Voltammetry Fundamentals

A microelectrode dipped in a stirred solution can report a concentration from a single current reading. That is the quiet promise of voltammetry: apply a controlled potential, measure the resulting current. The current-potential relationship reveals both the identity of a species and its concentration. Identity comes from the potential at which current appears. Concentration comes from the size of that current. The idea sounds almost too simple. Yet making it work reliably required solving a real instrumentation problem. That problem was how to control the potential of a working electrode. It had to be controlled without the applied current itself corrupting the measurement. This matters because voltammetry underlies an enormous swath of modern chemical sensing. Examples include blood-glucose meters and dissolved-oxygen probes, and trace-metal analysis in environmental water samples. Another example is the Cyclic Voltammetry experiments that determine redox potentials of new battery materials. Unlike potentiometry, which measures an equilibrium potential at zero current, voltammetry deliberately drives current. It studies current's dependence on potential and time. That dependence is governed by two competing physical processes at the electrode surface. One is the faradaic electron-transfer reaction the chemist wants to measure. The other is the non-faradaic charging of the electrode's electrical double layer. This charging behaves like a capacitor in parallel with the reaction. The big picture is a three-way negotiation between electrochemistry, electronics, and mass transport. Electrochemistry, via the Nernst Equation and its kinetic generalizations, sets how the surface concentrations and electron-transfer rate depend on potential. Electronics — specifically the three-electrode cell driven by a potentiostat — makes it possible to fix that potential accurately. This holds even as current flows and solution resistance would otherwise distort it. Mass transport — diffusion, migration, and convection — sets how fast fresh analyte reaches the electrode surface. This sustains the current. It therefore sets the current's magnitude and its dependence on time. This article builds the three-electrode cell from first principles. It separates faradaic current from the charging current that competes with it. It also derives the mass-transport laws that convert a measured current into a concentration. These are Fick's first law, the Cottrell equation, and the Levich equation. Every relationship is anchored to a worked numerical example with realistic magnitudes. These include microamp currents, millimolar concentrations, and cm²/s diffusion coefficients. This way, the numbers, not just the algebra, can be checked.

Key takeaways

A three-electrode cell (working, reference, auxiliary) lets a potentiostat fix the working electrode's potential accurately while current flows, because the reference carries no current and the auxiliary absorbs it.
Total measured current is the sum of faradaic current (electron transfer at the surface) and charging current (double-layer capacitance responding to a changing potential).
Faradaic current is limited by how fast analyte reaches the surface — mass transport — described by Fick's first law and, for time-dependent experiments, the Cottrell equation.
Charging current decays exponentially with the cell's $RC$ time constant, while faradaic current from diffusion decays as $t^{-1/2}$ — this different time dependence is what lets instruments separate the two.
Forced convection (a rotating disk electrode) gives a steady, calculable mass-transport-limited current via the Levich equation, replacing a decaying transient with a constant signal.
Migration of charged analyte in the electric field is suppressed by adding excess inert electrolyte, so that diffusion alone controls transport of the analyte ion.
Key idea: voltammetry converts a controlled potential into a diffusion-limited current whose magnitude and time-dependence, once separated from capacitive charging, are direct, calculable measures of concentration.

Where this lives today

Voltammetric principles built on the three-electrode cell and mass-transport theory are the operating basis of the personal blood-glucose meter, in which an enzyme-mediated amperometric sensor (a fixed-potential relative of the methods here, foreshadowing Amperometry and the Clark Electrode) reports blood glucose from a diffusion-limited current at a disposable screen-printed electrode, calibrated against known Cottrell-type or thin-layer transport equations. Dissolved-oxygen sensors used in wastewater treatment, aquaculture, and clinical blood-gas analyzers are built on the Clark oxygen electrode, whose steady-state current is governed by oxygen diffusion through a polymer membrane in direct analogy to the mass-transport equations derived here. Trace heavy-metal analysis in environmental and food samples (Pb²⁺, Cd²⁺, Hg²⁺ at parts-per-billion levels) uses anodic stripping voltammetry, in which a controlled deposition step (mass-transport-enhanced, often with a rotating or stirred electrode) preconcentrates metal at the electrode before a stripping potential sweep — governed by the same faradaic/charging-current competition and the same Levich-type convective transport theory. In battery and fuel-cell research, rotating disk electrode voltammetry with the Koutecký–Levich analysis of Example 3 is the standard method for separating intrinsic electrocatalytic kinetics from mass-transport limitations when screening new catalyst materials for oxygen reduction or hydrogen evolution. And essentially every modern potentiostat, from benchtop research instruments to portable field units, is built around the three-electrode feedback architecture described here, with hardware iR-compensation circuits specifically engineered to minimize the $R_u$ that would otherwise distort the applied potential — a direct descendant of the charging-current analysis worked through above.

