Project Topics
Get Complete Project Material File(s) Now! »
Hot-electron Generation
As was described in the previous section the LSPR band appears at different energies for different metallic nanoparticles, and can be tuned by controlling the particle size, shape, density, etc. Therefore, it is possible to fabricate plasmonic nanostructures with LSPR resonance covering the electromagnetic spectra from the UV to the infrared. This together with the enhancement of the electric field around the particle surface can be exploited for energy harvesting. In fact, the integration of plasmonic nanostructures into semiconductor and photovoltaic devices has been proposed as a way of improving device efficiency. Particularly, the LSPR allows direct conversion of light into electricity through hot-electron generation, as will be discussed in the following.
After LSPR excitation, surface plasmons decay radiatively, by re-emission of photons, or non-radiatively through emission of hot-electrons. These electrons can be excited above the Fermi level, as is shown in the schematic of Figure 1.13. If the particle is in contact with a semiconductor material (forming a Schottky junction), these electrons can be injected into the semiconductor if their energy is higher than the Schottky barrier, 𝜑𝑆𝐵=𝜑𝑀−𝜒𝑠, where 𝜑𝑀 is the work function of the metallic particle, and 𝜒𝑠 is the electron affinity of the semiconductor. Therefore, plasmonic nanoparticles, with LSPR in the visible or near-infrared (lower than the semiconductor bandgap) can be excited by the low energy photons of the solar spectrum, and produce hot-electrons (E ~ 1 – 4 eV) that can be injected into the conduction band of the semiconductor. Hence, the low energy portion of the absorption spectrum of the semiconductor material is enhanced (by itself, the semiconductor can only absorb photons of energy higher than its band gap). This effect has been exploited for the fabrication of optical antennas for photodetection, for photocatalysis and photovoltaic applications [3,110].
The LSPR gives place to a highly localized electric field at the nanoparticle surface, that can be 103 times higher than the electric field of the incoming light [111]. Electron-hole pairs generation in a semiconductor, in close proximity with the plasmonic nanoparticle, is increased by few orders of magnitude, since it depends on |𝐸2|. This mechanism has been used for increasing the surface enhanced Raman scattering (SERS) signal, in the detection of certain molecules [112] and photocurrent enhancement of 20% have been reported in Si nanowires decorated with Au nanoparticles, studied through variable-wavelength scanning photocurrent microscopy [113].
In systems composed of larger particles (2R > 50 nm) or aggregates, in addition to the LSPR there is an increase of the light scattering by the plasmonic nanostructures. Due to the strong scattering of the resonant photons, the optical photon path inside the device is increased and, in consequence, the generation of electron-hole pairs is boosted. This mechanism is depicted in Figure 1.14, where Cu nanoparticles are used to enhance the photocurrent in a Si photovoltaic device, by scattering and trapping light inside the device [114].
Hot-carrier Photodetectors
The LSPR phenomena observed in some metallic nanostructures that was described in section 1.4.1, has been widely exploited for photodection. There are two main photodetection mechanisms for devices based on plasmonic metallic nanostructures. First, through the enhanced electric field created at the nanoparticle surface that, when it is in contact with a semiconductor, can serve to increase the hole-carrier generation and the photocurrent, and secondly, the generation of hot-carriers at the plasmonic nanostructure and injection into the semiconductor. Here, we will focus on the hot-carrier plasmonic photodetectors, that is the main interest for this thesis.
As was already explained before, when a plasmonic metallic nanostructure is brought in contact with a semiconductor material and a Schottky junction is formed, thanks to LSPR, hot- carriers are generated at the metallic nanostructure and if they possess enough energy they can be transferred to the semiconductor. In a traditional Schottky photodetector, the photodetection mechanism is based on the generation of hole-electron pairs at the semiconductor and separation at the depletion layer. If the metallic film is replaced by plasmonic metallic nanoparticles, the photocurrent can be increased considerably by the injection of electron from the nanoparticle to the conduction band of the semiconductor, in addition to the electron generated in the semiconductor from band-to-band transitions. Furthermore, the spectral response of the device can be extended by combining the absorption spectra of the semiconductor with that of the metallic nanostructure.
