Table of contents

See pdf for details

We report on the development of optical fibres which guide light in a solid core using a photonic bandgap effect. The photonic bandgap cladding consists of a two-dimensional array of isolated high-index regions in a lower-index matrix, with a relatively low index contrast. The core is one or more unit cells of the matrix material without the inclusions. The frequency bands for photonic bandgap guidance can be predicted by considering the cut-off frequencies of the guided modes of the high-index rods in the cladding using the weakly-guiding approximation. We demonstrate the basic properties of such fibres and their use as a wavelength-selective element in a fibre laser cavity.

We experimentally and theoretically demonstrate the negative refraction and focusing of electromagnetic (EM) waves by two-dimensional photonic crystal slabs at microwave frequencies. The negative refraction is observed both for transverse magnetic (TM) and transverse electric (TE) polarized incident EM waves. Gaussian beam shifting method is used to verify the negative refractive index. The Subwavelength imaging and flat lens behavior of photonic crystals are succesfully demonstrated. We have been able to overcome the diffraction limit and focus the EM waves to a spot size of 0.21λ. Metallodielectric photonic crystals are employed to increase the range of angle of incidence that results in negative refraction. Experimental results and theoretical calculations are in good agreement throughout the work.

Nanophotonic waveguides and components are promising for use in the large-scale integration of photonic circuits. Coupling light between nanophotonic waveguides and a single-mode fiber is an important problem and many different solutions have been proposed and demonstrated in recent years. In this paper, we discuss a grating coupler approach. Grating couplers can be placed anywhere on a circuit and can easily be integrated. We have experimentally demonstrated >30% coupling efficiency with a 1 dB bandwidth of 40 nm on standard wafers. Theoretically, the coupling efficiency can be improved to >90% using an optimized grating design and layer stack. The fabrication of the couplers in silicon-on-insulator and in indium phosphide membranes is also discussed.

In-plane channel-drop filters consisting of point and line defects in a two-dimensional photonic-crystal slab are investigated. A clear channel-drop operation was successfully demonstrated by employing an ultrahigh quality factor nanocavity and a suitably designed waveguide bend. Moreover, improvements in the channel-drop efficiency of the device were made using a feature of heterostructure photonic crystals, as proposed previously. It is shown theoretically that drop efficiencies much higher than the maximum of 25% for a conventional configuration are achievable using reflection at the heterostructure interface. Drop operations with efficiencies of almost 100% were experimentally demonstrated by the fabricated devices. Moreover, a four-channel drop operation with a high drop efficiency and a constant quality factor was demonstrated by in-plane channel-drop filters in a two-dimensional photonic-crystal slab, with a simple configuration based on heterostructure photonic crystals.

We fabricated a 12-fold symmetric quasiperiodic photonic crystal (QPC) into a GaInAsP slab, and obtained a lasing action in a single-defect nanocavity and a defect-free area at room temperature. Four defect modes observed by a lithographic tuning were clearly explained by finite-difference time-domain calculation. The lowest threshold was obtained for the quadrupole mode, which has a small modal volume of 0.58 times the cubic wavelength in the cavity, by the fine-tuning of adjacent airholes to the defect. The lasing in the defect-free area was also suggested to be due to the localization of two-dimensional Bragg modes at the photonic gap edge. Because of its high structural flexibility, the QPC nanocavity has the potential of becoming a good alternative to the PC nanocavity.

We have investigated cavity resonant excitation effects on InGaAs quantum dots (QDs) embedded in high-quality-factor photonic crystal nanocavities. The light emission of the lowest-order cavity mode at 1.06 µm is enhanced by more than a factor of ten, as compared with nonresonant excitation, when the excitation wavelength is resonant with higher order cavity modes. This result can be attributed to an enhancement in the effective absorption coefficient due to a local enhancement of the excitation light, which couples with the cavity mode. The on-resonant excitation technique selectively and efficiently excites only the QDs in the cavity. On-resonant excitation at energies below the wetting layer band gap energy can achieve stronger light emission from the cavity mode than excitation at energies greater than the wetting layer band gap energy with much less undesirable background emission. It will be shown that this is primarily due to the enhancement in the effective absorption by the cavity resonant effect and the direct carrier generation in QDs.

The two-dimensional photonic crystal cavities employing local three-dimensional structures are investigated. Donor-type cavities with small and shallow holes are examined as examples of cavities with a local three-dimensional structure. It is shown that both the far-field and near-field patterns of the cavities can be controlled by the addition of small and shallow holes. The fabrication of such structures using a focused ion beam is proposed and investigated. Residual gallium ions, used in fabrication with a focused ion beam, lead to optical absorption; these gallium ions can be removed by heat treatment at 800 °C. The near-field pattern of a fabricated three-dimensional cavity is measured, and is shown to be consistent with a theoretical analysis.

Second-harmonic generation (SHG) and sum-frequency mixing (SFM) are investigated in a two-dimensional photonic crystal (PhC) waveguide formed using a nonlinear optical (NLO) polymer with metallic cladding. The improved waveguide structure of the NLO polymer PhC enables the realization of ultraviolet and blue SHG and SFM radiation at wavelengths ranging from 368 to 400 nm. Large SHG enhancement at 368 nm was observed, originating from resonance between the external pump laser field and a photonic band mode. The SHG enhancement results agree well with the experimental photonic band structure obtained by angle-dependent optical reflectivity and the theoretical structure generated by three-dimensional finite-difference time-domain calculations.