Ask the tutor: How would the Cottrell and Levich equations need to be modified if the electrode reaction were only partially reversible (mixed kinetic-diffusion control) rather than fully diffusion-limited?

The physical picture — thorough prose

Consider a small platinum disk (the working electrode, WE) immersed in a solution containing an electroactive species, say Fe³⁺, plus a large excess of an inert supporting electrolyte such as 0.1 M KNO₃. A potentiostat holds the WE at a controlled potential $E$ relative to a reference electrode (RE, typically Ag/AgCl or SCE) that draws essentially zero current and therefore maintains a fixed, known potential exactly as it does in potentiometry. Current cannot flow through the reference without polarizing it, so a third electrode — the auxiliary or counter electrode (CE), often a platinum wire or coil — completes the circuit and carries whatever current the cell produces. The potentiostat's feedback electronics continuously adjust the CE potential so that the WE–RE potential difference tracks the commanded value $E$ , regardless of the ohmic drop $iR_u$ that would otherwise appear across the uncompensated solution resistance $R_u$ between WE and RE. When $E$ is made sufficiently negative (for a reduction), Fe³⁺ near the electrode surface is reduced to Fe²⁺. This consumes Fe³⁺ at the surface faster than diffusion from bulk solution can replace it, so a concentration gradient develops — a thin diffusion layer of depleted Fe³⁺ adjacent to the electrode. Current flowing because of this net chemical transformation is faradaic current, $i_f$ , and it is what the chemist wants to measure, since by Faraday's law it is stoichiometrically tied to a rate of reaction and hence, at steady state, to bulk concentration. But there is a second contribution. Every time the potential $E$ is changed, charge must rearrange at the electrode–solution interface: solvent dipoles reorient, ions in the diffuse and compact layers adjust their positions, and the electrode surface itself accumulates or sheds charge. This interface behaves electrically like a capacitor of capacitance $C_d$ (the double-layer capacitance, typically 10–40 μF/cm²) in series with the solution resistance $R_u$ . Charging or discharging this capacitor requires charging current (also called capacitive or non-faradaic current), $i_c$ , which flows even in a solution with no electroactive species at all — it is a pure instrumental artifact of changing the potential, not a chemical signal. The measured current at any instant is the sum:

i_{\text{meas}}(t) = i_f(t) + i_c(t)

Separating these two is the central practical challenge of dynamic voltammetry. The key is that they have different time signatures. Charging current, driven by an $RC$ circuit, decays exponentially with time constant $\tau = R_u C_d$ (typically microseconds to low milliseconds). Faradaic current sustained by semi-infinite linear diffusion decays much more slowly, as $t^{-1/2}$ (the Cottrell decay). At long enough times after any potential step, $i_c$ has vanished while $i_f$ persists — this is exploited directly in techniques like normal-pulse and differential-pulse polarography, and it is the reason Polarography samples current late in a drop's life.

Figure

The three-electrode cell: potentiostat feedback loop

Mass transport of Fe³⁺ to the electrode can occur by three mechanisms: diffusion (motion down a concentration gradient), migration (motion of charged species in the electric field between WE and CE), and convection (bulk fluid motion, natural or forced). Analytical voltammetry is deliberately engineered so that only diffusion carries the analyte signal. Migration is suppressed by adding a large excess (typically ≥ 50-fold) of supporting electrolyte, which carries essentially all the migrational current and collapses the electric field gradient outside a thin double layer, leaving the bulk solution field-free for the trace analyte. Convection is either eliminated (quiescent solution, for transient techniques) or imposed in a controlled, calculable way (rotating disk electrode, for steady-state techniques).

Figure

Diffusion layer growth and the resulting current-time decay

The core ideas — governing equations

Fick's first law relates the flux of an electroactive species to its concentration gradient at the electrode surface, and Faraday's law converts that flux into current:

i_f(t) = nFAD\left(\frac{\partial C(x,t)}{\partial x}\right)_{x=0}

where:

$i_f(t)$ = faradaic current at time $t$ (A)
$n$ = number of electrons transferred per molecule/ion reacting
$F$ = Faraday constant, $96{,}485\ \text{C mol}^{-1}$
$A$ = electrode area (cm²)
$D$ = diffusion coefficient of the electroactive species (cm² s⁻¹)
$\left(\partial C/\partial x\right)_{x=0}$ = concentration gradient evaluated at the electrode surface (mol cm⁻⁴)

Reading: current is proportional to how steep the concentration profile is right at the surface — a sharper gradient (more depletion, faster resupply needed) means more current. The gradient itself must be found by solving Fick's second law (the diffusion equation) with the correct boundary conditions for the experiment; this is done below for the potential-step (chronoamperometric) case.

