Hydrogen evolution reaction
The hydrogen evolution reaction (HER) is a two-electron electrochemical reduction process that generates molecular hydrogen from protons or water at a cathode, prototypically expressed as $2\mathrm{H}^+ + 2\mathrm{e}^- \rightarrow \mathrm{H}_2 in acidic electrolytes or $2\mathrm{H}_2\mathrm{O} + 2\mathrm{e}^- \rightarrow \mathrm{H}_2 + 2\mathrm{OH}^- in alkaline media. This reaction proceeds through adsorbed hydrogen intermediates via stepwise mechanisms—Volmer (proton discharge), Heyrovsky (electrochemical recombination), or Tafel (chemical desorption)—with kinetics governed by the free energy of hydrogen adsorption on the catalyst surface. HER efficiency is quantified by metrics such as overpotential (the excess voltage beyond the thermodynamic potential required to achieve practical current densities, e.g., 10 mA/cm²), Tafel slope (indicating rate-limiting steps), and exchange current density, which collectively determine turnover frequency and stability under operational conditions.[1] As the cathodic counterpart to the oxygen evolution reaction in water electrolysis, HER underpins scalable hydrogen production for energy storage and fuel synthesis, converting intermittent renewable electricity into chemically storable fuel with minimal carbon emissions when powered by clean sources.[2] Platinum serves as the benchmark catalyst due to its near-zero overpotential and favorable adsorption energetics near the volcano peak of the Sabatier principle, enabling high catalytic activity across pH regimes, yet its high cost and scarcity—exacerbated by automotive and jewelry demands—limit industrial viability.[3] Intensive research since the 1970s has targeted earth-abundant alternatives, including transition metal sulfides (e.g., MoS₂ edge sites mimicking Pt's basal planes), phosphides, carbides, and single-atom catalysts on supports like nitrogen-doped carbon, which achieve competitive overpotentials (e.g., <100 mV at 10 mA/cm²) through tailored d-band centers and defect engineering.[4] Defining challenges persist in alkaline environments, where slower water dissociation and oxophilicity mismatches elevate barriers compared to acid, prompting hybrid strategies like bifunctional catalysts integrating HER with oxidation reactions for full electrolyzer efficiency.[5] Notable achievements include nanostructured Ni-Mo alloys sustaining industrial currents (>500 mA/cm²) with durability exceeding 1000 hours, advancing prospects for gigawatt-scale deployment amid global hydrogen targets.[6]Fundamentals
Definition and Basic Principles
The hydrogen evolution reaction (HER) is the cathodic half-reaction in water electrolysis that produces molecular hydrogen gas (H₂) through the electrochemical reduction of protons (H⁺) or water (H₂O).[7] In acidic electrolytes, the overall process is represented by the equation 2H⁺ + 2e⁻ → H₂, with a standard reversible potential of 0 V versus the standard hydrogen electrode (SHE) at pH 0 and 25°C.[1] In alkaline media, the reaction shifts to 2H₂O + 2e⁻ → H₂ + 2OH⁻, maintaining the same thermodynamic potential but adjusted for pH-dependent activities, where the equilibrium potential becomes -0.059 × pH V versus SHE.[8] This two-electron transfer process is fundamental to electrolytic hydrogen production, serving as the counterpart to the oxygen evolution reaction (OER) at the anode in overall water splitting.[9] At its core, the HER involves the adsorption of hydrogen intermediates on the catalyst surface, followed by their recombination into H₂ gas, governed by the interplay of adsorption free energy and electron transfer kinetics.[1] The reaction's efficiency is limited by overpotential—the extra voltage beyond the thermodynamic minimum required to drive appreciable current densities—arising from kinetic barriers such as slow proton discharge or H₂ desorption.[7] Catalysts, typically metals or alloys, facilitate the process by optimizing the binding strength of atomic hydrogen (H_ads) to neither too weak (hindering adsorption) nor too strong (impeding release), as described by the Sabatier principle.[9] Platinum exemplifies an ideal catalyst with near-zero overpotential at 10 mA/cm² in acid, achieving turnover frequencies exceeding 100 s⁻¹ under standard conditions, though scalability is constrained by scarcity.[1] The HER's principles extend to non-aqueous or microbial systems, but in conventional electrolysis, pH influences the rate-determining step: proton reduction dominates in acid, while water dissociation becomes rate-limiting in base due to higher H₂O activation energy.[8] Exchange current densities, a measure of intrinsic activity, vary by catalyst; for instance, Pt yields ~1 mA/cm², while non-precious alternatives like MoS₂ achieve ~10⁻³ mA/cm², highlighting the need for active sites mimicking Pt's d-band center position.[9] Faradaic efficiency approaches 100% under optimized conditions, ensuring minimal side reactions like hydride formation, though gas bubble evolution can introduce mass transport limitations at high currents (>100 mA/cm²).[1]Thermodynamic Aspects
The hydrogen evolution reaction (HER) in acidic media is represented by the half-cell reduction 2H⁺(aq) + 2e⁻ ⇌ H₂(g), with a standard reduction potential E° = 0.00 V versus the standard hydrogen electrode (SHE) defined at 25°C, unit activity of H⁺ (a_{H⁺} = 1), and H₂ fugacity of 1 bar.[10] This definition anchors the thermodynamic scale for electrochemical potentials, such that the standard Gibbs free energy change ΔG° = -nFE° = 0 kJ/mol for the reaction under these conditions, indicating equilibrium with no net driving force.[10] The equilibrium potential E_eq for HER deviates from E° according to the Nernst equation: E = E° - (RT/2F) ln(P_{H₂} / a_{H⁺}^2), where R is the gas constant, T is temperature in Kelvin, F is the Faraday constant, P_{H₂} is the partial pressure of H₂, and activities reflect non-standard conditions.