Researchers Achieve Unprecedented Efficiency in Photoelectrochemical Production of Hydrogen from Water: 42%
04 December 2006
Michael Grätzel and his colleagues have developed a device that sets a new benchmark for efficiency in splitting water into hydrogen and oxygen using ordinary sunlight. The research will be published in the 13 December issue of the Journal of the American Chemical Society.
Previously, the best water photooxidation technology had an external quantum efficiency of about 37%. The new technology’s efficiency is 42%, which the researchers term “unprecedented.” The efficiency is due to an improved positive electrode and other innovations in the water-splitting device.
Grätzel and collaborators developed the Grätzel Cell, a dye-sensitized photoelectrochemical cell that uses photo-sensitization of wide-band-gap mesoporous oxide semiconductors. The work originally appeared in a paper in Nature in 1991.
As reported in the Nature article, the original overall light-to-electric energy conversion yield of a Grätzel cell was 7.1–7.9% in simulated solar light and 12% in diffuse daylight.
Iron oxide (-Fe2O3, or hematite) is an especially attractive photoanode due to its abundance, stability, and environmental compatibility, as well as suitable band gap and valence band edge position. Unfortunately, the reported efficiencies of water oxidation at illuminated hematite electrodes are notoriously low.
Grätzel and his team tackled that in this most recent work by producing Fe2O3 photoanodes via deposition of silicon-doped nanocrystalline hematite films by APCVD (atmospheric pressure chemical vapor deposition).
The result was a highly developed dendritic nanostructure of 500 nm thickness having a feature size of only 10-20 nm at the surface. The dendritic nanostructure minimizes the distance photogenerated holes have to diffuse to the Fe2O3/electrolyte interface in a film that is thick enough for strong light absorption.
The efficiency is further enhanced by deposition of a thin insulating SiO2 layer below and a cobalt monolayer on top of the Fe2O3 film.
Under illumination in 1 M NaOH, water is oxidized at the Fe2O3 electrode with higher efficiency (IPCE [incident photon to current efficiencies] = 42% at 370 nm and 2.2 mA/cm2 in AM 1.5 G sunlight of 1000 W/m2 at 1.23 VRHE) than at the best reported single crystalline Fe2O3 electrodes.
Resources:
“New Benchmark for Water Photooxidation by Nanostructured
-Fe2O3 Films”; Andreas Kay, Ilkay Cesar, and Michael Grätzel; J. Am. Chem. Soc., ASAP Article 10.1021/ja064380l S0002-7863(06)04380-0
Can someone here translate this into something a garden variety energy geek can understand? For example, is this breakthrough more or less efficient than using a 15% efficient solar cell to electrolyze water?
(I did some research online about quantum efficiency, but that didn't lead me to anything that would help me answer the above question on my own.)
Posted by: Lou Grinzo | 04 December 2006 at 12:57 PM
Quantum efficiency is not the same as energy efficiency. Quantum efficiency refers to the fraction of photons that lead to the desired reaction, rather than the fraction of the photons' energy that is used rather than dissipated as heat.
Posted by: Paul Dietz | 04 December 2006 at 01:02 PM
I went through the Resources: link, and this seems to be more efficient at the blue-UV part of the spectrum. 370nm, which is already long wave UVA, might prove to be useful if combined with a photovoltaic/concentrator system that efficiently use longer wavelengths(500-700nm).
Posted by: allen_Z | 04 December 2006 at 02:06 PM
I'm a little confused about the term "photooxidation". Normally, taking oxygen out of a compound is considered a chemical reduction. How is splitting H20 to yield H2 and O2 an oxidation process?
Posted by: Rafael Seidl | 04 December 2006 at 02:57 PM
In simple terms it means its using 13 or so % more phtons. That is fairlybig.
Posted by: wintermane | 04 December 2006 at 02:57 PM
I'm a little confused about the term "photooxidation".
Oxidation in general means 'removing electrons from'. Here, the photogenerated holes remove electrons from water. I guess the reaction is something like:
H2O --> 1/2 O2 + 2 H* + 2e-
or
2 OH- --> 1/2 O2 + H2O + 2e-
Posted by: Paul Dietz | 04 December 2006 at 07:41 PM
Can someone here translate this into something a garden variety energy geek can understand? For example, is this breakthrough more or less efficient than using a 15% efficient solar cell to electrolyze water?
Well, 2.2mA/cm^2 @1.23V (for a kW/m^2 of "AM 1.5G" (??) sunlight) is 27 Watts for a 1 square meter cell. A 15% conventional solar cell would give you 150 W in that situation, or 5.6 times as much power. I'm guessing that it would be more efficient to use the solar cells and run an electrolyzer, unless I'm misinterpretting the meaning of the specs they gave.
