Introductory Fourier transform spectroscopy

It is well known that trapping effects can limit the output power performance of microwave field-effect transistors (FETs). This is particularly true for the wide bandgap devices. In this paper we review the various trapping phenomena observed in SiC- and GaN-based FETs that contribute to compromised power performance. For both of these material systems, trapping effects associated with both the surface and with the layers underlying the active channel have been identified. The measurement techniques utilized to identify these traps and some of the steps taken to minimize their effects, such as modified buffer layer designs and surface passivation, are described. Since similar defect-related phenomena were addressed during the development of the GaAs technology, relevant GaAs work is briefly summarized.

Trapping effects in GaN and SiC microwave FETs

The crystallization of undoped amorphous silicon films deposited by low‐pressure chemical vapor deposition in the temperature range 580–530 °C and annealed from 550 to 950 °C has been studied by transmission electron microscopy. The average grain size of the crystallized films depends on the annealing temperature and the deposition conditions. The nucleation rate of new grains during annealing decreases as the deposition temperature decreases from 580 to 545 °C and/or when the deposition rate increases. The final grain size is also influenced by the annealing temperature with the largest grain size obtained at low annealing temperatures. A simple model is described which explains the dependence of grain size on the annealing temperature. An average grain size of 500 nm has been obtained in a 200‐nm film deposited at 545 °C and annealed at 550 °C.

Large grain polycrystalline silicon by low‐temperature annealing of low‐pressure chemical vapor deposited amorphous silicon films

This paper critically investigates the advantages and limitations of the current-transient methods used for the study of the deep levels in GaN-based high-electron mobility transistors (HEMTs), by evaluating how the procedures adopted for measurement and data analysis can influence the results of the investigation. The article is divided in two parts within Part I. 1) We analyze how the choice of the measurement and analysis parameters (such as the voltage levels used to induce the trapping phenomena and monitor the current transients, the duration of the filling pulses, and the method used for the extrapolation of the time constants of the capture/emission processes) can influence the results of the drain current transient investigation and can provide information on the location of the trap levels responsible for current collapse. 2) We present a database of defects described in more than 60 papers on GaN technology, which can be used to extract information on the nature and origin of the trap levels responsible for current collapse in AlGaN/GaN HEMTs. Within Part II, we investigate how self-heating can modify the results of drain current transient measurements on the basis of combined experimental activity and device simulation.

Deep-Level Characterization in GaN HEMTs-Part I: Advantages and Limitations of Drain Current Transient Measurements

GalliumNitridebasedhighelectronmobilitytransistors(HEMTs)areattractiveforuseinhighpowerandhighfrequencyapplications, with higher breakdown voltages and two dimensional electron gas (2DEG) density compared to their GaAs counterparts. Specific applications for nitride HEMTs include air, land and satellite based communications and phased array radar. Highly efficient GaNbased blue light emitting diodes (LEDs) employ AlGaN and InGaN alloys with different compositions integrated into heterojunctions and quantum wells. The realization of these blue LEDs has led to white light sources, in which a blue LED is used to excite a phosphor material; light is then emitted in the yellow spectral range, which, combined with the blue light, appears as white. Alternatively, multiple LEDs of red, green and blue can be used together. Both of these technologies are used in high-efficiency white electroluminescent light sources. These light sources are efficient and long-lived and are therefore replacing incandescent and fluorescent lamps for general lighting purposes. Since lighting represents 20‐30% of electrical energy consumption, and because GaN white light LEDs require ten times less energy than ordinary light bulbs, the use of efficient blue LEDs leads to significant energy savings. GaN-based devices are more radiation hard than their Si and GaAs counterparts due to the high bond strength in III-nitride materials. The response of GaN to radiation damage is a function of radiation type, dose and energy, as well as the carrier density, impurity content and dislocation density in the GaN. The latter can act as sinks for created defects and parameters such as the carrier removal rate due to trapping of carriers into radiation-induced defects depends on the crystal growth method used to grow the GaN layers. The growth method has a clear effect on radiation response beyond the carrier type and radiation source. We review data on the radiation resistance of AlGaN/GaN and InAlN/GaN HEMTs and GaN‐based LEDs to different types of ionizing radiation, and discuss ion stopping mechanisms. The primary energy levels introduced by different forms of radiation, carrier removal rates and role of existing defects in GaN are discussed. The carrier removal rates are a function of initial carrier concentration and dose but not of dose rate or hydrogen concentration in the nitride material grown by Metal Organic Chemical Vapor Deposition. Proton and electron irradiation damage in HEMTs creates positive threshold voltage shifts due to a decrease in the two dimensional electron gas concentration resulting from electron trapping at defect sites, as well as a decrease in carrier mobility and degradation of drain current and transconductance. State-of-art simulators now provide accurate predictions for the observed changes in radiation-damaged HEMT performance. Neutron irradiation creates more extended damage regions and at high doses leads to Fermi level pinning while 60 Co γ-ray irradiation leads to much smaller changes in HEMT drain current relative to the other forms of radiation. In InGaN/GaN blue LEDs irradiated with protons at fluences near 10 14 cm −2 or electrons at fluences near 10 16 cm −2 , both current-voltage and light output-current characteristics are degraded with increasing proton dose. The optical performance of the LEDs is more sensitive to the proton or electron irradiation than that of the corresponding electrical performances. © The Author(s) 2015. Published by ECS. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 License (CC BY, http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse of the work in any

https://iopscience.iop.org/article/10.1149/2.0251602jss/pdf

Review—Ionizing Radiation Damage Effects on GaN Devices

In order to characterize the electron transport properties of the two-dimensional electron gas (2DEG) in AlGaN/GaN modulation-doped field-effect transistors, channel magnetoresistance has been measured in the magnetic field range of 0–12 T, the temperature range of 25–300 K, and gate bias range of +0.5 to −2.0 V. By assuming that the 2DEG provides the dominant contribution to the total conductivity, a one-carrier fitting procedure has been applied to extract the electron mobility and carrier sheet density at each particular value of temperature and gate bias. Consequently, the electron mobility versus 2DEG sheet density has been obtained for each measurement temperature. Theoretical analysis of these results suggests that for 2DEG densities below 7×1012 cm−2, the electron mobility in these devices is limited by interface charge, whereas for densities above this level, electron mobility is dominated by scattering associated with the AlGaN/GaN interface roughness.

