Electrocatalysis and water splitting: the oxygen-evolution reaction, the hydrogen-evolution reaction, and the search for non-precious-metal catalysts

Intuition Beginner

Water is H2O: two hydrogen atoms bonded to one oxygen. Pass electricity through water and the bonds break, producing hydrogen gas (H2) at one electrode and oxygen gas (O2) at the other. This is water splitting, and the hydrogen it produces is a zero-carbon fuel whose only combustion product is water again. The bottleneck is the oxygen half. Pairing oxygen atoms from water into O2 requires pulling off four electrons per molecule, and the chemistry is sluggish. Without a catalyst the reaction refuses to run at useful rates.

The best catalyst for the oxygen half is iridium oxide, but iridium is one of the rarest elements in the crust and sells for thousands of dollars per ounce. Platinum, the best catalyst for the easier hydrogen half, is also scarce. A global hydrogen economy built on these metals is impossible. The central research challenge of the field is finding catalysts from abundant, cheap elements (iron, nickel, cobalt, molybdenum) that approach the activity of the precious metals.

Why this exists: hydrogen made from renewable electricity and water (green hydrogen) is the cleanest route to decarbonising steel, ammonia, shipping, and long-duration energy storage — sectors that batteries and direct electrification cannot reach. Catalyst cost and durability at industrial scale are the limiting factors.

Visual Beginner

Picture a tank of water with two metal plates (electrodes) dipping in. Wire the plates to a power source. At the negative plate (the cathode), hydrogen ions gain electrons and pair up as H2 bubbles. At the positive plate (the anode), water molecules surrender electrons, oxygen atoms pair up as O2, and the leftover protons drift back through the solution.

Beside the cell sits a volcano-shaped curve: catalytic activity on the vertical axis, binding strength of the key intermediate on the horizontal. Platinum sits at the peak for hydrogen; nickel-iron oxyhydroxide sits near the peak for oxygen in alkaline water. Materials on the left bind too weakly; materials on the right poison themselves by binding too strongly.

Worked example Beginner

A solar water-splitting device (the Nocera artificial leaf).

In 2011 Reece, Nocera, and co-workers reported in Science a wireless device that splits water under illumination. The structure is a triple-junction amorphous-silicon solar cell, with a cobalt-borate (Co-Bi) oxygen-evolution catalyst deposited on one face and a NiMoZn hydrogen-evolution catalyst on the other. Drop it in a beaker of neutral-pH water, shine light on it, and oxygen bubbles from one face while hydrogen bubbles from the other.

Step 1. The solar cell absorbs sunlight. The triple-junction a-Si stack provides an open-circuit photovoltage of about 2.0 volts, enough to drive both half-reactions in series.

Step 2. The photovoltage drives the oxygen-evolution reaction at the Co-Bi face. Co-Bi self-assembles from Co2+ ions and borate in solution under operating bias, so the catalyst repairs itself if damaged.

Step 3. The complementary photovoltage drives the hydrogen-evolution reaction at the NiMoZn face. Hydrogen gas collects at the surface.

The reported solar-to-hydrogen efficiency was 2.5 percent in the wireless configuration and 4.7 percent in a wired variant. What this tells us: artificial photosynthesis is technically feasible with earth-abundant catalysts; the remaining challenges are efficiency, durability, and scale, all areas of intense industrial and academic effort.

Check your understanding Beginner

Formal definition Intermediate+

A water-splitting electrocatalyst is a material that lowers the activation barrier of one of the two half-reactions of water electrolysis. The two half-reactions, written for acidic and alkaline media, are:

Reaction	Acidic medium	Alkaline medium
HER (cathode, reduction)	$2{\text{H}}^{+}+2e^{-}\to {\text{H}}_2$	$2{\text{H}}_2\text{O}+2e^{-}\to {\text{H}}_2+2{\text{OH}}^{-}$
OER (anode, oxidation)	$2{\text{H}}_2\text{O}\to {\text{O}}_2+4{\text{H}}^{+}+4e^{-}$	$4{\text{OH}}^{-}\to {\text{O}}_2+2{\text{H}}_2\text{O}+4e^{-}$

The HER is a two-electron reduction producing H2; the OER is a four-electron oxidation producing O2. The net reaction $2{\text{H}}_2\text{O}\to 2{\text{H}}_2+{\text{O}}_2$ is medium-independent.

Thermodynamics. The standard potential of the OER is $E^{\circ }=+1.23\text{V}$ vs. the standard hydrogen electrode (SHE); the HER potential is $0\text{V}$ by definition. The thermodynamic minimum cell voltage is therefore $1.23\text{V}$ at 25 °C. Practical operation at industrial current densities (1 to 2 A/cm${}^2$) requires 1.5 to 1.8 V; the excess is overpotential at both electrodes plus ohmic losses. System-level energy efficiency of commercial alkaline and PEM electrolyzers is 70 to 80 percent; solid-oxide electrolysis cells (SOEC) operating at 700 to 850 °C reach 85 to 92 percent by drawing part of the energy as heat.

