Search for: All records

β-Ga2O3 has attracted much recent attention as a promising ultrawide bandgap semiconductor. Hydrogen can affect the conductivity of β-Ga2O3 through the introduction of shallow donors and the passivation of deep acceptors. The introduction of H or D into β-Ga2O3 by annealing in an H2 or D2 ambient at elevated temperature produces different classes of O–H or O–D centers. This work is a study of the interaction of D with VGa1 and VGa2 deep acceptors as well as other impurities and native defects in Ga2O3 by infrared spectroscopy and the complementary theory. (We focus primarily on the deuterium isotope of hydrogen because the vibrational modes of O–D centers can be detected with a higher signal-to-noise ratio than those of O–H.) O–D centers in β-Ga2O3 evolve upon annealing in an inert ambient and are transformed from one type of O–D center into another. These reactions affect the compensation of unintentional shallow donors by deep acceptors that are passivated by D. Defects involving additional impurities in β-Ga2O3 compete with VGa deep acceptors for D and modify the deuterium-related reactions that occur. The defect reactions that occur when D is introduced by annealing in a D2 ambient appear to be simpler than those observed for other introduction methods and provide a foundation for understanding the D-related reactions that can occur in more complicated situations. 

In this work, we demonstrate the rejuvenation of Ti/4H-SiC Schottky barrier diodes after forward current-induced degradation, at room temperature and in a few seconds, by exploiting the physics of high-energy electron interactions with defects. The diodes were intentionally degraded to a 42% decrease in forward current and a 9% increase in leakage current through accelerated electrical stressing. The key feature of our proposed rejuvenation process is very high current density electrical pulsing with low frequency and duty cycle to suppress any temperature rise. The primary stimulus is, therefore, the electron wind force, which is derived from the loss of the momentum of the high energy electrons upon collision with the defects. Such defect-specific or “just in location” mobilization of atoms allows a significant decrease in defect concentration, which is not possible with conventional thermal annealing that requires higher temperatures and longer times. We show evidence of rejuvenation with additional improvement in leakage current (16%) and forward current (38%) beyond the pristine condition. Transmission electron microscopy, geometric phase analysis, Raman spectroscopy, and energy dispersive x-ray-spectroscopy reveal the enhancement of defects and interfaces. The ultrafast and room temperature process has the potential for rejuvenating electronic devices operating in high power and harsh environmental conditions. 

There are numerous applications for deep UV AlGaN Light-Emitting Diodes (LEDs) in virus inactivation, air and water purification, sterilization, bioagent detection and UV polymer curing. The long-term stability of these LEDs is also of interest for long-duration space missions such as the Laser Interferometer Space Antenna (LISA), the first gravitational wave detector in space. We review the literature on long-term aging of these devices as a function of drive current, temperature and dc versus pulsed operation. The LEDs typically show a gradual decline in output power (up to 50%) over extended operating times (>100 h) and the rate of decline is mainly driven by current and temperature. Experimentally, the degradation rate is dependent on the cube of drive current density and exponentially on temperature. The main mechanism for this decline appears to be creation/migration of point defects. Pre-screening by considering the ratio of band edge-to-midgap emission and LED ideality factor is effective in identifying populations of devices that show long lifetimes (>10,000 h), defined as output power falling to 70% of the initial value.

 

While radiation is known to degrade AlGaN/GaN high-electron-mobility transistors (HEMTs), the question remains on the extent of damage governed by the presence of an electrical field in the device. In this study, we induced displacement damage in HEMTs in both ON and OFF states by irradiating with 2.8 MeV Au⁴⁺ion to fluence levels ranging from$1.72\times {10}^{10}$to$3.745\times {10}^{13}$ions cm⁻², or 0.001–2 displacement per atom (dpa). Electrical measurement is donein situ, and high-resolution transmission electron microscopy (HRTEM), energy dispersive x-ray (EDX), geometrical phase analysis (GPA), and micro-Raman are performed on the highest fluence of Au⁴⁺irradiated devices. The selected heavy ion irradiation causes cascade damage in the passivation, AlGaN, and GaN layers and at all associated interfaces. After just 0.1 dpa, the current density in the ON-mode device deteriorates by two orders of magnitude, whereas the OFF-mode device totally ceases to operate. Moreover, six orders of magnitude increase in leakage current and loss of gate control over the 2-dimensional electron gas channel are observed. GPA and Raman analysis reveal strain relaxation after a 2 dpa damage level in devices. Significant defects and intermixing of atoms near AlGaN/GaN interfaces and GaN layer are found from HRTEM and EDX analyses, which can substantially alter device characteristics and result in complete failure.