The Cottrell equation is the solution of Fick's second law for a potential step large enough to reduce the surface concentration of analyte to essentially zero (a diffusion-limited step), in unstirred solution:

i_f(t) = \frac{nFAC^*\sqrt{D}}{\sqrt{\pi t}}

where:

$C^*$ = bulk concentration of the electroactive species (mol cm⁻³)
$t$ = time elapsed since the potential step (s)
all other symbols as above

Reading: the faradaic current decays as $t^{-1/2}$ because the diffusion layer thickens with time (roughly as $\sqrt{Dt}$ ), so the gradient — and hence the flux — weakens. The current is directly proportional to bulk concentration, which is what makes chronoamperometry quantitative.

The double-layer charging current, for a potential step of size $\Delta E$ applied to a cell with uncompensated resistance $R_u$ and double-layer capacitance $C_d$ , decays exponentially:

i_c(t) = \frac{\Delta E}{R_u}\, e^{-t/R_uC_d}

where:

$\Delta E$ = magnitude of the potential step (V)
$R_u$ = uncompensated solution resistance (Ω)
$C_d$ = double-layer capacitance (F), often expressed per unit area (F cm⁻²) times $A$

Reading: immediately after the step the charging current is large ( $\Delta E/R_u$ , an ohmic instantaneous response) but collapses on the $R_uC_d$ timescale — typically microseconds for a well-designed cell — while the faradaic Cottrell current is just getting started and persists far longer.

The Levich equation gives the steady-state, mass-transport-limited current at a rotating disk electrode (RDE), where forced convection continuously replenishes the diffusion layer, producing a constant (not decaying) current:

i_{\ell} = 0.620\, nFAD^{2/3}\omega^{1/2}\nu^{-1/6}C^*

where:

$i_\ell$ = limiting current (A), constant in time
$\omega$ = angular rotation rate of the disk (rad s⁻¹), $\omega = 2\pi \times (\text{rpm}/60)$
$\nu$ = kinematic viscosity of the solution (cm² s⁻¹), typically $\approx 0.01\ \text{cm}^2\text{s}^{-1}$ for dilute aqueous electrolyte
$0.620$ = numerical constant from the exact hydrodynamic solution of convective-diffusion at a rotating disk

Reading: spinning the electrode thins the diffusion layer to a fixed steady-state thickness set by $\omega$ , so instead of a current that decays forever, one gets a stable plateau current proportional to $\sqrt{\omega}$ and to $C^*$ — ideal for calibration curves and for kinetic studies (Levich and Koutecký–Levich analysis).

Derivation AThe Cottrell equation from Fick's second law

Goal: find $i_f(t)$ for a potential step at a planar electrode in an unstirred, semi-infinite solution, when the step drives the surface concentration of O (the oxidized species being reduced) to zero.

Step 1 — set up the diffusion equation and boundary conditions. Fick's second law for one-dimensional diffusion along $x$ (distance from the electrode) is

\frac{\partial C(x,t)}{\partial t} = D\frac{\partial^2 C(x,t)}{\partial x^2}

The physical situation imposes three conditions: initially the solution is uniform, $C(x,0) = C^*$ for all $x > 0$ ; far from the electrode the bulk is undisturbed at all times, $C(x\to\infty, t) = C^*$ ; and at the electrode surface, once the potential step is applied and diffusion limitation sets in, $C(0,t) = 0$ for $t>0$ .

Step 2 — nondimensionalize with a similarity variable. Because the problem has no intrinsic length scale, the concentration must depend on $x$ and $t$ only through the combination $\eta = x/(2\sqrt{Dt})$ . Substituting $C(x,t) = C^*f(\eta)$ into Fick's second law converts the partial differential equation into an ordinary differential equation in $\eta$ :

f''(\eta) + 2\eta f'(\eta) = 0

Step 3 — solve the ordinary differential equation. Writing $g = f'$ , this is first order: $g'/g = -2\eta$ , so $\ln g = -\eta^2 + \text{const}$ , giving $g(\eta) = f'(\eta) = A_1 e^{-\eta^2}$ . Integrating,

f(\eta) = A_1\int_0^\eta e^{-u^2}\,du + A_2

The boundary conditions $f(0) = 0$ (surface depletion) and $f(\infty) = 1$ (bulk value) fix the constants using the known integral $\int_0^\infty e^{-u^2}du = \sqrt{\pi}/2$ , yielding

f(\eta) = \operatorname{erf}(\eta) = \frac{2}{\sqrt{\pi}}\int_0^\eta e^{-u^2}\,du

so that $C(x,t) = C^*\operatorname{erf}\!\left(\dfrac{x}{2\sqrt{Dt}}\right)$ .