[10] At 25°C and P_{H₂} = 1 bar, this simplifies to E ≈ -0.0592 pH V versus SHE, yielding a pH-dependent shift of -59.2 mV per unit increase in pH; for example, at pH 14 (corresponding to 1 M OH⁻ in alkaline media, where the effective reaction is 2H₂O(l) + 2e⁻ ⇌ H₂(g) + 2OH⁻(aq)), E_eq ≈ -0.828 V.[10] This pH dependence arises from the low proton activity in alkaline solutions, maintained by water autoprotolysis (K_w = 10^{-14} at 25°C), and implies a corresponding ΔG = -2FE_eq > 0 kJ/mol for reduction at potentials positive of E_eq, rendering the process non-spontaneous without applied voltage.[10] [11] Thermodynamic favorability for HER is thus governed by ΔG = ΔG° + RT ln Q, where Q is the reaction quotient, linking directly to the electrode potential via ΔG = -nFE; at E < E_eq (more negative), ΔG < 0, driving spontaneous evolution, though kinetic barriers necessitate overpotentials in practice.[10] Temperature influences E_eq through the RT term in the Nernst equation and subtle changes in activity coefficients, with the pH slope increasing to approximately -71 mV/pH at 80°C due to altered entropy contributions.[10] In electrocatalysis contexts, surface thermodynamics further modulate HER via the adsorption free energy of atomic hydrogen (ΔG_{H^*}), ideally near 0 eV for balanced Volmer step energetics, as deviations lead to either weak binding (high Volmer barrier) or strong binding (hindered desorption), per density functional theory assessments.[12] Electrolyte composition, such as high salt concentrations, can shift E_eq positively by enhancing proton activity, reducing ΔG for HER onset by up to 71 mV in concentrated aqueous solutions.[12]Kinetic Mechanisms
The hydrogen evolution reaction (HER) kinetics are governed by a sequence of elementary steps involving proton reduction and hydrogen desorption, which vary with pH, catalyst surface coverage of adsorbed hydrogen (θ_H), and electrode potential. In acidic media, the Volmer step entails the electrochemical discharge of a hydronium ion to form adsorbed atomic hydrogen: H₃O⁺ + e⁻ → H_ads + H₂O, with a rate often expressed as i_V = f_0 θ_{(1-θ)} exp(-α f η), where f_0 is the frequency factor, θ is the H_ads coverage, α is the transfer coefficient, f = F/RT, and η is the overpotential.[13] The subsequent desorption can occur via electrochemical (Heyrovsky) or chemical (Tafel) recombination, determining the dominant pathway.[14] In alkaline media, the Volmer step modifies to H₂O + e⁻ → H_ads + OH⁻, introducing water dissociation as a potential precursor influencing kinetics.[15] The Volmer-Heyrovsky mechanism predominates at low θ_H (<0.5), where the rate-determining step (RDS) is typically the Volmer reaction or Heyrovsky desorption: H_ads + H₃O⁺ + e⁻ → H₂ + H₂O (acidic). This pathway yields a Tafel slope of approximately 120 mV/decade if Volmer is RDS, reflecting single-electron transfer limitations, or ~40-50 mV/decade if Heyrovsky limits at higher overpotentials.[13] Experimental validation often employs electrochemical impedance spectroscopy (EIS) to deconvolute charge transfer resistance and confirm Heyrovsky dominance on catalysts like NiMo in alkaline conditions, where increasing pH accelerates Volmer relative to Heyrovsky.[15] [16] Conversely, the Volmer-Tafel mechanism prevails at high θ_H (≈1), favoring chemical recombination as the RDS: 2 H_ads → H₂, which is potential-independent and exhibits a low Tafel slope of ~30 mV/decade due to second-order dependence on coverage.[13] This pathway is observed on highly active platinum surfaces under high overpotentials or specific single-atom catalysts, where switching from Volmer-Heyrovsky enhances efficiency by reducing activation barriers in Tafel via bond-exchange.[17] [18] Microkinetic modeling, incorporating explicit solvent effects, reveals that Tafel RDS correlates with near-optimal H binding energies (ΔG_H ≈ 0 eV), minimizing overpotential.[19] Mechanism identification relies on Tafel plots (log|i| vs. η), where slopes diagnose RDS: 120 mV/dec for Volmer, 40 mV/dec for Heyrovsky or Tafel under specific coverages, though deviations arise from mass transport or pH gradients.[20] Electrolyte effects, such as anion adsorption or cation promotion of water dissociation, modulate θ_H and thus pathway preference, with alkaline HER often showing slower Volmer kinetics due to limited H₂O availability.[21] Quantitative models emphasize causal links between surface θ_H dynamics and current density, avoiding assumptions of equilibrium without empirical verification via rotating disk electrode data.[5]Historical Development
Early Discoveries in Electrolysis
The voltaic pile, invented by Alessandro Volta in early 1800, provided the first sustained electric current necessary for practical electrolysis experiments, enabling the observation of gas evolution at electrodes.[22] In May 1800, English chemists William Nicholson and Anthony Carlisle immersed platinum wires in water connected to Volta's device and observed bubbles forming at both electrodes: a larger volume of inflammable gas (hydrogen) at the wire linked to the pile's negative terminal (cathode) and a smaller volume of combustible gas (oxygen) at the positive terminal (anode).[23] [24] This marked the initial documented instance of the hydrogen evolution reaction (HER), where protons or water are reduced to hydrogen gas, though the underlying electrochemical mechanism remained unrecognized at the time. Nicholson and Carlisle's experiments confirmed a stoichiometric ratio of approximately 2:1 for hydrogen to oxygen volumes, consistent with water's composition as later quantified, and highlighted the role of electrode material in gas production efficiency.[22] On platinum, hydrogen evolved steadily without electrode corrosion, whereas base metals like iron promoted hydrogen release but underwent dissolution, suggesting early, albeit qualitative, insights into catalytic selectivity for HER.[25] These findings, reported in Nicholson's Journal of Natural Philosophy, Chemistry and the Arts, demonstrated electrolysis as a method to decompose water, laying empirical groundwork for HER as a cathodic process dependent on applied voltage exceeding the theoretical decomposition potential. Preceding these by over a decade, Dutch chemists Jan Rudolph Deiman and Adriaan Paets van Troostwijk achieved transient water decomposition in 1789 using a high-voltage electrostatic generator and Leyden jar, producing small quantities of hydrogen and oxygen gases, but the discontinuous nature limited sustained HER observation.[26] The 1800 breakthrough with galvanic current thus represented a causal advancement, shifting from sporadic sparks to controllable electrolytic hydrogen production and inspiring subsequent quantitative studies by Humphry Davy, who scaled up electrolysis in 1806–1807 to isolate alkali metals while noting persistent hydrogen evolution challenges on certain surfaces.[22]Evolution of Mechanistic Understanding
