Blood Protein Used to Split Water

brian0918 writes "The Imperial College in London is reporting that genetically-engineered blood protein can be used to split water into oxygen and hydrogen. The abstract can be viewed for free from the Journal of the American Chemical Society." From the article: "Scientists have combined two molecules that occur naturally in blood to engineer a molecular complex that uses solar energy to split water into hydrogen and oxygen. This molecular complex can use energy from the sun to create hydrogen gas, providing an alternative to electrolysis, the method typically used to split water into its constituent parts. The breakthrough may pave the way for the development of novel ways of creating hydrogen gas for use as fuel in the future."

Comments

Energy output = input?

[Disoriented]

Now we just have to figure out if the amount of energy needed to synthesize the blood protein (say, X liters of hydrogen in a fuel cell) is less than
the energy of the hydrogen produced from this process... :)

--
Rare 680X0 and PowerPC posters!

Re:Energy output = input?

[CorSci81]

Well, we're getting pretty good at genetically engineering simple organisms to produce things like this on their own.... (think BT corn).

Re:Energy output = input?

[Chris Burke]

I'd be willing to bet that this compound can be used to break multiple water molecules, just like our hemoglobin can carry another oxygen molecule after it drops one off. The source of energy that allows continued hydrogen production is the sun.

In which case, the main question is the rate at which you can produce hydrogen. How much of the substance do you need, and how much solar energy, to produce how much hydrogen over what period of time? That is what will define whether or not this is a practical method of producing hydrogen. One obvious point of comparison would be an equal-sized photovoltaic solar cell and water electrolysis machine. If it doesn't do better than that, it's pretty worthless. On the other hand it might be a very efficient way to convert solar energy into hydrogen gas for fuel cells, which would be sweet.

Not to mention the other possibilities it opens up in biochemistry. These proteins are fascinating, as is the idea of swapping out the bound metal atom to get different effects.

Re:Energy output = input?

[Foofoobar]

Actually they say it far surpasses the current method of separation and assuming this is a passive process (much like solar power), unless the production costs are over a million dollars for one unit, the time it would take to pay for itself is nominal.

Re:Energy output = input?

[radtea]

Actually they say it far surpasses the current method of separation and assuming this is a passive process

They say nothing of the kind. Quote from the abstract, "The efficiency of the photoproduction of H2 was greater than that of the system using the well-known organic chromophore, tetrakis(1-methylpyridinium-4-yl)porphinatozinc(II ) (ZnTMPyP4+), under the same conditions."

Note the complete lack of superlatives.

Re:Energy output = input?

[Pedrito]

Though it doesn't specify, it's highly unlikely that albumin or porphyrin is used up in the reaction. Instead, it likely cleaves the water molecules (the substrate). Not quite physically tearing it apart, but that probably isn't an entirely inaccurate description either. Many proteins perform functions like this on other molecules. They'll attach to part of the substrate and remove, say an -OH hydroxyl group, or some other piece of the molecule. This is how liver enzymes breaks down certain drugs so that the byproducts (called metabolites) can be removed from the blood by the kidneys. As someone else mentioned catalase from yeast, it works in a similar way and removes an oxygen molecule off of hydrogen peroxide, leaving water and oxygen, but the catalase isn't "used up" in the process.

That said, proteins don't usually last forever and how long they last largely depends on how hostile their environment is, and what constitutes a hostile environment for a protein varies from protein to protein.

Re:Energy output = input?

[sillybilly]

Yes, but most biological matter is easily degraded by UV radiation, plus infection - bacteria would love to feast on your new solar cells.
As a sidenote, to quote the artilce: "Dr Stephen Curry Opens in new window, a structural biologist from Imperial College London's Division of Cell and Molecular Biology who participated in the research explains: "This work has shown that it is possible to manipulate molecules and proteins that occur naturally in the human body by changing one small detail of their make-up, such as the type of metal at the heart of a porphyrin molecule, as we did in this study.
Naturally occur in a human? I was hoping they'd be talking about cow-derived materials, unless they are interested in genetically engineering photosynthetic human beings? Maybe one of our great great grand children will be engineered enough to be vacuum resistant and fully photosynthetic, then he can fly around in outer space while living off of sunshine.

