GaAs Fermi Level Research Articles

Controlled Growth of Heterostructured Ga/GaAs Nanowires with Sharp Schottky Barrier

Because of the inevitable Fermi level pinning on surface/interface states of nanowires, achieving high-performance nanowire devices with controllable nanoscale contacts is always challenging but important. Herein, single-crystalline heterostructured Ga/GaAs nanowires with sharp hetero-Schottky interfaces have been successfully synthesized on amorphous substrates by utilizing Au nanoparticles as catalytic seeds via chemical vapor deposition. These nanowires are found to grow with the hemispherical Au7Ga2 catalytic tips following the vapor–liquid–solid mechanism. During the growth, simply by manipulating the source and growth temperatures, the Ga precipitation rate from Au–Ga alloy tips as well as the reaction rate of Ga precipitates with As can be reliably controlled in order to tailor the length (0–170 nm) of Ga nanowire segments obtained in the heterostructure. When configured into field-effect transistors, these Ga/GaAs NWs exhibit the p-type conductivity with a sharp hetero-Schottky barrier of ∼1.0 eV at the atomically connected Ga segment/GaAs NW body interface, in which this barrier height is close to the theoretical difference between the GaAs Fermi level (5.1–5.3 eV) and the Ga work function (∼4.3 eV), suggesting the effective formation of nanoscale contact by minimizing the Fermi level pinning, being advantageous for advanced nanoelectronics.

Detecting Fermi-level shifts by Auger electron spectroscopy in Si and GaAs

In this paper, changes in surface Fermi-level of Si and GaAs, caused by doping and cleaning, are investigated by Auger electron spectroscopy. Based on the Auger voltage contrast, we compared the Auger transition peak energy but with higher accuracy by using a more accurate analyzer and an improved peak position determination method.For silicon, a peak shift as large as 0.46 eV was detected when comparing a cleaned p-type and n-type wafer, which corresponds rather well with the theoretical difference in Fermi-levels. If no cleaning was applied, the peak position did not differ significantly for both wafer types, indicating Fermi-level pinning in the band gap.For GaAs, peak shifts were detected after cleaning with HF and (NH4)2S-solutions in an inert atmosphere (N2-gas). Although the (NH4)2S-cleaning in N2 is very efficient in removing the oxygen from the surface, the observed Ga- and As-peak shifts are smaller than those obtained after the HF-cleaning. It is shown that the magnitude of the shift is related to the surface composition. After Si-deposition on the (NH4)2S-cleaned surface, the Fermi-level shifts back to a similar position as observed for an as-received wafer, indicating that this combination is not successful in unpinning the Fermi-level of GaAs.

Compound semiconductor MOSFETs

Enhancement mode, high electron mobility MOSFET devices have been fabricated using an oxide high- κ gate dielectric stack developed using molecular beam epitaxy. A template layer of Ga 2O 3, initially deposited on the surface of the III-V device unpins the GaAs Fermi level while a (Gd xGa 1−x) 2O 3 bulk ternary layer forms the highly resistive layer to reduce leakage current through the dielectric stack. A midgap interface state density of ∼2 × 10 11 cm −2 eV −1 and a dielectric constant of 20 are determined using electrical measurements.. N-channel MOSFETs with a gate length of 1 μm and a source-drain spacing of 3 μm show a threshold voltage, saturation current and transconductance of 0.11 V, 380 mA/mm and 250 mS/mm, respectively.

Development of GaAs-based MOSFET using molecular beam epitaxy

Enhancement mode, high electron mobility PHEMT-based metal oxide semi conductor field effect transistor (MOSFET) devices have been fabricated using an oxide high- κ gate dielectric stack developed using molecular beam epitaxy. A template layer of Ga 2 O 3 , initially deposited on the surface of the III–V device unpins the GaAs Fermi level while the deposition of a bulk ternary ( Gd x Ga 1 - x ) 2 O 3 layer forms the highly resistive layer to reduce leakage current through the dielectric stack. The use of molecular beam epitaxy allows for the control and uniformity of the oxide layers along the growth direction and deposition conditions were optimized for oxide surface morphology and interface quality. A midgap interface state density for the high- κ stack on GaAs of ≅ 2 × 10 11 cm - 2 eV - 1 and a dielectric constant of κ ≅ 20 are determined using electrical measurements. Enhancement-mode n-channel MOSFETs with a gate length of 1 μ m and a source-drain spacing of 3 μ m show a threshold voltage, saturation current, transconductance, and on-resistance of 0.11 V, 380 mA mm - 1 , 250 mS mm - 1 , and 4.5 Ω mm , respectively.