We propose a photonic molecule consisting of multiple nanocavities in a photonic crystal and demonstrate its lasing and modal characteristics by finite-difference time-domain analysis and by experiments. In the analysis, we show that a point-shift defect and a point-missing defect have s- and p-orbital cavity modes, respectively, and that two adjacent defects exhibit σ- and π-orbital bonding and antibonding modes, which are similar to electronic states in chemical molecules. For such structure, we verify the modal characteristics as photonic molecules, i.e., mode splitting and its dependence on coupling strength, through their fabrication into GaInAsP photonic crystal slabs and the observation of lasing characteristics.

Organic-based photonic crystal (PC) nanocavities emitting in the visible-light range have been demonstrated. Three types of organic layer were used as emitting layers for PCs with emission spectra in the blue, green and red light wavelength ranges. Emission peaks caused by the resonant modes of the PC nanocavities were observed. The wavelength of the peak was controlled by changing the lattice constant of the PC. The emission peaks were observed in the wavelength range of 435 to 747 nm.

An ultralow group velocity of light in a III–V semiconductor photonic crystal (PC) waveguide is expected to enhance the optical gain per unit length of the waveguide. We discuss the possibility of a compact semiconductor optical amplifier that utilizes a single-line-defect PC waveguide fabricated from an all semiconductor low-Δ slab with deep airholes. From photonic band and traveling-wave rate equation analysis, a gain of larger than 20 dB is expected in a short active PC waveguide of 10 µm length, although the waveguide mode has a large radiation loss. We experimentally demonstrate such a waveguide in the passive wavelength regime. The waveguide is fabricated into a GaInAsP/InP wafer with high-mesa input/output waveguides. We show that the observed transmission characteristics of light correspond well to the photonic band theory.

We succeeded in fabricating novel nanoparticle structures on KNbO₃–TeO₂ glass induced by ultraviolet (UV) laser irradiation and found that the size and size distribution could be controlled by the conditions of laser irradiation, such as fluence, pulse repetition rate, and the temperature of the glass specimen by heat assistance (HA). Nano-sized particles are an interesting new material, because they will be able to open the door to the invention of photonic circuits of extremely small size, which transfer optical signals through optical near-field energy operating by a completely different principle. The particle diameters may be controlled from approximately 200 nm up to 500 nm by modifications of UV laser fluence and the repetition rate of pulses. When using HA with a glass specimen during laser irradiation, it has been found that nanoparticle formation is activated and particle size increased as the temperature of the glass increased, and the uniformity of particle size is markedly improved by HA treatment. The standard deviation of particle size is 20 nm at 100 °C. A 60% improvement in the size variation of particles is possible in comparison with laser irradiation without HA. These results suggest that heat is involved in nanoparticle formation during UV laser irradiation, which means that the overall history of heating with pulsed laser irradiation is critically important to control the form of the particles. Furthermore, periodic structures with an 800 nm pitch constructed by lines of ordered nanoparticles with diameters of approximately 150–300 nm were successfully fabricated by laser irradiation through a phase mask. This is a significant advancement towards the realization of nanophotonic circuits.

We designed and fabricated 70×75 µm² arrayed waveguide grating demultiplexer consisting of Si slab and wire waveguides on a silicon-on-insulator substrate. By optimizing the connection between components and the layout of arrayed waveguides, internal light scattering and the increase in phase error were suppressed. As a result, clear demultiplexing characteristics were observed with a channel spacing of 8 nm and a sidelobe level of -22 dB in the wavelength range from 1.5 to 1.6 µm.

We investigate the time-resolved photoluminescence (PL) spectra of Si-nanocrystal (Si-nc)-doped SiO₂ on Au thin films. It is shown that PL intensity within several tenth of µs after excitation is increased in the presence of Au films. The data suggest that the radiative recombination rate of excitons in Si-nc's is increased, and the degree of increase depends strongly on the emission photon energy. We show that the enhancement is caused by the modification of the local photonic mode density in the presence of Au thin films.

Negative bias temperature instability (NBTI) degradation of p-channel metal–oxide–semiconductor field effect transistors ( p-MOSFETs) is significantly underestimated by using the conventional characterization techniques due to a severe NBTI recovery during the measurements. In this work, a simple characterization method based on the single-point measurement of the saturated drain current is proposed to minimize the unwanted recovery effect. This method is accurate as is proven by a carefully-designed experiment. With this new proposed method, the measurement time is reduced to tens of milliseconds. This method gives a closer-to-real threshold voltage shift, and yields a more reliable power-law factor and thus a more realistic NBTI picture.

In this paper, we present the detailed fabrication process and high-frequency characterization of a new silicon through-wafer via interconnection and a low pass filter module flip-chip bonded to these via interconnections. An oxide liner of 18 µm thick for the via was fabricated on a complementary metal–oxide–semiconductor (CMOS)-grade low-resistivity 5 Ω·cm silicon wafer using the oxidized porous silicon (OPS) process. The through-wafer vias were filled with copper by electroplating. For a via interconnection of 240 µm length and 70 µm diameter, the series inductance and resistance are 0.079 nH and 0.059 Ω each. A coplanar waveguide (CPW) and a RF low pass filter (LPF) module were assembled on this through-wafer via interconnection substrate.

