Elongated titanates and TiO2 nanostructures have been examined for use as an electrode for dye-sensitized solar cells (DSSC). Figure 5.1a shows the principle of DSSC utilising nanotubular titanates. The photoexcited molecule of dye adsorbed onto the surface of nanotubes injects an electron to the semiconductor, which then diffuses towards the electron sink. The oxidised form of the dye is reduced by iodide ions in the electrolyte solution, and the iodine released is further reduced on the platinum counter electrode.
The potential advantage of titanate nanotubes as an electrode for DSSC is realised by exploiting the phenomenon of improved adsorption of the positively charged dyes from aqueous solution onto the surface of negatively charged titanate nanotubes (compared to TiO2NP).2,3 This enables a compact monolayer of dye to be deposited with a capacity of over 1000 molecules per
RSC Nanoscience & Nanotechnology No. 12
Titanate and Titania Nanotubes: Synthesis, Properties and Applications By Dmitry V. Bavykin and Frank C. Walsh © Dmitry V. Bavykin and Frank C. Walsh 2010 Published by the Royal Society of Chemistry, www.rsc.org
nanotube. Such a dense loading of dye allows the thickness of the light-adsorbing layer of the electrode to be significantly reduced from typical values of several microns, and the consequent reduction in the electron diffusion distance can potentially improve the charge collection. The second advantage of nanotubes is the elongated morphology of these semiconductors, which can provide a direct electron pathway from the point of injection to the electron sink, allowing improved electron transport and charge collection efficiency. This effect is particularly pronounced in TiO2 nanotubular arrays.
The design and manufacture of conventional DSSCs have been optimised for a nanoparticulate TiO2 electrode, which can withstand high-temperature calcination and is a more suitable adsorbent for negatively-charged dye molecules, such as the cis-di(thiocyanate)bis(2,2-bipyridyl-4,4-di-carboxylate) ruthenium(ii) complex. By contrast, titanate nanotubes are more suitable for the adsorption of positively-charged dyes2 and are less stable during calcination at 450 °C (ref. 4,5), which is required for the conventional doctor blade method of electrode manufacture. Early data shows that titanate nanotubes can be readily applied as electrodes in a DSSC; however, no significant benefits were observed when compared with TiO2 nanoparticle-based electrodes.6 In most studies, however, calcination of the electrode at 450 °C was used to remove the polyethyleneglycol binder, meaning that TiO2 anatase nanorods instead of H-TiNT were used in the DSSCs.
An alternative method for immobilising elongated titanates is the direct alkaline hydrothermal synthesis of nanotubes on the surface of titanium metal.7,8 Prior to dye deposition at 500 °C, calcination of the sample was used, resulting in conversion of H-TiNT to TiO2 anatase NR.4 The advantage of the increased dye adsorption on the surface titanate nanotubes was not fully utilised.
in order to benefit from an improved electron transport in elongated nanostructures, it is necessary to assemble nanostructures on the surface of electrode. in most reports, elongated titanates are randomly oriented, which diminishes the advantage of direct transport. More recently, the successful alignment of titanate nanofibres on the surface of titanium under alkaline hydrothermal synthesis at 180 °C has resulted in an improved efficiency of DSSCs compared to the use of P25 TiO2.9
Detailed studies of electron relaxation kinetics in elongated TiO2 anatase nanorods have shown that the electron diffusion coefficient in nanorods is similar to that in P25. However, the electron lifetime was 3 times higher than that in P25, possibly due to a suppression of the recombination between electron and I ions on the surface of nanorods,10 resulting in a higher value of the open circuit potential. Transient studies, using intensity-modulated photo-current spectroscopy, show the improvement in charge collection in following order: P25, titanate nanotubes then anatase nanorods.11 The charge collection ratio, which is defined as a ratio of the recombination time constant to the electron time collection (tr/tc), was found to be ca. 150, 50 and 10 for TiO2 NR, H-TiNT and P25, respectively. These results were obtained for randomly oriented nanotubes and nanorods. The improvement in nanotube alignment on the electrode should offer further improvements in the electron collection efficiency.
