Effect of Dimensionality and Polymorphism on the properties of ZnO

Abstract

[cat] El treball de recerca desenvolupat en aquesta tesi es centra en ZnO, un dels semiconductors de tipus II-VI amb un ampli ventall d’aplicacions. En les estructures (ZnO)n suportades, s’observa que la presència del suport afecta l’ordre d’estabilitats dels mateixos però de manera molt més dràstica afecta selectivament les estructures bidimensionals (2D) que, a partir d’una certa grandària, en fase gas són menys estables que les tridimensionals (3D). Els càlculs per a la làmina 2D-ZnO aïllada interaccionant amb l’hidrogen proporcionen una forta evidència per a la formació d’un estat d’enllaços multi-centres de baixa energia quan passa a través de l’anell de Zn3O3 de la làmina 2D-ZnO, permetent així de forma relativament fàcil el transport d’hidrogen a través de la làmina. Quan canviem a models amb illes mes grans, observem reconstruccions estructurals a l’interior i sota l’illa formada per una nova capa incompleta. L’interior de les illes triangulars adopta estructura WZ i esta rodejada per vores amb estructures BCT i cantonades amb estructura T1. S’ha observat que aquests models presenten en un millor acord estructural amb les dades experimentals per el cas de les lamines formades per 2.7 ML que no pas respecte als models que assumeixen una estructura purament grafítica o purament WZ. Hem generat un ampli rang de polimorfs de ZnO basats en lamines hexagonals inspirades en l’enumeració de les seves xarxes subjacents característiques i evaluant l’estabilitat del sòlid “bulk” i les nano-lamines d’aquestes estructures mitjançant calculs ab initio. Hem observat un ampli polimorfisme d’estructures de baixa energia en les nano-lamines amb un ordre d’estabilitat totalment diferent al del sòlid “bulk”. A partir d’aquestes bases generals hem pogut tenir un millor coneixement de les transicions estructurals observades durant el creixement epitaxial i les prediccions d’estabilitat de les nano-lamines en variar-ne el gruix i la pressió exercida. Hem conclòs els nostres resultats explicant que la nanoporositat està inextricablement connectada tant amb la Erel com amb el ΔEgap i hem predit que la nanoporositat pot induir un increment en el band gap de fins a ~1.5 eV relatius a la wurtzita ZnO. Comprovant també la generalitat d’aquest fenomen, pe’l CdS i pel CdSe suggerim que la nanoporositat pot ser emprada com un mètode genèric d’enginyeria de band gap per materials funcionals morfològicament i electrònicament.

