betaFeedback

WorksPhysics and Astronomy

Profile-Preserving Phase-Field Model for Surfactant Transport and Adsorption-Desorption in Two-Phase Flow Systems

Abstract

The diffuse-interface model for two-phase flows with soluble surfactants has garnered considerable attention due to its ability to circumvent the need for Robin boundary condition in the bulk surfactant transport equation. However, the coupling between surfactant concentration and the phase field within this framework underscores the importance of accurately resolving interfacial equilibrium profiles. To address this limitation, we have developed a profile-preserving phase-field model for simulating surfactant transport and adsorption-desorption in two-phase flow systems. This approach iteratively refines interfacial profiles and delta functions, removing concentration singularities and improving mass conservation. The effectiveness of the model is demonstrated through two benchmark simulations: surfactant transport in a vortex-deformed droplet, which quantitatively reveals reduced mass error over time, and adsorption-desorption dynamics on a stationary spherical interface, showing strong agreement with one-dimensional analytical solutions for surfactant concentration distributions. We further highlight the model's capability by simulating the settling behavior of a surfactant-laden droplet, underscoring the critical role of adsorption-desorption kinetics in governing droplet dynamics.

Introduction

Surfactant molecules, which possess both hydrophobic and hydrophilic components, readily adsorb at fluid interfaces – especially those involving water. This adsorption changes the energy required to expand the interface (i.e., the interfacial surface tension), thereby exerting an inevitable influence on surface tension-driven dynamics and interfacial flows [1, 2, 3, 4, 5, 6, 7].

Because surfactants are widely used across various scientific and engineering disciplines [8, 9, 10, 11, 12, 13, 14, 15, 16], understanding their behavior in complex fluid systems is crucial. However, the mass transport of surfactants—including interfacial convection and diffusion, bulk convection and diffusion, as well as adsorption–desorption between the interface and the bulk—makes it challenging to experimentally measure surfactant concentrations at both the interface and within the bulk.

To address these experimental limitations, various numerical techniques have been developed for modeling two-phase flows with surfactants. These methods include the boundary integral method [2], the front-tracking method [17, 18, 19, 20], the immersed boundary method (IBM) [21], the arbitrary Lagrangian-Eulerian finite element method (ALE-FEM) [22], the smoothed particle hydrodynamics method (SPH) [23], the volume of fluid method (VOF) [24, 25], the level-set method (LS) [26, 27, 28], and the phase field method (PF) [29, 30, 31]. Additionally, hybrid approaches – such as combining the level-set and front-tracking methods [32] or integrating the immersed boundary and immersed interface methods [33] – have also been proposed. Among these, the diffuse-interface method has attracted particular attention for its simplicity and its success in modeling both insoluble and soluble surfactants, especially in two-phase flows with complex interface topologies [34, 35, 36, 37, 38, 39, 40, 41].

In this diffuse-interface framework, surface quantities can be directly captured on a fixed grid. Specifically, instead of solving the interfacial surfactant transport equation (Eq. (A.1)) for the surfactant concentration Γ\Gamma on the deforming interface SS and the bulk surfactant transport equation (Eq. (A.2)) for the concentration CC within the evolving and irregular domain Ω2\Omega_2 (Figure 1 (a)), the corresponding diffuse-interface formulations (Eqs. (A.4) and (A.5)) are solved over the entire domain Ω\Omega [28, 34, 35]. This is achieved by introducing Γ^=Γδ\hat{\Gamma} = \Gamma\delta and C^=Cϕ\hat{C} = C\phi, where δ\delta is a regularized delta function and ϕ\phi is the phase field function (Figure 1 (b)). These models have been successfully applied to simulating interfacial flows with solute surfactants, further enhancing the method’s versatility [30, 31, 36, 37, 42, 43, 44]. Recent developments, such as the immiscible-scalars model and interface-confined model, have improved the robustness and stability of the diffuse interface method [36, 37, 44].

Image not extracted

Figure 1. (a) Schematic of interfacial and bulk surfactant concentrations Γ\Gamma and CC. The white wine gradient color at the interface SS denotes the interfacial surfactant concentration, while the white blue gradient color in Ω2\Omega_2 (phase 2) represents the bulk surfactant concentration. (b) Profiles of phase field function ϕ\phi, surface delta function δ\delta, interfacial and bulk surfactant concentrations Γ\Gamma and CC, and their diffuse-interface forms Γ^\hat{\Gamma} and C^\hat{C} in one dimension. For simplicity, we assume the values of Γ\Gamma and CC are constant in the normal direction of the interface.

The diffuse-interface framework provides two notable advantages. First, this framework eliminates the need to enforce the Robin boundary condition (Eq. (A.3)) in a dynamically evolving, irregular domain when solving the bulk surfactant transport equation. Second, because Γ^\hat{\Gamma} is defined as an Eulerian variable across the entire domain Ω\Omega, it naturally handles complex topological changes of the interface. However, a key assumption underlying this method is that the phase field function ϕ\phi remains in equilibrium with the interfacial profile. The interplay between convection and diffusion terms, along with numerical diffusion from discretization errors in the interface-capturing equation (i.e., advective Cahn-Hilliard and Allen-Cahn equations), can cause the deviation from the equilibrium state of the interface profile in practical simulations [45, 46, 47, 48]. The out-of-equilibrium profile of ϕ\phi can cause singularities and mass non-conservation of both interfacial and bulk surfactant concentrations, as demonstrated in § 4.1. These deviations can result in inaccuracies in surface tension calculations and potentially lead to unphysical numerical outcomes.

We have recently developed an enhanced profile-preserving phase-field model, integrated with a hybrid Free-energy/Continuum surface force model (FECSF), to simulate two-phase flows involving insoluble surfactants [49]. By restoring the equilibrium interface profile at each time step [46], our model ensures the assumption of interfacial profile equilibrium is consistently satisfied. This strategy effectively prevents surfactant concentration singularities and enhances the accuracy of surface tension calculations and surfactant mass conservation, even when dealing with coarse mesh resolutions.

In this work, we extend our numerical model to address the transport of soluble surfactants, which is critical for a wide range of applications [5, 6, 7, 8, 11, 16]. This is achieved by integrating both interfacial and bulk surfactant transport equations within the profile-preserving Cahn-Hilliard framework. The surfactant adsorption and desorption processes are modeled using the Langmuir isotherm. To calculate the surface tension forces while minimizing spurious currents, we employ the hybrid FECSF model [49]. To mitigate time-step constraints, a semi-implicit scheme is adopted for solving both interfacial and bulk surfactant transport equations, alongside the finite volume method for conservative spatial discretization.

As a case study, we investigate the impact of surfactant solubility in the bulk phase on droplet dynamics [20, 50, 51, 52, 53, 54], with a particular focus on surfactants dissolved within droplets – a process that has received limited attention. Specifically, we examine the effects of the Biot and Damköhler number numbers on the settling behavior of a surfactant-laden droplet, along with the resulting impact on surfactant distribution and Marangoni stress at the interface.

The structure of this paper is as follows. § 2 briefly introduces the numerical model. § 3 describes the numerical methods for solving the interfacial and bulk surfactant transport equations, along with the solution procedure. § 4 provides two test cases to evaluate the accuracy of the numerical algorithm and model performance. In § 5, we present a demonstration with a particular emphasis on the influence of the Marangoni effects. Finally, we conclude the paper in § 6.

Methodology

Modeling Surfactant Transport with Adsorption and Desorption

The transport of soluble surfactants, which undergo an adsorption-desorption process, is modeled by considering both the interfacial concentration Γ(x,t)\Gamma(\mathbf{x},t) and the bulk concentration C(x,t)C(\mathbf{x},t). To frame the problem within a dimensionless context, we nondimensionalize the concentrations by the critical micelle concentration CCMC∗C_{\mathrm{CMC}}^{*} for the bulk and the corresponding saturation interfacial concentration Γ∞∗\Gamma_{\infty}^{*} for the interface, with ∗* denoting dimensional quantities.

The evolution of interfacial surfactant concentration is described by a transport equation on a deforming interface [3, 55], coupled with the fluid velocity field u\mathbf{u}. To get its diffuse-interface formulation, we follow the approach proposed by [28, 34, 35]. Specifically, we employ Γ^=∣∇ϕ∣Γ\hat{\Gamma} = \lvert\nabla\phi\rvert\Gamma (Figure 1b), where ϕ\phi is the phase field function tracking the evolving interface [49]. Within this framework, the interfacial surfactant transport equation, incorporating the Langmuir adsorption isotherm [56], is expressed as

∂Γ^∂t+∇⋅(uΓ^)=1PeΓ∇⋅[∇Γ^−(1−2ϕ)Γ^2Cn∇ϕ∣∇ϕ∣]+∣∇ϕ∣j(1)\frac{\partial\hat{\Gamma}}{\partial t} + \nabla\cdot\left(\mathbf{u}\hat{\Gamma}\right) = \frac{1}{Pe_{\Gamma}} \nabla\cdot\left[\nabla\hat{\Gamma} - \frac{(1-2\phi)\hat{\Gamma}}{\sqrt{2}C_{n}}\frac{\nabla\phi}{\lvert\nabla\phi\rvert}\right] + \lvert\nabla\phi\rvert j \tag*{(1)}

where the net flux term jj represents the mass transfer between the interface and the bulk, and is given by the relation,

j(Γ,Cs)=Bi(kCs(1−Γ)−Γ),(2)j(\Gamma,C_s)=Bi\left(kC_s(1-\Gamma)-\Gamma\right), \tag*{(2)}

with CsC_s representing the surfactant concentration in the sublayer of the bulk phase. The Biot number Bi=kdL/UBi=k_dL/U characterizes the ratio of the interfacial convection time scale L/UL/U to the desorption time scale kd−1k_d^{-1} with the characteristic length LL and the characteristic velocity UU of the system, while the parameter k=kaCCMC∗/kdk=k_aC^*_{CMC}/k_d denotes the ratio of the desorption time scale kd−1k_d^{-1} to the adsorption time scale (kaCCMC∗)−1(k_aC^*_{CMC})^{-1}. At the initial time (t=0)(t=0), we assume that surfactants at the interface and in the sublayer are in equilibrium, implying a zero net flux (j=0)(j=0). This equilibrium condition yields the relationship

Γ0=kC01+kC0,(3)\Gamma_0=\frac{kC_0}{1+kC_0}, \tag*{(3)}

where the subscript 0 denotes the equilibrium value.