Tafel kinetics. The current density $j$ at an electrode under pure activation control obeys the Tafel equation ^[Tafel1905]

$$\eta =a+b{\log }_{10}\left(\frac{|j|}{j_0}\right),$$

where $\eta =E-E_{\text{eq}}$ is the overpotential, $b$ is the Tafel slope in mV per decade of current density, and $j_0$ is the exchange current density. The Tafel slope diagnoses the rate-determining step (RDS). For HER on Pt in acid: $b=30\text{mV/dec}$ indicates Tafel recombination ($2{\text{H}}_{\text{ads}}\to {\text{H}}_2$) as RDS; $b=40\text{mV/dec}$ indicates Heyrovsky electrochemical desorption; $b=120\text{mV/dec}$ indicates the initial Volmer discharge is rate-limiting.

Sabatier principle and volcano plots. Catalytic activity peaks when the binding energy of the rate-limiting intermediate (H${}_{\text{ads}}$ for HER; *OH or *OOH for OER) is neither too strong nor too weak. Too weak: the intermediate desorbs before reacting further. Too strong: the surface saturates and the next elementary step cannot proceed. Activity plotted against binding energy traces a volcano with the optimal catalyst at the apex ^{[Sabatier1912]}.

Benchmark catalysts. HER: Pt is the gold standard, with $j_0\approx 10^{-3}{\text{A/cm}}^2$ on polycrystalline Pt in acid. Alternatives include MoS${}_2$ edge sites (Hinnemann 2005), NiMo alloys, and transition-metal phosphides (Ni${}_2$P, Co${}_2$P, MoP, FeP). OER: IrO${}_2$ and RuO${}_2$ are the gold standards in acid. In alkaline media, NiFe oxyhydroxide (NiFeOOH) is the most active known OER catalyst; the Fe site, not Ni, is the active centre ^{[Trotochaud2014]}. CoPi (Nocera 2008) self-assembles from Co${}^{2+}$ and phosphate in neutral water and self-heals under operating bias ^[Nocera2012].

Mechanisms. HER proceeds in three elementary steps. The Volmer step ${\text{H}}^{+}+e^{-}+\ast \to {\text{H}}_{\text{ads}}\ast$ (acid) or ${\text{H}}_2\text{O}+e^{-}+\ast \to {\text{H}}_{\text{ads}}\ast +{\text{OH}}^{-}$ (base) is the initial discharge. This is followed by either the Heyrovsky step ${\text{H}}_{\text{ads}}\ast +{\text{H}}^{+}+e^{-}\to {\text{H}}_2+\ast$ or the Tafel step $2{\text{H}}_{\text{ads}}\ast \to {\text{H}}_2+2\ast$. OER proceeds by a four-step adsorbate mechanism: ${\text{H}}_2\text{O}+\ast \to \ast \text{OH}+{\text{H}}^{+}+e^{-}$; $\ast \text{OH}\to \ast \text{O}+{\text{H}}^{+}+e^{-}$; $\ast \text{O}+{\text{H}}_2\text{O}\to \ast \text{OOH}+{\text{H}}^{+}+e^{-}$; $\ast \text{OOH}\to \ast +{\text{O}}_2+{\text{H}}^{+}+e^{-}$. The maximum of the four reaction free energies is the potential-determining step.

Device integration. Three electrolyzer architectures dominate. Proton-exchange-membrane (PEM) electrolyzers use a Nafion membrane, acidic environment, Pt HER catalyst, and IrO${}_2$ OER catalyst; they deliver high-purity H${}_2$ at high current density but require precious metals. Alkaline electrolyzers use concentrated KOH, nickel steel electrodes, and a porous separator; they are cheap and mature but limited to lower current densities. Solid-oxide electrolysis cells (SOEC) operate at 700 to 850 °C on a yttria-stabilised zirconia electrolyte, with Ni-YSZ cathodes and perovskite anodes; they reach the highest efficiency but face materials-stability challenges under thermal cycling.

Counterexamples to common slips

• “The best catalyst is the one that holds the intermediate most strongly.” No. Too-strong binding poisons the surface (Sabatier). The optimum sits at intermediate binding energy, where the intermediate engages long enough to react but releases the product. The volcano's descending strong-binding arm is the experimental signature.

• “Pt is essential for HER.” Almost. In acid at low loading, nothing matches Pt. But in alkaline media, well-structured MoS${}_2$ edges, NiMo alloys, and certain transition-metal phosphides approach Pt activity per active site. The remaining gap is at industrial current density, where base-metal catalysts still fall short.

• “IrO${}_2$ is essential for OER.” No. NiFe oxyhydroxide outperforms IrO${}_2$ in alkaline media at a fraction of the cost. IrO${}_2$'s role is as the only catalyst that is simultaneously active and stable in acid, where base metals dissolve. The constraint is medium, not absolute.

