A model was developed for the operation of a GaN pn junction vertical diode which includes rate equations for carrier capture and thermally activated emission by substitutional carbon impurities and carrier generation by ionizing radiation. The model was used to simulate the effect of ionizing radiation on the charge state of carbon. These simulations predict that with no applied bias, carbon is negatively charged in the n-doped layer, thereby compensating n-doping as experimentally observed in diodes grown by metal-organic chemical vapor deposition. With reverse bias, carbon remains negative in the depletion region, i.e., compensation persists in the absence of ionization but is neutralized by exposure to ionizing radiation. This increases charge density in the depletion region, decreases the depletion width, and increases the capacitance. The predicted increase in capacitance was experimentally observed using a pulsed 70 keV electron beam as the source of ionization. In additional confirming experiments, the carbon charge-state conversion was accomplished by photoionization using sub-bandgap light or by the capture of holes under forward bias.
Accurate predictions of device performance in 14-MeV neutron environments rely upon understanding the recoil cascades that may be produced. Recoils from 14-MeV neutrons impinging on both gallium nitride (GaN) and gallium arsenide (GaAs) devices were modeled and compared to the recoil spectra of devices exposed to 14-MeV neutrons. Recoil spectra were generated using nuclear reaction modeling programs and converted into an ionizing energy loss (IEL) spectrum. We measured the recoil IEL spectra by capturing the photocurrent pulses produced by single neutron interactions with the device. Good agreement, with a factor of two, was found between the model and the experiment under strongly depleted conditions. However, this range of agreement between the model and the experiment decreased significantly when the bias was removed, indicating partial energy deposition due to cascades that escape the active volume of the device not captured by the model. Consistent event rates across multiple detectors confirm the reliability of our neutron recoil detection method.
This report documents work done at the Sandia Ion Beam Laboratory to develop a capability to produce 14 Me neutrons at levels sufficient for testing radiation effects on electronic materials and components. The work was primarily enabled by a laboratory directed research and development (LDRD) project. The main elements of the work were to optimize target lifetime, test a new thin- film target design concept to reduce tritium usage, design and construct a new target chamber and beamline optimized for high-flux tests, and conduct tests of effects on electronic devices and components. These tasks were all successfully completed. The improvements in target performance and target chamber design have increased the flux and fluence of 14 MV neutrons available at the test location by several orders of magnitude. The outcome of the project is that a new capability for testing radiation-effects on electronic components from 14 MeV neutrons is now available at Sandia National Laboratories. This capability has already been extensively used for many qualification and component evaluation and development tests.
An experimental study of migration of tungsten in the DIII-D divertor is described, in which the outer strike point of L-mode plasmas was positioned on a toroidal ring of tungsten-coated metal inserts. Net deposition of tungsten on the divertor just outside the strike point was measured on graphite samples exposed to various plasma durations using the divertor materials evaluation system. Tungsten coverage, measured by Rutherford backscattering spectroscopy (RBS), was found to be low and nearly independent of both radius and exposure time closer to the strike point, whereas farther from the strike point the W coverage was much larger and increased with exposure time. Depth profiles from RBS show this was due to accumulation of thicker mixed-material deposits farther from the strike point where the plasma temperature is lower. These results are consistent with a low near-surface steady-state coverage on graphite undergoing net erosion, and continuing accumulation in regions of net deposition. This experiment provides data needed to validate, and further improve computational simulations of erosion and deposition of material on plasma-facing components and transport of impurities in magnetic fusion devices. Such simulations are underway and will be reported later.
The formulation of carrier transport through heterojunctions by tunneling and thermionic emission is derived from first principles. The treatment of tunneling is discussed at three levels of approximation: numerical solution of the one-band envelope equation for an arbitrarily specified potential profile; the WKB approximation for an arbitrary potential; and, an analytic formulation assuming constant internal field. The effects of spatially varying carrier chemical potentials over tunneling distances are included. Illustrative computational results are presented. The described approach is used in exploratory physics models of irradiated heterojunction bipolar transistors within Sandia's QASPR program.
The energy-dependent probability density of tunneled carrier states for arbitrarily specified longitudinal potential-energy profiles in planar bipolar devices is numerically computed using the scattering method. Results agree accurately with a previous treatment based on solution of the localized eigenvalue problem, where computation times are much greater. These developments enable quantitative treatment of tunneling-assisted recombination in irradiated heterojunction bipolar transistors, where band offsets may enhance the tunneling effect by orders of magnitude. The calculations also reveal the density of non-tunneled carrier states in spatially varying potentials, and thereby test the common approximation of uniform- bulk values for such densities.