Water is torn apart. During photosynthesis, water molecules are split into three components: oxygen gas, which is released into the atmosphere; electrons, which flow through the cell’s energy-producing machinery; and protons (hydrogen ions), which accumulate and help drive the production of energy-storage molecules. This splitting happens inside a protein complex called Photosystem II, powered by sunlight, and it is one of the most chemically demanding reactions in all of biology. The process is also the source of virtually all the oxygen you breathe.
Where Water Splitting Happens
Water splitting takes place in chloroplasts, the small green compartments inside plant cells. Within each chloroplast are stacks of internal membranes called thylakoids, and Photosystem II sits embedded in these membranes. The actual site where water gets broken apart is a tiny cluster of metal atoms: four manganese atoms and one calcium atom, bridged by oxygen atoms. Researchers often refer to this as the oxygen-evolving complex or the water-oxidizing complex. A high-resolution crystal structure published in 2011 finally revealed its atomic arrangement, and that structure has guided much of the research since.
The cluster works like a catalyst, cycling through a series of charge states as it strips electrons from water. It is remarkable that biology settled on manganese for this job. Manganese can exist in multiple oxidation states, which lets the cluster accumulate the electrical charge needed to rip apart one of the most stable small molecules in nature. The protein surrounding the cluster fine-tunes its chemistry, keeping the manganese atoms at precisely the right energy levels to do their work.
How Light Drives the Splitting
Sunlight does not split water directly. Instead, light energy is absorbed by chlorophyll molecules and funneled to a special chlorophyll at the heart of Photosystem II, called P680. When P680 absorbs enough energy, it ejects an electron within a few trillionths of a second to a nearby acceptor molecule. This leaves P680 in a highly oxidized state, making it one of the strongest biological oxidants known. That oxidized P680 pulls an electron from a nearby tyrosine amino acid, and the tyrosine in turn pulls an electron from the manganese-calcium cluster, which in turn pulls an electron from water.
This chain of electron transfers, each one tugged along by the next, is what ultimately forces water to give up its electrons. The process does not happen all at once. Splitting two water molecules requires the removal of four electrons total, and Photosystem II accomplishes this by absorbing four photons of light, one at a time, cycling the manganese cluster through progressively higher charge states. Only after the fourth photon’s energy has been harvested does the cluster have enough oxidizing power to break apart two water molecules simultaneously, releasing one molecule of oxygen gas.
The Kok Cycle
The stepwise accumulation of charge in the manganese cluster follows a pattern known as the Kok cycle (sometimes called the S-state cycle), named after the scientist who first described it in the 1970s. The cluster passes through five states, labeled S0 through S4. Each absorption of a photon advances the cluster by one state. At S4, the cluster has stored enough oxidizing equivalents to catalyze the formation of an oxygen-oxygen bond from two water molecules, releasing Oâ‚‚ and resetting the cycle back to S0.
Recent work using ultrafast X-ray crystallography has captured snapshots of this cycle as it happens. These experiments revealed that during the final, oxygen-releasing step, the oxygen-oxygen bond forms rapidly, within about 50 millionths of a second, and that no further oxidation of the manganese atoms beyond their highest stable state was detected before the bond formed. That finding has reshaped ideas about the exact mechanism of the last step, suggesting the bond forms before the final electron transfer to the tyrosine residue is complete.
Where the Pieces of Water End Up
Once water is split, its three products each follow a different path. The oxygen atoms pair up and leave the leaf as Oâ‚‚ gas, exiting through tiny pores on the leaf surface called stomata. This is the oxygen that sustains aerobic life on the planet.
The electrons stripped from water enter an electron transport chain embedded in the thylakoid membrane. They first reduce a molecule called plastoquinone, converting it to plastoquinol, which shuttles the electrons to the next protein complex in the chain. Eventually the electrons reach Photosystem I, where a second round of light absorption boosts their energy again, and they are finally used to produce NADPH, a molecule that carries chemical energy to the sugar-building reactions of the Calvin cycle.