Step 4 — differentiate to get the surface gradient. Using $\dfrac{d}{d\eta}\operatorname{erf}(\eta) = \dfrac{2}{\sqrt{\pi}}e^{-\eta^2}$ and the chain rule,

\left(\frac{\partial C}{\partial x}\right)_{x=0} = C^*\cdot\frac{2}{\sqrt{\pi}}\cdot\frac{1}{2\sqrt{Dt}} = \frac{C^*}{\sqrt{\pi D t}}

Step 5 — insert into Fick's first law / Faraday's law. Substituting into $i_f = nFAD(\partial C/\partial x)_{x=0}$ ,

i_f(t) = nFAD\cdot\frac{C^*}{\sqrt{\pi D t}} = \frac{nFAC^*\sqrt{D}}{\sqrt{\pi t}}

which is the Cottrell equation. Limiting-case check: as $t\to\infty$ , $i_f \to 0$ — the diffusion layer grows without bound in unstirred solution, the gradient flattens, and the current must vanish, consistent with there being no steady state for pure diffusion to a planar electrode with no stirring.

Go deeper · Derivation B · 2 minThe Levich equation from convective-diffusion at a rotating disk

Derivation BThe Levich equation from convective-diffusion at a rotating disk

Goal: find the steady-state limiting current at a rotating disk electrode, where forced convection prevents the diffusion layer from growing indefinitely, sketching the key physical steps that lead to the $\omega^{1/2}$ dependence (full derivation requires the von Kármán similarity solution for rotational flow; the physically load-bearing steps are shown).

Step 1 — steady-state convective-diffusion equation. At steady state, ${\partial C}/{\partial t}=0$ , and the one-dimensional convective-diffusion equation normal to the disk is

D\frac{d^2C}{dx^2} - v_x(x)\frac{dC}{dx} = 0

where $v_x(x)$ is the fluid velocity component perpendicular to (toward) the disk surface, generated by the disk's rotation.

Step 2 — von Kármán's result for the flow field. Solving the Navier–Stokes equations for a rotating disk (von Kármán, 1921) shows that near the disk surface the normal velocity grows as

v_x(x) \approx -0.51\,\omega^{3/2}\nu^{-1/2}x^2

for small $x$ , where $\omega$ is the angular rotation rate and $\nu$ the kinematic viscosity. This quadratic-in- $x$ dependence is the hydrodynamic input that makes the diffusion layer thickness finite and calculable rather than growing with $\sqrt{t}$ as in the unstirred case.

Step 3 — define an effective diffusion layer thickness. Balancing diffusive and convective transport across a layer of thickness $\delta$ shows $\delta$ scales as

\delta \approx 1.61\, D^{1/3}\nu^{1/6}\omega^{-1/2}

Physically: faster rotation ( $\omega\uparrow$ ) sweeps depleted solution away more effectively, thinning the diffusion layer; a larger diffusion coefficient partly compensates by letting analyte cross that layer faster on its own.

Step 4 — write the flux across this fixed-thickness layer as if it were simple linear diffusion. Approximating the concentration profile in the thin layer as linear from $C=0$ at the surface to $C=C^*$ at $x=\delta$ gives a steady gradient $C^*/\delta$ , so

i_\ell = nFAD\frac{C^*}{\delta} = nFAD\,C^*\cdot\frac{\omega^{1/2}}{1.61\,D^{1/3}\nu^{1/6}}

Step 5 — collect terms and match the exact numerical prefactor. Simplifying the power of $D$ : $D/D^{1/3} = D^{2/3}$ , giving

i_\ell = \frac{1}{1.61}\,nFAD^{2/3}\nu^{-1/6}\omega^{1/2}C^*

The exact solution of the full convective-diffusion boundary-value problem (rather than the linear-approximation shortcut used here) replaces $1/1.61 = 0.621$ with the precisely-derived constant $0.620$ , giving the standard Levich equation

i_\ell = 0.620\,nFAD^{2/3}\omega^{1/2}\nu^{-1/6}C^*

Limiting-case check: as $\omega \to 0$ (no rotation), $\delta \to \infty$ and $i_\ell \to 0$ — with no convection at all, there is no steady-state current, consistent with the Cottrell result that unstirred diffusion current decays to zero; the Levich equation is the "rescue" that forced convection provides.

Worked examples

Example 1 (foundational) — Cottrell current from a potential step. A quiescent solution contains $C^* = 2.50\ \text{mM}$ Fe(CN)₆³⁻ in 0.10 M KNO₃. A 2.0 mm-diameter Pt disk electrode ( $n=1$ , $D = 7.20\times10^{-6}\ \text{cm}^2\text{s}^{-1}$ for ferricyanide in this medium) is stepped to a potential on the diffusion-limited plateau. Find the faradaic current 0.50 s and 5.0 s after the step. Setup. Electrode area: radius $r = 0.10\ \text{cm}$ , so

A = \pi r^2 = \pi (0.10\ \text{cm})^2 = 3.1416\times10^{-2}\ \text{cm}^2

Convert concentration to mol/cm³:

C^* = 2.50\times10^{-3}\ \text{mol L}^{-1}\times\frac{1\ \text{L}}{1000\ \text{cm}^3} = 2.50\times10^{-6}\ \text{mol cm}^{-3}

Apply the Cottrell equation at $t=0.50$ s.