Re:Energy output = input?

[jcr]

What really matters is how much this process drops the cost of separating water, not the energy efficiency.
-jcr

Re:Energy output = input?

[Foofoobar]

Note the complete lack of superlatives.

So greater!=surpasses? I'd suggest a dictionary if I thought you literate enough to even understand what a superlative is.

Re:Energy output = input?

[wetfeetl33t]

Yes, that's true if this were a closed system. However, this isn't a closed system. Energy is entering it (from the sun), and mass is entering and leaving (water, hydrogen and oxygen). So yes, total entropy is increasing somewhere, but it just isn't in this hypothetical system

Re:Energy output = input?

[Tumbleweed]

Now we just have to figure out if the amount of energy needed to synthesize the blood protein (say, X liters of hydrogen in a fuel cell) is less than
the energy of the hydrogen produced from this process... :)

This statement _really_ old, and completely misses the point.

It takes more energy to find, drill for, pump, process, and ship oil, than you ever get out of it. That ain't the point, either.

The POINT is to have a MOBILE fuel (or energy transfer medium, or whatever you want to call it; it's semantics at this point). Energy you can use to move your butt from one place to another in a vehicle.

Laptop computers aren't as powerful as desktop computers, AND they're a lot more expensive as well. That doesn't mean they're worse than desktop computers, just that that is how MOBILE computers work. Same thing with oil/hydrogen/what-have-you.

We expend energy to have a fuel (oil, hydrogen, natural gas) that we can take with us. No, it's not as energy efficient as plugging into the electric grid, but then again, I don't have an extension cord that long. :)

I'm sure it won't be less than the energy of the hydrogen produced by the process, but, what SHOULD we be comparing it to?

How much energy does it take to produce hydrogen via cracking natural gas?
How much energy does it take to produce hydrogen via electrolysis?
How much energy does it take to produce hydrogen via splitting water with a high-temp gas cooled nuclear reactor?
How much energy does it take to produce hydrogen via some genetically-engineered seagrass or other plant?

The efficiency of whatever hydrogen fuel cell or Hydrogen internal combustion engine is another part of the 'efficiency' equation, but then you get to compare that to a fossil-fuel ICE and get to take into account the efficiencies and hard dollar costs of removing pollution from the exhaust, and from pollution mitigation efforts of the stuff that makes it out of the tailpipe, anyway.

Please, everyone, just stop it with the 'will it take more energy to get it into a portable form than you get from the portable energy'? Yes, it will always take more energy to put the energy source into a portable form. It's worth it because YOU'RE PUTTING IT IN A PORTABLE FORM.

Ahem. Hopefully I haven't been too obnoxious here, but this is just really getting old and tired (just like me).

And now I'm hungry.

Re:Energy output = input?

[DrFalkyn]

Actually there is somewhat of a point, since the ultimate source of biological energy is the sun, comparing its efficiency to other methods that involve solar is fair game. Unless they are actually talking about putting the protein *in the vehicle*, to produce hydrogen on the fly, then I would be impressed. If you have an efficient way to produce hydrogen from water on the fly, you wouldn't have to have store hydrogen directly which is extremely difficult to stora at the necessary energy densities for vehicular automotion (either needs extremely low temperature or extremely high pressures)

Re:Energy output = input?

[icebrain]

Are we going to look at this from a pure physics standpoint, or a "practical use" standpoint?

Yes, the laws of physics and thermodynamics say that we need to put more energy into the water/methane/$other_hydrogen_source to "crack" it and get hydrogen than we will get back from burning it or recombining it in fuel cells. However, that's not the point. As other posters have said, _all_ fuels take more energy to create or store than they produce when consumed.

