Gallium Nitride (GaN) based materials have emerged as the leading technology for a wide range of optoelectronic applications in the visible range. In particular, blue emitting lasers and light emitting diodes (LEDs) have led to the development of widely used “white LEDs” and “Blu-Ray” technologies. A next step, which represents one of the most important challenges for nitride LEDs, is to replicate in the ultra-violet (UV) range, i.e. below 360 nm, the performances obtained in the blue. Indeed, AlxGa1-xN materials, which allow covering an emission spectrum between 360 nm and 200 nm, are well adapted to become the next technology in the UV, in replacement of the mercury-vapour lamps, which are hindered by environmental issues (toxicity, hazardous waste) and technological limitations (large size equipments, low efficiency and lifetime, etc.). However, the internal quantum efficiency (IQE) of UV LEDs – which is given by the product of the radiative recombination efficiency (RE) and the injection efficiency (IE) - is rapidly decreasing when going towards shorter wavelengths. Therefore, the general objective of NANOGANUV project is to develop alternative solutions to the current technologies and designs, adopted by most R&D laboratories, by focusing on two major key elements of AlxGa1-xN LED structures: - 1) the active region (on which depends the RE); - 2) the p-type region (on which depends the IE). In order to address these locks, different approaches will be developed. Presently, RE and EI are limited in AlxGa1-xN materials by high defect densities and an impediment to the doping, respectively. Regarding the doping issue, which is due to the dopant ionization energy increase with Al content, the main lock concerns the case of p-type doping. Three paths will be followed to improve the IQE: - 1) the use of quantum dots (QDs), grown by Molecular Beam Epitaxy (MBE, the most mature epitaxial technique for QD fabrication). Indeed, owing to the spatial confinement of carriers in 3 dimensions (D) instead of the 1D spatial confinement in the case of quantum wells, QDs strongly improve the RE by reducing the influence of defects; - 2) the optimization of Mg doping in AlxGa1-xN layers by MBE, for which p-type GaN doped layers have been demonstrated with the largest hole concentrations (close to 10^19 cm^-3). In order to reach the level of optimized doping conditions and associated electrical characteristics, the fabrication of a high-stability dopant evaporation cell dedicated to Mg will also be developed; - 3) the use of AlN bulk substrates and the development of a specific high-temperature growth furnace to improve the structural quality of AlxGa1-xN. These two approaches will allow determining the potential of MBE to reach high-quality AlxGa1-xN layers. The overall ambition is to develop a novel route towards the fabrication of efficient UV LEDs, by using alternative scientific and technical solutions at both the nanoscale level, i.e. involving QD modelling, fabrication and quantum engineering, and the micro/macroscopic level i.e. investigating optical and transport properties, to identify, design and assemble the building blocks for the fabrication of QD-based UV sources. The final targets are to design and fabricate UV LEDs operating in the 260 – 360 nm spectral region. This large UV range (from UV-A to UV-C regions) should allow addressing a wide range of applications, from UV curing and counterfeit analysis to medical phototherapy, water and air purification. At the end of the project, devices presenting the best performances will be further processed in LED packages with the aim of performing a series of tests on experimental workbenches by companies specialized in LED testing for UV applications (which will be defined by the specific UV region covered by the LED prototypes).

High frequency/high power microwave devices can contribute to answer to the increasing demand in terms of security, sensors and communications. The aim of this project is to demonstrate GaN based devices on Silicon substrate with high output power density and efficiency as well as reduced propagation losses opening the way to the fabrication of Monolithic Microwave Integrated Circuits (MMICs). Substrate choice is motivated by the low cost and the availability of Silicon in large size, potentially leading to process cost reduction. However, the starting material necessarily needs to satisfy requirements in terms of structural and electrical quality, which are still yet difficult to reach in production environment. Indeed, Metal-Organic Chemical Vapor deposition (MOCVD) is the preferred tool for the crystal growth of GaN. It has led to high-quality GaN, large size production tools (up to several times 8 inch wafers capacity) but the high temperature involved in the growth process is a drawback for the growth on Silicon which necessitates trade-offs depending on the target application. A first objective will be to develop such knowhow not available in the literature to demonstrate MOCVD grown GaN high electron mobility transistors (HEMTs) on Silicon for power applications at 40 GHz and beyond. First, the project will focus on the reduction of microwave propagation losses through