Granulation of hydrophobic particles: production of hollow granules from liquid marbles

This thesis investigates granulation of hydrophobic powders. Recently discovered hollow granules formed via spreading of hydrophobic particles around a liquid droplet appear promising to solve the problematic wetting behavior of hydrophobic powders. This new way of granulating hydrophobic powders has not been previously investigated. This thesis focuses on the nucleation stage of hydrophobic powders and investigates the spreading mechanism of particles around a liquid droplet, which is known as liquid marble. The formation of hollow granules from liquid marbles, and the mass production of hollow granules in a mixer granulator are also a main point in this thesis. An entirely physical flow mechanism for the nucleation stage of liquid marbles was proposed instead of the spreading coefficient theory. Experimental work for testing this mechanism was carried out by studying the effect of binder and powder properties as well as the effect of kinetic energy of impact on powder coverage of the drop. Based on experimental results, a proportional relationship was found between increasing kinetic energy and the percentage of liquid marble coverage. Moreover, calculation of the solid-liquid spreading coefficient and thermodynamic analysis of the existing solid-liquid spreading coefficient revealed that the existing solid-liquid spreading coefficient theory is not a predictive approach. This analysis also pointed out some serious flaws in the assumptions used in definition of the solid-liquid spreading coefficient (Rowe, 1989) in terms of simply exchanging the solid-liquid subscript in liquid-solid spreading coefficient equation, as the thermodynamic conditions taking place during spreading solid over a liquid are quite different to spreading liquid over a solid substrate. Drying of liquid marbles and the formation of hollow granules was investigated in small scale experiments by employing several types of binder with different concentrations as well as various types of powder with different particle sizes at two selected drying temperatures. Higher drying temperature, smaller or nano-sized particles and higher binder concentration tend to promote the formation of perfect hollow granules. It was found that the survival rate was directly proportional to binder viscosity for HPMC and PVP. However for HPC binder, the survival rate was essentially constant regardless of HPC concentration due to precipitation of the HPC binder above the cloud point temperature. In the next step, the insight gained in small scale experiments helped to successfully produce hollow granules in large scale experiments using a 4 L laboratory scale granulator. The optimum liquid to solid ratio (L:S) ratio for Aerosil and 5% HPC was found by studying the effect of different liquid to solid ratio on the morphology of formed granules. At the optimum L:S ratio, more of the raw fine particles were granulated and fewer flattened or stretched hollow granules were produced. In addition, the experimental observations of different granule morphologies revealed the importance of performing X-ray tomography during hollow granule formulation. If the nucleation process starts with a preformed template droplet (e.g. spraying liquid), the final granule size increases as the L:S ratio increases as well as the amount of un-granulated fine particles decreases. Finally, a framework for liquid marble formation via solid spreading nucleation (droplet template) and via mechanical dispersion was proposed by outlining the sequential steps and possible controlling groups for each step. The first step of the framework via droplet template starts with droplet diameter (dd )> 25 particle diameter (dp) then it follows with Bo<1 and We<1000 in order to produce a spherical cap without shattering on powder bed. The next step requirement is that the contact angle of the liquid on the powder must be above 90º. If spray flux is low enough (Ψa<<1, the dimension spray flux (Ψa=3 /2 dd ) is a measure of spray density and drop overlap in the spray zone where (m3/s) is a volumetric spray flow rate, dd is an average drop size, and (m2/s) is powder flux through wetted spray area) and there is enough kinetic energy, an individual spherical liquid marble will be formed, providing that the energetic contribution Ep of system is within the acceptable range. The main difference between droplet template and mechanical dispersion frameworks are contact angle and spray flux requirement. When spray flux becomes greater than one (Ψa>1), granulation switches from droplet template regime to mechanical dispersion regime. Liquid marble can form successfully in mechanical dispersion regime if the contact angle is above 110 º. A framework for hollow granule formation from liquid marbles was also developed. In this framework, the importance of presence of compatible polymeric binder at the optimum concentration was emphasized as the first requirement for hollow granule formation. The optimum L:S ratio and drying temperature are additional criteria for a successful formation of single cavity hollow granule. These frameworks give a better understanding of the controlling mechanisms in granulation of hydrophobic particles and provide guidance on how to mass produce liquid marbles and hollow granules.