The pharmaceutical industry intends to switch from batch to continuous production. Continuous manufacturing offers many benefits when the process is constantly within control, ensuring appropriate end-product quality. This can be achieved when the process and formulation are well-understood since then, deviations can be rapidly detected and real-time adjustment of process parameters becomes feasible. Twin-screw hot melt granulation (TS HMG) is a valuable and promising technique for the pharmaceutical industry due to its versatile applications and various advantages over other techniques (such as batch melt granulation, twin-screw wet granulation, dry granulation). However, the TS HMG process is still in its infancy and is for this reason still an unexplored black-box. In order to allow implementation of QbD and strategically placed PAT technologies, this process needs thorough examination. Chapter 3 showed that caffeine anhydrous undergoes a polymorphic transition during twin screw melt granulation with Soluplusr which is stimulated depending on the binder concentration and/or granulation temperature. Inline Raman monitoring during a HMG process proved to be useful to monitor this polymorphic transformation during processing. Moreover, a correlation was seen between these in-line collected Raman spectra and FTIR spectra obtained during rheological oscillatory temperature ramps. The polymorphic conversion of caffeine anhydrous could be detected in the same temperature range with both techniques, proving the comparability of plate-plate rheometry and HMG for this case with the used parameter settings. The Raman spectra, thermal analysis as well as the FTIR results demonstrated a change of caffeine Form II into Form I which could be detected at 120 C for C1, whereas this conversion took place at 110 C for blend C2. For C3 to C6, the polymorph transition did already occur between 40 C and 60 C. Conversion to Form I improved the granulation efficiency owing to the enhanced deformability in comparison with the stable Form II. In situ FTIR during rheological tests allowed investigating a formulation’ chemical properties while applying a defined stress and utilizing a rheometer proved to be relevant to gain knowledge of the material behavior during a twin screw HMG process. In chapter 4, the combination of thermal analysis, rheological characterization, and microscopic images were used to unravel the melt granulation mechanism of an immiscible drug-binder blend (caffeine-anhydrous/Soluplusr) during continuous HMG. Thermal analysis of the granules showed a double Tg and rheological characterization demonstrated a clear loss peak in the tan(d) curve which revealed that the binder did behave as a separate phase allowing efficient binder distribution. Hence, the nucleation step, initiated by the immersion mechanism, is considered to be continued by a distribution step during melt granulation. While distributing, a thin binder layer with restricted mobility is formed on the surface of the primary drug particles which is covered by a second layer with improved mobility if the concentration is adequate. NIR-CI images in combination with torque, size and shape values revealed that the formation of the high mobility layer in this blend is established from 15% (w/w) Soluplusr. When the used Soluplusr concentration was 20% (w/w) or more, this high mobility layer became thickened and, consequently the granule properties were modified. The granules manufactured with a binder concentration of 20% (w/w) or higher became smaller and more spherical in function of temperature whereas a lower binder concentration resulted in larger and more needle-shaped granules in function of temperature. It is clear that the binder distribution highly affects the granule characteristics. Therefore, the combination of the applied analytical techniques was advantageous to provide insight in the granule mechanism as function of process temperature and binder concentration for an immiscible drug-binder blend. In chapter 5, it is demonstrated that utilizing a rheometer allowed to distinguish different granulation mechanisms. For the miscible MPT/SLP blend, no clear loss peak but a broad and high tan(d) curve was observed owing to the plasticizing effect of MPT on the SLP binder and to the hydrogen bonds formed between both. These phenomena impeded proper binder distribution avoiding friction and, hence, no clear peak could be discovered. These observations revealed that the binder distribution step, which followed the immersion step is, is restricted for the miscible formulation. Hence, instead of a continuous binder film (which was the case for the immiscible blend), the binder is rather distributed in discontinuous patches due to the low spreading efficiency. This consequently affected the influence of temperature and binder concentration on the granule characteristics. The restricted binder distribution resulted in a low drug homogeneity within the granules, but also in between the different granule size fractions. At low binder content, granule growth occurred by kneading small nuclei together, whereas at high binder concentration larger nuclei were produced. Elevated granulation temperature narrowed the granule size distribution and lowered the fraction oversized granules. The executed research revealed a different granulation mechanism for a miscible and immiscible drug-binder formulation due to the different drug-polymer interactions and hence, the processes of both systems should be interpreted differently. In chapter 6 it is examined how varying process and formulation parameters influence the different granulation mechanisms (miscible vs immiscible formulation) and how these affect granule and tablet characteristics after continuous hot melt granulation. Drug-binder miscibility induced plasticizing of the binder which limited the maximum possible granulation temperature. For these blends, granule growth was rather attained by kneading granules together and, thus the extent of barrel filling was highly influencing the granule and tablet properties. Granule growth of immiscible drug-binder blends was acquired after distribution of the binder allowing nuclei to stick together. Since the binder did behave as a separate phase, the granulation temperature could exceed the Tg/Tm of the binder. For these blends, binder concentration was the most influencing factor. Moreover, these blends contained caffeine anhydrous as a drug, which is able to convert from the commercial Form II into the more deformable Form I, which facilitated granule growth. This conversion was favored at elevated granulation temperature and/or higher binder concentration and was more expressed when Soluplusr was used as a binder compared to PEG. Furthermore, also the differences between an amorphous and brittle binder were investigated. When using an amorphous binder, the tablet tensile strength depended on granule size and deformability. In contrast, granule strength was more important regarding tablet quality when using a brittle binder. However, this was not the case for caffeinecontaining blends, since these phenomena were dominated by the enhanced compression properties of caffeine Form I. This research emphasized the importance to comprehend the material behavior during processing before the influence of the process parameters onto the granule and tablet properties can be understood. Based on the DOE models, optimal TS HMG parameter settings could be calculated in order to obtain granules having required characteristics to produce tablets with desired quality. Chapter 7 illustrated the impact of interaction between two matrix formers and the drug on the drug release from the melt granules. This study demonstrated that SA and PEO can be used as matrix-formers during twin screw melt granulation where increasing the PEO content extended the release of a highly water soluble drug, MPT. The solid state analysis revealed a preferred interaction of the MPT molecules with stearic acid impeding the PEO to form hydrogen bonds with the stearic acid chains, which was the case in drug-free granules. However, this allowed the PEO chains to crystallize inside the stearic acid matrix and, hence, elevating the sustained release characteristics of the stearic acid matrix. If water penetrated into the high molecular weight PEO, swelling of the polymer occurred with the formation of a gel-structure making diffusion of the drug more complicated. The increased crystallinity of the polymer hindered the hydration step of the polymer, increasing the release sustaining properties even more. Furthermore, the intact crystalline fatty acid matrix covered the PEO domains inside the granules which made hydration even harder. However, due to the interaction between the MPT and stearic acid molecules, MPT existed in the amorphous form in the fatty acid matrix.