i_f = \frac{nFAC^*\sqrt{D}}{\sqrt{\pi t}} = \frac{(1)(96485)(3.1416\times10^{-2})(2.50\times10^{-6})\sqrt{7.20\times10^{-6}}}{\sqrt{\pi(0.50)}}

Compute the numerator piece by piece: $nFA = (96485)(3.1416\times10^{-2}) = 3031.4\ \text{C cm}^2\text{mol}^{-1}$ (units track through with the rest). Then $nFAC^* = 3031.4\times2.50\times10^{-6} = 7.579\times10^{-3}$ . And $\sqrt{D} = \sqrt{7.20\times10^{-6}} = 2.683\times10^{-3}\ \text{cm s}^{-1/2}$ . Numerator $= 7.579\times10^{-3}\times2.683\times10^{-3} = 2.034\times10^{-5}$ . Denominator: $\sqrt{\pi(0.50)} = \sqrt{1.5708} = 1.2533$ .

i_f(0.50\ \text{s}) = \frac{2.034\times10^{-5}}{1.2533} = 1.62\times10^{-5}\ \text{A} = 16.2\ \mu\text{A}

At $t = 5.0$ s: the current scales as $t^{-1/2}$ , so

i_f(5.0\ \text{s}) = i_f(0.50\ \text{s})\times\sqrt{\frac{0.50}{5.0}} = 16.2\ \mu\text{A}\times\sqrt{0.10} = 16.2\times0.3162 = 5.12\ \mu\text{A}

Sanity check via limiting behavior: a 10-fold increase in time gives a $\sqrt{10} \approx 3.16$ -fold decrease in current — exactly what is observed (16.2/5.12 = 3.16). The current never truly reaches a steady nonzero plateau in unstirred solution, matching the $t\to\infty \Rightarrow i_f\to0$ limiting case checked in the derivation.

Example 2 (applied) — separating charging current from faradaic current, and back-calculating concentration. A cell has uncompensated resistance $R_u = 150\ \Omega$ and double-layer capacitance $C_d = 20\ \mu\text{F}$ (a typical small Pt disk). A potential step of $\Delta E = 0.30$ V is applied. (a) Find the $RC$ time constant and the charging current 5 time constants after the step. (b) A chronoamperometric measurement taken at $t = 20$ ms after the step gives a total measured current of $8.90\ \mu\text{A}$ for a solution of unknown Cd²⁺ concentration ( $n=2$ , $D_{\text{Cd}^{2+}} = 7.2\times10^{-6}\ \text{cm}^2\text{s}^{-1}$ , electrode area $0.020\ \text{cm}^2$ ). Estimate the true Cd²⁺ concentration, assuming charging current is negligible by 20 ms. (a) Time constant.

\tau = R_uC_d = (150\ \Omega)(20\times10^{-6}\ \text{F}) = 3.0\times10^{-3}\ \text{s} = 3.0\ \text{ms}

Charging current at $t=0$ : $i_c(0) = \Delta E/R_u = 0.30/150 = 2.0\times10^{-3}\ \text{A} = 2.0\ \text{mA}$ — large initially. After 5 time constants ( $t=15$ ms):

i_c(15\ \text{ms}) = (2.0\ \text{mA})\,e^{-5} = (2.0\times10^{-3})(6.74\times10^{-3}) = 1.35\times10^{-5}\ \text{A} = 13.5\ \mu\text{A}

Wait — this is not negligible yet compared to microamp-scale faradaic currents; that is precisely why the problem statement specifies waiting until $t=20$ ms, nearly 7 time constants:

i_c(20\ \text{ms}) = (2.0\ \text{mA})\,e^{-20/3} = (2.0\times10^{-3})(1.23\times10^{-3}) = 2.46\times10^{-6}\ \text{A} = 2.46\ \mu\text{A}

This is still a meaningful fraction of an 8.90 μA total signal — a reminder that "negligible" charging current requires either longer waiting times, smaller $R_uC_d$ , or explicit background subtraction; here we proceed with the problem's stated assumption for illustration but flag the residual 2.46 μA as a real source of systematic error (roughly 28% of the total signal), which a careful analyst would subtract rather than ignore. (b) Faradaic current and concentration, subtracting the estimated charging contribution.

i_f = i_{\text{meas}} - i_c = 8.90\ \mu\text{A} - 2.46\ \mu\text{A} = 6.44\ \mu\text{A}

Rearrange the Cottrell equation for $C^*$ :

C^* = \frac{i_f\sqrt{\pi t}}{nFA\sqrt{D}}

Compute each piece: $\sqrt{\pi t} = \sqrt{\pi(0.020)} = \sqrt{0.06283} = 0.2507$ . $nFA = (2)(96485)(0.020) = 3859.4$ . $\sqrt{D} = \sqrt{7.2\times10^{-6}} = 2.683\times10^{-3}$ .

