Search

(Sc2O3)x(Ga2O3)1−x was grown by molecular beam epitaxy at low temperatures (100 °C) using a variety of growth sequences to avoid surface segregation of Ga. Continuous and digital growth techniques always produced Ga segregation. This surface segregation was attributed to the stronger bond between the Sc and O compared to the Ga and O. A digital growth technique (alternate opening of Sc and Ga shutters with the O shutter open continuously during the growth) was unsuccessful in eliminating this effect. The segregation was eliminated using a growth technique in which the Ga shutter was closed for a set amount of time toward the end of the growth while the O and Sc shutters remained open. Characterization with reflection high energy electron diffraction, x-ray diffraction, and transmission electron microscopy revealed the growth of a fine-grained polycrystalline film under these conditions. A third growth technique was used that involved closing the Ga shutter for a set amount of time toward the end of the growth while the O and Sc shutters were open continuously. This technique was successful in depositing a uniform film. However, the breakdown field was only 1.40 MV/cm (at 1 mA/cm2). The addition of Ga to Sc2O3 diminished the insulating properties of the film. These initial experiments indicate that phase segregation is likely to be a major issue with most growth techniques and that alloying Ga2O3 with elements other than Sc, such as Gd or Al, might be a more successful approach.

The characteristics of sputtered NiO for use in pn heterojunctions with Ga2O3 were investigated as a function of sputtering parameters and postdeposition annealing temperature. The oxygen/ nickel and Ni2O3/NiO ratios, as well as the bandgap and resistivity, increased as a function of O2/Ar gas flow ratio. For example, the bandgap increased from 3.7 to 3.9 eV and the resistivity increased from 0.1 to 2.9 Ω cm for the O2/Ar ratio increasing from 1/30 to 1/3. By sharp contrast, the bandgap and Ni2O3/NiO ratio decreased monotonically with postdeposition annealing temperatures up to 600 °C, but the density of films increased due to a higher fraction of NiO being present. Hydrogen is readily incorporated into NiO during exposure to plasmas, as delineated by secondary ion mass spectrometry measurements on deuterated films. The band alignments of NiO films were type II-staggered gaps with both α- and β-Ga2O3. The breakdown voltage of NiO/β-Ga2O3 heterojunction rectifiers was also a strong function of the O2/Ar flow ratio during deposition, with values of 1350 V for 1/3 and 830 V for 1/30.

The (SmxGa1−x)2O3 alloy system is a potential new dielectric for compound semiconductors such as GaAs. Using molecular beam epitaxy under metal-modulated growth conditions, we grew the binary oxide, Sm2O3, at two substrate temperatures (100 and 500 °C) and optimized the structural, morphological, and electrical properties of the films. Decreasing the Sm cell temperature suppressed the formation of the monoclinic phase and promoted the growth of the cubic phase. Next, the ternary oxide, (SmxGa1−x)2O3, was deposited to investigate the effects of Ga incorporation. Optimization experiments were used to determine the effects of substrate temperature and samarium cell temperature (i.e., growth rate) on film stoichiometry, phase distribution, and microstructure in these films. Films grown at 500 °C showed significant surface roughness and the presence of multiple crystalline phases. Since all of the Sm-based oxides (i.e., samarium oxide with and without gallium) were found to have unbonded Sm metal, annealing experiments were carried out in oxygen and forming gas to determine the effects of annealing on film stoichiometry. The motivation behind annealing in forming gas was to see whether this commonly used technique for reducing interface densities could improve the film quality. GaAs metal-oxide-semiconductor diodes with (SmxGa1−x)2O3 showed breakdown fields at 1 mA/cm2 of 4.35 MV/cm, which decreased with increasing Sm unbonded metal content in the films.

β-Ga2O3 is a wide bandgap semiconductor (~4.8 eV) with properties suited to power electronics, truly solar-blind UV detection, and extreme environment applications. Additional bandgap tunability can be achieved through incorporation of Al into β-Ga2O3, leading to β-(Al,Ga)2O3 monoclinic phase alloys with the bandgap tuned from 4.8- 6 eV. This enables heterostructure designs such as modulation-doped electron channels, quantum wells, and superlattices. Out of these possible structures, β-(AlxGa1-x)2O3/Ga2O3 heterostructure field-effect transistors (HFETs) have shown impressive and improving performance compared to Ga2O3 based metal semiconductor field effect transistor (MESFETs). For these heterostructures, other wide bandgap oxides can be used as gates dielectrics, and narrow energy bandgap conductive oxides can be utilized to reduce contact resistance between the ohmic contacts and semiconductor surface. For gate dielectrics on (AlxGa1-x)2O3, one must utilize oxide materials with a bandgap of 7 eV of higher to have adequate offsets in both the conduction and valence bands. In this study, we utilize x-ray photoelectron spectroscopy (XPS) to report on the determination of the band alignment of various insulating and semiconducting dielectrics onto (AlxGa1-x)2O3 at various aluminum concentrations. The dielectrics were deposited using atomic layer deposition (ALD) or sputtering. For SiO2 and Al2O3, there are differences of up to 1 eV in band alignments on single crystal (AlxGa1-x)2O3, depending on whether they are deposited by sputtering or ALD. For the conductive oxides, AZO/ β-(AlxGa1-x)2O3 and ITO/ β-(AlxGa1-x)2O3 heterojunctions have a nested gap (type I) band offset. The valence and conduction band offsets are negative in both cases and can be used to enhance carrier transport across the heterointerface.

