Si based nanostructures for solar cells

Abstract

Photovoltaic technology has received increased attention as one of the most promising approach to carbon-free energy production. Bulk silicon cells, which convert between 14 and 17% of incident light into electricity, make up 90% of the solar cell market. Silicon is widely used because it is the second most abundant element in the earth's crust, and because the electronics industry has already developed infrastructure to process it. Yet the pricey and complicated manufacturing makes these photovoltaic (PV) systems more expensive per kW/h than conventional energy sources. These limitations have driven efforts to develop inexpensive solar modules with efficiencies equivalent to, or better than, existing devices. Thin films solar cells are widely recognized as a key solution to reducing the manufacturing cost of PV cells in the near to medium term. Thin film solar cells are able to be produced at low cost by removing the bulk active substrate and using an additive deposition process on top of a low cost substrate such as glass, metal foil, plastic, etc. Compared with other thin-film solar cell technologies, thin-film silicon has the advantage of constituting an industrially mature technology and of being based on raw materials which are present in abundance in the earth's crust.[1] During the last three decades, hydrogenated amorphous silicon has been studied extensively as basic material for thin film solar cells due to the natural abundance of source material, environmental safety, potential high performance and the capability of low cost production. However, the defect density of hydrogenated amorphous silicon (a-Si:H) increases with light exposure, to cause an increase in the recombination current and leads to the reduction in the sunlight to electricity conversion efficiency. This phenomenon is known as Staebler-Wronski effect [2] For this reason, the solar cells based on a-Si:H have always been associated with efficiency losses due to the light-induced degradation over the time. For amorphous silicon the photoinduced degradation due to the creation of dandling bond defect has been the most important issue for improving the efficiency [3] Extensive research has been carried out by many laboratories to improve the conversion efficiency of a-Si:H solar cells [4]. In this thesis an intensive study on thin film a-Si:H solar cells will be shown.The fundamental photodiode inside an amorphous silicon' based solar cell has three layers deposited in either the p-i-n or the n-i-p sequence. The realization of a pin solar cell will be shown, including the investigation of the key physical properties necessary to design the solar cell (sheet resistance, activation energies for conductivity, light absorption). The electrical characteristics of the solar cells in dark condition and under standard illumination will be presented and compared to the theoretical models.

In order to make photovoltaic technology cost effective we need to improve the efficiency and minimize the light induced degradation of photoconductivity caused by Staebler' Wronski effect [2]. A material that presents very promising features is hydrogenated nanocrystalline silicon (nc-Si:H)[6]. The nc-Si:H is made using PECVD.

In order to make photovoltaic technology cost effective we need to improve the efficiency and minimize the light induced degradation of photoconductivity caused by Staeblerà à à à à à à à à à à ¢ Wronski effect [2]. A material that presents very promising features is hydrogenated nanocrystalline silicon (nc-Si:H). The nc_Si:H is made using PECVD, using a gas mixutre of SiH4 and H2, the cristalline dimension can be tailored by variyng the hydrogen dilution of Silane in PECVD camera. The transition from amorphous to nanocrystalline silicon is induced by Hydrogen. Vettel et al [5]demonstrated an important result, the deposition condition near the transition between amorphous an cristalline growth have been found to be the most beneficial to the solar cells properties. Nc-Si:H films show stability under light soaking, suggesting that the disorder in the amorphous Si network plays a major role, and thus Nc-Si:H have high potential as stable solar cell material. The parameter that could play a role in the stability under light soaking and also in the crystalline fraction is believed to be the hydrogen concentration and its complex bonding mechanism. The major technical challenge with such nanocrystalline solar cells is the fact that the surface area of grain boundaries is very high, significantly increasing the density of recombination centers as well as the probability of recombination due to charge carriers having to pass through so many boundaries, indicating that the crystalline fraction and the grain size of these materials determine also the electrical properties of the final device. It is believed that the hydrogen content and its complex bonding behavior is the responsible for the structural characteristics and as a consequence for the stability under illumination. In order to control the optical and electrical properties and then the stability under light soaking effect, the hydrogen contribution must be deeply understood. In this thesis a study of the morphological behaviors of nc-Si:H films as a function of the hydrogen content will be showed. Nc-Si:H thin films were deposited by ST microelectronics under different conditions by PECVD deposition by varying the ratio H2/SiH4 flow. The variation of morphology and hydrogen content was studied in detail, using TEM analysis, ERDA and FT-IR measurements. The study of this material was also made using the radial distribution function (RDF), which starting from the diffraction pattern of TEM analysis, gives the number of atoms at a given distance from a central atom. The contribution of the hydrogen in the transition from an a-Si:H layer, with low R, to a nc-Si:H layer, with high R, will be demonstrated.