GaN based buffers and various templates on highly resistive silicon. An optimization loop will be necessary to satisfy this criterion as well as others such as crystal quality and low strain to avoid cracks in the films. The project is leaded by CRHEA which is in charge of crystal growth. With the help of its partners GREMAN and IEMN, the critical interface between the nucleation layer and the substrate will be studied. The growth conditions for the nucleation layer will be monitored to satisfy the requirements for a high electrical resistivity interface and buffer layer. The state of the art results previously obtained with the molecular beam epitaxy (MBE) technique will be a reference for this work. More, nucleation at low temperature with MBE will be investigated as an alternative solution. Compared to Silicon, the larger band gap and the chemical inertness of cubic Silicon Carbide are advantages. The alternative solution consisting in an intercalated Silicon Carbide buffer layer grown on Silicon will be evaluated. To benefit the advantage of the small transit time of a large carrier density under a short gate (less than 100 nm) and then to get a good aspect ratio, the transistors must present a small distance separating the carriers from the gate (thin barrier) as well as small access resistances. Today, the trade-off between barrier thinning and resistance increasing has led to thicknesses of the order of 15 nm and sheet resistances around 300 Ohm/sq in the case of AlGaN barrier. This induces a noticeable access resistance in the device which becomes the main limitation for operation beyond 30 GHz. The aim of the second task of the project will be to develop solutions to mitigate this limitation. To do so, the Aluminum content in AlGaN and InAlGaN barriers will be increased to enhance the carrier density and to reduce the sheet resistance. The barriers will be protected by a GaN cap or by in-situ grown SiN able to passivate the surface. More, in order to further reduce the access resistances, N+ doped GaN contacts will be selectively regrown by MBE in growth windows after etching of the barrier. This will avoid the high temperature annealing of the contacts and then permits to get abrupt edges of the ohmic contact, facilitating the fabrication of transistors with gates very close to the source. In order to validate the HEMT structures grown with low propagation losses, IEMN will perform the technology developments and the device electrical characterizations, including the load pull measurements of power performance at 40 GHz and beyond.

Gallium Nitride (GaN) devices are foreseen to play a major role in next generation of power electronic applications. This is due to its outstanding material properties and cost-effective manufacturing when grown on silicon substrate. Thus, GaN-based power switches have the potential to enable high efficiency connection of renewable energy sources to the electricity grid. The new technology would enable higher efficiency and less complexity as well as being light-weight with greater functionality, robustness and the ability to operate in a wide ambient temperature range. The ultimate goal for renewable energy companies is to supply/interface their energy to the national grid with minimal loss. This will ultimately mean there will be less demand for energy derived from fossil-fuel sources, and so this way will protect the environment from CO2 emissions. GaN devices are expected to reliably operate at elevated junction temperatures up to at least 225°C (presently used Si-based devices cease to function at ~150°C), easing the constraints on current cooling requirements. Similarly, GaN-based power conversion circuits can operate at higher efficiencies and high frequencies enabling compact converter and inverter designs, up to a 10? reduction in size, cost and weight. This will translate into significant energy saving (~10%), overall cost reduction, increased adoption of renewable energy sources, and improvements in profit for the renewable energy companies. In this frame, we have developed a new concept enabling to boost significantly the GaN-on-silicon device breakdown voltage above 3000 V. The key feature of this concept lies in a backside local removal of the silicon substrate around the drain electrode. One of the main challenges of power devices is the thermal management. This project aims at developing an innovative thermal management solution integrated within the backside trenches, which should generate unique substrate grounded GaN power devices operating at voltages far beyond existing GaN-based devices. Four public institutions IEMN, ESIEE, CRHEA and LAAS will combine their skills to reach this ambitious goal that would lead to a real technological breakthrough for power applications. The DESTINEE research effort addresses key critical components based on the emerging GaN material for next generation of power electronics. It will benefit from existing collaborations with partners that have leadership in their respective domain. The project covers a large added value chain from epitaxy, integrated device technology developments, to prototype device characterisation and preliminary reliability.