The wide-bandgap ternary (AlxGa1−x)2O3 forms a heterostructure system with Ga2O3 that is attracting attention for modulation-doped field-effect transistors. The options for gate dielectric on (AlxGa1−x)2O3 are limited by the need for adequate band offsets at the heterointerface. Al2O3 deposited by atomic layer deposition (ALD) is one option due to its large bandgap (6.9 eV). We measured the valence-band offset at the Al2O3/(Al0.14Ga0.86)2O3 heterointerface using x-ray photoelectron spectroscopy (XPS). Al2O3 was deposited by ALD onto single-crystal β-(Al0.14Ga0.86)2O3 (bandgap 5.0 eV) grown by molecular beam epitaxy (MBE). The valence-band offset was determined to be 0.23 ± 0.04 eV (straddling gap, type I alignment) for ALD Al2O3 on β-(Al0.14Ga0.86)2O3. The conduction-band offset was 1.67 ± 0.30 eV, providing good electron confinement.

The valence band offsets in rf-sputtered Indium Tin Oxide (ITO)/single crystal β-Ga2O3 (ITO/Ga2O3) heterostructures were measured with X-Ray Photoelectron Spectroscopy using the Kraut method. The bandgaps of the component materials in the heterostructure were determined by Reflection Electron Energy Loss Spectroscopy as 4.6 eV for Ga2O3 and 3.5 eV for ITO. The valence band offset was determined to be −0.78 ± 0.30 eV, while the conduction band offset was determined to be −0.32 ± 0.13 eV. The ITO/Ga2O3 system has a nested gap (type I) alignment. The use of a thin layer of ITO between a metal and the Ga2O3 is an attractive approach for reducing contact resistance on Ga2O3-based power electronic devices and solar-blind photodetectors.

This work tested the feasibility of a silicon-on-insulator microdosimeter, which mimics the size and shape of specific cells within the human body, to determine dose equivalent from neutron irradiation. The microdosimeters were analyzed in terms of their basic diode characteristics, i.e., leakage current as a function of bias voltage. Lineal energy spectra were acquired using two different converter layers placed atop the microdosimeter: a tissue-substitute converter made from high-density polyethylene, and a boron converter consisting of epoxy coated with boron powder. The spectra were then converted into absorbed dose and dose equivalent. Experimental results were compared to Monte Carlo simulations of the neutron irradiations, revealing good agreement. Uncertainty in the dose equivalent determinations was 7.5% when using the cell-shaped microdosimeter with the tissue-substitute converter and 13.1% when using the boron converter. This work confirmed that the SOI approach to cell-mimicking microdosimetry is feasible.

X-Ray Photoelectron Spectroscopy was used to determine the valence band offset at Al 2 O 3 /β-Ga 2 O 3 heterointerfaces. The Al 2 O 3 was deposited either by Atomic Layer Deposition (ALD) or rf magnetron sputtering and the synthesis method was found to have a very significant effect on the resulting band alignment. The bandgaps of the materials were determined by Reflection Electron Energy Loss Spectroscopy as 4.6 eV for Ga 2 O 3 and 6.9eV for Al 2 O 3 deposited by either method. The valence band offset was determined to be 0.07eV ± 0.20 eV (straddling gap, type I alignment) for ALD Al 2 O 3 on Ga 2 O 3 and -0.86 ± 0.25 eV (staggered gap, type II alignment) for sputtered Al 2 O 3 . This led to conduction band offsets of 2.23 ± 0.60 eV for ALD Al 2 O 3 and 3.16 ± 0.80 eV for sputtered Al 2 O 3 , respectively. The choice of deposition method for the dielectric alters the type of band alignment for the Al 2 O 3 /Ga 2 O 3 system from type I alignment to type II. Since the main difference is expected to be the disorder at the dielectric/Ga 2 O 3 interface, this shows how synthesis method can affect the resulting band alignment.

We report on the determination of band offsets in rf-sputtered Aluminum Zinc Oxide (AZO)/single crystal β-Ga2O3 (AZO/Ga2O3) heterostructures using X-Ray Photoelectron Spectroscopy. The bandgaps of the materials were determined by Reflection Electron Energy Loss Spectroscopy as 4.6 eV for Ga2O3 and 3.2eV for AZO. The valence band offset was determined to be −0.61eV ± 0.23eV, while the conduction band offset was determined to be −0.79 ± 0.34 eV. The AZO/Ga2O3 system has a nested, or straddling, gap (type I) alignment and provides a convenient method for reducing contact resistance on Ga2O3-based device structures.