As previously described, in contrast to amorphous silicon (a-Si), nc-Si exhibits high stability against photo-induced degradation. However, nc-Si thin-film solar cells require an intrinsic layer of in thickness to absorb sufficient amount of sunlight due to its indirect optical transition. For this reason in literature another type of nanostructure, has been recently proposed, the Si nanowires. Silicon nanowires offer several performance and manufacturing benefits that may impact future PV applications. Their main advantage is based on the fact that they orthogonalize the pathways for light absorption and carrier collection. Recently Atwater et al. reported the potential of radial p-n junction wire array solar cells, which consist of a dense array of semiconducting wires, each having a p-n junction in the radial direction, and oriented with the nanowires axis parallel to the incident light direction. By fabricating p-n junctions conformally around the wires structure, in a radial configuration, the absorption of light can be decoupled from minority carrier diffusion. This approach is useful to the creation of efficient solar cells using low lifetime materials. [7,8]. Although fabrication of wires in micrometer sized dimension have been proposed in many works to prevent the surface recombinarion, the approach proposed in this thesis is the use of nanometer sized dimention to the nanowires syntesis. The choice of using these nano-structures is driven by numerous factors. The possibility of a switch to an almost direct gap with the ability to vary the energy gap by varying the diameter of the wires is one of the main factors. Zhao et [10] found that for a nanowires oriented in [100] direction with diameter of about 10-12 nm the resulting energy gap is about 1.4 eV. This result is very close to the Shockley' Queisser limit [9]. The theoretical maximum almost corresponds to the GaAs band gap. ( 1.42 eV), However, due to high production costs GaAs is reserved for solar applications, for this reason in very interesting the use of a technique that is able to achieve an Eg of 1.4 with low cost production. There are two basic approaches of synthesizing nanowires: top-down and bottom-up approach. A top-down approach combines lithographic steps or etching process to produces nanowires from a flat surface. The major drawback of the top-down method is that the surfaces of the structure are damaged during the process resulting in nanowires with a poor crystal quality. In addition the lithographic techniques may not be able to produce sufficiently small structures for further downscaling of devices. In order to produce small enough nanowires of high enough crystal quality, the bottom-up approach is thought to be a potential alternative. The idea is to build-up nanosized structures and devices by using nanoscale building blocks to initiate growth directly at desired positions and with designed dimensions and properties. In contrast to the lithographic and etching techniques used in the top-down methodology, the bottom-up approach involves the direct growth of one-dimensional nanostructures onto a substrate. The typical method to fabricate NWs by bottom-up approach is the catalytic growth with random metallic nanoparticles. A better understanding of the catalytic nanowire growth process is necessary to pin down the growth mechanism and to be able to rationally control their compositions, sizes, crystal structures, and growth directions. Gold was used as a catalyst for the nucleation of the Si nanophase. The choice to use gold relies on the advantages of this metal respect to others, like Ni, Sn, Al and Cu, such as its low eutectic temperature with Si, the opportunity to easily form alloys with the growth precursor and the fast inter-diffusion of Si through gold nanodot. Gold, however, presents some drawbacks, because it creates deep band gap defects and for this reason the data shown represent a proof of concept, and alternative metals must be investigated for device integration. The catalytic growth of silicon nanowires is commonly described either by the vapor-liquid-solid (VLS) or solid-liquid-solid (SLS) process. The VLS process was first suggested by Wagner and Ellis [11]who showed that micrometer-scale silicon whiskers could be grown from metal droplet catalysts under Chemical Vapor Deposition conditions at about 1000à à à à à à °C. A typical VLS mechanism starts with the dissolution of