Theoretical Study Of The Electronic Structures And Related Properties Of Plasmonic Photocatalyst Based On AgCl And TiO₂

Among the many novel photocatalytic systems developed in very recent years, plasmonic photocatalyst possesses great potential for applications and is one of the most intensively investigated systems owing to its high solar energy utilization efficiency. This special photocatalytic system is generally constructed by noble metal nanoparticles and conventional semiconductor photocatalyst, where the noble metal nanoparticles strongly absorb solar light through the localized surface plasmon resonance （LSPR）. However, as a relatively new area, the detailed photocatalytic mechanism of plasmonic photocatalyst and the factors affecting the overall photocatalytic performance are not well resolved. For example, because Ag@AgX（X = Cl, Br, I）, one of the most representative plasmonic photocatalyst, decomposes easily under high-energy particle irradiation, it is experimentally impossible to investigate the detailed physical and chemical processes that occur during photocatalytic reactions. Thus the understanding of Ag@AgX is very limited. In addition, as a commonly used semiconductor in plasmonic photocatalyst, the studies of TiO2 as well as the interface properties between TiO2 and LSPR metals are tremendous. However, compared with recent experimental results, the theoretical understanding seems to be inadequate. These include the effects of nitrogen （N） and hydrogen （H） doping on the electronic structures and related properties of TiO2, the mechanism about how fluorine （F） adsorption modulates the surface stabilities of TiO2, the relative photo-oxidation and photo-reduction abilities of some typical surfaces of TiO2 as well as the effect of structure distortion on them, the factors that can undermine and modulate the interface carrier transfer in Au@TiO2 composite.Aiming at solving the problems mentioned above, in this dissertation we systematically studied the electronic structures, surface stabilities and related properties of pristine and doped AgCl and TiO2, the LSPR properties of Ag nanoparticles, and the interfacial properties of metal-semiconductor composites. Furthermore, the underlying physical and chemical mechanisms were fully investigated and discussed. The dissertation is divided into five chapters. In the first chapter, we briefly introduced the related physical mechanism, recent research background and progress of plasmonic photocatalyst and the research contents of this dissertation. In the second chapter, we introduced the theoretical basis, methods and computational codes for the first-principles calculations and electromagnetic simulations that have been used. In the third chapter, we presented the results and discussion for the properties of AgCl and Ag@AgCl. In the fourth chapter, we presented the results and discussion for the properties of TiO2 and Au@TiO2. In the last chapter, we summarized the main conclusions and innovations of this dissertation, and provided an outlook for the future of plasmonic photocatalyst. The main work and results are listed as follows:（1） The underlying mechanism and crystal growth dynamics in tuning the crystal morphology and surface exposing of AgCl with NaCl are elucidated. We show that the excess Cl provided by NaCl can readily adsorb on the AgCl surfaces during crystal growth, and the binding strength between Cl and surface changes significantly for different surfaces, which changes their relative stabilities. Therefore, one can modulate the crystal morphology and surface exposing of AgCl by adjusting the initial NaCl concentration. However, Cl adsorption on all surfaces decreases their oxidation potential, thus one should eliminate the surface adsorbates as much as possible to reach higher photocatalytic activities. This study provides some valuable guides for preparing Ag@AgCl with higher photocatalytic performance.（2） We investigate the adsorption properties and electronic structures of Ag clusters on AgCl （100） surface. The results show that because of the stronger binding between Ag atoms than between Ag and Cl atoms, the adsorption structure that can maximize the number of Ag-Ag bonds formed between Ag cluster and AgCl surface is the most stable for anyone of the investigated Ag clusters. This finding is very instructive for determining the adsorption structures of much larger Ag clusters on AgCl surface and similar systems. In addition, in comparison with planar adsorption, it is energetically more stable when Ag atoms adsorb on the AgCl surface in the form of cluster. Moreover, small Ag clusters tend to combine with each other to form larger ones. The adsorption of Ag cluster on AgCl （100） surface decreases its work function, which further affects its photo-redox potential. This study provides some basic insights into the physical and chemical properties of Ag clusters on AgCl surface.（3） We explore how the visible light absorbed by Ag nanoparticles （NPs） is converted to electrons and holes in plasmonic photocatalyst Ag@AgCl. The results suggest that the LSPR absorption of Ag NPs induces significant electric- field enhancement in the nearby AgCl, which strongly increases the optical transitions of AgCl involving the defect states, thus generating electrons and holes for photocatalytic reactions. This theoretical suggestion is further experimentally verified by preparing Ag@AgCl samples possessing different degrees of bulk and surface defects in AgCl and subsequently by carrying out photocatalytic experiments with them. This work deepens our understanding of the photocatalytic mechanism in plasmonic photocatalyst and provides theoretical guidelines for preparing high-efficiency photocatalytic materials.（4） The photocatalytic properties of Ag@AgX are discussed by studying the band structures and carrier effective masses of AgX. Our results show that the electron effective masses of AgX are spatially isotropic and much smaller than that of the commonly used photocatalyst TiO2, which may contribute to the excellent performance of Ag@AgX. However, their hole effective masses are rather large and anisotropic. Furthermore, both the direct bandgap and carrier effective masses alter when the anion in AgX changes from Cl to Br, and then to I. These results explain the observed remarkable dependence of Ag@AgX photocatalytic activities on the surface exposing and anion species of AgX.