M. E. Zvanut

Density of interface states, electron traps, and hole traps as a function of the nitrogen density in SiO2 on SiC

Nitridation of the SiO2/SiC interface yields a reduction in interface state density, immunity to electron injection, as well as increased hole trapping. It is shown that the accumulation of nitrogen at the oxide/semiconductor interface is solely responsible for these three effects. The evolution of the density of interface states, electron traps, and hole traps is measured in metal-oxide-semiconductor capacitors as a function of the nitrogen content which is varied by adjusting the gate oxide NO annealing time. A rate equation is derived to model the change in the interface state density, observed at various energy levels, in terms of nitrogen binding cross-sections. While the generation of acceptor interface states upon electron injection is suppressed after minimum N incorporation, the density of oxide hole traps appears to scale linearly with the amount of nitrogen. The origin and the properties of the N-induced hole traps resembles those of the defects responsible for enhanced negative bias temperatur...

Characterization of the luminescence center in photo- and electroluminescent amorphous silicon oxynitride films

We have studied the photoluminescence(PL) mechanism of photo- and electroluminescent amorphous silicon oxynitride filmsgrown by plasma enhanced chemical vapor deposition. The composition of the films was determined by Rutherford backscattering spectrometry and monitored by the index of refraction with single-wavelength ellipsometry. Two sets of samples were grown, each with different reactant gas residence times in the deposition chamber. For samples grown with a residence time of about 5 s, the energy of the PL peak for 2.54 eV excitation is 2.3 eV for stoichiometric films and redshifts with increasing silicon content to 1.7 eV for the most silicon-rich films. The energy of the PL peak for 3.8 eV excitation is 2.8 eV for stoichiometric films and 2.1 eV for the most silicon-rich films. For stoichiometric films, the PL intensity is independent of temperature between 80 and 300 K using 2.54 eV excitation, but the PL intensity decreases by a factor of two over the same temperature range for 3.8 eV excitation. The authors interpret these aspects of the PL as consistent with tail-state recombination. Other results imply the PL is due to a specific luminescence center related to Si–Si or Si–H bonding. A 450 °C anneal reduces the paramagnetic defect density in the films, as detected by electron paramagnetic resonance, by an order of magnitude, but does not increase the PL intensity, while a 950 °C anneal increases both the defect density and the PL intensity. In addition, films in a second set of samples, grown with a residence time of 1.8 s, display very different PL behavior than samples in the first set with the same composition. Samples near stoichiometry in the second set have a PL peak at 2.06 eV and are 20 times less intense than stoichiometric samples in the first set. Optical absorption measurements indicate both types of samples contain Si–Si bonds, with the second set containing many more Si–Si bonds than the first. Fourier-transform infrared measurements indicate the presence of a Si–H bond that is stable at temperatures of 950 °C in the first set, but not in the second set. Thus, the study as a whole suggests a complete picture of luminescence in our silicon oxynitride films must incorporate elements of both tail-state and luminescence center models. The relation of the results to other PL studies in silicon alloys and porous silicon is discussed.

A compensating point defect in carbon-doped GaN substrates studied with electron paramagnetic resonance spectroscopy

Electron paramagnetic resonance (EPR) spectroscopy was used to investigate a type of point defect present in 1019 cm−3 carbon-doped GaN substrates grown by hydride vapor phase epitaxy. A broad, isotropic resonance at g ∼ 1.987 was observed at 3.5 K, and the EPR intensity increased with illumination at energies greater than 2.75 eV and decreased with photon energies greater than 0.95 eV. The latter is consistent with a deep level of 0.95 eV above the valence band maximum and implies that the associated defect likely participates in donor compensation. The ionization energy for this defect is close to the predicted value for the (−/0) transition level of CN and transition levels associated with Ga vacancies such as VGa and VGa-ON-2H.

Charge transfer process for carbon-related center in semi-insulating carbon-doped GaN

Electron paramagnetic resonance (EPR) spectroscopy was used to study the point defects in 2 × 1017–1019 cm−3 C-doped GaN substrates grown by hydride vapor phase epitaxy. The intensity of an isotropic signal with g = 1.987 ± 0.001 increased monotonically with the carbon concentration, indicating that the EPR signal represents a carbon-related defect. In each sample, the signal intensity increased under illumination with photon energy greater than 2.75 eV, and the photo-induced signal decreased with subsequent illumination at 0.95 eV. A second signal, well-documented to be a shallow donor, appeared along with the g = 1.987 signal in the most lightly doped samples. The appearance of the donor confirms that the photo-induced increase is caused by excitation of an electron to the conduction band and implies that a defect level for the carbon-related center is about 1 eV above the valence band edge, consistent with temperature-dependent Hall measurements.

Charge transfer in Fe-doped GaN: The role of the donor

Several nitride-based device structures would benefit from the availability of high quality, large-area, freestanding semi-insulating GaN substrates. Due to the intrinsic n-type nature of GaN, however, the incorporation of compensating centers such as Fe is necessary to achieve the high resistivity required. We are using electron paramagnetic resonance (EPR) to explore charge transfer in 450 um thick GaN:Fe plates to understand the basic mechanisms related to compensation so that the material may be optimized for device applications. The results suggest that the simple model based on one shallow donor and a single Fe level is insufficient to describe compensation. Rather, the observation of the neutral donor and Fe3+ indicates that either the two species are spatially segregated or additional compensating and donor defects must be present.

The Mg impurity in nitride alloys

Although several magnetic resonance studies address the Mg acceptor in GaN, there are few reports on Mg doping in the alloys, where hole production depends strongly on the Al or In content. Our electron paramagnetic resonance (EPR) measurements of the p-type alloys suggest that the Mg impurity retains the axial symmetry, characteristic of a p-type dopant in both alloys; however, In and Al produce additional, different characteristics of the acceptor. In InGaN, the behavior is consistent with a lowering of the acceptor level and increasing hole density as In concentration increases. For AlGaN, the amount of neutral Mg decreases with increasing Al content, which is attributed to different kinetics of hydrogen diffusion thought to occur in samples with higher Al mole fraction.