We measured the band offsets of sputtered Sc 2 O 3 on thin film InGaZnO 4 (IGZO) using X-Ray Photoelectron Spectroscopy and obtained the bandgaps of the materials using reflection electron energy loss spectroscopy and UV/Vis absorption. The valence band offset was determined to be −1.33 eV ± 0.13 eV for Sc 2 O 3 on sputtered IGZO (bandgap 3.16 eV). The conduction band offset for Sc 2 O 3 /IGZO was then determined to be 4.07 eV. The Sc 2 O 3 /IGZO system has a staggered, type II alignment, meaning that it is not suitable for thin film transistors but it may still be useful for surface passivation to prevent the commonly observed changes in conductivity of the IGZO resulting from atmospheric exposure.

Selection of gate dielectrics in metal-oxide-semiconductor field-effect transistor/metal-insulator-semiconductor filed effect transistor (MOSFET/MISFET) design is of the upmost importance to optimize fully device performance. Understanding the band alignment of the gate dielectric to the semiconductor is the first step in device design, where the conduction and valence band offsets must provide sufficient energy barrier to both electron and hole transfer. Thus far band alignments for SiO2, HfO2, HfSiO4, Al2O3, LAO, ZrO2, AZO, ITO, GaN, 6H-SiC, a-Si, and others have been reported for Ga2O3. Al2O3 with its a wide bandgap of 8 eV has been the most used gate dielectric thus far for all β-Ga2O3-based power electronics due to its large conduction band offset and potentially commensurate structure. Much work is still needed in understanding the interface between the dielectric and Ga2O3, yet the results to date on band alignment provide a good starting point.

We have measured the band offsets of sputtered ZrSiO 4 on bulk ZnO wafers using X-Ray Photoelectron Spectroscopy and obtained the bandgaps of these two materials using reflection electron energy loss spectroscopy. The valence band offset was determined to be −0.60 eV ± 0.04 eV for ZrSiO 4 on ZnO, while the respective bandgaps were 3.22 eV for ZnO and 5.9 eV for ZrSiO 4 . The conduction band offset for ZrSiO 4 /ZnO was then determined to be 3.28 eV. The ZrSiO 4 /ZnO system has a staggered, type II alignment. This means that while ZrSiO 4 may be useful for surface passivation applications on ZnO, it is not suitable for thin film transistors where a positive valence band offset is needed.

We measured the valence band offset of sputtered ZrSiOx (bandgap 5.9 eV) on InGaZnO4 (IGZO) using X-ray photoelectron spectroscopy and the bandgaps of these materials using either absorption or reflection electron energy loss spectroscopy. This enabled us to additionally derive the conduction band offsets. We examined the effect of venting the samples prior to dielectric deposition on the band offsets compared to keeping them in vacuum. The ZrSiOx/IGZO system has a staggered, type II alignment, with a valence band offset of −0.12 eV ± 0.04 eV for samples in which all the layers were kept in vacuum or −0.05 eV for samples exposed to atmosphere prior to the measurement. The conduction band offset for ZrSiOx/IGZO was derived as 2.86 eV for the vacuum samples from the measured bandgap of the IGZO (3.16 eV) and 2.93 eV for vented samples. The small valence band offsets in this system are consistent with past reports that ZrO2/IGZO has essentially no valence band offset.

Amorphous InGaZnO 4 (a-IGZO) thin film transistors have been demonstrated on a variety of substrates, including flexible template like paper, tape and polyethylene terephthalate. Charge trapping at the a-IGZO/dielectric interface has been identified as a primary source of instability in these transistors. The valence and conduction band offsets are important parameters of these interfaces and should be sufficiently large to inhibit the injection of electrons and holes into the dielectric where they can cause threshold shifts and gate leakage. In this paper we report on the determination of band offsets in the HfO 2 /InGaZnO 4 heterostructure using X-Ray Photoelectron Spectroscopy. The valence band offset was determined to be ∼0.38 eV ± 0.03 eV, while the conduction band offset was determined to be 2.26 eV. The HfO 2 /InGaZnO 4 system has a nested or straddling gap (type I) alignment. We compare these bad offsets to previously reported values for HfO 2 and other dielectrics on a-IGZO.