Liu et al. reported the integration of Ag nanoplates into a flexible Si-nanowire-based photovoltaic device that allows a considerable enhancement of the external quantum efficiency of 59 % in the near-infrared, by LSPR in the Ag nanoplates and hot-electron injection into the n-type Si nanowire [116]. Luo et al. presented a photodetector made of Au nanoparticles deposited on CdTe nanowires. The Au nanoparticles improved the device responsivity by 100 times, and extended the device sensitivity to larger wavelengths, to which the virgin CdTe nanowires were blind. Furthermore, the response time of the device was considerably reduced with the incorporation of the nanoparticles from 6 to 2 seconds, approximately [117]. They also reported the fabrication of a plasmonic photodetector based on hollow gold nanoparticles (HGNs) deposited on CdSe nanoribbons (NR). All the figures of merit of the photodetector were improved by the inclusion of the HGNs. The experimental results, together with theoretical calculations show that the enhanced performance observed in the device is due to the increase of the photocurrent due to localized field enhancement at the particle surface and to direct carrier injection into the CdSe nanoribbons [118]. Knight et al. showed that embedding the plasmonic structures into the semiconductor material enhances the hot-electron transfer since more Schottky interfaces are created. As a proof of concept Au nanobelts were embedded in a Si substrate. The efficiency in the embedded architecture was 25 times higher compared with the planar structure [119].
Chemisorption Mechanisms and Self-limiting Reactions
Three main chemisorption mechanisms have been identified for the self-limiting ALD reactions: association, dissociation and ligand exchange reaction (see a schematic in Figure 2.2). When association occurs, the precursors molecule is completely adsorbed on a reactive site of the surface without any ligand release (a). In dissociation, the precursor molecule is split, a fragment is chemisorbed at the surface and the ligand released is bound to another reactive site at the surface (b). Finally, for the ligand exchange reaction, the precursor is also split, but the released ligand is bound with a surface group forming a volatile compound that is released as a gaseous reaction by-product (c) [121].
The factors causing the saturation of the surface in the self-limiting ALD reactions are the steric hindrance effect and the limitation in the number of reactive sites available at the surface (see Figure 2.3). In the case of steric hindrance of the ligands, some reactive sites at the surface can be shielded by the large precursor molecules already attached at neighboring sites, as is schematically represented in Figure 2.3 a. Under this condition, there is no space available for the new incoming precursor molecules, therefore the surface is considered full and saturation is reached. In the second case, when all the reactive sites at the surface are occupied by the precursor molecules, saturation is achieved because there are no more suited bonding sites available for new incoming molecules, even if there is still plenty of space at the surface (see Figure 2.3 b). Steric hindrance effect dominates when precursor formed by large molecules are employed. Regardless of the mechanism causing saturation of the surface reactions most of the ALD processes lead to the growth of less than one monolayer per cycles. Even if saturative conditions are achieved, in both cases, the chemisorbed layer doesn’t necessarily contain enough atoms to form a full monolayer of the desired compound [121], and rarely GPC reach values higher than 50 % of a monolayer. Typical values of GPC range from 0.01 to 0.3 nm/cycle [122].
ALD Temperature Window
Typical ALD processes are conducted at modest temperature (210-350°C) compared to the equivalent CVD processes. The temperature range, in which the conditions of self-limiting surface reactions are fulfilled, is called ALD temperature window. This temperature range is dependent on the specific ALD process (precursor employed, nature of the substrate, and reactions taking place). Figure 2.4 shows a schematic of the different possibilities for the GPC evolution as a function of the substrate temperature. The central region (orange) corresponds to the ALD window, in which the GPC is constant, independently of substrate temperature. Outside the ALD window, the self-limiting reaction mechanism is lost and there is a dependency of the GPC on the temperature.
For lower substrate temperatures, the self-limiting conditions are not reached because of slow reaction kinetics (Figure 2.4, curve A) or precursor condensation (Figure 2.4, curve B). In case A, the temperature is not high enough to efficiently activate the involved chemical reactions and complete saturation of the surface reaction sites is not achieved, which results in slow reaction kinetics. Therefore, the GPC increases with substrate temperature, until temperature is high enough to achieve complete saturation. In case B, the low substrate temperature results in the condensation of the reactants or the by-products, which produces more than one monolayer per cycle [123]. The reduction of condensation by increasing the substrate temperature produces the decrease in GPC observed in Figure 2.4, curve B.