The bulk surfactant concentration C(x,t)C(\mathbf{x},t) evolves according to the advection-diffusion equation, which is solved in an evolving and irregular domain. To avoid explicitly tracking domain boundaries or implementing Robin boundary conditions (as detailed in Appendix § 8), we use the diffuse-interface formulation. Specifically, we define C^=Cϕ\hat{C}=C\phi (Figure 1b), following the method in [35, 36]. Thus, the dimensionless advection–diffusion equation is

∂C^∂t+∇⋅(uC^)=1PeC∇⋅[∇C^−(1−ϕ)C^∇ϕ2Cn∣∇ϕ∣]−Da∣∇ϕ∣j,(4)\frac{\partial\hat{C}}{\partial t}+\nabla\cdot(\mathbf{u}\hat{C})=\frac{1}{Pe_C}\nabla\cdot\left[\nabla\hat{C}-\frac{(1-\phi)\hat{C}\nabla\phi}{\sqrt{2}Cn\lvert\nabla\phi\rvert}\right]-Da\lvert\nabla\phi\rvert j, \tag*{(4)}

where the Damköhler number Da=Γ∞∗/(CCMC∗L)Da=\Gamma^*_{\infty}/(C^*_{CMC}L) characterizes the ratio of the thickness of the depleted region Γ∞∗/(CCMC∗)\Gamma^*_{\infty}/(C^*_{CMC}) to the characteristic length LL.

Therefore, we have equations (1) and (4) describing surfactant transport both at the interface and in the bulk. Ensuring equilibrium in the interfacial profile (Figure 1b) is essential for accurate surface tension calculations and model stability, as the diffuse-interface formulations (Eqns. (1) and (4)) are inherently dependent on this equilibrium.

Interface Dynamics and Flow Modeling with the Profile-Preserving Method

The surfactant transport equations (Eqns. (1) and (4)) are integrated with governing flow dynamics by utilizing the profile-preserving phase-field method. The governing equations that apply the profile-preserving phase-field method [46, 49] are as follows

∂ϕ∂t+∇⋅(ϕu)=1Peϕ∇2η,(5)\frac{\partial\phi}{\partial t}+\nabla\cdot(\phi\mathbf{u})=\frac{1}{Pe_{\phi}}\nabla^{2}\eta, \tag*{(5)}
∂ϕ∂τ=∇⋅{[2Cn(∇ϕ⋅n(ψ))−14(1−tanh⁡2(ψ22Cn))]n(ψ)},(6)\frac{\partial\phi}{\partial\tau}=\nabla\cdot\left\{\left[\sqrt{2}Cn\left(\nabla\phi\cdot\mathbf{n}(\psi)\right)-\frac{1}{4}\left(1-\tanh^{2}\left(\frac{\psi}{2\sqrt{2}Cn}\right)\right)\right]\mathbf{n}(\psi)\right\}, \tag*{(6)}
ρ(ϕ)(∂u∂t+u⋅∇u)=−∇p+1Re{∇⋅[μ(ϕ)(∇u+∇uT)]+FstCa}+ρ(ϕ)Fre,(7)\rho(\phi)\left(\frac{\partial\mathbf{u}}{\partial t}+\mathbf{u}\cdot\nabla\mathbf{u}\right)=-\nabla p+\frac{1}{Re}\left\{\nabla\cdot\left[\mu(\phi)\left(\nabla\mathbf{u}+\nabla\mathbf{u}^{\mathrm{T}}\right)\right]+\frac{\mathbf{F}_{\mathrm{st}}}{Ca}\right\}+\frac{\rho(\phi)}{Fr}\mathbf{e}, \tag*{(7)}
∇⋅u=0,(8)\nabla\cdot\mathbf{u}=0, \tag*{(8)}
FstCa=62η∇ϕCaCn[1+Eln⁡(1−Γ)]+Ma∇sln⁡(1−Γ)∣∇ϕ∣.(9)\frac{\mathbf{F}_{\mathrm{st}}}{Ca}=\frac{6\sqrt{2}\eta\nabla\phi}{CaCn}\left[1+E\ln(1-\Gamma)\right]+Ma\nabla_{s}\ln(1-\Gamma)\lvert\nabla\phi\rvert. \tag*{(9)}

To capture the interfacial evolution, we apply the advective Cahn-Hilliard equation (Eqn. (5)) [29, 45]. The equilibrium interfacial profile is derived by minimizing the interfacial free energy [57], yielding the equilibrium form ϕe(x,t)=12{1+tanh⁡[ψ(x,t)/(22Cn)]}\phi^{e}(\mathbf{x},t)=\frac{1}{2}\left\{1+\tanh\left[\psi(\mathbf{x},t)/(2\sqrt{2}Cn)\right]\right\} for the order parameter ϕ(x,t)\phi(\mathbf{x},t), where ψ(x,t)\psi(\mathbf{x},t) is the signed-distance function [29, 45]. To maintain equilibrium in the interfacial profile, the profile-preserving phase field equation (Eqn. (6)) is solved iteratively until ϕ(x,τ)\phi(\mathbf{x},\tau) reaches a steady state, with τ\tau representing the iteration time.

The flow field u(x,t)\mathbf{u}(\mathbf{x},t) in the two-phase system, which include contributions from surface tension and gravity, is governed by the dimensionless Navier–Stokes equations (Eqns. (7) and (8)).

We assume a dilute surfactant concentration, such that the variation in surfactant concentration does not affect other fluid properties, except for surface tension. The surface tension effect, incorporating Marangoni effects, is modeled using the FECSF model (Eqn. (9)).

The involved dimensionless parameters include Péclet number PeϕPe_{\phi}, Cahn number CnCn, Reynolds number ReRe, Capillary number CaCa, Froude number FrFr, Elasticity number EE, and Marangoni number MaMa. A comprehensive description of the model can be found in 2.2, and the detailed derivation is available in our previous work [49].

Numerical Discretization and Solution Procedure

Numerical Discretization

We employ a finite-volume method on a uniform rectangular staggered mesh in space and a semi-implicit strategy to discretize governing equations in time. Specifically, the scalar fields (pp, μ\mu, ρ\rho, ϕ\phi, ψ\psi, Γ^\hat{\Gamma}, Γ\Gamma, C^\hat{C}, and CC) are defined at cell centers, the vector fields (the velocity u\mathbf{u} and the surface tension Fst\mathbf{F}_{\mathrm{st}}) are defined at cell faces, and the interface normal n\mathbf{n} is defined at cell vertices. We refer the readers to [49] for more details of the numerical scheme used to solve the advective Cahn-Hilliard equation (Eq. (5)), the profile-preserving equation (Eq. (6)), and the Navier-Stokes equation (Eqs. (8) and (7)).

We implement a semi-implicit scheme [58] for the temporal discretization of the interfacial and bulk surfactant transport equations. In this scheme, the diffusion terms (i.e., ∇2Γ^\nabla^2\hat{\Gamma} and ∇2C^\nabla^2\hat{C}) are discretized using a Crank-Nicolson scheme, while the other terms are discretized using a second-order Adams-Bashforth scheme. The discretized form of interfacial surfactant transport equation is formulated as follows

3Γ^n+1−4Γ^n+Γ^n−12Δt=−[32W(Γ^n,un,ϕn+1)−12W(Γ^n−1,un−1,ϕn)]+12PeΓ[(∇2Γ^)n+1+(∇2Γ^)n],(10)\frac{3\hat{\Gamma}^{n+1}-4\hat{\Gamma}^{n}+\hat{\Gamma}^{n-1}}{2\Delta t} = -\left[ \frac{3}{2}\mathcal{W}(\hat{\Gamma}^{n},\mathbf{u}^{n},\phi^{n+1}) -\frac{1}{2}\mathcal{W}(\hat{\Gamma}^{n-1},\mathbf{u}^{n-1},\phi^{n}) \right] +\frac{1}{2Pe_{\Gamma}}\left[ (\nabla^2\hat{\Gamma})^{n+1}+(\nabla^2\hat{\Gamma})^{n} \right], \tag*{(10)}

where

W(Γ^n,un,ϕn+1)=∇⋅(uΓ^)n+1PeΓ∇⋅{Γ^n[(1−2ϕ)∇ϕ2Cη∣∇ϕ∣]n+1}−∣∇ϕ∣n+1jn,(11)\mathcal{W}(\hat{\Gamma}^{n},\mathbf{u}^{n},\phi^{n+1}) = \nabla\cdot(\mathbf{u}\hat{\Gamma})^{n} +\frac{1}{Pe_{\Gamma}}\nabla\cdot\left\{ \hat{\Gamma}^{n} \left[ \frac{(1-2\phi)\nabla\phi}{\sqrt{2}C\eta|\nabla\phi|} \right]^{n+1} \right\} -|\nabla\phi|^{n+1}j^{n}, \tag*{(11)}

and Δt\Delta t is the time step. Similar to Eq. (10), the discretized form of bulk surfactant transport equation can be expressed as