AlGaN/GaN high electron mobility transistors (HEMTs) showed its promising performance in high power and high frequency, which can be used for applications such as satellite-based communication networks, inverter units in hybrid electric vehicles and advanced radar systems. However, intrinsic defects and defects generated during the device fabrication degraded HEMT performance, such as drain current collapse, high gate leakage, and lower rf power density and power add efficiency. Furthermore, subsequent electrical stressing of the HEMTs during operation leads to creation of more traps and further device degradation through various mechanisms, including gate contact sinking, shallow trap formations, and the inverse piezoelectric effect. It is highly desirable to have non-destructive methods available to identify the activation energies of the defects and spatial location of trap states in HEMT. A sub-bandgap optical pumping technique was developed to identify trap locations in AlGaN/GaN HEMTs. By varying photon fluxes, the traps with different activation energies appeared at different photon flux levels. This implies that the defects originate at different physical locations in the HEMT. The locations of the traps identified with the sub-bandgap optical pumping methods confirmed by gate pulse measurements under optical pumping.

The valence band discontinuity (ΔE(v)) of Y2O3/InGaZnO4 (IGZO) heterojunctions was measured by a core-level photoemission method. The Y2O3 exhibited a band gap of -6.27 eV from absorption measurements. A value of ΔE(v) = 0.44 ± 0.21 eV was obtained by using the Ga 2p3/2, Zn 2p3/2 and in 3d5/2 energy levels as references. Given the experimental bandgap of 3.2 eV for the IGZO, this would indicate a conduction band offset ΔE(c) of - 2.63 eV in the Y2O3/IGZO heterostructures and a nested interface band alignment.

AlGaN/GaN high electron mobility transistor (AlGaN/GaN HEMT) has attracted considerable interest as high power electronics for use in the electric utility industry, defense and space applications, and hybrid vehicles. However, the impact ionization in gate-to-drain region induced by hot electrons causes serious issues such as carrier trapping, gate leakage, and efficiency degradation. Many researches have studied the location and properties of these traps, however, to separate out trap population and trap energy tends to be rather tedious task. We have created a characterization system that employs below bandgap illumination to effectively separate trap energy and population. We investigated the traps affecting 2DEG density in AlGaN/GaN HEMTs by sub-bandgap optical pump. The device is fabricated on 6H-SiC substrate passivated by 120nm SiNx. The barrier and the buffer layer are 25 nm Al0.25Ga0.75N and 2 μm GaN, respectively. During optical pumping, the VDS and VGS applied are 0.5V and 0V to minimize the self-heating effect and the change in drain current is monitored. The wavelength of monochromatic source we used were 671 and 447 nm, with the same power density, 29 mW/mm2, the power density that the number densities of photons for both wavelengths are in the same order (1016 sec-1mm-2). After optical pumping, the device was then stressed at semi-on state (VDS=20V, VGS=-1V) under an elevated temperature of 80 °C until the drain current decreased by 10%. High electric field during the stress allows electrons to leave the channel and induces traps in the buffer layer, which induces current degradation. Figure 1 shows the result of optical pumping for the devices before and after the stress. The illumination starts at 10 seconds and lasts for 2 minutes. From Table I we can see that the amount of current increase when the device illuminated by 447 nm is larger than that illuminated by 671 nm, indicating deeper traps being affected due to the photon with higher energy. Larger difference between illuminated current (Ihυ) and dark current (Idark) can be observed after the stress, and we can see that there is 0.21mA and 0.57mA increase in the amount of difference when the device is under the exposure of 671 nm and 447 nm photons, respectively. This suggests that there are more deep defects than the shallow ones created due to stress. Based on the current difference induced by photons, we can learn the number of carrier detrapped, and the estimated value is included in table I. The work described above is useful for evaluating the type of traps and their approximate population during DC device stress and measurement. The traps measured by this method of below bandgap optical pumping are found in the entire AlGaN/GaN structure, both surface regions and buffer layer. It is well know that surface traps lead to a current collapse effect which is detrimental to device performance. The presences of these surface traps can be determined from gate-lag measurements. It is our goal to optically enhance the gate-lag technique by incorporating the two into one measurement system. This will allow for characterization of trap energy, population and location in the HEMT structure. This work performed at the University of Florida was supported by an AFOSR MURI monitored by James Huang.

The electrical and structural properties of sputter-deposited W Schottky contacts with Au overlayers on n-type Ga2O3 are found to be basically stable up to 500 °C. The reverse leakage in diode structures increases markedly (factor of 2) for higher temperature annealing of 550–600 °C. The sputter deposition process introduces near-surface damage that reduces the Schottky barrier height in the as-deposited state (0.71 eV), but this increases to 0.81 eV after a 60 s anneal at 500 °C. This is significantly lower than conventional Ni/Au (1.07 eV), but W is much more thermally stable, as evidenced by Auger electron spectroscopy of the contact and interfacial region and the minimal change in contact morphology. The contacts are used to demonstrate 1.2 A switching of forward current to −300 V reverse bias with a reverse recovery time of 100 ns and a dI/dt value of 2.14 A/μs. The on/off current ratios were ≥106 at −100 V reverse bias, and the power figure-of-merit was 14.4 MW cm−2.