The protons released from water accumulate inside the thylakoid’s interior space, called the lumen. This compartment was once thought of as little more than an empty bag for collecting protons, but it is now recognized as a functionally important space that influences how the whole photosynthetic apparatus operates. The buildup of protons in the lumen creates a concentration gradient across the thylakoid membrane, and this gradient powers an enzyme called ATP synthase, which generates ATP. So the hydrogen atoms from water ultimately drive the production of both major energy currencies, NADPH and ATP, that the plant uses to fix carbon dioxide into sugar.
Why Splitting Water Is So Difficult
Water is an extremely stable molecule. Breaking its bonds requires a substantial input of energy, and doing so in a controlled, four-electron process without producing dangerous intermediate molecules is a serious chemical challenge. Free radicals and reactive oxygen species can destroy cellular components, so the reaction has to be tightly managed. The manganese-calcium cluster and its surrounding protein scaffold accomplish this by holding the reaction intermediates in place and carefully timing each electron-removal step so that partially oxidized water species do not escape and wreak havoc.
The energy needed to split water comes entirely from visible light. A standard photon absorbed by P680 has a wavelength of about 680 nanometers. Recent research on certain cyanobacteria that live in shaded environments has shown that water oxidation can occur using far-red light with wavelengths up to about 727 nanometers. These organisms use a modified chlorophyll called chlorophyll d in place of the usual chlorophyll a at the critical electron-donor position. A photon at 727 nanometers carries about 120 millielectronvolts less energy than one at 680 nanometers, which appears to be close to the minimum energy needed to drive the reaction. This finding has raised questions about just how little energy is fundamentally required for biological water oxidation.
The Self-Destructive Nature of Water Splitting
The very chemistry that makes Photosystem II so powerful also makes it fragile. The highly oxidizing conditions needed to split water inevitably cause damage to the protein complex itself, particularly to a core protein called D1. The strongly oxidizing radicals generated during normal operation, along with reactive oxygen species produced as occasional byproducts, chew up the D1 protein over time. Under normal light, this damage happens at a manageable rate, but it accelerates significantly under high light intensity or when combined with other environmental stresses like heat or drought.
Plants have evolved an elaborate repair system to deal with this. The damaged Photosystem II complex is partially disassembled, the broken D1 protein is chopped up by enzymes and removed, a freshly made D1 is inserted, and the complex is reassembled and returned to active duty. This repair cycle involves dozens of assisting proteins and shares many steps with the process of building new Photosystem II complexes from scratch. It is an ongoing, energy-costly maintenance operation that every photosynthetic organism must carry out constantly during daylight hours.
Under very high light, the rate of damage can outpace the rate of repair, leading to a condition called photoinhibition. This is why a shade-loving plant placed in direct midday sun can suffer visible bleaching and reduced growth: its Photosystem II complexes are being destroyed faster than they can be rebuilt.
What Happens When Water Supply Runs Low
Since water is literally a raw material consumed by the light reactions, drought directly threatens the photosynthetic machinery. When soil moisture drops, plants close their stomata to conserve water. This reduces water loss through evaporation but also cuts off the inflow of carbon dioxide needed for sugar production. The Calvin cycle slows, which means the energy products of the light reactions (NADPH and ATP) start to back up with nowhere to go. The electron transport chain becomes over-reduced, and excess energy spills over into the production of reactive oxygen species that damage chloroplasts, membranes, and proteins.
Drought also directly impairs the light reactions themselves. The photosystems suffer structural damage, and the electron transport chain is partially blocked. The combination of COâ‚‚ starvation and oxidative stress creates a feedback loop: less carbon gets fixed, more damaging byproducts accumulate, and the plant’s photosynthetic capacity declines further. For major food crops, this cascade is one of the primary reasons drought so severely reduces yield.
Before Water Splitting Existed
Photosynthesis is older than water splitting. The earliest photosynthetic organisms, which still exist today as various groups of bacteria, used molecules other than water as their electron source. Green sulfur bacteria, for example, use hydrogen sulfide. These organisms carry out what is called anoxygenic photosynthesis because they do not produce oxygen. Their photosynthetic machinery is simpler and does not need to generate the extreme oxidizing power required to break apart water.