3C^n+1−4C^n+C^n−12Δt=−[32G(C^n,un,ϕn+1)−12G(C^n−1,un−1,ϕn)]+12PeC[(∇2C^)n+1+(∇2C^)n],(12)\frac{3\hat{C}^{n+1}-4\hat{C}^{n}+\hat{C}^{n-1}}{2\Delta t} = -\left[ \frac{3}{2}\mathcal{G}(\hat{C}^{n},\mathbf{u}^{n},\phi^{n+1}) -\frac{1}{2}\mathcal{G}(\hat{C}^{n-1},\mathbf{u}^{n-1},\phi^{n}) \right] +\frac{1}{2Pe_{C}}\left[ (\nabla^2\hat{C})^{n+1}+(\nabla^2\hat{C})^{n} \right], \tag*{(12)}

where

G(C^n,un,ϕn+1)=∇⋅(uC^)n+1PeC∇⋅{C^n[(1−ϕ)2Cn∇ϕ∣∇ϕ∣]n+1}−Da∣∇ϕ∣n+1jn.(13)\mathcal{G}(\widehat{C}^{n},\mathbf{u}^{n},\phi^{n+1})=\nabla\cdot\left(\mathbf{u}\widehat{C}\right)^{n}+\frac{1}{Pe_{C}}\nabla\cdot\left\{\widehat{C}^{n}\left[\frac{(1-\phi)}{\sqrt{2C_{n}}}\frac{\nabla\phi}{\lvert\nabla\phi\rvert}\right]^{n+1}\right\}-Da\lvert\nabla\phi\rvert^{n+1}j^{n}. \tag*{(13)}

To evaluate fluxes at the cell faces for the advection terms in Eqs. (10) and (12), we employ a fifth-order weighted essentially non-oscillatory (WENO) scheme [59] with the local flow velocity determining the upwind direction. For other terms in these two equations, we apply a second-order central difference scheme to approximate the gradient operator ∇\nabla, and a second-order linear interpolation scheme to approximate the values of variable at cell face.

Solution Procedure

The comprehensive solution procedure for each time step loop can be summarized as follows

  1. Advance the phase field function ϕ(x,t)\phi(\mathbf{x},t) using Eqn. (5) after initialization.

  2. Correct the profile of the phase field ϕ(x,t)\phi(\mathbf{x},t) using the profile-preserving equation Eqn. (6) and update the phase field function ϕ(x,t)\phi(\mathbf{x},t).

  3. Compute the mass transfer term j(Cs,Γ)j(C_{s},\Gamma) using Eqn. (2).

  4. Solve interfacial surfactant transport equation Eqn. (10) and compute the interfacial surfactant concentration Γ(x,t)\Gamma(\mathbf{x},t).

  5. Solve bulk surfactant transport equation Eqn. (12) and compute the bulk surfactant concentration C(x,t)C(\mathbf{x},t).

  6. Compute the capillary force and Marangoni force using Eqn. (9).

  7. Solve Navier-Stokes equation Eqn. (10) and obtain the velocity u(x,t)\mathbf{u}(\mathbf{x},t) and pressure p(x,t)p(\mathbf{x},t) by solving Eqns. (8) and (7).

  8. Update above functions.

Numerical Tests

This section presents two numerical tests to evaluate the model’s performance and algorithm accuracy. The first test examines a drop with soluble surfactant in a single vortex, comparing interfacial and bulk surfactant profiles with and without iterative correction of the interface profile to equilibrium. Following the study [18, 34], we then simulate surfactant mass transfer between the interface and bulk on a stationary spherical drop.

Transport of surfactants in a drop deformed in a single vortex

We examine the effect of preserving the equilibrium of interface profile on surfactant distributions in a drop deformed by a single vortex. Surfactant exchange between the interface and bulk is neglected. The vortex flow is prescribed as u=−sin⁡2(πx/L)sin⁡(2πy/L)cos⁡(πt/T)u=-\sin^{2}(\pi x/L)\sin(2\pi y/L)\cos(\pi t/T) and v=sin⁡(2πx/L)sin⁡2(πy/L)cos⁡(πt/T)v=\sin(2\pi x/L)\sin^{2}(\pi y/L)\cos(\pi t/T), with L=1L=1 and T=4T=4. A drop of radius R=0.15R=0.15 is initially placed at (0.5,0.75)(0.5,0.75) in a unit square domain. The dimensionless parameters are Δt=4×10−4\Delta t=4\times10^{-4}, L/Δx=256L/\Delta x=256, Cn=0.75ΔxCn=0.75\Delta x, Γ0=0.5\Gamma_{0}=0.5, C0=0.5C_{0}=0.5, Peϕ=1/CnPe_{\phi}=1/Cn, PeΓ=100Pe_{\Gamma}=100, and PeC=100Pe_{C}=100.

Figure 2(aFigure 1-(cFigure 1 show the phase field ϕ\phi, interfacial surfactant concentration Γ^\hat{\Gamma}, and bulk concentration C^\hat{C} after one full vortex rotation period (t=Tt=T) without profile correction, while Figure 2(aFigure 2-(cFigure 2 show corrected results based on Eq. (6). Profile correction maintains a drop shape closer to the initial circle by minimizing deviations of ϕ\phi from equilibrium [46]. Without correction, singularities appear in Γ^\hat{\Gamma} and C^\hat{C} near the interface, as highlighted along the blue dashed line (ss) in the inset of Figure 2(bFigure 1. Applying profile correction eliminates these singularities and confines the surfactant distributions to their respective regions, restoring the expected profiles.

Image not extracted

Figure 2. Surfactant transport in a vortex-deformed drop. Profiles of (a) ϕ\phi and ∣∇ϕ∣|\nabla\phi|, (b) Γ\Gamma and Γ^\widehat{\Gamma}, and (c) CC and C^\widehat{C} after one full vortex rotation period (t=Tt=T). Insets show the phase field ϕ\phi, interfacial concentration Γ^\widehat{\Gamma}, and bulk concentration field C^\widehat{C}. For comparison, results without profile correction (a1-c1) and with correction (a2-c2) are plotted along the blue dashed line (ss) indicated in the inset of (b1). (d) and (e) show the temporal evolution of profiles without correction over the interval t/T∈[0.95,1]t/T \in[0.95,1]. (f) show the time evolution of mass errors of interfacial and bulk surfactant concentrations (Em,ΓE_{m,\Gamma} in orange and Em,CE_{m,C} in blue) with profile correction (solid lines) and without correction (dashed lines).

The origin of these singularities is analyzed via the interfacial and bulk surfactant transport equations (Eq. (1) and Eq. (4)). Under equilibrium with zero velocity, the steady-state solutions are Γ^(x)=Γ0∣∇ϕ(x)∣\hat{\Gamma}(x)=\Gamma_{0}\lvert\nabla\phi(x)\rvert [37] and C^(x)=C0ϕ(x)\hat{C}(x)=C_{0}\phi(x) [36], indicating that accurate surfactant profiles rely on preserving phase field equilibrium. However, as shown in Figure 2(aFigure 1, discretization errors disturb the interface, creating spurious gradients that drive artificial surfactant accumulation (Figure 2(d) and (Figure 2e)) and singularity formation. These inaccuracies distort Γ^\hat{\Gamma} and C^\hat{C} (black lines in Figure 2(bFigure 1 and (cFigure 1), further leading to numerical errors in surfactant mass transfer (Eq. (2)) and surfactant tension forces (Eq. (9)). Profile correction effectively suppresses these errors (Figure 2(bFigure 2 and (cFigure 2). The temporal evolution of mass errors (Em,ΓE_{m,\Gamma} and Em,CE_{m,C}) is shown in Figure 2(f), where Em,Γ(t)=∫Ω∣Γ^(t)−Γ^0∣ dV/∫ΩΓ^0 dVE_{m,\Gamma}(t)=\int_{\Omega}|\widehat{\Gamma}(t)-\widehat{\Gamma}_0|\,\mathrm{d}V/\int_{\Omega}\widehat{\Gamma}_0\,\mathrm{d}V and Em,C(t)=∫Ω∣C^(t)−C^0∣ dV/∫ΩC^0 dVE_{m,C}(t)=\int_{\Omega}|\widehat{C}(t)-\widehat{C}_0|\,\mathrm{d}V/\int_{\Omega}\widehat{C}_0\,\mathrm{d}V. With profile correction (solid lines), mass errors (∼10−6\sim10^{-6}) are two orders of magnitude smaller than without correction (∼10−4\sim10^{-4}), indicating negligible surfactant mass loss and the improvement of our method [28, 35, 39].

