Mass Transfer to Atmospheric Particles

Overview

This module implements the fundamental equations describing mass transfer between gas and particle phases in the atmosphere. The treatment covers three transport regimes based on the Knudsen number (Kn = λ/Rp, ratio of mean free path to particle radius):

  • Continuum regime (Kn ≪ 1): Applicable to large particles where Fick's law of diffusion governs transport
  • Kinetic regime (Kn ≫ 1): Applicable to small particles where molecular kinetic theory governs transport
  • Transition regime (Kn ~ 1): Requires interpolation formulas bridging the two limiting cases

The implementation includes:

  • Transition regime correction factors (Fuchs-Sutugin, Dahneke)
  • Maxwellian flux for particle growth/evaporation
  • Mass transfer coefficients for gas-particle exchange
  • Characteristic timescales for various mass transport processes
  • Aqueous-phase chemistry and mass transport coupling

Reference: Seinfeld, J.H. and Pandis, S.N. (2006) "Atmospheric Chemistry and Physics: From Air Pollution to Climate Change", 2nd Edition, John Wiley & Sons, Chapter 12, pp. 537-587.

Aerosol.MassTransfer — Function
MassTransfer(; name=:MassTransfer)

Complete mass transfer model combining transition regime correction, gas-phase diffusion, and interfacial transport.

This component provides the transition-regime corrected mass flux for particle growth/evaporation, applicable across all Knudsen numbers.

Reference: Seinfeld & Pandis (2006) Chapter 12, Sections 12.1.1-12.1.4

source
Aerosol.FuchsSutugin — Function
FuchsSutugin(; name=:FuchsSutugin)

Fuchs-Sutugin transition regime correction for mass transfer.

This is the most widely used formula for transition regime mass transfer due to its simplicity and accuracy across the full Knudsen number range.

Returns the ratio J/Jc where Jc is the continuum regime flux.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eq. 12.43 (with accommodation coefficient)

source
Aerosol.Dahneke — Function
Dahneke(; name=:Dahneke)

Dahneke transition regime correction for mass transfer.

An alternative interpolation formula for the transition regime.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eq. 12.42

source
Aerosol.MaxwellianFlux — Function
MaxwellianFlux(; name=:MaxwellianFlux)

Calculate the Maxwellian (continuum regime) molar flux to a particle.

This is the steady-state diffusive flux in the continuum regime (Kn ≪ 1).

Reference: Seinfeld & Pandis (2006) Chapter 12, Eq. 12.12

source
Aerosol.MassTransferCoefficient — Function
MassTransferCoefficient(; name=:MassTransferCoefficient)

Calculate the combined gas-phase and interfacial mass transfer coefficient.

This coefficient accounts for both gas-phase diffusion and interfacial resistance to mass transfer.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eq. 12.116

source
Aerosol.UptakeCoefficient — Function
UptakeCoefficient(; name=:UptakeCoefficient)

Calculate the uptake coefficient γ for gas-particle mass transfer.

The uptake coefficient is the probability that a molecule striking the particle surface is taken up by the particle.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eq. 12.121

source

Implementation

Transition Regime Mass Transfer

The Fuchs-Sutugin formula is the most widely used expression for mass transfer across the full range of Knudsen numbers:

using Aerosol
using ModelingToolkit
using ModelingToolkit: mtkcompile
using NonlinearSolve
using Plots

# Create the Fuchs-Sutugin system
sys = FuchsSutugin()
compiled = mtkcompile(sys)

# Calculate correction factor across Knudsen number range
Kn_range = 10 .^ range(-2, 2, length = 100)
f_FS_α1 = Float64[]
f_FS_α01 = Float64[]
f_FS_α001 = Float64[]

for Kn in Kn_range
    prob = NonlinearProblem(compiled, Dict(compiled.Kn => Kn, compiled.α => 1.0))
    push!(f_FS_α1, solve(prob)[compiled.f_FS])

    prob = NonlinearProblem(compiled, Dict(compiled.Kn => Kn, compiled.α => 0.1))
    push!(f_FS_α01, solve(prob)[compiled.f_FS])

    prob = NonlinearProblem(compiled, Dict(compiled.Kn => Kn, compiled.α => 0.01))
    push!(f_FS_α001, solve(prob)[compiled.f_FS])
end

plot(Kn_range, f_FS_α1, label = "α = 1.0", xscale = :log10,
    xlabel = "Knudsen Number (Kn)", ylabel = "J/Jc (Correction Factor)",
    title = "Fuchs-Sutugin Transition Regime Correction")
plot!(Kn_range, f_FS_α01, label = "α = 0.1")
plot!(Kn_range, f_FS_α001, label = "α = 0.01")
savefig("fuchs_sutugin.svg");

