Arthur B. Ellis

p-GaN surface treatments for metal contacts

The chemical bonding and electronic properties of wet, chemically treated p-GaN surfaces were studied using synchrotron radiation photoemission spectroscopy. Chlorine-based chemical bonding was identified on the conventional HCl-treated p-GaN surface, which is associated with a shift of the surface Fermi level toward the conduction band edge by ∼0.9 eV with respect to the thermally cleaned surface. Compared to the HCl-treated surface, the surface Fermi level on the KOH-treated surface lies about ∼1.0 eV closer to the valence band edge, resulting in a much smaller surface barrier height to p-type materials than the HCl-treated surface. The smaller surface barrier height to p-GaN after KOH treatment can lead to a lower contact resistivity and can play an important role in lowering the metal contact resistivity to p-GaN.

Shape memory effect in nanoindentation of nickel–titanium thin films

In this study, a series of nanoindentations was made on NiTi shape memory alloy thin films at millinewton loads with a Berkovich indenter. Mapping of the indentation topography using atomic force microscopy reveals direct evidence that the thermally induced martensitic transformation of these films allows for partial indent recovery on the nanoscale. Indeed, recovery is nearly complete at indentation depths of less than 100 nm. A hemispherical cavity model is presented to predict an upper limit to shape memory recovery of sharp indentations.

Oriented nickel‐titanium shape memory alloy films prepared by annealing during deposition

Nickel‐titanium shape memory alloy films, between 2 and 10 μm thick, were sputter deposited onto (100) silicon substrates. Films deposited onto a substrate at ambient temperature were amorphous; however, several post‐deposition annealing procedures produced crystalline films exhibiting the B2‐to‐B19’ phase transition that gives rise to the shape memory effect. Films that were deposited onto a heated substrate, 350–460 °C, crystallized during deposition, eliminating the need for a separate annealing step. Powder x‐ray diffraction indicated that these films were highly oriented, with the NiTi (110)B2 face parallel to the silicon substrate (100) face.

X-ray photoemission determination of the Schottky barrier height of metal contacts to n-GaN and p-GaN

Synchrotron radiation-based x-ray photoemission spectroscopy was used to study the surface Fermi level position within the band gap for thin metal overlayers of Au, Al, Ni, Ti, Pt, and Pd on n–GaN and p–GaN. Nonequilibrium effects were taken into account by measuring the Fermi edge of the metal overlayer. There are two different behaviors observed for the six metals studied. For Au, Ti, and Pt, the surface Fermi level lies about 0.5-eV higher in the gap for n-type than for p-type GaN. For Ni, Al, and Pd, the surface Fermi level position is independent of doping, but varies from one metal to the other. Results for Ni, Pd, and Al fit a modified Schottky–Mott theory, while Au, Ti, and Pt demonstrate a more complex behavior. Atomic force microscopy was used along with photoemission to investigate the growth mode of each metal on the GaN surface.

n-GaN surface treatments for metal contacts studied via x-ray photoemission spectroscopy

The surface chemistry and electronic properties of n-GaN surfaces were studied via x-ray photoemission spectroscopy before and after wet chemical treatments. Shifts of the surface Fermi level were measured with the change in position of the Ga 3d core level peak. HCl treatment of n-GaN led to a 0.9 eV shift of the surface Fermi level toward the conduction band minimum, while KOH treatment led to a 0.3 eV shift of the surface Fermi level toward the valance band maximum. These shifts lead to a reduction in the surface barrier for HCl-treated n-GaN and for KOH-treated p-GaN, potentially improving contact resistance. The changes in surface chemistry indicate that a N (or Ga) deficiency with HCl(KOH) treatment alters the surface state density through the formation of donor (acceptor)-like states.

X-ray photoemission spectroscopic investigation of surface treatments, metal deposition, and electron accumulation on InN

The effects of surface chemical treatments and metal deposition on the InN surface are studied via synchrotron-based photoemission spectroscopy. Changes in the In 4d core level as well as the valence band spectra are reported. The surface Fermi level position, EF, relative to the valence band maximum was determined for both Au and Ti Schottky barriers. EF lies at an energy of 0.7 eV above the valence band maximum for Au deposited on annealed InN and 1.2 eV above the valence band maximum for Ti deposited on Ar-sputtered InN. These results that the surface Fermi level lays at or above the conduction band maximum when a value of InN band gap of 0.7–0.9 eV is assumed.

