Ionic Strength and the Concept of Activity

Every pH meter calibration relies on activity, not concentration: certified pH buffer standards (e.g., NIST/NBS phosphate or phthalate buffers) are assigned pH values computed from the activity of $\mathrm{H^+}$, using an ionic-strength-based convention, because a glass electrode genuinely responds to $a_{\mathrm{H^+}}$, not $[\mathrm{H^+}]$. Ignoring this distinction is why "0.0100 M HCl has pH 2.00" is only approximately true — the measured pH of such a solution is closer to 2.04, the difference being exactly the activity correction discussed here. In environmental and geochemical analysis, natural waters — from freshwater lakes ($\mu \sim 10^{-3}$ M) to estuarine and seawater samples ($\mu \sim 0.7$ M) — span the entire range over which activity corrections go from negligible to dominant, and models like the Davies equation or Pitzer's ion-interaction formalism (extensions of the ideas here) are load-bearing in predicting mineral solubility, trace-metal speciation, and carbonate equilibria in the ocean's buffering system. In clinical chemistry, ion-selective electrodes for $\mathrm{Na^+}$, $\mathrm{K^+}$, and $\mathrm{Ca^{2+}}$ in blood plasma report activities directly; because physiological ionic strength is fixed and well characterized (roughly $0.15$ M), clinical reference ranges are calibrated against activity standards rather than raw concentration standards, and confusing the two has historically caused real diagnostic discrepancies between "total calcium" and "ionized (free, active) calcium" measurements. In pharmaceutical formulation, solubility and dissolution testing of drug substances is performed in simulated physiological fluids at fixed ionic strength precisely because the diverse-ion effect can shift apparent solubility (and hence bioavailability predictions) by tens of percent relative to solubility measured in pure water — a direct, high-stakes descendant of the AgCl-in-KNO₃ experiment that opened this topic. In precise potentiometric and electroanalytical work — the Nernst equation itself, ion-selective electrode calibration, and redox titrations — every formal potential $E^\circ{}'$ tabulated for "real" laboratory conditions already bakes in an assumed or measured ionic strength, because the Nernst equation, like every equilibrium expression here, is fundamentally written in activities.

Ask the tutor: How do I actually calculate a numerical value for an activity coefficient γᵢ once I know the ionic strength, rather than just being told its approximate value?

Picture a $\mathrm{Ba(NO_3)_2}$ solution with a trace of dissolved $\mathrm{BaSO_4}$ in equilibrium with the solid. Each $\mathrm{Ba^{2+}}$ ion in solution is not isolated — it sits, on time average, inside a diffuse cloud of excess $\mathrm{NO_3^-}$ and $\mathrm{SO_4^{2-}}$ charge that partially cancels its own charge as seen from a distance. This cloud is the ionic atmosphere. Building it costs no work at infinite dilution (there's no cloud to build), but at finite ionic strength the electrostatic attraction between an ion and its atmosphere lowers the ion's free energy relative to the ideal (no-interaction) reference state. A lower free energy means a lower "escaping tendency" — the ion is thermodynamically more comfortable in solution, less inclined to leave it (e.g., by precipitating, by neutralizing an acid, or by being reduced at an electrode) than a naive concentration count would suggest. The quantity that measures "escaping tendency" rigorously is the chemical potential,

and because the electrostatic stabilization described above lowers $\mu_i$ below the ideal value $\mu_i^\circ + RT\ln[i]$, the activity $a_i$ must be smaller than the concentration $[i]$ whenever the ion is stabilized by its atmosphere — which, for the moderate ionic strengths of ordinary lab solutions, is essentially always the case for charged species. We write this as $a_i = \gamma_i [i]$ with $\gamma_i < 1$ typically, approaching $\gamma_i = 1$ only as the solution becomes so dilute that ions rarely feel one another at all. The strength of this atmosphere effect obviously depends on how much charge is floating around and how highly charged each species is — a $2+$ ion polarizes its surroundings far more violently than a $1+$ ion at the same molar concentration, because the electrostatic interaction energy scales with the product of charges, and building an atmosphere around a $z=2$ ion involves twice the charge doing twice the work. This is why ionic strength weights each ionic species by $z_i^2$ rather than by $z_i$ or not at all.

The single number engineers and chemists use to summarize "how much charge is floating around" is the ionic strength,

summed over every ion in solution — not just the ones appearing in the equilibrium of interest. A spectator salt like KNO₃ raises $\mu$ just as effectively as NaCl does, which is precisely why adding "inert" KNO₃ to a AgCl suspension increases its solubility: it thickens every ion's atmosphere and lowers every $\gamma_i$, including those of $\mathrm{Ag^+}$ and $\mathrm{Cl^-}$, even though $\mathrm{K^+}$ and $\mathrm{NO_3^-}$ never appear in the solubility reaction itself. This is traditionally called the diverse-ion effect or salt effect, to distinguish it from the common-ion effect.

