Role of solutal free convection on interdiffusion in a horizontal microfluidic channel

TL;DR

This paper investigates how buoyancy-driven free convection influences solute diffusion in microfluidic channels, combining numerical simulations and analytical solutions to identify regimes where convection impacts diffusion.

Contribution

It provides a comprehensive analysis of solutal free convection effects in microfluidics, deriving asymptotic solutions and dispersion coefficients, and clarifies when buoyancy effects can be neglected.

Findings

Different convection regimes identified based on Rayleigh number.

Analytical expressions for mixing zone width over time.

Transverse flow effects are negligible in most microfluidic setups.

Abstract

We theoretically investigate the role of solutal free convection on the diffusion of a buoyant solute at the microfluidic scales, $\simeq 5$--$500~\mu$m. We first consider a horizontal microfluidic slit, one half of which initially filled with a binary solution (solute and solvent), and the other half with pure solvent. The buoyant forces generate a gravity current that couples to the diffusion of the solute. We perform numerical resolutions of the 2D model describing the transport of the solute in the slit. This study allows us to highlight different regimes as a function of a single parameter, the Rayleigh number $\text{Ra}$ which compares gravity-induced advection to solute diffusion. We then derive asymptotic analytical solutions to quantify the width of the mixing zone as a function of time in each regime and establish a diagram that makes it possible to identify the range of $\text{Ra}$ and times for which buoyancy does not impact diffusion. In a second step, we present numerical resolutions of the same model but for a 3D microfluidic channel with a square cross-section. We observe the same regimes as in the 2D case, and focus on the dispersion regime at long time scales. We then derive the expression of the 1D dispersion coefficient for a channel with a rectangular section, and analyse the role of the transverse flow in the particular case of a square section. Finally, we show that the impact of this transverse flow on the solute transport can be neglected for most of the microfluidic experimental configurations.