X-ray photoelectron spectroscopy (XPS) was used to measure the energy discontinuity in the valence band (ΔEV) and conduction band (ΔEC) of various dielectric/IGZO heterostructures. ΔEV values of 0.48{plus minus}0.025 eV for HfO2/IGZO, 0.44{plus minus}0.21 eV for Y2O3/IGZO, 0.81{plus minus}0.17 eV for MgO/IGZO, 0.95{plus minus}0.17 eV for the Al2O3/IGZO, and 1.43{plus minus}0.15 eV for the SiO2/IGZO heterojunctions were obtained, respectively. The bandgap differences between these dielectrics and IGZO would lead to corresponding conduction band offset (ΔEC) values of 2.39-4.74 eV. These results are important for designing structures with good carrier confinement in transparent thin film transistors based on IGZO.

Sub-micron AlGaN/GaN high electron mobility transistors were RF stressed at various drain bias conditions at 10 GHz under 3 dB and 3.7 dB compression. Rapid degradation was observed above a drain bias of 20 V, with significant degradation of the Schottky contact. Additionally, electroluminescence and cathodoluminescence was performed on stressed devices. Localization of 2.2 eV defect emission was observed on a device suffering from infant mortality.

AlGaN/GaN high electron mobility transistors with 0.125 μm gate length were rf stressed at 10 GHz. Depth resolved cathodoluminescence (CL) was employed to investigate the role of defects in devices that exhibited typical mean time to failure (MTTF) compared to those that suffered from infant mortality. The CL spectra exhibited both laterally and vertically localized defect emission within the channel of devices that failed early, indicating the presence of an increased concentration of VGa-based defects within the device. Additionally, residual compressive stress was observed in all devices after electrical stress.

AlGaN/GaN High Electron Mobility Transistors (HEMTs) with gate lengths of 0.14–0.17 μm were electrically stressed under on-state (VG = 0 V), off-state (VG = −5 V) and typical operating conditions (VG = −2 V) and subsequently illuminated with below band-gap light to monitor changes in drain current corresponding to the changes in trapping and de-trapping of carriers within the band-gap. The changes in drain current are indicators of a change in trap density as a result of electrical stressing, since the energy from a specific wavelength of light fills traps whose activation energies are less than or equal to that of the light source. Changes in trap densities were minimal after both off-state and on-state stressing but significant trap creation in the range EC-0.4–0.6 eV were observed in HEMTs exhibiting gradual degradation during stressing. Energy levels corresponding to these values in the literature have been suggested to correlate with GaN and NGa substitutional defects, as well as GaI interstitials.

We present a review of reliability issues in AlGaN/GaN and AlGaAs/GaAs high electron mobility transistors (HEMTs) as well as Heterojunction Bipolar Transistors (HBTs) in the AlGaAs/GaAs materials systems. Because of the complex nature and multi-faceted operation modes of these devices, reliability studies must go beyond the typical Arrhenius accelerated life tests. We review the electric field driven degradation in devices with different gate metallization, device dimensions, electric field mitigation techniques (such as source field plate), and the effect of device fabrication processes for both DC and RF stress conditions. We summarize the degradation mechanisms that limit the lifetime of these devices. A variety of contact and surface degradation mechanisms have been reported, but differ in the two device technologies: For HEMTs, the layers are thin and relatively lightly doped compared to HBT structures and there is a metal Schottky gate that is directly on the semiconductor. By contrast, the HBT relies on pn junctions for current modulation and has only Ohmic contacts. This leads to different degradation mechanisms for the two types of devices.

Materials modification and nanofabrication results are reported from the use of a new multi-ion beam lithography and processing system developed by the University of Florida (UF) and Raith Inc. The UF system utilizes liquid metal alloy ion sources and an ExB filter to produce nanometer-dimension, mass selected ion beams from 15 to 40 kV that can be used for direct-write ion beam lithography, sputter profiling, maskless ion implantation, ion beam mixing, and spatial and temporal ion beam assisted writing and processing over (100 × 100 mm2) – all with nanometer precision. The initial materials modification results reported here utilized an AuSi eutectic source to fabricate lithographically patterned arrays of Au and Si nanocrystals in SiO2 substrates by direct-write, maskless implantation and thermal annealing. The potential for nanofabrication using this system was illustrated by comparing the advantages for developing a prototype GaAs device by: (1) surface patterning GaAs with Au versus Ga ions; and by the ability to switch to Si for in-situ implantation doping of the GaAs device without removing the sample or applying additional lithography or processing steps. Capabilities and future possibilities for the system are discussed.

AlGaN/GaN High Electron Mobility Transistors were found to exhibit a negative temperature dependence of the critical voltage ( V CRI ) for irreversible device degradation to occur during bias-stressing. At elevated temperatures, devices exhibited similar gate leakage currents before and after biasing to V CRI , independent of both stress temperature and critical voltage. Though no crack formation was observed after stress, cross-sectional TEM indicates a breakdown in the oxide interfacial layer due to high reverse gate bias.