We have prepared silicon nanoparticles by an arc discharge method in liquid nitrogen. This system is easily adoptable in the preparation of large quantities of Si nanoparticles and has the potential for low-cost production. Etching Si nanoparticles with hydrofluoric acid reduced their size and passivated their surface such that they exhibited bright visible luminescence under UV illumination. The slow HF etching allows the formation of Si nanoparticles with a narrow size distribution. After etching, nanoparticles as small as 3–6 nm in diameter were obtained. Raman spectroscopy was used to determine the average diameter of Si nanoparticles during etching. The photoluminescence of Si nanoparticle was measured as a function of size and could be associated with a quantum confinement model.

To clarify the role of grain boundaries in iron sinks and carrier recombination centers, iron distributions and their chemical states were studied before and after gettering. They were measured by the X-ray microprobe fluorescence and the X-ray absorption in the near-edge structure using the beamline 37XU at the SPring-8 third-generation synchrotron facility. To determine the crystallographic orientation of the grain boundaries, electron backscatter diffraction measurements were performed. The distribution of electric active defects was characterized by electron-beam-induced current measurements. Before gettering, the iron was distributed in the small grain and its chemical state was similar to that of iron oxide. After gettering, the iron was redistributed along the small angle grain boundary, and its chemical state was similar to the iron silicide complexed with the iron oxide. Regarding the electrical activity, high carrier recombination was observed along the small-angle grain boundary. On the contrary, Σ3 grain boundaries were relatively weak impurity sinks and showed low recombination activity.

We have observed for the first time the large temperature dependence of Coulomb blockade (CB) oscillation peak current in a room-temperature-operating silicon single-hole transistor (SHT) with high peak-to-valley current ratio (PVCR). The large temperature dependence is not explained by the classical CB theory for single-dot single-electron transistors (SETs). The SHT is fabricated in the form of an ultranarrow-wire-channel metal–oxide–semiconductor field-effect transistor (MOSFET), which acts as a single-dot SHT at room temperature. It is found that, considering the result of numerical calculation, this large temperature dependence is caused by stochastic Coulomb blockade in the SHT, which has multiple-dot behavior at low temperatures. Other possible origins, such as thermally activated current and parasitic MOSFETs, are also discussed.

In this study we developed a phototransistor from a pseudomorphic high-electron-mobility transistor (pHEMT) for ultrawide-band fiber-radio communications. The phototransistor can be used not only as a photodetector for absorbing optical signals, but also as an optoelectronic mixer for up-converting intermediate frequency (IF) signals into radio carrier signals. Experimental results show that the proposed phototransistor can provide good mixing performance for the lightwave and microwave signals. It can also simplify the internal structure of a radio access unit (RAU) for ultrawide-band (UWB) fiber-radio systems.

Microcrystalline silicon (µc-Si) based single junction solar cells have been deposited by very high-frequency plasma-enhanced chemical vapor deposition (VHF-PECVD) using a showerhead cathode at high pressures in depletion conditions. The i-layers are made near the transition from amorphous to crystalline. An energy conversion efficiency of 9.9% is obtained with a single junction solar cell that is deposited on a texture-etched ZnO:Al front contact. The µc-Si i-layer is 1.5 µm thick, deposited at a rate of 0.5 nm/s. In order to control the material properties in the growth direction, the hydrogen dilution of silane in the gas phase is graded following different profiles with a parabolic shape. Materials with higher deposition rates were developed by increasing the RF power and the total gas flow such that the depletion condition is constant. At a deposition rate of 4.5 nm/s, a stabilized conversion efficiency of 6.7% is obtained for a single junction solar cell with a µc-Si i-layer of 1 µm. It is found that the defect density increases one order of magnitude upon the increase in deposition rate from 0.45 to 4.5 nm/s. This increase in defect density is partially attributed to the increased energy of the ion bombardment during the plasma deposition. We have introduced an additional method to limit the ion energy by controlling the DC self bias voltage using an external power source. In this way, the defect density in the µc-Si layers is decreased and the performance of the solar cells is further improved. It is observed that the performance of solar cells deposited at high rate improves under light soaking conditions at 50 °C, which we attribute to post deposition equilibration of a fast deposited transition material.

A compact and empirical model of subthreshold swing S and body factor γ is developed in bulk metal oxide semiconductor field effect transistor (MOSFET) in the short-channel regime. Although the relation between S and γ is simply given as S=60(1+γ) in the long-channel regime, this relation no longer holds in the short-channel regime due to the short channel effect (SCE). The model is derived using the capacitance network model and by fitting analytical equations to two-dimensional device simulation results. It is confirmed that the model is valid not only at low drain voltage but also at high drain voltage considering the effect of drain-induced barrier lowering. This model is very effective in the design of variable threshold-voltage complementary metal oxide semiconductor (VTCMOS) circuits.

A confined-chalcogenide (CC) cell structure for reducing the reset current of phase-change random access memory (PRAM) is proposed in this investigation. Both single normal-bottom-contact (NBC) (for reference) and proposed CC PRAM cells are simulated by two-dimensional finite element modelling. The simulated amorphous region of the NBC cell after reset operation is generally a semiellipse, which agrees very well with the reported experimental results. The CC cell has a rectangular amorphous region after reset operation. The reset operation current of the CC cell is much lower than that of the NBC cell. The CC cell structure needs a low reset current and a low power consumption and has a simple configuration.

