Barrier Height Measurements Research Articles

Photoemission spectroscopy of GaAs:As photodiodes

We report on photoemission measurements of molecular-beam-epitaxy-grown GaAs p-i-n structures, in which the optically active insulating GaAs layer contains As precipitates (GaAs:As). GaAs:As is formed by low-temperature growth of GaAs at 225 °C, followed by an anneal at 600 °C. Layers grown in this way have been reported to be sensitive to subband-gap light. The measured barrier height of 0.7 eV, extracted from a well-behaved Fowler plot, indicates that the mechanism for photodetection involves arsenic clusters embedded in GaAs acting as internal Schottky barriers.

The interaction of platinum with GaP(110): band bending and surface photovoltage effects

The study of metals deposited on semiconductor surfaces is important to determine the chemical and electronic properties of these interfaces. In the present study, the formation of Pt GaP(110) interfaces at 150 and 300 K was monitored by core and valence level photoelectron spectroscopy using synchrotron radiation. Detailed analysis of band-bending reveals a strong surface photovoltage (SPV) effect, even at the higher temperature. The measured barrier height (φ bn = 1.5eV) is similar to values for chemically very different metal contacts on n-GaP(110). The temperature, metal coverage and photon flux dependence of the SPV is determined and interpreted in terms of calculations which consider the relevant current transport mechanism across the developing interface.

Imaging of Metal/Semiconductor Interface by Ballistic-Electron-Emission Microscopy (Beem)

AbstractStudies in ballistic-electron-emission spectroscopy (BEES) have enabled precise energy measurements of Schottky barrier heights with excellent spatial resolution and, more recently, it was shown that even scattering at the metal/semiconductor interface affects the BEES spectrum [1]. Monte Carlo simulations have been done to predict the spatial resolution of ballistic-electron-emission microscopy (BEEM) [2]. In this paper, we will discuss the experimental spatial resolution of BEEM, and we will also give some of our BEES results for Au/Si and for Au/PtSi/Si. Our experimental BEEM studies indicate that, for Au/Si, hot electron transport is diffusive rather than ballistic, because the inelastic mean free path length (∼100 nm) is much larger than the elastic mean free path length (∼10 nm). This is in agreement with existing theories and with the literature on the internal photoemission method of studying the transport. Even in this diffusive regime, the spatial resolution of BEEM is still expected to be very good, being on the order of 10 nm [2]. Our preliminary work on PtSi shows that it has an attenuation length of 4 nm, which differs significantly from that of Au.

Photoreflectance study of the surface Fermi level at (001) n- and p-type GaAs surfaces

We report a photoreflectance study of Fermi level pinning (VF) on (001) n- and p-type GaAs with large, uniform electric fields. Surface photovoltage (Vs) effects were evaluated as a function of temperature (77 K<T<450 K), pump beam wavelength (633 and 407 nm) and W–metal coverage (in situ). The dependence of the measured barrier height, Vb (=VF−Vs), on T can be explained by a modification of the theory of M. Hecht [Phys. Rev. B 41, 7918 (1990)] yielding values of VF =0.77±0.02 V and VF =0.75±0.02 V for n- and p-type GaAs, respectively, at 300 K. In addition, by introducing the ratio (r) of the area of the surface states to the illuminated area into the theory of Hecht we have been able to estimate the density of surface states on the GaAs surface. The effect of metal coverage is to increase r and reduce the influence of Vs.

Thermionic emission of Γ and L electrons in the GaSb/InAs hot-electron transistors

We have fabricated GaSb/InAs hot-electron transistors with base widths of 30 to 200 Å. The barrier height of thermionic emission was evaluated with two techniques; deducing from values of collector current density itself, and from its temperature dependence. The difference between the two results reveals that electrons are emitted from both Γ and L valleys of the GaSb conduction band. From the measured barrier height, the energy of the ground quantum level formed in InAs base was also deduced. Its dependence on the base width was compared with simple calculation where nonparabolicity of the InAs conduction band was considered. It indicates electron density in the InAs base is enhanced due to electron’s transfer from GaSb to InAs.