Fuchs-Sutugin correction factor

The correction factor approaches 1 in the continuum limit (Kn → 0) and decreases in the kinetic regime (Kn → ∞). Lower accommodation coefficients (α) reduce the mass transfer rate.

State Variables

using DataFrames, Symbolics, DynamicQuantities

sys = MassTransfer()
vars = ModelingToolkit.unknowns(sys)
DataFrame(
    :Name => [string(Symbolics.tosymbol(v, escape = false)) for v in vars],
    :Units => [string(DynamicQuantities.dimension(ModelingToolkit.get_unit(v)))
               for v in vars],
    :Description => [ModelingToolkit.getdescription(v) for v in vars]
)
7×3 DataFrame
RowNameUnitsDescription
StringStringString
1c_barm s⁻¹Mean molecular speed
2ρ_airm⁻³ kgAir density
3λmMean free path
4KnKnudsen number (dimensionless)
5f_FSFuchs-Sutugin correction factor (dimensionless)
6J_cs⁻¹ molContinuum regime molar flow rate
7Js⁻¹ molTransition regime corrected molar flow rate

Parameters

params = ModelingToolkit.parameters(sys)
DataFrame(
    :Name => [string(Symbolics.tosymbol(p, escape = false)) for p in params],
    :Units => [string(DynamicQuantities.dimension(ModelingToolkit.get_unit(p)))
               for p in params],
    :Description => [ModelingToolkit.getdescription(p) for p in params]
)
13×3 DataFrame
RowNameUnitsDescription
StringStringString
1TKTemperature
2k_Bm² kg s⁻² K⁻¹Boltzmann constant
3N_Avmol⁻¹Avogadro's number
4M_Akg mol⁻¹Molecular weight of species A
5M_airkg mol⁻¹Molecular weight of air
6Pm⁻¹ kg s⁻²Pressure
7Rm² kg s⁻² K⁻¹ mol⁻¹Universal gas constant
8μm⁻¹ kg s⁻¹Dynamic viscosity of air
9R_pmParticle radius
10αAccommodation coefficient (dimensionless)
11D_gm² s⁻¹Gas-phase diffusivity
12c_infm⁻³ molBulk gas-phase concentration
13c_sm⁻³ molSurface concentration

Equations

equations(sys)
7-element Vector{Symbolics.Equation}:
 c_bar(t) ~ sqrt((8N_Av*T*k_B) / (M_A*π))
 ρ_air(t) ~ (M_air*P) / (R*T)
 λ(t) ~ (2μ) / (sqrt((8R*T) / (M_air*π))*ρ_air(t))
 Kn(t) ~ λ(t) / R_p
 f_FS(t) ~ (0.75(1 + Kn(t))*α) / (Kn(t) + 0.75α + Kn(t)^2 + 0.283Kn(t)*α)
 J_c(t) ~ 12.566370614359172D_g*R_p*(c_inf - c_s)
 J(t) ~ J_c(t)*f_FS(t)

Characteristic Timescales

The module also provides components for calculating characteristic timescales of various mass transport processes:

Aerosol.GasDiffusionTimescale — Function
GasDiffusionTimescale(; name=:GasDiffusionTimescale)

Characteristic timescale for gas-phase diffusion to a particle.

This is the time required for a gas-phase concentration perturbation to diffuse to the particle surface.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eq. 12.49

source
Aerosol.AqueousDiffusionTimescale — Function
AqueousDiffusionTimescale(; name=:AqueousDiffusionTimescale)

Characteristic timescale for aqueous-phase diffusion within a droplet.

This is the time for a dissolved species to diffuse throughout the droplet interior.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eq. 12.75

source
Aerosol.InterfacialTimescale — Function
InterfacialTimescale(; name=:InterfacialTimescale)

Characteristic timescale to achieve equilibrium at the gas-particle interface.

