Mass transfer by migration & diffusion (Ch.
4)
Mass transfer equation
Migration
Mixed migration & diffusion near an electrode
Mass transfer during electrolysis
Effect of excess electrolyte
Diffusion
Microscopic view
Fick’s laws
Boundary conditions in electrochemical problems
Solution of diffusion equations
Mass transfer equation
Mass transfer by diffusion, migration, convection
- Diffusion & migration result from a gradient in electrochemical potential, μ
- Convection results from an imbalance of forces on the solution
Two points in solution; r & s
→ difference of μj
due to conc. & electric field (φ) differences
Flux Jj ∝ gradμj or Jj ∝∇μj
1-D: ∇ = i(∂/∂x)
3-D: ∇ = i(∂/∂x) + j(∂/∂y) + k(∂/∂z)
Jj = -(CjDj/RT)∇μj
Minus sign: flux direction opposite the direction of increasing μj
If solution moving with a velocity v
Jj = -(CjDj/RT)∇μj + Cjv
Nernst-Planck equations
Jj(x) = -Dj(∂Cj(x)/∂x) –(zjF/RT)DjCj(∂φ(x)/∂x) + Cjv(x)
In general
Jj = -Dj∇Cj – (zjF/RT)DjCj∇φ + Cjv
diffusion migration convection
Convection absent in this Chapter (Ch.9 for convection)
→ in an unstirred or stagnant solution
For linear system
-Jj (mols-1cm-2) = ij/zjFA [C/s per (Cmol-1cm2)] = id,j/zjFA + im,j/zjFA
With id,j/zjFA = Dj(∂Cj/∂x)
im,j/zjFA = (zjFDj/RT)Cj(∂φ/∂x)
Id,j & im,j: diffusion & migration currents of species j
Total current i = ∑ij
Migration
In the bulk soln (away from the electrode), conc gradient small: migration
ij = (zj2F2ADjCj/RT)(∂φ/∂x)
Einstein-Smoluchowski equation
mobility uj = ׀zj׀FDj/RT
ij = ׀zj׀FAujCj(∂φ/∂x)
For a linear electric field
∂φ/∂x = ΔE/l
ij = ׀zj׀FAujCjΔE/l
Total current i = ∑ij = (FAΔE/l)∑׀zj׀ujCj
Conductance (L) L = 1/R = i/ΔE = (FA/l)∑׀zj׀ujCj = Aκ/l
κ: conductivity (Ω-1cm-1)
κ = F∑׀zj׀ujCj
Resistivity ρ = 1/κ
Transference number
tj = ij/i = ׀zj׀ujCj/∑׀zk׀ukCk = ׀zj׀Cjλj/∑׀zk׀Ckλk
Mixed migration and diffusion near an active electrode
i = i d + im
Balance sheets for mass transfer during electrolysis
e.g., 4.1. Electrolysis of a solution of HCl at Pt electrode
λ+ (conductance of H+), λ- (conductance of Cl-): λ+ ~ 4λ- → t+ = 0.8, t- = 0.2
Assume total current of 10e/unit time producing 5H2 (cathode) & 5Cl2 (anode)
Total current in bulk soln: 8H+ + 2Cl-
→ diffusion of 2 additional H+ to cathode with 2Cl- for electroneutrality
diffusion of 8Cl- with 8H+
For H+: id = 2, im = 8, for Cl-: id = 8, im =2 → total current i = id + im = 10
(same direction of id & im)
For mixture of charged species, current by jth species ij = tji
→ # of moles of jth species migrating per sec = tji/zjF
→ # of moles arriving at the electrode per sec by migration = ±im/nF
(positive sign for reduction of j, negative sign for oxidation)
±im/nF = tji/zjF
im = ±(n/zj)tji
id = i – im = i(1 -/+ ntj/zj)
Negative sign for cathodic current, positive sign for anodic current)
e.g., 4.2. Electrolysis of a solution of 10-3 M Cu(NH3)42+, 10-3 M Cu(NH3)2+,