You say that "[h]ydrogen is a non-starter, even with this technology. Why? Simple physics: it takes more energy to unbond water than you get back from burning the hydrogen and thusly re-bonding it back into water. Period, end of story. It's a little thing called the Second Law of Thermodynamics. Deal." Again, all fuels require that. Yes, I know that most of that has already been done for us (plants into animals into fossil fuels), but sunlight is FREE. Once we get infrastructure in place, it is (from a _practical_ standpoint) self-sustaining. In other words, we don't have to sit there and pay for every Joule of solar energy we use, because it's going to be there regardless. Might as well take advantage of it.

It's kinda like designing a rocket vs. designing an airplane. When you design a rocket, you have to carry all of your propellant (oxidizer and fuel) with you, and it all has to be accounted for. Every bit of extra fuel or inefficiency hurts you in overall performance. Similarly, when you design a plane, you know that you need air (your oxidizer) to run your engine and to fly. The difference is, however, that you don't need to worry about carrying the air with you. It's everywhere, and you don't have to worry about where it's going to come from. Essentially, it is free. There's a reason rockets define efficiency in terms of total propellant used, while airplanes define it only by fuel used--I don't really care how much air the plane uses, because again--I don't have to pay for it.

I guess what I'm trying to get at is that, at the real-world practical level of things, efficiency is defined as "what you get" over "what you paid for." We don't have to continually pay money for the sunlight to produce hydrogen (which could then power its own distribution costs) like we have to continually pay for the coal/oil/uranium that would be used for the same thing (or that is used in the process of collecting, refining, and distributing themselves).

protestors...

[User 956]

The Imperial College in London is reporting that genetically-engineered blood protein can be used to split water into oxygen and hydrogen.

I can hear it now... "No blood for oil! or hydrogen!"

Very exciting!

[__aagmrb7289]

I eagerly await the return to the days of human/animal sacrifice. "It's for the good of the country! We need to have more SUVs on the road!" Bow down, I say!

Re:Very exciting!

[Skidge]

The Red Cross can just roll it into their blood drives: "Give a pint of blood and fill up your hydrogen tank!"

Comment removed

[account_deleted]

Comment removed based on user account deletion

Catalase (cool experiment)

[b0s0z0ku]

Blood also contains a protein called catalase. It makes the hydrogen peroxide that you put on a wounds bubble up with little oxygen bubbles. Yeast contains the same protein. Mix yeast and 3% peroxide solution and you get ------ oxygen and water. Stick a burning match in it and it burns with a bright white flame like a welding torch.

-b.

efficiency

[drDugan]

The mention efficiency many times in the article, but do not mention the most important efficiency number - that is total energy in/out.

So, a quick calculation of efficiency:

FTA

Light in:
6 hours, 450 W light = 2.7 kWh

H energy out:
0.044 mL H ... at 4.7 MJ/L (Wikipedia) * 1/1000 (L/mL) * 1/3.6e6 (kWh/J) * 1e6 (J/MJ) =

= 5.7 e -5 kWh

Disclaimer:

This probably has an error, please help me correct it.

It has been a really long time since I did physics or dimensional analysis.

I could not find in the paper the pressure for the 0.044 ml of generated hydrogen, nor it's weight, so I made a gross assumption the energy density listed in Wikipedia (at 700 bar) was close enough.

Regardless, if you put in 2.7 units of energy and get out 0.000057 units... that seems really (s)low.

Re:efficiency

[TubeSteak]

Regardless, if you put in 2.7 units of energy and get out 0.000057 units... that seems really (s)low

Even if your math is off, it might not matter if the process can be scaled up, since solar power is cheap/free.

The important question is how cheaply can they synthesize the needed protein.

Doomsday weapon?