Surfactant mass transfer between interface and bulk on a sphere

To assess the accuracy of the diffusion and source terms in the surfactant transport equations (Eqs. (1) and (4)), we simulate surfactant mass transfer between the interface and bulk on a sphere of radius RR [18]. The initial conditions assume a uniform bulk concentration C(r,t)∣t=0=C0C(r,t)|_{t=0}=C_0 and a clean interface Γ(r,t)∣t=0=0\Gamma(r,t)|_{t=0}=0. Only the adsorption term (via source term jj) is considered, with the diffusion time scale in the bulk assumed much longer than the adsorption kinetics. Neglecting convection, the bulk surfactant concentration evolves as,

∂C∂t=DCr2∂∂r(r2∂C∂r),(14)\frac{\partial C}{\partial t}=\frac{D_C}{r^2}\frac{\partial}{\partial r}\left(r^2\frac{\partial C}{\partial r}\right), \tag*{(14)}

with the boundary condition ∂C∂r∣r=R=kaCs\left.\frac{\partial C}{\partial r}\right|_{r=R}=k_a C_s and C∣r=∞=C0\left.C\right|_{r=\infty}=C_0. The surfactant mass conservation equation gives the interfacial mass MΓ(t)M_\Gamma(t) and concentration Γ(t)\Gamma(t) as MΓ(t)=MC(0)−4π∫R4RC(r,t)r2drM_\Gamma(t)=M_C(0)-4\pi\int_R^{4R}C(r,t)r^2dr and Γ(t)=MΓ(t)/4πR2\Gamma(t)=M_\Gamma(t)/4\pi R^2 [18]. We use a finite volume method (FVM) and explicit Euler method [34] to discretize the diffusion equation (Eq. (14)) and compare the result with the 1D solutions derived from the phase-field framework. Figure 3(a) shows the computational domain of size 4R×8R4R\times8R, with the fixed sphere centered at z=4Rz=4R. Parameters are chosen as: mesh size R/Δx=50R/\Delta x=50, time step Δt=5×10−4\Delta t=5\times10^{-4}, Cahn number Cn=0.75ΔxCn=0.75\Delta x, initial bulk concentration C0=1C_0=1, diffusion coefficients DC=DΓ=1D_C=D_\Gamma=1, and adsorption rate ka=1k_a=1. An axisymmetric boundary condition is applied at r=0r=0, while Neumann boundary conditions on the remaining boundaries.

Image not extracted

Figure 3. Surfactant mass transfer between interface and bulk on a stationary spherical drop. (a) Schematic representation of surfactant adsorption on a spherical interface. (b) Numerical snapshots of distributions of Γ\Gamma at the interface, and bulk surfactant concentration with C^\hat{C} shown on left side and CC on the right side at t=0.1t = 0.1, 0.30.3, and 0.50.5. (c) Comparison of profiles of CC (dots) with 1D solution (solid lines) at times t=0.1t = 0.1, 0.20.2, 0.30.3, 0.40.4, and 0.50.5. (d) Comparison of temporal evolution of Γ\Gamma with 1D solution (solid lines). (e) Convergence analysis of the infinity norm error L∞L_{\infty} at t=Tt = T with respect to the interface thickness (Cn=ΔxCn = \Delta x) and the spatial resolution (Cn=0.02Cn = 0.02).

As surfactant adsorbs onto the interface, the bulk concentration near the interface decreases (Figure 3(b) and (c)). At the same time, surfactant diffuses toward the interface, preserving spherical symmetry in the concentration contours and driving a continuous increase in Γ\Gamma (Figure 3(b) and (d)). Our numerical results show good agreement with 1D solutions for both bulk concentration CC profiles and interfacial concentration Γ\Gamma, confirming the accuracy of the numerical algorithm. To assess mesh convergence, we introduce the infinity norm error [34, 38] L∞=max⁡∣Γ(t)−Γ1D(t)∣L_\infty=\max\left|\Gamma(t)-\Gamma_{1D}(t)\right|, where Γ1D(t)\Gamma_{1D}(t) represents the 1D analytical solution. Figure 3(e) presents convergence rates of L∞L_\infty at t=0.5t=0.5 for two cases: variable numerical interface thickness (Cn=ΔxCn=\Delta x) and fixed numerical interface thickness (Cn=0.02Cn=0.02). The results indicate sub-first-order convergence for variable interface thickness, while fixed thickness achieves convergence between first and second order accuracy.

Influence of Surfactant Adsorption-Desorption on Droplet Settling

In this section, we apply the proposed method to investigate the settling of a surfactant-laden droplet. The droplet, assumed axisymmetric with radius RR, is initially placed at z=24Rz = 24R in a cylindrical domain of radius 6R6R and height 30R30R, as shown in Figure 4(a). Unlike previous studies where surfactants were dissolved in the external phase [6, 20, 53, 60, 61], we investigate a scenario where the surfactant is initially confined within the droplet but not in the external phase (Figure 4(b)).

Image not extracted

Figure 4. (a) Sketch of a surfactant-laden droplet of radius RR settling due to gravity. Surfactants are dissolved inside the droplet with initially uniform distribution of concentration (blue region). (b) Sketch of distribution of interfacial and bulk surfactant concentrations. The interfacial fluid motion convects the surfactant from the front toward the rear of the droplet, thereby establishing the gradient of interfacial surfactant concentration. (c) Sketch of surfactant mass transfer through adsorption and desorption processes and Marangoni stress arising from the interfacial surfactant gradient. Adsorption occurs at the front of the droplet owing to the decrease in the surfactant concentration, while desorption takes place at the rear due to surfactant accumulation.

This setup highlights the coupling between interfacial adsorption-desorption dynamics and the convective transport inside the droplet, both influencing the droplet’s settling behavior. As illustrated in Figure 4(c), settling-induced convection transports surfactant from the front to the rear of the droplet, creating an interfacial concentration gradient, ∇Γ\nabla\Gamma, that not only generates retarding Marangoni stresses [3, 41, 62] but also drives the adsorption-desorption process. Surfactant desorbs from the droplet surface when the interfacial concentration exceeds that in the sublayer, while adsorption occurs when the interfacial concentration is lower. To examine how the adsorption-desorption process affects the settling motion, we vary the Biot number Bi∈[10−3,10]Bi \in[10^{-3}, 10] and the Damköhler number Da∈[0.01,0.5]Da \in[0.01, 0.5], assuming that adsorption and desorption timescales are identical, i.e., (kaCCMC∗)−1=kd−1(k_a C^{*}_{CMC})^{-1} = k_d^{-1}, which leads to an adsorption number of k=1k = 1. This assumption simplifies the analysis while preserving the core dynamics of the coupling.

We define the characteristic length as the droplet diameter, L=2RL = 2R, and the characteristic velocity as U=(1−ζd)gLU = \sqrt{(1-\zeta_d)gL} with the density ratio ζd=ρ2/ρ1\zeta_d = \rho_2/\rho_1, as gravity drives the settling process. The Reynolds number is then defined via the Galileo number Re=GaRe = \sqrt{Ga}, where Ga=ρ12(1−ζd)gL3/μ12Ga = \rho_1^2(1-\zeta_d)gL^3/\mu_1^2 quantifies the relative effects of gravity and viscosity and is set as Ga=100Ga = 100 in the simulation. The hydrodynamics of the settling droplet are influenced by gravity, viscosity, and surface tension [41, 53, 63, 64].

Other relevant hydrodynamic parameters include the Froude number Fr=1−ζd=0.2Fr = 1-\zeta_d = 0.2, the Marangoni number Ma=RTΓ∞∗/(μ1(1−ζd)gL)=1Ma = \mathcal{R}T\Gamma_\infty^*/(\mu_1\sqrt{(1-\zeta_d)gL}) = 1, and the Capillary number Ca=μ1(1−ζd)gL/σ0=0.1Ca = \mu_1\sqrt{(1-\zeta_d)gL}/\sigma_0 = 0.1. The Péclet number for interfacial and bulk surfactant transport are set to PeΓ=100Pe_\Gamma= 100 and PeC=100Pe_C = 100, respectively, with the viscosity ratio ζμ=1\zeta_\mu= 1.

A far-field boundary condition at the top, a no-slip condition at the bottom, and a free-slip condition on the sides are applied (Figure 4(a)). We assume initial uniform and equilibrium surfactant concentrations for both the interfacial and bulk phases, with C0=0.25C_0 = 0.25 and Γ0=0.2\Gamma_0 = 0.2 according to Eq. (3). To ensure numerical stability and prevent non-physical negative values of surface tension, we impose a constraint on the surface tension, i.e., σ(Γ)=max⁡(σε,1+Eln⁡(1−Γ))\sigma(\Gamma) = \max(\sigma_\varepsilon, 1 + E\ln(1-\Gamma)), where σε=0.05\sigma_\varepsilon= 0.05 is a threshold. The Péclet number for the order parameter PeϕPe_\phi is defined as 1/Cn1/Cn, where Cn=0.75ΔxCn = 0.75\Delta x is the interface thickness. Finally, a mesh convergence study, as demonstrated in Appendix § 10, ensures the accuracy of the results. Based on this, the mesh size is set to 2R/Δx=1602R/\Delta x = 160 for the simulations presented hereafter.

Effect of Biot Number

The Biot number Bi=(2R/U)/kd−1Bi = (2R/U)/k_d^{-1} in Equation (1) quantifies the relative mass transfer rate of adsorption-desorption kinetics compared to interfacial advection, and the simulation results indicate that the Biot number (BiBi) affects the settling dynamics of surfactant-laden droplets.

Figure 5 (a) shows the droplet velocity UdU_d as a function of time tt for varying Biot numbers, ranging from 10−310^{-3} to 1010, while maintaining a constant Damköhler number Da=0.1Da = 0.1. The presence of surfactants leads to a lower final settling velocity compared to the surfactant-free case (black solid line). The evolution of settling velocity across different BiBi follows a similar trend, as highlighted in the inset. However, as the Biot number increases, the droplets reach a higher final settling velocity, as shown by the variation in the normalized final velocity relative to the surfactant-free value in Figure 5(b), where the dashed line represents the surfactant-free condition.