C^* = \frac{(6.44\times10^{-6})(0.2507)}{(3859.4)(2.683\times10^{-3})} = \frac{1.615\times10^{-6}}{10.354} = 1.560\times10^{-7}\ \text{mol cm}^{-3}

Converting to molarity: $C^* = 1.560\times10^{-7}\ \text{mol cm}^{-3}\times1000\ \text{cm}^3/\text{L} = 1.56\times10^{-4}\ \text{M} = 0.156\ \text{mM}$ . This example demonstrates the practical necessity of the charging-current correction — ignoring it here would have inflated the apparent concentration by roughly 28%, from 0.156 mM to a spurious 0.216 mM.

Example 3 (exam / competition level) — combining Levich analysis with a Nernstian assumption to find a diffusion coefficient and check electrode kinetics. A rotating disk electrode ( $A = 0.196\ \text{cm}^2$ , corresponding to a 5.0 mm diameter disk) is used to study the one-electron reduction of a nitroaromatic compound, $C^* = 1.00\ \text{mM}$ , in a solution with kinematic viscosity $\nu = 0.0100\ \text{cm}^2\text{s}^{-1}$ . Limiting currents are measured at four rotation rates:

rotation rate (rpm)	$\omega^{1/2}$ (rad $^{1/2}$ s $^{-1/2}$ )	$i_\ell$ (μA)
400	6.481	42.3
900	9.721	63.4
1600	12.96	84.6
2500	16.20	105.7

(a) Confirm the Levich prediction that $i_\ell \propto \omega^{1/2}$ and extract $D$ . (b) Predict $i_\ell$ at 1200 rpm as an internal consistency check. (a) Convert rpm to $\omega$ (rad/s) and confirm linearity. For 400 rpm: $\omega = 2\pi(400/60) = 41.89\ \text{rad s}^{-1}$ , so $\omega^{1/2} = 6.472$ (matches the tabulated 6.481 to rounding). The ratio $i_\ell/\omega^{1/2}$ should be constant if the Levich equation holds:

\frac{42.3}{6.481} = 6.527,\quad \frac{63.4}{9.721}=6.523,\quad\frac{84.6}{12.96}=6.528,\quad\frac{105.7}{16.20}=6.525\ (\mu\text{A}\,\text{s}^{1/2}\text{rad}^{-1/2})

These four values agree to within 0.1%, confirming the predicted $\omega^{1/2}$ dependence — strong evidence that the reduction is mass-transport-limited (not kinetically limited) across this entire rotation range, since a kinetically-limited or mixed-control current would show a non-linear or intercept-shifted dependence on $\omega^{1/2}$ (the Koutecký–Levich test). Extract $D$ from the slope. The slope of $i_\ell$ vs. $\omega^{1/2}$ is $\text{slope} = 0.620\,nFAD^{2/3}\nu^{-1/6}C^*$ . Taking the average ratio, $\text{slope} = 6.526\ \mu\text{A s}^{1/2}\text{rad}^{-1/2} = 6.526\times10^{-6}\ \text{A s}^{1/2}\text{rad}^{-1/2}$ . Solve for $D^{2/3}$ :

D^{2/3} = \frac{\text{slope}}{0.620\,nFAC^*\nu^{-1/6}}

Compute the denominator pieces: $nFA = (1)(96485)(0.196) = 18911.1$ . $C^* = 1.00\times10^{-6}\ \text{mol cm}^{-3}$ . $\nu^{-1/6} = (0.0100)^{-1/6}$ . Since $0.0100 = 10^{-2}$ , $\nu^{-1/6} = 10^{2/6}=10^{0.3333}=2.1544$ .

0.620\times18911.1\times1.00\times10^{-6}\times2.1544 = 0.620\times18911.1\times2.1544\times10^{-6}

= 0.620\times40{,}751\times10^{-6} = 25{,}266\times10^{-6} = 2.527\times10^{-2}

D^{2/3} = \frac{6.526\times10^{-6}}{2.527\times10^{-2}} = 2.583\times10^{-4}

D = (2.583\times10^{-4})^{3/2}

Compute: $(2.583\times10^{-4})^{1/2} = 1.607\times10^{-2}$ , then $D = (2.583\times10^{-4})\times(1.607\times10^{-2}) = 4.152\times10^{-6}\ \text{cm}^2\text{s}^{-1}$ . This value, $D \approx 4.15\times10^{-6}\ \text{cm}^2\text{s}^{-1}$ , is physically reasonable for a modest-sized organic anion in aqueous solution (typical range $10^{-6}$ – $10^{-5}\ \text{cm}^2\text{s}^{-1}$ ), which is an important internal-consistency check on the whole analysis. (b) Predict $i_\ell$ at 1200 rpm. $\omega = 2\pi(1200/60) = 125.66\ \text{rad s}^{-1}$ , so $\omega^{1/2} = 11.21$ .

i_\ell(1200\ \text{rpm}) = 6.526\ \mu\text{A s}^{1/2}\text{rad}^{-1/2}\times11.21\ \text{rad}^{1/2}\text{s}^{-1/2} = 73.2\ \mu\text{A}

This falls neatly between the tabulated values at 900 rpm (63.4 μA) and 1600 rpm (84.6 μA), as expected from the monotonic $\omega^{1/2}$ trend — a useful internal check that no arithmetic slip occurred in extracting $D$ .