[Reality Master 101]

I wonder if you could bioengineer a plant that could survive in the ocean similar to seaweed, which would secrete this chemical. Eventually all the oceans would turn into Hydrogen and Oxygen... and LIFE WOULD BE DOOMED! Bwahahaha

Re:Desalinization

[dextromulous]

If we're lucky, you'd not only get clean water, you'd get an abundance of (clean, perhaps?) energy that could be converted to electricity.

Thank you!

[mapkinase]

I would like to praise the submitter for providing a link to a peer-reviewed article. Does not happen very often, worth mentioning.

Next:

[jafac]

We'll need one of these that can split Oxygen and Carbon.

(ie - remove Carbon Dioxide from the atmosphere, and plant the Carbon somewhere safe - like maybe in empty petroleum resevoirs, where it came from).

Re:Next:

[Jherico]

You mean these?

Re:We really don't want to do that.

[catbutt]

Given that its from a living thing anyway, it seems like if breaking down hydrogen and oxygen in mass had any survival benefit, natural selection would have figured it out already.

Obviously, caution is always needed in genetic tinkering, but still....I think the knee jerk "OMG its going to zap all our oceans!" is unwarranted.

Re:How effecient is this?

[kebes]

I'm reading over the actual article right now. It seems that process is quite efficient. In the conclusion of the paper they note:

Currently, rHSA(wt) is manufactured in an industrial scale, which allows us to use this zinc-protein photosensitizer in practical applications Thus the raw materials are cheap enough that one could imagine scaling this up significantly. Moreover since its behavior is catalytic, the protein isn't used up, so you wouldn't need to replace it very often.

With regard to efficiency, in the Abstract they also point out that their system is more efficient than the previous standard in organic photo-synthesis:

The efficiency of the photoproduction of H2 was greater than that of the system using the well- known organic chromophore, tetrakis(1-methylpyridinium-4-yl)porphinatozinc(II ) (ZnTMPyP4+), under the same conditions. Since the discovered system is a photosensitized catalyst, it effectively is a new kind of solar power. However it is one that directly generates H2 from incident light, without requiring one to harvest light energy as electricity, store it, and then use it to split water. So this discovery, coupled with cars/devices that run on H2 efficiently, seems like a viable idea. Of course we'll have to wait and see whether this really pans out, but from this paper it does indeed seem that this is a feasible way to harvest solar power (and store it as H2).

Re:How effecient is this?

[jalet]

> Do you have to refuel this?

Yes, but they are still wondering if it's better to refill this stuff with water, or with human bodies...

Biochemical isn't the only approach

[quoll]

I'm pleased to see alternative technologies to split water using sunlight, but the idea is not new.

There is a group at UNSW who have been working on ceramics which use sunlight to split water (via a process of electrolysis). It's still in research (mostly due to efficiency), but it's an interesting option if you're interested in this stuff.

Their website is pretty sparse, but there is a story on them here.

Re:We really don't want to do that.

[CorSci81]

It has, it's called photosynthesis. Granted, here you're not liberating free hydrogen. But to counter the GP argument of using up all water on earth... can you imagine how incredibly unstable the local environment would become for one of these organisms in the wild? They'd be very liable to kill themselves off either through pH changes or simply setting their environment on fire if they reproduced unchecked. That combined with the fact you could never split all the water on earth faster than it will recombine if sunlight is your only energy input.

Re:We really don't want to do that.

[SquareOfS]

Umm . . . we already do this? Check it out.

Net loss of 1 H2O molecule in the Krebs Cycle. And plenty of other places as well, I assume.

It's impossible, one presumes, for any standard cellular organism to destroy all water in its environment, because then no biochemical processes could occur and it would be dead.

I presume the way this works is that they isolate the protein, rather than adding the organism to the water. And proteins don't self-replicate.

Re:We really don't want to do that.

[Coryoth]

I think we (collectively) need to reevaluate the risk factors here; such an organism in the wild could very well turn our planet into a dustbowl in such a shockingly short time we wouldn't even have time to lynch the scientists who created it (think: hours - the mathematics of unchecked reproduction are truly alarming).

