Echéancier

Echéance du projet

PHASES

The ambition of NANOGANUV project is to define and develop strategic building blocks for the fabrication of UV LEDs emitting in the 260 – 360 nm range. The project is planned on a four year schedule. In order to reach the objectives, the project is divided into four workpackages (WP) corresponding to the different scientific and technical targets of the project, which are described below.

Workpackage WP1: “self-assembled Al_yGa_1-yN nanostructures: from the modelling to the growth”

This phase will focus on the modelling, the growth and the structural characterization of superlattices (SLs) of self-assembled GaN and Al_yGa_1-yN-based nanostructure arrays on Al_xGa_1-xN alloys. Indeed, semiconductor SLs are commonly used in a number of modern optoelectronic applications. From a fundamental point of view, the modelling, fabrication, structural and optical characterizations of self-assembled SL of Al_yGa_1-yN nanostructure arrays have never been associated. In the case of GaN, it is known that the 2D-3D strain-induced transition leads to different nanostructure shape (quantum- dots, wires, wells). Consequently, it is of prime importance to identify the factors (thermodynamics, kinetics, elasticity) leading to the shape, size and density of the Al_yGa_1-yN nanostructures (which can directly impact on their optical properties). The introduction of artificial periodicities by repeating a pattern of layers of different materials modifies the electronic band structure and the electron transport properties, thereby providing an approach to fine-tune material properties. For most applications, the ideal SL consists of a periodic array of uniform distinct layers separated by sharp, atomically flat, coherent interfaces. However, misfit elastic stress causes deviations from this ideal structure, due to the elastic relaxation, which commonly occurs during the growth process. The origin of elastically driven relaxations in a multilayer film lies in the instability of a film under stress with respect to the formation of shape perturbations, namely the Asaro-Tiller-Grinfeld instability. As each layer of material composing the multilayer film is deposited, the growing surface of the film develops an undulating profile. In the case of single-layer films, there is extensive experimental evidence of this phenomenon. As more layers are deposited, the undulated surfaces become buried under subsequent layers of materials, resulting in wavy pattern interfaces. These undulations can be seen as quantum dots (QDs) which stand on a wetting layer. These structures can be exploited to form self-organizing QDs for UV light emissions. In this project, we will apply the theory of continuum elasticity and growth to the case of strained semi-conductor Al_xGa_1-xN multilayers. We will investigate the effect of the misfit between the lattices, the effect of surface diffusion and of the flux, on the morphology and the spatial distribution of the QDs. Our goal is to find the experimental parameters (flux, misfit, etc.) which lead to a maximum QD density while being below the threshold of plastic relaxation in order to avoid the creation of dislocations. In particular, we will investigate the effect of the surface diffusion coefficient (of the different atomic species) on the position of the QDs in the successive layers (of the SL) since the position and the shape of the QDs are important factors for UV light emission. In order to model the growth and the morphology of the QDs which appears in the multi-layer system, we will use the techniques and methods of continuum elasticity and semi-conductor growth which we have developed in Nice (INLN) and Paris (INSP). Also, first models on the GaN QD formation have started to be developed. Based on our experience, we plan to produce results which will be compared to experiments at CRHEA on Al_yGa_1-yN QDs for UV LED applications. In order to fabricate the nanostructures, we will take advantage of the possibility to form self-assembled Al_yGa_1-yN nanostructures on Al_xGa_1-xN surfaces by using molecular beam epitaxy (MBE). The lateral and vertical spatial organization of the SLs with nanostructure arrays will be investigated, aiming at the precise control of the main growth parameters in order to obtain high densities (in the 10¹¹ cm^-2 range) and small size (< 2nm high) nanostructures with minimized size dispersion and composition inhomogeneities.Haut de page

Workpackage WP2: “fundamental properties of Al_yGa_1-xN materials”

