Croissance des COuches Semipolaires de semiconducteurs NItrures de qualité optimale pour applications opto-électroniques – COSNI

In order to carry out the objectives of COSNI, the growth of layers and heterostructures was performed by the two main national laboratories in the field of nitrides (CEA-INAC and CRHEA). The three growth technologies providing better material quality (hydride vapour phase epitaxy, metalorganic vapour phase epitaxy and molecular beam epitaxy) have been applied to the growth of semipolar nitride layers. A strongly correlated effort between partners involved in growth and characterization has permitted a rapid optimization of the material synthesis. By the middle of the project, the doping and the layer quality was mastered enough to initiate the realization of devices for visible emission and nanostructures with infrared intersubband absorption.

During the COSNI collaboration, the consortium has set up an original process for the growth of semipolar nitride layers, namely the asymmetric epitaxial lateral overgrowth. This has allowed us to decrease the defect densities by more than two orders of magnitude. We also identified the mechanism which stops the basal stacking faults during this type of growth, and showed that prismatic stacking faults play a critical role in the process. The doping of nitride layers along the semipolar direction has been optimized and heterostructures with emission up to the yellow range have been fabricated. Regarding ternary alloys, we have noticed differences in the incorporation of Indium in comparison to the polar [0001] orientation. By MBE, indium incorporation is inhibited, whereas it is enhanced by MOVPE. Moreover, we have observed a systematic phase separation in MBE AlGaN (11-22) as soon as the Al composition is brought above 10%. Finally, we have observed for the first time intersubband absorption from semipolar nanostructures.

During the 42 months of the COSNI project, the partners have investigated the possibility that are to be brought about by the use of semipolar nitrides for the visible emission as well and as the intersubband transitions from the infrared up to THz. A very close collaboration has been under operation for the characterization of the materials and devices. Thick semipolar GaN nitride layers have been produced and even large pieces of auto supported GaN up to a few centimeters have been obtained. A the end of this project, it is obvious for our partnership that ways should be found in order to continue this research effort. Indeed, it critically important that the acquired knowledge is rapidly exploited in order to be able to fabricate nitride based efficient emitters past the green as well as the THz range.

The COSNI project has given us an opportunity to make 17 contributed original publications in international scientific journals with high impact factor and a book review chapter on nitride semiconductors. It also has allowed the partners to fund 27 attendance to scientific conferences and present the project results, amongst these , 10 were upon explicit invitation. During this period of collaboration, eight seminars have also been given by the partners, at Uinversities and in the industry. Moreover, COSNI has served as an important scope fr the prepartion of three PhD research works which all went on perfectly and the candidates presented thay examination during our collaboration.

III-nitride semiconductors crystallizing in the wurtzite structure exhibit a strong spontaneous polarization along the [0001] direction (c-axis), in addition to any polarization that may arise as a consequence of their piezoelectric properties. In GaN-based quantum well structures grown along the c-axis, the polarization discontinuity can cause electric fields of up to 10 MV/cm, leading to low radiative recombination efficiency, redshift and broadening of the luminescence lines, and anomalies in vertical carrier transport. These effects can be reduced by choosing growth on nonpolar m-(10-10} or a-(11-20) surfaces. However, growth of nonpolar III-nitrides remains challenging due to the strong anisotropy of the surface properties, resulting in epitaxial layers with an extremely high density of crystalline defects. An alternative approach is the growth on semipolar surfaces, i.e. those (hkil) planes with nonzero h, k or i, and l indices. The target of this project is to assess the possibilities of a semipolar nitride technology, particularly using the (11-22) crystallographic orientation, for optoelectronic applications both in the visible (interband) and in the infrared (intersubband) spectral ranges. To face the challenge of the material quality, the project involves the three main growth techniques available for III-nitride materials (metalorganic vapour phase epitaxy, hydride vapour phase epitaxy and plasma-assisted molecular-beam epitaxy), within a strongly correlated experimental and modelling effort. By the end of this project, it is forecast to produce high quality crystalline semipolar GaN thin films, nanostructures, and test devices: electroluminescent diodes and quantum well infrared photodetectors.