Characterization of Suspension Networks Using Confocal Microscopy

In 2019, we started to investigate the microstructure and shear properties of capillary suspensions using confocal microscopy and rheology. The chosen model system was an index-matched mixture of Hexamoll DINCH/dodecane (oily bulk phase), aqueous glycerol (watery secondary liquid) and silica particles (Kromasil 100-7-SIL, average radius 3.21 m). Confocal measurements were performed on a Leica TCS SP8 inverted confocal microscope equipped with a linear shear cell (RheOptiCAD). Rheological measurements were performed on a stress-controlled MCR702 rheometer using an 8 mm parallel plate geometry. For the analysis of the confocal image stacks, a particle detection program based on Canny edge detection and Hough transform was written in IDL, which has an improved detection accuracy for concentrated and polydisperse suspensions compared to the classical Crocker and Grier algorithm [13].

The rst experiments were designed to examine the in uence of compression and shear on the microstructure of capillary suspensions with varying solid volume fraction, while keeping the relative ratio of secondary liquid and particles constant. The confocal microscopy study revealed an increase in coordination number with increasing particle volume fraction, compression, and shear. The clustering coe - cient only exhibited only slight variation. The increase in coordination number was also re ected in a higher storage modulus for higher solid volume fraction samples. By compressing capillary suspensions in the shear cell and the rheometer, it appeared that a transition in the relative cluster compaction could be observed around an e ective solid volume of 30 % to 35 %. The increase in coordination number was signi cantly higher after compression for samples above this boundary and a normal force was measured on the rheometer when going to the gap.

Figure 1: Close-up micrographs of porous Kromasil particles wetted by a) lms and b) patches of secondary liquid might suggest a critical volume fraction determining the change from a oppy to a rigid network. In the work of Domenech and Velankar [18], a change towards a more heterogeneous microstructure was reported around a similar solid volume fraction. However, the change in structure did not a ect the scaling of the elastic modulus or yield stress in their study.

While the transition between a oppy and rigid network appears to be compelling, the porosity of the Kromasil particles induced large variations in the wetting behaviour. Porous particles were chosen due to the need for fully-dyed spheres in the Crocker-Grier-based particle detection program. After dyeing the particles, their porosity was reduced using a modi ed St ober synthesis to prevent secondary liquid imbibition in the pores. The poor reproducibility of this procedure was the reason for a multitude of observed bridge shapes, as shown in Figure 1. The desired toroidal bridge shape was not observed for the prepared capillary suspensions, which predominantly showed the liquid lm behaviour of Figure 1a. Hence, these experiments were repeated in 2020 using non-porous particles showing toroidal bridges, shown on Figure 2, which was now possible due to the new particle detection program.