Au and Al Schottky barrier formation on GaAs (100) surfaces prepared by thermal desorption of a protective arsenic coating

Soft x-ray photoemission spectroscopy has been used to investigate the initial stages of Schottky barrier formation on GaAs (100) surfaces prepared by the thermal desorption of an As cap. This work was motivated by a previous study [Brillson et al., J. Vac. Sci. Technol. B 6, 1263 (1988)] of identically grown and capped samples which reported ‘‘unpinned’’ Schottky barrier formation, with barrier heights falling over a wide range (0.75 eV) of energies. This large energy range is a striking result, as a considerable number of prior studies on both (110) and (100) surfaces have found that all metals will pin in a narrow (0.25 eV) range near midgap. Since Au and Al are the extremes of the larger 0.75 eV span of Schottky barriers, we have studied the deposition of these two metals. We found that the barrier height measurements on the low doped n-type samples used in this work and in the paper referenced above are affected by photovoltaic effects, even at room temperature. These photovoltaic effects cause shifts in the band bending, which are an artifact of the measurement. We also performed measurements on more heavily doped samples, and the photovoltaic effects were removed. In addition, we point out that Au–Ga alloying makes the case of Au potentially misleading. With the photovoltaic effects removed, and the Au–Ga alloying carefully accounted for, we found that the barriers heights for Au and Al differ by only 0.25 eV.

Tip-surface forces during imaging by scanning tunneling microscopy

The effect of compressive and shear forces between tip and surface during the operation of the scanning tunneling microscope (STM) is illustrated with examples obtained both in air and vacuum environments. We show that at typical gap resistances used in STM (≤20 GΩ) these forces can have significant effects. Compressive or repulsive forces give rise to anomalous topographic corrugations (elastic deformations) as well as to permanent damage (inelastic or plastic deformation). These forces also cause the anomalously low values obtained in measurements of the tunneling barrier height. The effects of shear forces when imaging weakly bound material will also be demonstrated.

Role of atomic force in tunneling-barrier measurements

Experimental measurements of the apparent barrier height as a function of tip-sample separation using a scanning tunneling microscope (with clean W tips and clean Si surfaces in ultra high-vacuum) show that the barrier height starts at 3.5 eV at large separations, increases to 4.8 eV at about 1.5 Å before the mechanical contact, and then drops to below 0.3 eV within a fraction of an ångström. At the distances encountered in scanning tunneling microscopy, forces between sample and tip can be significant. Using a simple model of this system including tip-sample forces leads to a calculated apparent barrier height which quantitatively reproduces the observed variation in apparent barrier height over the entire range of tip-sample separations.

Plasma etching of III–V semiconductors in CH4/H2/Ar electron cyclotron resonance discharges

We have investigated the etch rates, residual lattice damage, surface morphologies, and chemistries of InP, InGaAs, AlInAs, and GaAs plasma etched in electron cyclotron resonance (ECR) CH4/H2/Ar discharges. The etch rates of InP and InGaAs increase linearly with additional rf biasing of the substrate, and are approximately a factor of 2 faster than for GaAs. Under our conditions the etch rate of Al0.52Ga0.48As is very low (∼25 Å min−1) even for the addition of 100 V rf bias. In all of these materials the residual damage layer remaining after dry etching is very shallow (∼20 Å) as evidenced from Schottky barrier height and photoluminescence measurements combined with wet chemical etching. InP shows significant P depletion with the addition of rf biasing during the ECR etching while GaAs retains a near-stoichiometric surface. Hydrogen passivation of shallow donors in n-type GaAs occurs to a depth of ∼3000 Å during exposure to the CH4/H2/Ar discharge for long periods (60 min). The surface morphologies in the In-based materials become roughened for etching with the addition of rf biasing while GaAs displays smooth, residue-free surfaces under these conditions.

Growth of CoSi2 on Si(001): Structure, defects, and resistivity

Epitaxial films of CoSi2 on Si(001) have been grown by molecular-beam epitaxy, and their electrical and structural properties studied by resistivity and Schottky barrier height measurements, by Rutherford backscattering spectroscopy (RBS), and by transmission electron microscopy (TEM). Most growth conditions successful on Si (111) have been found to result in misoriented grains when applied to Si(001). However, single-orientation (001) CoSi2 films, with low resistivities (16 μΩ cm) and low ion channeling minimum yields (χmin=5%) have been obtained by direct codeposition of Co and Si at ∼500 °C at Co-rich stoichiometries. Single orientation (001) CoSi2 films have also been obtained by using a template method on epitaxially grown Si buffer layers, but the resistivities of these films have not been as good. A Schottky barrier height of 0.71 eV has been measured for single-orientation CoSi2(001)/Si(001). This is significantly higher than the barrier height of 0.64 eV for CoSi2(111)/Si(111).