The mechanistic understanding of the hydrogen evolution reaction (HER) emerged in the early 20th century through empirical studies of cathodic overpotentials. In 1905, Julius Tafel reported a logarithmic relationship between overpotential and current density during hydrogen evolution on platinum electrodes in acidic media, yielding the Tafel equation \eta = a + b \log j, where \eta is the overpotential, j is the current density, a is a constant, and b (the Tafel slope) reflects kinetic parameters such as electron transfer coefficients.[27] This empirical relation highlighted the non-ideal kinetics of HER but lacked a detailed atomic-level description, attributing deviations to surface processes without specifying intermediates.[28] By the 1930s, foundational mechanistic models were proposed, incorporating adsorbed hydrogen intermediates. Max Volmer and coworkers introduced the Volmer step, describing the electrochemical adsorption of protons (\ce{H+ + e- -> H_{ads}}), while the Heyrovsky step outlined electrochemical desorption (\ce{H_{ads} + H+ + e- -> H2}), and the Tafel step captured recombinative desorption (\ce{2 H_{ads} -> H2}).[28] These elementary steps enabled two primary pathways—Volmer-Tafel (recombination-dominated) and Volmer-Heyrovsky (desorption-dominated)—with the rate-determining step inferred from Tafel slopes: ~120 mV/dec for Volmer-limited kinetics, ~40 mV/dec for Tafel-limited, and ~30 mV/dec for Heyrovsky-limited under low coverage assumptions.[29] Early kinetic analyses assumed steady-state adsorption isotherms, such as Langmuir, to model surface coverage \theta, where HER rate r \propto \theta (1 - \theta) for recombination paths.[29] Post-1950s advancements refined these models through electrochemical impedance spectroscopy and microelectrode techniques, revealing deviations from ideal Tafel behavior due to mass transport, pH effects, and catalyst surface heterogeneity. Studies distinguished mechanisms on metals like Pt (often Volmer-Heyrovsky) versus Ni (Volmer-Tafel), with Tafel slopes varying by electrolyte pH and overpotential, indicating shifts in rate-determining steps or coverage-dependent kinetics.[30] In alkaline media, slower Volmer steps arise from water dissociation requirements (\ce{H2O + e- -> H_{ads} + OH-}), prompting bifunctional catalyst designs to enhance proton supply.[31] Contemporary insights, accelerated since the 2000s, integrate density functional theory (DFT) simulations and in situ spectroscopy (e.g., Raman, XAS) to validate adsorbed H binding energies near the volcano peak for optimal activity, per Sabatier principle, while addressing non-Faradaic processes and double-layer influences on proton transfer.[32] These developments underscore that early discharge-recombination frameworks remain core but require extensions for real-world conditions like nanoparticle facets and electrolyte interfaces.[31]Catalysis
Precious Metal Catalysts
Platinum (Pt) is the benchmark precious metal catalyst for the hydrogen evolution reaction (HER), demonstrating exceptional activity in acidic electrolytes due to its hydrogen adsorption free energy (ΔG_H) near zero, which balances adsorption and desorption kinetics according to Sabatier principle-derived volcano plots. In 0.5 M H₂SO₄, polycrystalline Pt electrodes achieve an overpotential (η) of approximately 30 mV at 10 mA cm⁻² with a Tafel slope of 29–30 mV dec⁻¹, reflecting a mechanism where the Volmer step (H⁺ + e⁻ → H_ads) is fast, and the Heyrovsky step (H_ads + H⁺ + e⁻ → H₂) is rate-limiting under low overpotential.[33][34] Nanostructured Pt, such as atomic layer deposition variants or nanowires alloyed with nickel, enhances mass activity to 10–12 A mg_Pt⁻¹ while maintaining low Tafel slopes, though single-crystal facets like Pt(111) show intrinsic activities up to 10-fold higher than polycrystalline forms due to reduced oxide formation and optimized active sites.[34] Ruthenium (Ru) catalysts exhibit HER performance comparable to or occasionally surpassing Pt, particularly in pH-universal conditions, with η values of 22 mV in acidic media and 17 mV in alkaline at 10 mA cm⁻², and Tafel slopes of ~38 mV dec⁻¹ in base, attributed to modulated d-band center enabling facile water dissociation in alkaline HER.[34] Ru nanostructures, such as Ru@C₂N or RuCu nanosheets on carbon supports synthesized at 250 °C, leverage electronic tuning and support interactions to suppress dissolution and aggregation, yielding stable operation over thousands of cycles.[34] Iridium (Ir) provides high stability against corrosion in harsh electrolytes but typically requires higher overpotentials than Pt (~50–100 mV at 10 mA cm⁻²); however, alloys like IrW nanoparticles or N-doped graphene-supported Ir achieve η lower than Pt in both acid and base via weakened H binding and improved charge transfer.[34][35] Ir-based systems follow similar Volmer-Heyrovsky mechanisms, with Tafel slopes optimized to ~40 mV dec⁻¹ through doping or alloying that shifts the d-band for better ΔG_H alignment.[34] Palladium (Pd) catalysts show moderate HER activity with Tafel slopes of 35–50 mV dec⁻¹ and η ~50 mV at 10 mA cm⁻² in acid, benefiting from strong H adsorption but suffering from slower desorption compared to Pt; Pd nanoparticles confined in carbon nanoreactors or alloyed with Pt reduce this gap, achieving apparent Tafel slopes as low as 10–25 mV dec⁻¹ at low η via site-specific effects.