Nanotubular belles

A breakthrough development in photovoltaic (PV) cell nanotechnology at the US Department of Energy's Argonne National Laboratory (ANL) has scientists looking forward to the possibility of manufacturing solar panels that generate electricity more cheaply and more efficiently than the present crop of conventional crystalline silicon (cSi) or cadmium telluride PV cells.

BRISTOL: ANL has been working to refine the design of a so-called hybrid solar cell - in which less-expensive organic and inorganic semiconducting materials are combined to create electricity - that would solve many of the shortcomings hampering commercialisation of current hybrid technologies.

"The potential of solar power will always be limited until its levelised cost of energy is equivalent to, say, coal," says ANL assistant scientist Seth Darling. "So the challenge now - given that, generally speaking, the low-cost materials and processes are too inefficient in terms of how well they convert sunlight into electricity - is how to improve efficiency.

"Hybrid organic-inorganic materials are very promising as future alternatives to traditional technologies such as crystal silicon and cadmium telluride, if we can improve their performance." As Darling explains: "When silicon absorbs light, you immediately get charges: an electron and its partner charge, a 'hole' - but with organics, you get this problem where the electron and the hole, which have opposite charges, are stuck together, and it takes energy to pull them apart."

Research over the past decade that explored the use of wide band-gap nanomaterials such as titanium dioxide (TiO 2 ) primed with light-absorbing dye molecules, for instance, has led to cells with energy-conversion efficiencies of better than 10%. But take-up has been hindered by the need for corrosive liquid electrolytes in the process, which meant the technology relied on elaborate and expensive sealing techniques.

Pairing an organic conjugated polymer, which absorbs light as the electron-donor and transport 'holes', with inorganic materials, which serve as the acceptor and electron-transporter, has proved promising, if very tricky.

When semiconductors are placed too far apart, the electron-hole pair, known as an 'exciton', dies before creating a charge; packed too closely together, the separate changes won't escape the cell. Getting synthesised polymer into the nanotubes - each 1/10,000th as wide as a human hair - requires no less deft a touch. Wet-processing deposition techniques such as spin-coating, doctor-blading and screen- printing, according to Darling, all share the same problem, which he describes as being similar to "trying to stuff wet spaghetti into a table full of tiny holes". "Titanium dioxide is a fantastic electron-acceptor - it really wants those electrons - and it is a relatively low-cost, scalable material," he states. "And titanium dioxide tubes are particularly good because they offer clear pathways for the electrons to get out. So, ideally, we would just fill these tubes with polymer, which will absorb the visible light and donate the electrons to the titanium dioxide, and we are all set," says Darling. "The problem is that it is very hard to get polymer into these very small tubes." Polymer chains also bend and twist in the filling process, leading to 'dead space' in the tubes and, by knock-on effect, poor conductivity and energy-conversion efficiency.

Now, ANL has taken up the baton in its recent lab work by devising a means of 'growing' organic conjugated polymer inside inorganic TiO 2 nanotubes, themselves grown, row upon row, on a film of titanium in an electrochemical bath. Instead of trying to force the polymer into the tubes, Darling and his colleagues have taken a photochemical tack where the tubes are filled with a polymer precursor - a single molecule that can be converted into a polymer chain by a chemical process - and spend the night bathed in ultraviolet (UV) light to allow the polymers to grow inside.

"Using this technique, your filling becomes much better because the polymer chains grow straight and the polymer doesn't shy away from the tubes like it does normally (because it channels through titanium dioxide), so it forms a better interface," he says.

ANL's solid state - that is, no liquid electrolyte - hybrid-cell technology has shown that the tubes are able to absorb light at wavelengths that neither superconductor material can capture alone.

Early tests have shown that nanotube prototypes filled with 'home-grown' polymer - which Darling stresses have not been optimised "by any stretch" for efficiency - produce around 10 times more electrical current than those packed with pre-grown polymer.

"In terms of scaleability, you can scale this technology to your heart's content," offers Darling. "The process by which we grow the tubes is a quick, solution-based electrochemical process that can be grown to almost any scale. And then the polymerisation is, likewise, infinitely scalable, as it only needs polymer precursors and a big UV lamp." Being solution-based, the ANL process will be typically much cheaper than growing single cSi crystals such as are used in high-efficiency solar cells. "There are many, many axes upon which one could optimise the prototype devices that have been made," he says, noting that ANL is looking at alternate polymers that absorb more of the solar spectrum, as well as other wide-band metal oxides that are "energetically better matched" to the various polymers that might be used in the future.

"We are looking to the long run. If we are to meet the targets for terawatts of solar power required by, say, 2050, we need as many ways of getting photovoltaic energy out that are low-cost - and this is one of the pathways towards doing that," reasons Darling.