I have proposed and developed dual-gate polycrystalline silicon thin-film transistors (poly-Si TFTs) with an intermediate lightly doped region (LDR) for the reduction of leakage current. The proposed poly-Si TFTs are easily fabricated and have a symmetric structure less sensitive to misalignment than the conventional LDD poly-Si TFTs. In the proposed TFTs, it is proved that a decrease in leakage current is due to a reduction in lateral electric field at the drain edge and a reduction in on-current is caused by an increase in the resistance of the LDR. The leakage current of the proposed TFTs is significantly reduced and the maximum ON/OFF current ratio is obtained with a 2 µm LDR length and a 2×10¹³/cm² LDR implant dose.

We prepared micro-thermoelectric hydrogen sensors (micro-THSs) operating with a combination of the thermoelectric effect of SiGe thin film and Pt-catalyzed exothermic hydrogen oxidation of the catalyst. The membrane area of each sensor was heated by a micro heater. The responses of the micro-THSs with three different types of membranes to hydrogen in air were investigated. The micro-THS with a single membrane, which had a better thermal isolation than those of other membranes, reduced the power consumption of 251 mW and the time required to heat the membrane by 5 s. This micro-THS with a single membrane showed the optimal gas sensing property of voltage signal of 1.13 mV for 1% hydrogen in air with the low detection limit of 50 ppm.

To investigate the surface properties of silicon germanium (SiGe) as a chemically sensitive membrane, SiGe layers deposited at Si₂H₆ and GeH₄ ambience by low-pressure chemical-vapor-deposition (LPCVD) at 475 °C were prepared. First, the effect of ac measuring frequency on capacitance–voltage (C–V) curves obtained from a SiGe electrolyte–insulator–semiconductor (EIS) structure was studied. The pH sensitivity of SiGe-EIS structure was 59.8 mV/pH, calculated from reference voltages in the pH range from 10 to 2. A voltage shift was observed in the same pH buffer solution for different measurement loops. To analyze the depth profile of a SiGe membrane, Auger electron spectroscopy (AES) was used. The formation of a native oxidation layer on the surface of SiGe was observed, and the layer had a thickness of 20 Å.

Structural changes of Y₂O₃ films and La₂O₃ films deposited on some oxidized silicon substrates were studied using X-ray photoelectron spectroscopy (XPS), Secondary ion mass spectrometry (SIMS), and Fourier transform infrared spectroscopy attenuated total reflection method (FT-IR ATR). Y₂O₃ and La₂O₃ films on chemical oxide and NH₃ annealed oxy-nitride were prepared by the Low-pressure chemical vapor deposition (LPCVD) method using an lanthanide–dipivaloyl-methanate (Ln–DPM) complex. The Y₂O₃ film and the La₂O₃ film on the both kinds of substrate already contained a partly silicate structure at the interface side as a result of an interface reaction during the deposition process. During post deposition annealing, the whole film structure of the Y₂O₃ and the La₂O₃ on the chemical oxide changed to a silicate structure due to silicon diffusion with interface reaction. In the case of the Y₂O₃ film, this interface reaction can be suppressed using thermal oxy-nitride as the interfacial layer. In the case of the La₂O₃ film, the suppression effect using oxy-nitride was smaller than the case with the Y₂O₃ film. Also, it was found that there was a strong correlation between the structural change of the films and the change of flat-band-voltage of both Y₂O₃ and La₂O₃ MIS diodes during post-deposition-annealing.

Depth profiles of composition and chemical structures in radical nitrided silicon oxynitride films formed with Ar/N₂, Xe/N₂, or Ar/NH₃ plasma excited by microwave have been investigated by X-ray photoelectron spectroscopy combined with step etching in HF solution. The relationship between the intensities of emission from N₂⁺ radical in these plasmas and the concentration of nitrogen atoms forming Si₃≡N configuration near the silicon oxynitride film/Si substrate interface nitrided using these plasmas was studied. The emission intensities from N₂⁺ radical generated in Xe/N₂ or Ar/NH₃ plasma are a quarter or one-sixth of that from N₂⁺ radical generated in Ar/N₂ plasma respectively. However, the emission from NH radical is also detected in Ar/NH₃ plasma. Although the nitrogen concentration of Xe/N₂ plasma is smaller than that of Ar/N₂ plasma at the film/substrate interface, that of Ar/NH₃ plasma is larger than that of Ar/N₂ plasma at the interface. It is important for the reduction of the nitrogen concentration near the film/substrate interface to use Xe/N₂ plasma in which both of the generation efficiencies of N₂⁺ and NH radicals are low.

An abnormal program disturbance mode was found in 32-string NAND flash memory which was fabricated with 0.09 µm complimentary metal–oxide–semiconductor shallow trench isolation (CMOS STI) process technology. This new disturbance mainly occurs in cells next to source select line (SSL) transistor and is not suppressed even when program bias is not applied. This is strongly related to the boosted channel bias level and the interface state which is located between the tunnel oxide and the silicon substrate. This unexpected program disturbance is a hot carrier program which results from the high electric field between the SSL transistor and its nearest cell and the leakage current to the boosted channel. The leakage current to the channel is due to the interface state of the SSL transistor, not the gate induced drain leakage (GIDL) effect. We present a brief model of this abnormal program disturbance mechanism.

