Double-junction, or tandem, solar cells perform better than their single-junction counterparts with power conversion efficiencies of around 42% compared with just over 30%. As their name suggests, these devices contain two junctions (rather than just one), each of which absorb light of different wavelengths from the Sun. For example, the junctions at the front of the cell can be made of a wider bandgap material that harvests high-energy photons while more abundant lower-energy photons can be collected by a smaller-bandgap material situated at the back of the cell.

The critical ingredient in these cells is an efficient transparent intermediate layer sandwiched between the junctions. This layer allows photogenerated electrons and holes from neighbouring junctions to meet and efficiently recombine. In conventional compound semiconductor multijunction solar cells, tunnel junctions are employed for this purpose: highly doped p- and n-type materials produce an extremely thin junction just several nanometres thick in which the valance band on the p-side is aligned – in terms of energy – with the n-side. The depletion region is thin enough so that carriers can tunnel from one side to the other. In organic photovoltaics, electrons and holes recombine in metal nanoparticles inserted between the electron-transport and hole-transport layers.

Graded recombination layer

Ted Sargent’s group have moved away from these traditional approaches and recently proposed tandem solar cells with a new type of recombination layer based on colloidal quantum dots that were tuned – through quantum size effects – to have bandgaps that absorbed in a wide range of the electromagnetic spectrum. The researchers dubbed this layer a “graded recombination layer (GRL)”.

The team has now put forward a clear set of generalized design rules for making such an ideal GRL and says that multiple graded intervening layers based on low-doped oxides are the best options for achieving low optical losses in solar cell devices. “We describe our concept as a ‘rainbow within a rainbow’,” explained Sargent, “because the multi-spectral layers of the multijunction cell are themselves spanned using a ‘spectrum’ of electronic material that provides a graduated progression among the constituent subcells.”

Transparent conductive oxides

The Toronto team’s strategy relies on using a wide range of readily available transparent conductive layer n-type oxide materials, such as TiO2, ZnO, AZO and ITO. These oxides have work functions that span large energy barriers, from 1 to 1.6 eV. “Such a GRL design would enable electrons to survive these energy barriers and to flow across the graded intervening layers with acceptable resistance in a photovoltaic device context,” team member Ghada Koleilat told

The researchers, who present their design rules in Nano Letters, are now busy further improving the performance of each constituent junction in their colloidal quantum dot solar cells. “We would also like to find a simpler way of manufacturing each layer in the materials stack because this would help us to make a triple junction quantum-tuned solar cell more easily,” said Sargent.