A low-temperature synthesis has been developed to make single crystals of titanium dioxide that contain pores tens to hundreds of nanometres in size. This opens the way to cheap, highly efficient optoelectronic devices. See Letter p.215
The discovery that nanoparticles of metal-oxide semiconductors can be used to make highly efficient solar-cell devices using low-cost processes has attracted a tremendous research effort worldwide1,2. One of the main challenges is to make semiconductors that have a high surface area, but maintain good charge transport. Nanoparticles have high surface areas, but when used in the composite materials that are required for third-generation solar cells, a large number of interfaces are generated between particles; the more interfaces there are, the greater is the hindrance of electron transport across the nanoparticle network, and the less efficient is the device. In this week's issue, Crossland et al.3 (page 215) report that this problem can be solved if the network is replaced by a porous single crystal of a semiconductorFootnote 1.
Ideal mesoporous single crystals (MSCs) have an ordered atomic lattice that develops around cavities of sizes in the range of tens to hundreds of nanometres. Until now, methods for preparing mesoporous materials — particularly semiconductors — have achieved only porous assemblies of nanometre-scale crystals. By contrast, Crossland and colleagues' technique yields MSCs of titanium dioxide (TiO2) up to 1 micrometre in size, but with high surface areas because of their large porosity.
This is important in solar-cell applications because the high surface area of the MSC maximizes the probability of generating free electrons, and the large crystal size means that electrons travelling through an MSC-based active layer (the photoanode) of a solar cell would cross just 2–5 interfaces. By comparison, electrons travelling through a photoanode containing 20-nanometre-sized particles might have to cross as many as 50 interfaces. The chances of an electron generating a useful current are therefore much higher in the MSC system.
The key to the success of Crossland and colleagues' method is tight control of nucleation (localized formation of tiny crystals that act as 'seeds' for crystallization processes) and of crystal growth. Crossland et al. prepared a 'sacrificial' template structure from close-packed arrays of silica beads, then treated it with titanium tetrachloride to seed it with TiO2 nuclei4. The template was then exposed to a solution of titanium tetrafluoride, using a proven recipe5 for synthesizing TiO2 in its 'anatase' crystal form. Finally, they removed the template using a selective etching process, leaving behind porous crystals of TiO2 (Fig. 1). The networks of pores in the crystals were ordered and interconnected, and the size of the cavities was determined by the size of the beads in the sacrificial template.
Crucially, their approach works at low temperatures: the highest temperatures required are 500 °C, with the main steps requiring 210 °C at most. This should allow the crystals to be integrated into plastic substrates, and minimize the environmental impact and production costs of the material.
The authors also performed a detailed analysis of their synthetic process to study the effects of seed density, temperature and size of the silica spheres on the size and morphology of the resulting crystals (see the Supplementary Information for the paper3). Crossland et al. then provided the first demonstration of the electronic properties of anatase TiO2 MSCs, and of the crystals' photovoltaic properties — their ability to convert light into electrical power. Even though these are only preliminary results, the authors show that MSCs do favour electron transport and are characterized by higher electron mobilities than nanoparticle networks.
Internal surfaces of MSCs probably act as electron-scattering centres, thereby complicating electron transport. Further investigations might show how porosity affects such transport. If, as the authors hint, it is possible to explore different template geometries (for example, by making templates from self-assembling polymers), then it may be possible to optimize surface area and charge-carrier transport independently.
An exciting development that will probably follow on from Crossland and co-workers' study is the exploration of methods to fine-tune the electronic and optical properties of TiO2 MSCs. In both cases, it may be beneficial to substitute oxygen atoms in the crystal lattice for other atoms, or to coat the internal surfaces with appropriate (metallic) overlayers. One of the immediate benefits of adding other elements to anatase TiO2 MSCs is that it may then be possible to modify the crystals' light-absorption properties, allowing the material to respond to visible light rather than just to ultraviolet light. This has obvious implications for photovoltaic and catalytic applications of TiO2 nanoparticles.
Crossland et al. show that their MSCs can be coupled with light-harvesting materials (known as light sensitizers) to make solar cells in which all the components were prepared at temperatures below 150 °C. These cells achieved a record light-to-electricity power conversion efficiency of 7.2%, although this was partly as a result of using a state-of-the-art light sensitizer6. The cells' performance is markedly better than that of nanoparticle-based systems that are processed at similarly low temperatures.
The development of low-temperature processes for making optoelectronic devices has already enabled the construction of transparent, flexible and portable gadgets, such as smartphones and tablets, and will facilitate many other important advances. For example, scientists are developing inexpensive solar cells that can be integrated into the fabric of buildings. These devices need to be efficient, light, unobtrusive, self-contained, easily maintained and readily combined with existing architectural elements. MSCs of anatase TiO2 might be ideal components for these cells.
Although the successful synthesis of TiO2 MSCs allows us to speculate on how they might be used in energy-harvesting applications, fundamental questions remain to be answered. How are the internal concave surfaces of the crystals structured? What material defects ensue from the unusual geometry of the crystals? And could this geometry affect the chemistry of TiO2 — especially its chemical response to light — in ways that might improve the material's performance for other applications, such as hydrogen production and the degradation of toxic substances and pollutants in the environment? If Crossland and colleagues' synthetic method can be scaled up to meet potential demands, then the significance and impact of TiO2 MSCs could be tremendous.
*This article and the paper under discussion3 were published online on 6 March 2013.
O'Regan, B. & Grätzel, M. Nature 353, 737–740 (1991).
Grätzel, M. Acc. Chem. Res. 42, 1788–1798 (2009).
Crossland, E. J. W. et al. Nature 495, 215–219 (2013).
O'Regan, B. C., Durrant, J. R., Sommeling, P. M. & Bakker N. J. J. Phys. Chem. C 111, 14001–14010 (2007).
Yang, H. G. et al. Nature 453, 638–641 (2008).
Lee, M. M., Teuscher, J., Miyasaka, T., Murakami, T. N. & Snaith, H. J. Science 338, 643–647 (2012).
About this article
Materials Letters (2019)
Effect of incorporation of multiwalled carbon nanotubes on photodegradation efficiency of mesoporous anatase TiO2 spheres
Materials Chemistry and Physics (2017)
Nature Communications (2017)
Journal of Materials Research (2016)
Experimental and theoretical investigation of a mesoporous KxWO3 material having superior mechanical strength