[36][34] Rhodium (Rh) and its alloys, such as Rh-Ir nanoparticles synthesized via microwave methods (sub-10 nm, tunable composition), offer enhanced durability in acidic media with activity improvements from bimetallic synergy that tunes H binding and resists poisoning, though pure Rh typically demands η >100 mV at 10 mA cm⁻² and performs better in alkaline via optimized nanoparticle sizes (~3 nm).[37] Strategies to mitigate the high cost and scarcity of these metals include single-atom dispersion, core-shell architectures, and low-loading supports (e.g., <2 wt% Pt or Ru), which maximize turnover frequencies (up to 10 s⁻¹ for Pt sites) while preserving stability over 10,000 cycles, though dissolution under operational potentials remains a kinetic challenge requiring causal analysis of surface reconstruction.[34] These catalysts set performance benchmarks, with empirical data from rotating disk electrode tests confirming their superiority over non-precious alternatives in turnover and onset potential.[33]Earth-Abundant and Non-Precious Catalysts
Earth-abundant catalysts for the hydrogen evolution reaction (HER) encompass transition metal compounds such as sulfides, phosphides, and hydroxides derived from nickel, cobalt, iron, and molybdenum, offering cost-effective alternatives to platinum by leveraging inexpensive precursors and scalable synthesis methods.[38] These materials achieve viable activity through optimized hydrogen adsorption free energies (ΔG_H ≈ 0 eV) at active sites, often enhanced by nanostructuring to expose edges or defects.[39] Performance metrics include overpotentials (η) at 10 mA/cm² typically ranging from 100–300 mV in acidic or alkaline media, with Tafel slopes of 40–120 mV/dec indicating Volmer-Heyrovsky or Volmer-Tafel mechanisms.[38] Molybdenum disulfide (MoS₂) exemplifies non-precious sulfides, with HER activity localized at basal-plane sulfur vacancies and edge sites that promote H⁺ reduction without strong binding biases.[40] Edge-engineered MoS₂ nanoparticles on graphene supports yield η ≈ 160 mV and a Tafel slope of 41 mV/dec in 0.5 M H₂SO₄, surpassing bulk MoS₂ due to increased edge exposure and conductivity.[41] Doping or heterostructuring, such as Se-co-confined MoS₂ with inner Co layers, further reduces η to below 100 mV at high current densities (>1000 mA/cm²), attributed to modulated electronic structure and suppressed Mo-S bond weakening.[42] Recent variants, like ball-milled MoS₂-graphene composites, maintain stability over 10 hours at 10 mA/cm² with minimal degradation.[43] Nickel-based catalysts, including metallic Ni, Ni(OH)₂ nanosheets, and Ni sulfides, excel in alkaline conditions (e.g., 1 M KOH) via hydroxide-intermediate pathways that lower the energy barrier for water dissociation.[44] Hydrotalcite-derived Ni(OH)₂ on Ni foam achieves η = 120 mV at 10 mA/cm² and sustains 1000 mA/cm² at η < 300 mV for extended operation, owing to in-situ restructuring into active Ni clusters.[45] Dynamic restructuring in NiS electrodes under HER conditions generates Ni-Ni₃S₂ interfaces, yielding Tafel slopes of ~50 mV/dec and turnover frequencies up to 1 s⁻¹ at 100 mV overpotential.[46] Iron, cobalt, and bimetallic phosphides (e.g., CoP, Ni₂P, FeP) provide additional options, with P atoms tuning d-band centers to weaken M-H bonds and enhance proton acceptance.[47] CoP nanowires exhibit η = 50 mV at 10 mA/cm² in acid, with stability exceeding 20 hours, due to surface reconstruction forming phosphide-metal hybrids.[47] Heterostructure designs, such as Ni-Fe phosphosulfides, combine synergies for η < 100 mV and Tafel slopes ~45 mV/dec across pH ranges.[48] Strategies like multi-interface engineering (e.g., core-shell or alloying) and supports (e.g., carbon nanotubes) mitigate limitations such as corrosion in acid or sluggish kinetics in alkali, though long-term durability under industrial currents (>500 mA/cm²) remains inferior to Pt.[49][38] Ongoing efforts focus on single-atom dispersions and high-entropy alloys to approach Pt benchmarks while ensuring elemental abundance.[39]Computational and Theoretical Approaches
Density functional theory (DFT) constitutes the primary computational framework for elucidating the hydrogen evolution reaction (HER), facilitating the determination of adsorption energies, free energy barriers, and reaction pathways on catalyst surfaces. The hydrogen adsorption free energy (ΔG_H*) emerges as the key descriptor, where values approaching 0 eV correlate with minimal overpotentials, as deviations lead to either weak binding (limiting Volmer step) or strong binding (hindering Tafel or Heyrovsky desorption).[50] In a foundational 2005 study, Nørskov et al. applied DFT to compute ΔG_H* for various transition metals, constructing a volcano plot of exchange current density versus ΔG_H*, which positioned Pt near the peak due to its near-thermoneutral H adsorption (ΔG_H* ≈ -0.09 eV), explaining its benchmark activity while highlighting poorer performance for metals like Fe (ΔG_H* ≈ -0.4 eV) or Au (ΔG_H* ≈ 0.2 eV). Theoretical models employ the computational hydrogen electrode (CHE) to construct Pourbaix diagrams and free energy profiles for HER mechanisms, evaluating the Volmer (H⁺ + e⁻ → H*), Heyrovsky (H* + H⁺ + e⁻ → H₂), and Tafel (2H* → H₂) steps under applied potential U via ΔG = ΔE_DFT + ΔZPE - TΔS + eU. Microkinetic simulations, informed by these profiles, predict turnover frequencies, Tafel slopes (e.g., 30 mV/dec for Tafel-limited, 120 mV/dec for Volmer-limited), and pH-dependent shifts, revealing that alkaline HER often faces higher barriers due to water dissociation prerequisites.[50] Scaling relations impose inherent thermodynamic constraints, manifesting as linear correlations (slope ≈ 1) between ΔG_H* and other intermediates like OH* adsorption, preventing simultaneous optimization of all steps and yielding a theoretical overpotential floor of approximately 0.3 V even for ideal descriptors.