• “The Tafel slope alone identifies the rate-determining step.” Often combined with isotope effects, pH dependence, and DFT modelling to disambiguate. Tafel slope is necessary but not sufficient evidence for mechanism assignment.

• “Water splitting is energy-efficient.” Commercial electrolyzers operate at 70 to 80 percent system efficiency (LHV); the remainder goes to overpotentials, ohmic losses, gas separation, and balance-of-plant. SOEC reaches 85 to 92 percent by using high-temperature heat to drive the entropy change.

• “H${}_2$ is the future of all energy.” Hydrogen is part of the future — for hard-to-electrify sectors (steel via direct reduction, ammonia synthesis, shipping, long-duration storage). Passenger cars are dominated by batteries; H${}_2$ loses there on round-trip efficiency.

Key mechanism: Sabatier-volcano rational catalyst design Intermediate+

Sabatier's 1912 principle ^{[Sabatier1912]} holds that catalytic activity is a peaked function of the binding energy $\Delta G$ of the rate-limiting intermediate. Trasatti's 1972 volcano for HER ^{[Trasatti1972]}, plotting $\log j_0$ against the metal-hydrogen bond strength, placed Pt at the apex; metals on the weak-binding arm (Ta, W, Nb) failed to engage H, while metals on the strong-binding arm (Au, Cu in the corrected picture) bound H too strongly for the Tafel recombination to proceed. The empirical volcano preceded any microkinetic derivation by half a century.

The volcano from microkinetics. Consider HER on a metal surface $M$. The Volmer step ${\text{H}}^{+}+e^{-}+\ast \to {\text{H}}_{\text{ads}}\ast$ has forward rate $k_1(1-{\theta }_{\text{H}})$, reverse rate $k_{-1}{\theta }_{\text{H}}$, and equilibrium constant $K_{\text{V}}=k_1/k_{-1}\propto \exp (-\Delta G_{\text{H}}/kT)$. The Heyrovsky step ${\text{H}}_{\text{ads}}\ast +{\text{H}}^{+}+e^{-}\to {\text{H}}_2+\ast$ has rate $k_2{\theta }_{\text{H}}$; the Tafel step $2{\text{H}}_{\text{ads}}\ast \to {\text{H}}_2+2\ast$ has rate $k_3{\theta }_{\text{H}}^2$. Steady state on ${\theta }_{\text{H}}$ gives

$${\theta }_{\text{H}}=\frac{K_{\text{V}}}{1+K_{\text{V}}+k_2/k_1+2k_3{\theta }_{\text{H}}/k_1}\approx \frac{K_{\text{V}}}{1+K_{\text{V}}}$$

in the Tafel-Volmer regime with $k_2$ small. The current $j\propto k_3{\theta }_{\text{H}}^2\propto [K_{\text{V}}/(1+K_{\text{V}}){]}^2$ rises from zero at weak binding ($K_{\text{V}}\to 0$), reaches a maximum at intermediate $K_{\text{V}}\sim 1$ (corresponding to $\Delta G_{\text{H}}\approx 0$), then saturates and falls as $K_{\text{V}}\gg 1$ makes the Tafel step rate-limiting on a saturated surface. The peak corresponds to $\Delta G_{\text{H}}\approx 0$ — the computational hydrogen-electrode criterion.

The computational hydrogen electrode. Nørskov and co-workers (2004) ^{[Norskov2004]} introduced a thermodynamic framework that converts DFT-computed adsorption energies into HER and OER activity predictions. At electrode potential $U$ vs. SHE, the free energy of a proton-electron pair equals $-eU$ at standard conditions, so the reaction free energy of any elementary step involving $n$ protons and $n$ electrons is

$$\Delta G(U)=\Delta E_{\text{DFT}}+\Delta E_{\text{ZPE}}-T\Delta S-neU.$$

Setting $U=0$ V vs. SHE identifies $\Delta G=0$ with the reversible condition. A catalyst with $\Delta G_{{\text{H}}_{\text{ads}}}\approx 0$ at $U=0$ is predicted to lie at the volcano apex — a purely computational criterion that placed Pt, Pd, and certain MoS${}_2$ edge sites at the peak, in quantitative agreement with experiment ^{[Hinnemann2005]}. For OER, Nørskov's framework evaluates the four-step mechanism and identifies the catalyst minimising ${\max }_i\Delta G_i$ as optimal. This maximum is the potential-determining step, and the corresponding minimum overpotential is

$${\eta }_{\text{OER}}=\frac{{\max }_i\Delta G_i}{e}-1.23\text{V}.$$

The OER volcano is peaked when $\Delta G_{\ast \text{OOH}}-\Delta G_{\ast \text{O}}\approx 1.6$ eV (the scaling-relation value), with RuO${}_2$ and IrO${}_2$ at the peak in acid and NiFe oxyhydroxide near the peak in base. The scaling relation $\Delta G_{\ast \text{OOH}}-\Delta G_{\ast \text{O}}\approx$ const is the principal obstacle to closing the residual overpotential: no known catalyst breaks it cleanly.

