Synthesis Of Novel Nano-TiO₂Complex Thin Films And Their Application In Dye-sensitized Solar Cells

Nano-TiO2has been widely used in dye sensitized solar cells (DSSC) as theacceptor of photoelectrons from excited dyes, and the morphology, size andconfiguration of TiO2influence the DSSC performance significantly. An ideal TiO2photoanode of DSSC is required to have huge surface area and superior electrontransport. One dimensional TiO2nanotube/nanowire array favors the electrontransport; however, the surface area is reduced largely because of the easy bundling.On the other hand, TiO2film composed of nanoparticles has larger surface area thanthat of nanotube array, but it suffers poor electron transport due to the numerousinterfaces among nanoparticles. In this context, it is a big challenge to developnanostructured TiO2photoanodes with both big surface area and good electrontransport. In this thesis, we report a novel TiO2nanostructure called “wire-in-tube”,which has a huge surface area since the inner nanowires are well separated from theouter nanotubes. We also prepared a novel TiO2/CNTs (or graphene) complexphotoanode by planting continuous CNTs (or graphene) network into the TiO2filmcomposed of nanoparticles so as to enhance the electron transport property whilekeeping the large surface area. The performance of DSSC fabricated usingabove-mentioned photoanodes is significantly improved.The TiO2nano “wire-in-tube” array was prepared by soaking the pre-anodizedTiO2nanotube array into KOH and HCl alternately. The obtained TiO2nanowires arehighly separated from each other owing to the protection of surrounding TiO2nanotube walls. In this study, we systematically investigated the effects of variousparameters including anodization conditions (voltage, current, etc), concentration ofKOH and HCl, and the time of KOH(HCl) treatment on the configuration of TiO2nano “wire-in-tube” array. Moreover, freeze drying and electrochemical impedancespectroscopy (EIS) were employed to study the formation mechanism of TiO2nanowires. The performance of DSSC using TiO2nano “wire-in-tube” array as thephotoanode was also evaluated. The main results are as follows:(1) The TiO2nano“wire-in-tube” array is formed in two steps, i.e., alkali treatment firstly destroys theinner layer of the tube wall to form TiO2nanoclusters inside the tubes, then theseclusters assemble to form nanowires through dehydration of the hydroxyl group onthe surface of clusters catalyzed by protons;(2) The pre-formation of TiO2nanoclusters during the alkali treatment is important to the final formation ofnanowires. The time of alkali treatment influences the amount of clusters and the concentration of electrolyte for anodization determines the formation rate of TiO2nanoclusters.(3) Acid treatment plays a crucial role in the assembly of TiO2nanoclusters to form nanowire. During the acid treatment, the–OH radicals on thesurface of nanoclusters react with protons and dehydrate quickly (ca.10min in0.5MHCl) to form nanowires. TiO2nanoclusters including their precursor (inner layer oftube wall) and their derivative (nanowires) could be dissolved by HCl, and thedissolving rate is proportional to the concentration of HCl, allowing us to control thediameter of TiO2nanowires to a large extent by adjusting the concentration of HCl;(4)TiO2nano “wire-in-tube” array exhibits a much better DSSC performance (3.48%conversion efficiency) than TiO2nanotube array (0.92%conversion efficiency) underthe same conditions.The TiO2/CNTs(graphene) complex photoanodes were prepared by a three-stepmethod. Firstly, cracked TiO2film was fabricated on FTO conductive glass bychapping, then CNTs or graphene was filled in the cracks to form a continuousnetwork, and finally the cracks were filled by TiO2to form a complete film. TheCNTs (graphene) network contacts the FTO substrate (electron collector) directly,making the electron transport easier. Various parameters including the content ofethylcellulose, drying process for film formation and chapping, calcination andsoaking were studied to clarify their influence on the crack size. The performance ofTiO2/CNTs(graphene) complex photoanode was also evaluated. The results show that:(1) A sufficient content of ethylcellulose in the TiO2slurry is important to thechapping. By adjusting the content of ethylcellulose, the mean width of TiO2islandsin the cracked film can be achieved as26m, indicating that the longest way forelectrons to transport from the TiO2island to the CNTs(graphene) network is13m,which is very close to the mean free path distance of electrons in TiO2film;(2)Soaking in H2O is crucial to the chapping, and the least soaking time is6h forobtaining a cracked TiO2film;(3) The calcination and drying process influences themorphology of chapping more or less. Chapping can occur in the TiO2filmscalcinated at300~400, and the calcination time for chapping becomes shorter asthe temperature increases;(4) Addition of CNTs, especially graphene, cansignificantly improve the DSSC performance by two times increase in thephoto-to-electrical conversion efficiency.