Limiting cases, scope & common traps

Situation	What happens	Governing relation
$t\to 0^+$ immediately after a potential step	Charging current dominates ( $i_c \gg i_f$ ); faradaic signal is swamped	$i_c(0)=\Delta E/R_u$
$t \gg R_uC_d$ , unstirred solution	Charging current has decayed to ~0; faradaic current persists but keeps falling as $t^{-1/2}$	Cottrell equation
$t\to\infty$ , unstirred	Diffusion layer infinitely thick; $i_f\to 0$ — no true steady state without convection	Cottrell equation limiting case
Forced convection (RDE), steady state	Diffusion layer thickness fixed at $\delta\propto\omega^{-1/2}$ ; current constant in time, scales as $\sqrt{\omega}$	Levich equation
$\omega\to 0$ (RDE stopped)	Recovers the unstirred, decaying-current regime; $i_\ell\to0$	Levich equation limiting case
No supporting electrolyte	Migration contributes to mass transport; current no longer cleanly proportional to $C^*$ via diffusion alone	Nernst–Planck equation (migration term retained)
Excess (≥ 50×) supporting electrolyte	Migration of analyte suppressed; diffusion-only transport, Cottrell/Levich apply directly	—
Kinetically sluggish (irreversible) electron transfer	Current is not simply mass-transport-limited even at large overpotential on typical timescales; Koutecký–Levich plot shows curvature/intercept	Butler–Volmer kinetics superimposed on transport

The chemist behind it

Jaroslav Heyrovský developed Polarography at Charles University in Prague in 1922, using a dropping mercury electrode whose constantly renewing, reproducible surface made quantitative current-potential curves practical for the first time — for which he received the 1959 Nobel Prize in Chemistry, the technique that historically launched all of modern voltammetry. The three-electrode potentiostatic cell used throughout modern voltammetry, and the systematic separation of faradaic from charging current exploited in pulse techniques, were built out gradually afterward by electrochemists and instrument designers (notably including work associated with Archie Hickling, who introduced the term "potentiostat," and later by Allen Bard and Larry Faulkner, whose synthesis of the mass-transport and kinetic theory in the standard modern textbook made the three-current-component picture — migration, diffusion, convection, faradaic vs. charging — the common language of the field.

Origins · 1 minJaroslav Heyrovský and how this was discovered

Jaroslav Heyrovský

1890-1967 · Charles University, Prague; developer of polarography and the dropping mercury electrode

Heyrovský trained in chemistry and physics in Prague and London, and in 1922 combined a controllable, continuously renewed mercury drop electrode with a simple applied-voltage sweep to record current-voltage curves ("polarograms") that revealed both the identity and concentration of dissolved ions and organic species. He later collaborated with instrument maker Masaryk-era engineers to build the automatic polarograph, turning a laboratory curiosity into a routine analytical tool used across chemistry, biochemistry, and industry for decades.

History

- 1922 — Heyrovský introduces polarography with the dropping mercury electrode.

- 1925 — Heyrovský and Shikata build the first automatic polarograph recording instrument.

- 1942 — Kolthoff and Lingane's monograph systematizes polarographic theory and practice.

- 1959 — Heyrovský awarded the Nobel Prize in Chemistry for the discovery and development of polarography.

- 1980 — Bard and Faulkner's Electrochemical Methods consolidates modern three-electrode, mass-transport-based theory into the standard reference framework still used today.

Why is a three-electrode cell needed instead of just two electrodes?

With only two electrodes, all the cell current must pass through the same electrode used to sense potential. Passing current through a reference electrode polarizes it away from its equilibrium potential (via its own finite exchange current and any ohmic drop across it), so the "reference" potential drifts unpredictably as current flows, corrupting the very quantity being controlled. Splitting the roles — a reference that senses potential at essentially zero current, and a separate auxiliary electrode that carries the reaction current — lets the potentiostat's feedback loop hold the working electrode's potential accurately and reproducibly even while substantial faradaic current flows. This is the electroanalytical analog of using a high-impedance voltmeter in potentiometry so the measurement itself does not disturb the system.

What physically is the "double-layer capacitance," and why does it produce a current with no chemical reaction at all?