Electro-optical evidence for the chelate effect at semiconductor surfaces.

Monoamines and diamines dissolved in cyclohexane solution reversibly enhance the band-edge photoluminescence (PL) intensity of immersed n-type cadmium sulfide (n-CdS) and n-type cadmium selenide (n-CdSe) substrates through adsorption. The magnitude of the PL increase is used to estimate amine-induced contractions in the semiconductors depletion width, and the dependence of the PL intensity on amine concentration provides an estimate of the adduct formation constant. Two diamines, ethylenediamine and o-phenylenediamine, exhibit unusually low reductions in depletion width and substantially larger adduct equilibrium constants relative to the other amines studied, consistent with chelation to surface Cd2+ ions. These studies demonstrate that PL can be used as a contactiess, in situ technique for characterizing the steric and electronic landscape of semiconductor surfaces and for correlating molecular and surface chemistry.

Photoluminescent Properties of n-GaAs Electrodes: Applications of the Dead-Layer Model to Photoelectrochemical Cells.

Single-crystal samples of n-type GaAs have been used as electrodes in photoelectrochemical cells (PECs) employing aqueous ditelluride electrolyte. Photoluminescence (PL) from the electrodes can be quenched by the electric field present in the semiconductor during PEC operation. The extent of PL quenching, studied as a function of carrier concentration, excitation wavelength, and applied potential, is consistent with the dead-layer model previously used to describe PL quenching in semiconductor/metal, Schottky-barrier systems. PL quenching curves calculated by assuming that the dead-layer thickness varies with applied potential in the same manner as the depletion width differ from the experimental data, particularly in the region near the flat-band potential. Sources of these discrepancies are discussed, including the possibility that relative PL intensity reflects the manner in which applied potential is partitioned across the semiconductor-electrolyte interface.

Photoluminescent Properties of n-GaAs Electrodes: Simultaneous Determination of Depletion Widths and Surface Hole-Capture Velocities in Photoelectrochemical Cells.

Steady‐state photoluminescence measurements performed on n‐GaAs electrodes used in photoelectrochemical cells (PEC’s) employing a stabilizing, aqueous telluride electrolyte yield values for the electrode’s depletion width W and surface hole‐capture velocity S. Between −1.0 V (a potential near short circuit) and −1.5 V vs an SCE reference electrode (a potential near open circuit at the photon flux of 1×1015 photons/s cm2 employed), the interface behaves ideally: virtually all of the applied potential appears in the semiconductor space‐charge region. Over this potential regime S is determined to be constant to within 10% and has a value, using literature values for hole lifetime and diffusion length, of approximately 2×105 cm/s for n‐GaAs electrodes having carrier concentrations of (1–4)×1017 cm−3. Similar values of S obtained in air and in the PEC suggest a common rate‐limiting mechanism for hole consumption in the two media.

LUMINESCENT PHOTOELECTROCHEMICAL CELLS - 7. PHOTOLUMINESCENT AND ELECTROLUMINESCENT PROPERTIES OF CADMIUM SULFO-SELENIDE ELECTRODES.

Photoelectrochemical cells (PECs) are being widely studied as devices for optical energy conversion. The excited-state properties of the semiconductors around which PECs are constructed are crucial to efficient energy conversion. We have employed luminescence as a probe of these excited-state-properties, generally using materials such as n-type CdS:Te(Te-doped CdS) which exhibit subband gap emission. Recently we examined emission of band gap energy from n-type CdS and CdSe, two materials which have been used extensively in PEC studies. Since these two compounds form solid solutions over the entire composition range, the mixed compounds represent a natural extension of our emissive studies. We report herein that luminescence from samples of n-type, single-crystal CdSXSe1-X can be used to probe interfacial charge-transfer events relevant to PECs. Specifically, we demonstrate that photoluminescence (PL) can be perturbed and electroluminescence (EL) initiated by interfacial charge-transfer processes.