Active-site identification in NiFe oxyhydroxide. Trotochaud, Young, Ranney, and Boettcher (2012-2014) ^{[Trotochaud2014]} showed that Fe — not Ni — is the active site in NiFe oxyhydroxide. Fe(III) sites embedded in a NiOOH host exhibit faster OER kinetics than either pure NiOOH or pure FeOOH. The Ni host tunes the Fe $d$-band centre toward the volcano peak, pushing Fe to the apex without losing Fe redox activity. This is the computational-volcano prediction realised in an earth-abundant catalyst. The structural insight is that the as-prepared material (NiFe-layered double hydroxide, NiFe-LDH) reconstructs under anodic bias to the active phase (NiFeOOH); operando spectroscopy (XAS, Raman) is required to identify the in-situ structure, since ex-situ characterisation captures the wrong phase.

Bridge. The computational hydrogen electrode builds toward 16.10.01 pending the catalysis survey as the DFT-grounded expression of the Sabatier principle, and appears again in 14.11.04 batteries and fuel cells as the same framework that predicts Pt-activity for the oxygen-reduction reaction (ORR) at the fuel-cell cathode. The foundational reason DFT-based catalyst design works at all is that adsorption energies are continuous functions of the metal $d$-band centre (Hammer-Nørskov 1995), and this is exactly the parameter that tunes a catalyst from the weak-binding to the strong-binding arm of the volcano. The central insight is that real catalysts are dynamic: NiFe oxyhydroxide reconstructs under operating bias, CoPi self-assembles from solution, and the active phase is generally not the as-prepared phase. Putting these together identifies operando DFT modelling of the electrochemically transformed surface, not the resting surface, as the substrate for rational catalyst design, and the bridge is between surface science and electrochemical engineering.

Exercises Intermediate+

Advanced results Master

Theorem 1 (Tafel 1905). Julius Tafel, working in Wurzburg on organic electrochemistry, established empirically that the overpotential $\eta$ at a polarised electrode is linear in ${\log }_{10}|j|$ over many decades of current density ^[Tafel1905]. The Tafel equation $\eta =a+b{\log }_{10}(|j|/j_0)$ held for HER on nine metals studied (Pt, Pd, Ni, Fe, Cu, Ag, Au, Hg, Pb), with $b$ in the range 0.04 to 0.12 V per decade and $j_0$ varying by ten orders of magnitude. The result preceded the Butler-Volmer formalism by half a century but contained its essential content: an exponential current-voltage relation under activation control. Tafel's slope parameter $b$ remains the standard diagnostic for the rate-determining step in electrocatalysis.

Theorem 2 (Bockris 1956). John O'M. Bockris, working at the University of Pennsylvania, developed the modern mechanistic framework for HER kinetics, identifying the Volmer, Heyrovsky, and Tafel elementary steps and showing how each combination produces a characteristic Tafel-slope signature. Bockris's textbook Modern Electrochemistry (with Reddy, 1970; second edition 2001) consolidated electrochemical thermodynamics and kinetics into a unified theoretical structure, providing the conceptual language in which the rest of the field operates. Bockris's "21st century" predictions of an electrified hydrogen economy (1962 onward) were three decades premature but framed the modern case for green hydrogen as a decarbonisation pillar.

Theorem 3 (Trasatti 1972 — HER volcano). Sergio Trasatti, working at the University of Milan, plotted $\log j_0$ for HER on transition metals against the metal-hydrogen bond energy $D_{\text{M-H}}$ and obtained a volcano-shaped curve with the Pt-group metals at the apex ^{[Trasatti1972]}. The weak-binding arm (Ta, W, Nb, Ti) was limited by the rate of H adsorption; the strong-binding arm (Au, Cu, in Trasatti's corrected picture) was limited by the rate of H${}_2$ desorption. The result placed Sabatier's 1912 optimal-binding principle on a quantitative footing for electrochemistry and remains the canonical empirical volcano. Subsequent corrections (Khan, Rheenen, Srinivasan) refined the strong-binding arm without disturbing the peak position.

Theorem 4 (Hinnemann-Norskov 2005 — MoS${}_2$ edges). Hinnemann, Moses, Bonde, Jorgensen, Nielsen, Horch, Chorkendorff, and Norskov combined DFT calculations with experimental synthesis to identify theedges of layered molybdenum disulfide (MoS${}_2$) as HER-active sites with $\Delta G_{{\text{H}}_{\text{ads}}}\approx 0.08$ eV, close to the Pt apex of the HER volcano ^{[Hinnemann2005]}. The basal plane of MoS${}_2$ is catalytically dead ($\Delta G_{{\text{H}}_{\text{ads}}}\approx 1.8$ eV), but the (10-10) edge termination sits near the volcano peak. The prediction was confirmed experimentally by measuring HER activity on MoS${}_2$ nanoparticles whose edge-site density was tuned independently of basal-plane area. The work established DFT-driven materials discovery as a productive strategy for non-precious-metal electrocatalysis and triggered a decade of research into MoS${}_2$, MoSe${}_2$, WS${}_2$, and other layered transition-metal chalcogenides.