III-V MOSFET's with Advanced High k Dielectrics

Ga2O3(Gd2O3), the novel oxide, which was electron-beam evaporated from a gallium- gadolinium-garnet target in UHV, has unpinned the GaAs Fermi level. Systematic heat treatments under various gases and temperatures were studied to achieve low leakage currents of 10-8 - 10-9 A/cm2 and low Dit's in the range of 4-9 x 1010 cm-2V-1 for Ga2O3(Gd2O3) on InGaAs. By removing moisture from the oxide, thermodynamic stability of the Ga2O3(Gd2O3)/GaAs hetero-structures was achieved with high temperature annealing, which is needed for fabricating inversion-channel MOSFET's. The oxide remains amorphous and the interface remains intact with atomic smoothness and sharpness. The mechanism of Fermi-level unpinning and the band parameters in atomic layer deposition (ALD)-Al2O3 ex-situ deposited on InGaAs were studied and unveiled. The work of inversion- channel, depletion- mode, and power GaAs MOSFET's using Ga2O3 (Gd2O3) as the gate dielectric was briefly reviewed.

Advanced High K Dielectrics for Nano Electronics - Science and Technologies

Our research activities on high k dielectrics on GaAs, Si, and GaN are reviewed. Novel Ga2O3(Gd2O3) has unpinned the GaAs Fermi level for the first time. Systematic heat treatments under various gases and temperatures were studied to achieve low leakage currents of 10-8 - 10-9 A/cm2 and low Dit's in the range of 4-9 x 1010 cm-2V-1. By removing moisture from the oxide, thermodynamic stability of Ga2O3(Gd2O3)/GaAs is achieved. The mechanism of Fermi-level unpinning in ALD-Al2O3 on InGaAs was unveiled. Single crystal Gd2O3 was found to grow epitaxially on GaAs and unpin its Fermi level. Inversion- channel, depletion-mode, and power GaAs MOSFET's using Ga2O3 (Gd2O3) as the gate dielectric are discussed. The advances in MBE-HfO2 on Si and the MOSFET's with metal gate of TiN and gate dielectric of HfO2 are discussed. Recent work of nano- thick single crystal oxides of Al2O3 and Sc2O3 epitaxially grown on Si (111) is discussed.

Gate dielectric on compound semiconductors by molecular beam epitaxy

An oxide-based high-κ gate dielectric stack for GaAs has been developed using molecular beam epitaxy. A template layer of Ga2O3, initially deposited on a GaAs(001) surface, serves to unpin the GaAs Fermi level while the deposition of a bulk ternary (GdxGa1−x)2O3 layer forms the highly resistive layer to reduce leakage current through the dielectric stack. The use of molecular beam epitaxy allows for the control and uniformity of the oxide layers along the growth direction and deposition conditions were optimized for oxide surface morphology and interface quality. The midgap interface state density for the high-κ stack on GaAs of ≅2×1011cm−2eV−1 and a dielectric constant of k≅20 are determined using electrical measurements. Metal oxide semiconductor field effect transistor layer structures were produced by replacing the GaAs cap layer in a pseudomorphic high electron mobility transistor-type layer structure with the high-κ gate dielectric stack.

Gate dielectrics on compound semiconductors

An oxide-based high-κ gate dielectric for GaAs has been developed using molecular beam epitaxy. A template layer of Ga2O3 is initially deposited on a GaAs(001) surface followed by the deposition of a bulk (GdxGa1−x)2O3 layer. The Ga2O3 layer serves to unpin the GaAs Fermi level while the GdGaO layer forms the high resistive layer to reduce leakage current through the dielectric stack. Deposition conditions during MBE were optimized for oxide surface morphology and interface quality. The use of molecular beam epitaxy also allowed for the control and uniformity of the ternary GdGaO layer along the growth direction. The midgap interface state density for the high-κ stack on GaAs of≅2×1011 cm−2 eV−1, and a dielectric constant of k≅20 were determined using electrical measurements. MOSFET devices were produced by replacing the GaAs cap layer in a PHEMT-type structure with a thick oxide layer. A contactless method was used to determine the mobilities from such devices which were comparable to those measured on a GaAs PHEMT using conventional Hall measurement.

