Tomohiro Kubota

Studies on interface states at ultrathin SiO2/Si(100) interfaces by means of x-ray photoelectron spectroscopy under biases and their passivation by cyanide treatment

The energy distribution of interface states at ultrathin oxide/Si(100) interfaces is obtained using a new method, i.e., x-ray photoelectron spectroscopy measurements under biases between the metal overlayer and the Si substrate of the metal-oxide-semiconductor (MOS) devices. Ultrathin thermal oxide layers formed at 450u2009°C in oxygen have an interface state peak near the midgap and it is attributed to isolated Si dangling bonds with which no atoms in the oxide layer interact. On the other hand, thermal oxide layers formed at 650u2009°C have a two-peaked structure, one peak above and the other below the midgap, and they are attributed to Si dangling bonds with which an oxygen or Si atom in the oxide layer interacts weakly. The density of the interface states, especially that near the midgap, decreases drastically by cyanide treatment, i.e., the immersion of Si in a KCN solution for a few seconds followed by a rinse in boiling water, performed before the oxide formation. It is suggested that cyanide ions penetrat...

OXIDE THICKNESS DEPENDENCE OF ENERGY SHIFTS IN THE SI 2P LEVELS FOR THE SIO2/SI STRUCTURE, AND ITS ELIMINATION BY A PALLADIUM OVERLAYER

The energy difference between the oxide and substrate Si 2p peaks for silicon oxide/Si structures increases with the oxide thickness. The dependence of the energy shift on the oxide thickness almost disappears with the deposition of a thin palladium overlayer, because of the avoidance of the surface charging effect due to photoemission and because of the nearly constant energy shift resulting from extra atomic relaxation. The true chemical shift of silicon oxide layers thicker than 2 nm is determined to be ∼3.8u2009eV. For the thickness dependence of the oxide Si 2p energy, the extra atomic relaxation and charging effect are dominant for oxide layers thinner than ∼2u2009nm and thicker than ∼4u2009nm, respectively. In the intermediate thickness region, both the effects are important.

Theoretical and spectroscopic studies of gap-states at ultrathin silicon oxide/silicon interfaces

The energy distribution of interface states in the Si forbidden gap at ultrathin thermal oxide/Si(111) interfaces is obtained from x-ray photoelectron spectroscopy measurements under bias. All the observed interface state spectra have peaked structure, indicating that they are due to Si dangling bonds. For thermal oxide layers formed at 350u200a°C, only one interface state peak is present near the midgap. The interface state peak has ∼0.1 eV width, showing that the effective correlation energy of the Si dangling bond interface state is less than ∼0.1 eV. For oxide layers produced above 550u200a°C, on the other hand, two peaks are observed, one above and the other below the midgap. It is found using a density functional theory method by employing clusters containing 27 bulk-like Si atoms (interior atoms, without H passivation) that an isolated Si dangling bond, with which no atoms in the oxide layer interact, has an energy level near the midgap. It is also found from the calculations that weak interaction of the S...

Deoxidation of gallium arsenide surface via silicon overlayer : A study on the evolution of the interface state density

The GaAs surface with the native oxide formed by wet etching has been gradually deoxidized via evaporation of a silicon overlayer. Both chemical and electronic properties of such structures have been examined by x-ray photoelectron spectroscopy (XPS) and “XPS under biases,” respectively. The latter technique enables a direct assessment of the interface state density of insulator∕semiconductor interfaces. We have concluded that gap states incident to the native oxide∕GaAs interface have annihilated due to replacement of Ga–O bonds by Ga–Si and As–Si bonds.

Effects of an ultrathin silicon oxynitride buffer layer on electrical properties of ferroelectric Bi4Ti3O12 thin films on p-Si(100) surfaces

Electrical properties of ferroelectric Bi4Ti3O12u2002(BiT) films on Si(100) using a 1 nm thick silicon oxynitride (SiON) buffer were investigated. The capacitance–voltage (C–V) characteristics of Au/BiT/SiON/Si(100) exhibited hysteresis loops with a memory window of 2 V due to the ferroelectricity, and did not show large carrier injections. The effects of the SiON buffer were demonstrated in current–voltage characteristics. In the reverse bias region, a leakage current density of the specimen without the SiON buffer was much larger than that of the specimen with the buffer. Apart from these electrical measurements, anomalous features appeared in C–V characteristics of the illuminated specimen, which were likely to be due to the ac response of the optically generated electrons in some trap states at the interface.