The polycrystalline nanoparticles of anatase, obtained from titanate nano-tubes by hydrothermal treatment at 240 °C in the presence of dilute HNO3, have exhibited a slightly higher short circuit current, Isc, but a lower open circuit potential, Voc in DSSCs, compared to P25. This is most likely due to the higher recombination rate of charge carriers in these nanoparticles.12
The typical efficiency of a DSSC using elongated TiO2 NR with standard dye and electrolyte is approximately 7.1% (ref. 10), and the shape of its voltam-metric curve is shown in Figure 5.1 b. As a consequence of the leakage problems associated with a liquid electrolyte DSSC, recent attention has been focussed on solid or gel electrolyte DSSCs. The addition of H-TiNT filler to a poly-ethyleneglycol gel electrolyte up to a 10% of nanotubes has facilitated an improved ionic conductivity in gels.13
The other type of TiO2 nanotubular array,14,15 produced by anodising a Ti surface with a smaller specific surface area, also attracts attention as a candidate material for DSSC electrodes.16 This interest is arises from the ordered structures within the electrode which allow an improvement in electron transport; the efficiency of this cell is claimed to be approximately 4.1%.
The classical Gratzel cell operates with sintered compressed layers of spheroidal TiO2 NP as the electron harvesting material. A several micron thick agglomerate layer contains a high number of grain boundaries that act as recombination sites, reducing current collection efficiency. The ordered layer of aligned nanotubes can potentially show a lower rate of recombination due to the direct transport of electrons, but it is important to optimise the length of the nanotubes.
Figure 5.2 shows the dependence of photocurrent collected from the TiO2 nanotube array electrode on the length of nanotubes. The nanotubes were functionalized with a Ru-complex deposited onto their surface and were illuminated at the 650 nm. It is evident that for thin layers, an increase in tube length results in a linear increase in the photocurrent, indicating the independence of current collection efficiency on nanotube length. When the thickness of the nanotube layer exceeds a certain value, a further increase in the nanotube length leads to a decrease in the photocurrent. Such a drop in photocurrent can be attributed to an increase in the electron diffusion length, resulting in an increase in the recombination rate. In the thick electrode layer, the absorption of light can occur unevenly causing a shadowing of areas having efficient electron transport.
The adsorption capacity of standard Ru-complex dye on the surface of anodic TiO2 nanotubes is similar to that on the surface of TiO2 anatase NP, resulting in relatively low coverage and making it necessary to use a thick layer of TiO2 in order to absorb all incident light. For example,17 the surface coverage of a 6 mm i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i
i i I i i i i i i i i i I i i i i i i i i i I i i i i i i i i i
Nanotube length / ^m
Figure 5.2 A graph showing the photocurrent collected from a Ru-complex dye-sensitised array of TiO2 nanotubes as a function of nanotube length. Illumination at 650 nm. (Data adapted from ref. 15).
film of TiO2 NP with standard N719 dye is ca. 0.5nmolcm"2, resulting in 80% absorption at 535 nm. In contrast, titanate nanotubes are characterised by a higher adsorption capacity, particularly towards negatively-charged dyes.2
The manufacture of DSSCs using TiO2 nanotube arrays can be achieved either by the anodic oxidation of a titanium layer vacuum-deposited onto the surface of a conductive glass forming a transparent conductive glass/TiO2 nanotube electrode, or by the anodic oxidation of titanium foil to form nontransparent electrodes, which are illuminated from the reverse side during operation.14 The TiO2 nanotubes obtained by template hydrolysis were also studied in a DSSC recently, with a reported efficiency of ca. 8.4%.18
It is useful to review the advantages and limitations of titanate and TiO2 nanotubes when used in DSSC applications. These nanostructures can provide a high capacity for dye adsorption onto their surfaces, potentially allowing a reduction in the thickness of the light harvesting layer and a reduced electron diffusion time. However, it has proved difficult to prepare films of titanate nanotubes with a high degree of alignment, which has resulted in reduced current collection efficiency due to ineffective electron transport in the light harvesting layer. On the other hand, TiO2 nanotube arrays are characterised by high nanotube alignment and relatively high current collection efficiencies. A low surface area of nanotubes and a poor dye adsorption capacity can result in an increased thickness of the light harvesting layer, which can lower the current collection efficiency.
Was this article helpful?