Effect of arsenic precipitates on Fermi level in GaAs grown by molecular-beam epitaxy at low temperature

A simple model is presented to discuss the effect of As precipitates on the Fermi level in GaAs grown by molecular-beam epitaxy at low temperature (LT-GaAs). This model implements the compensation between point defects and the depletion of arsenic precipitates. The condition that the Fermi level is pinned by As precipitates is attained. The shifts of the Fermi level in LT-GaAs with annealing temperature are explained by our model. Additionally, the role of As precipitates in conventional semi-insulating GaAs is discussed.

On the validity of the amphoteric-defect model in gallium arsenide and a criterion for Fermi-level pinning by defects

Using the theoretically calculated point-defect total-energy values of Baraff and Schluter in GaAs, anamphoteric-defect model has been proposed by Walukiewicz to explain a large number of experimental results. The suggested amphoteric-defect system consists of two point-defect species capable of transforming into each other: the doubly negatively charged Ga vacancyVGa2− and the triply positively charged defect complex (ASGa+VAs)3+, with AsGa being the antisite defect of an As atom occupying a Ga site andVAs being an As vacancy. When present in sufficiently high concentrations, the amphoteric defect systemVGa2−/(AsGa+VAs)3+ is supposed to be able to pin the GaAs Fermi level at approximately theEv+0.6 eV level position, which requires that the net free energy of theVGa/(AsGa+VAs) defect system to be minimum at the same Fermi-level position. We have carried out a quantitative study of the net energy of this defect system in accordance with the individual point-defect total-energy results of Baraff and Schluter, and found that the minimum net defect-system-energy position is located at about theEv+1.2 eV level position instead of the neededEv+0.6 eV position. Therefore, the validity of the amphoteric-defect model is in doubt. We have proposed a simple criterion for determining the Fermi-level pinning position in the deeper part of the GaAs band gap due to two oppositely charged point-defect species, which should be useful in the future.

Energy distribution of interface states in the band gap of GaAs determined from x-ray photoelectron spectra under biases.

The energy distribution of interface states in the GaAs band gap is determined for metal-oxide-semiconductor devices with an ultrathin thermal oxide layer of \ensuremath{\sim}3.8 nm, from measurements of x-ray photoelectron spectra under biases. The energy distribution has a peaked structure with four peaks at \ensuremath{\sim}0.15, \ensuremath{\sim}0.5, \ensuremath{\sim}0.75, and \ensuremath{\sim}1.1 eV above the valence-band maximum (VBM). The 0.75-eV peak has the highest density of \ensuremath{\sim}1.9\ifmmode\times\else\texttimes\fi{}${10}^{12}$ ${\mathrm{cm}}^{\mathrm{\ensuremath{-}}2}$ and is attributed to a (+/0) transition of ${\mathrm{As}}_{\mathrm{Ga}}$ antisite defects. The weak 0.5-eV peak is tentatively attributed to a (++/+) transition of the ${\mathrm{As}}_{\mathrm{Ga}}$ antisite defects. The 0.15- and 1.1-eV peaks that have densities of 1.3\ifmmode\times\else\texttimes\fi{}${10}^{12}$ and 0.8\ifmmode\times\else\texttimes\fi{}${10}^{12}$ ${\mathrm{cm}}^{\mathrm{\ensuremath{-}}2}$, respectively, are attributed to ${\mathrm{Ga}}_{\mathrm{As}}$ antisite defects and Ga vacancy defects, respectively. The interface Fermi level of GaAs is located at 0.85 eV above the VBM, indicating that it is strongly affected by the ${\mathrm{As}}_{\mathrm{Ga}}$ antisite defects. From the density of the interface states near the Fermi level, i.e., \ensuremath{\sim}1\ifmmode\times\else\texttimes\fi{}${10}^{13}$ ${\mathrm{cm}}^{\mathrm{\ensuremath{-}}2}$ ${\mathrm{eV}}^{\mathrm{\ensuremath{-}}1}$, it is shown that d\ensuremath{\varphi}/d${\mathrm{\ensuremath{\chi}}}_{\mathit{M}}$ (\ensuremath{\varphi}: barrier height in GaAs, ${\mathrm{\ensuremath{\chi}}}_{\mathit{M}}$: metal electronegativity) is 0.24, indicating that the Fermi level is pinned partly by the ${\mathrm{As}}_{\mathrm{Ga}}$ antisite defects and that fixed oxide positive charges with a density of (2--3)\ifmmode\times\else\texttimes\fi{}${10}^{12}$ ${\mathrm{cm}}^{\mathrm{\ensuremath{-}}2}$ are present at the GaAs/oxide interface.