Image not extracted

Figure 5. (a) The droplet velocity UdU_d as a function of time tt for different Biot numbers BiBi with Ma=1Ma = 1, Da=0.1Da = 0.1, and k=1k = 1. (b) The droplet velocity ratio Uds/Uds,cleanU_{ds}/U_{ds,\mathrm{clean}} versus BiBi. UdsU_{ds} and Uds,cleanU_{ds,\mathrm{clean}} denote the steady velocity of droplet with and without surfactants (clean interface), respectively. (c) Contours of the interfacial and bulk surfactant concentrations Γ\Gamma (at the right half of the droplet interface) and CC, the adsorption-desorption flux jj (at the left half of the droplet interface), and the velocity vector (in the reference frame of the droplet) for Bi=10−2Bi = 10^{-2} and Bi=1Bi = 1 at t=9t = 9. Profiles of interfacial quantities as a function of the polar angle θ\theta for Bi=10−2Bi = 10^{-2} and Bi=1Bi = 1 at t=9t = 9: (d) Interfacial surfactant concentration Γ\Gamma, (e) Adsorption-desorption flux jj, (f) Marangoni stress FMaF_{Ma}.

To understand how the Biot number affects the velocity reduction, we examine the relationship between BiBi and the changes in surfactant distribution on the droplet surface. In Figure 5(c), we present two numerical snapshots of surfactant-laden droplets at t=9t=9 for Bi=10−2Bi=10^{-2} and Bi=1Bi=1, once they have reached their final settling velocity. The arrows indicate the flow field with the reference frame fixed on the droplet’s center of mass, and the bluish shading represents the surfactant concentration, CC. The color coding along the droplet’s left and right profiles corresponds to the adsorption-desorption flux, jj, and the surfactant interfacial concentration, Γ\Gamma, respectively.

In both cases (Figure 5(c)), the interfacial concentration is lower at the droplet’s front and higher at the rear due to convection. For Bi=10−2Bi=10^{-2}, however, Γ\Gamma exhibits a steeper distribution with both a lower minimum and higher maximum concentration, as shown in Figure 5(d), where θ\theta is the angular coordinate. This occurs because a small Biot number (Bi=2kdR/U≪1Bi=2k_dR/U\ll1) implies the adsorption-desorption time-scale (kd−1k_d^{-1}) is much larger than the advective timescale (2R/U2R/U), leading to a significantly weaker adsorption-desorption flux at Bi=10−2Bi=10^{-2}. Figure 5(e) further reveals that adsorption occurs at the droplet’s front (θ<π/2\theta<\pi/2) with positive fluxes, while desorption occurs at the rear with negative fluxes. For small BiBi, both adsorption and desorption rates are relatively negligible. Consequently, the sharper gradient in Γ\Gamma induces a stronger Marangoni stress (Figure 5(f)), which retards the droplet’s settling and results in a lower final settling velocity for small BiBi, as shown in Figure 5(a). Additionally, since the disruption of the interfacial concentration has little impact on bulk concentration variation for small BiBi, the bulk concentration becomes more uniform, as evidenced by the more homogeneous bluish color inside the droplet in Figure 5(c).

For the extreme case where Bi=0Bi=0, there is no adsorption-desorption process, which effectively corresponds to the scenario of an insoluble surfactant. In this case, the surfactant remains at the interface, and its distribution is solely governed by the interfacial convection and diffusion without any exchange with the bulk phase [41, 64, 65].

Effect of Damköhler Number

The Damköhler number Da=(Γ∞∗/CCMC∗)/(2R)Da = (\Gamma_{\infty}^{\ast}/C_{\mathrm{CMC}}^{\ast})/(2R) in Equation (4) characterizes the relative thickness of the depletion region compared to the droplet size. A higher Damköhler number indicates a thicker depletion region, which corresponds to a relatively higher surfactant capacity at the interface or a lower surfactant solubility in the bulk phase. The simulation results confirms that the Damköhler number DaDa influences the droplet settling dynamics as well.

As shown in Figure 6(a), we calculated the evolution of droplet velocity by increasing the Damköhler number from 0.01 to 0.1 and then to 0.5, while keeping the Biot number constant at Bi=0.5Bi = 0.5. For reference, the surfactant-free case (black solid line) is included in the figure. The normalized final velocity relative to the surfactant-free case, shown in Figure 6(b), further confirms the reduction in the final settling velocity as the Damköhler number increases. To provide further insights, we present two numerical snapshots for Da=0.01Da = 0.01 and Da=0.5Da = 0.5 at t=9t = 9, when the system reaches a steady state, in Figure 6(c). The arrows, shading, and color codings are consistent with those in Figure 5(c). The simulation results reveal that a large Damköhler number (Da=0.5Da = 0.5) results in a steeper distribution of the interfacial concentration Γ\Gamma (Figure 5(d)) and a corresponding weaker adsorption-desorption flux jj (Figure 5(e)). This occurs because a higher DaDa favors a greater surfactant presence at the interface, as opposed to in the bulk, implying higher interfacial capacity and lower bulk solubility. Consequently, convection transports more interfacial surfactants toward the droplet’s rear, leading to a higher interfacial concentration at the rear for Da=0.5Da = 0.5. Meanwhile, the reduced bulk solubility not only depresses the desorption flux at the rear but also limits the adsorption flux at the front, creating a larger concentration gradient and stronger Marangoni stresses (Figure 5(f)). These effects result in a lower settling velocity in the steady state. In contrast, for Da=0.01Da = 0.01, the increased surfactant solubility in the bulk enhances both adsorption and desorption fluxes. This causes a less steeper gradient in Γ\Gamma, weakening the Marangoni stresses and ultimately leading to a slightly higher settling velocity in the steady state.

Image not extracted

Figure 6. (a) The droplet velocity UdU_d as a function of time tt for different Damköhler numbers DaDa with Ma=1Ma=1, Bi=0.5Bi=0.5, and k=1k=1. (b) The droplet velocity ratio (Ud/UdcU_d/U_{dc}, where UdcU_{dc} is the steady velocity of droplet without surfactant (clean interface)) versus DaDa. (c) Contours of the interfacial and bulk surfactant concentrations Γ\Gamma and CC, the adsorption-desorption flux jj, and the velocity vector (in the reference frame of the droplet) for Da=10−2Da=10^{-2} and Da=0.5Da=0.5 at t=9t=9. Profiles of interfacial quantities as a function of the polar angle θ\theta for Da=10−2Da=10^{-2} and Da=0.5Da=0.5 at t=9t=9: (d) Interfacial surfactant concentration Γ\Gamma, (e) Adsorption and desorption flux jj, (f) Marangoni stress FMaF_{Ma}.

Conclusion

In summary, we have developed a diffuse-interface model for incompressible two-phase flow with sorption isotherms of dilute surfactants, utilizing the profile-preserving phase field method. This model maintains the profiles of the interfacial and bulk surfactants in diffuse-interface formulation since the assumption of interfacial profile equilibrium is precisely fulfilled by using the profile-preserving approach. Numerical test involving a droplet deformed in a single vortex demonstrates the model’s capability to eliminate interfacial surfactant concentration singularities while enhancing profile accuracy and mass conservation. Further validation through simulations of surfactant mass transfer on a stationary spherical droplet verifies the model’s accuracy.

We apply the model to investigate the settling of a surfactant-laden droplet. We observe increasing the Biot number enhances adsorption/desorption fluxes, relaxing interfacial surfactant gradients. This reduces Marangoni stress and increases droplet velocity. We also find increasing Damköhler number suppresses adsorption-desorption fluxes, amplifies interfacial surfactant gradients and Marangoni stress, and ultimately decreases steady-state droplet velocity.

In the future, we will extend the model to incorporate surfactant micelle formation at bulk surfactant concentrations exceeding the critical micelle concentration (CMC) In addition, at high Péclet numbers, bulk surfactant concentration exhibits sharp gradients across small length scales [19], necessitating adaptive mesh refinement (AMR) to resolve these multiscale concentration fields effectively.

Acknowledgements

This work has received financial support from the Natural Science Foundation of China (Grants No. 12472271), and the Guangdong Basic and Applied Basic Research Foundation (Grants No. 2024A1515010509 and 2024A1515010614).

Surfactant transport equations in sharp-interface and diffuse-interface formulations

The evolution of surfactant on the interface is governed by an advection-diffusion equation with a source term accounting for the surfactant mass transport between the interface and the bulk (i.e., adsorption and desorption) [3, 66]. Its dimensionless form in sharp-interface formulation can be written as

∂Γ∂t+∇s⋅(Γus)+Γ(∇s⋅n)(u⋅n)=1PeΓ∇s2Γ+j,(15)\frac{\partial\Gamma}{\partial t}+\nabla_s\cdot\left(\Gamma\mathbf{u}_s\right)+\Gamma\left(\nabla_s\cdot\mathbf{n}\right)\left(\mathbf{u}\cdot\mathbf{n}\right)=\frac{1}{Pe_{\Gamma}}\nabla_s^2\Gamma+j, \tag*{(15)}

where Γ=Γ∗/Γ∞∗\Gamma=\Gamma^*/\Gamma^*_{\infty} denotes the dimensionless surfactant concentration at the interface, Γ∞∗\Gamma^*_{\infty} represents its saturation interfacial concentration, ∇s=(I−nn)⋅∇\nabla_s=(\mathbf{I}-\mathbf{n}\mathbf{n})\cdot\nabla is the surface gradient operator, us=(I−nn)⋅u\mathbf{u}_s=(\mathbf{I}-\mathbf{n}\mathbf{n})\cdot\mathbf{u} denotes the tangential velocity along the interface, u\mathbf{u} is the dimensionless fluid velocity, and n\mathbf{n} is the normal direction of the interface. Here, PeΓ=UL/DΓPe_{\Gamma}=UL/D_{\Gamma} is the Péclet number representing the rate ratio between advective and diffusive surfactant transport at the interface.

