Search

Electrical properties of AlxGa1-xN/GaN heterostructures with an Al content below 15% and carrier concentrations as low as 1.0 × 1012 cm-2 were investigated by Hall effect measurements and capacitance-voltage profiling. The nominally undoped GaN capped structures were grown by low-pressure metal-organic vapor-phase epitaxy. The threshold voltage of transistor devices follows the trend already found for high Al-containing structures, which are described by a model indicating a surface potential independent of Al content. Photoreflectance spectroscopy confirms the results for as-grown heterostructures. The Hall effect measured on the as-grown samples, however, shows a stronger decrease in carrier concentration than expected from the effect of polarization and constant surface potential. In contrast, Hall effect data determined on samples with Ni Schottky contacts and capacitance-voltage profiling on as-grown samples yield the expected behavior, with surface potentials of 0.86 eV and 0.94 eV, respectively. The inconsistency is eliminated by describing the results of the Hall effect on as-grown samples by a two-carrier model. Self-consistent Schrödinger-Poisson calculations support these considerations if we take into account a transition range at the AlxGa1-xN/GaN interface. [ABSTRACT FROM AUTHOR]

The influence of the growth method on the surface potential and thus on the sheet carrier concentration of GaN capped AlxGa1-xN/GaN heterostructures was evaluated. Nominally undoped low pressure metal-organic vapor-phase (MOVPE) and plasma-assisted molecular beam epitaxial (PA-MBE) grown structures with an Al-content between 12% and 30% yield carrier concentrations from 3.6×1012 to 1.2×1013 cm-2. A difference of the concentrations for a fixed Al-content was found between the different epitaxial techniques. This result indicates unambiguously different surface potentials determined quantitatively from the carrier concentration, and is verified in addition by the results of photoreflectance spectroscopy. The GaN surface potentials of MOVPE and PA-MBE grown samples amounts to (0.26±0.04) and (0.61±0.10) eV irrespective of the Al-content of the barrier layer. After device fabrication, we find that due to the identical surface potential defined by the Ni Schottky gate, the threshold voltage for a given Al-content is the same for samples grown with different techniques. Thus, the interplay between epitaxy and process technology defines the threshold voltage. [ABSTRACT FROM AUTHOR]

We systematically investigate Al0.22Ga0.78N/GaN high electron mobility transistors with GaN cap layer thicknesses of 0, 1, and 3 nm. All samples have electron mobilities around 1700 cm2/Vs and sheet carrier concentrations around 8×1012 cm-2 as determined by Hall effect measurements. From photoreflectance measurements we conclude that the electric field strength within the AlGaN barrier increases with GaN cap layer thickness leading to a broadening of the transition peaks as determined by spectroscopic ellipsometry. The surface potential as determined by photoreflectance varies in the range between 0.585 and 0.249 eV dependent on the thickness of the GaN cap. Device results show a significant decrease in Ohmic contact resistance, an increase in ideality factor, a decrease in gate and drain leakage currents, an increase in gain, and an increase in power added efficiency with increasing cap layer thickness. Finally, devices with GaN cap show an improved direct current reliability compared to their counterparts without GaN cap. [ABSTRACT FROM AUTHOR]

A detailed analysis of the piezoelectric response of (GaN/)AlGaN/GaN heterostructures is reported. The electromechanical properties of two types of heterostructures with an Al content of 31% are compared. Only a single two-dimensional electron gas (2DEG) is formed for samples with thin GaN cap layers, while both a 2DEG and a two-dimensional hole gas coexist in the case of thick GaN caps. The lower GaN layer represents the mechanically supporting layer, while the AlGaN film, and in some cases an additional GaN cap layer, serves as the piezoelectrically active layers for actuation. The 2DEG (at the lower AlGaN/GaN interface) provides the conducting channel which was used as back electrode for the applied external voltage. Electroreflectance spectroscopy is applied in order to determine the electric field distribution across the whole structure as a function of the applied voltage. It is found that only a part of the modulation voltage drops across the active region. Piezoelectric force microscopy yields the field (voltage)-dependent actuation of the layers. By correlating the results of the two experimental techniques we are able to determine the piezoelectric modulus d33 with considerably improved reliability. A value for Al0.31Ga0.69N of 5 pm/V is found which is higher than an estimation based on previously reported data for GaN and AlN. [ABSTRACT FROM AUTHOR]

The effect of the growth method on the surface potential and on the sheet carrier concentration of GaN capped [Al.sub.x][Ga.sub.1-x]/GaN heterostructures is analyzed. The studies have shown that after the fabrication, the threshold voltage for a given Al-content is the same for samples grown with different methods because of the identical surface potential defined by the Ni Schottky gate.