[50] These relations, derived from d-band center models and frontier orbital overlaps, explain activity trends across metals and alloys but underscore the need for nanostructuring or dopants to decouple bindings via ensemble or ligand effects. DFT limitations, including underestimation of band gaps and neglect of dynamic solvation, are mitigated by hybrid functionals (e.g., HSE06 for better charge transfer), van der Waals corrections (e.g., optB86b-vdW), and solvation models—implicit (e.g., VASPsol, error ~0.2 eV) or explicit via ab initio molecular dynamics (AIMD) to capture hydrogen-bonding networks influencing proton transfer.[51] For non-precious catalysts, such as MoS₂ basal planes or single-atom catalysts (SACs), DFT identifies active sites (e.g., undercoordinated S edges with ΔG_H* ≈ 0.1 eV) and evaluates stability under operational potentials.[50] Machine learning (ML) augments DFT by surrogate modeling large datasets, with graph neural networks (e.g., CGCNN, DimeNet++) trained on thousands of DFT-computed ΔG_H* achieving mean absolute errors of 0.04–0.15 eV, enabling high-throughput screening of over 10⁶ candidates.[52] Active learning loops iteratively refine predictions, as in the discovery of Pt₀.₆₅Ru₀.₃₀Ni₀.₀₅ alloys (overpotential 54 mV at 10 mA/cm²) or Co SACs on N-doped graphene (65% optimal Co-4N-P sites), while Bayesian optimization targets high-entropy alloys outperforming Pt/C (mass activity 3.25 mA/μg).[52] These hybrid DFT-ML pipelines reveal feature importance (e.g., d-band filling via SHAP analysis) and guide synthesis, though validation against experiment remains essential due to ensemble effects absent in periodic slab models.[52]Applications
In Water Electrolysis Systems
The hydrogen evolution reaction (HER) serves as the primary cathodic process in water electrolysis systems, facilitating the production of molecular hydrogen from water under applied electrical potential. In these systems, electrolysis decomposes water into hydrogen at the cathode via HER and oxygen at the anode via the oxygen evolution reaction (OER), with the overall reaction being 2H₂O → 2H₂ + O₂ at standard conditions requiring a thermodynamic minimum voltage of 1.23 V. However, practical cell voltages exceed 1.8 V due to kinetic overpotentials, particularly for HER in non-acidic environments, ohmic losses, and mass transport limitations.[53][2] In alkaline water electrolysis (AWE), which dominates industrial-scale hydrogen production with systems operating at 60–80°C in 20–40% KOH electrolytes, HER proceeds via 2H₂O + 2e⁻ → H₂ + 2OH⁻, exhibiting slower kinetics compared to acidic conditions because of the initial water dissociation step preceding hydrogen adsorption. Nickel-based cathodes, such as Raney nickel or Ni foams, are commonly employed for HER in AWE due to their cost-effectiveness and stability in alkaline media, achieving current densities up to 0.5 A/cm² at overpotentials around 200–300 mV, though efficiency drops at higher densities needed for gigawatt-scale deployment.[54][55] Recent advancements have pushed AWE HER performance to over 1 A/cm² using optimized Ni alloys, but persistent challenges include bubble-induced resistance and catalyst degradation from OH⁻ adsorption.[56] Proton exchange membrane (PEM) electrolyzers, operating in acidic environments at 50–80°C with perfluorosulfonic acid membranes, enable HER through 2H⁺ + 2e⁻ → H₂, benefiting from faster proton discharge and lower overpotentials (typically 30–50 mV at 1 A/cm² with Pt/C catalysts). These systems achieve higher current densities (up to 2 A/cm²) and purity (>99.99% H₂), making them suitable for dynamic renewable energy integration, but reliance on scarce platinum-group metals limits scalability, with HER catalyst loading often exceeding 0.5 mg/cm² Pt.[53][57] Anion exchange membrane (AEM) electrolyzers bridge alkaline and PEM advantages by using hydroxide-conducting membranes, allowing non-precious HER catalysts like Ni or Fe-based materials in near-neutral to alkaline conditions, with reported cell efficiencies approaching 60% at 1 A/cm² as of 2023. However, AEM systems face HER challenges from membrane degradation and lower ionic conductivity, restricting commercial viability despite potential cost reductions below $500/kW.[58][59] Across these systems, HER efficiency directly impacts overall hydrogen yield, with industrial targets aiming for <50 mV overpotential at 1 A/cm² to enable stack efficiencies >70% when coupled with renewable electricity. Ongoing efforts focus on mitigating HER-related losses through nanostructured electrodes and electrolyte optimization to support terawatt-scale green hydrogen production by 2050.[60][61]In Photoelectrochemical and Photocatalytic Processes
In photoelectrochemical (PEC) water splitting, the hydrogen evolution reaction (HER) occurs at the photocathode, where photogenerated electrons from a p-type semiconductor reduce protons to form H₂ gas, contributing to the overall half-reaction 2H⁺ + 2e⁻ → H₂.[62] This process relies on efficient charge separation and transfer, with cocatalysts often deposited on the semiconductor surface to minimize overpotential and enhance kinetics by facilitating Volmer-Heyrovsky or Volmer-Tafel mechanisms.[62] Typical onset potentials for HER in PEC systems range from 0.2 V to 0.53 V vs. reversible hydrogen electrode (RHE), influenced by semiconductor band alignment and catalyst loading.[62] Earth-abundant cocatalysts have shown promise in reducing HER overpotentials without precious metals. For instance, molybdenum disulfide (MoS₂) cocatalysts on silicon-based photocathodes achieved photocurrent densities of -15.2 mA cm⁻² at 0 V vs. RHE in 2022, attributed to edge-site active centers improving electron transfer.[62] Cobalt phosphide (Co-P) on inverted pyramid silicon yielded -35.2 mA cm⁻² at 0 V vs. RHE with 150 hours of stability in 2019, demonstrating resilience against photocorrosion.[62] Nickel diphosphide (NiP₂) delivered -12 mA cm⁻² at 0 V vs. RHE for 6 hours in 2016, highlighting phosphide structures' role in stabilizing interfaces.