麻豆传媒

TAKE a long, deep breath. Now spare a moment to think about the precious oxygen you鈥檝e just inhaled. Without it you鈥檇 be unconscious in a couple of minutes, and dead soon afterwards. It鈥檚 no mystery where this life-sustaining gas comes from: plants churn it out as a by-product of photosynthesis. If they didn鈥檛, neither we nor any of the Earth鈥檚 other animals would be around right now.

But if you look a bit deeper at how plants do this, things start to get seriously puzzling. The oxygen is released by splitting water molecules, a reaction that in the lab requires such violence that it would tear any living system apart. So how can plants do it, using only energy from the Sun? 鈥淚t is the most important biological reaction on Earth,鈥� says Jim Penner-Hahn, a chemist at the University of Michigan in Ann Arbor. 鈥淚t is also one of the most complicated reactions in the biosphere.鈥�

Now Gary Brudvig and Robert Crabtree of Yale University believe they have the key to understanding the plant鈥檚 secrets. They are not biologists, though. They are chemists who have created an artificial version of the plant鈥檚 oxygen machine. They hope that by studying the way it works they can learn more about how plants split water molecules. The dream is that it could eventually lead to new ways of harvesting the power of the Sun, producing not only oxygen but also hydrogen, the gas touted as the fuel of the future.

With its perfectly orchestrated sequence of reactions, photosynthesis eventually creates all the fuel and chemical building blocks that plants need to live and grow. Biochemists studying the inner workings of green plants already have a pretty good grasp of the first stages in this sequence. When a plant absorbs light, each photon triggers a cycle of reactions in which an electron is shuttled from one molecule to the next. In the first of these steps, a photon flicks an electron out from a specially aligned pair of chlorophyll molecules-known as P680-and passes it to the first electron acceptor.

It wouldn鈥檛 be much good if each P680 only did this once; to capture useful amounts of energy it has to absorb photon after photon, over and over again. But before it can absorb another photon, and send another electron down the line, it must be 鈥渞eset鈥� with a new electron.

This replacement comes from a small complex of proteins and metal ions that nestles alongside P680. Somehow, this complex manages an astonishing task. It pulls an electron out of a water molecule-and in the process breaks down the water into hydrogen ions and oxygen. The assembly that keeps P680 supplied with electrons is known as the oxygen-evolving complex, and it is thanks to the OEC鈥檚 tireless efforts that we have air to breathe.

Given the vital role of the OEC, we know surprisingly little about the way it works. We do know some of the basics, though. For the OEC to release a single molecule of oxygen, biochemists have discovered that light must trigger the plant鈥檚 electron-transfer cycle four times, splitting two water molecules, releasing four hydrogen ions and a single molecule of oxygen, and passing on four electrons. This complicated chemistry seems to occur in the very heart of the OEC, at a single cluster of four manganese ions.

These ions have a vital job to do. Try breaking a water molecule apart in one go and you鈥檇 need a huge blast of energy. To do the job with heat energy alone, for example, you would need to raise the temperature of water by thousands of degrees-more than enough to vapourise a geranium. The great thing about manganese ions is that they have several oxidation states: in other words they can lose electrons one at a time, becoming a more powerful oxidising agent or 鈥渆lectron grabber鈥� as each one is removed.

At the moment, that鈥檚 pretty much all we know about the OEC, because it鈥檚 notoriously difficult to study. Along with other components of the plant鈥檚 electron-transport system, it is attached to proteins that fit together into larger complexes like the pieces of a jigsaw. 鈥淲e can only study it as part of a large protein complex,鈥� says Brudvig. 鈥淚t is not possible to remove an intact OEC.鈥� If you try to isolate the complex from the other components, the proteins that hold it together change shape and the manganese cluster breaks apart, leaving individual manganese ions and a completely useless mess of proteins.