Simulations for the optimization of the mask error enhancement factor (MEEF) by using Taguchi's design of experiment (DOE) method in both dry and immersion ArF lithography have been demonstrated here. By DOE, MEEF has been successfully reduced, and the process window has also been enlarged. Furthermore, in immersion lithography, MEEF has been significantly reduced, and resolution is also enhanced. In this study, we determined the optimal process and optical parameters to enlarge the process window and reduce MEEFs for different types of mask by DOE.

HfO_xN p-channel metal–oxide–semiconductor field-effect transistors (MOSFETs) with a low threshold voltage (|V_th|) were successfully fabricated using a partially silicided (PASI) platinum gate electrode for scaled complementary MOS (CMOS). The PASI platinum (PASI-PtSi) gate electrode is composed of a monoclinic-Pt₃Si phase in the vicinity of a HfO_xN dielectric. The reduced silicon content of the PASI gate electrodes is effective in suppressing the Fermi-level pinning on the Hf-based gate dielectrics which induces a significant |V_th| shift in p-channel MOSFETs. It is shown that the |V_th| of HfO_xN p-channel MOSFETs with the PASI-PtSi gate electrode is sufficiently low and applicable to low-standby-power devices. The mobility of the holes at 0.8 MV/cm is as high as about 90% of the universal mobility. It is concluded that the PASI technology in which the gate electrode has a reduced silicon content is useful for scaled CMOSs.

It was demonstrated that recovery from dry etching and ashing damage in porous silica low-k films occurred by 1,3,5,7-tetramethylcyclotetrasiloxane (TMCTS) vapor annealing. The increase in k-value after Ar/C₅F₈/O₂ plasma etching was reduced from 35 to 6.5% of the initial value (k=2.25) by TMCTS vapor annealing. Leakage current also returned to the initial level. Hydrofluoric acid wet etching revealed the sidewall damaged region in a porous silica trench due to plasma processes. The TMCTS vapor annealing was found to be effective for recovery from the sidewall damage. Fourier transformed infrared absorption spectroscopy indicated that the replacement of Si–CH₃ bonds in low-k films by Si–O and Si–OH bonds occurred during plasma processes. The recovery mechanism involves hydrophobic bond (–CH₃) reintroduction into the film followed by stable cross-linked poly(TMCTS) network formation on pore wall surfaces by TMCTS vapor annealing.

A 5.2 GHz upconversion Gilbert mixer with single-ended intermediate frequency (IF), local oscillator (LO), and RF ports is demonstrated using 0.35 µm SiGe heterojunction bipolar transistor (HBT) technology. The upconverter has a single-ended IF port. In addition, a lumped-element rat-race hybrid is inserted in the LO input stage to maintain balanced LO signals. The physics of the lumped-element rat-race is discussed in this paper. A passive LC current mirror is used to convert the mixer's differential outputs into a single-ended output and double the output current. The design methodology of the LC current mirror and a new approach to improve the power gain with the output buffer are developed in this paper. The fully matched Gilbert upconverter has the conversion gain of -1 dB, OP_{1 dB} of -10 dBm and OIP₃ of 6 dBm when input IF=300 MHz, LO=4.9 GHz and output RF=5.2 GHz. The LO-IF isolation is -36 dB and the LO-RF isolation is -39 dB. The supply voltage is 3.3 V, the current consumption is 11.5 mA, and the die size is 0.98×0.83 mm² with six integrated on-chip inductors.

A pyramidal-shaped GaAs (tip-GaAs) photocathode for a polarized electron source (PES) was developed to improve beam brightness and negative electron affinity (NEA) lifetime by field emission. The emission mechanism also enables the photocathode to extract electrons from the positive electron affinity (PEA) surface into vacuum, and alleviates the NEA lifetime problem. The measured electrical characteristics of tip-GaAs and its polarization exhibited distinctive field-emission behavior. The polarization of the electron beam extracted from tip-GaAs was 20–38% under irradiation with circularly polarized light of 700–860 nm, and the peak polarization was 37.4±1.4% at a wavelength of 731 nm. These experimental results indicate that spin-polarized electrons can be extracted from the conduction band into vacuum by a field-emission mechanism. This, in turn, shows that this type of photocathode has the prospect of generating a low-emittance spin-polarized electron beam.

The reaction of CH₄ on a Si(111)-7×7 surface is investigated by reflection high-energy electron diffraction (RHEED) analysis, quadrupole mass spectroscopy (QMS) and scanning electron microscopy/energy dispersive analysis of X-ray (SEM/EDAX). The RHEED patterns during CH₄ exposure indicate the evolution of structures such as δ-7×7, 1×1, √3×√3 and SiC for various exposures at temperatures from room temperature (RT) up to approximately 800 °C. A mass spectrum shows that CH₄ decomposition products are primarily CH₃, CH₄, and H₂ at a tungsten filament temperature of 1500 °C. A SEM image of the SiC formed at 825 °C and 7920 L shows that the surface is rugged and consisted of hills and valleys. On the basis of the EDAX measurements, it is determined that the SiC layers are C-rich at the hill and Si-rich at the valley. It is found that the SiC layer is generated only after the formation of the √3×√3 structure.

