Butler-Volmer & Tafel

A tug-of-war between two reactions

Dip a metal electrode into a solution containing both an oxidized species O and its reduced partner R. Even when you do nothing, the surface is busy: electrons are constantly hopping off the metal to reduce O (a cathodic current) and other electrons are hopping back on as R gives them up (an anodic current). At equilibrium these two flows are exactly equal and opposite, so your ammeter reads zero — but the reactions are anything but stopped. They are running at a balanced speed called the exchange current density, i₀.

Now move the electrode's potential away from its equilibrium value by an amount η, the overpotential. This is like tilting a see-saw. Push the potential more negative and you tilt the energy landscape to favor reduction — the forward (cathodic) flow surges while the reverse (anodic) flow withers. The net current you measure is just the difference between the two. The Butler-Volmer equation is the precise statement of that difference, and its punchline is that both flows respond exponentially to η. A swing of a tenth of a volt can change the rate by orders of magnitude.

This is the single most important equation in electrode kinetics. It is to electrochemistry what the Arrhenius equation is to thermal chemistry — except here the knob you turn is voltage, not temperature, and voltage is far easier to control than temperature.

The two exponentials and where they come from

Consider the elementary one-electron step O + e⁻ ⇌ R. Each direction has its own activation barrier and therefore its own Arrhenius-like rate. The trick of electrode kinetics is that applying an overpotential η shifts those barriers electrostatically: a fraction α_c of the energy change lowers the barrier for reduction, while the complementary fraction α_a = 1 − α_c raises the barrier for oxidation (for a simple single step). Writing each partial current as a rate × concentration × charge gives:

i = i₀ · [ exp( α_a·F·η / RT )  −  exp( −α_c·F·η / RT ) ]
         └─── anodic (oxidation) ───┘   └── cathodic (reduction) ──┘

where F = 96,485 C/mol (Faraday's constant), R = 8.314 J/(mol·K), T is absolute temperature, α_a and α_c are the anodic and cathodic charge transfer coefficients, and i₀ is the exchange current density. At η = 0 both exponentials equal 1, so i = i₀(1 − 1) = 0. The ammeter reads zero, exactly as it must at equilibrium, while both half-reactions churn away at i₀ underneath.

The quantity F·η/RT is the natural scale of the problem. At 25 °C, RT/F = 25.7 mV, so an overpotential of one "thermal voltage unit" (25.7 mV) changes each exponent by α. Because the dependence is exponential, an overpotential of about 60/α mV multiplies a partial current by ten.

Exchange current density: the electrode's intrinsic speed

i₀ packs the chemistry of the interface — the catalyst, the reactant concentrations, the transition-state energy — into one number. It is the rate of the forward and reverse reactions when they are balanced. A large i₀ means the electrode barely needs any overpotential to deliver useful current; a tiny i₀ means you must push hard before anything happens. The spread across real systems is staggering:

Reaction / electrode	i₀ (A/cm²)	Practical consequence
H⁺/H₂ on platinized Pt	≈ 10⁻³	Fast — basis of the standard hydrogen electrode
H⁺/H₂ on smooth Pt	≈ 10⁻⁴	Still excellent; benchmark HER catalyst
H⁺/H₂ on iron	≈ 10⁻⁶	Sluggish — large overpotential needed
H⁺/H₂ on mercury	≈ 10⁻¹³	Catastrophically slow; used to suppress HER
O₂ evolution on Pt	≈ 10⁻⁹	The OER is the efficiency killer in water splitting
Fe³⁺/Fe²⁺ on Pt	≈ 10⁻²	Nearly reversible outer-sphere redox couple

That ten-order-of-magnitude range between Pt and Hg for the same hydrogen reaction is why catalyst choice dominates electrochemical engineering. The thermodynamics (the equilibrium potential) is identical on both metals; only the kinetics — i₀ — differs, and it differs by a factor of ten billion.