For very soluble gases, the timescale is controlled by interfacial transport. For low solubility gases, the timescale is controlled by aqueous diffusion.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eqs. 12.61-12.62

source
Aerosol.ReactionTimescale — Function
ReactionTimescale(; name=:ReactionTimescale)

Characteristic timescale for aqueous-phase chemical reactions.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eqs. 12.76-12.78

source
Aerosol.SolidEquilibrationTimescale — Function
SolidEquilibrationTimescale(; name=:SolidEquilibrationTimescale)

Characteristic timescale for a gas to equilibrate with solid aerosol particles.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eqs. 12.135, 12.139

source
Aerosol.AqueousEquilibrationTimescale — Function
AqueousEquilibrationTimescale(; name=:AqueousEquilibrationTimescale)

Characteristic timescale for a gas to equilibrate with aqueous aerosol particles.

This accounts for the additional buffering effect of Henry's law dissolution.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eq. 12.147

source

Timescale Comparison

using Aerosol

# Calculate timescales for a 10 μm droplet
sys_dg = GasDiffusionTimescale()
sys_da = AqueousDiffusionTimescale()

compiled_dg = mtkcompile(sys_dg)
compiled_da = mtkcompile(sys_da)

R_p = 1.0e-5  # 10 μm
D_g = 2.0e-5  # m²/s (gas-phase diffusivity)
D_aq = 1.0e-9  # m²/s (aqueous-phase diffusivity)

prob_dg = NonlinearProblem(compiled_dg, Dict(compiled_dg.R_p => R_p, compiled_dg.D_g => D_g))
τ_dg = solve(prob_dg)[compiled_dg.τ_dg]

prob_da = NonlinearProblem(compiled_da, Dict(compiled_da.R_p => R_p, compiled_da.D_aq => D_aq))
τ_da = solve(prob_da)[compiled_da.τ_da]

println("Gas-phase diffusion timescale: $(round(τ_dg * 1e6, digits=2)) μs")
println("Aqueous-phase diffusion timescale: $(round(τ_da * 1e3, digits=2)) ms")
println("Ratio τ_da/τ_dg: $(round(τ_da/τ_dg, digits=0))")
Gas-phase diffusion timescale: 1.25 μs
Aqueous-phase diffusion timescale: 10.13 ms
Ratio τ_da/τ_dg: 8106.0

Aqueous-Phase Chemistry Coupling

For droplet chemistry applications, the module provides components for mass transport limitations and coupled gas-aqueous dynamics:

Aerosol.AqueousDiffusionReaction — Function
AqueousDiffusionReaction(; name=:AqueousDiffusionReaction)

Calculate the correction factor Q for aqueous-phase reaction rate when diffusion within the droplet limits the reaction.

The factor Q accounts for concentration gradients within the droplet due to fast reactions near the surface depleting reactants.

Reference: Seinfeld & Pandis (2006) Chapter 12, Eq. 12.110

source
Aerosol.MassTransportLimitation — Function
MassTransportLimitation(; name=:MassTransportLimitation)

Evaluate mass transport limitation criteria for aqueous-phase chemistry.

Computes the limiting values of k₁H* below which mass transport does not limit the aqueous-phase reaction rate (within 10%).

Reference: Seinfeld & Pandis (2006) Chapter 12, Eqs. 12.85, 12.86, 12.93

source
Aerosol.DropletMassBalance — Function
DropletMassBalance(; name=:DropletMassBalance)

Coupled gas-phase and aqueous-phase mass balance for cloud droplet chemistry.

This model couples:

  • Gas-phase partial pressure evolution
  • Aqueous-phase concentration evolution
  • Mass transfer between phases
  • Aqueous-phase chemical reaction

Reference: Seinfeld & Pandis (2006) Chapter 12, Eqs. 12.122-12.123

source

Mass Transport Limitation Criteria

The following example calculates the limiting reaction rates below which mass transport does not limit aqueous-phase chemistry:

using Aerosol

sys = MassTransportLimitation()
compiled = mtkcompile(sys)

