Copper nanoflower illustration. Credit: Virgil Andrei.
Tiny copper ‘nanoflowers’ have been attached to an artificial leaf to produce clean fuels and chemicals.
The researchers, from the University of Cambridge and the University of California, Berkeley, developed a practical way to make hydrocarbons – molecules made of carbon and hydrogen – powered solely by the sun. The device they developed combines a light absorbing ‘leaf’ made from a high-efficiency solar cell material called perovskite, with a copper nanoflower catalyst, to convert carbon dioxide into useful molecules.
Here ScienceVega talks to Cambridge’s Dr Virgil Andrei to get a better idea of the science…
ScienceVega: What work had been undertaken on artificial leaves prior to your new research?
VA: Artificial leaves as a field began developing a couple of decades ago, but there have always been challenges. One issue has been that products obtained using artificial leaves could only split water into oxygen and hydrogen. To solve this challenge, we started introducing a perovskite light absorber, which is a relatively new and promising material. The perovskite device can give you a high photovoltage, which means you can perform more challenging reactions such as carbon dioxide reduction to small molecules like carbon monoxide or formic acid, which is what we have been doing in the past.
These small molecules can be used as industrial intermediates for liquid fuel production, but we also wanted to directly obtain value-added products. These complex molecules have a higher energy density and can be used for industrial or fuel purposes.
We’re looking at ways to find the right catalyst and design the right system in order to, for instance, produce hydrocarbons like ethane and ethylene directly from carbon dioxide.
This is what I was doing during my exchange visit to the University of California, Berkeley. I am based in Cambridge, but I used the Winton-Kavli visiting fellowship to work in the group of Professor Peidong Yang for six months and look at value-added chemical production.
Over there, I took advantage of their expertise in catalysis and firstly combined the copper nanoflower catalysts with our own artificial leaves, consisting of perovskite and bismuth vanadate light absorbers. You need two light absorbers because bismuth vanadate will oxidise water to oxygen and the perovskite will use the rest of the light to produce hydrogen or reduce carbon dioxide, so it will do reductive reactions.
With the copper nanoflowers interfaced to these artificial leaves, we could get some ethane and ethylene, but only very small amounts because oxygen is still quite a thermodynamically challenging reaction to perform.
That is why we had to look at alternatives to gain some extra voltage and overcome this thermodynamic barrier. We turned our attention to the oxidative reaction on the other side, and it just so happens that organic substrates are much easier to oxidise than water. For instance, if you take glycerol, which is a common industrial by-product, you can easily oxidise it into glycerate, lactate or acetate. These are also value-added chemicals that can be used in the fine chemicals, pharmaceuticals or cosmetics industries.
In this way, we replaced a product which is not really valuable like oxygen with something that you can gain economic value by producing it. We still have the perovskite and the copper flowers on the reduction side, but now we have these silicon nanowires for glycerol oxidation. They work quite well because there is a higher surface area, so you can perform your reaction at higher rates. They are also really good light absorbers, so you can drive quite high currents and therefore obtain a high product rate.
By combining these two systems together, you are able to perform both oxidative and reductive reactions at higher rates, so you can convert water and carbon dioxide into hydrocarbons on one side and glycerol and water into useful chemicals on the other side of the reactor.
Artificial nanoflowers. Credit: Virgil Andrei.
ScienceVega: Where did the idea for this specific research come from?
VA: Sunlight is our most abundant renewable energy source, so plenty of progress has been made on developing efficient solar cells. Although you can store solar energy with batteries, these have a very low energy density; about 50 times lower than that of chemical fuels. This means you can’t use batteries for everything: you can’t use them for aircraft, or for heavy shipping, so you still need those chemical fuels.
In nature, plants store energy in the form of sugars by using water and carbon dioxide.
In our field we wanted to take inspiration from nature for our design, using two light absorbers and two catalysts, but simplifying the whole system and utilising synthetic materials that would help us increase the solar to chemical conversion and target useful products. For example, instead of sugars, we would be making hydrogen or ethylene.