The evolution of water splitting was a pivotal event in the history of life. Phylogenetic studies of the D1 protein, the core subunit of Photosystem II that coordinates the manganese cluster, suggest that several ancestral forms of D1 existed before water oxidation evolved. These early forms lacked some of the amino acid residues needed to bind the manganese-calcium cluster. At some point, the right set of mutations appeared, the cluster could be assembled, and water oxidation became possible. The oxygen this process released gradually transformed Earth’s atmosphere from an oxygen-free environment to one where oxygen-breathing organisms could thrive.
Artificial Photosynthesis and Why Scientists Want to Copy Water Splitting
The ability to split water using sunlight is essentially the ability to produce clean fuel. If you can break water into oxygen and hydrogen (or use the electrons and protons for other chemical reactions), you have a solar-powered energy system that takes in only water and light and produces no carbon emissions. This is the core idea behind artificial photosynthesis, a field that has grown enormously in recent decades.
Researchers have paid particular attention to manganese-based catalysts, since nature itself chose manganese for the job. Various synthetic manganese oxide materials have been developed and tested as functional models of the biological water-oxidizing complex. Nano-sized manganese oxides have been explored as catalysts for water oxidation because manganese is cheap, abundant, and environmentally benign compared to the precious metals (like iridium and ruthenium) that dominate current industrial water-splitting catalysts.
The biological system also inspires the overall design strategy, not just the catalyst material. In plants, the light-harvesting step, the charge-separation step, and the catalytic water-splitting step are physically separated and optimized independently. Artificial systems aim to replicate this modular architecture. Some semi-artificial approaches go a step further, directly wiring natural photosynthetic proteins to synthetic electrodes or coupling algal photosystems with hydrogen-producing enzymes called hydrogenases. Proof-of-concept systems have been demonstrated in which Photosystem I is fused to a hydrogenase in living cyanobacteria, so the electrons extracted from water by Photosystem II flow through Photosystem I and are immediately used to produce hydrogen gas.
None of these systems yet match the efficiency, stability, or scalability needed for commercial deployment. The biological water-oxidizing complex, despite its self-destructive tendencies, operates under mild conditions (room temperature, neutral pH, ordinary water), repairs itself, and has been running reliably for over two billion years. Matching all of those features simultaneously in a synthetic device remains one of the grand challenges of energy research.
Far-Red Photosynthesis and the Energy Floor
Most textbook descriptions of photosynthesis focus on chlorophyll a, which absorbs red light at around 680 nanometers. But certain cyanobacteria that live in low-light environments, such as underneath other photosynthetic organisms or inside rocks, have adapted to use far-red light that penetrates where shorter wavelengths cannot. When these organisms acclimate to far-red conditions, they swap out some of their standard chlorophyll for chlorophyll d and chlorophyll f, which absorb at longer wavelengths.
Structural studies of Photosystem II from one such organism, Chroococcidiopsis thermalis, showed that chlorophyll d sits in the critical electron-donor position (ChlD1) at the heart of the reaction center. The primary donor bleaching wavelength in this system was measured at 727 nanometers, meaning the complex is driving water oxidation with photons carrying substantially less energy than the 680-nanometer photons used by conventional photosynthesis. The amino acid residues that hold chlorophyll d in place at this site are well conserved across far-red-adapted cyanobacteria, suggesting this is a widespread adaptation rather than a quirk of one species.
This raises a genuine puzzle: how close is 727 nanometers to the absolute minimum energy needed to split water? The thermodynamic requirement for water oxidation is fixed by physics, but the biological system needs extra energy beyond the bare minimum to drive the reaction at useful speeds and overcome losses. Finding organisms that push the limit further toward the infrared would have practical implications, since far-red and near-infrared light make up a substantial fraction of the solar spectrum that conventional crops largely waste. If crops could be engineered to use some of that light for water splitting and carbon fixation, it could theoretically increase the total solar energy captured by agriculture.