Posted by: George | 04 December 2006 at 09:16 PM
i can't help but agree that the given spec does appear to mean 27 W/m^2.
of course, that is from an early experimental setup so may not represent the technologies real efficiency.
most importantly, though, is the expected system cost, which could be significantly lower than Si-based tech, since the most expensive material listed is iron.
Posted by: shaun mann | 04 December 2006 at 11:34 PM
Could the oxydation be: (using energy against the gradient)
H2O + O -> H2 + O2
effectively being: 2H2O -> 2H2 + O2
Posted by: Neil | 05 December 2006 at 12:18 AM
Energy geek observations:
"Deposition of silicon-doped nanocrystalline hematite films by APCVD produces Fe2O3 photoanodes that oxidize water under visible light with unprecedented efficiency. The dendritic nanostructure minimizes the distance photogenerated holes have to diffuse to the Fe2O3/electrolyte interface in a film that is thick enough for strong light absorption."
Most interesting is the process' dendritic nanostructure mimicking more commonly recognized neurological biology. And the layered process allowing wide spectrum absorption.
Assuming the cost of sunlight at near zero - a hematite semiconductor producing H2 @ +20% efficiency might be a business - if scaled manufacture costs are reasonable and H2 capture and storage have no significant problems.
Posted by: gr | 05 December 2006 at 07:11 AM
Using anything to make hydrogen as a way of storing energy is daft due to the problems of storing energy as hydrogen. Posts like this show that.
The sooner money is stoped being wasted on hydrogen the sooner the world can focus at solving its energy problems in a sensible way.
Posted by: John Baldwin | 05 December 2006 at 07:49 AM
To clarify the "oxidation" discussion, Paul Dietz is correct that the use of the term is in reference to the loss of electrons by the oxygen atoms, and everyone is correct to point out that it's not the most accurate description.
In terms of formal oxidation states, the oxygen in water is at oxidation state -2, while the hydrogen atoms are at +1. After the reaction, which is balanced as:
2 H2O --> 2 H2 + 1 O2
the hydrogen atoms in H2 and the oxygen atoms in O2 are all at oxidation state 0, so the oxygen atoms have been oxidized (from -2 to 0), and the hydrogen atoms have been reduced (from +1 to 0).
This reaction is more accurately called a disproportionation since the electron donor and acceptor are within the same molecule, but "oxidation of water" seems to have caught on as a general term for this reaction.
Posted by: Andrew | 05 December 2006 at 08:08 AM
Water splitting by (photo)electrolysis requires two electrodes. In alkaline solution (NaOH --> Na+ + OH-) the reactions are:
1. oxidation at the anode (positive electrode):
2 OH- --> 1/2 O2 + H2O + 2 e-
2. reduction at the cathode (negative electrode):
2 (H2O + e- --> 1/2 H2 + OH-)
which combined gives :
H2O --> H2 + 1/ O2
At room temperature this reaction requires theoretically a voltage of 1.23 Volt as driving force. However, due to an overvoltage of several 100 mV, especially at the anode, a voltage of more than 1.6 V is needed in practice. In photoelectrolysis part of the driving force is delivered by the energy of the light.
Posted by: Emosson | 05 December 2006 at 09:07 AM
In response to Baldwin, who called this project daft - I must disagree.
For one, the research into this sort of cell has much scientific overlap with the design of solar -> electron conversion cells -- as it does electron -> light (LED) circuits. Gratzel is known for the Gratzel cell - a solar -> electron device. Photon -> hydrogen are his new hobby.
Secondly... hydrogen might very well come in handy to power long distance big rig trucks, long distance cargo ships, and possibly jets. Basically, anything that would be impractical to power via batteries. It may not be hydrogen that powers these -- it may be methanol. You can make methanol by reacting with ambient CO2 (read about the methanol economy)
Lastly.. This man is a genius -- I think we can trust he'll choose what is best to work on.
Take care --
Matt
Posted by: Matt | 05 December 2006 at 09:38 AM
Here is interview of Michael Grätzel from September 2006
http://www.technologyreview.com/printer_friendly_article.aspx?id=17490
Posted by: argod | 05 December 2006 at 11:17 AM
John,
Roger made a good suggestion in another post re: H2 fired power plants, derived from on site bioreactor farms. Such a scheme would surely require buffering if not outright storage of H2 prior to combustion. Granted it's not an immediate broad-based energy solution for transportation - but there will be vertical applications for H2 in the energy future.
www.greencarcongress.com/2006/12/arizona_public_.html
Posted by: gr | 05 December 2006 at 12:42 PM
Isn't there enough pressure on the world's (fresh) water? To suggest breaking it down to it's molecular components as fuel is crazy. Even if we use sea water it's needs more processing therefore requires more energy so what's the point. I think pursuing Hydrogen is a waste of time since is doesn't exist on this planet in a usable state. We should be directing our resources elsewhere.
