Ultrafast Momentum-Resolved Free-Electron Probing of Optically Pumped Plasmon Thermal Dynamics

Current advances in ultrafast electron microscopy make it possible to combine optical pumping of a nanostructure and electron beam probing with sub-Angstrom and femtosecond spatiotemporal resolution. We present a theory predicting that this technique can reveal a rich out-of-equilibrium dynamics of plasmon excitations in graphene and graphite samples. In a disruptive departure from the traditional probing of nanoscale excitations based on the identification of spectral features in the transmitted electrons, we show that the measurement of angle-resolved, energy-integrated inelastic electron scattering can trace the temporal evolution of plasmons in these structures and provide momentum-resolved mode identification, thus avoiding the need for highly monochromatic electron beams and the use of electron spectrometers. This previously unexplored approach to study the ultrafast dynamics of optical excitations can be of interest to understand and manipulate polaritons in 2D semiconductors and other materials exhibiting a strong thermo-optical response.

Nanoparticle sensitisation of solid-state nanocrystalline solar cell

Robert Plass

Over the past fifteen years dye-sensitised nanocrystalline solar cells have been the subject of intense research and development efforts. These systems provide a technically and economically credible alternative to classical p-n junction solar cells, reaching over 10 % certified efficiency under standard solar illumination conditions (AM 1.5, 1000 W/m2). Recently, the liquid electrolyte, commonly used in these dye-sensitised solar cells, could successfully be replaced by a novel solid hole-conducting material (spiro-OMeTAD). The absence of volatile solvents and corrosive components like iodine presents a clear advantage of these solid-state devices over their photoelectrochemical counterparts. Yet, their maximum overall efficiency of about 3 % clearly lacks behind the performance features of classical dye-sensitized solar cells. The objective of this work is the replacement of the usually used sensitiser (transition metal complexes and organic dyes) by quantum-dots. It was motivated by the possibility to achieve panchromatic light absorption due to the size-dependent properties of the quantum-dots. Some studies have been published using quantum-dots of inorganic low-bandgap semiconductors as sensitisers for mesoporous wide bandgap semiconductor electrodes in conjunction with liquid electrolytes. These systems often suffer from corrosion and photo-corrosion mainly due to the aggressive nature of the electrolyte employed. Solid-state organic-inorganic heterojunctions might provide the desired environment for quantum-dot sensitisation, as the environment is much less aggressive. A large variety of inorganic quantum-dot materials like PbS, CdS, PbSe and CdSe were scrutinized for their use as sensitisers. All particles were synthesised in situ on the TiO2 surface using two different techniques: dip-coating and chemical bath deposition. Lead sulphide was the most investigated due to its superior photovoltaic characteristics. High Resolution Transmission Electron Microscopy (HRTEM) of PbS sensitised TiO2 electrodes fabricated via dip-coating clearly showed the distinct particles on the surface of the semiconductor. The electron transfer dynamics following optical excitation of quantum-dot sensitized TiO2/spiro-OMeTAD heterojunctions were thoroughly studied. Fluorescence spectroscopy was used to prove the possible electron injection from the PbS into the TiO2. Nanosecond laser spectroscopy was applied to monitor the interfacial recombination of the injected electron and the oxidised hole-conductor. Femtosecond laser spectroscopy allowed measuring the ultra-fast kinetics in the system, including electron trapping, exciton recombination and PbS regeneration. An overall efficiency of 0.5 % at 0.1 sun (AM 1.5) has been reached with such systems. Chemical bath deposition (CBD) of PbS leads to the formation of a layer rather than distinct particles on the TiO2. Solar cells fabricated via CBD exhibited open-circuit voltages which were generally 100 mV higher then those of comparable devices made via dip-coating. The PbS layer acts as blocking layer, hindering the contact between TiO2 and the hole-conductor. This technique allowed achieving devices highly linear with respect to the illumination power. Different strategies were pursued to impede interfacial recombination, among which the introduction of self-assembled monolayers at the interface between the organic and inorganic phases was proven to be the most efficient. Nanosecond laser spectroscopy showed a strong decrease of the interfacial recombination kinetic rates in the presence of hexadecylmalonic acid or decylphosphonic acid monolayers. Short-circuit currents could be raised by a factor two to four using such type of molecules. The overall efficiency of the device could be strongly increased, reaching 1 % at 0.1 Sun (AM 1.5). CBD was also used to deposit PbSe and CdSe layers at the surface of TiO2. Fluorescence spectroscopy was used to probe electron injection from CdSe quantum dots to TiO2. Generally it was observed that metal sulfides were more efficient in sensitizing TiO2/OMeTAD heterojunctions then the respective selenides.

EPFL2004

Realization and study of InAs/GaAs quantum dot superluminescent diodes emitting at 1.3µm

