Atomistic investigation of morphology and optoelectronic properties of bismuth sulfide nonostructures

Nanostructured metal sulfides (MS) have attracted great interest in recent years because of the possibility to synthesize nanoparticles from solution and tuning their optoelectronic properties through quantum confinement phenomena. A large number of applications have been reported in the field of solar cells, light-emitting diodes, lithium-ion batteries, thermoelectric devices, sensors, fuel cells and nonvolatile memory devices. Among the large family of semiconductor sulfides, the present Thesis is focussed on bismuthinite. The choice was motivated by a combination of factors. Its non-toxicity, low cost synthesis and high absorption properties make Bi2S3 a promising material for several application, such as solid-state semiconductor-sensitized solar cells. The intrisic anisotropy in the crystal structure of this material facilitates the formation of elongated nanostructures, in particular nanorods, nanoribbons, and nanowires. These structures find important applications in many nanodevices, for example field emitters, solar cells, and lithium-ion batteries. On the other hand, researchers are still far from a complete understanding of Bi2S3 properties. The colloidal synthesis of bismuthinite nanostructures, although cheap and environmentally friendly, does not allow a perfect control on stoichiometry and surface passivation. The large majority of the experimental studies does not report photoluminescence of the nanocrystals, which indicates the presence of trap states and a low defect tolerance of the material. These facts cause a lower efficiency of the devices based on Bi2S3 nanoparticles with respect to anologous system based on other nanomaterials (e.g Sb2S3 in solid-state sensitized solar cells). The absence of linear optical data makes more difficult to investigate the electronic structure of the material by means of spectroscopic techniques. Ab initio atomistic simulations represent a valid alternative to get an insight on the optoelectronic properties of bismuth sulfide. Despite some computational work on bulk Bi2S3 is already present in literature, there are currently no ab initio studies concerning nanostructures of this material. Such lack of information motivates the work of the present thesis that focuses on the investigation of morphology and electronic properties of Bi2S3 nanocrystals. The thesis is organized as follows. In the first chapter the main differences and advantages of nanostructured materials over the bulk counterpart are presented. I put the accent on metal sulfide nanocrystals, in particular those of the Bi2S3 family (pnictogen chalcogenide) and their application in several fields of physics, environmental science, and engineering. A section is reserved to report the development of elongated semiconductor nanostructures and their peculiarity with respect to nanocrystals with lower aspect ratio. The second chapter describes the computational and experimental methods used in this study. The basic concepts of density functional theory and its implementation in quantum-chemistry codes are reported. A description of the synthesis and spectroscopic methods used to check the validity of the theoretical predictions is also given. For the detailed list of the basis sets, pseudopotentials, and exchangecorrelation functionals used in each calculation I refer to the end of Chapter 3 and 4. Chapter 3 deals with the bulk properties of Bi2S3. Atomic and crystal cell relaxation are performed. Also I investigated electronic properties from the calculation of the band structure, density of states, and efficient mass. These simulations are an important preliminary to the study of Bi2S3 nanostructures. By comparing my results with the previous studies present in literature it is possible to validate the method (functionals, pseudopotentials, etc) and proceed with the study of unexplored systems. Chapter 4 investigates the properties of Bi2S3 nanostructures. First, I focus on elongated nanoribbons (that are the building blocks of the crystal structure) and study saturated and unsaturated nanocrystals of finite size in comparison with one-dimensional infinite ones. By means of (time-dependent) density functional theory calculations it is demonstrated that the optical gap can be tuned through quantum confinement with sizable effects for nanoribbons smaller than three nanometers. A comparison with Sb2S3, shown that Bi2S3 nanostructures have similar tunability of the band gap and a better tendency of passivating defects at the (010) surfaces through local reconstruction. Then, the focus shifts over ultrathin nanowires formed by the aggregation of a small number of nanoribbons, with lateral sizes as small as 3 nm as in fact observed by transmission electron microscopy. Their electronic properties are investigated finding that surfaces induce peculiar 1D-like electronic states on the nanowire edges that are lo- cated 300 meV above the valence band. Sulfur vacancies are also responsible for localized states a few hundreds meV below the conduction band. The possibility to remove the surface-induced intragap states is further investigated by passivating the surfaces of the nanowires with carboxylic and amine groups that are commonly employed in colloidal synthesis. The small methylamine and acetic acid molecules are expected to fully passivate the surfaces of the nanowires removing the edge states and restoring a clean band gap. Conclusions are finally reported in Chapter 5. The results of the present Thesis provide a characterization of the energetics and optoelectronic properties of bismuth sulfide nanostructures showing the relevance of surface defects and suggesting a possible route for improving optoelectronic properties of Bi2S3 nanostructures by tuning the size of the ligand molecules.