TiO2 gas phase photocatalysis: from morphological design to plasmonic enhancement

Datum: 5 mei 2014

Locatie: UAntwerp - Stadscampus - Hof van Liere - Francis de Tassiszaal - Prinsstraat 13 - 2000 Antwerp

Tijdstip: 15.30 uur

Organisatie / co-organisatie: Department of Bioscience Engineering

Promovendus: Sammy Verbruggen

Promotor: Prof Silvia Lenaerts, Johan Martens

Korte beschrijving: PhD defense Sammy Verbruggen - Faculty of Science, Department of Bioscience Engineering


The past decades, photocatalysis has emerged as a powerful technology for pollution abatement and energy applications, but its use in gaseous environment is less obvious than in water borne applications. Therefore this thesis specifically aims at understanding and improving TiO2-based photocatalysis in all its facets towards gas phase processes.

In first instance a suitable photoreactor is developed and validated. A simple suspension-based coating strategy enables to immobilize powderous photocatalytic materials onto glass bead supports that are packed around a light source in a cylindrical glass reactor tube. The presented design offers several advantages such as good catalyst immobilization, efficient light utilization, intimate contact with gaseous pollutants and a catalyst weight gain by a factor of 25 compared to self-supporting pellets. This glass bead photoreactor is used in a comparative study on TiO2-based photocatalytic materials, in which both technological and economical parameters are considered. By performing a cost effectiveness analysis, PC500 (Cristal Global) is determined to be a very active and cost efficient photocatalyst in the degradation of gaseous acetaldehyde and even outperforms the typical benchmark Aeroxide P25 catalyst (Evonik).

In the next part of the work, fundamental insight is gathered in the driving factors for gas phase photocatalytic reactions by investigating the different nature of the P25 and PC500 catalysts. The high surface area and sophisticated yet accessible pore system of PC500 seem to dominate over the superior electronic efficiency of P25. These findings are supported by photocatalytic gas phase experiments, photocurrent measurements, N2-sorption data, X-ray diffraction patterns, UV-VIS spectroscopy, thermo-gravimetric analysis and acetaldehyde adsorption measurements, amongst others.

The dominating effect of surface area on gas phase photocatalytic activity is exploited in the development of well-immobilized, spacious TiO2 thin films. These films are prepared by depositing a thin, conformal layer of TiO2 onto sacrificial carbonaceous templates by means of atomic layer deposition. The carbonaceous templates are either carbon nanosheets or multi-walled carbon nanotubes, both grown on silicon wafers. After application of a calcination step, the sacrificial template is removed, TiO2 is crystallized into the anatase phase and the as-deposited continuous TiO2 layer has transformed into an interconnected network of nanoparticles. Electron microscopy images show that the overall appearance of the films is entirely commensurate to the original template morphologies. The employed strategy allows to fabricate spacious thin films with surface area enhancement factors of up to 260 with regard to a dense, flat TiO2 film. Thus obtained thin films exhibit superior photocatalytic activity compared to a reference film consisting of the photoactive PC500 catalyst. For the testing of these thin films another type of photoreactor was developed. The use of the constructed single pass flow through, slit-shaped, flat bed photoreactor is visualized by computational fluid dynamic simulations.

As an intermission the use of surface photovoltage measurements is discussed in relation to photocatalytic activity. The research question is simply whether surface photovoltage measurements can be used as a quick screening tool for assessing photocatalytic performance. The answer is less straightforward. Based on several practical case studies, the trend between photovoltage and photoactivity is investigated using a custom-made photovoltage set-up. For TiO2-based materials with variable anatase/rutile ratios, catalytic activity is determined to be proportional to photovoltage readings. In contrast, in the case of variable amounts of silver nanoparticles deposited on the TiO2 surface that act as electron traps under UV illumination, an inverse relation is observed. Furthermore, a significant effect of the catalyst humidity is detected, but the most important conclusion to be drawn from this study is that surface photovoltage measurements can only probe electronic properties. The vast impact of morphological parameters on photocatalytic activity is not accounted for by this technique. Therefore care should be taken when interpreting surface photovoltage data in relation to photocatalysis.

Finally, successful attempts have been made to extend the TiO2 photoactivity window towards the visible light region of the spectrum. This is achieved by exploiting surface plasmon resonance effects of gold-silver alloy nanoparticles. In the first part of the research a theoretical, predictive model is established that enables to predict the plasmon resonance wavelength of such alloy nanoparticles, based on the combined effect of particle size and alloy composition. The model is developed using theoretical simulations of extinction spectra based on Mie theory and dielectric data from literature. The proposed model indicates that mainly the alloy composition determines the resonance wavelength, while particle size is of minor importance. In the second part of the investigation, gold-silver alloy nanoparticles are deposited on the TiO2 surface. By merely altering the alloy composition of the deposited nanoparticles, plasmonic photocatalysts can be prepared that display surface plasmon resonance in a 70 nm broad window in the visible light wavelength range that is roughly centered at the maximum intensity wavelength of solar radiation. Thus obtained plasmonic photocatalysts are tested towards their self-cleaning performance in the degradation of a solid layer of stearic acid located at the catalyst-air interface. The highest quantum efficiency is obtained when the resonance wavelength of the plasmonic catalyst exactly matches that of the incident light. This is demonstrated for the case of Au0.3Ag0.7 nanoparticles on TiO2 under 490 nm illumination, provided by a custom-made LED array.

In conclusion, in this work TiO2 gas phase photocatalysis is investigated in its broadest sense. Different aspects are discussed ranging from reactor development, over techno-economic analysis, morphological catalyst design, characterization, structure-activity relation, interpretation of photovoltage measurements, to visible light activity using plasmonics. We are hopeful that this altogether leads to better understanding and novel insight into this fascinating research domain.