The Tafel limit: when one branch wins

Tafel's law actually predates the full Butler-Volmer equation by two decades — Julius Tafel published his empirical η-vs-log(i) straight line for hydrogen evolution in 1905. Butler-Volmer explains why it is a straight line. Push the overpotential far enough negative and the anodic exponential becomes negligible; the cathodic term alone survives:

large cathodic η:   |i| ≈ i₀ · exp( −α_c·F·η / RT )

take logs:          ln|i| = ln(i₀) − α_c·F·η / RT
rearrange:          η = (RT / α_c·F)·ln(i₀)  −  (RT / α_c·F)·ln|i|

in base-10 form:    η = a + b·log|i|
   with   b = −2.303·RT / (α_c·F)   (the Tafel slope)
          a = −b·log(i₀)            (the Tafel intercept → gives i₀)

So a plot of overpotential against log|current| is a straight line. Its slope b tells you α (and hence the mechanism), and its intercept extrapolated back to η = 0 hands you i₀. At 25 °C with α = 0.5 the slope is 2.303 × 25.7 mV / 0.5 ≈ 118 mV per decade — the textbook "120 mV/decade." A one-decade jump in current (ten times more current) costs only about 120 mV of extra overpotential. That is the exponential made visible.

Worked example: reading a Tafel plot

Suppose you measure the hydrogen evolution reaction on a new catalyst at 25 °C and find that drawing 1 mA/cm² requires η = −150 mV, while drawing 10 mA/cm² requires η = −270 mV.

Tafel slope  b = Δη / Δ(log|i|)
              = (−270 − (−150)) mV / (log 10 − log 1)
              = −120 mV / 1 decade
              = −120 mV/decade

Charge transfer coefficient:
  |b| = 2.303·RT/(α·F)  ⇒  α = 2.303·RT/(|b|·F)
  α = 2.303 · 8.314 · 298 / (0.120 · 96,485)
  α = 5,706 / 11,578
  α ≈ 0.49        (a near-symmetric barrier)

Exchange current density (extrapolate η → 0, where |i| = i₀):
  η = a + b·log|i|   ⇒   η = b·(log|i| − log|i₀|)
  rearrange:  log|i₀| = log|i| − η/b
  use the η = −150 mV point (|i| = 1 mA, b = −120 mV/decade):
  log|i₀| = log(1) − (−150)/(−120) = 0 − 1.25 = −1.25
  i₀ = 10^(−1.25) mA/cm² ≈ 0.056 mA/cm² ≈ 5.6×10⁻⁵ A/cm²

From two data points you have extracted both kinetic parameters: a symmetric transfer coefficient (α ≈ 0.5) and an exchange current density of ~5.6 × 10⁻⁵ A/cm² — respectable, in the same ballpark as smooth platinum. A steeper slope would have implied a lazier electrode and a different rate-determining step.

Where the exponential stops: mass-transport limits

Butler-Volmer is a charge-transfer law — it assumes the reacting species is always plentiful right at the electrode surface. That assumption fails at high current. Once electrons are consumed faster than fresh reactant can diffuse in, the surface concentration sags toward zero and the current can no longer rise. It saturates at the diffusion-limited plateau:

i_lim = n·F·D·C_bulk / δ

where D is the diffusion coefficient (≈ 10⁻⁵ cm²/s for small ions), C_bulk the bulk concentration, and δ the diffusion-layer thickness (≈ 10–500 µm depending on stirring). A fuller "mixed-control" form multiplies the Butler-Volmer kinetics by a concentration factor (1 − i/i_lim). On a real Tafel plot this shows up as the line bending over at high overpotential: the straight Tafel region is the charge-transfer-controlled middle, and the curvature at the top is diffusion taking the wheel. Rotating-disk electrodes are popular precisely because they fix δ and let you separate the two regimes cleanly.

The other limit: the linear, near-equilibrium region

At the opposite extreme — very small η, within a few millivolts of equilibrium — both exponentials are close to 1 and you can linearize them (e^x ≈ 1 + x). The two leading terms combine into a strikingly simple result:

small η:   i ≈ i₀ · (α_a + α_c)·F·η / RT  =  i₀·F·η / RT   (for α_a + α_c = 1)

So near equilibrium the electrode behaves like a resistor:
   R_ct = η / i = RT / (i₀·F)        the charge-transfer resistance

This is why electrochemists love impedance spectroscopy: at small AC perturbations the electrode is Ohmic, and the measured charge-transfer resistance R_ct is inversely proportional to i₀. A high-i₀ electrode has a tiny R_ct; a sluggish one has a huge R_ct. So the same Butler-Volmer equation gives you a resistor near equilibrium (linear region) and a logarithmic Tafel line far from it — two faces of one exponential.