Photoellipsometry determination of surface Fermi level in GaAs (100)

Photoellipsometry, a new contactless optical method, is presented. Related to photoreflectance, this method utilizes spectroscopic ellipsometry with the addition of a laser pump beam directed at nearly normal incidence onto the sample surface, where photoexcited electron-hole pairs modify the built-in electric field as a result of the photovoltaic effect. The field-induced changes in the real and imaginary parts of the complex pseudodielectric function can be directly measured, and analyzed in terms of the pump beam power density or the probe beam photon energy to determine parameters, such as surface Fermi level, built-in field strength, depletion depth, broadening, critical point energies, etc., of a semiconductor under study. Demonstration of this method was made with two n-type GaAs (100) substrates, where the measurements were taken in room air near the band-gap E0 region, for determination of surface Fermi level.

Time dependence of the surface Fermi level of GaAs in atmosphere

This letter reports the time dependence of the surface Fermi level of GaAs grown by molecular-beam epitaxy and then exposed to atmosphere. The sheet resistance of sample structures for field effect transistors alternately increased, decreased, increased, and decreased to become nearly constant after about 500 h. These changes correspond to the surface Fermi level varying between 0.3 and 0.7 eV and finally settling 0.7 eV above the valence band maximum. Comparison between annealed and unannealed samples with low-temperature-grown GaAs layers showed that the pinning of the surface Fermi level at 0.7 eV above the valence band maximum is caused by arsenic antisite defects. The result supports the advanced unified defect model.

Photoreflectance characterization of surface Fermi level in as-grown GaAs(100)

Photoreflectance (PR) spectroscopy has been used to study the surface Fermi level in GaAs by taking account of photovoltaic properties at the semiconductor surface. The PR signal amplitude ‖ΔR/R‖ is proportional to a modulating photovoltage Vm generated by a modulation light irradiation. To modify the surface potential, the sample was irradiated by a third perturbing light, i.e., a continuous bias light Pb, together with the modulation light. The modulation light power dependence of ‖ΔR/R‖ is extremely sensitive to bias light intensity, surface Fermi level, and temperature. From the analysis of the temperature dependence of ‖ΔR/R‖ on modulation-light power, the surface Fermi level of 0.47±0.09 eV below the conduction band was determined for a molecular beam epitaxially-grown GaAs(100).

Photoreflectance study of surface Fermi level in GaAs and GaAlAs

Franz–Keldysh oscillations from GaAs and AlGaAs structures have been studied and we find that the electric field obtained from the oscillations is in agreement with that derived from electrostatic calculations. Our results show that illumination from pump and probe beams in a normal photoreflectance experiment can significantly affect the measurement and thus erroneously lead to a reduced value of the electric field. The Fermi level on the bare surface of AlGaAs with different Al mole fraction has also been determined.

Changes in surface composition and Fermi level position during thermal desorption of ultraviolet radiation/ozone formed oxides on GaAs

X-ray photoelectron spectroscopy (XPS) has been used to study the native oxide structure and surface Fermi level of GaAs exposed to ultraviolet light (UV)/ozone, as well as to monitor changes that occur during thermal desorption of the surface oxide. The XPS results confirm that the surface oxide consists of a mixture of gallium and arsenic oxide phases which desorb at two different temperature ranges: desorption of arsenic oxide phases and oxygen transfer from arsenic to gallium occur at 250–500 °C; and desorption of gallium oxide phases occurs at 550–600 °C. In addition, we found that the surface Fermi level of GaAs below the UV/ozone formed oxide was always at about 0.8 eV from the valence- band maximum (VBM) for n-type substrates and 0.5 eV for p-type substrates. Similar results were found for wet chemically treated and then air exposed GaAs surfaces. However, the surface Fermi levels of both n- and p-type substrates, during the desorption of arsenic oxide in the temperature range of 250–500 °C, moved further away from VBM by as much as 0.4 eV. When the samples were heated to higher temperatures, the surface Fermi level moved back to midgap even before complete desorption of the gallium oxide phases. Fermi level pinning on samples before and after oxide desorption has also been characterized by in situ deposition of 5 nm of silicon nitride on the sample surface in the XPS system and by monitoring the surface potential of the semiconductor with XPS as a function of charging potential on the dielectric. It was found that surfaces both before and after arsenic oxide desorption were pinned within 0.1 eV with a negative voltage of about 2 V on the thin dielectric.