This is part of the reason why researchers are so keen to build synthetic versions of the OEC. By comparing the behaviour of the artificial complexes with that of the real thing, they hope to unravel some of the mysteries of how it works. Already, various research groups have put together an assortment of exotic water-splitting catalysts using all sorts of metal ions. The trouble is that few of them release bubbles of oxygen. And of those that do, most use metals other than manganese.

The Yale team鈥檚 achievement is to have engineered an artificial OEC with manganese ions at its core. And although it isn鈥檛 identical to its natural counterpart-whereas a natural OEC has four manganese ions, the artificial complex has just two ions connected by two bridging oxygen atoms-it certainly seems to work in much the same way. But most importantly, Brudvig and Crabtree have shown that their complex is reasonably long-lasting. It is not destroyed every time a molecule of oxygen is formed but, like a true catalyst, survives to work over and over again.

The secret of the Yale researchers鈥� success lies in their choice of ligands-the molecules that surround the core of metal ions. Manganese ions often form complexes by binding to six ligands, and in this state the bonds to the ligands tend to be so strong that no further reactions can take place. The researchers knew they had to avoid this: 鈥淥ur goal was to fill five of the binding sites with tightly bound ligands and have one site filled with a weakly bound ligand,鈥� says Brudvig. Using fewer than five sites would make the complex unstable, allowing it to rearrange itself to form an inactive structure. So Brudvig and Crabtree added a pair of semicircular terpyridine ligands which bind to each manganese in three places (see Diagram). The two oxygen bridges take up another two sites, leaving just one available. The researchers hoped this would allow a molecule of water to stick on and react.

The terpyridine ligand鈥檚 big advantage, says Crabtree, is that it鈥檚 a tough, rigid molecule. This helps it survive attack by the oxygen radicals and other reactive species that are produced as water splits, and which would quickly destroy the catalyst given the chance.

All they needed now was a source of energy. Instead of sunlight, Brudvig and Crabtree decided to rely on brute force, in the shape of sodium hypochlorite (NaOCl), the main ingredient of household bleach. Hypochlorite (OCl^鈥�) is a powerful oxidising agent-which means that it tends to drag electrons out of other molecules. The idea was that it would suck electrons from the catalyst, playing the part that P680 plays in plants.

For their first tests, the researchers added sodium hypochlorite to a solution containing the catalyst, and waited. Gradually they began to see tiny bubbles appearing, Brudvig recalls. When they analysed them, the researchers were delighted to find they contained oxygen. Best of all, the catalyst didn鈥檛 fall apart immediately after the reaction had begun. They calculate that over a period of six hours, each manganese complex in the solution produced about five oxygen molecules. And when Brudvig replaced the bleach with oxone, one of the strongest commercially available oxidants, each complex produced more than 100 molecules of oxygen before it was destroyed.

So how does their catalyst work? Brudvig and Crabtree believe the sequence goes like this. First, a molecule of water binds to the one available site on each manganese ion in the complex. Next, a hypochlorite ion from the bleach displaces the water molecule from one of these sites, losing a chloride (Cl^鈥�) ion in the process. This leaves the oxygen atom attached to the manganese ion by a highly reactive double bond.

In the final stage, this oxygen grabs onto an oxygen atom on a passing water molecule to form oxygen gas. The researchers believe this oxygen-manganese complex is the key intermediate in the natural OEC, so they designed their complex to mimic it.

To back up their ideas with some solid evidence, they labelled the oxygen atoms in the water by replacing them with the heavier-than-normal isotope oxygen-18. When they examined the oxygen given off by the artificial complex, they discovered that the proportions of isotope in the gas matched their predictions for the reaction. This tallies with experiments on plants suggesting that the two oxygen atoms in the gas they release come from different places-one from the manganese complex and one from a nearby molecule of water. Brudvig claims that his synthetic OEC is the first that both resembles the crucial part of the plant鈥檚 manganese cluster and is able to produce more than a single molecule of oxygen per complex.

