Powder Technology, 2019, vol 355pp. 738-754
The complex multiscale structure of spray dried detergent granules has been characterized using a complementary range of techniques including scanning electron microscopy (SEM), image analysis, wide angle and small angle X-ray scattering (WAXS, SAXS) and X-ray microtomography. A complex composite structure is revealed due to the multiphase nature of the starting slurry and its evolution during the drying process. Four simple model formulations based on linear alkylbenzene sulphonate (NaLAS) and sodium sulphate were used to probe the influence of initial slurry water content and binder, sodium silicate (SiO_2:nNa_2 O), on the structure. The structure revealed can be viewed as a continuous matrix consisting of NaLAS, sodium sulphate and binder in which large, dense, crystals of sodium sulphate are embedded. These large crystals were initially undissolved in the slurry and are consequently reduced in number in the product made from higher water content slurry. Air is also dispersed in this matrix at two length scales, large vacuoles, typically larger than 0.1x the particle diameter driven by ‘puffing’ i.e. water boiling, and micro-scale porosity which evolved during the crystallisation of the drying matrix. The matrix is formed from the dried slurry liquid phases and is a composite structure of binder, dried surfactant liquid crystalline phase and sub-micron scale crystallites of sodium sulphate. The liquid phase compositions change with the slurry water content, as well as the addition of binder; this changes both the sulphate polymorphs present and the spacing of the liquid crystalline lamellae, i.e. d-spacing, of the NaLAS phase. The composition and crystalline structure of the surface is also affected. The particle morphology can also be significantly affected with the high initial moisture content particle having a distinct agglomerated and blistered structure. The multiscale influence of formulation on particle structure, and consequently product properties and performance, highlights the challenge in predictive design of these important detergent products.