The dissolution of solid drug formulations depends on the solubility and dissolution rate of the drug substances. In addition, several factors influence release kinetics. For example, drug interaction with excipients, compression force and hardness in tablets, and the type of binder in granulates, pellets, or polymer coatings all play a role.

Understanding the dissolution rate and release kinetics is essential for optimal pharmacotherapy. Dissolution of solid substances typically follows first-order kinetics due to diffusion processes. However, some formulations show zero-order release, where equal amounts are released in equal time intervals.

Multiple processes often occur simultaneously. These include wetting of the dosage form, drug dissolution, diffusion of drug molecules out of the dosage form, swelling of matrix formulations, and water uptake by insoluble films. As a result, the kinetics may not fit simple zero-, first-, square-, or cubic-root order equations.

To evaluate the best approach, release data are linearized using various models. The coefficient of determination (CoD) of the linearized curve indicates the best fit and suggests which process likely dominates [1,2,3,4,5,6].

The model-independent parameters, difference factor f₁ and similarity factor f₂, are used to compare release profiles. f₁ describes the relative error between two release profiles. It is calculated from the cumulative amounts released at a specific time T for a test and a reference formulation, or more generally, between any two formulations—for example, during drug development. On the other hand, f₂ is based on the sum of squared deviations of the released drug amounts from the two profiles [4,5,7,8,9].

Increasing attention has therefore focused on drug-loaded pellets and their controlled release. Specifically, this control is achieved by slowly swelling matrix systems or, alternatively, a final functional coating. Consequently, researchers have investigated the release of drugs from matrix pellets prepared by extrusion/spheronization and, moreover, coated with different amounts and types of insoluble ethylcellulose [10,11,12].

In addition, other studies report additional factors influencing drug release. For example, these include the filler type [13], the pH of the release fluid [4], storage conditions of the drug and methylcellulose matrix pellets [14], the amount of enteric polymer coating [15], and, finally, the salt concentration in the release fluid [16].

Further investigations examine the effects of talc and hydrogenated castor oil on the dissolution of metformin-loaded matrix pellets with an acrylic-based sustained-release coating [17]. Researchers also studied the sustained release of Lisinopril from mucoadhesive matrix pellets [18] and sinomenine hydrochloride from pellets produced using a novel whirlwind fluidized bed process [19].

Drug-layered inert pellets coated with polymer (heterogeneous pellets) were studied in order to assess the influence of release kinetics. Specifically, researchers investigated modifications of ethylcellulose coatings [20]; furthermore, they studied ethylcellulose mixed with polyvinylpyrrolidone (PVP) as a pore former [21], alternating layers of ethylcellulose and polyvinylacetate [22], various coating levels with final curing [23], and additionally, acetaminophen-layered sugar pellets coated with ethylcellulose [24]. Moreover, with polyacrylate coatings, drug release from layered pellets was delayed [7,25]. Therefore, changing the polymer type and layer thickness allowed control of the release rate over a wide range [8].

heterogeneous pellets coated with sodium benzoate

In our previous studies, heterogeneous pellets were manufactured using fluidized bed technology with a Wurster inlet. Initially, inert microcrystalline cellulose pellets (Cellets®175, median 170 µm), which offer a large specific surface area, were first coated with excipients as well as the water-soluble model drug sodium benzoate [26,27,28]. Consequently, these sodium benzoate (SB) pellets showed narrow particle size distribution, high sphericity, homogeneous layers, and additionally, rapid drug release. Subsequently, to achieve retarded release, the SB pellets were coated a second time with different amounts of ethylcellulose using the same fluidized bed technique [29]. As expected, the release rate decreased with increasing coating thickness. Moreover, the process was monitored in-line using spatial filter velocimetry (SFV) probes [27,28] to ensure control over particle size, distribution, and ethylcellulose layer thickness.

The present project aimed to produce heterogeneous pellets in a fluidized bed with a Wurster inlet while controlling the process using in-line particle size and coating thickness measurements. We studied sodium benzoate release kinetics, interpreted the partial processes affecting release, correlated release rate with polymer thickness, and determined the coating process endpoint to improve pellet quality.

PVP binder to improve layer stability

For pellet manufacturing, we followed a similar experimental approach to [26,27,28]. Small initial inert pellets (Cellets®175, median 170 µm) with large specific surface areas were coated with a sodium benzoate solution containing a small amount of PVP binder to improve layer stability. In the second step [29], SB pellets received varying amounts of insoluble but slowly swelling polyacrylate for retarded release. Agglomeration risk during coating was minimized by adjusting process parameters and adding talcum as an anti-stick agent. The SFV probe monitored particle size and detected agglomerates in real time.

Drug release was analyzed using zero-order, first-order, square-root, and cubic-root kinetic models. We identified the most likely release model by calculating area under the curve (AUC), dissolution efficiency (DE), and mean dissolution time (MDT), and by comparing the CoD of different models. We also calculated the difference factor f₁ and similarity factor f₂ to compare release profiles of different polyacrylate-layered pellet batches. Linearization works well for first-order kinetics. For other release profiles, nonlinear methods describe dissolution curves more accurately and reduce standard deviation in fitting parameters [30].

SB pellets are received as a free-flowing material. The coating process proceeded smoothly, and the Wurster inlet created a homogeneous fluidized bed pattern. As a result, the product shows a narrow particle size distribution [27]. The median x50.3 increased from 170 µm (uncoated Cellets® 175) to 200 µm. The sphericity of both the initial Cellets®175 and SB pellets remained above 0.9.

The polyacrylate coating of SB pellets caused no significant agglomeration. Only a few twins and triplets appeared under microscopic observation (Figure 1). The median size of polyacrylate-coated pellets grew to 232.2 µm, with a layer thickness of 16.1 µm (Table 4, P3, 17.6% polyacrylate content). Yield losses and incomplete sodium benzoate recovery resulted from material precipitation at the textile filter and inner chamber wall. Nevertheless, a sphericity above 0.9 confirms the formation of spherical products and indicates homogeneous processing.