 

Abstract Films of α-Ga2O3 (Sn) grown by Halide Vapor Phase Epitaxy (HVPE) on sapphire with starting net donor densities in the range 5×1015- 8.4×1019 cm-3 were irradiated at room temperature with 1.1 MeV protons to fluences from 1013 -1016 cm-2. For the lowest doped samples, the carrier removal rate was ~35 cm-1 at 1014 cm-2 and ~1.3 cm-1 for 1015 cm-2 proton fluence. The observed removal rate could be accounted for by the introduction of deep acceptors with optical ionization energies of 2 eV, 2.8 eV and 3.1 eV. For doped samples doped at 4x1018 cm-3, the initial electron removal rate was 5×103 cm-1 for 1015 cm-2 proton fluence and ~300 cm-1 for 1016 cm-2 proton fluence. The same deep acceptors were observed in photocapacitance spectra, but their introduction rate was orders of magnitude lower than the carrier removal rate. For the heaviest doped samples, an electron removal rate could be measured only after irradiation with the highest proton fluence of 1016 cm-2 and was close to that measured for the 4×1018 cm-3 sample after exposure to the same fluence. Possible reasons for the observed behavior are discussed and radiation tolerances of lightly doped α-Ga2O3 films is higher than for similarly doped β-Ga2O3 layers.  

While a number of O-H and O-D vibrational lines have been observed for hydrogen and deuterium in β-Ga2O3, it has been commonly reported that there is no absorption with a component of the polarization E parallel to the [010], or b, axis. This experimental result has led to O-H defect structures that involve shifted configurations of a vacancy at the tetrahedrally coordinated Ga(1) site [VGa(1)] and have ruled out structures that involve a vacancy at the octahedrally coordinated Ga(2) site [VGa(2)], because these structures are predicted to show absorption for E//[010]. In this Letter, weak O-D lines at 2475 and 2493 cm−1 with a component of their polarization with E//[010] are reported for β-Ga2O3 that had been annealed in a D2 ambient. O-D defect structures involving an unshifted VGa(2) are proposed for these centers. An estimate is made that the concentration of VGa(2) in a Czochralski-grown sample is 2–3 orders of magnitude lower than that of VGa(1) from the intensities of the IR absorption lines.

 

Strain localization in microelectronic devices commonly arises from device geometry, materials, and fabrication processing. In this study, we controllably relieve the local strain field of AlGaN/GaN HEMTs by milling micro-trenches underneath the channel and compare the device performance as a function of the relieved strain as well as radiation dosage. Micro-Raman results suggest that the trenches locally relax the strain in device layers, decreasing the 2DEG density and mobility. Intriguingly, such strain relaxation is shown to minimize the radiation damage, measured after 10 Mrads of 60Co-gamma exposure. For example, a 6-trench device showed only ∼8% and ∼6% decrease in saturation drain current and maximum transconductance, respectively, compared to corresponding values of ∼15% and ∼30% in a no-trench device. Negative and positive threshold voltage shifts are observed in 6-trench and no-trench devices, respectively, after gamma radiation. We hypothesize that the extent of gamma radiation damage depends on the strain level in the devices. Thus, even though milling a trench decreases 2DEG mobility, such decrease under gamma radiation is far less in a 6-trench device (∼1.5%) compared to a no-trench device (∼20%) with higher built-in strain.

 

Radiation damage in electronic devices is known to be influenced by physics, design, and materials system. Here, we report the effects of biasing state (such as ON and OFF) and pre-existing damage in GaN high electron mobility transistors exposed to γ radiation. Controlled and accelerated DC biasing was used to prestress the devices, which showed significant degradation in device characteristics compared to pristine devices under ON and OFF states after γ irradiation. The experiment is performed in situ for the ON-state to investigate transient effects during irradiation until the total dose reaches 10 Mrad. It shows that threshold voltage, maximum transconductance, and leakage current initially decrease with dosage but slowly converge to a steady value at higher doses. After 10 Mrad irradiation, the OFF-state device demonstrates larger RON and one order of magnitude increased leakage current compared to the ON-state irradiated device. The micro-Raman study also confirms that the ON-state operation shows more radiation hardness than OFF and prestressed devices. Prestressed devices generate the highest threshold voltage shift from −2.85 to −2.49 V and two orders of magnitude higher leakage current with decreased saturation current after irradiation. These findings indicate that high electric fields during stressing can generate defects by modifying strain distribution, and higher defect density can not only create more charges during irradiation but also accelerate the diffusion process from the ionizing track to the nearest collector and consequently degrade device performances.

 

As electronic systems become larger and more complex, detection of the most vulnerable regions (MVR) to radiation exposure becomes more difficult and time consuming. We present a heuristic approach where the mechanical and thermal aspects of devices are exploited to quickly identify MVRs. Our approach involves the topological mapping of two device conditions. The first condition identifies regions with the highest mechanical strain or density of defects and interfaces via thermal wave probing and phase analysis. The second condition identifies regions with high electrical field. It is hypothesized that the region with the highest thermal wave penetration resistance and electrical field will exhibit the highest sensitivity to incoming radiation for single events and potentially, total ionizing dose. Our approach implements a simplistic design that improves analysis time by ∼2–3 orders of magnitude over current radiation sensitivity mapping methods. The design is demonstrated on the well-studied operational amplifier LM124, which shows agreement with the literature in identifying sensitive transistors–QR1, Q9, and Q18–with relatively high phase percentile values (>70%) and ΔT percentiles (>50%), satisfying conditions for elevated radiation susceptibility. This is followed by experimental results on a static random access memory (HM-6504) and a Xilinx Artix-7 35 T system on a chip.