The surfactant that is soluble in one phase, is governed by the time-dependent advection-diffusion equation [3, 6]. In sharp-interface formulation, the dimensionless form of this equation can be written as

∂C∂t+∇⋅(Cu)=1PeC∇⋅(∇C),(16)\frac{\partial C}{\partial t}+\nabla\cdot(C\mathbf{u})=\frac{1}{Pe_C}\nabla\cdot(\nabla C), \tag*{(16)}

where C=C∗/CCMC∗C=C^*/C^*_{CMC} represents the dimensionless surfactant concentration at the bulk, and PeΓ=UL/DCPe_{\Gamma}=UL/D_C is the Péclet number representing the rate ratio between advective and diffusive surfactant transport in the bulk. Here, CCMC∗C^*_{CMC} is its saturation bulk concentration and DCD_C is the diffusion coefficient of surfactant in the bulk. The dimensionless form of the balance between surfactant diffusion to the interface and adsorption-desorption can be expressed as [3, 6]

1PeC∂C∂n∣S=−Daj,(17)\left.\frac{1}{Pe_C}\frac{\partial C}{\partial n}\right|_S=-Da j, \tag*{(17)}

where Da=Γ∞∗/(CCMC∗L)Da=\Gamma^*_{\infty}/(C^*_{CMC}L) is the Damköhler number which characterizes the ratio of the thickness of the depleted region Γ∞∗/(CCMC∗)\Gamma^*_{\infty}/(C^*_{CMC}) to the characteristic length LL.

The above surfactant transport equations (Eqs. (A.1) and (A.2)) in sharp-interface formulation can be rewritten in a diffuse-interface one, as [35]

∂(δΓ)∂t+∇⋅(δΓu)=1PeΓ∇⋅(δ∇Γ)+δj,(18)\frac{\partial(\delta\Gamma)}{\partial t}+\nabla\cdot(\delta\Gamma\mathbf{u})=\frac{1}{Pe_{\Gamma}}\nabla\cdot(\delta\nabla\Gamma)+\delta j, \tag*{(18)}
∂(ϕC)∂t+∇⋅(ϕCu)=1PeC∇⋅(ϕ∇C)−Daδj.(19)\frac{\partial(\phi C)}{\partial t}+\nabla\cdot(\phi C\mathbf{u})=\frac{1}{Pe_{C}}\nabla\cdot(\phi\nabla C)-Da\delta j. \tag*{(19)}

Profile-preserving phase-field model

We employ the phase field method coupled with a profile-preserving approach [46] to implicitly capture the interface of two immiscible, incompressible fluids. This method can keep the profile of interface in a equilibrium state, maintain the mass conservation of each phase, and thus is more accurate than the tranditional phase-field method that adopts a single interface-capturing equation, e.g., the Cahn-Hilliard equation [29, 45]. The interface is characterized by the order parameter ϕ(x,t)\phi(\mathbf{x},t), governed by the advective Cahn-Hilliard equation [29, 45], as

∂ϕ∂t+∇⋅(ϕu)=1Peϕ∇2η,(20)\frac{\partial\phi}{\partial t}+\nabla\cdot(\phi\mathbf{u})=\frac{1}{Pe_{\phi}}\nabla^{2}\eta, \tag*{(20)}

where u\mathbf{u} is the dimensionless fluid velocity, η=ϕ3−1.5ϕ2+0.5ϕ−Cn2∇2ϕ\eta=\phi^{3}-1.5\phi^{2}+0.5\phi-Cn^{2}\nabla^{2}\phi represents the dimensionless chemical potential, and Peϕ=(M∗/Cn)−1Pe_{\phi}=(M^{*}/Cn)^{-1} denotes the Péclet number characterizing the ratio of convection to diffusion for the order parameter ϕ\phi. Here, M∗M^{*} is the mobility number and Cn=ℓ/LCn=\ell/L is the Cahn number, where ℓ\ell is the typical interface thickness at mesoscopic scales, and LL is a characteristic length scale for the macroscopic size of the fluid system. To attain the sharp-interface limit, where the diffuse interface converges to a sharp interface as CnCn approaches zero, we employ the relations M∗∼Cn2M^{*}\sim Cn^{2} and Peϕ∼Cn−1Pe_{\phi}\sim Cn^{-1} [67]. A specific form of equilibrium profiles for the order parameter ϕ(x,t)\phi(\mathbf{x},t) [57] can be determined by minimizing the interfacial free energy, expressed as [29, 45]

ϕ(x,t)=12{1+tanh⁡[ψ(x,t)22Cn]},\phi(\mathbf{x},t)=\frac{1}{2}\left\{1+\tanh\left[\frac{\psi(\mathbf{x},t)}{2\sqrt{2}Cn}\right]\right\},

where ψ(x,t)\psi(\mathbf{x},t) is the signed-distance function.

To correct the interfacial profile to maintain its equilibrium state, we use an interfacial profile-preserving approach [46] and the profile-preserving equation is given by

∂ϕ∂τ=∇⋅{[2Cn(∇ϕ⋅n(ψ))−14(1−tanh⁡2(ψ22Cn))n(ψ)]}.(21)\frac{\partial\phi}{\partial\tau}=\nabla\cdot\left\{\left[\sqrt{2}Cn\left(\nabla\phi\cdot\mathbf{n}(\psi)\right)-\frac{1}{4}\left(1-\tanh^{2}\left(\frac{\psi}{2\sqrt{2}Cn}\right)\right)\mathbf{n}(\psi)\right]\right\}. \tag*{(21)}

where τ\tau is an iteration time, n(ψ)=∇ψ∣τ=0/∣∇ψ∣τ=0\mathbf{n}(\psi)=\left.\nabla\psi\right|_{\tau=0}/\left|\nabla\psi\right|_{\tau=0} represents the normal direction of the interface, and the signed distance function is expressed as ψ=2Cnln⁡[ϕ/(1−ϕ)]\psi=\sqrt{2}Cn\ln[\phi/(1-\phi)]. We iteratively solve this equation with respect to τ\tau until the phase field function ϕ\phi satisfies the steady state criteria, defined by ∫Ω∣ϕm+1−ϕm∣ dΩ≤TOL⋅Δτ\int_{\Omega}|\phi_{m+1}-\phi_m|\,\mathrm{d}\Omega\leq TOL\cdot\Delta\tau, where the subscript notation mm-th denotes the mm-th artificial correction step and TOLTOL is the threshold value. Unless otherwise stated, we set TOL=1TOL=1 and Δτ=0.005Δx\Delta\tau=0.005\Delta x, with Δx\Delta x representing the grid size.

The motion of two-phase flows, incorporating the effects of surface tension and gravity, is governed by the dimensionless Navier–Stokes equations, expressed as follows

∇⋅u=0,(22)\nabla\cdot\mathbf{u}=0, \tag*{(22)}
ρ(ϕ)(∂u∂t+u⋅∇u)=−∇p+1Re{∇⋅[μ(ϕ)(∇u+∇uT)]+FstCa}+ρ(ϕ)Frj,(23)\rho(\phi)\left(\frac{\partial\mathbf{u}}{\partial t}+\mathbf{u}\cdot\nabla\mathbf{u}\right)=-\nabla p+\frac{1}{Re}\left\{\nabla\cdot\left[\mu(\phi)\left(\nabla\mathbf{u}+\nabla\mathbf{u}^{T}\right)\right]+\frac{\mathbf{F}_{\mathrm{st}}}{Ca}\right\}+\frac{\rho(\phi)}{Fr}\mathbf{j}, \tag*{(23)}

where pp represents pressure, Fst\mathbf{F}_{\mathrm{st}} denotes the surface tension force, and j\mathbf{j} indicates the direction of gravitational acceleration. We use the Langmuir equation of state [6] to characterize the relationship between the surface tension coefficient σ(Γ)\sigma(\Gamma) and the interfacial surfactant concentration Γ\Gamma, i.e., σ(Γ)=1+Eln⁡(1−Γ)\sigma(\Gamma)=1+E\ln(1-\Gamma), where E=RTΓ∞∗/σ0E=RT\Gamma^{*}_{\infty}/\sigma_{0} is the elasticity number. Here, RR is the ideal gas constant, TT is the temperature, and σ0\sigma_{0} is the surface tension coefficient. We adopt the hybrid Free-Energy/Continuum Surface force model (FECSF) [49], and the surface tension force in dimensionless form is written as

FstCa=62η∇ϕCaCn[1+Eln⁡(1−Γ)]+Ma∇sln⁡(1−Γ)∣∇ϕ∣,(24)\frac{\mathbf{F}_{\mathrm{st}}}{Ca}=\frac{6\sqrt{2}\eta\nabla\phi}{CaCn}\left[1+E\ln(1-\Gamma)\right]+Ma\nabla_{s}\ln(1-\Gamma)|\nabla\phi|, \tag*{(24)}

where Ma=E/Ca=RTΓ∞∗/(μ1U)Ma=E/Ca=RT\Gamma^{*}_{\infty}/(\mu_{1}U) is the Marangoni number which quantifies the relative strength of the Marangoni stress and the viscous stress.

In Eq. (B.4), the properties of fluid 1 are characterized by density ρ1\rho_{1} and viscosity μ1\mu_{1}, while fluid 2 is defined by ρ2\rho_{2} and μ2\mu_{2}. Using the properties of fluid 1, we define the dimensionless fluid density ρ(ϕ)\rho(\phi) and dynamic viscosity μ(ϕ)\mu(\phi) as ρ(ϕ)=ϕ+(1−ϕ)ζd\rho(\phi)=\phi+(1-\phi)\zeta_{d} and μ(ϕ)=ϕ+(1−ϕ)ζμ\mu(\phi)=\phi+(1-\phi)\zeta_{\mu}, respectively, where ζd=ρ2/ρ1\zeta_{d}=\rho_{2}/\rho_{1} and ζμ=μ2/μ1\zeta_{\mu}=\mu_{2}/\mu_{1} represent the density ratio and the viscosity ratio, respectively. Three dimensionless parameters used are Reynolds number Re=ρ1UL/μ1Re=\rho_{1}UL/\mu_{1}, Capillary number Ca=μ1U/σ0Ca=\mu_{1}U/\sigma_{0}, and Froude number Fr=U2/(gL)Fr = U^2/(gL), where gg and LL represent the gravitational acceleration and the characteristic length, respectively.