In today’s fast evolution and expansion of wireless communications, the GaN HEMT is a good candidate, taking advantage of its high frequency performance, high breakdown voltage and material robustness. For the next generations of RF power amplifiers, GaN-based technology is the most promising to satisfy more demanding specifications of the market. The standard GaN-HEMT is a depletion mode device (normally-on). RF circuit designers claim for normally-off devices (enhancement-mode) integrated with depletion-mode ones in the same chip to reduce circuit complexity and power consumption. According to literature, many research teams have worked in the development of enhancement-mode devices by different means. Among many proposed methods to deplete the channel at gate bias = 0 V, the most promising for device commercial production are: A) Fluorine (F- ions) implantation in the upper part of GaN barrier layer to deplete the channel B) Gate recess on a thick barrier layer C) Etching of the barrier layer using an etch-stop layer D) Local epitaxy of p-GaN layer on the top of the barrier layer. All of these aforementioned techniques are currently being investigated in research labs. Each technique has its own advantages and disadvantages. So far, it is not clear which of them (A-D) is the one that provides the best combination of device performance, reproducibility and reliability. Furthermore, the monolithic integration of enhancement and depletion mode GaN-based HEMTs, fabricated on the same wafer is a great challenge. The reason is that the epitaxial structures are generally designed to foster one of the two modulation modes (enhancement or depletion), which are sometimes conflicting. This situation generates risks for any foundry that wants to bring the integrated ED-mode devices into the RF market without making the choice of the best depletion-mode technology. The objective of this project is integrating high performance, reproducible and reliable enhancement-mode GaN HEMTs with the depletion-mode devices on the same chip. The consortium will perform an assessment of these four depletion mode approaches (A-D) to bring RF device innovation based on a nonpartisan investigation. To accomplish this, the industrial partner (OMMIC) and the four research institutions (LN2, IEMN, CRHEA, IMS) have designed a 42 month scientific program that includes device design, epitaxial growth, microfabrication, performance evaluation and reliability testing. Then, the consortium will select the best integrated enhancement-depletion mode approach answering OMMIC’s specification. This path represents a major difference compared to what is presently reported at the state-of the art. After the project is completed, the technology will be transferred to OMMIC’s foundry. One of the most important advantages of this project is that the most promising technologies to fabricate ED-mode GaN HEMTs are going to be tested under the same conditions, reducing the risk of making an inadequate technology choice. Therefore, OMMIC will be in a winning position and will have an excellent opportunity to make a breakthrough, becoming a world-class leader in GaN RF circuits. The commercial success of the product will have a positive impact to the manufacturer (OMMIC) and on upstream and downstream companies, allowing the creation of new jobs and prosperity in the related sectors. Defense, health equipment, consumer electronics, the automotive industries are vast markets for this technology. They are strategically important for economic development, security, safety and well-being of our society. Furthermore, due to the highly experimental characteristic of this project and the exceptional capabilities of the consortium, the project will produce an outstanding scientific impact in the wide-bandgap semiconductor domain.

The terahertz (THz) frequency domain, between infrared and microwave spectral range, is referred as the « THz gap » because of the lack of compact and efficient semiconductor devices. Exploration of this spectral region is considerably developing due to the wide range of applications: medical diagnostic, security, detection of molecules, astronomy, non-destructive control of material, high data rate secured communications in open space… One pre-requisite for scientific applications like spectroscopy, astrophysics, space imaging, is the availability of compact, fast and low-noise detectors. A necessary large detectivity imposes the use of cryogenic cooling for the detector and the main challenge is to increase the working temperature. However, for a widespread use of the THz technology, compact sources working at non-cryogenic temperatures are necessary. The OptoTeraGaN project is intended to tackle both issues, namely the development of high-detectivity THz quantum detectors with increased operating temperature with respect to existing technologies as well as THz light emitters, both based on the quantum cascade (QC) concept and on high-quality polar and semipolar GaN/AlGaN semiconductors. We intend to benefit from the large energy of optical phonons in GaN materials for demonstrating QC devices in a much broader spectral range (1-15 THz) and in particular in the 5-12 THz range, which cannot be covered with other III-V semiconductors. The large optical phonon energy is also one key point for increasing the operating temperature of THz QCD and for achieving room temperature operation of THz QC lasers, which appears to be out of range of current GaAs-based QC laser technology. The first objective of the project is to demonstrate THz quantum cascade detectors (QCD). QCDs are formed by the repetition of active and extractor quantum well (QW) regions and rely on intersubband (ISB) absorption in the active QW and photo-excited electron transfer through the extractor from one period to the other. In contrast to existing THz quantum detectors such as GaAs QWIPs, these photovoltaic devices operate under zero bias and do not suffer from any dark current, which is one main advantage for increasing the operation temperature while benefiting from the maximum detectivity limited by the background (BLIP). Our target is to demonstrate THz QCDs with a responsivity larger than 100 mA/W and a BLIP temperature of 77 K at 12 THz. A second related objective of the project is to make significant progress towards THz QC lasers in the GaN/AlGaN material system. We will make use of the advanced know-how acquired on the design, growth and processing of GaN-based THz QCD devices to develop electroluminescent sources. One first goal is to develop spectrally narrow THz light emitting devices at room temperature based on in-plane transport, which can find a number of applications because of their fast modulation capabilities. Our final target is to demonstrate stimulated gain under vertical transport using plasmonic waveguide resonators and lasing at cryogenic temperatures. The consortium, which regroups teams with a world-class expertise on GaN-based material growth by MBE and MOVPE, GaN ISB devices, as well as on QCL/QCD and THz technology and characterization, has been specifically assembled to meet the objectives of this basic research project. The recent demonstration by members of the consortium of GaN-based ultrafast QCDs in the mid-IR spectral range as well as the first observation of reproducible resonant tunnelling and THz ISB electroluminescence from GaN/AlGaN QWs are preliminary results, which constitute the major building blocks required for the present project. The project OptoTeraGaN is of major technologic and scientific impact in agreement with “défi n°7 société de l’information et de la communication: micro et nanotechnologie, optoélectronique”.