Theorem 5 (Norskov 2004 — computational hydrogen electrode). Norskov, Bligaard, Logadottir, Kitchin, Chen, Pandelov, and Stimming introduced a thermodynamic framework that converts DFT-computed adsorption free energies into electrochemical reaction energies at arbitrary electrode potential $U$ ^{[Norskov2004]}. The key identity is that at $U=0$ V vs. SHE, one proton-electron pair is in equilibrium with H${}_2$ at standard pressure, so $\frac{1}{2}{\text{H}}_2\rightleftharpoons {\text{H}}^{+}+e^{-}$ at $\Delta G=0$. Therefore $\Delta G$ for any proton-coupled electron-transfer step at potential $U$ equals $\Delta G_{\text{DFT}}-neU$, eliminating the need to model the electrode potential explicitly. The framework reproduces the Trasatti volcano for HER from first principles and predicts the OER volcano on rutile oxides (Norskov 2005 Nature Mater. on "Origin of Overpotential"), identifying IrO${}_2$ and RuO${}_2$ at the peak. The computational hydrogen electrode is the foundational tool of modern computational electrocatalysis.

Theorem 6 (Nocera 2008-2012 — CoPi and the artificial leaf). Daniel Nocera and co-workers at MIT reported in 2008 that electrodeposition of Co${}^{2+}$ from a phosphate-buffered neutral solution under anodic bias produces a thin film of amorphous cobalt-oxyhydroxide-phosphate ("CoPi") that catalyses OER at $\eta \approx 0.4$ V in neutral water, self-assembling from solution and self-healing under operating conditions ^[Nocera2012]. The catalyst operates at room temperature and neutral pH, conditions compatible with biological systems and ambient water. In 2011, Reece, Nocera, and co-workers wired CoPi (anode) and a NiMoZn alloy (cathode) to a triple-junction amorphous-silicon solar cell, producing a wireless artificial-leaf device that splits water under illumination at 2.5 percent solar-to-hydrogen efficiency ^[Reece2011]. The result demonstrated that earth-abundant catalysts can support unassisted solar water splitting, with self-healing kinetics compensating for the catalyst's modest thermodynamic activity. The Chem. Rev. 2012 article consolidated the underlying chemistry and articulated the "personalised energy" framing.

Theorem 7 (Trotochaud-Boettcher 2012-2014 — Fe active site in NiFe oxyhydroxide). Trotochaud, Young, Ranney, and Boettcher at the University of Oregon used a combination of thin-film synthesis, electrochemical kinetics, and operando conductivity measurements to identify the Fe(III) site — not Ni — as the OER active centre in NiFe oxyhydroxide ^{[Trotochaud2014]}. Pure NiOOH is a poor OER catalyst (overpotential > 0.45 V at 10 mA/cm${}^2$); adding 10 to 40 percent Fe reduces the overpotential to < 0.30 V, the lowest for any OER catalyst in alkaline media. The Ni host provides conductivity and redox flexibility; the Fe site provides the optimal *O and *OOH binding energies. Subsequent operando XAS and Raman work (Klaus, Cai, Strasser groups) confirmed that the Fe site remains Fe(III) under OER operation. This is the computational-volcano prediction realised in an earth-abundant catalyst, and is the basis for the modern NiFe-LDH OER catalysts used in commercial alkaline and AEM electrolyzers.

Theorem 8 (Jin 2014-2018 — transition-metal phosphides as HER catalysts). Song Jin and co-workers at Wisconsin-Madison, building on the early work of Liu and Rodriguez on Ni${}_2$P (2005), developed nanostructured transition-metal phosphides (Ni${}_2$P, Co${}_2$P, FeP, MoP, WP) as acid-stable HER catalysts with overpotentials below 100 mV at 10 mA/cm${}^2$ in acidic media, rivalling Pt at low loading. The phosphorus sites act as proton acceptors and the metal sites as hydride acceptors, mimicking the bifunctional active site of hydrogenase enzymes. The class is now the leading non-precious-metal HER catalyst family for PEM electrolyzers, with durability exceeding 10${}^4$ hours in single-cell tests (Plymouth Laboratory, 2022).

Synthesis. The Tafel 1905 empirical rate law builds toward the modern microkinetic derivation of volcano plots from elementary-step rate constants, and appears again in 14.08.04 catalysis mechanisms as the same Arrhenius kinetics applied to thermal catalysis. The foundational reason DFT-based catalyst design works at all is that adsorption energies scale linearly across transition-metal surfaces (the scaling relations of Abild-Pedersen 2007), and this is exactly the regularity that lets a single descriptor ($\Delta G_{{\text{H}}_{\text{ads}}}$ for HER; $\Delta G_{\ast \text{O}}$ and $\Delta G_{\ast \text{OOH}}$ for OER) predict activity for an entire family of materials. Putting these together with the Boettcher operando spectroscopy identifying Fe as the active site in NiFe oxyhydroxide identifies the field's design loop: compute the volcano, identify a candidate near the apex, synthesise and test operando, refine the descriptor.

