Photocatalytic water splitting offers a route to store solar energy in chemical form by producing hydrogen from water. For engineers and technical decision-makers, its potential depends on more than choosing a material that absorbs light: the semiconductor must generate useful charge carriers, keep them from recombining, and support the surface reactions that form hydrogen and oxygen. These linked requirements shape the design and evaluation of a photocatalytic water splitting system.
How photocatalytic water splitting works
Water splitting is an energetically uphill reaction. In a semiconductor photocatalyst, light with sufficient energy excites electrons from the valence band to the conduction band, leaving positively charged holes behind. The electrons and holes must then move to suitable reaction sites. Electrons drive proton reduction to form hydrogen, while holes drive water oxidation to form oxygen.
The process can be understood as three connected stages:
- Light harvesting: the photocatalyst absorbs photons and generates electrons and holes.
- Charge separation and transport: the carriers move through the material to its surface rather than recombining.
- Surface reactions: reaction sites use electrons and holes to carry out hydrogen production and water oxidation.
Each stage constrains the overall result. A material that absorbs a broad range of sunlight may still perform poorly if its charge carriers recombine quickly or cannot reach active surface sites. Conversely, improving surface reactions can help consume carriers before they recombine. These processes are therefore interdependent, not separate design problems.
Band structure sets the operating window
A semiconductor’s band gap determines which photons it can absorb, while the positions of its conduction and valence bands determine whether its electrons and holes have the energetic driving force needed for the reduction and oxidation reactions. The minimum thermodynamic energy requirement for water splitting is 1.23 eV; practical systems also need to account for reaction overpotentials, particularly for water oxidation.
This creates a materials-design trade-off. A wider band gap can provide energetic headroom for the reactions but limits absorption to higher-energy light. Many established metal-oxide photocatalysts respond primarily to ultraviolet light. Because visible light represents a larger portion of sunlight, researchers have explored ways to extend absorption, including modifying valence or conduction bands and forming solid solutions.
Introducing elements such as nitrogen or sulfur can change the valence-band structure, while changing the composition of a semiconductor can tune its conduction band. Oxynitrides and oxysulfides are examples of materials designed to absorb visible light. However, an absorption shift alone does not guarantee better performance: dopants or other structural changes can create charge-recombination centers and reduce photocatalytic activity.
Charge separation is central to performance
Photogenerated electrons and holes compete between reaching reaction sites and recombining. Defects in a material can provide locations where recombination occurs, while the distance carriers must travel also affects the likelihood that they reach the surface in time. High crystallinity can reduce defect-related recombination, and smaller particles can shorten carrier diffusion paths.
Several structural approaches aim to direct electrons and holes along different pathways. A p–n junction forms an interface between p-type and n-type semiconductors. The resulting built-in electric field can promote movement of electrons and holes in opposite directions. Heterojunctions similarly bring together semiconductors with different band structures or Fermi levels to influence charge transfer.
Charge separation can also occur between different crystalline phases of the same material, forming a phase junction. In titanium dioxide, for example, anatase–rutile combinations have been associated with improved charge separation compared with individual phases. Research on bismuth vanadate has also shown spatial separation between crystal facets, with reduction- and oxidation-related deposits appearing preferentially on different surfaces. These approaches illustrate that both interfaces and the internal structure of a particle can influence where charges accumulate.
Cocatalysts support surface reactions
Even when charges reach the photocatalyst surface, the chemical reactions must proceed at a sufficient rate. Cocatalysts provide reaction sites for proton reduction and water oxidation. Their role is not limited to catalysis: the interface between a cocatalyst and semiconductor can also affect charge separation, helping direct carriers toward the reactions that consume them.
Because hydrogen production and oxygen evolution are distinct half reactions, the catalyst system must support both sides of the overall process. A material that produces hydrogen in the presence of a sacrificial reagent has demonstrated a reduction half reaction, but this does not by itself establish overall water splitting. Sacrificial reagents consume electrons or holes to make one half reaction easier to study. In an overall water-splitting test without such reagents, hydrogen and oxygen should be produced together in a 2:1 stoichiometric ratio.
Measure performance with the right metric
Photocatalytic results depend on experimental conditions, including light intensity and reaction temperature, so activity values are not automatically comparable. Apparent quantum efficiency (AQE) relates reacted electrons to incident photons at a specified wavelength. Internal quantum efficiency (IQE) instead uses absorbed photons, but determining absorption in a dispersed powder system can be difficult because of light scattering and losses.
Solar-to-hydrogen (STH) efficiency provides a system-level measure by comparing the energy stored in produced hydrogen with the incident solar energy. When assessing results, technical teams should distinguish between half-reaction activity and overall water splitting, and check whether reported efficiency refers to incident or absorbed light. These distinctions clarify what a material or system has actually demonstrated.
Practical implications for system development
Particulate photocatalyst systems, in which powders are dispersed in water and illuminated, offer a comparatively simple equipment concept. Other approaches include Z-scheme photocatalyst systems and hybrid natural–artificial photosynthesis systems. Across these configurations, the same core questions remain: does the material absorb useful sunlight, can it separate and transport charges, and can its surface sites carry out both reactions?
For industrial automation and process teams evaluating solar hydrogen pathways, the practical lesson is to assess the complete conversion chain rather than a single material property. Band engineering, junction design and cocatalyst selection must work together, while testing should clearly identify the reaction conditions and efficiency metric. Progress in photocatalytic water splitting depends on integrating light capture, charge management and surface chemistry into a system that produces hydrogen and oxygen reliably as an overall reaction.