The in-situ doping of sputtered titanium dioxide thin-films for their efficient visible light sensitization and use for the electro-photocatalytic degradation of pollutants in water.
At a Glance
Section titled āAt a Glanceā| Metadata | Details |
|---|---|
| Publication Date | 2018-01-01 |
| Journal | EspaceINRS (National Institute for Scientific Research (Canada)) |
| Authors | Nazar Delegan |
Abstract
Section titled āAbstractāThe symbols and special characters used in the original abstract could not be transcribed due \nto technical problems. Please use the PDF version to read the abstract.The purpose of this work is to develop photoactive TiO2 films for the electro-photocatalytic degradation \nof water pollutants under sunlight. Radio frequency magnetron sputtering (RF-MS) was \nused to synthesize in-situ N-doped TiO2:N films in order to shift their photoactivity from the UV \nto visible light domain. Careful tuning of the nitrogen-to-argon mass flow rates during sputterdeposition \nallowed the growth of TiO2:N films with a wide range of N-dopant contents ranging from \n0 at.% to 13 at.%. These studies indicated that most of the N-doping atoms were in the desired \nsubstitutional states (i.e. NO). UV-Vis absorption and spectroscopic ellipsometry measurements \nwere used to quantify the bandgap (Eg) variation of the TiO2:N thin films as a function of their N \ncontent. This was used to point out the existence of an optimal nitrogen loading level that reduces \nthe bandgap from 3.2 eV for TiO2 to 2.2 eV for ~3.5 at.% nitrogen doped TiO2:N. This gain in \nvisible-light photosensitivity was assessed for the 1.5AM light driven electro-photocatalytic (EPC) \ndegradation of an emerging pollutant, chlortetracycline (CTC). Thus, TiO2:N photoanodes have \nbeen shown to achieve CTC degradation efficiencies of up to 98% (within 2 h treatments under \nsimulated solar light). Moreover, the EPC performance of the TiO2:N was shown to be directly correlated \nto their optoelectronic properties, sharing a common optimal point between Eg shrinkage and \nEPC degradation efficiency. However, a more exhaustive analysis revealed that the increase in EPC \nperformance was disproportional with the photosensitization, with many photogenerated charges \nseemingly lost. Further studies revealed that nitrogen doping was synonymous with the formation of \noxygen vacancy (VO) defects that can reduce the per-photon efficiency of the material by acting as \ncharge recombination centers. In order to circumvent this limitation, tungsten doped TiO2:W thin \nfilms were developed, as they were theoretically proposed to increase VO defect formation energy. \nTo this end, RF-MS deposition was used to fabricate undoped TiO2, oxygen deficient (TiO2-x), \nand tungsten doped (TiO2:W) films with varying dopant levels. The compositional analysis of the \nW-dopant bonding states revealed the presence of substitutional WVI (W00 \nTi) and WIV (WĆ \nTi) type \ndopants with the total concentration in from 0 at.% to 10 at.%. Additionally, XPS studies revealed a \nsignificant recovery of oxygen stoichiometry upon small W incorporations. On the other hand, high \nfrequency spectroscopy measurements (HF-DS) confirmed an optimal tungsten doping of ~2.5 at.% \nassociated with the lowest εā contribution from the 2 TiIII V00 \nO defect pair (two orders of magnitude \nreduction of the VO dielectric contribution as compared to TiO2-x). Consequently, this reduction \nin VO was exploited by integrating the optimally doped TiO2:W films as photoanodes in visiblelight- \ndriven electro-photocatalytic degradation of atrazine (ATZ), another emerging pollutant. The \npseudo-first order degradation kinetic constants were shown to increase from 0.027 minā1 for TiO2 \nand VO-doped TiO2-x to 0.053 minā1 for the optimally doped TiO2:W photoanodes as a direct result \nof the reduction in VO. Interestingly, it was revealed that optimally doped TiO2:W photoanodes \nperformed on par with the visible-light photosensitized TiO2:N ones. This confirmed that while \nTiO2:N had higher EPC performance due to an increased amount of usable photons, TiO2:W had \nbetter photocharge transport properties, resulting in higher relative per-photon efficiency. In light of these results, and guided by theoretical models, RF-MS was used to synthesize acceptor-donor \npassivated, in-situ WN-codoped TiO2:WN thin films. Thus, by varying the reactive RF-MS deposition \nparameters, we were able to tune the in-situ incorporation of both N and W dopants in the \nTiO2:WN films over a wide concentration range. The objective of the co-doping was twofold: (i) \nnarrow the bandgap of the TiO2:WN films through N-doping to extend their photosensitivity as far \nas possible in the visible, and (ii) passivate the N-doping induced VO defect centers through appropriate \nWN-codoping. Systematic analysis by means of XPS and XRD techniques revealed that both \nW and N dopants were mostly of substitutional nature. Nitrogen doping was found to be the key \ncomponent in narrowing the optical-band-gap down to 2.3 eV for both TiO2:N & TiO2:WN. Most \nimportantly, XPS analysis hinted that the codoping approach greatly reduced the density of VO in \nthe TiO2:WN films as compared to TiO2:N ones. This reduction in defects translated into improved \ncrystalline structure, and increased dopant solubility. The suppression of VO via the acceptor-donor \npassivating approach was directly confirmed by HF-DS measurements showing a marked reduction \nin the density of 2 TiIII V00 \nO defect pairs with the codoping of tungsten and nitrogen as compared \nto N-monodoped samples. This defect reduction was shown to increase photocharge characteristic \nlifetimes using visible light flash photolysis time-resolved microwave conductivity measurements \n(FP-TRMC). Photocharge lifetime analysis indicated the presence of three distinct decay processes: \ncharge trapping, recombination, and surface reactions. These characteristic lifetimes of the codoped \nTiO2:WN films (i.e. 0.08 μs, 0.75 μs, and 11.5 μs, respectively) were found to be about 2.5 times \nlonger than those of their nitrogen monodoped TiO2:N counterparts (i.e. 0.03 μs, 0.35 μs, and \n6.8 μs). This quantitatively confirms the effective passivation of the WN-codoping approach developed \nhere. Finally, the developed TiO2:WNās practicality was confirmed by integrating them as \nphotoanodes for the visible-light driven EPC decontamination of ATZ. A significant increase in the \ndegradation kinetics, resulting in up to a four-fold increase in the pseudo-first order degradation \nconstant k for the optimally codoped TiO2:WN photoanodes (k = 0.106 minā1), in comparison \nwith the undoped TiO2-x & TiO2 ones (both having a k = 0.026 minā1). Most importantly, the \noptimal TiO2:WN photoanodes showed a twofold increase in degradation kinetics as compared to \nTiO2:W (k = 0.057 minā1) & TiO2:N (k = 0.047 minā1) as a direct consequence of both increased \nphotocharge lifetimes and visible light photosensitivity.
Tech Support
Section titled āTech SupportāOriginal Source
Section titled āOriginal Sourceā- DOI: None