Elimination of interface states in the GaAs band-gap by cyanide treatment: XPS measurements under bias

Abstract Energy distribution of interface states for GaAs-based metal-oxide-semiconductor structure with an ultra-thin silicon oxide layer is obtained from “XPS measurements under bias.” The interface state spectra have peaked-structure at 0.7 and 0.9 eV above the valence band maximum and they are attributed to (++/+) and (+/0) transitions of As Ga antisite defects at the interface. When cyanide treatment (i.e., the immersion in a KCN aqueous solution followed by the rinse in boiling water) is performed after the deposition of the silicon oxide layer, the interface state density is decreased to ∼50%, resulting in the partial unpinning of the Fermi level.

Dependence of interface states for ultra-thin SiO2/Si interfaces on the oxide atomic density determined from FTIR measurements

Abstract The atomic density of ultra-thin thermal silicon oxide layers formed on Si(111) and (100) substrate is determined from Fourier transform infrared (FTIR) measurements, which well explains a dependence of the energy distribution of interface states obtained from X-ray photon spectroscopy measurements under bias on the oxidation temperature. In the case of the thermal oxide layers formed below 450°C, observation of the LO and TO phonons shows that the oxide atomic density is low and an interface state peak is present near the midgap, where the peak is attributable to isolated Si dangling bonds at the interface. With an increase in the oxide formation temperature, the atomic density of the oxide layer increases and interface state peaks are observed above and below the midgap, which peaks are attributable to Si dangling bonds interacting weakly with a Si or oxygen atom in the oxide layer. The energy separation between the interface state peaks for the oxide/Si(111) interfaces is much smaller than that for the oxide/Si(100) interfaces. The small energy separation arises from the long distance between a Si dangling bond and the interacting atom in the oxide layer, originated from the interfacial structure.

Mechanism of open circuit photovoltages for silicon | methanol junction solar cells

Photoelectrochemical cells (PECs) are made by immersing semiconductor electrodes in a redox solution. Therefore, a high energy barrier can be formed simply by employing a redox couple with a sufficiently positive (or negative) redox potential for n-type (or p-type) semiconductors, leading to the generation of high photovoltages. We obtained a high photovoltage of 695 mV from the highly-doped Si electrodes in a methanol solution [1]. In this case, the majority carrier dark current was suppressed by the formation of a high energy barrier due to the adsorption of methanol and by the low concentration of oxidized species [2]. The minority carrier dark current also became low because of the high donor density. Rosenbluth and Lewis [3] obtained a high photovoltage of 670 mV from a PEC of similar structure. We showed that both the photovoltage and the photocurrent density could be made high by employing an acetone redox solution containing a small amount of methanol [4]. Lewis and coworkers [5,6] observed the dependence of the open circuit photovoltage Voc for Si electrodes in methanol on the temperature T, and obtained almost linear Voc-T plots in the temperature range investigated (197294 K [5], 218-293 K [6]). However, we obtained Voc-T plots with slopes decreasing towards the lower temperature, and the plots were well explained by the dark current consisting of the minority carrier diffusion current and the majority carrier thermionic emission current [7]. In contrast, Kumar and Lewis [6] have claimed that nonlinear Voc-T plots could result from uncontrolled experimental conditions, such as poor stirring of the redox solution, and

A new spectroscopic method for determination of energy distribution of interface states in the semiconductor band-gap

Abstract Measurements of X-ray photoelectron spectra (XPS) are performed for (3 nm-Pt/2.6 nm-oxide/p-InP(100)) and (2.5 nm-Pt/3.8 nm-oxide/n-GaAs(100)) MOS devices under biases. Upon applying a negative bias voltage to the InP (or GaAs) substrate with respect to the Pt layer, the substrate In 3d 5 2 (or As 3d 5 2 ) peak is shifted toward the lower binding energy, while it is shifted toward the higher binding energy by applying a positive bias. These shifts are caused by a change in the potential drop across the oxide layer due to accumulation of charges in interface states. The energy distribution of the interface states in the band-gap is obtained by analyzing the magnitudes of these shifts measured as a function of the bias voltage. The energy distribution has peaked-structure and the peaks are attributed to antisite and vacancy defects. The interface Fermi level of InP (or GaAs) is located near the energy level of the (+/0) transition of the PIn (or AsGa) antisite defects.