The activity of species $i$ is defined relative to the molar concentration $[i]$ by

where:

• $a_i$ — the (dimensionless) activity of species $i$, the thermodynamically "correct" quantity for equilibrium expressions;
• $\gamma_i$ — the activity coefficient of species $i$, dimensionless, a function of ionic strength (and, weakly, temperature and specific ion identity);
• $[i]$ — the molar (equilibrium) concentration of species $i$, mol L⁻¹;
• $c^\circ = 1$ mol L⁻¹ — the standard-state concentration, included to make $a_i$ dimensionless; conventionally suppressed in numerical work by simply writing $a_i = \gamma_i[i]$ with $[i]$ understood as a pure number equal to the molarity.

Reading: activity is concentration scaled by a "how ideal is this ion's environment" factor. When $\gamma_i = 1$ the ion behaves exactly as concentration predicts; as $\gamma_i$ falls below 1, the ion is less reactive/available per mole present than an ideal picture would suggest, because part of its free energy is already "spent" stabilizing its ionic atmosphere. The ionic strength that governs every $\gamma_i$ in the solution is

where:

• $\mu$ — ionic strength, mol L⁻¹ (some texts use $I$ for this symbol);
• $c_i$ — molar concentration of the $i$-th ion (all ions, not just reactants), mol L⁻¹;
• $z_i$ — charge number of the $i$-th ion (sign irrelevant since it is squared);
• the factor $\tfrac12$ compensates for summing over both cations and anions, so that a 1:1 electrolyte at concentration $c$ gives $\mu = c$ exactly, matching the historical convention.

Reading: ionic strength is a charge-weighted total ion concentration; doubly and triply charged ions contribute out of proportion to their molarity (a factor of 4 or 9), because they build thicker, more stabilizing atmospheres. The true equilibrium law, exact at any concentration, is written in activities:

where:

• $K$ — the thermodynamic equilibrium constant, a true constant depending only on temperature (and pressure), independent of ionic strength;
• $\nu_i$ — the stoichiometric coefficient of species $i$ in the balanced reaction (positive for products, negative for reactants);
• the product runs over every species with a nontrivial activity (pure solids/liquids and solvent water are assigned $a=1$ in dilute solution, so they drop out).

Reading: everything you were taught as "the equilibrium constant expression" using bare concentrations is really this expression with every $\gamma_i$ silently set to 1. The concentration constant actually measured or used in a specific solution,

is therefore not a universal constant — it depends on $\mu$ through the $\gamma_i$'s, and only converges to $K$ in the limit $\mu \to 0$.

Example 1 (foundational) — Ionic strength of a simple mixed solution. Compute the ionic strength of a solution that is $0.0150$ M in $\mathrm{KBr}$ and $0.0050$ M in $\mathrm{K_2SO_4}$. Step 1: list every ion and its concentration. KBr fully dissociates to $\mathrm{K^+}$ and $\mathrm{Br^-}$, each at $0.0150$ M. $\mathrm{K_2SO_4}$ dissociates to $2(0.0050) = 0.0100$ M $\mathrm{K^+}$ and $0.0050$ M $\mathrm{SO_4^{2-}}$. Total $\mathrm{K^+}$ concentration (both sources add) is $0.0150 + 0.0100 = 0.0250$ M. Step 2: apply the definition.

Sanity/limiting check: if the $\mathrm{K_2SO_4}$ were replaced by an equal molar amount of a 1:1 salt like KCl ($0.0050$ M, contributing $0.0050+0.0050=0.0100$ to the bracket instead of $0.0200$), $\mu$ would drop to $0.0250$ M — confirming that the $z^2=4$ weighting of $\mathrm{SO_4^{2-}}$, not merely its molarity, is responsible for the extra ionic strength. The arithmetic is fully asymmetric (three different concentrations, mixed charge types), so no accidental cancellation is hiding an error.

Example 2 (applied) — Correcting a buffer pH calculation for ionic strength. An acetate buffer is prepared at $0.0400$ M acetic acid ($K_a = 1.75\times10^{-5}$, a thermodynamic constant) and $0.0600$ M sodium acetate, then $0.100$ M KCl is added as supporting electrolyte (common in electrochemical work to fix ionic strength for a reference electrode). Estimate the pH, accounting for activity effects on the acetate/acetic acid equilibrium. Take $\gamma_{\mathrm{H^+}} = 0.83$ and $\gamma_{\mathrm{OAc^-}} = 0.77$ at the resulting ionic strength (typical single-ion activity coefficients at $\mu \approx 0.13$ M, as would be obtained from an extended Debye–Hückel expression); acetic acid is neutral, so $\gamma_{\mathrm{HOAc}} \approx 1$. Step 1: write the thermodynamic equilibrium exactly.