Experimental Setup and Deposition Parameters
For the deposition of the metallic Cu thin films, Cu NP and Cu2O thin films an ALD PICOSUN™ R-200 Advanced reactor was employed (see Figure 2.6 a, b). This ALD reactor is composed of a dual chamber structure: an exterior chamber (the vacuum chamber) and an interior one which houses the substrate holder and where deposition takes place. Copper (II) hexafluoro acetyl acetonate, known as Cu(hfac)2, was used as Cu precursor (99.99+%-Cu, Stream Chemicals) and deionized water was used as reactant. The precursor choice for this study was done on the basis of the versatility of Cu(hfac)2 for the growth of metallic copper and copper oxide films, based on several CVD processes reported in the literature [130–132] and in a few ALD studies [38,41]. Both precursors were introduced into the reaction chamber by separate gas lines (see schematic in Figure 2.6 b). Nitrogen was used as carrier gas to transport the precursor into the reaction chamber, and as purge gas to evacuate the reactor between precursor pulses. The N2 flow rate of each line was controlled by a mass flow controller and the pressure was measured by a pressure transducer. For the Cu(hfac)2 and H2O lines, N2 flow rates of 120 and 150 sccm were used, respectively. In order to enhance the vapor pressure of Cu(hfac)2 the boosting mode was used with N2 flow rate of 300 sccm. The Cu and Cu2O films were deposited on sputtered ZnO, Al-doped ZnO thin films, monocrystalline -Al2O3 (CrystTec) and monocrystalline ZnO (PI-KEM) substrates.
The procedure followed for the deposition is described below:
Cleaning the substrates with ethanol and distilled water.
Introducing the substrates in the sample holder inside the reaction chamber.
Pumping down the ALD reactor to 10 hPa.
Heating the reaction chamber and the Cu(hfac)2 bottle to the selected temperature (see Table 2.1).
When temperature is stabilized, the 4-step ALD deposition process starts as described in Figure 2.1 (just before starting the deposition of the films the substrate was exposed to a remote-plasma of O2 for 2 min, in order to prepare the surface for the deposition).
Reactive Magnetron Sputtering
So far we have described the sputtering of a metallic target to deposit a metallic film of the same material. There is also the possibility of depositing compound films such as oxides, nitrides or carbides. One possibility is the use of a ceramic target; however, this requires working in radiofrequency (RF) mode due to the insulating character of this type of targets to avoid the charge accumulation at the surface and the occurrence of arcs. RF power supplies are more expensive than DC generators, and also ceramic target are usually more expensive that metallic ones. Moreover, the growth rate with RF sputtering is considerably smaller compared with DC power supplies working at the same power. Finally, it is difficult to control and monitor individually voltage and current. Another possibility for depositing a compound is the use of reactive magnetron sputtering. It consists in the sputtering of a metallic target, like in traditional sputtering, but in a reactive gas mixture formed by Ar and reactive species (e.g. O2, N2, CH4….). The reactive gas reacts at the film surface with the metallic atoms to form the desired compound. The amount of the reactive gas is controlled in order to produce a fully reacted compound.
UV Visible Near-Infra Red (UV-VIS-NIR) Spectroscopy
UV-VIS-NIR spectroscopy is a very useful non-destructive optical characterization technique based on the analysis of the intensity of the light reflected or transmitted by a solid to obtain information on the optical and electronic properties of the material. When electromagnetic radiation reaches a solid, a fraction is reflected by the surface; the reflectance (𝑅) of a material is defined as the ratio of the reflected and incident light intensities (𝐼𝑅𝐼0⁄). Another part of the light is transmitted through the material and the transmittance (𝑇) is defined as the ratio of the transmitted to incident light intensities (𝐼𝑇𝐼0⁄). Finally, a part of the light is absorbed by the film, and the absorbance (𝐴) is defined in analogy with the previous cases. From the law of conservation of energy: 𝐼0=𝐼𝑅+𝐼𝑇+𝐼𝐴 (2.6).
Photoluminescence spectroscopy
Photoluminescence (PL) spectroscopy is a nondestructive and versatile characterization technique that is widely employed to study the electronic structure and defect levels in semiconductor materials. In this technique light is shined onto the material under analysis and electrons are excited to higher energy levels. When the electrons return to their equilibrium state, the excess energy is emitted in the form of light if the process is radiative. The energy emitted by these electrons is given by the difference between the states involved in the transition, so the wavelength emitted by the material gives us information of its electronic structure.
