Donald J. Suntrup

Barrier height inhomogeneity and its impact on (Al,In,Ga)N Schottky diodes

III-N materials, especially ternary and quaternary alloys, are profoundly affected by barrier height inhomogeneity as evidenced by great variability in reported barrier height and Richardson constant values for Schottky diode samples involving epilayers with identical material composition. Research into AlInGaN-based devices is gaining traction due to its usefulness for strain engineering, polarization engineering, and vertical device design. Thus it is important to characterize the Schottky barrier height between AlInGaN and technologically relevant metals like nickel. It is proposed that alloy composition fluctuations inherent to low-temperature III-N alloys result in a Schottky barrier height inhomogeneity, and that the Schottky barrier height follows a Gaussian distribution. Current vs voltage data as a function of temperature was measured for three AlInGaN samples of varying composition. Utilizing a model tailored to thermionic emission over a Gaussian distribution of barriers, both the average barri...

Measurement of the hot electron mean free path and the momentum relaxation rate in GaN

We present a method for measuring the mean free path and extracting the momentum relaxation time of hot electrons in GaN using the hot electron transistor (HET). In this device, electrons are injected over a high energy emitter barrier into the base where they experience quasi-ballistic transport well above the conduction band edge. After traversing the base, high energy electrons either surmount the base-collector barrier and become collector current or reflect off the barrier and become base current. We fabricate HETs with various base thicknesses and measure the common emitter transfer ratio (α) for each device. The mean free path is extracted by fitting α to a decaying exponential as a function of base width and the relaxation time is computed using a suitable injection velocity. For devices with an injection energy of ∼1 eV, we measure a hot electron mean free path of 14 nm and calculate a momentum relaxation time of 16 fs. These values are in agreement with theoretical calculations where longitudinal optical phonon scattering is the dominant momentum relaxation mechanism.

In situ metalorganic chemical vapor deposition of Al2O3 on N-face GaN and evidence of polarity induced fixed charge

The in situ metalorganic chemical vapor deposition (MOCVD) of Al2O3 dielectrics on N-face GaN is reported. The near-interface fixed charges are measured by using capacitance-voltage techniques on a metal-oxide-semiconductor (MOSCAP) structure, and the results are compared with those obtained on Ga-face MOSCAPs with the same in situ MOCVD Al2O3 dielectrics. The influence of GaN polarity as well as other possible mechanisms on the formation of fixed charge are identified and discussed.

Fiber-connectorized micropillar cavities

We present a cryogenically compatible method for permanently connecting and coupling a single mode fiber to a single mode of a micropillar cavity with embedded quantum dots (QDs). Efficient coupling of up to 40% was measured which requires a 300 nm positioning accuracy that remains preserved during the fiber attachment procedure and during cool-down to 4 K. Fiber coupling, as opposed to conventional free space coupling, makes it possible to connect many such QD-cavity systems within the same cryostat which can interact through an external optical network, facilitating the implementation of hybrid photon/confined-electron schemes for quantum communication and information processing.

Measuring the signature of bias and temperature-dependent barrier heights in III-N materials using a hot electron transistor

We present a measurement technique to study barrier height inhomogeneity (BHI) in III-N materials. This technique is enabled by a hot electron transistor (HET), a vertical, unipolar device that works by injecting hot electrons over the emitter barrier and into a very thin base layer. After traversing the base, high energy electrons surmount the collector barrier and contribute to the collector current while low energy electrons reflect off the collector barrier and become the base current. The prevailing theory of BHI prescribes the replacement of a constant emitter barrier height with one that depends on both bias and temperature (i.e. ). Because the magnitude of the collector current is a strong function of the emitted electron energy and, therefore, the emitter barrier height, measuring the change in collector current with emitter bias and temperature allows us to determine the dependence of on these quantities. This advance will help provide a more thorough understanding of the physical sources of BHI in the III-Ns and assist in the diagnosis of key device nonidealities.

Design Space of III-N Hot Electron Transistors Using AlGaN and InGaN Polarization-Dipole Barriers

Transistor operation by common emitter (CE) current modulation is shown for the first time in III-N hot electron transistors (HETs). The emitter and collector barriers (φ_BE and φ_BC) are implemented using Al_0.45 Ga_0.55N and In_0.1Ga_0.9N layers as polarization dipoles, respectively. CE modulation is achieved by increasing the E-B barrier height beyond the B-C barrier height by increasing the Al_0.45Ga_0.55N thickness (t). Similar CE performance is seen in the identical HET structures grown on both bulk GaN and sapphire. A maximum α of ~0.3 is achieved using a GaN base thickness of 10 nm. The InGaN dipole used as the collector barrier is shown to be instrumental in enabling ohmic base contacts, low base sheet resistance, and low collector leakage, simultaneously.

Barrier height fluctuations in InGaN polarization dipole diodes

We present an analysis of the effects of lateral barrier height fluctuations on the transport properties of an InGaN polarization dipole diode (PDD). Typical diode analysis proceeds by performing a linear fit to the data contained in a Richardson plot in order to extract the zero-bias barrier height ( ϕB) and the Richardson constant ( A*). The experimental PDD Richardson plot, however, is highly nonlinear and we attribute this to a spatially nonuniform distribution of barrier heights. After modeling the barrier height distribution using a Gaussian, we fit the Richardson data with a modified, second-order function from which we extract the Richardson constant, mean barrier height, and standard deviation simultaneously. We propose that the physical cause of the observed barrier height inhomogeneity in the PDD is statistical nanoscale fluctuations in indium composition.

Monitoring the formation of oxide apertures in micropillar cavities

An imaging technique is presented that enables monitoring of the wet thermal oxidation of a thin AlAs layer embedded between two distributed Bragg reflector mirrors in a micropillar. After oxidation we confirm by white light reflection spectroscopy that high quality optical modes confined to a small volume have been formed. The combination of these two optical techniques provides a reliable and efficient way of producing oxide apertured micropillar cavities for which the wet thermal oxidation is a critical fabrication step.

Common emitter operation of III-N HETs using AlGaN and InGaN polarization-dipole induced barriers

In conclusion, we show that injection energy and base thickness are the critical parameters for achieving gain in III-N HETs. It is important to make sure that high leakage currents and large base resistance do not result in inaccurate extraction of gain and transfer ratio. A better understanding of the design space is required to achieve CE current gain and move towards our goal of building a high-frequency III-N HET.

Fine tuning of micropillar cavity modes through repetitive oxidations

Repetitive wet thermal oxidations of a tapered oxide aperture in a micropillar structure are demonstrated. After each oxidation step the confined optical modes are analyzed at room temperature. Three regimes are identified. First, the optical confinement increases when the aperture oxidizes toward the center. Then, the cavity modes shift by more than 30 nm when the taper starts to oxidize through the center, leading to a decrease in the optical path length. Finally, the resonance frequency levels off when the aperture is oxidized all the way through the micropillar, but confined optical modes with a high quality factor remain. This repetitive oxidation technique therefore enables precise control of the optical cavity volume or wavelength.