[62] Solar-to-hydrogen (STH) efficiencies in such PEC configurations have reached up to 11.07% using MoS₂-modified titanium foil in 2022, though charge recombination and material degradation remain barriers to higher yields.[62][62] In photocatalytic processes, HER proceeds via light-excited electron-hole pairs in suspended semiconductor particles, with electrons migrating to surface cocatalysts or defect sites to drive proton reduction, while holes oxidize water or scavengers. Common catalysts include TiO₂ loaded with Pt nanoparticles, CdS/ZnO composites, and Ni-modified sulfides, where Pt enhances HER rates by lowering activation barriers but increases costs. Apparent quantum efficiencies for visible-light-driven HER typically fall below 1% under solar conditions, limited by the narrow UV portion (4%) of the spectrum and rapid charge recombination. Pure photocatalytic overall water splitting (without sacrificial agents) faces severe challenges, including reverse H₂-O₂ recombination and low intrinsic activities, yielding efficiencies often orders of magnitude below electrocatalytic benchmarks. Systems relying on sacrificial electron donors (e.g., alcohols) achieve higher H₂ evolution rates but do not represent sustainable closed-cycle operation. Stability issues, such as photocorrosion in CdS-based materials, and scalability constraints in reactor designs further hinder practical deployment, with recent nanostructured semiconductors showing incremental improvements in charge separation but no breakthrough in STH exceeding 1% for pure systems as of 2024.As a Competing Reaction in Other Electrochemical Systems
In electrochemical CO2 reduction (CO2RR), the hydrogen evolution reaction (HER) serves as a primary competing cathodic process, particularly on metal electrodes where the reduction potentials for CO2-derived products and H+ or H2O overlap, leading to faradaic efficiencies for CO2 products often below 50% at industrially relevant currents due to preferential HER kinetics.[63] For instance, on copper electrodes in acidic media (pH 2.5), water reduction dominates over proton reduction as the HER pathway, exacerbating competition even under buffered conditions.[64] Mass transport limitations further modulate this rivalry, with local pH gradients near the electrode surface enhancing HER rates and suppressing CO2RR selectivity, as demonstrated in rotating ring-disk electrode studies where HER contributions exceed 70% of total current at potentials below -0.5 V vs. RHE.[65] The nitrogen reduction reaction (NRR) for ammonia synthesis faces analogous challenges, with HER outcompeting N2 reduction due to the latter's higher activation barriers and the thermodynamic favorability of H2 formation in aqueous electrolytes, resulting in NRR faradaic efficiencies typically under 10-20% and ammonia yield rates limited to micrograms per hour per milligram catalyst.[66] Microkinetic modeling reveals potential-dependent shifts, where at overpotentials greater than -0.3 V vs. RHE, HER dominates via Volmer-Heyrovsky mechanisms on sites that could otherwise bind N2, with coadsorption of NRR intermediates (e.g., NH2) and HER species (H) dictating selectivity on single active sites.[67][68] In metal electrowinning processes, such as zinc or gallium recovery from sulfate electrolytes, HER competes directly with metal ion reduction, reducing current efficiencies to 80-90% or lower by evolving H2 gas, which also causes electrode polarization and energy losses exceeding 20% of input power.[69] For gallium electrowinning from zinc hydrometallurgy leachates, cathodic HER predominates alongside Ga3+ deposition at potentials around -1.2 V vs. SHE, necessitating additives like glue or animal glue derivatives to inhibit H2 evolution without compromising metal deposition rates.[70] Organic electrosynthesis, including reductions of carbonyls or halides, encounters HER as an unwanted solvent-based side reaction that diminishes product yields, particularly in protic media where electrode microenvironments favor H+ discharge over organic acceptor reduction, with selectivity tunable via kinetic control of local pH and overpotential.[71] In near-critical water systems, HER pathways compete with desired organic conversions like adiponitrile formation, where suppressing H2 evolution requires balancing electrode potentials to prioritize substrate adsorption over proton reduction.[72]Challenges
Overpotential and Activity Limitations
The overpotential (η) in the hydrogen evolution reaction (HER) represents the additional voltage required beyond the reversible hydrogen electrode (RHE) potential (0 V under standard conditions) to achieve a desired current density, arising primarily from kinetic barriers that impede proton reduction and hydrogen desorption despite thermodynamically favorable ΔG ≈ 0 eV for H₂ formation.[73] These barriers stem from the multi-step nature of HER, involving the Volmer step (H⁺ + e⁻ → H_{ads} in acidic media, or H₂O + e⁻ → H_{ads} + OH⁻ in alkaline), followed by either the Heyrovsky (H_{ads} + H⁺ + e⁻ → H₂) or Tafel (2 H_{ads} → H₂) step, where the rate-determining step (RDS) dictates the Tafel slope (b): b ≈ 120 mV/dec for Volmer RDS (high H_{ads} binding energy barrier), b ≈ 40 mV/dec for Tafel/Heyrovsky RDS (desorption-limited), and intermediate values for mixed control.[74] [75] Platinum (Pt) serves as the benchmark catalyst, exhibiting low η due to optimal H_{ads} binding energy (ΔG_H ≈ 0 eV) per the Sabatier principle, with typical η ≈ 30–50 mV at 10 mA/cm² in acidic electrolytes, enabling high turnover frequencies (TOF > 1 s⁻¹ at low η).[33] [38] However, activity limitations persist even for Pt, including mass transport constraints at higher current densities (>100 mA/cm²) and sensitivity to impurities like CO, which poison active sites by competitive adsorption.[76] In alkaline media, HER overpotentials increase by 100–200 mV compared to acidic conditions for the same catalysts, attributed to the kinetically slower water dissociation step preceding Volmer, requiring additional energy input (≈0.4–0.6 eV barrier) and leading to higher RDS activation energies.