鈥淭he functional model we have really does seem to mimic the natural system both structurally and mechanistically,鈥� says Crabtree. 鈥淔or me, this is its most exciting property.鈥� This has allowed the researchers to put together a model for the final steps in oxygen evolution in the plant鈥檚 OEC, a process that also involves a calcium ion. Just months ago, in a paper in the Journal of the Chemical Society-Dalton Transactions, Brudvig suggests that molecules of oxygen form when one of the oxygen atoms attached directly to a manganese ion reacts with another oxygen attached to a nearby calcium ion (see Diagram).

But many mysteries remain. For example, if an artificial catalyst with only two manganese ions can do the job, why does the natural OEC need another pair as well? 鈥淭hey may be a bit like innocent bystanders,鈥� says Brudvig. Tinkering with the chemical structures of synthetic OECs might reveal their role, as well as helping researchers to discover exactly what part the proteins play in the chemistry of the real system. 鈥淲e can use the models as proving grounds for our ideas on the photosynthetic reaction centres,鈥� explains Brudvig. 鈥淸Models] can be altered systematically and many variables can be controlled that cannot be in the plant protein itself.鈥�

Another key problem is to protect the artificial OEC against damage from the highly reactive water-splitting chemicals. Brudvig claims that his catalyst is able to produce more than a single molecule of oxygen before breaking down. But there is still plenty of room for improvement. To make the catalyst last longer, the researchers want to design an inflexible ligand that will bind more tightly to the manganese ions. The OEC in the plant is also battered by a variety of highly reactive species produced during water splitting, but it can get around this by constantly making new proteins. 鈥淭his is one of the drawbacks of the artificial system,鈥� admits Brudvig. 鈥淚t can鈥檛 constantly regenerate itself.鈥�

The team is also experimenting with alternative oxidising agents to power the water-splitting reaction, including cerium, cobalt and ruthenium ions. While hypochlorite is moderately alkaline, these agents operate at a pH that is closer to the environment of a plant cell. Electricity, delivered with a pair of electrodes, is another possible source of oxidising power. 鈥淭his would provide a great deal more control because we can set the driving force ourselves,鈥� says Brudvig. But in the long term, Brudvig and Crabtree dream of using light as the power source, just as nature does, and are working with other researchers to design a catalyst tethered to a chlorophyll-like molecule that will be able to harvest energy from sunlight.

Brudvig and Crabtree believe that what they鈥檝e learnt about the plant鈥檚 oxygen machine could also be useful to the chemicals industry. Oxygen plays a part in a multitude of industrial processes, and what could be cleaner or more environmentally friendly than an artificial system that splits water using little more than sunlight? 鈥淏y developing a catalyst that can convert water to oxygen, the Yale group has opened a potential avenue for efficiently splitting water with solar radiation,鈥� says Terry Collins of Carnegie Mellon University in Pittsburgh.

Eventually, it might even be possible to make a catalyst that produces hydrogen gas. During water splitting, the OEC releases hydrogen ions but these are immediately mopped up by other molecules. Instead, if chemists could design a catalyst to bring pairs of these hydrogen ions together to form hydrogen gas, they would end up with a cheap, efficient source of fuel which has no end of uses. 麻豆传媒, 23 November 1996, p 40).

And all sorts of industrial processes rely on a precise sequence of electron-transfer reactions just like the ones that take place in the OEC. 鈥淚t鈥檚 possible that synthetic clusters might be adapted to do multielectron chemistry on things other than water to produce economically important chemicals,鈥� says Charles Yocum, a chemist at the University of Michigan. As a first step, Crabtree is planning to use the catalyst to pull electrons from hydrocarbons such as toluene.

Whatever the long-term prospects for their catalyst, Brudvig and Crabtree are pretty satisfied with the insights it is providing into one of the most crucial biochemical reactions on Earth. 鈥淚 think we are a lot closer to understanding how plants make oxygen,鈥� says Crabtree. Of course, the water-splitting process still has its secrets. 鈥淚t鈥檚 still very much a blackbox,鈥� says Penner-Hahn. 鈥淏ut I think that we鈥檝e succeeded in making the black box smaller.鈥