Radajewski, D., Roblin, P., Bacchin, P., Meireles, M., & Hallez, Y. (2025). Microfluidic osmotic compression with operando meso-structure characterization using SAXS. Lab on a Chip, 25(12), 2851-2861. PDF
Contact: Yannick Hallez
We have developed a microfluidic chip for the osmotic compression of samples at the nanoliter scale, enabling the in situ and operando acquisition of structural features through small-angle X-ray scattering throughout the compression process. The design builds upon a previous setup allowing high-throughput measurements with minimal sample quantities. The updated design is specifically tailored for compatibility with a laboratory beamline, taking into account factors such as reduced photon flux and increased beam size compared to synchrotron beamlines. As a proof of concept, we performed on-chip compression of well-documented silica colloidal particles (Ludox TM-50). We demonstrated that the volume fraction could be tracked over time during compression, either by monitoring X-ray absorbance or by modeling the scattered signal. With precise control of the osmotic pressure and salt chemical potential, equations of state can be determined unambiguously from the volume fraction measurements and be interpreted with the help of the scattered intensity. These microfluidic chips will be valuable for understanding the behavior of colloidal suspensions, with applications in areas such as crystallization, nucleation, soil mechanics, control of living matter growth and interaction conditions, as well as the measurement of coarse-grained colloidal interaction potentials.
Radajewski, D., Roblin, P., Bacchin, P., Meireles, M. & Hallez, Y.. (2023). In situ structural analysis with a SAXS laboratory beamline on a microfluidic chip. Lab on a Chip. 23(14), 3280-3288. PDF
Contact: Yannick Hallez
In this work, we discuss how improvements of an X-ray laboratory beamline and an optimal design of a microfluidic device allow reliable structural information to be obtained without the need for a synchrotron. We evaluate the potential of these new developments by probing several well known dispersions. These include dense inorganic gold and silica nanoparticles that scatter photons quite intensely, the bovine serum albumin (BSA) macromolecule, with moderate contrast, to highlight possible applications in biology, and latex nanospheres with only weak contrast with the solvent to show the limits of the setup. We established a proof of concept for a versatile setup that will open the way for more complex lab-on-a-chip devices suitable for in situ and operando structural analysis by small angle X-ray scattering analysis without the necessity for a synchrotron source.
We showed, using a model coupling mass transport and liquid theory calculations for a charge-stabilized colloidal dispersion, that diffusion significantly limits measurement times of its equation of state (EOS), osmotic pressure vs composition, using the osmotic compression technique. Following this result, we presented a microfluidic chip allowing one to measure the entire EOS of a charged dispersion at the nanoliter scale in a few hours. We also show that time-resolved analyses of relaxation to equilibrium in this microfluidic experiment lead to direct estimates of the collective diffusion coefficient of the dispersion in Donnan equilibrium with a salt reservoir.
Keïta, C., Hallez Y. & Salmon, J.B. (2021). Microfluidic osmotic compression of a charge-stabilized colloidal dispersion:
Equation of state and collective diffusion coefficient. Phys. Rev. E, 104, L062601. PDF
Contact: Yannick Hallez
Emulsion stabilization is most often achieved using simple surfactants, proteins, or particles. We investigated a complex natural stabilizer, gum arabic, widely used in food formulations and known for its remarkable metastability even at high dilution. Gum arabic is a heterogeneous mixture of partially conjugated protein–polysaccharide complexes, spanning a continuum of architectures revealed by chromatography and small-angle scattering. In solution, this diversity leads to hierarchical self-organization controlled by concentration and electrostatic repulsions.
At the oil–water interface, selective partitioning occurs: protein-rich species adsorb irreversibly and dominate the interfacial film. While film composition is weakly dependent on formulation parameters, interfacial density is strongly controlled by ionic repulsions and directly correlates with emulsion metastability. At high coverage, the adsorbed species aggregate into a percolated two-dimensional interfacial network. The interface can thus be described as a hydrophobic protein network anchored at the interface and coupled to a dense, water-swollen polysaccharide brush. This architecture explains the exceptional robustness of gum-arabic-stabilized emulsions and enabled, in practice, a tenfold reduction in stabilizer content.
Emulsions used in parenteral nutrition rely on phospholipid-stabilized nanoemulsions, selected for their biocompatibility. This choice is intrinsically paradoxical: phospholipids naturally favor near-zero curvature assemblies, such as lamellar phases, which are poorly suited to stabilizing highly curved oil droplets. In practice, small compositional changes often lead to catastrophic losses of metastability.
We addressed this problem using model lipid/oil systems and a combination of complementary approaches. Ternary phase diagrams were used to quantify the spontaneous curvature of amphiphilic films, neutron scattering to probe interfacial structure, and separation coupled to Raman spectroscopy to determine the effective localization of phospholipids. Emulsion metastability was quantified through kinetic measurements. Our results show that spontaneous curvature controls metastability at leading order, but that its precise value is critical. A slightly positive curvature is sufficient to stabilize nanodroplets via an adsorbed monolayer, whereas macrodroplets require interfacial multilayers. Conversely, excessively positive curvature promotes vesicle formation in the aqueous phase and depletes the interface. This work identifies a narrow formulation optimum, explaining the extreme industrial sensitivity while providing a rational framework for formulation design.