[55] [77] For earth-abundant catalysts (e.g., Ni, MoS₂, or transition metal phosphides), overpotentials are markedly higher—often η > 200 mV at 10 mA/cm²—due to suboptimal ΔG_H (either too exergonic or endergonic), resulting in volcano plot deviations from Pt's peak activity; for instance, Ni exhibits η ≈ 250 mV in acid, limited by strong H_{ads} binding that hinders desorption.[73] [38] These limitations are exacerbated by surface reconstruction under operational potentials, where active sites (e.g., edge defects in MoS₂) degrade or passivate, reducing intrinsic activity (measured as exchange current density j₀, typically 10⁻³–10⁻⁶ A/cm² for non-Pt vs. 10⁻³ A/cm² for Pt).[9] Reporting inconsistencies, such as unaccounted iR drops or non-standard current densities, further obscure true activity comparisons, emphasizing the need for standardized metrics like η at 10 mA/cm² (solar-relevant) or 1 A/cm² (industrial).[78]| Catalyst | Electrolyte | η at 10 mA/cm² (mV) | Tafel Slope (mV/dec) |
|---|---|---|---|
| Pt/C | Acidic | 30–50 | 30–40 |
| Pt/C | Alkaline | 50–100 | 40–60 |
| Ni | Acidic | ≈250 | 100–120 |
| MoS₂ | Acidic | 150–200 | 50–80 |
Stability and Durability Issues
Stability and durability of hydrogen evolution reaction (HER) electrocatalysts represent critical barriers to practical implementation, particularly for non-precious metal variants, where degradation often limits operational lifetimes to mere hours under industrially relevant conditions. Primary degradation mechanisms include metal dissolution, Ostwald ripening, agglomeration, phase reconstruction, and active site poisoning by impurities or reaction intermediates, exacerbated by harsh electrochemical environments such as acidic or alkaline electrolytes.[79][80] In acidic media, transition metal phosphides like Ni₂P and CoP exhibit rapid elemental leaching, with up to 50% Ni loss observed over 16 hours of operation due to thermodynamic instability at open-circuit potentials, leading to surface oxidation and reduced active sites.[81] Similarly, sulfides such as MoS₂ undergo dissolution and phase reversion from metastable 1T to stable 2H structures, diminishing catalytic activity.[38] Precious metal catalysts, exemplified by platinum, demonstrate superior durability—maintaining performance beyond 100 hours in acidic conditions owing to resistance to leaching and corrosion—but are not immune to issues like particle agglomeration and carbon support degradation under prolonged high-current operation.[81][79] In alkaline electrolytes, non-precious catalysts face phosphorus leaching in phosphides (e.g., Ni₂P activity halving within 48 hours) and selective dissolution of components like molybdenum in NiMo alloys above -150 mV versus reversible hydrogen electrode.[38] These processes are intensified by factors including high current densities (>10 mA cm⁻²), elevated temperatures, and electrolyte impurities (e.g., Cl⁻ or SO₄²⁻ ions), which promote local corrosion and bubble-induced mechanical stress.[79][80] Testing protocols further complicate durability assessments, as uncontrolled immersion into electrolytes induces preemptive oxidation or dealloying in non-noble metals, altering catalyst morphology before HER initiation and yielding irreproducible results.[80] Ex situ characterizations post-operation often fail to capture dynamic degradation, such as redeposition at cathodic potentials, underscoring the need for operando techniques to reveal true long-term viability. Overall, while precious metals offer benchmarks with minimal degradation, earth-abundant alternatives' susceptibility to compositional and structural instability hinders scalability for sustained hydrogen production.[38][81]Scalability and Economic Constraints
The scalability of catalysts for the hydrogen evolution reaction (HER) to industrial electrolyzer stacks is limited by performance gaps between bench-scale testing and operational demands. Laboratory assessments typically evaluate activity at low current densities (e.g., 10 mA/cm² under 1 atm H₂), which do not replicate the high-current densities (>1 A/cm²) and dynamic conditions required for commercial production rates exceeding 100 Nm³/h per stack. Under such regimes, many nanostructured or molecular catalysts degrade rapidly due to mechanical instability, bubble-induced detachment, or active site reconfiguration, with electrode lifetimes often falling short of the 40,000–80,000 hours needed for economic feasibility.[82] [83] Economic barriers stem primarily from material costs and energy inefficiencies. Platinum (Pt), the standard HER catalyst in proton exchange membrane (PEM) electrolyzers, incurs high capital expenditures due to its scarcity and price, averaging $1,638 per troy ounce (approximately $52,000/kg) in 2024, with typical loadings of 0.3–0.5 mg Pt/cm² contributing 5–15% to total stack costs in systems targeting gigawatt-scale deployment. Earth-abundant alternatives, such as Ni- or Mo-based chalcogenides, reduce material expenses to <$10/g but demand higher overpotentials (often >100 mV at 1 A/cm²), elevating the cell voltage beyond the thermoneutral potential of 1.48 V and increasing electricity consumption—which accounts for 45–70% of levelized hydrogen costs (LCOH)—by 5–10% per 50 mV increment.[84] [85] [86] Synthesis and manufacturing scalability further constrain viability, as lab-scale methods like electrodeposition or chemical vapor deposition yield non-uniform coatings unsuitable for large-area electrodes (>1 m²), leading to yield losses >50% during upscaling and inconsistent performance. Techno-economic models project that HER innovations must achieve <30 mV overpotential at 2 A/cm² with >90% capacity retention over 50,000 cycles to drive LCOH below $2/kg H₂ using renewable electricity at $20/MWh, a threshold unattained in 2024 demonstrations due to trade-offs between activity and durability. Membrane and balance-of-plant costs compound these issues, with PEM systems' reliance on perfluorosulfonic acids adding $200–500/kW in acidic environments favoring Pt.[87] [88] Recent efforts in continuous-flow synthesis for non-precious catalysts show promise for cost parity with Pt at <$1/g, but validation at pilot scales (>10 MW) remains pending as of 2025.[86]Recent Advances
