Multiscale X-ray scattering microscopy maps nanoparticles in transitional flow
A new multiscale imaging method combining X-ray scattering and polarized light reveals how anisotropic nanoparticles influence transitional flow.
A new multiscale imaging technique has provided the first time-resolved view of how anisotropic nanoparticles influence transitional flow, bridging nanoscale orientation and macroscale vortex dynamics. The research combines small-angle X-ray scatter (SAXS) microscopy and polarized light imaging (PLI) to resolve nanoparticle orientation and flow structure across seven orders of magnitude in length and at millisecond timescales.
SAXS microscopy achieved its temporal resolution using frequency-domain analysis and the extreme X-ray flux from diffraction-limited synchrotron sources. The method maps the preferential orientation, shape, and size of nanoscale constituents within a Taylor-Couette flow cell, where the inner cylinder rotates and the outer cylinder is stationary. Polarized light imaging visualizes the macroscopic flow patterns by measuring the intrinsic flow-induced birefringence of the nanoparticle suspensions, avoiding tracer particles that could perturb the nanoscale flow. Results from the two techniques were verified to be quantitatively comparable.
For PLI, space-time diagrams were constructed from single-pixel lines along the flow axis, with Fourier analysis yielding spatial wavenumbers and temporal frequencies versus Reynolds number. For SAXS, two-dimensional scattering patterns were integrated to construct orientation-time diagrams. The Hermans orientation parameter served as an instantaneous, ensemble-averaged, local measure of nanoparticle preferential alignment. Scans were scaled to cover one and a half adjacent vortices.
Particle shape and concentration shift flow instability thresholds
The method was applied to study platelet-like graphene oxide nanoparticle suspensions at 0.7 wt% and rod-like cellulose nanocrystal suspensions at 3 wt%. The shape and concentration caused dramatic shifts in the flow instability thresholds.
| Nanoparticle | Concentration | Rotational Péclet Number (Pe) | Effect on Taylor Vortex Onset (Re_cr1) | Effect on Wavenumber (κ_TVF) |
|---|---|---|---|---|
| Graphene Oxide (platelet) | 0.7 wt% | ~10² | Lowers (destabilizes): Re_cr1^GO < Re_cr1^Newt ≈ 184 | Increases: κ_TVF^GO > κ_TVF^Newt |
| Cellulose Nanocrystal (rod) | 3 wt% | ~1 | Raises (stabilizes): Re_cr1^CNC > Re_cr1^Newt | Decreases: κ_TVF^CNC < κ_TVF^Newt |
Graphene oxide suspensions, with a high rotational Péclet number around 100, destabilize the flow. The platelets follow the macroscopic dynamics of secondary flows. In contrast, cellulose nanocrystal suspensions, with a moderate Péclet number around one, stabilize the flow. The rod-like particles exhibit characteristic high-frequency orientational motion that correlates with the turnover frequency of vortex instabilities. A simple kinematic model found this CNC orientational frequency consistent with the vortex turnover frequency.
Above the instability threshold, flow patterns for both suspensions remained similar to a Newtonian fluid, but the onset and range of Reynolds numbers differed significantly. The supercritical-flow rotational Péclet number-comparing vortex-wave timescale with particle orientational memory loss-is about 10 for GO and 0.1 for CNC. This explains why GO tracks the vortex wave while CNC loses orientational memory many times per wave cycle.
The Péclet-number interpretation alone cannot explain distinct nanoscale spectral peaks in CNC suspensions across different instabilities. The higher reduced number density of CNC implies stronger interparticle interactions than in GO, which could further influence orientation. In the simple model, experimentally estimated rotational diffusion suppressed the coherent CNC peak, suggesting a mismatch pointing to unmodeled interparticle interactions.
Bifurcation sequence and nanoscale response mapped across flow regimes
The laminar-turbulent transition in Taylor-Couette flow follows a defined sequence: laminar Couette flow (LCF) to Taylor vortex flow (TVF) to wavy vortex flow (WVF) and onward to chaotic and turbulent states. The LCF-TVF transition is a pitchfork bifurcation where axial symmetry breaks and periodic counter-rotating vortices form. The TVF to WVF transition is a Hopf bifurcation, manifesting as a travelling wave of vortices.
The rotational Péclet number measures the competition between Brownian rotational diffusion and flow-induced rotational advection. At large Pe_r, rotational advection dominates and particles align with the flow field. At small Pe_r, Brownian motion dominates and orientation is nearly isotropic. At moderate Pe_r around one, the interplay causes complex alignment dynamics.
In LCF, GO nanoplatelets align with surface normals orthogonal to the flow-vorticity plane, while CNC rods align with their director parallel to the velocity. In TVF, GO nanoplatelets align vertically in the vorticity direction within radial inflow and outflow regions, while inside vortices they show orientation variations. At TVF onset, the CNC nanoparticle director follows the macroscopic vortex in its toroidal trajectory.
The study concludes that the detailed role of multiscale behavior in flow stability and its relationship to classical effects like elasticity and shear thinning remains unresolved.