（5） We re-investigate the electronic structures of N doped TiO2 by proposing a novel electronically coupled N doping model. The results show that conventional N doping models by maximizing the mutual distances between dopants only introduce localized gap states irrespective of the doping concentration, which agrees with previous theoretical results but cannot explain recent experiments. In comparison, the electronically coupled N doping model, which is almost as stable as the conventional doping models, can not only narrow the overall band gap of TiO2 but also decrease the carrier recombination rate. Moreover, when pristine and the N doped TiO2 are combined with each other, perfect type-â…¡-like homojunction is formed, which can further decrease the carrier recombination rate. This study conclusively accounts for the recent experimental results and indicates that the final electronic structure of doping system is very sensitive to the electronic coupling between adjacent dopants, which rationalizes the distinct conclusions about N-doped TiO2 in previous theoretical studies.（6） We advance an effective bonding level method to examine the stabilities of pure and F-adsorbed TiO2 surfaces. The results show that the high stability of pure TiO2 （101） surface mainly benefits from its surface atomic structures and local potential, which strengthen the Ti-O binding in the surface. However, F-adsorption significantly weakens the Ti-O binding in （101） surface but strengthens them in （001） surface, so that （001） becomes more stable than （101） for the F-adsorbed surfaces. On the basis of these results, we further demonstrate that n-type doping in TiO2 decreases the ability of F-adsorption in modulating relative stability of the two surfaces. This study not only provides new insights into the physical and chemical properties of pure and F-adsorbed TiO2 surfaces, but also explains related experiments.（7） On the basis of the polaron model of photo-generated carriers and first-principles calculations, we identify that both the relative photo-oxidation and photo-reduction abilities of anatase TiO2 （100）, （101） and （001） surfaces satisfy the order （100）> （101）> （001）, which rectifies conventional understanding and agrees with recent experiments. Moreover, it is demonstrated that the surface under-coordinated atoms have dual functions in photocatalytic process:（ⅰ） Acting as active adsorption and reaction sites, thus promoting photocatalytic reactions. （ⅱ） Preventing photogenerated electrons from transferring to the outermost surface atomic layer, thus being detrimental to photo-reduction process. These results deepen our knowledge of the physical and chemical properties of under-coordinated atoms in TiO2 surface.（8） On the basis of first-principles calculations we investigate the effect of surface distortion on the separation and transfer of photo-generated carriers in anatase TiO2. The results show that proper distortion in （101） surface can promote the transfer of electrons in the bulk region to the surface, but prevent holes from transferring to it. Therefore, surface distortion helps to separate holes from electrons and reduce carrier recombination rate （which is believed to result in the low activities of （101） surface）. On the other hand, distortion in （001） surface facilitates the transfer of electrons from the subsurface atomic layer （the original electron trapping sites） to the outermost atomic layer （where photocatalytic reactions take place） by eliminating the energy barrier. In practical photocatalytic process, similar surface distortion is induced by surface adsorbates, the present study thus present some insights into the real properties of photo-generated carriers.（9） We explore the factors that can affect the hot electron injection efficiency （HEBE） in plasmonic photocatalyst Au@TiO2 by first-principles calculations and electromagnetic simulations. We find that the occupation of TiO2 surface oxygen vacancies by gold atoms increases the hot electron transfer barrier and expands the space charge region, thus decreasing the HEIE. Moreover, the Au@TiO2 microstructure that is not beneficial to the absorption of light propagating perpendicular to the interface normal can decrease the generation of hot electrons and their transfer to TiCh, thus decreasing the HEIE. These results can provide some guidelines for improving the activities of Au@TiO2.（10） The interfacial carrier transfer properties, which are dominated by the interface Schottky barrier height （SBH）, generally play crucial roles in the performance of metal-semiconductor heterostructures. Herein, we investigate the effect of strain on the interface SBH of heterostructures using Au/TiO2 （001） as a model system. The results demonstrate that strain can effectively modulate the interface SBH and that the n-type SBH can be more effectively tuned than the p-type SBH. Astonishingly, strain modulates the SBH mostly by changing the intrinsic properties of Au and TiO2, whereas the interfacial potential alignment is almost unchanged. These results can be understood on the basis of the general free-electron gas model of typical metals, tight-binding theory and crystal-field theory, which suggests that similar results may be obtained for many other metal-semiconductor heterostructures. Given the commonness and tunability of strain in heterostructures, we anticipate that the modulation of interface SBH with strain presented here can provide an alternative way for realizing more efficient applications of relevant heterostructures.（11） We explore the possibility of introducing LSPR in TiO2 by H doping. The results show that H interstitial doping in TiO2 can reach a very high concentration and its donor level is resonant with the TiO2 conduction band, thus introducing an electron carrier concentration as high as 1022/cm3. Moreover, H atoms can readily occupy the oxygen vacancies that often exist in as-prepared TiO2 and eliminate the Ti3+ ions. These results conclusively explain the recent experiments. Furthermore, on the basis of Drude model, we demonstrate that the H introduced carrier concentration is high enough to induce LSPR effect for TiO₂, and the LSPR is located in the infrared region, making it possible to develop noble-metal-free infrared plasmonic photocatalyst. This study may also have a significant impact in infrared bio-imaging and spectroscopy where infrared LSPR is involved.