When an electrode's potential is held away from the solution's equilibrium potential, charge redistributes at the interface: excess or deficient electrons accumulate on the metal, and a compensating layer of solvent dipoles and ions arranges itself in solution within roughly a nanometer of the surface — the electrical double layer. This structure stores charge just as a parallel-plate capacitor does, with capacitance $C_d$ typically 10–40 μF per cm² of electrode area. Any change in potential requires charging or discharging this capacitor, and moving charge means current flows — but no chemical species is oxidized or reduced. This is why charging current appears even in pure supporting electrolyte with no electroactive species present, and why it must be subtracted or allowed to decay before interpreting current as faradaic.

Why does the faradaic current from an unstirred solution never reach a true steady state, while the RDE current does?

In unstirred solution the only way for analyte to reach a depleted region near the electrode is diffusion, and the diffusion layer simply keeps growing thicker over time (as $\sqrt{Dt}$ ) with nothing to stop it; a thicker layer means a shallower gradient and hence smaller current, so $i_f$ falls forever as $t^{-1/2}$ and never levels off. At a rotating disk electrode, forced convection continuously sweeps solution past the disk and replenishes the outer edge of the diffusion layer, capping its thickness at a fixed value set by the balance of convection and diffusion (roughly $\delta \propto D^{1/3}\nu^{1/6}\omega^{-1/2}$ ). Because the diffusion layer thickness stops changing, the concentration gradient — and hence the current — becomes constant in time, giving the steady plateau the Levich equation predicts.

How much supporting electrolyte is "enough" to suppress migration, and why does that number matter for quantitative work?

A common rule of thumb is at least a 50- to 100-fold excess of inert electrolyte over the analyte concentration, though the precise ratio needed depends on the analyte's charge and the desired accuracy. The physical reasoning comes from the Nernst–Planck equation: the migrational contribution to an ion's flux is proportional to its own concentration and charge relative to the total ionic strength carrying the applied field. With enough inert ions present, they carry essentially all of the migrational current, and the electric field in the bulk solution — outside a very thin double layer — collapses to nearly zero, leaving diffusion as the only transport mechanism for the trace analyte. Skipping this precaution means the Cottrell and Levich equations, both derived assuming pure diffusion, will systematically misestimate concentration because part of the measured current is migrational rather than diffusional.

Why does Example 2's charging current correction matter so much even at 20 milliseconds — isn't that "long" compared to microsecond RC times?

The specific numbers matter: with $R_u = 150\ \Omega$ and $C_d = 20\ \mu\text{F}$ , the time constant is $\tau = 3.0$ ms, not microseconds — a realistic value for a modestly-sized electrode in moderately resistive solution, not an idealized ultramicroelectrode. At $t=20$ ms, only about 6.7 time constants have elapsed, and while $e^{-6.7}\approx 1.2\times10^{-3}$ makes the fraction remaining of the initial charging current small, the initial charging current ( $\Delta E/R_u = 2.0$ mA) was so large that even this small remaining fraction (2.46 μA) is comparable to the microamp-scale faradaic signal. The lesson generalizes: whether charging current is "negligible" depends on the ratio of $i_c(t)$ to $i_f(t)$ at the sampling time, not merely on how many time constants have passed — both need to be computed explicitly rather than assumed away.

Does the Levich equation apply to any electrode reaction, or only to fast, reversible ones?

The Levich equation itself is purely a statement about mass transport — it describes the current that would flow if the electrode reaction were fast enough that surface concentration of the electroactive species dropped essentially to zero at the applied potential (mass-transport control). If electron-transfer kinetics are sluggish, the actual measured current will be smaller than the Levich prediction and will not extrapolate to zero as $1/\omega^{1/2}\to 0$ in a Koutecký–Levich plot ( $1/i$ vs. $1/\sqrt{\omega}$ ) — instead showing a nonzero intercept related to the kinetic (charge-transfer) resistance. Example 3's clean $\omega^{1/2}$ -linearity through the origin (constant slope, effectively no intercept) is itself evidence that the nitroaromatic reduction studied there is under mass-transport control across the whole rotation-rate range tested.

How does this connect to cyclic voltammetry and polarography, and why isn't it covered here?

Cyclic voltammetry sweeps the potential linearly back and forth through the redox couple's potential and records the resulting current-potential curve, producing characteristic peak-shaped responses whose height and separation encode both thermodynamic (formal potential) and kinetic (electron-transfer rate) information — it is built directly on the faradaic-versus-charging-current and diffusion concepts developed here, but its peak-current theory (the Randles–Ševčík equation) requires solving a more involved, time-dependent boundary-value problem specific to a linear potential sweep. Polarography, historically the first voltammetric technique, uses a continuously renewing dropping mercury electrode and exploits the same Cottrell-type diffusion physics, with the added twist of an expanding spherical (drop) electrode geometry. Both techniques are natural next steps built entirely on the three-electrode cell, the faradaic/charging current distinction, and the mass-transport equations developed here.

Voltammetry Fundamentals

Where this lives today

The physical picture — thorough prose

The core ideas — governing equations

Worked examples

Limiting cases, scope & common traps

The chemist behind it

Before this

Where this leads