Scanning Electron Microscopy
Scanning electron microscopy (SEM) is a technique used for the acquisition of high resolution topographic images of the samples. A high-energy electron beam is accelerated by a voltage of a few kV and focused on the sample surface by a system of magnetic lenses. The electron beam scans the sample surface by moving along a path of parallel lines [145]. Due to the interactions of the electron beam with the sample, different signals are produced such as: secondary electrons, backscattered electrons, characteristic X-rays, visible light, among others (Figure 2.18 a). Secondary electrons (SE) originate due to inelastic interactions between the primary electron beam and the sample atoms. Since they are emitted with low energy, only those originated close to the sample surface can escape form the sample, which allows the acquisition of high resolution images. Backscattered electrons (BE) are a part of the primary electrons that have been reflected by the sample surface due to elastic collisions with the sample atoms. They originate from a wider region within the interaction volume compared with SE. The signal produced by BE can also be used to obtain images, although the SE are better for obtaining topographic images and the BE to show differences in the composition of the sample through
chemical contrast.
Table of contents :
1 General Introduction
1.1 Basic Properties of Cu2O
1.1.1 Atomic Layer Deposition of Copper Oxide Thin Films
1.2 Basic Properties of ZnO
1.3 All-oxides Photovoltaics
1.3.1 Cu2O-based Solar Cells
1.3.2 Cu2O/ZnO Heterojunctions
1.4 Noble Metal Nanoparticles
1.4.1 Localized Surface Plasmon Resonance
1.4.2 Effects of Size, Shape and Interparticle Distance on the LSPR Signal
1.4.3 Hot-electron Generation
1.5 Photodetectors
1.5.1 Hot-carrier Photodetectors
2 Experimental Techniques
2.1 Introduction
2.2 Atomic Layer Deposition
2.2.1 Basic Principles
2.2.2 Chemisorption Mechanisms and Self-limiting Reactions
2.2.3 ALD Temperature Window
2.2.4 Experimental Setup and Deposition Parameters
2.3 Sputtering Deposition Process
2.3.1 Basic Principles
2.3.2 Magnetron Sputtering
2.3.3 Reactive Magnetron Sputtering
2.3.4 ZnO and Al-doped ZnO Thin Films Deposition
2.3.5 Cu2O Thin Film Deposition
2.4 X-ray Diffraction
2.5 UV Visible Near-Infra Red (UV-VIS-NIR) Spectroscopy
2.6 Ellipsometry
2.7 Photoluminescence spectroscopy
2.8 Scanning Electron Microscopy
2.9 Transmission Electron Microscopy and Electron Energy Loss Spectroscopy
2.10 X-ray Photoelectron Spectroscopy
2.11 Hall Effect and 4-point Probe Method
2.12 Conductive Atomic Force Microscopy
2.13 Current-density Characteristics
3 Local Structure and Point-Defect-Dependent Area-Selective Atomic Layer Deposition of Copper Oxide and Metallic Copper Thin Films
3.1 Introduction
3.2 Microstructure and Morphology of the Cu2O and Cu Thin Films
3.3 Temperature-driven Selective Deposition of Cu and Cu2O on α-Al2O3
3.4 Cu2O Deposition on Monocrystalline ZnO Substrates
3.4.1 ZnO Polar Surfaces
3.4.1 ZnO Non-Polar Surfaces
3.5 Mechanism for the Area-Selective ALD of Cu2O and Cu
3.6 Fabrication of p-Cu2O/n-ZnO Microjunctions
3.6.1 Electrical Characterization of the Microjunctions
3.7 Chapter Conclusions
4 Tunable Localized Surface Plasmon Resonance and Broadband Visible Photoresponse of Cu Nanoparticles / ZnO Surfaces
4.1 Introduction
4.2 Evolution of the NP with the Number of ALD Cycles
4.3 Metallic Cu Thin Film
4.4 Localized Surface Plasmon Resonance
4.5 Cu/ZnO Schottky Diode
4.6 Photo-response of Cu NP under Visible Light
4.7 Preliminary Results on the Deposition of Cu NP on ZnO Nanowires
4.8 Chapter Conclusions
5 Fabrication of Semi-Transparent p-Cu2O/n-ZnO Thin Film Heterojunctions for Photovoltaics and Photo-detecting Applications
5.1 Introduction
5.2 Structural, Morphological and Chemical Characterization
5.3 Optical Characterization of the Cu2O and ZnO Films
5.4 Electrical Characterization of the Films
5.5 Electrical and Optical Characterization of the Heterojunctions
5.6 Characterization of the Sputtered Cu2O films
5.7 Improving the Electrical Properties of the Heterojunction
5.8 Preliminary Results in the Cu2O Deposition on ZnO Nanowires
5.9 Chapter Conclusions
Conclusions and Outlook
Appendixes
Bibliography