As AlGaN/GaN high electron mobility transistors gain commercial acceptance for use in high power and high frequency applications, it is becoming ever important to understand the degradation mechanisms that drive failure in the field. Because of the complex nature and multi-faceted operation modes of these devices, reliability studies must go beyond the typical Arrhenius accelerated life tests. To do this, we investigated the electric field driven degradation in devices with different gate metallization, device dimensions, electric field mitigation techniques (such as source field plate), and the effect of device fabrication processes for both dc and RF stress conditions.

The critical degradation voltage of AlGaN/GaN High Electron Mobility Transistors with Ni/Au gate during off-state electrical stress was studied. Devices with different gate length and gate-drain distances were found to exhibit the onset of degradation at different source-drain biases but similar electric field strengths, showing that the degradation mechanism is primarily field-driven. The calculated degradation field was ~1.8MV/cm by ATLAS simulations. Transmission Electron Microscopy imaging showed creation of defects under the gate after DC stress.

The critical voltage for degradation of AlGaN/GaN high electron mobility transistors (HEMTs) employed with the Pt/Ti/Au gate metallization instead of the commonly used Ni/Au was significantly increased during the off-state stress. The typical critical voltage for HEMTs with Ni/Au gate metallization was around -60V. By sharp contrast, no critical voltage was observed for the HEMTs with Pt/Ti/Au gate metallization, even up to -100V, which was the instrumental limitation in this experiment. Both Schottky forward and reverse gate characteristics of the Ni/Au degraded once the gate voltage passed the critical voltage of around -60V. There was no degradation exhibited for the HEMTs with Pt-gated HEMTs.

Semiconductor ZnO and GaN nanowires have shown the ability to detect many types of gases, biological and chemical species of interest. In this review, we give some recent examples of using these nanowires for pH sensing, glucose detection and hydrogen detection at ppm levels. In addition, new ion beam systems with nm resolution capability are now available. This opens up the possibility of directly writing doped regions into semi-insulating materials such as undoped GaAs or GaN for creation of arrays of nm channel transistors and other active elements. Initial work in this area is discussed.

AlGaN/GaN High Electron Mobility Transistors (HEMTs) with various gate lengths have been step-stressed under both on- and off-state conditions. On-state, high power stress tests were performed on 0.17 μm gate length HEMTs and a single 5 μm spaced TLM pattern. Significant degradation of the submicron HEMTs as compared to the excellent stability of the TLM patterns under the same stress conditions reveal that the Schottky contact is the source of degradation. Off-state stress showed a linear relationship between the critical degradation voltage and gate length, though two dimensional ATLAS/Blaze simulations show that the maximum electric field is similar for all gate lengths. Additionally, as the drain bias was increased, the critical voltage decreased. However, the cumulative bias between the gate and drain remained constant, further indicating that the electric field is the main mechanism for degradation.

The energy discontinuity in the valence band (deltaE(v)) of Y2O3-stabilized ZrO2 (YSZ)/InGaZnO4 (IGZO) heterostructures was obtained from X-ray photoelectron spectroscopy (XPS) measurements. The YSZ exhibited a bandgap of 4.4 eV from absorption measurements. A value of deltaE(v) = 0.57 +/- 0.12 eV was obtained by using Ga 2P3/2, Zn 2p3/2 and In 3d5/2 energy levels as references. This implies a conduction band offset (deltaE(c)) of 0.63 eV in YSZ/InGaZnO4 heterostructures and a nested interface band alignment.

The (AlxGa1−x)2O3/Ga2O3 system is attracting attention for heterostructure field effect transistors. An important device design parameter is the choice of gate dielectric on the (AlxGa1−x)2O3 and its band alignment at the heterointerface. The valence band offset at the SiO2/(Al0.14Ga0.86)2O3 heterointerface was measured using x-ray photoelectron spectroscopy. The SiO2 was deposited by atomic layer deposition (ALD) onto single-crystal β-(Al0.14Ga0.86)2O3 grown by molecular beam epitaxy. The bandgap of the SiO2 was determined by reflection electron energy loss spectroscopy as 8.7 eV, while high resolution XPS data of the O 1s peak and onset of elastic losses were used to establish the (Al0.14Ga0.86)2O3 bandgap as 5.0 eV. The valence band offset was determined to be 1.60 ± 0.40 eV (straddling gap, type I alignment) for ALD SiO2 on β-(Al0.14Ga0.86)2O3. The conduction band offset was 2.1 ± 0.08 eV, providing for a strong electron transport restriction.The (AlxGa1−x)2O3/Ga2O3 system is attracting attention for heterostructure field effect transistors. An important device design parameter is the choice of gate dielectric on the (AlxGa1−x)2O3 and its band alignment at the heterointerface. The valence band offset at the SiO2/(Al0.14Ga0.86)2O3 heterointerface was measured using x-ray photoelectron spectroscopy. The SiO2 was deposited by atomic layer deposition (ALD) onto single-crystal β-(Al0.14Ga0.86)2O3 grown by molecular beam epitaxy. The bandgap of the SiO2 was determined by reflection electron energy loss spectroscopy as 8.7 eV, while high resolution XPS data of the O 1s peak and onset of elastic losses were used to establish the (Al0.14Ga0.86)2O3 bandgap as 5.0 eV. The valence band offset was determined to be 1.60 ± 0.40 eV (straddling gap, type I alignment) for ALD SiO2 on β-(Al0.14Ga0.86)2O3. The conduction band offset was 2.1 ± 0.08 eV, providing for a strong electron transport restriction.