Marco Rossetti

A quantum dot is a semiconductor nanostructure that confines the motion of conduction band electrons and valence band holes in all three spatial directions, thus creating fully discrete energy levels. The confinement in the InAs/GaAs material system is generated by the bandgap difference for the two materials (∼ 0.4 eV for InAs, and ∼ 1.4 eV for GaAs) which provides a minimum of potential energy for both electrons and holes inside the InAs nanoparticle. The appearance of new nanoscale effects modifies the electro-optical properties of such a system which makes possible the improvement of existing device performance or even fascinating novel device applications. For the creation of coherent structures with very high purity and high density, as it is required for application in light emitting devices, self-assembling is a very important pathway. The quantum dot creation mostly relies on the relaxation of the strain accumulated during the growth of lattice mismatched materials, which happens in a non-equlibrium condition. This results in a statistical distribution of all the parameters characterizing the InAs nanocrystals as size, composition and strain, whose direct manifestation is the inhomogeneously broadened light emission deriving from quantum dot ensembles. At the dawning of the quantum dot development the broadening was considered as a limiting factor for application in semiconductor lasers. Today, it is considered favorably for all the applications where a broad gain spectrum is required as optical amplifiers, monolithic and external-cavity tunable lasers, and superluminescent diodes. The objective of this thesis is the application of InAs/GaAs self-assembled quantum dots to the active region of high power and broad-band superluminescent diodes emitting in the 1.3 μm wavelength region. Superluminescent diodes are edge-emitting semiconductor light sources based on the amplification of spontaneous emission along a waveguide where optical gain is achieved through current injection. They combine the high power and brightness of laser diodes with the low coherence of LEDs. The latter is a fundamental feature for the application in optical coherence tomography medical imaging, where the image resolution is related to the spectral bandwidth of the light source used. The achievement of bandwidths larger than 100 nm are demonstrated in this work through the optimization of molecular beam epitaxy growth conditions. This may be done optimizing the statistical size-dispersion of the dot ensemble, or through the use of different dot layers emitting at slightly different wavelengths. The large bandwidth emission is obtained also due to the peculiar energy structure of this material, for which a multi-state emission appears under particular injection conditions. As a term of comparison, best commercial single-chip superluminescent diodes for the imaging application at 1.3 μm show bandwidths around 60 nm. Moreover, the discrete nature of the energy levels in InAs quantum dots together with the implementation of a GaAs-based technology can be beneficial for the device temperature stability (the competing technology, based on InP, suffers from strong thermal degradation). The analysis and understanding of the device temperature characteristics will be detailed in the last chapter of the manuscript. Even though standard devices show an important temperature dependence, a better stability may be achieved through the use of p-doped active regions. In the last few years quantum dot devices have shown outstanding properties if compared to the competing quantum well technology. Low threshold currents, high temperature stability and low chirp in lasers, large bandwidths and low gain saturation in amplifiers have been demonstrated. However, in spite of the large interest among the scientific community, many of the microscopic processes at the origin of the electro-optical characteristics of this material are not completely understood yet. The physics of the quantum dot active material, will be modeled in this work through the use of systems of rate equations with different complexity depending on the system they are applied to. We will provide a mean-field rate equation model allowing to get insights about carrier dynamics in QD lasers. Also, we will develop two different traveling-wave rate equation models, one for the modeling of the L-I characteristics and the other for the modeling of the spectral characteristics of quantum dot superluminescent diodes. Considering the carrier and photon density distributions across the device cavity is essential in superluminescent diodes, where the high single pass gain results in large non-homogeneities of photon and carrier distributions. The work is motivated by an industrial cooperation with EXALOS AG, a leading company in the development, manufacturing, and sales of broadband superluminescent diodes.

EPFL2007

Investigation of Two-Dimensional Electron Gases with Angular Resolved Photoemission Spectroscopy

Isabella Gierz

In this thesis we study the electronic structure of different two-dimensional (2D) electron systems with angular resolved photoemission spectroscopy (ARPES). This technique is based on the photoelectric effect and directly probes the electronic structure of a system. By carefully analyzing the measured band structure with respect to peak position and line width we can determine the complex self-energy Σ that describes the renormalization of the electron's energy and the change in lifetime due to many-body interactions. The 2D electron systems investigated in this work are surface alloys on Ag(111), bismuth trimers on Si(111) and epitaxial graphene monolayers grown on SiC(0001). Surface alloys on Ag(111) are formed by depositing 1/3 of a monolayer of bismuth, lead or antimony (alloy atoms) on the clean silver surface. Although (Bi,Pb,Sb) and Ag atoms are immiscible in the bulk they form long-range ordered surface alloys, where every third Ag atom is replaced by an alloy atom. These systems as well as the Bi trimers on Si(111) show a spin splitting of the 2D band structure due to the Rashba-Bychkov (RB) effect. The RB model states that in a symmetry broken environment (such as the surface of a semi-infinite crystal) the spin-orbit interaction will lift the spin-degeneracy of the band structure. Such a spin-split band structure bares great potential for applications in the field of spintronics, e.g. in a Datta-Das spin field effect transistor. In the present work we investigate the origin of the observed giant spin splitting in surface alloys, especially the interplay between structural parameters and the atomic spin-orbit interaction. Furthermore, we will show that it is possible to transfer these concepts to a semiconducting substrate, which is better suited for spintronics applications. The third system under investigation — graphene — is an ideally two-dimensional crystal. It consists of a single layer of carbon atoms arranged in a honeycomb lattice, and its charge carriers are confined within a plane that is just one atom thick. These charge carriers behave like massless Dirac particles and possess extremely high carrier mobilities. This makes graphene a promising material system for high-speed electronic devices. In order to reach this ambitious goal one needs reliable methods for the large-scale production of high quality graphene films. Epitaxial growth on silicon carbide (0001) substrates is the method of choice in this case, as it offers the advantage of a precise thickness control and a semiconducting substrate at the same time. However, the presence of the substrate reduces the carrier mobility of graphene's charge carriers considerably. Therefore, it is necessary to decouple the graphene layer from the substrate after epitaxial growth. A second issue that needs to be addressed, are viable doping methods for graphene. As graphene's peculiar band structure results from a sensible interplay between electrons and crystal lattice it is not an option to replace single atoms of the graphene lattice by dopants as is common practice when doping silicon. In order to preserve its band structure, graphene is usually doped by adsorbing atoms or molecules on its surface. As graphene grown on SiC is n-doped due to charge transfer from the substrate, appropriate means for p-type doping are clearly required. In this thesis, we will present a new growth method for quasi free-standing graphene on SiC(0001) and viable means for p-type doping.

EPFL2011