Who said anything about reproduction, let alone unchecked reproduction? The article says it is a molecular complex, not a living organism capable of reproduction. I expect it is just an enzyme to catalyse the reaction, so I wouldn't worry about this any more than you would be inclined to worry about naturally occuring cellulase suddenly going rampant and destroying all plant life on earth in a matter of hours. Generally being somewhat informed is a prerequisite critical analysis of risks and any ensuing scaremongering (okay, that's not true, i just think it should be a prerequisite!).

Finally reading about...

[fahrbot-bot]

...something on the bleeding edge of technology here on /.

Ok, ok, OK. I promise not to post for the entire weekend, sigh.
Damn.

Problem with large scale use?

[The Step Child]

From the abstract:

In the presence of the colloidal PVA-Pt as a catalyst and triethanolamine (TEOA) as a sacrificial electron donor, the photosensitized reduction of water to H2 takes place.

My chemistry knowledge isn't really up to the point where I can fully understand the whole abstract, but it sounds like we still need triethanolamine as a source of electrons in order to reduce water to H2. So the energy needed to produce more triethanolamine could put a big dent in the net energy gained from the H2 produced when we're talking about the practical large-scale usefulness of this. Maybe one day we could use another (renewable) electron donor like NADPH, so that we can couple it to another biochemical process like photosynthesis in order to renew our electron donors :)

Re:Problem with large scale use?

[Anonymous Coward]

Right on the money, it's hard to believe the Journal of the American CHEMICAL Society would publish this as is. First, water is not reduced; protons are reduced. If they had added a little acid, they could have claimed "photosensitized reduction of acid X" just as well. Second, as you mentioned, triethanolamine is a sacrificial reductant. So, the Slashdot title is misleading--water is not being split (which in the scientific jargon means reversing combustion of hydrogen, i.e. net reaction 2H2O --> 2H2 + O2, no sacrifical reductants allowed). Their work leaves the oxygen in its reduced state. Third, they require a platinum catalyst. There are already means to use expensive Pt catalysts or electrodes wired into a circuit with a photovoltaic module. What they've accomplished is light-driven oxidation of TEA to H2. From an industrial-scale energy conversion standpoint, this work doesn't seem terrifically relevant.

It is a neat use of HSA to solubilize Zn protoporphyrin IX and prevent collisional self-quenching of the excited triplet state, though. And, they achieved a greater efficiency in this non-water-splitting reaction than another known catalyst, so that's worth knowing too, I suppose.

Re:Splitting CO2

[noigmn]

Yeh was thinking the same. If we could break down carbon and sulphur compounds in the air, it would be a big step forward in fixing global warming. And also in atmospheric engineering, which we might need if we decide to create an atmosphere on Mars.

Imagine if photosynthesis could work with whatever compound we wanted. We could have it on space ships to break the CO2 breathed out back into O2 to rebreath also. Might also work for divers.

Re:Desalinization

[oni]

Actually, I thought the abstract said that the compound used is oxidized, meaning that the oxygen is captured and only the hydrogen is released. If I read the abstract wrong, please correct me.

My (admittedly layman's) understanding is thus: they have a molecule that sticks to oxygen. Put the molecule into water and it grabs the oxygen away from H2O, releasing H2. That by itself is not very impressive. Sodium does something similar. So here's the cool part, when exposed to sunlight, the molecule releases its oxygen - thus the process will go on so long as you have sunlight and water. This is only interesting because the molecule works like a catalyst.

IF it really works (I am cautiously optimistic) this could be the biggest discovery in the history of the world. It could mean that our civilization is no longer on the road to oblivion. It could mean no more energy wars (but don't worry, we'll still have to fight the United Atheist Alliance).

Re:How effecient is this?

[smackt4rd]

Just catch some small animals and stuff them in the gas-tank. :)

Where do the electrons come from?

