Kazuyoshi Tsukamoto

Deposition and extensive light soaking of highly pure hydrogenated amorphous silicon

We have developed an ultrahigh vacuum plasma‐enhanced chemical‐vapor deposition system, and deposited high‐purity device‐quality hydrogenated amorphous silicon films. High sensitivity secondary ion mass spectrometry measurements show that impurity contents in the bulk of the present films are reduced to 2×1015 cm−3 for O, 7–10×1015 cm−3 for C, and 5×1014 cm−3 for N; these impurities are normally present at fairly high levels. Nevertheless, extensive light soaking of the films resulted in a defect density as high as 5×1017 cm−3, which is well above the impurity content. This result excludes those models of photoinduced degradation that postulate one‐to‐one correlation between light‐induced defects and O, C, or N impurity atoms.

Hydrogen Concentration and Bond Configurations in Silicon Nitride Films Prepared by ECR Plasma CVD Method

The influence of deposition conditions on the hydrogen concentration and bond configurations in silicon nitride (SiN) films prepared by the electron cyclotron resonance plasma CVD(ECR P-CVD) method has been studied in comparison with those prepared by a conventional plasma CVD(P-CVD) method, by means of secondary ion mass spectrometry (SIMS), infrared absorption (IR) spectra and electron spin resonance (ESR). SiN films deposited at low microwave powers and a relatively high SiH4 flow rate had higher H concentrations than those deposited at high microwave powers. The hydrogen concentration was found to be in the range from 1.3×1022-5×1021/cm3 (13.8-4.7%) at microwave powers between 50 and 500 W. The bond configurations in the films are markedly dependent upon the deposition conditions. SiN films with only N-H bonds are thermally stable, even after annealing at 900°C in a dry N2 atmosphere, while almost all the H atoms are lost after annealing for films with both N-N and Si-H bonds. Annealing studies by ESR measurements revealed that the H concentration in the films has little influence on the spin density.

The presence of a SO molecule in [NiFe] hydrogenase from Desulfovibrio vulgaris Miyazaki as detected by mass spectrometry.

The active site of [NiFe] hydrogenase is a binuclear metal complex composed of Fe and Ni atoms and is called the Ni-Fe site, where the Fe atom is known to be coordinated to three diatomic ligands. Two mass spectrometric techniques, pyrolysis-MS (pyrolysis-mass spectrometry) and TOF-SIMS (time-of-flight secondary ion mass spectrometry), were applied to several proteins, including native and denatured forms of [NiFe] hydrogenase from Desulfovibrio vulgaris Miyazaki F, [Fe4S4]2-ferredoxin from Clostridium pasteurianum, [Fe,S2]-ferredoxin from Spirulna platensis, and porcine pepsin. Pyrolysis-MS revealed that only native hydrogenase liberated SO/SO2 (ions of m/z 48 and 64 at an equilibrium ratio of SO and SO2) at relatively low temperatures before the covalent bonds in the polypeptide moiety started to decompose. TOF-SIMS indicated that native Miyazaki hydrogenase released SO/SO2 (m/z 47.97 and 63.96) as secondary ions when irradiated with a high-energy Ga+ beam. Denatured hydrogenase, clostridial ferredoxin, and pepsin did not release SO as a secondary ion. The FT-IR spectrum of the enzyme suggested the presence of CO and CN. These lines of evidence suggest that the three diatomic ligands coordinated to the Fe atom at the Ni-Fe site in Miyazaki hydrogenase are SO, CO, and CN. The role of the SO ligand in helping to cleave H2 molecules at the active site and stabilizing the Fe atom in the diamagnetic Fe(II) state in the redox cycle of this enzyme is discussed.