Copper nanoflowers for carbon dioxide reduction to hydrocarbons. Credit: Virgil Andrei.
ScienceVega: In that case, why does the copper nanoflower catalyst enable more complex hydrocarbons than previously?
VA: Copper is the only element that can perform carbon dioxide reduction to these complex hydrocarbons. This depends on the binding of carbon dioxide and the binding of intermediate species on the surface. For copper, the binding is neither too strong nor too weak, which allows the carbon dioxide to react on the surface and the carbon atoms to couple. Copper is the only metal that can do this, but then it also depends a lot on how you nanostructure this copper; if you have a flat surface, you will still mostly produce hydrogen, but if you start nanostructuring it, you can steer selectivity towards different products.
These nanoflower clusters are made out of thin lamella plates that give a lot of porosity. This means, carbon dioxide can react at the surface and products linger in the pores for a length of time, so ethylene can also be reduced further to ethane.
Cu-tree-flower. Credit Virgil Andrei.
ScienceVega: Can you give me an idea of the amounts of energy required for these processes?
VA: We don’t actually need to input external energy. Our artificial leaves are bias free, so in principle you only need water, carbon dioxide, glycerol, and sunlight. The sunlight energy is enough to drive the reaction, but that happens because we have the two light absorbers which provide us an additional voltage. From perovskite you get around one volt and from silicon, you can get around 0.5 volts. With 1.5 V provided by your light absorbers you have enough energy to drive the whole process unassisted.
ScienceVega: How do the silicone nanowires overcome the limits of water splitting?
VA: Silicon nanowires give you the extra voltage needed to produce this reaction at higher rates, but the type of oxidation reaction also matters.
For oxygen evolution, you need to input at least 1.23 V versus the reversible hydrogen electrode. For these organic oxidations you need much less voltage because the compounds are easier to oxidise than water. You need around 0.4 V to 0.5 V and part of that voltage is also supplied by your silicon nanowires. In the end, you need very small energy inputs to drive this reaction on the oxidation side.
But the overall process also depends on the reduction side, because you do need around 0.7 V to 0.8 V to drive the hydrocarbon formation. That’s why you need the perovskite.
Silicon nanowires for organic oxidations. Credit: Virgil Andrei.
ScienceVega: This is the production of clean fuels – what would it take to scale this operation up to an industrial level?
VA: In this study we showed a proof of principle for how to combine hydrocarbon synthesis with glycerol oxidation. What we would need to look at is how best to scale this up. In the past we have probed the scale-up of artificial leaves, but it was only for simple products like water splitting or syngas (carbon monoxide plus hydrogen) production.
We have shown previously that you can deposit light absorbers onto thin flexible substrates and fabricate them on a slightly larger scale. With our existing lab equipment, we could make 100 cm2 panels, which were flexible and around 15 times lighter than our other small scale prototypes. In principle, you could use similar roll-to-roll or spray coating techniques to produce those on a larger scale for water splitting and syngas production.
In this case we would need to think about the best way to scale up the silicon nanowires, which can be produced on a wafer scale. The perovskite component can be also scaled-up with the technology available at the moment, but we need to think more about how to integrate the two systems.
Design of an artificial leaf. Credit Virgil Andrei.
ScienceVega: What is your next research in this area?
VA: Over the years we have been working on scaling-up these artificial leaves and improving their stability, because right now we have systems which typically run for a few hours or days. We do have a system which worked for a few weeks, but we need to develop these systems to the point where they operate for a year whilst also minimising the amount of material utilised. The flexible devices can be in principle mass-produced, so even if they don’t last for ten years you can replace them cheaply.
It would be interesting to also look at other types of reactions, because we know that you can oxidise glycerol, this industrial by-product, to make chemicals. In principle you could use the same system to perform more complex organic reactions, and this is where it could become very useful for fine chemical synthesis or the pharmaceutical industry. In this type of system, you can selectively convert a complex molecule to another valuable molecule just by using sunlight.
The link to the original paper can be found here.
Portrait credit: Virgil Andrei.