[Draka]

The abstract also mentions "In the presence of the colloidal PVA-Pt as a catalyst and triethanolamine (TEOA) as a sacrificial electron donor, the photosensitized reduction of water to H2 takes place." This basically means that electron fro TEOA is being used to reduce water to hydrogen. This chemical (TEOA) is oxidized and has to be replenished to maintain the H2 production rate. I am not disparaging their results (they are valuable, otherwise it would not be published in such a reputed journal), but trying to put things in perspective. Compare this to the reports of water splitting using titanium dioxide and other ceramics ( http://adsabs.harvard.edu/abs/2006ApPhL..89p3106P, http://edu.chem.tue.nl/6KM11/files/Project%20repor ts%202003%202004/Photocatalytic%20water%20splittin g.pdf ) where water is split to yield hydrogen and oxygen without the need for any "sacrificial electron donor".

Re:We really don't want to do that.

[Qzukk]

(The molecule would "eat" it's own host up.)

This might be of interest to you.

The Brits will love this.

[Roadkills-R-Us]

I can hear it now.

``Bloody hydrogen!''

Re:We really don't want to do that.

[_argonauta]

now immagine that water molecules could fight back! what a mess! war of the worlds! (hypothetical joke, of course)

Re:Desalinization

[hadhad69]

The NaCl in the sea water may interfere with the catalytic pathway in question, its another story altogether really

Re:How effecient is this?

[Hubristically Yours]

Not sure about the recombinant albumin, but part of my job involves pharmaceutical purchasing, and a vial of 20mL of 25% human serum albumin can be obtained for approximately $13. The human version is produced by precipitation from donated blood and is used quite routinely in the hospital to treat various conditions such as shock or malnutrition. Also, many medications are packaged with albumin in the vial (to provide a binding surface for the drug molecules).

I would say though that the "manufactured in an industrial scale" statement is a bit misleading. Purified blood proteins in general are ungodly expensive. For instance, immunoglobins, which you might get to protect you against infection if you've been exposed to, say, Hepatitis B or C, are some of the most expensive drugs we have, ranging up into the thousands of $ per shot. Most of these are refined from human blood, but even if you have trillions of bacteria slaving away for you producing recombinant proteins, it's the purification and quality control steps that are the killer.

Re:age old mistery finally solved !

[ibn_khaldun]

Moses would have happily run the service on Sunday; it was Saturday he would have skipped. Hmmm, except he didn't even have those laws yet, and when he did, Version 1.0 crashed (literally) and only Version 2.0 was widely marketed. Success of the product is still debated.

Porphyrin chemistry is very interesting...

[alchemist68]

Porphyrin chemistry is very interesting and has been studied for over 100 years. This news is both exciting and old news, because porphyrins and related isomers have been the subject of continued research. For very detailed information about porphyrin chemistry, refer to The Porphyrins edited by David Dolphin. Also, review the research of Martin Gouterman. In biological systems, porphyrins are found commonly in heme-type proteins used for oxygen transport and cytochrome P450 in the liver for metabolizing biological compounds including pharmaceutical products, and as chlorophyll in plants. Porphyrins have served as catalysts for organic reactions in industry, photodynamic therapy for cancer, molecular devices including sensors and switches, and model compounds for the active sites of enzymes. My thesis, which available for download through OhioLink:

http://www.ohiolink.edu/etd/view.cgi?akron11339504 18

details the photophysical characterization of N-Confused tetraphenylporphyrin and characterization of zinc N-Confused tetraphenylporphyrin.

Upon reading this post on Slashdot, I was pleasantly surprized that the subject of my thesis has some similarities to a related compound that could be used for further research into catalyzing an energy source. In one way I'm surprized, and in another I'm not, and I'm glad that one of the Slasdot crowd submitted the post. Porphyrin chemistry is vast, interesting, and complex.

Happy reading!

Sacrifical Donor

[Roxton]

n the presence of the colloidal PVA-Pt as a catalyst and triethanolamine (TEOA) as a sacrificial electron donor, the photosensitized reduction of water to H2 takes place. [Emphasis mine]
Isn't this a problem? How do you restore the triethanolamine without using energy?