Secondary ion yields of C, Si, and Ge in InP, and Cs surface density and concentration, in SIMS

Abstract In secondary ion mass spectrometry (SIMS), the relationship between negative secondary ion yield ( γ − ) of C, Si, and Ge in an InP substrate and sputtering yield ( Y ), and that between the secondary ion yield and the surface density of reactive primary ion Cs ( n Cs (0); zero refers to substrate surface) were examined by changing the incident angle of the primary ion to the surface. The γ − value was derived by dividing the corresponding useful yield by the instrumental transmission factor, which was obtained by measuring the useful yield of an element with high electron affinity, such as fluorine. Monte Carlo simulations using TRIM code and theoretical analysis were performed to obtain the Cs density (atoms/unit volume) depth distribution, n Cs ( x ), in the sputtered substrate. The density depth distribution was measured directly using SIMS, under conditions where the energy of the O 2 + primary ion was sufficiently reduced. n Cs (0) was obtained by setting x of n Cs ( x ) to zero. The Cs ions used for sputtering were distributed mostly according to a Gaussian function. The n Cs ( x ) profile estimated on the basis of the Gaussian distribution showed good accordance with the measured one. At the sputtered surface, n Cs (0) was closely proportional to 1/ Y . The values of γ − for C, Si and Ge changed proportionally to the power of n Cs (0), or 1/ Y . For C, Si, and Ge, the lower the γ − , the more sensitively γ − changed in response to n Cs (0). Furthermore, the composition at the Cs-sputtered surface was examined using Auger electron and wavelength dispersive X-ray spectrometry. These measurements showed that n In (0), n P (0) and n Cs (0) changed almost linearly with 1/ Y InP , suggesting that the Cs surface concentration (at.%), C Cs (0), is proportional not to 1/ Y itself, but to a function of 1/ Y . However, under certain special conditions, C Cs (0) is also proportional to 1/ Y itself, as is n Cs (0).

Structure and electrical properties of interfaces between silicon films and n+ silicon crystals

We have studied the morphology and electrical characteristics of the interfaces between silicon‐implanted, low‐pressure chemical‐vapor‐deposited (LPCVD) silicon films with an n+ single‐crystal silicon substrate. Using high‐resolution transmission electron microscopy, it is shown that the silicon implant causes a ‘‘balling up’’ of the native oxide layer at the interface and epitaxial growth occurs in the LPCVD silicon film even after rapid thermal annealing at only 940  °C for 30 s. This morphological change results in a realization of a low‐ohmic‐resistivity LPCVD silicon/ n+ single‐crystal silicon contact even at a sub‐half‐μm size, although the unimplanted contact becomes nonohmic. The leakage current for the implanted contact is as low as that for the unimplanted one in shallow junctions.

Doping of Trench Side-Walls Using an Arsenic Planar-Type Solid-Diffusion Source (S-D Source) and Analysis of Doping Uniformity by Secondary Ion Mass Spectroscopy (SIMS)

We developed a new method of doping side-walls of sub-micron-width trenches using an arsenic planar-type solid-diffusion source at a low pressure. The sheet resistivity, which was measured on the wafer surfaces, is found to be controlled by the oxygen concentration in the tube. Arsenic was doped uniformly in the 6-inch diameter wafer with a standard deviation of 2.7%. The doping was reproducible; the standard deviation of the doping for many batches was 3.0%. A new method of evaluating uniformity of trench side-wall doping using SIMS was also developed. The doping uniformity is confirmed to be within the accuracy of the SIMS measurement.

SIMS Applications for Electronic Materials.

Current problems and applications of SIMS analysis for electronic materials, especially for semiconductors are discussed from the viewpoint of in-depth profile, micro area and trace analysis and quantification. Static-SIMS and SNMS which are presently being developed as new analytical techniques are also reviewed and those applications for semiconductor materials are shown.

A 2 kW Photovoltaic Power Generating System Using a-Si Solar Cells

We have developed integrated amorphous silicon (a-Si) solar panels and have applied them to an electric power generating system. The integrated a-Si solar panel consists of 20 integrated type a-Si solar cell modules, each with a size of 10 cm ×10 cm, which have been developed for use in an electric power generating system. The power generating system is set in an experimental model house. The output performance of the solar panels has been measured, and a peak amount of 2 kW was obtained. The conversion efficiency decreased about 10% within one month and then became stable. This stability has continued until the present time. The initial degradation of the conversion efficiency mainly depends on the degradation of the fill factor. It has been confirmed that electric power generating systems using a-Si solar panels are already effective enough for practical use.