© 2000 The Materials Research Society
This E-MRS 2000 Symposium C on Group-III nitrides has been held annually since 1996 with the only exception for the 1999.
The scientific program covered the entire field from growth and characterization to processing and devices. The very fast development of the III-nitrides, which after a very short period of laboratory research came to the stage of mass production in selected areas, was underlined in the plenary talk given by J.V. Duboz. Based on a critical analysis of the advantages and disadvantages of GaN in comparison with other semiconductors, a forecast of the scientific and technological evolution of the GaN research field was given.
The search for suitable substrates for GaN-based technology remains an important goal. Several alternatives to conventionally used sapphire or SiC substrates have been extensively discussed, including LEO quasi-substrates, high pressure grown GaN single crystalline platelets, or silicon substrates. An example is growth of cubic GaN on silicon using a cubic BP buffer. Comprehensive characterization results with high resolution CL microscopy, micro-Raman scattering, TEM and AFM, as well as optical spectroscopies were presented to evaluate the material quality. Each approach has particular advantages and disadvantages and thus can be beneficial for certain device applications.
Si, as a substrate material, is the most attractive candidate for potential integration of simple GaN-based devices into the existing device technology. The surface morphology of the GaN layer has been shown to be critically dependent on the III/V ratio. The flexibility of the MBE growth technique, were the layer structure can be easily varied from a flat layer to nano-columns has been demonstrated. A high doping efficiency demonstrated for both n-type (Si) and p-type (Be) dopants, as well as relatively high electron mobilities provide some hope for future device applications.
For the LEO growth, the significance of the mask orientation and proper choice of carrier gas for the growth process was directly visualized by using high resolution optical characterization, such as the CL and micro-Raman imaging techniques. As an alternative approach, antisurfactant-mediated growth has been suggested, resulting in 106-107 cm–2 dislocation density. The exact role of the anti-surfactant Si atoms as “dislocation-killers”, however, is not clear at the moment.
The progress in physical understanding of GaNAs with low As content was reviewed. New ODMR data were presented on the electron effective mass in this material, as well as a study of recombination processes in GaNAs/GaAs MQWs.
The recombination activity of different extended defects (threading dislocations, inversion domain boundaries, stacking faults) and their effect on device performance was analyzed in several contributions. A significant result is the reported PL efficiency in AlGaN/GaN QWs, grown on low dislocation density bulk GaN substrates as well as on sapphire. There is a definite correlation of PL efficiency with the dislocation density. Whereas threading dislocations are commonly agreed to be detrimental for the performance of laser diodes, for other defects the situation remains less understood.
Growth and characterization of the low-dimensional structures attracted increasing attention, reflected by the large number of the presented talks. Several growth approach have been successfully employed including Stranski-Krastanow growth of GaN QDs on AlN, self-assembled InGaN dot formation in the GaN matrix, anti-surfactant (Si) mediated growth of GaN dots on Ga(Al)N or more exotic sequential ion implantation of Ga and N into a dielectric matrix. Reasonable control of dot density and size uniformity has been achieved by optimizing the growth and reflects essential progress in the understanding of the strain-relaxation processes in the system. In addition to comprehensive studies of structural quality, new physical properties of the studied material systems have been underlined. In particular, strong internal polarization fields have been revealed experimentally from the time-resolved PL measurements. A theoretical analysis of the electronic structure of GaN/AlN QDs including built-in strain and electric effects was also given.
The progress in devices was discussed in a number of contributions. The problem of beam shape control for the applications of GaN LDs in optical storage systems has been addressed. The possibilities of realizing the desired index-guided structure and stabilization of the transverse optical mode by utilizing an AlGaN buried layer have been demonstrated.
An impressive improvement of the device characteristics has been reported for Schottky diodes and pin rectifies in the lateral AlGaN system, where reverse breakdown voltages up to 4.3 kV have been achieved.
The AlGaN/GaN FET performance has been also examined. The advantages of double heterostructure design as compared with single heterostructure have been clearly demonstrated by about a factor 2 increase of the 2DEG mobility, understood in terms of polarization effects. The possibility of major improvements is obvious from e.g. up to two orders of magnitude decrease in gate leakage current in HFETs when low dislocation LEO GaN is used as buffer layer instead of regular GaN.
I A Buyanova
|
|
|
|