The central insight is that the Sabatier principle is invariant across the half-reactions — HER, OER, ORR, CO${}_2$ reduction all sit on their own volcano, and the bridge is the universal peaked-in-$\Delta G$ shape with apex at intermediate binding. The pattern generalises to photoelectrochemical water splitting, where the catalyst is wired to a light absorber and the operating potential is set by the photovoltage; this is exactly the artificial-leaf architecture of Nocera 2011.

Full proof set Master

Proposition 1 (Tafel slope from Butler-Volmer). For a one-electron rate-determining step at an electrode, in the high-overpotential cathodic limit $|\eta |\gg RT/F$ with $\eta <0$, the Tafel slope is

$$b=\frac{2.303RT}{\alpha F}=\frac{0.05916}{\alpha }\text{V/decade at }T=298\text{K},$$

where $\alpha \in (0,1)$ is the symmetry (charge-transfer) coefficient. For $\alpha =1/2$, $b=118\text{mV/dec}$.

Proof. The Butler-Volmer equation for the net current density at an electrode undergoing a one-electron charge-transfer rate-determining step is

$$j=j_0\left[\exp \left(\frac{(1-\alpha )F\eta }{RT}\right)-\exp \left(-\frac{\alpha F\eta }{RT}\right)\right],$$

with $j_0$ the exchange current density, $\eta =E-E_{\text{eq}}$ the overpotential, and $\alpha$ the symmetry factor. Consider the cathodic branch ($\eta <0$, $|\eta |\gg RT/F$). Then $(1-\alpha )\eta <0$, so $\exp [(1-\alpha )F\eta /RT]\to 0$, and the cathodic term dominates:

$$|j|=j_0\exp \left(\frac{\alpha F|\eta |}{RT}\right).$$

Take ${\log }_{10}$:

$${\log }_{10}(|j|)={\log }_{10}(j_0)+\frac{\alpha F|\eta |}{2.303RT}.$$

Solve for $|\eta |$:

$$|\eta |=\frac{2.303RT}{\alpha F}{\log }_{10}\left(\frac{|j|}{j_0}\right).$$

Compare with the Tafel form $|\eta |=a+b{\log }_{10}(|j|/j_0)$ and identify $b=2.303RT/(\alpha F)$. At $T=298\text{K}$, $2.303RT/F=0.05916\text{V}$, giving $b=0.05916/\alpha \text{V/dec}$. For $\alpha =1/2$, $b=0.118\text{V/dec}=118\text{mV/dec}$. $\square$

Proposition 2 (Sabatier volcano from Volmer-Tafel microkinetics). In the Volmer-Tafel regime for HER, the steady-state current density satisfies

$$|j|\propto {\left(\frac{K_{\text{V}}}{1+K_{\text{V}}}\right)}^2,$$

where $K_{\text{V}}=k_1/k_{-1}\propto \exp (-\Delta G_{\text{H}}/RT)$ is the Volmer equilibrium constant. This function rises from zero at weak binding ($K_{\text{V}}\to 0$, $\Delta G_{\text{H}}\to +\infty$), reaches a maximum at $K_{\text{V}}\approx 1$ ($\Delta G_{\text{H}}\approx 0$), and falls off as $K_{\text{V}}^{-1}$ at strong binding ($K_{\text{V}}\to \infty$, $\Delta G_{\text{H}}\to -\infty$).

Proof. Steady-state balance on ${\theta }_{\text{H}}$:

$$\frac{d{\theta }_{\text{H}}}{dt}=k_1(1-{\theta }_{\text{H}})-k_{-1}{\theta }_{\text{H}}-2k_3{\theta }_{\text{H}}^2=0,$$

where the factor of 2 in the Tafel-step rate reflects that two adsorbed H atoms are consumed per event. Neglecting $k_3$ relative to $k_1$ in the steady-state balance (Tafel-Volmer regime: Tafel step is slow, Volmer step is in quasi-equilibrium), this reduces to $k_1(1-{\theta }_{\text{H}})\approx k_{-1}{\theta }_{\text{H}}$, so

$${\theta }_{\text{H}}=\frac{K_{\text{V}}}{1+K_{\text{V}}},K_{\text{V}}=k_1/k_{-1}.$$

The Tafel-step rate per unit area is $k_3{\theta }_{\text{H}}^2$, and each Tafel event releases one H${}_2$ molecule carrying two electrons through the external circuit, so $|j|=2e{k}_3{\theta }_{\text{H}}^2\cdot N_{\text{s}}$, with $N_{\text{s}}$ the density of active sites:

$$|j|=2e{k}_3{N}_{\text{s}}{\left(\frac{K_{\text{V}}}{1+K_{\text{V}}}\right)}^2.$$