Posted by: Mark | 05 December 2006 at 02:58 PM
Now how about working on the plutonium to store it?
Then there is the problem that you can't park it anywhere just like Natural Gas cars are prohibited.
Would you want to keep a hydrogen bomb in your air-tight garage? Only 1/10th of the energy required for gas to explode will ignite hydrogen. No more cellphones in cars - too dangerous. Static from rubbing your hands on your car seat when there is a microscopic leak in the hydrogen tank - explosion. Focus on non-flammable batteries like Saphion or nanosafe. Or improve biodiesel , svo , wvo engines. Moms want safety not just innovation.
Posted by: Ron D | 05 December 2006 at 02:59 PM
Isn't there enough pressure on the world's (fresh) water? To suggest breaking it down to it's molecular components as fuel is crazy.
Not if you actually compute how much water this would use.
Even if used on a large scale, it would be a tiny fraction of the volume of water already used for things like irrigation or evaporative cooling.
Posted by: Paul Dietz | 06 December 2006 at 08:52 AM
Garden variety means convert it to something normal people understand, like Dollars, or production.
Does this finding revolutionize the production of Hydrogen or just get closer to being marginally sub-standard to gasoline.
Will the people that sell Hydrogn switch from conversion of natural gas as feed stock to water because of this?
Posted by: Greg Woulf | 06 December 2006 at 10:11 AM
Secondly... hydrogen might very well come in handy to power long distance big rig trucks, long distance cargo ships, and possibly jets.
Hydrogen is already very useful. Something like 9 million tons of the stuff is made every year in the US, mostly for industrial purposes.
If (and that's a very big if) this process eventually leads to something that is economically competitive with conventional sources of hydrogen, there is a ready-made market waiting for it.
Posted by: Paul Dietz | 06 December 2006 at 11:47 AM
"We" aren't wasting our time. At the very worst, Gratzel and his associates are wasting their time; which is entirely their right. (Although I doubt that this example is a waste of time, this is useful basic science.)
The more important thing is that Gratzel et al. created a production process for making materials that can be used for other photoelectrochemical processes. This will turn out to be a very useful engineered material.
Also, H2 is needed in a lot of chemical processes where this technology can be fully controlled. I can think of two derivative uses for the technology in the making of methane with CO2 (instead of using natural gas or petroleum oils) which can then be used for monomers and plastic production; and in the use of H2 in fuel cells for residences (again, instead of natural gas).
Both examples of industry and residences are huge producers of GHGs. I think it is silly to require broad-spectrum GHG reductions across the board right out of the gate. Maybe a first good step is offsetting GHG production in chemical production and home use while engineers work on the personal vehicle sources of GHGs.
True, H2 use in vehicles is a tricky engineering problem (as compared to a basic research problem), but I have faith that smart, well-motivated people will prevail over the minor engineering issues that remain to make the technology economic in a "user with an 80 IQ" world.
Posted by: km519 | 07 December 2006 at 03:05 AM
Isn't there enough pressure on the world's (fresh) water? To suggest breaking it down to it's molecular components as fuel is crazy.
(Paul Dietz replies:)
Not if you actually compute how much water this would use.
Even if used on a large scale, it would be a tiny fraction of the volume of water already used for things like irrigation or evaporative cooling.
Not only that, but guess what you get when you burn the hydrogen to produce energy?
Posted by: George | 07 December 2006 at 08:48 PM
Now how about working on the plutonium to store it?
Then there is the problem that you can't park it anywhere just like Natural Gas cars are prohibited.
Would you want to keep a hydrogen bomb in your air-tight garage? Only 1/10th of the energy required for gas to explode will ignite hydrogen. No more cellphones in cars - too dangerous. Static from rubbing your hands on your car seat when there is a microscopic leak in the hydrogen tank - explosion. Focus on non-flammable batteries like Saphion or nanosafe. Or improve biodiesel , svo , wvo engines. Moms want safety not just innovation.
Sigh. The error density of this post is really impressive. Hydrogen's lower explosive limit is 4.0%. Gasoline is 1.2%. Hydrogen is light, so it disperses upward in air. Hydrocarbons creep along the ground untile they find an ignition source. Plutonium to store it? Huh? Hydrogen is stupid, but it isn't particularly dangerous. On the other hand, maybe we could exploit fear and ignorance to get the public to put a stop to the hydrogen nonsense. Sort of like the Bush administration, only for good instead of evil.
Posted by: George | 07 December 2006 at 09:03 PM
How much water does it take to make a pound of Hydrogen? When it burns the 'waste' byproduct is mostly water vapour, assuming it's pure. Is there a net water loss? How much energy does it take?
In laymans terms please.
Posted by: Mark | 08 December 2006 at 12:02 PM