Quantum dots boost solar cell efficiencies

Scientists in the UK and US have shown how to increase photovoltaic efficiencies by attaching nanocrystal quantum dots to patterned semiconductor layers. The approach exploits the phenomenon of non-radiative energy transfer and could, say the researchers, lead to a new generation of more efficient solar cells.

Semiconductor solar cells work by using the energy of incoming photons to raise electrons from the semiconductor s valence band to its conduction band. A potential barrier formed at the junction between p-type and n-type regions of the semiconductor forces the pairs to split up, thereby producing a current.

A solar cell s performance is measured by its efficiency; in other words how much electrical power it generates for a given incident solar power. Cells consisting of a single p–n junction that are made from bulk semiconductor have a maximum theoretical efficiency of 31 percent — and the best performing and affordable of such commercial devices are about 18 percent efficient. Carrier multiplication

One way in which scientists are trying to overcome this limit is to make cells from billions of tiny pieces of semiconductor known as quantum dots, rather than one large piece of semiconductor, because these can harness light more effectively and can also create multiple carriers from each incoming photon — a process known as carrier multiplication.

Unfortunately, carriers in quantum dots are not as mobile as in bulk semiconductors and are usually trapped in crystal impurities. In addition, immobile carriers are attracted by neighbouring carriers of opposite charge and by coupling together they annihilate and emit a photon in exactly the reverse process that created the carriers in the first place.

Pavlos Lagoudakis of Southampton University and colleagues say that they can overcome these problems by combining the light-absorption ability of quantum dots with the current-generating capacity of a bulk semiconductor.

To demonstrate this they etched an array of rectangular channels some 500 nm wide into a layered semiconductor structure. The structure comprised a multiple-quantum-well (MQW) layer sandwiched between a p-type layer and an n-type layer. The MQW itself comprised 20 layers of gallium arsenide, each about 7.5 nm thick.

They then deposited a solution of cadmium selenide quantum dots, each just a few nanometres across, onto the structure. Inspired by photosynthesis

The idea, says Lagoudakis, is to take advantage of the “non-radiative energy transfer” used in photosynthesis. The photo-generated carriers within the quantum dots, which are confined within the etched channels, are close enough to the quantum wells that they can exchange energy via a dipole–dipole interaction. “Appropriate engineering of the hybrid device allows for the coupling of the electronic properties of the different components in a way that we get the best properties from each system,” he adds.

To prove that their device was enhancing current output via non-radiative transfer, the researchers also deposited quantum dots onto a substrate without channels. They reasoned that this unpatterned device would not support non-radiative transfer because the photo-generated carriers would be too far apart to interact via dipole–dipole interaction, and that it would therefore produce a much smaller current for a given light input than the patterned device. This is what they found: the patterned device, they report, was six times more efficient than the unpatterned one.

Lagoudakis and colleagues are now designing devices that can combine this feature of non-radiative energy transfer with carrier multiplication by appropriate engineering of the p–n junction and choice of materials for the quantum dots. Such devices, he believes, will exceed the 31 percent efficiency limit. He admits that these devices would cost more than existing silicon solar cells because the molecular beam epitaxy technique used to make the layered structure is expensive. However, he maintains that because the biggest cost in manufacturing solar cells is in fact associated with housing the device, an improvement in efficiency, which would reduce the “active area” of the device, could lead to cheaper solar cells overall. Thinner and cheaper?

Solar cell developer Martin Green of the University of New South Wales in Australia believes that the research by Lagoudakis team is interesting because cells made this way may prove to be thinner and cheaper than traditional cells. But he cautions that the group may not have chosen the ideal reference device to demonstrate increased photocurrent. He says it would have been better to have used as a reference an identical device but with no quantum dots attached as this should also be more efficient than the unpatterned device.