# Parameters for a 10 μm droplet at 298 K
prob = NonlinearProblem(compiled,
    Dict(
        compiled.R_p => 1.0e-5, compiled.D_g => 2.0e-5, compiled.D_aq => 1.0e-9,
        compiled.T => 298.15, compiled.α => 1.0, compiled.M_A => 0.029))
sol = solve(prob)

println("Gas-phase diffusion limit for k₁H*: $(sol[compiled.k1H_gas_limit]) mol/(m³·Pa·s)")
println("Aqueous-phase diffusion limit for k₁: $(sol[compiled.k1_aq_limit]) s⁻¹")
println("Interfacial transport limit for k₁H*: $(sol[compiled.k1H_interface_limit]) mol/(m³·Pa·s)")
Gas-phase diffusion limit for k₁H*: 24.205074101615562 mol/(m³·Pa·s)
Aqueous-phase diffusion limit for k₁: 9.869604401089358 s⁻¹
Interfacial transport limit for k₁H*: 1411.5943281130376 mol/(m³·Pa·s)

Analysis

Comparison of Transition Regime Formulas

Figure 12.2 in Seinfeld & Pandis compares different transition regime formulas. Here we compare Fuchs-Sutugin and Dahneke:

sys_fs = FuchsSutugin()
sys_d = Dahneke()
compiled_fs = mtkcompile(sys_fs)
compiled_d = mtkcompile(sys_d)

Kn_range = 10 .^ range(-2, 2, length = 100)
f_FS = Float64[]
f_D = Float64[]

for Kn in Kn_range
    prob_fs = NonlinearProblem(compiled_fs, Dict(compiled_fs.Kn => Kn, compiled_fs.α => 1.0))
    push!(f_FS, solve(prob_fs)[compiled_fs.f_FS])

    prob_d = NonlinearProblem(compiled_d, Dict(compiled_d.Kn => Kn, compiled_d.α => 1.0))
    push!(f_D, solve(prob_d)[compiled_d.f_D])
end

plot(Kn_range, f_FS, label = "Fuchs-Sutugin", xscale = :log10,
    xlabel = "Knudsen Number (Kn)", ylabel = "J/Jc",
    title = "Comparison of Transition Regime Formulas (α = 1)")
plot!(Kn_range, f_D, label = "Dahneke", linestyle = :dash)
savefig("transition_comparison.svg");

Transition regime formula comparison

Both formulas agree in the limiting regimes but show slight differences in the transition regime (Kn ~ 1).

Particle Size Dependence

The mass transfer regime depends strongly on particle size through the Knudsen number:

sys = MassTransfer()
compiled = mtkcompile(sys)

# Calculate for particle diameters from 10 nm to 10 μm
D_p_range = 10 .^ range(-8, -5, length = 50)  # meters
R_p_range = D_p_range ./ 2

Kn_vals = Float64[]
f_FS_vals = Float64[]

for R_p in R_p_range
    prob = NonlinearProblem(compiled,
        Dict(
            compiled.R_p => R_p, compiled.α => 1.0,
            compiled.c_inf => 1.0e-6, compiled.c_s => 0.0))
    sol = solve(prob)
    push!(Kn_vals, sol[compiled.Kn])
    push!(f_FS_vals, sol[compiled.f_FS])
end

plot(D_p_range .* 1e6, Kn_vals, label = "Knudsen Number", xscale = :log10, yscale = :log10,
    xlabel = "Particle Diameter (μm)", ylabel = "Kn or J/Jc",
    title = "Particle Size Dependence")
plot!(D_p_range .* 1e6, f_FS_vals, label = "J/Jc (Fuchs-Sutugin)")
hline!([1.0], label = "Kn = 1 (transition)", linestyle = :dash, color = :gray)
savefig("size_dependence.svg");
┌ Warning: Assignment to `prob` in soft scope is ambiguous because a global variable by the same name exists: `prob` will be treated as a new local. Disambiguate by using `local prob` to suppress this warning or `global prob` to assign to the existing global variable.
└ @ mass_transfer.md:234
┌ Warning: Assignment to `sol` in soft scope is ambiguous because a global variable by the same name exists: `sol` will be treated as a new local. Disambiguate by using `local sol` to suppress this warning or `global sol` to assign to the existing global variable.
└ @ mass_transfer.md:238

Particle size dependence

Small particles (< 100 nm) are in the kinetic or transition regime where mass transfer is significantly reduced compared to the continuum limit.