gaseous precursors (SiH4 or SiH2Cl2) into nanosized liquid alloy (metal Au-Si) droplet considered as a catalytic site. Once the liquid droplet is supersaturated with silicon, then the precipitation of solid nanowire occurs. The process was named the VLS beacuse of the three phases involved. In this method, metallic nanoparticles are deposited randomically on a Si substrate and work as catalysist seeds for the nucleation of the Si phase from the vapor source, in the first case, and from the substrate source, in the second case. During the growth, in both cases however, some lateral diffusion of the metallic dots on the substrate can produce a coalescence of the metal nano-seeds, thus increasing the disorder level in the template substrate, and producing a final' forest-like' material. Therefore, when position control of spatially separated NWs is desired, nanopatterning techniques become essential. Some approaches have been demonstrated, or are potentially applicable, for position and size control of semiconductor NWs. [12] A number of patterning and templating methods can be applied for the controlled preparation of metal dots or metal-dot arrays on a substrate surface, including photo- or e-beam lithography, manipulation of single gold nanodots, arrangement of Au nanocrystals from suspensions, nanosphere lithography, gold deposition masks based on porous alumina templates, nanoimprint lithography, as well as other catalyst-positioning approaches. Since the final structural and geometrical characterisitcs of the NWs are strictly correlated to the catalyst dot properties, we want to control the catalytic seed characteristics, such as shape, position, dimension and we create a SiO2 barrier to the lateral diffusion during the annealing to control the final nanowires template. In this thesis we propose an approach to obtain these characteristics based on a two steps process. i) First of all the formation of a polymeric nanomask by diblock copolymer self-assembling [13-15] and successive dry etch to transfer the polymeric pattern on the oxide substrate [16-20]. This creates an ordered array of nanopores of controlled size and density and position. The obtained substrate is made of a silicon dioxide layer with nanopores 20 nm wide and separated by a 40 nm distance, etched down to the Si substrate. ii) Afterwards, catalyst nanodots was deposited and diffused over the template. A single layer of gold NDs, with coverage 20%, is deposited by sputtering all over the template, and subjected to thermal annealing at temperatures ranging between 600 and 1000à à à à à à °C, to diffuse the metallic NDs on the surface. In this thesis it is shown that the NDs deposited randomically over the template either inside and outside the pores, during the annealing diffuse and coalesce randomly over the substrate, but in correspondence of the nanopores they stop the diffusion, probably due to the presence of some surfacial defects. The final result of this process is that most of the nanopores are saturated with the gold NDs. The density of saturated nanopores can be tuned by changing the annealing temperature. The preferential diffusion of the gold NDs toward the nanopores is followed in situ by annealing the samples during TEM imaging, and ex-situ by annealing in furnace the samples and observing them by TEM and high-statistics SEM analysis.

The first chapter describes the pin cells realization. To explain the characteristic of hydrogenated amorphous silicon in the first section, we introduce some of the fundamental physical concepts required to interpret the scientific literature about amorphous silicon.The second section describes the electrical and optical characterization of the layer that compose the solar cells, and an accurate electrical characterization of the pin structure realized by ST.

The second chapter shows the characterization of nc-Si:H thin films made by PECVD for solar cells. In particular a structural characterization performed by TEM analysis will be showed, and the role of hydrogen in the transition from a-Si:H to nc-Si:H will be deeply analyzed and discussed.

The third chapter shows results on the fabrication of ordered nanowires. It uses an alternative lithography based on diblock copolymer self-assembling which generates ordered arrays of nanopores of controlled size, density and position. This can be used as a template for the catalytic mediated growth of controlled nanowires.