Nanostructured and Single-Atom Catalysts
Nanostructured catalysts enhance the hydrogen evolution reaction (HER) by increasing active surface area and exposing high-energy sites, such as edges and defects, which facilitate proton adsorption and hydrogen desorption with optimal free energy (ΔG_H* ≈ 0 eV).[89] Transition metal dichalcogenides like MoS₂ nanosheets exemplify this, where edge sites dominate activity over inert basal planes, achieving overpotentials as low as 155 mV at 10 mA cm⁻² in acidic media with Tafel slopes around 40-60 mV dec⁻¹.[90] Non-precious alternatives, such as Ni-based nanostructures including phosphides and oxides, have demonstrated overpotentials of 169 mV at 10 mA cm⁻² and Tafel slopes of 56 mV dec⁻¹, outperforming bulk counterparts due to synergistic electronic effects and strain at interfaces.[91] Recent designs incorporate hybrid architectures, such as Pt-NiFe oxide nanostructures, which reduce Pt loading while maintaining low overpotentials through optimized particle sizes (e.g., <5 nm) and support interactions that prevent agglomeration.[92] A 2022 Al-based catalyst featuring ~2 nm medium-entropy nanocrystals embedded in amorphous regions exhibited robust HER activity in alkaline conditions, with stability over thousands of cycles attributed to the amorphous-crystalline interface resisting dissolution.[93] Alloyed nanostructures like Pt₃Co on nitrogen-doped graphene further lower overpotentials to 13 mV at 10 mA cm⁻², though scalability is limited by synthesis complexity and precious metal content.[94] Single-atom catalysts (SACs) for HER achieve near-100% metal utilization by anchoring isolated atoms on supports like N-doped carbon or graphene, tuning coordination environments to modulate d-band centers and hydrogen binding energies via ligand effects.[95] Non-noble SACs, such as Co or Ni atoms in pyridinic N sites, rival Pt benchmarks with overpotentials below 100 mV at 10 mA cm⁻² in alkaline media, as seen in Co-P₂N₂-C configurations that stabilize *H intermediates through back-donation.[96] These SACs often exhibit Tafel slopes of 30-50 mV dec⁻¹, indicating Volmer-Heyrovsky mechanisms, with activity enhanced by microenvironment modulation that prevents clustering under operational potentials.[97] Stability in SACs hinges on strong metal-support bonds, with recent advances like axial ligand tailoring boosting durability to over 10,000 cycles by suppressing atom migration and dissolution, though high-loading variants (>5 wt%) risk aggregation without precise anchoring.[98] Computational screening predicts SAC stability via Pourbaix diagrams, revealing that pH-dependent dissolution limits many transition metal SACs in acidic HER, favoring alkaline operation for earth-abundant variants.[99] While lab-scale metrics are promising, industrial viability requires validation beyond low-current densities (<100 mA cm⁻²), as support corrosion and active site reconstruction often degrade performance over prolonged exposure.[100]High-Current Density and Industrial-Relevant Designs
To enable economically viable large-scale hydrogen production via water electrolysis, hydrogen evolution reaction (HER) electrocatalysts must sustain current densities exceeding 1 A/cm² (1000 mA/cm²) while maintaining low overpotentials (typically <300 mV) and operational stability beyond 1000 hours, as lower densities increase electrode footprint and system costs.[101] Current industrial alkaline electrolyzers operate at 200–500 mA/cm², but U.S. Department of Energy targets aim for 1600 mA/cm² by 2040 to achieve cost parity with fossil-based hydrogen.[101] At high current densities, challenges intensify, including accelerated bubble formation that impedes mass transport and active site coverage, necessitating designs that decouple intrinsic activity from transport limitations through enhanced conductivity, porosity, and superaerophobic surfaces.[102] Key strategies for industrial-relevant HER designs emphasize non-precious metal catalysts with earth-abundant elements like Ni, Mo, Co, and P or S, integrated into three-dimensional architectures such as foams, nanowires, or arrays to boost site density and electron/mass transfer.[102] Self-supporting electrodes on conductive substrates (e.g., Ni foam) eliminate binders, enhancing mechanical integrity and scalability via electrodeposition or hydrothermal synthesis amenable to meter-scale production.[102] Polymetallic alloys optimize adsorption free energies near zero for hydrogen intermediates, while defect engineering and heterostructuring (e.g., Ni₂P/Mo₂C interfaces) facilitate faster Volmer-Heyrovsky kinetics under alkaline conditions prevalent in low-cost electrolyzers.[101] These approaches prioritize causal factors like local electronic modulation over empirical screening, yielding catalysts that resist deactivation from H₂ bubble adhesion.[101] Notable examples demonstrate feasibility in half-cell configurations, often in 1 M KOH, though full-cell validation remains sparse due to anode limitations and ohmic drops.[102]| Catalyst | Material/Design | Current Density (mA/cm²) | Overpotential (mV) | Stability | Conditions | Citation |
|---|---|---|---|---|---|---|
| h-NiMoFe on Ni foam | Hierarchical alloy array, scalable synthesis | 1000 | 98 | Not specified | Alkaline half-cell | [102] |
| Ni₂P nanowire arrays | Self-supported phosphide | 1000 | 306 | Not specified | Alkaline half-cell | [101] |
| MoS₂ on Cu foam | Nanosheet-coated 3D foam | 1000 | 519 | Not specified | Alkaline half-cell | [101] |
| F-Co₂P/Fe₂P interfaced | Fluorinated phosphide heterostructure | 1000 (up to 3000) | 260.5 | Not specified | Alkaline half-cell | [102] |
| Int-Ni/MoO₂ | Interlayer-bonded blades | 1000 | Not specified | 100 h | Alkaline, high-rate | [82] |