[eng]Throughout this thesis, we have studied ZnO and its properties in a bottom-up manner through a dimensionality range starting from 0D nanoparticles to 3D bulk phases. For the 0D clusters and the 2D nanofilms studied we also considered the effect of a support in models designed to study ZnO thin film growth on the Ag(111) surface. In chapter 3, we have studied ZnO nanoclusters on a Ag support and compared their properties with free space ZnO nanoclusters. In this chapter we highlighted the importance of the presence of the support during the global optimization of the clusters (i.e. as opposed to global optimization of the clusters in free space and then introducing a support). Our results show that the presence of the support strongly affects the energetic stability ranking of the nanocluster isomers. More drastically, after a certain cluster size, the support also stabilizes selectively 2D type structures, which are not stable in free space, with respect to the 3D clusters. The extra stabilization of supported 2D clusters is attributed mainly to the contact area, which is evidently greater for 2D clusters where all the atoms can interact with the surface. The importance of the contact are is also observed for 3D clusters, as ellipsoid bubbles or inflated double layer clusters being of lower energy than more spherical clusters on the support whereas the latter, more symmetric 3D clusters are more stable in free space. Matching of the cluster structure with the surface morphology was found to be another factor determining cluster stability. The Zn3O3 sixmembered ring, which is one of the main motifs for both 3D and 2D clusters, matches best with the Ag(111) surface because it follows the same six-fold (C6) symmetry (or its trigonal C3 subgroup with a three-fold axis, taking into account the distinction between Zn and O atoms). However because of the lattice size differences, such matching dies away for larger ZnO clusters. The preferential stabilities of the 2D structures of ZnO clusters can be seen as the initial stages of thin film growth and is found to be in line with the experimentally observed layered ZnO sheets on the Ag(111) surface. In Chapter 4, we have considered a full 2D-ZnO sheet on Ag(111) surface and also investigated also how H atoms interact with it. Following our results for ZnO nanoclusters on the Ag surface, we highlighted the importance of the degree of 2D-ZnO:Ag(111) unit cell commensurability for calculating accurate sheeton-substrate binding energies. We have found a 8:9 commensurate monolayer to be more favored with interatomic potentials and a 7:8 commensurate monolayer with DFT calculations, where the latter is found in experiment. Our calculations showed no evidence of charge transfer or covalent bonding between the Ag(111) surface and the 2D-ZnO sheet, but did show that the ZnO sheet and the Ag(111) surface exhibit small structural distortions in order to maximize their mutual interaction. Calculations of the unsupported 2D-ZnO sheet interacting with hydrogen provided strong evidence for H forming a low energy Zn 4s–H 1s multi center bonding state when passing through a Zn3O3ring of the 2D-ZnO sheet, thus allowing for relatively facile H transport through the sheet. In chapter 5, we have extended our study of supported 2D ZnO nanofilms with higher coverage models, including triangular islands on top of two full monolayers, prepared to model the experimental system. Our results showed that the triangular adlayer islands induce a transition to the WZ structure in the island core and in local region in the two layers immediately below the island core. The islands are also found to have BCT-structured reconstructions on their edges and T1-structured reconstructions on their corners. These models are found to better match the experimental structural data for the experimental 2.7 ML Ag-supported ZnO film with respect to models assuming a purely layered or a purely WZ structure. In chapter 6, we focused on 4ML nanofilms and compared bulk and the 4 ML nanofilm poymorphism of ZnO. Our results revealed that the stability range of nanofilms and their energetic ordering are radically different than that of bulk polymorphs. We have developed a method to generate a wide range of new low energy nanofilm and bulk polymorphs using nets as a basis, and showed that there exist at least three nanofilm structures with trigonal basal plane symmetry compatible epitaxial growth on fcc metal (111) surfaces that are more stable than layered-ZnO. While confirming the previous theoretical studies predicting the BCT-ZnO phase as being the lowest energy free-standing nanofim for small thicknesses, we obtained a range of structurally related and near energetically degenerate phases, indicating there exists BCT polytypism. With increasing thickness we found that atomically reconstructed wz-ZnO becomes more stable than BCT-ZnO for ~14 MLs, and is always more stable than non-reconstructed wz-ZnO. We have also stressed the influence of strain on polymorphism by showing that BCT-ZnO and layered-ZnO nanofilms are unstable to novel polymorphs under in-plane strain. Together with the T1 structures and BCT structures which were also predicted as reconstructions on island corners in the previous chapter, our results strongly suggest that many new nanofilm polymorphs should be experimentally accessible, and in some cases, may have even already been observed. In chapter 7, we focused on bulk polymorphism and, specifically, investigated the effect of nanoporosity. Our results showed that both energetic instability and band gap increase with nanoporosity and we predicted that nanoporosity could induce band gap increases of up to ~1.5 eV relative to wurtzite ZnO. We showed that the band gap increase is related with bandwidth changes in the conduction band and the valance band. We suggested that the underlying physical mechanism for this effect is that introducing nanoporosity, and thus periodic internal void space, restricts extended orbital overlaps and thus decreases bandwidths. Due to the generality of this argument, we expect that nanoporosity could similarly affect bandgap values in a wide range of materials and could be employed as a band gap engineering method.