Hf adsorption on thin SiO₂ films was carried out using an ultrahigh-vacuum system with e-beam evaporation. The interaction of adsorbed Hf and SiO₂ films was studied by Auger electron spectroscopy (AES), atomic force microscopy (AFM), and current–voltage (I–V) measurements. The leakage current of the SiO₂ films increased with Hf adsorption temperature. The dielectric properties of the SiO₂ layer were markedly degraded by Hf adsorption even at 300 °C. Additionally, Hf adsorbed on a SiO₂ (6.3 nm)/Si substrate at 900 °C penetrated through the SiO₂ layer and reacted with the Si substrate forming large concavities. Otherwise, Si diffused from the SiO₂/Si interface to the surface of adsorbed Hf.

The effects of thermal annealing between Ni film and a p-type GaN layer have been investigated. The electrical and optical properties were measured by Hall effect, capacitance–voltage (C–V) and photoluminescence (PL) measurements. The samples activated with Ni film obtained higher effective carrier concentrations than those activated without Ni film. Effective carrier concentrations of 5×10¹⁵ and 1×10¹⁷ cm^-3 were achieved at an activating temperature of 400 °C without and with Ni film. The Ni film may act as a catalyst for the activation of Mg-doped GaN at a temperature less than 500 °C. At a temperature higher than 600 °C, the Ni film may react with the Mg-doped GaN. X-ray diffraction (XRD) analyses indicated that Ni film on Mg-doped GaN transforms to nickel oxide (NiO) and nickel nitride (Ni₃N) during thermal annealing in air. The peaks of the PL spectra at 15 K of the samples activated at 600 °C with and without Ni film were observed at approximately 3.2 and 2.9 eV. We suggest that Ni atoms not only enhance hydrogen desorption but also diffuse into the Mg-doped GaN layers to form Ni-compound materials. At a high annealing temperature, impurities such as Ni nitride, nitrogen vacancies or other defects may reduce the carrier mobility and provide an increase in the effective carrier concentrations in the surface region.

We are developing high energy resolution X-ray microcalorimeters based on iridium and gold (Ir/Au) phase transition thermometers. Here we analyze the signal behavior of a 200×200 µm² Ir/Au transition edge sensor (TES) with iridium and gold having the thicknesses of 100 and 25 nm, respectively. The energy resolution of this device was 15.4 eV full width at half maximum (FWHM) at 5.9 keV X-ray energy. For operation at bias resistances lower than 140 mΩ, we observed two distinct decay components, a slow component followed by a very fast component, which is thought to be due to the variation of current distribution inside the TES. However, taking into consideration the nonlinearity associated with the slow component, a very fast signal response makes the operation of this device promising at low-resistance bias points.

The third-order nonlinear optical properties of a chloroform solution and a neat film of poly(9,9-dioctylfluorenyl-2,7-yleneethynylene) (PDOFE) were investigated by forward degenerate four-wave mixing (DFWM) technique under resonant conditions. The temporal profiles of the DFWM signal of PDOFE were obtained with a time resolution of 0.3 ps (FWHM), and were found to consist of at least three components, i.e., a coherent instantaneous nonlinear response (electronic response) and two slow responses due to the formation of excited molecules for both the chloroform solution and the neat film of PDOFE. The electronic component of molecular hyperpolarizability per repeat unit of PDOFE in the chloroform solution was determined to be ca. (6.2±0.3)×10^-31 esu at 388 nm. The third-order nonlinear optical susceptibility of the neat film of PDOFE was estimated to be ca. (1.2±0.1)×10^-8 esu at 387 nm.

A passively Q-switched and mode-locked diode-pumped Nd:GdVO₄ laser was demonstrated using a low-temperature-grown GaAs wafer (LT-GaAs) as an intracavity saturable absorber. The maximal Q-switched mode-locked average output power was 750 mW with the Q-switched envelop having a repetition rate of 167 kHz. The mode-locked pulse trains inside the Q-switched pulse envelope had a repetition rate of ∼790 MHz.

We report a coupled quantum dot (QD) structure for long wavelength laser applications. The structure comprises an InAs seed layer and a second InAs QD layer capped with an In_0.33Ga_0.67As capping layer. Cross-sectional transmission electron microscopy (TEM) images show a vertical alignment between the QD stacks, which causes the coupled QD sample to have a larger dot size and a lower dot density than the control sample. The laser with the coupled QD structure exhibits a markedly longer emission wavelength and a slightly higher threshold current density than lasers with a conventional QD structure, indicating that the coupled QD structure has potential for long wavelength applications.

We report that a unique reliable process for manufacturing a 28 mm small disk for the blue-ray optical recording system has been developed successfully. To prevent the severe axis run-out, a stepped metal hub for magnetic chucking is directly inserted into the punched-center of a 0.6 mm thick disk. Instead of the conventional laser initializer, we make a new type of UV disk initializer, using Xe-type UV-lamps, which is expected to increase the productivity of rewritable optical disks, due to the short initialization time. Furthermore, the modified dynamic tester makes it possible to analyze small blue-ray disks effectively.

We report the optical modification of gallium nitride by femtosecond laser irradiation. A 5-nm-band-gap shift, observed by spectral ellipsometry, results a reduction of 0.01 in the refractive index at a wavelength of 400 nm. An indistinct circular pattern was also observed using nano-beam diffraction. These results suggest that femtosecond-laser irradiation induces stress–strain, rather than slight amorphization. We also observed that the angle of first-order diffraction from the grating with a 1.0 µm period, fabricated by interferometric irradiation using a femtosecond laser without any damage, corresponds to the calculated angle. The first-order diffraction efficiency for a wavelength of 403 nm was estimated to be 4.0×10^-3%.