Scattering mechanisms limiting two-dimensional electron gas mobility in Al0.25Ga0.75N/GaN modulation-doped field-effect transistors

Transient capacitance measurements of Schottky diodes fabricated on nominally undoped n-type GaN exposed to 60Co gamma irradiation indicate the introduction of two defect levels with thermal activation energies of 89±6 and 132±11 meV. While the emission characteristics of these defects manifest significant broadening, their parameters are consistent with reported electron-irradiation-induced nitrogen-vacancy related centers. Three deep-level defects present before irradiation exposure with activation energies of 265, 355, and 581 meV were found to remain unaffected for cumulative gamma-ray doses up to 21 Mrad(Si).

60Co gamma-irradiation-induced defects in n-GaN

The conduction and breakdown properties of thermally grown SiO 2 films on amorphous-deposited n+polycrystalline silicon (polysilicon) are evaluated using ramped current-voltage ( I-V ) measurements. It is shown that the inferior insulating properties of oxides on polysilicon (polyoxides), when compared to SiO 2 on bulk silicon, can be directly attributed to oxidation-induced interface roughness leading to localized enhancement of the oxide electric field. For example, 16.7-nm-thick polyoxides approach bulk SiO 2 properties since a breakdown field E BD of approximately 9.5 MV . cm-1and an effective barrier height for Fowler-Nordheim tunneling, φ Beff as high as 2.78 eV were measured. Both of these parameters are progressively degraded by increasing polyoxide thickness D OX such that for 165-nm-thick polyoxides E_{BD} \simeq 2.5 MV . cm-1and φ Beff is reduced to as low as 0.83 eV. The measured I-V curves are found to become more polarity dependent with increasing D OX due to a comparatively higher degree of oxidation-induced surface roughening at the lower interface, which renders it more "conductive," with regard to Fowler-Nordheim electron injection, than the upper oxide-polysilicon interface. Certain specific device applications require a relatively "conductive" polyoxide capable of carrying high current densities before failure. Consequently, a polysilicon texturing procedure was developed that has the effect of decreasing φ Beff , increasing breakdown current J BD , and eliminating the polarity dependence of I-V curves for any subsequently formed thin polyoxide. The particular process entails growing a predetermined thickness of "texturing" oxide D teox , and removal prior to formation of the device polyoxide of approximately 25-nm thickness. As D teox is increased from zero to 103 nm, the resultant J BD is found to increase by more than an order of magnitude for both polarities of bias. The corresponding decrease in φ Beff is from 1.7 and 2.4 eV for positive and negative gate bias, respectively, to a polarity-independent value of approximately 1.3 eV.

Thermal SiO 2 films on n + polycrystalline silicon: Electrical conduction and breakdown

For modern semiconductor heterostructures containing multiple populations of distinct carrier species, conventional Hall and resistivity data acquired at a single magnetic field provide far less information than measurements as a function of magnetic field. However, the extraction of reliable and accurate carrier densities and mobilities from the field-dependent data can present a number of difficult challenges, which were never fully overcome by earlier methods, such as the multicarrier fit, the mobility-spectrum analysis of Beck and Anderson, and the hybrid mixed-conduction analysis. More recently, to overcome the limitations of those methods, several research groups have contributed to development of the quantitative mobility-spectrum analysis (QMSA), which is now available as a commercial product. The algorithm is analogous to a fast Fourier transform in that it transforms from the magnetic-field (B) domain to the mobility (μ) domain. The QMSA converts the field-dependent Hall and resistivity data into a visually meaningful transformed output, comprising the conductivity density of electrons and holes in the mobility domain. In this article, we apply QMSA to both synthetic and real experimental data that are representative of modern multilayer HgCdTe structures. We discuss such features as the accuracy of the extraction of individual layer conductivities and average mobilities, reconstruction of the carrier mobility distribution within a particular layer, the resolution of two carriers with similar mobilities, and limits of the sensitivity.

Application of quantitative mobility-spectrum analysis to multilayer HgCdTe structures

In this paper, the Poisson's ratio of low-temperature plasma-enhanced chemical vapor deposited silicon nitride thin films has been determined by a modified double-membrane bulge test. This test method utilizes a square membrane and a large-aspect-ratio rectangular membrane that is fabricated alongside from the same thin film. The Poisson's ratio is determined from the ratio of the bulge deflections of the two membranes under an applied pressure. The method is suitable for determining of either stress-free thin films or those containing low tensile residual stresses. Poisson's ratio values of 0.23 0.02 and 0.25 0.01 were measured for films that were deposited at 125 and 205 , respectively.

L. Faraone

Papers

Scattering mechanisms limiting two-dimensional electron gas mobility in Al0.25Ga0.75N/GaN modulation-doped field-effect transistors

60Co gamma-irradiation-induced defects in n-GaN

Thermal SiO 2 films on n + polycrystalline silicon: Electrical conduction and breakdown

Application of quantitative mobility-spectrum analysis to multilayer HgCdTe structures

Poisson's Ratio of Low-Temperature PECVD Silicon Nitride Thin Films