Photoreflectance characterization of semiconductors and semiconductor heterostructures

The electromodulation method of photoreflectance (PR) is becoming an important tool for the characterization of semiconductors, semiconductor interfaces and semiconductor microstructures such as superlattices, quantum wells, multiple quantum wells and heterojunctions. Since PR is contactless, requires no special mounting of the sample and can be performed in a variety of transparent ambients it can be utilized for in-situ monitoring of growth at elevated temperatures, in-situ elevation at 300K before the samples are removed from the growth/processing chamber as well as convenient ex-situ characterization. This invited article discusses some recent uses of PR to measure (a) the direct gap (and spin-orbit split component) of InP up to 600° C, (b) strains in Si at the Si/SiO2 interface, (c) changes in the surface Fermi level of GaAs caused by photowashing, (d) quantized intersubband transitions in a GaAs/Gao0.82Al0.18As multiple quantum well and (e) two-dimensional electron gas effects in selectively dopedn-Gao.7Alo.3As/ GaAs heterostructures.

Unpinning the GaAs Fermi level with thin heavily doped silicon overlayers

A theoretical investigation is made into the unpinning of the GaAs Fermi level (E/sub f/) at Schottky contacts by thin interfacial layers of heavily doped Si. The pinning mechanism is assumed to be a plane of interface states in the GaAs. The method used is to solve Poisson's equation numerically as a two-point boundary problem across the semiconductor. The results show that E/sub f/ can be moved from its pinned position to near the edge of the silicon valence band maximum or conduction band minimum at the Si/GaAs heterojunction with heavily doped p-Si or n-Si overlayers, respectively. This unpinning is observed with and without thick metallization on the Si. Exactly analogous results are obtained for Ge interfacial layers. The results are in good agreement with the experimental results obtained by J.R. Waldrop and R.W. Grant (1988). Although the unpinning is seen to occur for interface-state densities sufficient to pin E/sub f/ at the free GaAs surface, interface-state densities high enough to result in significant E/sub f/ pinning at metal/GaAs contacts are seen to prevent such unpinning by Si interfacial layers. It is therefore suggested that Si (or Ge) deposition gives rise to fewer interface states in the GaAs than does metal deposition.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Increased dependence of Schottky barrier height on metal work functions due to a thin-oxide layer

Ultraviolet illumination has been used to form a few monolayers of GaAs oxide in dry oxygen on both p-type and n-type GaAs(100) surfaces. Some of the oxides have been annealed to form Ga-rich oxide surfaces. Six metals with a wide range of work functions and different reactivity with the oxide layer have been used to form Schottky contacts on clean GaAs and on the thin-oxide layers. Large variations from the Schottky barrier height on clean GaAs have been observed for both doping types due to the photo-oxidation of the GaAs surfaces. The trend in Schottky barrier change is toward the ideal Schottky limit where the metal work function determines the barrier height. The GaAs Fermi level is shown to move in a much wider region of the band gap for the oxidized surfaces than for clean surfaces, however, it is still confined to a midgap region.

Oxygen stabilization of molecular beam epitaxial Al-GaAs Schottky barrier heights

Aluminum Schottky barrier heights (φn) to GaAs depend heavily upon the deposition conditions and the gallium arsenide surface stoichiometry. After annealing to approximately 320 °C diodes show φb approximately 0.69 eV in association with oxygen at the Al/GaAs interface. We infer that these ’’stabilized’’ barrier heights are associated with oxygen-induced surface-state pinning of the GaAs Fermi level. Al deposition on arsenic stabilized surfaces show variable barrier heights and no oxygen peak indicating variable Al induced states in intermediate AlGaAs layers between the Al and GaAs.