In this paper, we present a monolithic integration of a microlens and an InP/InGaAs heterojunction bipolar phototransistor (HPT). Three types of HPTs were fabricated and their optical characteristics were investigated. The optical gain of the proposed HPT was improved 6-fold compared to the front type HPT. The HPT with a back-side microlens showed 0.61 A/W dc responsivity (PD mode) and that of an HPT without a lens showed 0.18 A/W (PD mode) at 1.55 µm illumination. Quantum efficiency was increased by 34.8%. The fiber alignment tolerance has also been improved. The HPT with a lens showed a 40 µm tolerance, while the HPT without a lens exhibited a 22-µm tolerance at 3 dB below the maximum values.

A waveguide lens, composed of multistage polymer-filled thin grooves in a silica planar lightwave circuit (PLC) is proposed and a low-loss structure has been designed. A waveguide lens in a silica slab waveguide has been fabricated using reactive ion etching (RIE) and formed by filling with polymer. Both an imagding optical system and a Fourier-transform optical system can be configured in a PLC using a waveguide lens. It renders the PLC functional and its design flexible. To obtain a shorter focal length with a low insertion loss, it is more effective to use a multistage lens structure. An imaging optical system and a Fourier-transform optical system with a focal length of less than 1000 µm were fabricated in silica waveguides using a multistage lens structure. The lens imaging waveguides incorporate a 16–24-stage lens, with insertion losses of 4–7 dB. A 4 ×4 optical coupler, using a Fourier-transform optical system, utilizes a 6-stage lens with losses of 2–4 dB.

A ZnS–SiO₂ composite dielectric is widely used in the optical stack designs of rewritable optical recording media as an index-matching medium and as a protection layer for the high-index chalcogenide (compound with sixth group element of S, Se, Te) phase change material used in these media. The addition of Si and O to ZnS is primarily intended to stabilize against crystalline grain growth of ZnS with high numbers of direct overwriting cycles. In this study, we carry out infrared (IR) spectroscopy to clarify the role of Si in this stabilization process. IR spectroscopy is performed on sputter as-deposited and annealed ZnS–SiO₂ dielectric protection layers. We find that Si exists not in the SiO₂ oxide phase but as [SiS_4-nO_n] tetrahedrons. Moreover, zinc and sulfur do not exist as ZnS, but in highly chemically disordered ZnS:O crystallites. The highly directional and rigid covalent bonds in the [SiS_4-nO_n] tetrahedrons are key to establishing thermal stability against the coalescence of ZnS. The importance of the Si–S bond also extends into a more thorough understanding of the low thermal conductivity of the ZnS–SiO₂ material. The consideration of elastic implications allows us to predict an average phonon velocity less than 50% compared to that in SiO₂. With this, we predict a thermal conductivity of 0.0067 W cm^-1 K^-1 for this material, which is in complete agreement with measured values.

A non-equilibrium heating situation occurs when the intensity of heat flux is high and the heating duration is shorter than the thermalization time of the substrate material. In the present study, the heating of gold due to time exponentially decaying surface heat flux is considered. Electron and lattice site temperatures are predicted from the electron kinetic theory approach. To quantify the thermodynamic irreversibility in the thermal system, entropy analysis in the electron and lattice subsystems is performed and thermal communication between both systems is formulated. It is found that electron and lattice subsystems separate thermally. Electron temperature rise is rapid while lattice temperature rise is gradual during the early heating period. Volumetric entropy generation in the electron subsystem is higher than it is in the lattice subsystem.

In this paper, we describe a second-harmonic generation scheme with a blue laser diode (LD) that generates tunable coherent deep ultra violet at approximately 219 nm. Using a short fundamental wavelength with this technique, a high second-harmonic (SH) power is hardly observed. At an approximately 8 µW SH power at 219 nm and an approximately 1 nm fundamental wavelength tunability can be obtained using a single-mode blue LD chip, a nonlinear optic crystal, and an external cavity.

Self-starting pulse generation from a ytterbium-doped fiber (YDF) ring laser is presented. The laser consisted of an all-polarization-maintaining (PM) fiber system, specifically a single-mode PM-YDF with a side hole. Environmentally stable pulses and a wide tuning range of 60 nm were obtained with a polarization extinction ratio >20 dB. The YDF oscillator described in this study is compact, reliably self-starting, and has excellent long-term stability in light of its robustness to thermal and mechanical disturbances.

The phase transformation of a liquid crystal lens with its focus movable not only along but also off the lens axis is discussed. A lateral translation of the phase transformation is found. The wave front of an incident light wave in the exit pupil of the lens is observed by the interference method. It is found that the focusing ability of the lens is preserved even when the focus is away from the lens axis. From the obtained wave front, the focus position and the appropriate voltage for maintaining the focus in the focal plane are determined.

An optical isolator employing a nonreciprocal phase shift is composed of a three-guide tapered coupler. The coupling characteristics of the three-guide tapered coupler for an integrated optical isolator with a Si guiding layer are demonstrated.