The function $f(K)=[K/(1+K){]}^2$ satisfies $f(0)=0$, $f(K)\to 1$ as $K\to \infty$, with a single inflection point; but the current decreases at strong binding because $k_3$ itself depends on the H desorption barrier, which rises with $\Delta G_{\text{H}}$ on the strong-binding arm. Including $k_3(\Delta G_{\text{H}})={\nu }_0\exp [-(\Delta G_{\text{Tafel}}^{‡})/RT]$ with $\Delta G_{\text{Tafel}}^{‡}$ increasing with $|\Delta G_{\text{H}}|$ on the strong-binding side, the full $|j|(\Delta G_{\text{H}})$ is peaked. The maximum is at $\Delta G_{\text{H}}\approx 0$, where H${}_2$ desorption is fast and H${}_{\text{ads}}$ coverage is high but not saturated, matching the Norskov computational hydrogen-electrode prediction. $\square$

Connections Master

• Catalysis survey 16.10.01 pending. This unit is the deep dive into electrocatalysis beneath the catalysis-survey treatment of 16.10.01 pending, which introduced the Sabatier principle in a single paragraph and showed the Wilkinson, olefin-metathesis, Haber-Bosch, three-way-automotive, and enzyme cycles as parallel examples. The foundational bridge from that unit to here is that the same Sabatier principle (optimal binding) governs thermal heterogeneous catalysis and electrocatalysis alike; putting these together identifies the volcano plot as a universal design tool, and identifies DFT-based catalyst design (Norskov 2004) as the unifying computational framework. This is exactly the framework that appears again in the catalysis survey's treatment of Ertl's ammonia-synthesis surface science, with adsorption energies as the common descriptor.

• Perovskite solar cells 16.07.05. The Nocera 2011 artificial leaf and the perovskite solar cell of 16.07.05 share the architecture of a light absorber driving electrochemical work, with the perovskite-on-silicon tandem ($E_{g1}\approx 1.68$ eV, $E_{g2}\approx 1.12$ eV) supplying the photovoltage required to exceed the 1.23 V thermodynamic minimum for water splitting. The central insight is that photoelectrochemical water splitting — using a perovskite photoanode directly wired to a NiFe oxyhydroxide OER catalyst — generalises the artificial-leaf concept from a-Si to perovskite absorbers, with higher photovoltage enabling higher solar-to-hydrogen efficiency. The pattern generalises further to perovskite-driven CO${}_2$ reduction, where the same catalyst-design loop (volcano + DFT + operando spectroscopy) applies.

• Batteries and fuel cells 14.11.04. The HER is the inverse of the hydrogen-oxidation reaction at the fuel-cell anode, and the OER is the inverse of the oxygen-reduction reaction (ORR) at the fuel-cell cathode; the same catalysts (Pt for HER/HOR, IrO${}_2$ and Pt for OER/ORR) appear in both electrolyzers and fuel cells, with the direction of current determining whether the cell stores or releases energy. The foundational reason is microscopic reversibility: a catalyst lowers the activation barrier by the same amount in both directions, leaving equilibrium unchanged. This is exactly the symmetry that the Norskov computational hydrogen electrode exploits, and the bridge is between the energy-storage and energy-release halves of the hydrogen cycle. The thermodynamic ceiling (1.23 V per cell) appears in 14.11.04 as the Carnot-style detailed-balance limit familiar from the Shockley-Queisser photovoltaic ceiling.

• Catalysis mechanisms 14.08.04. The Tafel kinetics, Butler-Volmer formalism, and Sabatier principle treated here are the electrochemical specialisation of the Arrhenius kinetics and Michaelis-Menten enzyme kinetics of 14.08.04, with the electrode potential as an additional thermodynamic drive that thermal catalysis lacks. The central insight is that every catalytic cycle is a closed loop of elementary steps, and the rate-determining-step identification (via Tafel slope, isotope effect, or Hammett substituent) uses the same kinetic toolkit across homogeneous, heterogeneous, enzymatic, and electrocatalytic regimes. The pattern generalises from chemical catalysis to electrochemical catalysis by adding the electrode potential as a continuous activation-energy tuner, which is exactly what makes computational electrocatalysis more predictive than computational thermal catalysis.

Historical & philosophical context Master

Julius Tafel, working in Wurzburg on the electrochemistry of organic substrates, established the empirical Tafel equation in 1905 ^[Tafel1905] by measuring overpotentials at metal electrodes across nine decades of current density. The relation $\eta =a+b{\log }_{10}(|j|/j_0)$ preceded the Butler-Volmer formalism by half a century but contained its essential content; Tafel's slope parameter $b$ remains the standard diagnostic for the rate-determining step in electrocatalysis. The mechanism underlying Tafel's law was clarified by John O'M. Bockris's school in the 1950s, which identified the Volmer, Heyrovsky, and Tafel elementary steps of HER and mapped Tafel slopes to rate-determining-step assignments; Bockris's Modern Electrochemistry (1970, with Reddy) consolidated the framework into the form taught today.

