Large area TMD-based van der Waals heterostructures featuring enhanced photoconversion in the flat optics regime

One of the most urgent technological needs of this century concerns the research of innovative nanomaterials for applications in optoelectronics, nanophotonics and photovoltaics for renewable energy conversion. In fact, downscaling of the silicon-based devices (e.g. field-effect transistors) has come to an end due to intrinsic physical limitations such as size, inadequate carrier mobility, short channel effects, atomic-scale interactions, heat generation, and energy consumption at atomic scale thickness, requiring innovative materials to overcome such difficulties. Graphene discovery in 2004 by Novoselov and Geim arose a tremendous interest for the two-dimensional (2D) materials, in which the reduced dimensionality brings new properties with respect to their bulk counterparts. Despite the extraordinary properties exhibited by graphene, its gapless nature strongly limits the fabrication of graphene-based optoelectronic devices. For this reason, the interest of researchers shifted towards the class of 2D semiconductors. Among these materials, Transition Metal Dichalcogenides (TMDs) represent the most important family due to suitable bandgap energy values which match the Schockley-Queisser efficiency criterion for solar photoconversion, and to their extraordinary optical absorption coefficient. Additionally, 2D materials can be vertically stacked to form the so-called van der Waals heterostructures that are endowed by pristine interface thanks to the relatively weak van der Waals interactions that keep the stack together. This offers the opportunity to fabricate heterostructures of arbitrary 2D materials, independently by their crystal structure, with no limitations on the engineering of the optoelectronic and photonic properties of the new stacked metamaterial. In particular, the possibility to realize van der Waals p-n junctions by coupling 2D-TMDs layers is very intriguing for photoconversion and photovoltaic applications. So far, TMDs employment has been mostly limited to the fabrication of prototypical devices such as field-effect transistors and photodetectors, most of which were realized via mechanical exfoliation from single crystal of flakes with thickness ranging from one to few atomic layers. Despite their intriguing properties, these materials do not represent an industrially relevant alternative to traditional semiconductors, being the exfoliation a randomic process with very low yields and limited areas in the micrometer range. Consequently, one of the key frontiers of TMDs research is the large area growth of homogeneous ultra-thin layers with controlled thickness over macroscopic areas. To meet this requirement, several large area techniques have been studied to synthesize TMDs layers extending over cm² areas. However, in contrast with the exfoliation process, these techniques result in polycrystalline layers where the high concentration of grain boundaries degrades the macroscopic conduction. The second crucial issue to be solved for ultra-thin TMD-based devices to be used in photoconductive and photodetection applications is the maximization of the optical absorption. Despite the excellent optical absorption coefficient, an ultra-thin TMD film indeed cannot absorb efficiently the incoming light due to the ultimately reduced thickness, which means a nanometric optical path. It is thus evident that new strategies for the optical absorption amplification in ultra-thin 2D semiconductors need to be developed to allow proper performances in TMD-based photodetection and photovoltaics applications. Because of the atomic thickness of the 2D layer, traditional solutions developed for the light harvesting amplification in conventional silicon-based photovoltaic devices, based on pyramidal microstructuring or on the addition of antireflective coatings, cannot be transferred to these materials. Recently, the research group where I carried out my activity worked on large area ultra-thin MoS₂ films conformally grown on self-organized rippled substrate fabricated by ion beam sputtering. Their results clearly showed a modification of the TMD optical and electronic properties grown on the nanostructured substrate with respect to a flat one. This observation was explained with the stress induced by the substrate morphology in the MoS₂ layer in correspondence to the high curvature regions given by the crests and valleys of the ripples, meaning that by control of the substrate morphology it is possible to engineer the material intrinsic properties. Additionally, the nanostructures anisotropy adds a polarization-dependent optical response, offering a way to engineer the optical absorption of the 2D material in view of photoconversion applications. However, self-organized nanostructures suffer from a relevant size dispersion and long range disorder, whereas more interesting optical effects are expected for periodic nanogratings in which the subwavelength TMD layers reshaping provide them the functionality of flat optic diffractive elements. Starting from here, I devoted the most of my research activity to face the two main challenges described above: developing a growth process for large area ultra-thin TMD films, and studying an efficient light harvesting strategy to maximize the optical absorption in few-layer semiconductor films. An overview of the state-of-the-art regarding 2D materials and nanophotonics approaches for light harvesting in ultra-thin films is given in Chapter 1, while the growth synthesis techniques are postponed in Chapter 2. At the beginning of my PhD, MoS₂ films were grown by external collaborators of my group, thus limiting the activities. In the first phase of my research activity, I developed a novel large area growth process based on the physical deposition of solid precursor films and sulfurization, as I will describe in Chapter 2. This novel technique enabled not only the in-house growth of ultra-thin MoS₂ films for the first time, but also allowed to extend the process to ultra-thin WS₂ layers. Having now the capability to control the deposition of two different TMDs layers, I moved to the growth of large area van der Waals heterostructures. Due to the type-II heterojunction formed by the band structure coupling of MoS₂ and WS₂, such heterostructures are expected to have a high potential in photoconversion applications. In Chapter 3 I will show the nanofabrication of planar MoS₂/WS₂ heterostacks, and their application in photocatalytic experiments and in a prototype of photonic device, featuring first evidence of photovoltage and photocurrent under illumination. This latter application also required me to develop large area transparent electrodes, so that I will dedicate a part of the chapter to indium tin oxide thin films and large area graphene. In the second phase of my research activity, I focused on light harvesting in ultra-thin TMDs layers. Thanks to the conformality achieved by the physical deposition process, I explored a nanophotonic approach based on the optical anomalies arising from periodic modulation of the TMD layer at the subwavelength scale, obtained by conformal growth of the TMD layers onto nanostructured substrates. To this end, periodic nanogratings have been used as a template for the growth of ultra-thin MoS₂ layers. Differently from the self-organized nanostructures mentioned before, the periodicity induces diffractive effects that are exploited to steer the light parallel to the active 2D material enhancing the optical absorption, as demonstrated in Chapter 4 both directly by absorption measurements and indirectly by a photo-to-chemical energy conversion experiment where we detected enhanced photocatalytic performances. In the final part of my project, I started preliminary studies on the elastic scattering properties of subwavelength periodical lattices based on nanostructured tilted TMD layers. By defocused ion beam sputtering, I was able to reshape the morphology of the initial nanograting templates to further engineer the TMD optical response. In particular, by off-normal incidence sputtering it is possible to tailor a specific slope of the tilted nanofacets, on top of which I deposited laterally disconnected MoS₂ nanostripes by grazing angle physical deposition. By developing a custom-made scatterometer, optical characterization of ultra-thin MoS₂ nanostripes and thicker MoS₂ films was performed, giving interesting preliminary results on the directional light scattering properties of these reshaped layers, as reported in Chapter 5. Finally, I adopted a similar deposition approach for the nanofabrication of large area heterostructures nanoarrays based on few-layer WS₂ nanostripes coated by a conformal MoS₂ layer, demonstrating further engineering of the optical response of few-layer TMD films with impact in photoconversion.