Thin films of copper iodide (CuI) were grown on (-201) bulk Ga2O3 and (010) epitaxial (Al0.14Ga0.86)2O3 using a copper film iodination reaction method. The valence band offsets for these heterostructures were measured by X-ray photoelectron spectroscopy (XPS). High resolution XPS data of the O 1s peak and onset of elastic losses were used to establish the (Al0.14Ga0.86)2O3 bandgap to be 5.0 ± 0.30 eV. The valence band offsets were −0.25 eV ± 0.07 eV and 0.05 ± 0.02 eV for CuI on Ga2O3 or (Al0.14Ga0.86)2O3, respectively. The respective conduction band offsets were 1.25 ± 0.25 eV for Ga2O3 and 1.85 ± 0.35 eV for (Al0.14Ga0.86)2O3. Thus, there is a transition from type-II to type-I alignment as Al is added to β-Ga2O3. The low valence band offsets are ideal for hole transport across the heterointerfaces.

A threshold reverse bias of ∼21 V was observed leading to a sharp increase in the gate current of AlGaN/GaN high electron mobility transistors biased at low source-drain voltage (5 V). The gate current increases by one to two orders of magnitude at this bias, corresponding to an electric field strength around 1.8 MV cm−1. The gate current increased by roughly five orders of magnitude after step-stressing the gate bias from 10 to 42 V in 1 V increments for 1 min at each bias. The drain current was also decreased by ∼20% after this step-stress cycle. The photoluminescence and electroluminescence intensity from the semiconductor is decreased along the periphery of the gate region after stressing and transmission electron microscopy shows a thin native oxide layer under the gate and this disappears as the gate metal reacts with the underlying AlGaN.

Peltier element cooling of ungated AlGaN/GaN high electron mobility transistors (HEMTs) is shown to be an effective method for condensing exhaled breath, enabling the measurement of the pH and glucose of the exhaled breath condensate (EBC). By comparison with standard solutions, the current change measured in the HEMTs with EBC shows that the sensitivity of the glucose detection is lower than the glucose concentration in the EBC of healthy human subjects and the pH of the condensate from the exhaled breath is within the range of 7-8, typical of that for human blood. The HEMT sensors can be integrated into a wireless data transmission system that allows for remote monitoring. Details of the transmitter and receiver design for the transmission system are given. Our work demonstrates the possibility of using AlGaN/GaN HEMTs for extended investigations of airway pathology without the need for clinical visits.

Pt-coatings sputtered on to multiple GaN nanowires is shown to enhance their sensitivity for hydrogen at the hundreds of ppm level at 25 °C. GaN nanowires without a metal coating showed no detectable change in current when exposed to hydrogen under the same conditions. Pt-coated multiple GaN nanowires gave non-linear relative responses of ∼1.7% at 200 ppm up to ∼1.9% at 2000 ppm H 2 in N 2 after a 10-min exposure. More than 80% of the initial GaN conductance was recovered within 2 min after removal of hydrogen from the measurement medium. At higher temperatures, larger resistance changes and faster response and recovery were obtained. The adsorption activation energy of the sensor was 7.3 kcal mol −1 at 2000 ppm H 2 in N 2 . Pt-coated GaN nanowire sensors resulted in much lower relative response than Pd-coated sensors.

Peltier element cooling is demonstrated to be an effective method for collecting exhaled breath condensate (EBC) on AlGaN/GaN High Electron Mobility Transistors (HEMT). The HEMT sensors have functionalized gate areas for glucose and pH measurement. The current change measured in the HEMTs with EBC shows that the sensitivity of the glucose detection is lower than the glucose concentration in the EBC of healthy person and the pH measurement range includes 7- 8, typical of that for human blood. The sensors can be integrated into a wireless data transmission system that allows remote monitoring. Details of the transmitter and receiver design for the transmission system are given. Our work demonstrates the possibility of using AlGaN/GaN HEMTs for extended investigations of airway pathology without the need for clinical visits.

