Hostname: page-component-7b7567c8c-wxvwp Total loading time: 0 Render date: 2026-11-16T06:09:31.442Z Has data issue: false hasContentIssue false

Multiphase modelling of precipitation-induced membrane formation

Published online by Cambridge University Press:  07 February 2020

P. S. Eastham*
Affiliation:
Department of Mathematics, Florida State University, Tallahassee, FL32304, USA
M. N. J. Moore
Affiliation:
Department of Mathematics, Florida State University, Tallahassee, FL32304, USA
N. G. Cogan
Affiliation:
Department of Mathematics, Florida State University, Tallahassee, FL32304, USA
Q. Wang
Affiliation:
Department of Chemistry and Biochemistry, Florida State University, Tallahassee, FL32306, USA
O. Steinbock
Affiliation:
Department of Chemistry and Biochemistry, Florida State University, Tallahassee, FL32306, USA
*
†Email address for correspondence: peastham@math.fsu.edu

Abstract

We formulate a model for the dynamic growth of a membrane developing in a flow as the result of a precipitation reaction, a situation inspired by recent microfluidic experiments. The precipitating solid introduces additional forces on the fluid and eventually forms a membrane that is fixed in the flow due to adhesion with a substrate. A key challenge is that, in general, the location of the immobile membrane is unknown a priori. To model this situation, we use a multiphase framework with fluid and membrane phases; the aqueous chemicals exist as scalar fields that react within the fluid to induce phase change. To verify that the model exhibits desired fluid–structure behaviours, we make simplifying assumptions to obtain a reduced form of the equations that is amenable to exact solution. This analysis demonstrates no-slip behaviour on the developing membrane without requiring fluid–membrane interface boundary conditions. The model has applications towards precipitate reactions where the precipitate greatly affects the surrounding flow, a situation appearing in many laboratory and geophysical contexts including the hydrothermal vent theory for the origin of life. More generally, this model can be used to address fluid–structure interaction problems that feature the dynamic generation of structures.

Information

Type
JFM Papers
Copyright
© The Author(s), 2020. Published by Cambridge University Press
Figure 0

Figure 1.Inorganic precipitate membrane formed in a microfluidic channel. (a) The photograph of the resulting membrane after 2 h with 0.5 M NaOH and 0.5 M NiCl2 solutions being injected simultaneously into the microfluidic device. The mixing part of the channel is 50 mm long, 2 mm wide and approximately $130~\unicode[STIX]{x03BC}\text{m}$ high. (b) A magnified view of a selected area from (a). (c–g) A sequence of micrographs showing the unidirectional thickening process after (c) 1 min, (d) 15 min, (e) 30 min, (f) 60 min and (g) 120 min. Scale bars correspond to (a) 1 cm, (b) 5 mm and (c) $200~\unicode[STIX]{x03BC}\text{m}$.

Figure 1

Figure 2.Schematic of precipitate reaction in control volume. Precipitation causes solution (white) to transform into membrane (shaded) after a certain concentration threshold is reached of aqueous product $C$. Aqueous chemicals $A$, $B$ and $C$ are volumeless scalar fields while the solvent and membrane are treated as a multiphase material. The volume of membrane gained is exactly equal to the volume of solvent lost.

Figure 2

Figure 3.Comparison of friction terms $\unicode[STIX]{x1D709}$. $\unicode[STIX]{x1D709}_{KC}$ (solid) is singular in the limit $\unicode[STIX]{x1D703}_{s}\rightarrow 0$, $\unicode[STIX]{x1D709}_{H}$ (dash) is non-singular in the porous limit ($K=0.5$, $n=2$) and $\unicode[STIX]{x1D709}_{B}$ (dot) yields maximum friction when both phases are present in equal amounts. All terms have been normalized by choosing $h$ such that $\unicode[STIX]{x1D709}(\unicode[STIX]{x1D703}_{S}^{\ast })=\unicode[STIX]{x1D709}^{\ast }$.

Figure 3

Figure 4.Dynamical system for $\unicode[STIX]{x1D713}_{C}$ and $\unicode[STIX]{x1D703}_{m}$. (a) Qualitative stability of $\dot{\unicode[STIX]{x1D713}}_{C}(\unicode[STIX]{x1D713}_{C})$ ODE. The left equilibrium $\unicode[STIX]{x1D713}_{C}^{-}$ is stable and exists for $\unicode[STIX]{x1D712}\geqslant 0$. (b) Visualization of planar dynamical system approaching the fixed point $(\unicode[STIX]{x1D713}_{C}^{-},1)$ for $\unicode[STIX]{x1D712}>0$. The thicker line corresponds to homogeneous initial conditions for $\unicode[STIX]{x1D713}_{C}$ and $\unicode[STIX]{x1D703}_{m}$. Shaded region is outside of the domain of $(\unicode[STIX]{x1D713}_{C},\unicode[STIX]{x1D703}_{m})\in [0,\infty )\times [0,1]$.

Figure 4

Figure 5.Developing membrane affects fluid flow. (a) Flow profile in a one-dimensional channel transitions from one-channel to two-channel flow. The reaction region is shaded. (b) Relevant variables evaluated in the reaction region at $x=L/2$, normalized for legibility; $\unicode[STIX]{x1D713}_{C}$ develops first, followed by $\unicode[STIX]{x1D703}_{m}$, which when large enough triggers the transition from one- to two-channel flow. The percolation threshold $\unicode[STIX]{x1D703}_{s}^{\ast }$ is set to $\unicode[STIX]{x1D703}_{s}^{\ast }=0.3$, which is why negligible change in fluid velocity is seen until $\unicode[STIX]{x1D703}_{m}\approx 0.7$. The specific $\unicode[STIX]{x1D703}_{s}^{\ast }$ was chosen based on the maximum volume fraction of an arrangement of packed spheres in three dimensions (Conway & Sloane 2013). The black diamond and square represent the system state inside the reaction region at $t=2.5$ and $t=5$, respectively.

Figure 5

Figure 6.Comparison of different models. (a–c) Use the stress in (2.32) with the friction term given by (a) Kozeny–Carman, (b) Hill function and (c) the ‘biofilm’ term; (d) uses a conventional multiphase stress $\unicode[STIX]{x1D64F}^{\prime }$ in (3.12) with $\unicode[STIX]{x1D709}_{B}$. The coefficients $h$ are scaled so as to make the three coefficients comparable in strength. The first two friction terms produce the desired no-slip on the membrane interface and the third, while affecting the fluid flow, does not generate the desired no-slip boundary condition. Additionally, the stress and friction combination used in (d), which is usually employed in multiphase models, does not yield the desired no-slip behaviour on the fluid–membrane interface.

Figure 6

Figure 7.Effect of increasing membrane thickness. As a percentage of the domain length, membrane width is (a) 5 %, (b) 15 % and (c) 33 %. The increasing maximum flow speed is due to the constant-flux constraint, and is analogous to that what would occur if the membrane boundaries and interface conditions were prescribed a priori in a single-phase flow.