Mesh convergence tests

We assess the accuracy of the settling velocity of the droplet and the distribution of the interfacial surfactant concentration with respect to the mesh size Δx\Delta x, as shown in Fig. C.7. Simulations with three different mesh sizes (2R/Δx=106.72R/\Delta x = 106.7, 160 and 213.3) for Bi=1Bi = 1 and Da=0.1Da = 0.1 are performed. From the enlarged view of Fig. C.7, we observe the convergence behavior with increasing mesh resolution. We also find that the mesh size of 2R/Δx=1602R/\Delta x = 160 ensures the accurate resolution of both interfacial and bulk surfactant concentrations.

Image not extracted

Figure C.7. Mesh convergence tests for settling of a surfactant-laden droplet for Bi=1Bi = 1 and Da=0.1Da = 0.1. (a) The droplet velocity UdU_d as a function of dimensionless time tt with different mesh resolutions. Profiles of surfactant concentrations as a function of the polar angle θ\theta at t=9t = 9: (b) Interfacial surfactant concentration Γ\Gamma, (c) Bulk surfactant concentration CC.

References

  1. [1]V. Levich, V. Krylov, Surface-tension-driven phenomena, Annual review of fluid mechanics 1 (1) (1969) 293–316. https://doi.org/10.1146/annurev.fl.01.010169.001453
  2. [2]H. A. Stone, L. G. Leal, The effects of surfactants on drop deformation and breakup, Journal of Fluid Mechanics 220 (1990) 161–186. https://doi.org/10.1017/S0022112090003226
  3. [3]L. G. Leal, Advanced Transport Phenomena: Fluid Mechanics and Convective Transport Processes, Cambridge Series in Chemical Engineering, Cambridge University Press, 2007. https://doi.org/10.1017/CBO9780511800245
  4. [4]R. Stocker, J. W. Bush, Spontaneous oscillations of a sessile lens, Journal of Fluid Mechanics 583 (2007) 465–475. https://doi.org/10.1017/S0022112007005940
  5. [5]S. Takagi, Y. Matsumoto, Surfactant effects on bubble motion and bubbly flows, Annual Review of Fluid Mechanics 43 (1) (2011) 615–636. https://doi.org/10.1146/annurev-fluid-122109-160756
  6. [6]H. Manikantan, T. M. Squires, Surfactant dynamics: hidden variables controlling fluid flows, Journal of Fluid Mechanics 892 (2020) P1. https://doi.org/10.1017/jfm.2020.170
  7. [7]D. Lohse, Surfactants on troubled waters, Journal of Fluid Mechanics 976 (2023) F1. https://doi.org/10.1017/jfm.2023.891
  8. [8]X. Zhang, M. H. Uddin, H. Yang, G. Toikka, W. Ducker, N. Maeda, Effects of surfactants on the formation and the stability of interfacial nanobubbles, Langmuir 28 (28) (2012) 10471–10477. https://doi.org/10.1021/la301851g
  9. [9]A. Perazzo, G. Tomaiuolo, V. Preziosi, S. Guido, Emulsions in porous media: From single droplet behavior to applications for oil recovery, Advances in Colloid and Interface Science 256 (2018) 305–325. https://doi.org/10.1016/j.cis.2018.03.002
  10. [10]A. Banerjee, H. Tan, T. M. Squires, Drop-in additives for suspension manipulation: Colloidal motion induced by sedimenting soluto-inertial beacons, Physical Review Fluids 5 (2020) 073701. https://link.aps.org/doi/10.1103/PhysRevFluids.5.073701DOI
  11. [11]H. Tan, A. Banerjee, N. Shi, X. Tang, A. Abdel-Fattah, T. M. Squires, A two-step strategy for delivering particles to targets hidden within microfabricated porous media, Science Advances 7 (33) (2021) eabh0638. https://www.science.org/doi/abs/10.1126/sciadv.abh0638DOI
  12. [12]D. Lohse, Fundamental fluid dynamics challenges in inkjet printing, Annual Review of Fluid Mechanics 54 (1) (2022) 349–382. https://doi.org/10.1146/annurev-fluid-022321-114001
  13. [13]S. Michelin, Self-propulsion of chemically active droplets, Annual Review of Fluid Mechanics 55 (1) (2023) 77–101. https://doi.org/10.1146/annurev-fluid-120720-012204
  14. [14]F. Temprano-Coleto, S. M. Smith, F. J. Peaudecerf, J. R. Landel, F. Gibou, P. Luzzatto-Fegiz, A single parameter can predict surfactant impairment of superhydrophobic drag reduction, Proceedings of the National Academy of Sciences 120 (3) (2023) e2211092120. https://www.pnas.org/doi/abs/10.1073/pnas.2211092120
  15. [15]M. Erinin, C. Liu, X. Liu, W. Mostert, L. Deike, J. Duncan, The effects of surfactants on plunging breakers, Journal of Fluid Mechanics 972 (2023) R5. https://doi.org/10.1017/jfm.2023.721
  16. [16]L. Yi, I. Girotto, F. Toschi, C. Sun, Divergence of critical fluctuations on approaching catastrophic phase inversion in turbulent emulsions, Physical Review Letters 133 (13) (2024) 134001. https://doi.org/10.1103/PhysRevLett.133.134001
  17. [17]J. Zhang, D. Eckmann, P. Ayyaswamy, A front tracking method for a deformable intravascular bubble in a tube with soluble surfactant transport, Journal of Computational Physics 214 (1) (2006) 366–396. https://doi.org/10.1016/j.jcp.2005.09.016
  18. [18]M. Muradoglu, G. Tryggvason, A front-tracking method for computation of interfacial flows with soluble surfactants, Journal of Computational Physics 227 (4) (2008) 2238–2262. https://doi.org/10.1016/j.jcp.2007.10.003
  19. [19]M. Booty, M. Siegel, A hybrid numerical method for interfacial fluid flow with soluble surfactant, Journal of Computational Physics 229 (10) (2010) 3864–3883. https://doi.org/10.1016/j.jcp.2010.01.032
  20. [20]M. Muradoglu, G. Tryggvason, Simulations of soluble surfactants in 3d multiphase flow, Journal of Computational Physics 274 (2014) 737–757. https://doi.org/10.1016/j.jcp.2014.06.024
  21. [21]K.-Y. Chen, M.-C. Lai, A conservative scheme for solving coupled surface–bulk convection–diffusion equations with an application to interfacial flows with soluble surfactant, Journal of Computational Physics 257 (2014) 1–18. https://doi.org/10.1016/j.jcp.2013.10.003
  22. [22]M. Rump, C. Diddens, U. Sen, M. Versluis, D. Lohse, T. Segers, Role of surfactants in droplet formation in piezoacoustic inkjet printing across microsecond-to-second timescales, Phys. Rev. Appl. 23 (2025) 024076. https://link.aps.org/doi/10.1103/PhysRevApplied.23.024076DOI
  23. [23]S. Adami, X. Hu, N. Adams, A conservative sph method for surfactant dynamics, Journal of Computational Physics 229 (5) (2010) 1909–1926. https://doi.org/10.1016/j.jcp.2009.11.015
  24. [24]C. Pesci, A. Weiner, H. Marschall, D. Bothe, Computational analysis of single rising bubbles influenced by soluble surfactant, Journal of Fluid Mechanics 856 (2018) 709–763. https://doi.org/10.1017/jfm.2018.723
  25. [25]T. Antritter, T. Josyula, T. Marić, D. Bothe, P. Hachmann, B. Buck, T. Gambaryan-Roisman, P. Stephan, A two-field formulation for surfactant transport within the algebraic volume of fluid method, Computers and Fluids 275 (2024) 106231. https://doi.org/10.1016/j.compfluid.2024.106231
  26. [26]L. Stricker, Numerical simulation of artificial microswimmers driven by marangoni flow, Journal of Computational Physics 347 (2017) 467–489. https://doi.org/10.1016/j.jcp.2017.07.007
  27. [27]C. Cleret de Langavant, A. Guittet, M. Theillard, F. Temprano-Coleto, F. Gibou, Level-set simulations of soluble surfactant driven flows, Journal of Computational Physics 348 (2017) 271–297. https://doi.org/10.1016/j.jcp.2017.07.003
  28. [28]J.-J. Xu, W. Shi, M.-C. Lai, A level-set method for two-phase flows with soluble surfactant, Journal of Computational Physics 353 (2018) 336–355. https://doi.org/10.1016/j.jcp.2017.10.019
  29. [29]H. Ding, P. D. Spelt, C. Shu, Diffuse interface model for incompressible two-phase flows with large density ratios, Journal of Computational Physics 226 (2) (2007) 2078–2095. https://doi.org/10.1016/j.jcp.2007.06.028
  30. [30]G. Soligo, A. Roccon, A. Soldati, Coalescence of surfactant-laden drops by phase field method, Journal of Computational Physics 376 (2019) 1292–1311. https://doi.org/10.1016/j.jcp.2018.10.021
  31. [31]H. Liu, Y. Zhang, Phase-field modeling droplet dynamics with soluble surfactants, Journal of Computational Physics 229 (24) (2010) 9166–9187. https://doi.org/10.1016/j.jcp.2010.08.031
  32. [32]S. Shin, J. Chergui, D. Juric, L. Kahouadji, O. K. Matar, R. V. Craster, A hybrid interface tracking – level set technique for multiphase flow with soluble surfactant, Journal of Computational Physics 359 (2018) 409–435. https://doi.org/10.1016/j.jcp.2018.01.010