Our Machine Overlords

[sylvainsf]

So scientists have invented a way for the machines to get cheap hydrogen power FROM OUR BLOOD?

check your physics, spoilsport

[bodrell]

Hydrogen is a non-starter, even with this technology. Why? Simple physics: it takes more energy to unbond water than you get back from burning the hydrogen and thusly re-bonding it back into water. Period, end of story. It's a little thing called the Second Law of Thermodynamics. Deal.

Hydrogen is a Really Bad Idea.

Nothing can contain it, storing it (as a supercold liquid) takes enormous amounts of energy, and, at root, it's got negative ER/EI. I don't care if it's in a portable form - IT'S NOT A SUSTAINABLE TECHNOLOGY. PERIOD.

I suggest bicycles.

Refuting your points, one-by-one:

1) Energy doesn't come from out of the ether; even oil comes from sunlight's energy, ultimately. All organic matter is fuel, and it took a lot more energy (from the sun) to produce that fuel than will be obtained from burning it. That would be the case even if extraction and separation were free, which is far from reality. It takes a LOT more energy to vaporize water into steam than is obtained from the mechanical energy in steam. Even a Carnot engine is less than 40% efficient. But guess what? That lack of efficiency doesn't matter when the heat is free, from geothermal to solar sources. Are you going to tell me geothermal is a "non-starter" because of the difference in energy input vs. output? Didn't think so.

2) I think you mean the First Law of Thermodynamics, conservation of energy. The Second Law simply states that the entropy of the universe will continue to increase.

3) Hydrogen need not be stored as cold liquid in a tank. The focus of hydrogen technology right now is matrices that can absorb hydrogen at one pressure / temperature, then release it with a pressure / temperature swing in a controllable fashion. Other ideas involve chemically releasing hydrogen (from ammonia, for example) as needed. No one said gaseous hydrogen was the be-all end-all.

4) Bicycles are great, but we should be riding those regardless of what fuel goes in the gas tank. It's also difficult to, say, move furniture with a bicycle.

Re:We really don't want to do that.

[maraist]

such an organism in the wild could very well turn our planet into a dustbowl

So why haven't trees stripped every ounce of Carbon Dioxide from the atmosphere?

Because there is more to a chemical process than one input (such as water).. For photosynthesis, there are many chemicals and input sources that are necessary. Sunlight being the most critical element, as it's what provides the energy.

You can do some simple math to figure out how much energy would be necessary in a 100% efficient environment to convert the ocean to Hydrogen and Oxygen.. Then take into account that very little of the high energy solar radiation actually gets to the earth's surface. Then take into account the starvation of constituent ingredients. In photo-synthesis, you need carbon dioxide, Oxygen and water. I don't recall the exact cycle. But for the engine to operate you need to efficiently feed all ingredients in the exact mixture. In nature, this happens through diffusion.. The "waste" products slowly ooze out, while the ingredients seep in (with sun-light permiating based on ideal geographic locations).

Then you have competition between the cells.. They fight over one another, thus starving one or more ingredients. But much like a database deadlock situation. If A blocks B for resource 1 and B blocks A for resource 2, then you have an inpass.

Finally, there are counter-weights in nature. As the chemical makeup of the surroundings change (due to super-saturation of new elements, and th starvation of others'), the ability to do business as usual degrades. The chemical engines themselves, eventually become the food source of some other mechanism.

Thus, even in a homogenous environment of some genetically engineered cellular factory, it would be nearly impossible for the oceans to run dry. SOOOO many factors would kick in LONG before any appreciable progress was made.

Now, it's possible under the right circumstances for a desert's lake to dry up, for example (assuming the right minerals exist to promote cellular replication).

But as other posters have noted, if this were an easy thing to occur, it would have already happened naturally and there wouldn't be water on earth today.