Ballistic-electron-emission microscopy of subsurface defects at the Au-GaAs(100) interface

Semiconductor interface electronic properties are strongly influenced by the presence of defects. Throughout the long history of interface investigation, the study of interface defect electronic properties has been seriously hindered by the fundamental experimental difficulty of probing subsurface structures. A new method, ballistic-electron-emission microscopy (BEEM), has been developed which not only enables probing of subsurface interface properties, but also provides nanometer-resolution imaging capabilities. BEEM employs scanning tunneling microscopy (STM) and a unique spatially localized ballistic electron spectroscopy method. The imaging of subsurface interface structure and spectroscopic measurement of local Schottky barrier height have been demonstrated by BEEM. Recent BEEM results include the first direct observation of a multiphase defect structure at the important Au-GaAs(100) Schottky barrier interface. This paper describes the application of BEEM to a study of the influence of GaAs(100) substrate conditions on the formation of this defect structure at the Au-GaAs(100) interface.

Plasma Etching of InP and Related Materials in Electron Cyclotron Resonance CH4/H2/Ar Discharges

We have investigated the etch rates, residual lattice damage, surface morphologies and chemistries of InP, InGaAs, AlInAs and GaAs plasma etched in electron cyclotron resonance (ECR) CH4/H2/Ar discharges. The etch rates of InP and InGaAs increase linearly with additional RF biasing of the substrate, and are approximately a factor of two faster than for GaAs. Under our conditions the etch rate of Al0.52Ga0.48As is very low (≈25γ · min−1) even for the addition of 100 V RF bias. In all of these materials the residual damage layer remaining after dry etching is very shallow (≈20γ) as evidenced from Schottky barrier height and photoluminescence measurements combined with wet chemical etching. InP shows significant P depletion with the addition of RF biasing during the ECR etching while GaAs retains a near-stoichiometric surface. Hydrogen passivation of shallow donors in n-type GaAs occurs to a depth of ≈3000γ during exposure to the CH4/H2/Ar discharge for long periods (60 mins). The surface morphologies of the In-based materials become roughened for etching with the addition of RF biasing while GaAs displays smooth, residue-free surfaces under these conditions.

Interface states at metal-compound semiconductor junctions

Direct observations of electronic states at metal-semiconductor junctions provide new information for understanding Schottky barrier formation. Low energy cathodoluminescence spectroscopy (CLS) offers a sensitive probe of near-surface and ‘buried’ interface states. These states are discrete and evolve with metal overlayer coverage to energies corresponding to Fermi level ( E F) positions of the Schottky contact. CLS and photoluminescence spectroscopy (PL) also reveal that deep levels resident in the semiconductor bulk can dominate E F movements. By controlling atomic-scale interface chemistry, one can affect deep level production near the junction and thereby alter E F stabilization energies. Soft X-ray photoemission spectroscopy (SXPS) and internal photoemission measurements of barrier heights confirm this macroscopic effect. Examples include metals on clean, ordered GaAs, InP, CdTe, and InAlAs-InP heterojunctions. The SXPS-derived GaAs(100) barrier heights vs metal work function yield a self-consistent electrostatic analysis in terms of interface states actually observed by CLS. For GaAs (100) surfaces at low temperatures, such states are minimized and classic Schottky barrier formation is evident. These results highlight the extrinsic nature of metal-semiconductor interface states and the role of interface chemistry and bulk crystal quality in achieving near-ideal electronic properties.

Damage in silicon caused by magnetron ion etching and its recovery effect

Damage in silicon exposed to an MIE (magnetron ion etching) plasma has been investigated. The damage was characterized by Schottky barrier height measurements using Al/n-Si diodes. The depths of the damaged layer are determined as a function of RF power. It is found that the damaged layer at an RF power of 2 kW (self-bias: 270 V) is about 12 nm, and that the damage depths correlate with the self-bias voltage, that is, the energy of ions impinging on the Si surface during plasma exposure. Several methods for removal of the damaged layer were examined. It was found that the damaged layer can be removed by a wet Si etching. >

Determination of the local electronic structure of atomic-sized defects on Si(001) by tunneling spectroscopy

Tunneling spectroscopy and voltage-dependent scanning tunneling microscopy have been used to study the geometry and electronic properties of atomic-sized defects on the Si(001) surface. Individual dimer vacancies are shown to be semiconducting, consistent with the π-bonded defect model of Pandey. Another type of characteristic defect is found which gives rise to strongly metallic tunneling I–V characteristics, demonstrating that it has a high density of states at the Fermi level and is likely active in Fermi level pinning on Si(001). Spatially dependent I–V measurements and tunneling barrier height measurements also directly reveal the spatial extent of this metallic character and provide direct measures of the ‘‘size’’ of the defects.