Butler-Volmer vs Tafel vs the linear approximation

	Butler-Volmer (full)	Tafel approximation	Linear (low-field)
Form	i = i₀[e^(α_aFη/RT) − e^(−α_cFη/RT)]	η = a + b·log|i|	i ≈ i₀Fη/RT
Valid when	All overpotentials (kinetic control)	|η| ≳ 50–100 mV (one branch dominates)	|η| ≲ 10 mV (near equilibrium)
Reverse reaction	Fully included	Neglected	Fully included (linearized)
Behavior near η = 0	i → 0 smoothly	Diverges (log blows up) — invalid	i → 0, Ohmic
What you extract	i₀, α_a, α_c by full fit	i₀ (intercept), α (slope)	R_ct = RT/(i₀F)
Typical use	Simulation, full polarization curve	Catalyst benchmarking, mechanism ID	Impedance spectroscopy
Breaks down when	Mass transport limits current	Approaching equilibrium or i_lim	η leaves the linear window

Where Butler-Volmer runs the world

• Fuel cells and water electrolyzers. The oxygen reaction (ORR in fuel cells, OER in electrolyzers) has an i₀ around 10⁻⁹ A/cm² — a billion times slower than the hydrogen reaction. That sluggish kinetics forces ~300–400 mV of overpotential just for oxygen, and that overpotential is wasted as heat. It is the single largest efficiency loss in green-hydrogen production, which is why platinum-group and iridium catalysts command their prices.
• Battery charging and rate capability. The fast-charge limit of a lithium-ion cell is partly set by the Butler-Volmer kinetics of Li⁺ intercalation at the anode. Drive too much current and the required overpotential pushes the electrode potential below 0 V vs Li/Li⁺, where metallic lithium plates out as dangerous dendrites instead of intercalating.
• Corrosion. The mixed-potential theory of corrosion superimposes two Butler-Volmer curves — metal dissolution and the cathodic reaction (O₂ reduction or H₂ evolution). Their intersection sets the corrosion potential and the corrosion current. Tafel extrapolation of a measured polarization curve is the standard laboratory method for measuring corrosion rates.
• Electroplating and electrowinning. Plating quality depends on keeping the deposition overpotential in the right Butler-Volmer regime — too little and deposition is uneven, too much and you hit diffusion control and get rough, dendritic, or burnt deposits.
• Electrocatalysis research. Tafel-slope measurement is the everyday diagnostic. A reported 120 → 40 mV/decade transition as overpotential rises is read as a change in rate-determining step, guiding which elementary reaction a new catalyst must accelerate.

Common misconceptions and pitfalls

• "Zero net current means nothing is happening." At η = 0 the net current is zero but both half-reactions run at i₀. Equilibrium is dynamic, not dead.
• Confusing overpotential with cell voltage. η is the deviation of a single electrode from its own equilibrium potential, not the voltage across the whole cell. A full cell's required voltage = thermodynamic E° + anode η + |cathode η| + iR (ohmic) drop. Butler-Volmer governs only the η terms.
• Reading i₀ off the wrong axis. i₀ is the Tafel-line intercept extrapolated to η = 0, not the current you measure at η = 0 (which is zero). The extrapolation is the only way to recover it.
• Forcing the Tafel equation near equilibrium. The η = a + b·log|i| form diverges as i → 0 and is simply invalid within ~50 mV of equilibrium — use the full Butler-Volmer or the linear form there.
• Assuming α = 0.5 always. α near 0.5 is common for simple symmetric single-electron transfers, but multistep reactions show apparent α values from below 0.5 to above 1 depending on which step limits. Fit it; don't assume it.
• Ignoring the iR drop. Uncompensated solution resistance adds a linear iR term that masquerades as extra overpotential and artificially steepens the apparent Tafel slope. Always correct for it (e.g. via current-interrupt or impedance) before quoting a slope.
• Treating the Tafel slope as a single fixed number. The slope can change across the current range as the mechanism's rate-determining step changes; report the region you measured.