We report on the growth and characterization of high-quality GaN nanowires for hydrogen sensors. We grew the GaN nanowires by catalytic chemical vapor deposition (CVD) using gold thin films as a catalyst on a Si wafer with an insulating SiO2 layer. Structural characterization of the as-grown nanowires by several methods shows that the nanowires are single-crystal wurtzite GaN.␣Photoluminescence measurements under 325 nm excitation show a near-band-edge emission peak around ∼3.4 eV. The hydrogen sensors are fabricated by contacting the as-grown GaN nanowires by source and drain electrodes and coating them with a thin layer of Pd. Hydrogen sensing experiments using the fabricated devices show high sensitivity response (ppm detection limit at room temperature) and excellent recovery. This work opens up the possibility of using high-quality GaN nanowire networks for hydrogen sensing applications.

The optical properties of N-doped ZnO films grown by pulsed laser deposition are examined for which zinc nitride is used as the source of nitrogen. The motivation for this study is to determine if nitrogen-related acceptor state formation can be achieved in ZnO films using Zn3N2 doping in the ablation target. The films were deposited in oxygen or nitrogen on c-plane sapphire. Photoluminescence measurements at 20 K reveal a 3.31 eV acceptor-bound exciton emission due to nitrogen substitution on the oxygen site, donor–acceptor pair emission at 3.23 ± 1 eV and free electron-acceptor at 3.27 eV. The binding energy of the N-related acceptor is estimated to be in the range of 170–15 meV. While the as-deposited films were n-type, thermal annealing in oxygen yielded insulating behavior, consistent with compensating acceptor states.

The optical and transport properties of As-doped ZnO films are examined. The films are grown by pulsed-laser deposition using As2O3 as the source of As and O3/O2 as the oxidant. The films are epitaxial, single phase, and oriented in the ZnO [002] direction. Low temperature photoluminescence spectra at 20 K show for the formation of acceptor states associated with the As, although Hall measurements indicate that the films doped with 0.2 at% As are n-type. From the photoluminescence spectra, the binding energy of the As-related acceptor was estimated to be 155 meV. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

Ohmic contacts to p-type CuCrO 2 using Ni/Au/CrB 2 /Ti/Au contact metallurgy are reported. The samples were annealed in the 200–700 °C range for 60 s in flowing oxygen ambient. A minimum specific contact resistance of ∼2 × 10 −5 Ω cm 2 was obtained after annealing at 400 °C. Further increase in the annealing temperature (>400 °C) resulted in the degradation of contact resistance. Auger Electron Spectroscopy (AES) depth profiling showed that out-diffusion of Ti to the surface of the contact stacks was evident by 400 °C, followed by Cr at higher temperature. The CrB 2 diffusion barrier decreases the specific contact resistance by almost two orders of magnitude relative to Ni/Au alone.

Ir-based electrical contacts to p-type GaN have been fabricated and characterized. Both GaN//Ni/Au/Ir/Au and GaN//Ni/Ir/Au contact structures were deposited, however, only the former produced Ohmic current–voltage characteristics. At an anneal temperature of 500 °C, the Ni/Au/Ir/Au contact had a specific contact resistance of ∼2 × 10−4 Ω cm2, comparable or superior to conventional Ni/Au contacts that are less thermally stable. Anneal temperatures above 500 °C caused the Ir-based contact to fail. Auger electron spectroscopy was used to obtain depth profiles of both types of contacts at a variety of temperatures in order to provide insight into the mechanism of Ohmic formation as well as potential reasons for failure. A comparison to other metallization schemes on p-GaN is also given.

Annealing at temperatures up to 1000 °C is shown to decrease band edge photoluminescence in bulk ZnO crystals and increase deep level-related emission. The surface roughens for anneals in the range of 600–800 °C as O is lost preferentially from the surface, but at 900 °C the morphology improves as excess Zn is also lost from the surface. Splitting of the peak in the rocking curve of the ZnO (0 0 2) plane after annealing at 900–1000 °C indicates that the substrate is a mosaic of two or more crystals oriented slightly differently from one another and we are detecting differences in orientation of some of the grains in different areas or small changes due to annealing. There was no significant change in bulk conductivity of the ZnO for anneals up to 1000 °C, suggesting that ion implantation followed by annealing may be an effective approach for doping in this.

The use of Ir diffusion barriers in Ni/Au-based Ohmic contacts to p-type CuCrO2 layers was investigated. A specific contact resistance of ~5 × 10−4 Ω cm2 was achieved after annealing at 500°C for the Ir-containing contacts, and the contacts were rectifying for lower anneal temperatures. In this case, the contact resistance was basically independent of the measurement temperature, indicating that tunneling is the dominant transport mechanism in the contacts. The morphology for the Ir-containing contacts was still smooth at 500°C although Auger electron spectroscopy depth profiling showed that some of the nickel had diffused to the surface and had oxidized. Contacts annealed at 800°C showed that some copper and most of the nickel had diffused to the surface and oxidized. The presence of the Ir diffusion barrier does increase the thermal stability of the contacts by ∼200°C compared to conventional Ni/Au contacts. By contrast, the use of other materials such as TaN, ZrN, and W2B5 as the diffusion barrier led to poorer thermal stability, with the contact resistance increasing sharply above 400°C.