This phase will concern the investigation of the fundamental (including material and optoelectronic) properties of Al_xGa_1-xN materials and nanostructures. It will include studies on Mg-, Si-, and non-intentionally doped (n.i.d.) layers (grown on sapphire and AlN substrates), from the growth by MBE to the characterization by an extensive panel of experimental techniques (spectroscopy, microscopy, optical and electrical measurements).The optical properties of the nanostructures will be investigated by temperature dependent photoluminescence (PL) and time-resolved PL experiments, together with computations of the oscillator strengths of band-edge transitions using the effective mass (envelope functions) approximation taking into consideration the spontaneous and piezoelectric polarization fields. The polarization selection rules for a given excitonic resonance will be calculated for the (0001) orientation. Due to electron-hole exchange interaction, which is relatively strong in GaN, excitonic states split into dipole forbidden (spin-triplet) and dipole allowed (spin-singlet) states. The small but finite exciton wave vector can contribute to relaxation of these selection rules. Using excitonic binding energies and envelope functions obtained via a variational approach in bulk, in QWs and in QDs, the radiative decay rates for the excitons polarized along the wave-vector and polarized normal to the QW or QD plane will be calculated. The results will be compared to the experimental data which will facilitate the optimization of the theoretical model. A variety of active region structures will be explored. Doping is a prerequisite to realize electrically driven devices. The incorporation of dopants and the formation of compensating defects depend on the atomic structure of the growth materials and the growth conditions. To evaluate the incorporation of both p-type (Mg) and n-type (Si) dopants as well as non-intentional doping different experimental techniques will be applied. High temperature Hall effect, C(V) characteristic and SIMS measurements will be combined to determine the doping level and the efficiency and ionization energies of the dopants. The quality of the epitaxial layers will be estimated by the comparison of experimental mobility with the theoretical predictions. Therefore, Hall Effect and conductivity will be measured in a wide temperature range (from cryogenic temperatures to 1000K). Despite technological efforts, both epitaxial GaN and related ternary Al_xGa_1-xN layers contain different types of crystallographic defects and electrically active non-homogeneities. Revealing and analysis of defects and assessment of their role for the GaN-based device performance constitute an indispensable part of the complex manufacturing technology. The photo-assisted etching method will be applied. The broad range of defects will be revealed with no limits on the size of samples and by a very quick data analysis. Radiative recombination efficiency would be the ultimate indicator of the quality of the material as the active defects cause non radiative recombinations. This is a pivotal issue in devices relying on optical emission, such as LEDs and LDs. In this vein, we plan to study theoretically and experimentally radiative recombination rates and oscillator strengths for the band edge transitions in both case of low Al composition (the crystal field splitting parameter is positive) and in the context of high Al composition (the crystal field splitting parameter is in this situation negative). We should mention that we already have a large amount of data for a range of GaN QW and QD structures on c-plane Al_0.5Ga_0.5N barrier layers but there is a need for in depth investigation of Al rich Al_xGa_1-xN confining layers or QDs. These parameters will be explored experimentally using temperature and polarization dependent CW and time resolved optical excitation experiments at L2C (over the complete Al concentration range by using a frequency-quintupled Ti:Sa laser at 200nm). Radiative and nonradiative recombination rates will provide information on the material quality, particularly the point defect densities. Knowledge on the recombination rates will provide the necessary feedback for growth optimization. Haut de page

Workpackage WP3: “Development of MBE components for Al_xGa_1-xN material optoelectronics”

This phase will be dedicated to the design, the fabrication and the development of novel MBE components aiming at the fabrication of Al_xGa_1-xN-based active layers for UV nitride optoelectronic applications. Indeed, MBE appears to be the best-suited technology for the fabrication of Al_xGa_1-xN layers incorporating nanostructures (in particular QDs) as the active region since it offers the possibility to control the formation of self-assembled Al_yGa_1-yN nanostructures by using a strain-induced epitaxial growth mode. This type of active region represents a central key point of the project for the fabrication of high-radiative active regions and LEDs emitting in the UV range. To reach this goal, two major technical locks will be addressed: the improvement of Al_xGa_1-xN material structural quality (using sapphire and AlN substrates), and the p-doping efficiency of Al_xGa_1-xN alloys (for x > 0.1). In this view, a high-temperature growth furnace (> 1000°C) and a high-stability dopant valved evaporation cell (for Mg doping) will be developed and tested at RIBER S.A. and then installed on a MBE reactor at CRHEA. The performances of these equipments will be fully tested and compared to the characteristics of MOCVD grown Al_xGa_1-xN layers (which are also fabricated in CRHEA). This point will allow us to assess MBE grown layers in comparison with the leading industrial technology (MOCVD) which sets the benchmark (in terms of structural and optical properties). Their performances will be continuously optimized based on the characterization results and finalized through the fabrication of a major device in the nitride technology: the LED. At the end of the project, the performances of UV LED prototypes, emitting between 260 and 360 nm, will be compared to the latest results reported in the literature and international conferences. Haut de page

Workpackage WP4: “design, fabrication and characterization of quantum dot based UV LEDs”

This phase will target the design, the fabrication and the characterization of UV LEDs (on sapphire and AlN substrates) including Al_xGa_1-xN nanostructures as the active region. LEDs emitting in the 260 – 360 nm spectral range, i.e. from UV-A to UV-C regions, will be fabricated. LEDs operating in the UV-A (315-360 nm) and UV-C (260-280 nm) will be particularly targeted since these spectral regions represent the largest markets in terms of potential applications, including curing, disinfection and purification technologies. Therefore, an important aspect will concern the design of the structure, in particular the active region and the p-type layer engineering which are the main scientific and technical locks addressed in this project. The LED processing will be performed in the clean room facility at CRHEA. The final goal is to evaluate the potential of the proposed technology by determining the figures of merit of LEDs (on sapphire and AlN substrates, by comparing the ratio LED performances over the fabrication cost), and comparing them to the state-of-the-art. Along these views, to study effects of the polarization field on escaping mechanisms of injected carriers from the LED active region, we will grow LED structures with various Al-rich Al_xGa_1-xN electron blocking layers and electron/hole injector layers. Finally, if the performances of LED demonstrators at the end of the project are reaching state-of-the-art values (typically ~ 1 - 2 mW @ 20 mA for simple “on wafer” measurements of LEDs), the fabrication of LED structures (epi-wafers) and LED packaging will be performed. The next step will be to perform a series of tests on experimental workbenches by companies specialized in LED testing for UV applications (French companies BMES (69) and LEDPOWER (72) with which we are in contact with the R&D service, MM. D. Chavanon and S. Simon, respectively). Haut de page