Sergio Trasatti, working at the University of Milan, plotted $\log j_0$ for HER against the metal-hydrogen bond energy in 1972 ^{[Trasatti1972]} and obtained the canonical volcano, placing Pt at the apex. The result placed Sabatier's 1912 optimal-binding principle ^{[Sabatier1912]} on a quantitative empirical footing for electrochemistry. Three decades later, Jens Norskov and co-workers at DTU Copenhagen derived the volcano from first principles using density-functional theory and the computational hydrogen electrode (2004) ^{[Norskov2004]}, reproducing Trasatti's volcano for HER and predicting the OER volcano on oxide surfaces (Nature Mater. 2005). The Norskov framework identified $\Delta G_{{\text{H}}_{\text{ads}}}\approx 0$ as the design rule for HER catalysts, prompting the experimental confirmation that MoS${}_2$ edge sites are HER-active (Hinnemann 2005) ^{[Hinnemann2005]} and initiating the modern era of DFT-driven catalyst discovery.

The non-precious-metal OER programme crystallised with Daniel Nocera's 2008 report of CoPi, a self-assembling cobalt-phosphate OER catalyst operating in neutral water ^[Nocera2012], and the 2011 "artificial leaf" ^[Reece2011] wiring CoPi and NiMoZn to a triple-junction silicon solar cell. The Fe-active-site identification in NiFe oxyhydroxide by Trotochaud and Boettcher (2012-2014) ^{[Trotochaud2014]} closed the loop for alkaline OER, identifying the most active known non-precious-metal catalyst and confirming the computational-volcano prediction in an earth-abundant material. The programme continues with transition-metal phosphides (Jin 2014-2018), single-atom M-N-C catalysts, and operando spectroscopic identification of in-situ active phases.

Bibliography Master

@article{Tafel1905,
  author = {Tafel, J.},
  title = {Ueber die Polarisation bei kathodischer Wasserstoffentwicklung},
  journal = {Z. Phys. Chem.},
  volume = {50},
  year = {1905},
  pages = {641--712}
}

@article{Trasatti1972,
  author = {Trasatti, S.},
  title = {The Work Function of Metals and the Energetics of the Hydrogen Electrode Reaction},
  journal = {J. Electroanal. Chem.},
  volume = {39},
  year = {1972},
  pages = {163--184}
}

@article{Norskov2004,
  author = {Norskov, J. K. and Bligaard, T. and Logadottir, A. and Kitchin, J. R. and Chen, J. G. and Pandelov, S. and Stimming, U.},
  title = {Origin of Overpotential in Hydrogen and Oxygen Electrocatalysis},
  journal = {J. Phys. Chem. B},
  volume = {108},
  year = {2004},
  pages = {17886--17892}
}

@article{Hinnemann2005,
  author = {Hinnemann, B. and Moses, P. G. and Bonde, J. and Jorgensen, K. P. and Nielsen, J. H. and Horch, S. and Chorkendorff, I. and Norskov, J. K.},
  title = {Biomimetic Hydrogen Evolution: {MoS$_2$} Nanoparticles as Catalyst for Hydrogen Evolution},
  journal = {J. Am. Chem. Soc.},
  volume = {127},
  year = {2005},
  pages = {5308--5309}
}

@article{Reece2011,
  author = {Reece, S. Y. and Hamel, J. A. and Sung, K. and Jarvi, T. D. and Esswein, A. J. and Pijpers, J. J. H. and Nocera, D. G.},
  title = {Wireless Solar Water Splitting Using Silicon-Based Semiconductors and Earth-Abundant Catalysts},
  journal = {Science},
  volume = {334},
  year = {2011},
  pages = {645--648}
}

@article{Nocera2012,
  author = {Nocera, D. G.},
  title = {The Artificial Leaf},
  journal = {Acc. Chem. Res.},
  volume = {45},
  year = {2012},
  pages = {767--776}
}

@article{Trotochaud2014,
  author = {Trotochaud, L. and Young, S. L. and Ranney, J. K. and Boettcher, S. W.},
  title = {Nickel-Iron Oxyhydroxide Oxygen-Evolution Electrocatalysts: Activity and Dissolution Pathways},
  journal = {J. Am. Chem. Soc.},
  volume = {136},
  year = {2014},
  pages = {6744--6753}
}

@article{Sabatier1912,
  author = {Sabatier, P.},
  title = {La Catalyse en Chimie Organique},
  journal = {B{\'e}raud, Paris},
  year = {1913}
}

@book{Bockris2001,
  author = {Bockris, J. O'M. and Reddy, A. K. N. and Gamboa-Aldeco, M.},
  title = {Modern Electrochemistry 2A: Fundamentals of Electrodics},
  publisher = {Kluwer Academic / Springer},
  edition = {2nd},
  year = {2001}
}

@book{Koper2009,
  author = {Koper, M. T. M.},
  title = {Fuel Cell Catalysis: A Surface Science Approach},
  publisher = {Wiley},
  year = {2009}
}