P-type PtSi Schottky-diode barrier height determined from I– V measurement

We have shown that the method of analyzing the fundamental barrier height (flat-band barrier height at zero field) for n-type platinum-silicide Schottky barrier diodes can be extended to p-type Schottky barrier diodes. A mean flat-band barrier height of 0.272 ± 0.005 eV is obtained for p-type Schottky barrier diodes by relating the measured barrier heights and their corresponding ideality factors from I- V measurements. Nine different combinations of processing conditions were used in the fabrication of these devices so that the effects of different processes on the device characteristics could be evaluated.

Si/Ti/TiB 2/Al structures investigated as contacts in microelectronic devices

The metallization system Ti/TiB 2/Al to silicon has been investigated. TiB 2 was reactively sputter deposited in an Ar/B 2H 6 mixture. With this technique it is possible to study the stability of both Ti/TiB 2 and TiB 2 contacts to both n- and p-type silicon. In general Ti/TiB 2 contacts showed significantly smaller contact resistances than TiB 2 contacts. Upon annealing we found that current crowding at the periphery became predominant. This current crowding effect was more pronounced on n-type silicon. Diode structures of Ti/TiB 2/Al annealed up to 500°C exhibited no junction leakage. The TiB 2/Al contact remained intact up to 525°C. Barrier height measurements on Schottky contacts annealed at different temperatures was also performed. The barrier height value of 0.6 eV, close to that reported for TiSi 2, was recorded after annealing at 525°C. The usefulness of these contact structures in microelectronic metallization was discussed. The difference between the two metallization systems was also investigated by Rutherford Backscattering Spectroscopy and X-Ray Diffraction measurements.

Control of the Schottky barrier using an ultrathin interface metal layer

Fermi level movements at Pt/GaAs and Ti/GaAs interfaces have been investigated using a direct measurement of Schottky barrier heights in a bimetal Schottky structure. Using thin interfacial layers, the Schottky barrier was smoothly varied from the characteristic value of the thick metal to that of the interfacial metal. The variation of barrier height versus the inner metal thickness was found to exhibit an exponential behavior extending over a few monolayers coverage. This experiment indicates a new approach to the fundamental study of metal-semiconductor interfaces and could be useful in device applications.

Characterization of TiSi2 Ohmic and Schottky Contacts Formed by Rapid Thermal Annealing Technology

The new trend for formation is rapid thermal annealing of sputtered titanium on silicon. Ohmic contacts of to n+ and p+ silicon and Schottky diodes to n‐type silicon were physically and electrically analyzed in order to characterize the RTA process. A four‐terminal Kelvin resistor structure has been used to accurately measure the contact resistance. The redistribution of arsenic and boron‐implanted atoms was monitored during silicidation. Sheet resistance measurements, secondary ion mass spectroscopy, and Rutherford backscattering were all performed to characterize the silicide‐silicon interface. All the results were discussed in comparison with to n+ to p+ contact characteristics. contact resistivity to n+ silicon is found to be equal to contact resistivity to n+ silicon. This result, in contradiction with the theory, can be explained by the difference in As redistribution for each silicide. to p+ contact resistivity is higher than that of to p+ silicon, but this result can be improved by increasing the p+ surface concentration. The measured barrier height of the Schottky diodes is 0.52V, and the diodes show good ideality and low dispersion.

Investigation of mercury adsorption on gold films by STM

SUMMARYScanning tunnelling microscopy, combined with complementary electrical and analytical measurements, provides a powerful method for examining the behaviour of mercury on gold‐film sensors under actual sensor operating conditions. The films exhibit a linear increase in resistance upon adsorption of mercury, and this resistive change is accompanied by a decrease in measured barrier height at submonolayer coverages of mercury. STM studies of changes in the effective barrier height upon mercury adsorption provide evidence that mercury migrates to grain boundaries for films exposed at submonolayer coverages. Coverages in excess of a monolayer result in a nearly constant, reduced effective barrier height across the entire surface, which is indicative of a more uniform distribution of mercury over the surface. This behaviour is interpreted in terms of the adsorption of mercury onto defective, contaminated gold surfaces.