  33. [33]W.-F. Hu, M.-C. Lai, C. Misbah, A coupled immersed boundary and immersed interface method for interfacial flows with soluble surfactant, Computers and Fluids 168 (2018) 201–215. https://doi.org/10.1016/j.compfluid.2018.04.013
  34. [34]K. E. Teigen, X. Li, J. Lowengrub, F. Wang, A. Voigt, A diffuse-interface approach for modeling transport, diffusion and adsorption/desorption of material quantities on a deformable interface, Communications in Mathematical Sciences 4 (7) (2009) 1009–1037. https://doi.org/10.4310/cms.2009.v7.n4.a10
  35. [35]K. Erik Teigen, P. Song, J. Lowengrub, A. Voigt, A diffuse-interface method for two-phase flows with soluble surfactants, Journal of Computational Physics 230 (2) (2011) 375–393. https://doi.org/10.1016/j.jcp.2010.09.020
  36. [36]S. S. Jain, A. Mani, A computational model for transport of immiscible scalars in two-phase flows, Journal of Computational Physics 476 (2023) 111843. https://doi.org/10.1016/j.jcp.2022.111843
  37. [37]S. S. Jain, A model for transport of interface-confined scalars and insoluble surfactants in two-phase flows, Journal of Computational Physics 515 (2024) 113277. https://doi.org/10.1016/j.jcp.2024.113277
  38. [38]S. Yamashita, S. Matsushita, T. Suekane, Conservative transport model for surfactant on the interface based on the phase-field method, Journal of Computational Physics 516 (2024) 113292. https://doi.org/10.1016/j.jcp.2024.113292
  39. [39]H. Liu, J. Zhang, Y. Ba, N. Wang, L. Wu, Modelling a surfactant-covered droplet on a solid surface in three-dimensional shear flow, Journal of Fluid Mechanics 897 (2020) A33. https://doi.org/10.1017/jfm.2020.416
  40. [40]Y. Ba, H. Liu, W. Li, W. Yang, Modelling the effect of soluble surfactants on droplet deformation and breakup in simple shear flow, Journal of Fluid Mechanics 988 (2024) A41. https://doi.org/10.1017/jfm.2024.463
  41. [41]P. K. Farsoiya, S. Popinet, H. A. Stone, L. Deike, Coupled volume of fluid and phase field method for direct numerical simulation of insoluble surfactant-laden interfacial flows and application to rising bubbles, Physical Review Fluids 9 (2024) 094004. https://link.aps.org/doi/10.1103/PhysRevFluids.9.094004DOI
  42. [42]H. Liu, Y. Ba, L. Wu, Z. Li, G. Xi, Y. Zhang, A hybrid lattice boltzmann and finite difference method for droplet dynamics with insoluble surfactants, Journal of Fluid Mechanics 837 (2018) 381–412. https://doi.org/10.1017/jfm.2017.859
  43. [43]Y. Hu, A diffuse interface–lattice boltzmann model for surfactant transport on an interface, Applied Mathematics Letters 111 (2021) 106614. https://doi.org/10.1016/j.aml.2020.106614
  44. [44]S. S. Jain, Modeling soluble surfactants in two-phase flows, Center for Turbulence Research Annual Research Briefs (2023) 135–145. https://web.stanford.edu/group/ctr/ResBriefs/2023/13_Jain.pdfDOI
  45. [45]D. Jacqmin, Calculation of two-phase navier–stokes flows using phase-field modeling, Journal of Computational Physics 155 (1) (1999) 96–127. https://doi.org/10.1006/jcph.1999.6332
  46. [46]H. Hao, X. Li, C. Jiang, H. Tan, An interfacial profile-preserving approach for phase field modeling of incompressible two-phase flows, International Journal of Multiphase Flow 174 (2024) 104750. https://doi.org/10.1016/j.ijmultiphaseflow.2024.104750
  47. [47]G. Soligo, A. Roccon, A. Soldati, Mass-conservation-improved phase field methods for turbulent multiphase flow simulation, Acta Mechanica 230 (2019) 683–696. https://doi.org/10.1007/s00707-018-2304-2
  48. [48]P.-H. Chiu, A coupled phase field framework for solving incompressible two-phase flows, Journal of Computational Physics 392 (2019) 115–140. https://doi.org/10.1016/j.jcp.2019.04.069
  49. [49]H. Hao, X. Li, T. Liu, H. Tan, Enhanced profile-preserving phase-field model of two-phase flow with surfactant interfacial transport and marangoni effects (2024). https://arxiv.org/abs/2409.19374
  50. [50]P. LAKSHMANAN, P. EHRHARD, Marangoni effects caused by contaminants adsorbed on bubble surfaces, Journal of Fluid Mechanics 647 (2010) 143–161. https://doi.org/10.1017/S0022112009993077
  51. [51]S. S. Dukhin, V. I. Kovalchuk, G. G. Gochev, M. Lotfi, M. Krzan, K. Malysa, R. Miller, Dynamics of rear stagnant cap formation at the surface of spherical bubbles rising in surfactant solutions at large reynolds numbers under conditions of small marangoni number and slow sorption kinetics, Advances in Colloid and Interface Science 222 (2015) 260–74. https://doi.org/10.1016/j.cis.2014.10.002
  52. [52]C. Pesci, A. Weiner, H. Marschall, D. Bothe, Computational analysis of single rising bubbles influenced by soluble surfactant, Journal of Fluid Mechanics 856 (2018) 709–763. https://doi.org/10.1017/jfm.2018.723
  53. [53]A. C. Castonguay, R. Kailasham, C. M. Wentworth, C. H. Meredith, A. S. Khair, L. D. Zarzar, Gravitational settling of active droplets, Physical Review E 107 (2023) 024608. https://doi.org/10.1103/PhysRevE.107.024608
  54. [54]S. N. Jadhav, S. Mandal, U. Ghosh, Coupled bulk and interfacial transport of surfactants governs the settling of a drop towards a wall, Journal of Fluid Mechanics 987 (2024) A41. https://doi.org/10.1017/jfm.2024.406
  55. [55]H. A. Stone, A simple derivation of the time-dependent convective-diffusion equation for surfactant transport along a deforming interface, Physics of Fluids A: Fluid Dynamics 2 (1) (1990) 111–112. https://doi.org/10.1063/1.857686
  56. [56]P. A. Kralchevsky, K. D. Danov, N. D. Denkov, Chemical physics of colloid systems and interfaces, Handbook of surface and colloid chemistry 2 (1997).
  57. [57]J. W. Cahn, J. E. Hilliard, Free Energy of a Nonuniform System. I. Interfacial Free Energy, The Journal of Chemical Physics 28 (2) (1958) 258–267. https://doi.org/10.1063/1.1744102
  58. [58]U. M. Ascher, S. J. Ruuth, B. T. R. Wetton, Implicit-explicit methods for time-dependent partial differential equations, SIAM Journal on Numerical Analysis 32 (3) (1995) 797–823. https://doi.org/10.1137/0732037
  59. [59]X.-D. Liu, S. Osher, T. Chan, Weighted essentially non-oscillatory schemes, Journal of Computational Physics 115 (1) (1994) 200–212. https://doi.org/10.1006/jcph.1994.1187
  60. [60]J. Chen, K. J. Stebe, Marangoni retardation of the terminal velocity of a settling droplet: The role of surfactant physico-chemistry, Journal of Colloid and Interface Science 178 (1) (1996) 144–155. https://doi.org/10.1006/jcis.1996.0102
  61. [61]S. N. Jadhav, S. Mandal, U. Ghosh, Coupled bulk and interfacial transport of surfactants governs the setting of a drop towards a wall, Journal of Fluid Mechanics 987 (2024). https://doi.org/10.1017/jfm.2024.406
  62. [62]N. Young, J. S. Goldstein, M. Block, The motion of bubbles in a vertical temperature gradient, Journal of Fluid Mechanics 6 (3) (1959) 350–356. https://doi.org/10.1017/S0022112059000684
  63. [63]O. S. Pak, J. Feng, H. A. Stone, Viscous marangoni migration of a drop in a poiseuille flow at low surface péclet numbers, Journal of Fluid Mechanics 753 (2014) 535–552. https://doi.org/10.1017/jfm.2014.380
  64. [64]S. N. Jadhav, U. Ghosh, Effect of surfactant on the settling of a drop towards a wall, Journal of Fluid Mechanics 912 (2021). https://doi.org/10.1017/jfm.2020.1101
  65. [65]D. Fernández-Martínez, M. Cabezas, J. López-Herrera, M. Herrada, J. Montanero, Transient bubble rising in the presence of a surfactant at very low concentrations, International Journal of Multiphase Flow 188 (2025) 105205. https://doi.org/10.1016/j.ijmultiphaseflow.2025.105205
  66. [66]L. E. Scriven, Dynamics of a fluid interface equation of motion for newtonian surface fluids, Chemical Engineering Science 12 (2) (1960) 98–108. https://doi.org/10.1016/0009-2509(60)87003-0
  67. [67]F. Magaletti, F. Picano, M. Chinappi, L. Marino, C. M. Casciola, The sharp-interface limit of the cahn–hilliard/navier–stokes model for binary fluids, Journal of Fluid Mechanics 714 (2013) 95–126. https://doi.org/10.1017/jfm.2012.461

Paper details

vlnc:W_ez7cja5h
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents