping is critical to ensuring the desired release profiles and gastric resistance. Various factors influence agglomeration tendencies, including process parameters, the physical properties of the carrier pellets, and the composition of the coating formulations. Agglomeration in a fluidised bed occurs when wetted particles collide and liquid bridges between particles can form.

The surface area of the pellets is a primary factor, while smaller particles have a larger surface area per unit of mass, which should increase the likelihood of particle-to-particle contact and therefore sticking. This study systematically investigates how pellet size, material composition and the choice of coating dispersion affect agglomeration and the achievement of gastric resistance.

Materials

The study utilized two types of starter pellets in three different size ranges:

• Microcrystalline cellulose (MCC) pellets (CELLETS®, ADD Advanced Drug Delivery Technologies Ltd; CH)
• Calcium phosphate (DCPA) und MCC pellets (PharSQ® Spheres CM, Budenheim, DE)

All pellets were layered with a water-soluble colourant (allura red colourant, AR) dissolved in a solution of polyvinyl alcohol/polyethylene glycol graft copolymer (Kollicoat IR, BASF, DE). Subsequently the colour-layered pellets were insulated (ISO) with a coating of the same polymer.

Two enteric coating formulations were compared:

• a) An optimized pellet-coating formulation with an enteric methacrylate polymer, a higher concentration of anti-tacking agents such as talc and silicon dioxide with triethylcitrate as plasticizer → AquaPolish® E (BIOGRUND GmbH, DE) (APE)
• b) A literature-based formulation of the same polymer containing also polyvinylpyrrolidone (< 3%), a lower content of anti-tacking agents, plasticized with propylene glycol → (REF)

Methods

Pellet characteristics such as size and shape were measured by 3D microscopy (VHX-X1, Keyence, DE). Additionally manual sieving was employed to measure the degree of pellet agglomeration using sieves of 355 µm, 600 µm and 800 µm. Bulk density was measured using a Scott-volumeter. To estimate the agglomeration from the microscopic images were analysed and filtered regarding the sphericity and the projected area of the object (single pellet or agglomerate). Two ways to calculate the agglomerate volume from the projected surface area were explored:

• One considered the agglomerates to be spherical and the subsequent calculations are based on the circle equivalent surface (ce)
• The other method assumed that the agglomerates are more plate-like in shape and the volume was calculated from the corresponding number of average circles (ac) that fitted into the projected area.

Both ways are illustrated in Figure 1 and Equations 1–3 show the calculations:

Equation 1: V_ag(ce) = (r(ce))³ · π · 4/3

Equation 2: V_ag(ac) = n · (r(ac))³ · π · 4/3

Equation 3: % Agglomerated Pellets = 100 · ΣV_ag / (ΣV_p + ΣV_ag)

Figure 1: Illustration of calculated volume of agglomerated pellets

Colourant Layer Coating

A Ventilus® multipurpose lab unit (Romaco Innojet, DE) was used for the first layering of the pellets at a constant batch size of 3.0 kg. Choosing a low spray rate of 2.3 g/min and a low solid content of the dispersion (8%) allowed to keep all process parameters. The parameters are shown in Table 1.

Enteric Layer Coating

Enteric coatings were applied on a SolidLab 1 fluid bed machine (Syntegon, DE). The batch size and the spray rates were adjusted to account for increasing surfaces with reduced pellet diameters. Table 2 shows the decreasing surface areas of the batches with increasing pellet size. Slight adjustments were required regarding the inlet air temperature. The process parameters are provided in Table 1. All experiments were carried out within a DoE framework set up in Modde Pro 13.

Content, Gastric Resistance (GR)

The content of AR and the release thereof were measured photometrically on an Agilent 8453 UV VIS Photometer (Agilent Technologies Deutschland GmbH, DE) at 506 nm. Release of the colourant from the pellets was tested with 0.1 mol/l HCl in 25 ml vials placed on a shaker at room temperature after evaluating that the temperature had no significant impact on the release of the substance.

Results

Pellet Properties

The results for the characterisation of uncoated pellets is provided in Table 2. Cellets seem to have more even particle size distribution than the pellets made of MCC and calcium phosphate. All pellets showed a high degree of roundness and aspect ratio. The values for the pellet surface were used to calculate the spray rate per surface area. Variations in diameter differed, with pellets made with MCC/calcium phosphate showing a wider particle size distribution.

AR + Iso Coating

The combined AR and ISO layers had a thickness around 10 μm as shown in Figure 2 A. Table 3 shows a comparison of pellet
agglomeration during the AR+ISO layering process, estimated by sieving and by microscopic imaging as well as the results of the two
calculation methods. Sieving yields lower agglomeration values in all cases, most likely due to insufficient separation of agglomerates.

It can be observed that pellets made solely from MCC show significantly higher agglomeration tendencies than pellets made from DCPA / MCC. It can be assumed that this is due to higher pellet densities and thus improved separation of wetted particles after collision. Improved binding of liquids onto / into the structure of these pellets might also contribute to the observed effect. The calculation method based on average circles (ac) always yields lower agglomerate concentrations. Considering that the number of pellets in agglomerates is rather low, as shown in Table 4, that estimation seems more reasonable. The results in Table 4 also underline the higher agglomeration tendency of Cellets in comparison to PSQ pellets.

Conclusion

Gastric acid resistance is primarily determined by the coating formulation, whilst pellet size plays only a minor role. However, both pellet size and the composition of the coating dispersion are key factors that influence pellet agglomeration during the coating process. As shown in Figure 8, these variables have a significant impact on key process outcomes. In particular, larger pellets, when combined with an optimised coating formulation, can significantly reduce the risk of agglomeration. Although process parameters can be adjusted to further minimise agglomeration, such adjustments often result in longer processing times and increased costs.

See the full poster on Investigation of Pellet Agglomeration in Fluidised Bet Coating Process here

(click the picture to download the poster)

Source: Biogrund, R. Jurek, M. Lachmann, E. Kempnich, K. Köhler
Poster: Investigation of Pellet Agglomeration in Fluidised Bet Coating Process

Enjoy also the other interesting posters of the PBP World Meeting 2026 here:

Excipients insights. Straight to your inbox. Subscribe now.

Interested in more information or a sample from Biogrund?