Process and Formulation Design in Roller Compaction / Dry Granulation (RC/DG)

Introduction

Roller compaction/dry granulation (RC/DG) is a key process in pharmaceutical manufacturing as it improves powder flowability, density, and segregation resistance. The interplay between process parameters and material attributes presents a complex challenge in terms of the manufacturability of the produced granules. The aim of our study was to use advanced statistical modeling to optimize RC/DG process parameters and subsequently binder compositions by employing process and formulation design experiments. We used microcrystalline cellulose (MCC), silicified MCC, and dicalcium phosphate as filler-binder examples in RC/DG experiments. The variables included feeder-screw speed, roller speed and compaction pressure in the first stage of the study followed by different blend ratios of the above filler binders in the second stage. Granule and tablet properties, including flowability, bulk and tapped densities, as well as resistance to crushing, were analyzed as response factors.

Materials and Methods

Materials

Three materials were used for RC/DG studies: Microcrystalline cellulose (MCC, VIVAPUR® 101, JRS Pharma, predominantly plastic deforming material), silicified microcrystalline cellulose (SMCC, PROSOLV® SMCC 50, JRS Pharma, predominantly plasticdeforming material), and dicalcium phosphate (DCP, EMCOMPRESS® Anhydrous Powder, JRS Pharma, brittle-deforming). Due to the dependency of DG/RC granule properties on the particle size of plastically deformable binders, the raw materials MCC and SMCC were selected with comparable particle sizes.

Roller Compaction and Tableting

The raw material blends were mixed for 20 min in a V-mixer (120 L, JRS). RC/DG was performed with a compactor (Walzenpresse WP 50 N/75, Alexanderwerk), equipped with press rolls (Cavex CHUA 99, Flender) and a feeding screw (H4V41, Heynau Gears Production Service). Compaction took place with a variable gap using a 7.5 cm broad, axially profiled roller. The uncompacted fine grains were discarded, whereas the slugs were coarsely crushed before granulating over an oscillating sieve with an 800 μm screen. The feeding screw speed N [min-1], the roll S speed NR [min-1] and the specific compaction force SCF [kN/cm] were varied as part of the experimental design (Table 1). For the purpose of this study, a non-automated compactor was selected in order to enable manual setting of all key variables. The screw-to-roll speed ratio ϑ was calculated as (N /NR). S

Tableting of flat face tablets (13 mm in diameter) was performed with an instrumented tablet press (Pressima, IMA KILIAN). Sodium stearyl fumarate (PRUV®, JRS Pharma) was added as lubricant (1 %) to the RC/DG granules resulting from Exp. 1 – 8a (MCC) and Exp. 1 – 8b (SMCC) as per Table 1 and mixed for 3 minutes in a cube mixer. RC/DG granules, and for comparison also the unprocessed materials as physical mixtures, were tableted with different compression forces in the range between 2 and 10 kN during the process design experiments and between 2 and 15 kN for the experiments of the formulation design experiments. Results are compared based on tablets produced at compression forces of 10 kN.

Tab. 1 Factors and levels for roll compaction/dry granulation process design studies

Physical and Data Analysis

Powder, granule and tablet properties were analyzed according to the corresponding methods of the European Pharmacopoeia.

Data of the process design experiments was evaluated with Minitab statistical software (21.4.0) and Python (3.12.0). The randomized 23 full factorial design was generated by Minitab, the measured data evaluated with Python by polynomial interpolation. The formulation design was generated with Stat-Ease360. Model equations for formulation design experiments can be generated using various mathematical methods, including multiple linear regression, where mixture proportions serve as independent variables to estimate the response. Scheffé model equations are commonly used in this context, which is why we developed Scheffé mixture models (Cornell, 2002) with Python (3.12.0) (Märkle et al., 2025).

Results and Discussion

Stage 1: Process Design Experiments

With all tableting experiments, the theory of loss in manufacturability due to initial RC/DG (Sun and Kleinebudde, 2016) was proven. A very strong, inverse correlation was found between tablet breaking force and the granular coarse fraction (e.g. R² = 0.97, R = -0.98 for the SMCC:DCP blend). The coarse fraction was defined as percentage greater than the d90 of the corresponding physical blend. While commonly the screw-to-roll speed ratio ϑ is used to characterize the roller compaction intensity, we found that for the purpose of our study, it was more meaningful to multiply ϑ by the SCF. Suitable processing conditions, balancing sufficient granulation with adequate manufacturability, were found in the range of 20 – 60 [kN/cm] for the roller compaction intensity SCF*ϑ (Fig. 1).

Fig. 1 Response of the coarse sieve fraction and the breaking force of tablets (tablet compaction force 10 kN) to the roller compaction intensity (SCF* ϑ [kN/cm]).

Stage 2: Formulation Design Experiments

Using the process parameters resulting from stage 1, ternary blends of MCC, SMCC, and DCP were roller compacted. The granules were tested for bulk density, powder flowability (as angle of repose), and manufacturability. The contribution level of each of the three components in the blends was 0 %, 16.5 %, 33.3 %, 67 %, or 100 % respectively.

The contour plots for the response factors revealed practically linear behavior in case of bulk density, whereas powder flow and breaking force showed non-linear patterns, indicating potential synergistic effects (Märkle et al., 2025). Our results indicate that tablets tend towards higher breaking forces the more SMCC is used instead of MCC as a dry binder in roller compaction. This might be of particular interest for the choice of the dry binder, as MCC is also known to be more sensitive towards the use of lubricants in tableting than SMCC.

In order to create a synopsis of the three parameters, an overlay plot was created following the suggestion of the Appendix 2 of the ICH Q8. The superimposed formulation design highlighted certain areas of optimum blending ratios with regard to all three target parameters (Fig. 2). The optimal combination was defined for granules resulting in an angle of repose of α < 38 °, bulk densities of ρ > 0.5 g/mL and (tablet) breaking forces of σ > 100 N.

Fig. 2 Superimposed formulation design results of combined target parameters bulk density, flow properties, and breaking force.

Conclusion

The process design experiments confirmed that RC/DG reduces manufacturability compared to direct compression. Optimal processing conditions, balancing sufficient tablet strengths and granule formation, were identified to be 20 – 60 (SCF * ϑ) [kN/cm]. Formulation design experiments revealed optimal mixtures balancing SMCC, MCC, and DCP to achieve desired properties like low angle of repose, high bulk density, and strong tablets. The findings of this study provide guidance for selecting formulations and process parameters in RC/DG applications.

See the full poster on Process and Formulation Design in Roller Compaction / Dry Granulation (RC/DG) here

(click the picture to download the poster)

Source: JRS Pharma, Niclas Märkle; Gernot Warnke; Miriam Pein-Hackelbusch, poster: Process and Formulation Design in Roller Compaction / Dry Granulation (RC/DG)

If you have any questions or need a sample by JRS Pharma, please feel free to contact us: