The Tafel Equation

What the Tafel equation tells you

Every electrode reaction — plating copper, splitting water, discharging a battery — needs a push beyond its thermodynamic equilibrium potential to actually run at a useful rate. That extra push is the overpotential, η. The Tafel equation answers the practical question: how much more overpotential do I have to spend to make the reaction go faster?

The answer is startlingly clean. Current density and overpotential are not linearly related. Instead, over a wide window the overpotential rises with the logarithm of the current:

    η = a + b · log₁₀|i|

Here η is the overpotential (V), i is the current density (A/cm², normalized to electrode area), b is the Tafel slope in volts per decade, and a is the intercept. The physical meaning of the slope is the punchline: every tenfold increase in current costs you a fixed chunk of voltage, b. Go from 1 to 10 mA/cm², pay b. Go from 10 to 100, pay b again. A good catalyst has a small b — it buys you a factor of ten in rate for the least voltage.

Plot η against log|i| and you get a straight line. That line — the Tafel plot — is the single most-reported measurement in electrocatalysis, because its slope and its intercept between them encode the mechanism and the intrinsic activity of the surface.

Where the logarithm comes from

The logarithm is not empirical folklore; it falls straight out of the kinetics of electron transfer. Start from the Butler-Volmer equation, the master law for a single electrode reaction:

    i = i₀ [ exp(α_a F η / RT)  −  exp(−α_c F η / RT) ]

    i₀    = exchange current density (both directions at equilibrium)
    α_a   = anodic transfer coefficient
    α_c   = cathodic transfer coefficient   (α_a + α_c = n for a simple step)
    F     = 96485 C/mol       R = 8.314 J/mol·K       T = 298.15 K

Read the two terms as a tug-of-war. The first exponential is the forward (oxidation) rate; the second is the reverse (reduction) rate. At equilibrium (η = 0) both equal i₀ exactly and cancel, so the net current is zero even though charge is sloshing back and forth. The mechanism of the Tafel regime is what happens when you pull hard on one side:

1. Apply a large anodic overpotential. Push η positive by more than about 100 mV. The activation barrier for oxidation drops; the barrier for the reverse reduction rises.
2. The reverse reaction switches off. With η large and positive, the second exponential, exp(−α_c F η/RT), collapses toward zero. At η = +118 mV and α_c = 0.5, it is already down to about 1% of the forward term. The back-reaction is effectively dead.
3. One exponential survives. The current is now governed by a single term: i ≈ i₀·exp(α_a F η / RT). The electrode is running one-way.
4. Take the logarithm. ln(i) = ln(i₀) + (α_a F / RT)·η. Solve for η, convert ln to log₁₀ (multiply by 2.303), and the straight line appears:

    η = −(2.303 RT / α_a F)·log(i₀)  +  (2.303 RT / α_a F)·log(i)
         └────────  a  ────────┘         └──────  b  ──────┘

    ⇒  b = 2.303 RT / (α_a F)          a = −b · log(i₀)

So the Tafel slope is the reciprocal of the transfer coefficient times a universal thermal factor, and the intercept is set entirely by the exchange current density. The whole equation is just Butler-Volmer with one arm amputated. The exact same algebra on the cathodic branch (large negative η) gives the mirror-image line with slope b_c = 2.303RT/(α_cF).

Reading real numbers off the slope

Plug the constants in at 25 °C. The thermal factor 2.303RT/F = 0.0592 V = 59.2 mV. So:

    b = 59.2 mV / α        (at 25 °C)

    α = 0.5  →  b ≈ 118 mV/decade   (one-electron transfer, symmetric barrier)
    α = 1.0  →  b ≈  59 mV/decade
    α = 1.5  →  b ≈  40 mV/decade
    α = 2.0  →  b ≈  30 mV/decade

These are not arbitrary. In the hydrogen evolution reaction (2 H⁺ + 2 e⁻ → H₂), each canonical slope maps to a different rate-determining step:

• ≈ 120 mV/decade → Volmer step limiting. The first electron transfer, H⁺ + e⁻ → H_ads, is slow (α = 0.5, one electron before the slow step).
• ≈ 40 mV/decade → Heyrovsky step limiting. H_ads + H⁺ + e⁻ → H₂ is slow, but the fast Volmer pre-equilibrium has already put an electron through.
• ≈ 30 mV/decade → Tafel-recombination step limiting. Two adsorbed H atoms combine, H_ads + H_ads → H₂ — no electron in the slow step, two in the fast pre-equilibria. (This chemical recombination step is what gave Julius Tafel's name to both the step and the plot.)

Measure the slope, and you have narrowed the mechanism to one candidate rate-determining step without ever isolating an intermediate. That diagnostic power is why the Tafel slope is quoted in essentially every electrocatalysis paper. Temperature matters too: because b scales with T, a slope measured at 60 °C is about 11% larger than the same mechanism at 25 °C, so slopes must always be reported with the temperature.

Worked example: reading a Tafel plot

You run a platinum electrode in acid and measure the cathodic current for the hydrogen evolution reaction:

    η (mV)     |i| (A/cm²)     log|i|
    ───────    ──────────      ──────
    −30        1.0 × 10⁻³      −3.00
    −60        1.0 × 10⁻²      −2.00
    −90        1.0 × 10⁻¹      −1.00

Step 1 — get the slope. Between −30 and −90 mV the current climbed by two full decades (10⁻³ → 10⁻¹). The overpotential changed by 60 mV. So:

    b = Δη / Δ(log|i|) = 60 mV / 2 decades = 30 mV/decade

A 30 mV/decade slope on platinum tells you the reaction is limited by the chemical recombination (Tafel) step, with a transfer coefficient α ≈ 2 — the classic signature of fast, plentiful adsorbed hydrogen on Pt.

Step 2 — get the exchange current density. Extrapolate the line back to η = 0. Using η = a + b·log|i| with b = −0.030 V (cathodic) and the point (−0.030 V, log|i| = −3):

    0 = a + b·log(i₀)                         (at η = 0, i = i₀)
    a = −0.030 − (−0.030)(−3) = −0.030 − 0.090 = ... solve:
    a = η − b·log|i| = −0.030 − (−0.030)(−3.00) = −0.120 V
    log(i₀) = −a / b = −(−0.120)/(−0.030) = −4.00
    i₀ = 10⁻⁴ A/cm²

An exchange current density of 10⁻⁴ A/cm² is high — hallmark of a superb catalyst. On mercury the same reaction has i₀ ≈ 10⁻¹² A/cm², eight orders of magnitude smaller, which is exactly why mercury is a terrible hydrogen catalyst and platinum is the benchmark. The overpotential to reach a given current is what you actually pay for: at 10 mA/cm² this Pt surface needs only ~60 mV, while a 120 mV/decade catalyst with a lower i₀ might need 400 mV or more — energy lost as heat in every electrolyzer cell.

Tafel vs the equations it lives among

Relation	What it connects	Regime	Form in η vs i
Nernst equation	Equilibrium potential vs concentrations	i = 0 (thermodynamics only)	Sets the zero point (η = E − Eeq)
Butler-Volmer	Net current vs overpotential (both directions)	All η — the full curve	Difference of two exponentials
Tafel equation	Overpotential vs log current	Large |η| (> ~50–100 mV)	Straight line, slope b
Linear (micro-polarization)	Current vs overpotential	Small |η| (< ~10 mV)	Ohmic-like, i ≈ (i₀F/RT)η
Diffusion-limited (Cottrell/Levich)	Current vs mass transport	Very large |η|	Plateau — plot bends over
Marcus theory	Rate constant vs reorganization energy	Molecular level	Predicts α, curved Tafel at extreme η

The mental model: Nernst tells you where η = 0. Near there, current is linear in η (the low-field limit of Butler-Volmer). Push harder and Butler-Volmer curves; push harder still and one exponential wins, giving the straight Tafel line. Push to the limit and mass transport caps the current, so the line bends flat. The Tafel equation owns the useful middle band.

Real-world use: benchmarking catalysts and sizing cells

• Ranking electrocatalysts. When a paper claims a new oxygen-evolution catalyst, the headline numbers are its Tafel slope (mV/decade) and its overpotential at 10 mA/cm² — a benchmark current density chosen to match a ~10%-efficient solar water-splitting device. IrO₂ and RuO₂ for oxygen evolution sit near 40–60 mV/decade; state-of-the-art NiFe layered hydroxides in base reach ~30–40 mV/decade at overpotentials around 250–300 mV.
• Water electrolysis energy budget. A commercial alkaline or PEM electrolyzer runs near 1 A/cm². Every extra 100 mV/decade of Tafel slope, or every decade of missing exchange current, adds hundreds of millivolts of overpotential — and at green-hydrogen scale, 100 mV of avoidable overpotential across a gigawatt stack is a multi-megawatt heat leak.
• Corrosion rate from polarization. Sweeping potential around a corroding metal and fitting the anodic and cathodic Tafel lines lets you extrapolate to the corrosion current i_corr at the open-circuit potential — the basis of the Stern-Geary equation and every "linear polarization resistance" corrosion monitor in a pipeline or refinery.
• Battery and fuel-cell modeling. The activation loss in a fuel-cell polarization curve is a Tafel term; the oxygen-reduction cathode's ~60–70 mV/decade slope is the dominant voltage loss at low current in a PEM fuel cell and the reason Pt loading matters.
• Chlor-alkali and electroplating. Choosing electrode coatings (e.g. RuO₂-based dimensionally stable anodes for chlorine) is largely a hunt for low Tafel slope and high i₀ to cut the industrial cell voltage — where a single company's chlorine plant may draw hundreds of megawatts.

Limitations and where the line lies

• Fails at low overpotential. Below ~50 mV both Butler-Volmer exponentials contribute; the current is linear in η, not logarithmic. Fitting a Tafel line through the near-equilibrium region gives a meaningless slope. Only fit the genuinely straight, high-η portion.
• Fails at high current — mass transport. Once electron transfer outruns the supply of reactant to the surface, the current plateaus at the diffusion-limited value i_L and the plot bends over. The apparent slope shoots up; the surface is starved, not sluggish. Stirring, rotating-disk electrodes, or gas-diffusion layers push this ceiling higher.
• Uncompensated resistance (iR drop). Solution and contact resistance R_u add an ohmic term i·R_u that is mistaken for overpotential, artificially steepening the measured slope. You must correct for R_u (measured by impedance and subtracted, or compensated in situ) before quoting a Tafel slope — an uncorrected slope is one of the most common errors in the literature.
• Changing mechanism. Real Tafel plots can show two linear regions with different slopes, signalling that the rate-determining step switches as overpotential rises (e.g. Volmer→Heyrovsky). One clean slope over a broad range is the exception, not the rule.
• Potential-dependent coverage. The derivation assumes constant coverage of adsorbed intermediates and a potential-independent α. When intermediate coverage changes with η (Frumkin isotherm effects) the plot curves, and a single b no longer describes it.

Who found it, and when

Julius Tafel, a Swiss-born organic chemist working at the University of Würzburg, published the empirical relationship in 1905 while studying the electrolytic reduction of organic compounds. He noticed that the overpotential for hydrogen evolution rose linearly with the logarithm of the current density across many metals, and he tabulated the slopes and intercepts — decades before anyone could explain why. Tafel was primarily a natural-products chemist (he worked with Emil Fischer on purines and uric-acid derivatives); the equation was almost a side-note in a career of synthetic organic chemistry.

The kinetic explanation came later. John Alfred Valentine Butler (1924) and Max Volmer with Tibor Erdey-Grúz (1930) built the electron-transfer theory — the Butler-Volmer equation — that showed Tafel's straight line was simply the high-overpotential limit of two competing exponential rates. So the timeline runs backward from the usual pattern: the empirical law (Tafel, 1905) preceded its theory (Butler-Volmer, 1924–1930) by a generation. Today the Tafel slope remains the workhorse diagnostic of electrode kinetics, and the constant 2.303RT/F ≈ 59 mV at 25 °C is one of the most-used numbers in the field.