3 x 10-3 M Cl- in 0.1 M NH3 at two Hg electrodes
Assume λCu(II) = λCu(I) = λCl- = λ
From tj = ij/i = ׀zj׀ujCj/∑׀zk׀ukCk = ׀zj׀Cjλj/∑׀zk׀Ckλk
tCu(II) = 1/3, tCu(I) = 1/6, tCl- = 1/2
Assume total current of 6e/unit time, i = 6, n = 1
For Cu(II) at cathode: ׀im( = ׀1/2)(1/3)(6) = 1, id = 6 - 1 = 5,
for Cu(I) at anode: ׀im( = ׀1/1)(1/6)(6) = 1, id = 6 + 1 = 7
Effect of adding excess electrolyte
e.g., 4.3. Electrolysis of a solution of 10-3 M Cu(NH3)42+, 10-3 M Cu(NH3)2+,
3 x 10-3 M Cl- in 0.1 M NH3 + 0.1 M NaClO4 (as excess electrolyte) at two
Hg electrodes
Assume λNa+ = λClO4- = λ
→ tNa+ = tClO4- = 0.485, tCu(II) = 0.0097, tCu(I) = 0.00485, tCl- = 0.0146
*Na+ & ClO4- do not participate in e-transfer rxns, but because their conc are high,
they carry 97% of the current in the bulk solution
→ Most of Cu(II) reaches the cathode by diffusion & 0.5% flux by migration
Addition of an excess of nonelectroactive ions (a supporting electrolyte):
1. nearly eliminates the contribution of migration to the mass transfer of the
electroactive species → eliminate ∇φ or ∂φ/∂x in mass transfer equations
2. Decreases the solution resistance, improve the accuracy of WE potential
Disadvantage: impurities, altered medium
Diffusion
A microscopic view-discontinuous source model
Diffusion occurs by a “random walk” process
Probability, P(m, r) after m time units (m = t/τ)
P(m, r) = (m!/r!(m – r)!)(1/2)m
where x = (-m + 2r)l with r = 0, 1,…m
Mean square displacement of the molecule
Δ2 = ml2 =(t/τ)l2 = 2Dt
D: diffusion coefficient (= l2/2τ), cm2/s (Einstein in 1905)
Root-mean square displacement
Δ = √2Dt
→ estimating the thickness of diffusion layer
e.g., D = 5 x 10-5 cm2/s: diffusion layer thickness 10-4 cm in 1 ms, 10-3 cm in 0.1
s, 10-2 cm in 10 s
For N0 molecules at the original position at t = 0 → Gaussian distribution later
N(x, t) in a segment Δx wide on x
N(x, t)/N0 = (Δx/2√πDt)exp(-x2/4Dt)
2-D: Δ = √4Dt, 3-D: Δ = √6Dt
Diffusional velocity (vd)
vd = Δ/t = (2D/t)1/2
Migration (v = uiE) vs. diffusion (vd)
Einstein-Smoluchowski equation
mobility uj = ׀zj׀FDj/RT
v = ׀zj׀FDjE/RT (E: electric field strength)
v << vd DiE/(RT/׀zi׀F) << (2Di/t)1/2
rearrange (2Dit)1/2E << 2RT/׀zi׀F
→ voltage drop over length scale of diffusion << 2RT/׀zi׀F : migration negligible
Fick’s laws of diffusion
Fick’s 1st law: flux ∝ conc gradient
-JO(x, t) = DO(∂CO(x, t)/∂x)
Fick’s 2nd law: change in concentration of O with time
∂CO(x, t)/∂t = DO(∂2CO(x, t)/∂x2)
→ solution gives concentration profiles, CO(x, t)
General formulation of Fick’s 2nd law
∂CO/∂t = DO∇2CO
Fig. (a): Planar electrode (linear diffusion equation)
∂CO(x, t)/∂t = DO(∂2CO(x, t)/∂x2)
Fig. (b): spherical electrode (hanging Hg drop)
∂CO(r, t)/∂t = DO[(∂2CO(r, t)/∂x2) + (2/r)(∂CO(r, t)/∂r)]
Consider O transported purely by diffusion to an electrode
O + ne = R
-JO(0, t) = i/nFA = DO[∂CO(x, t)/∂x]x = 0
