^{1}, the diffusion coefficient is defined as the proportionality constant between the molar flux of the diffusing species and the concentration gradient, as shown below, \begin{equation} \label{eq:eq18} J=-D \nabla C \end{equation} where, $$\begin{eqnarray} J \equiv flux \nonumber \\ \nabla c \equiv concentration\;gradient \nonumber \\ D \equiv diffusion\;coefficient \nonumber \end{eqnarray}$$ In the Einstein relation, the diffusion coefficient of a particle undergoing Brownian motion is given by, \begin{equation} \label{eq:eq19} D=\frac{k_B T}{\xi} \end{equation} ξ is defined as a friction factor of the diffusing particles, k_B is Boltzmann’s constant, and T is the absolute temperature. For diffusion of spherical particles through a liquid, equation (B2) can be rewritten as the Stokes-Einstein equation2, (B3). D=(k_B T)/6πηr (B3) The friction factor has now been replaced by the fluid viscosity, η, and a relationship with the particle radius, r. For small solute molecules in water, Wilke and Chang3 proposed an empirical change to equation (B3), resulting in equation (B4). D=K (T√ϕM)/(ηV^0.6 ) (B4) Where K is a fitting coefficient, ϕ is an empirical parameter to account for association between solvent molecules, Mis the molecular mass of the solvent, and V is the molar volume of the solute. For water, the best agreement with diffusion coefficients suggested that ϕ=2.6. While there is a temperature dependence for the diffusion coefficient included in equation (B4), the viscosity of the solution also changes as a function of temperature. This temperature dependence has two principal effects4. The first effect can be captured with an Arrhenius expression, equation (B5). η_0=C_1 e^(C_2/T ) (B5) C_1 and C_2 are fitting coefficients. The second effect occurs because adding solute molecules to a solvent can change the solvent structure and thereby influence its viscosity. From, Jones and Dole5, the effect of the solute on viscosity can be estimated using equation (B6). η_1=1+A'√c+B'c (B6) A' is a parameter to estimate ion-ion interactions6 and B', known as the Jones-Dole B coefficient7, captures solvent-solute interactions. The viscosity of the solvent is then estimated from equation (B7). η=η_0 η_1 (B7) In this approach, then, η_0 from equation (B5) must be unitless, with η_1in units of mPa s. Using equations (B4), (B5), (B6), and (B7), and fitting to diffusion coefficient data obtained from the literature8, results in the plot shown in Figure B1. Figure B1. Plots of dissolved O2 diffusion coefficient data at various temperatures and chloride concentrations (symbols) overlaid with estimated diffusion coefficients (solid lines). The temperature dependence of each of the parameters from equations (B4) - (B7) can then be estimated by fitting them to a Padé approximant model, as given in equation (B7). p_i=(b_1+b_2 x)/(1.0+b_3 x) (B7) p_i indicates the relevant parameter and b_iare the fitting coefficients. The temperature-dependent model fits for these parameters are shown in Figure B2. a. b. c. d. Figure B2. Plots of the temperature dependence of parameters a. K, b.A', c.B', and d. C_1 and C_2. Thus, the diffusion coefficient for dissolved O2 can be calculated for other temperatures and chloride concentrations. An estimate of the diffusion coefficient, compared with several literature values9, is shown in Figure B3. Figure B3. Plot of the modeled dissolved O2 diffusion coefficient at T = 25oC for a range of Cl- concentrations from 0.0 M to 5.0 M NaCl (solid line) compared with experimental data (+ symbols).

### Sources

- Aifantis, E. C., On the problem of diffusion in solids. Acta Mechanica 1980, 37 (3), 265-296.
- Einstein, A., Über die von der molekularkinetischen Theorie der Wärme geforderte Bewegung von in ruhenden Flüssigkeiten suspendierten Teilchen. Annalen der Physik 1905, 322 (8), 549-560.
- Wilke, C. R.; Chang, P., Correlation of diffusion coefficients in dilute solutions. AIChE Journal 1955, 1 (2), 264-270.
- Reinsberg, P. H.; Bawol, P. P.; Thome, E.; Baltruschat, H., Fast and Simultaneous Determination of Gas Diffusivities and Solubilities in Liquids Employing a Thin-Layer Cell Coupled to a Mass Spectrometer, Part II: Proof of Concept and Experimental Results. Analytical Chemistry 2018, 90 (24), 14150-14155.
- Jones, G.; Dole, M., THE VISCOSITY OF AQUEOUS SOLUTIONS OF STRONG ELECTROLYTES WITH SPECIAL REFERENCE TO BARIUM CHLORIDE. Journal of the American Chemical Society 1929, 51 (10), 2950-2964.
- Falkenhagen, H.; Vernon, E. L., LXII. The viscosity of strong electrolyte solutions according to electrostatic theory. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 1932, 14 (92), 537-565.
- Jenkins, H. D. B.; Marcus, Y., Viscosity B-Coefficients of Ions in Solution. Chemical Reviews 1995, 95 (8), 2695-2724.
- van Stroe, A. J.; Janssen, L. J. J., Determination of the diffusion coefficient of oxygen in sodium chloride solutions with a transient pulse technique. Analytica Chimica Acta 1993, 279 (2), 213-219.
- Ho, C. S.; Ju, L.-K.; Baddour, R. F.; Wang, D. I. C., Simultaneous measurement of oxygen diffusion coefficients and solubilities in electrolyte solutions with a polarographic oxygen electrode. Chemical Engineering Science 1988, 43 (11), 3093-3107.