Step 2: solve for $[\mathrm{H^+}]$, using the buffer concentrations directly as $[\mathrm{HOAc}]$ and $[\mathrm{OAc^-}]$ (a good approximation since the buffer ratio is not extreme).

Step 3: evaluate numerically.

Step 4: compare with the naive (activity-ignoring) calculation. Setting all $\gamma_i = 1$ gives $[\mathrm{H^+}] = K_a(0.0400/0.0600) = (1.75\times10^{-5})(0.667) = 1.167\times10^{-5}$ M, pH $=4.93$. The activity correction shifts the predicted pH by $+0.19$ unit here — a difference well beyond typical pH-meter reproducibility of $\pm 0.02$ — showing that ionic-strength effects on buffer pH are not academic even at "ordinary" lab concentrations. As $\mu\to0$ (no added KCl, dilute buffer), both $\gamma$'s $\to1$ and the two calculations converge, confirming the correction vanishes exactly in the ideal limit.

Example 3 (exam/competition level) — Iterative ionic strength for a solubility problem with self-consistency. Estimate the molar solubility $s$ of $\mathrm{PbI_2}$ ($K_{sp} = 7.1\times10^{-9}$, thermodynamic) in pure water, accounting for the fact that the ionic strength itself is generated by the dissolved $\mathrm{Pb^{2+}}$ and $\mathrm{I^-}$ — a self-consistency loop that appears often in real solubility work (and is exactly the kind of problem that motivates the systematic activity-coefficient equations of Debye-Huckel and Activity Coefficients). Step 1: zeroth iteration — ignore activity effects entirely.

Step 2: compute the ionic strength generated by this first estimate.

Step 3: obtain activity coefficients at this ionic strength. At $\mu = 3.63\times10^{-3}$ M, extended Debye–Hückel-type estimates give approximately $\gamma_{\mathrm{Pb^{2+}}} \approx 0.74$ and $\gamma_{\mathrm{I^-}} \approx 0.94$ (a $z=2$ ion is far more strongly affected than a $z=1$ ion even at this fairly low ionic strength — again the $z^2$ weighting at work, this time acting on $\gamma$ rather than on $\mu$ itself). Step 4: recompute solubility using activity-corrected $K_{sp}$.

Evaluate the denominator: $(0.94)^2 = 0.8836$; $4(0.74)(0.8836) = 2.616$.

Step 5: check convergence (one more loop). Using $s_1$ instead of $s_0$ changes $\mu$ only from $3.63\times10^{-3}$ to about $4.18\times10^{-3}$ M — a small shift that moves $\gamma$ values by only a percent or two, so $s$ converges to about $1.40\times10^{-3}$ M after a second iteration, essentially unchanged from $s_1$. The final answer, $s \approx 1.40\times10^{-3}$ M, is roughly 15% larger than the naive $s_0 = 1.21\times10^{-3}$ M — a real, measurable difference arising entirely from ionic-strength effects generated by the very ions being counted, with no outside electrolyte added at all. As a limiting-case check: if $K_{sp}$ described a pair of singly-charged ions instead (both $z=1$, so $\mu$ stays much smaller for the same $s$), the correction would shrink dramatically, consistent with $z^2$ weighting making $2{:}1$ and $2{:}2$ electrolytes far more sensitive to self-generated ionic strength than $1{:}1$ salts.

The modern quantitative language of ionic strength and activity coefficients was built by Peter Debye and Erich Hückel, but the conceptual groundwork — the recognition that concentration is not the same as thermodynamic "effective concentration" — belongs to Gilbert N. Lewis, who introduced the term activity itself in 1907 while trying to rescue the elegant mass-action law from the mounting experimental evidence that real electrolyte solutions disobeyed it. Lewis, working at MIT and later Berkeley, was relentlessly empirical: rather than abandon the beautifully simple $K = \prod[i]^{\nu_i}$ form, he proposed keeping its mathematical shape intact but replacing concentration with a new quantity, defined operationally through measurable chemical potentials, that would make the law exact again. This was a characteristically Lewis move — he did the same thing later with the generalized acid-base definition that bears his name — preferring to redefine a concept precisely rather than patch an equation with ad hoc correction terms. Lewis and his student Merle Randall then showed, empirically, that for dilute electrolytes the activity coefficient of an ion depends mainly on a single combined variable — the ionic strength — rather than on the detailed identity and concentration of every ion separately. This was a genuinely surprising simplification: two wildly different mixtures of salts, so long as they produced the same $\mu$, gave approximately the same $\gamma_i$ for a given ion. It was this empirical regularity, published in Lewis and Randall's 1921 paper and later their landmark 1923 textbook Thermodynamics and the Free Energy of Chemical Substances, that Debye and Hückel explained from first principles a few years later using a statistical-mechanical model of the ionic atmosphere.