Polylactic acid (PLA) is a biomass material made of starch. Compact disc recordable (CD-R) discs with a PLA substrate were prepared and their properties were measured. Although the glass transition point of PLA is lower than that of polycarbonate (PC), the PLA substrate is usable for CD-R discs. It was confirmed that the substrate is usable for recordable optical discs at temperatures under 50 °C.

A novel diffractive-pumping scheme is proposed to improve the evanescent amplification using blazed fiber grating for the first time. We also investigate the cw-pumped evanescent amplification at 1.55 µm wavelength with the relative optical gain pumped at 1480 nm of around 2 dB based on side-polished fiber with the effective interaction length as long as 16 mm and with a heavily Er³⁺-doped (N_Er³⁺>1.19×10²¹ ions/cm³), low refractive index (n₁₅₅₀<1.47) glass overlay, which has no concentration quenching (τ_f=9.0 ms).

The multilevel run-length-limited (ML-RLL) modulation recording can improve the capacity and the data transfer rate of the optical disc without changing the optical and mechanical units. A viterbi detector with feedback is used to improve the performance of the partial-response maximum likelihood (PRML) system for ML-RLL optical recording. It can avoid the complexity of viterbi detector caused by the overlength of the memory length of the partial-response mode. The principle, implementation and the simulated results of the viterbi detector with feedback used in ML-RLL system are all presented.

High speed recordable media often show only marginal modulation at the lowest recording speeds. This reduction is caused by bumps formed in the polycarbonate substrate during writing. After treating sliced discs with NaOH, holes were found underneath the written marks in the bottom of the grooves only in the case of high speed recording. This effect can be explained by postulating an additional polycarbonate layer with a lower refractive index. Simulations show that this extra layer increases modulation at high recording speeds. The simulation coincides well with experimental results.

An initialization-free Blu-ray disc was proposed as a candidate for multispeed blue laser recording and successfully fabricated. Experiment results of the initialization-free Blu-ray disc showed that it had a multispeed blue laser recording capability with speeds of 1 to 3×. The crystallization acceleration effect caused by additional layers and the rapid cooling caused by the initialization-free disc structure were suggested as the physical mechanism of multispeed recording in an initialization-free Blu-ray disc.

A UV-cured adhesive with a low glass transition temperature (T_g) below the room temperature was used as a substrate for manufacturing a film-like liquid crystal display. This film-like liquid crystal display shows a fully flexibility due to the low T_g property, and the manufacture processes are roll-to-roll compatible due to the coating processes.

The relationship between the electronic properties of grain boundaries and polycrystalline silicon (poly-Si) thin-film solar cell performance was investigated. The poly-Si thin films were directly deposited on foreign substrates by atmospheric pressure chemical vapor deposition. We evaluated the interface state density at the grain boundaries of poly-Si thin films deposited under different deposition conditions by measuring the temperature dependence of Hall mobility. We found a strong correlation between the interface state density and poly-Si thin film solar cell performance. The decrease in interface state density resulted in the improvement of the solar cell performance, which was caused by decreasing the minority carrier recombination velocity at grain boundaries.

A two-band model involving the A- and B-valence bands was adopted to analyze the temperature-dependent Hall effect measured on p-type ZnO. The hole transport characteristics (mobilities and effective Hall factor) are calculated using the "relaxation time approximation" as a function of temperature. It is shown that the lattice scattering by the acoustic deformation potential is predominant. In the calculation of the scattering rate for ionized impurity mechanism, an activation energy of 100 or 170 meV is used at different compensation ratios between donor and acceptor concentrations. The theoretical Hall mobility at an acceptor concentration of 7×10¹⁸ cm³ is about 70 cm² V^-1 s^-1 with the activation energy of 100 meV and the compensation ratio of 0.8 at 300 K. We also found that the compensation ratios conspicuously affected the Hall mobilities.

The microstructures and microwave dielectric properties of the (1-x)(Mg_0.95Zn_0.05)TiO₃–xCa_0.6La_0.8/3TiO₃ ceramic system were investigated. To achieve a temperature-stable material, we studied a method of combining a positive-temperature-coefficient material with a negative one. Ca_0.6La_0.8/3TiO₃ has the following dielectric properties: a dielectric constant of ε_r∼109, a Q×f of ∼17,600 GHz and a large positive τ_f of ∼213 ppm/°C. (Mg_0.95Zn_0.05)TiO₃ possesses a dielectric constant of ε_r∼16, a high Q×f of ∼210,000 (at 9 GHz) and a negative τ_f of -60 ppm/°C. As x varies from 0 to 0.9, the (1-x)(Mg_0.95Zn_0.05)TiO₃–xCa_0.6La_0.8/3TiO₃ ceramic system has the following dielectric properties: 16<ε_r<88, 32,800<Q×f<210,000 and -60<τ_f<205. By appropriately adjusting the x value in the (1-x)(Mg_0.95Zn_0.05)TiO₃–xCa_0.6La_0.8/3TiO₃ ceramic system, a zero τ_f value can be attained. With x=0.15, a dielectric constant ε_r of ∼26, a Q×f value of ∼86,000 GHz (at 9 GHz) and a τ_f value of ∼0.5 ppm/°C were obtained for 0.85(Mg_0.95Zn_0.05)TiO₃–0.15Ca_0.6La_0.8/3TiO₃ ceramics sintered at 1320 °C for 4 h.

Compositional design and properties of a lead-free, low-temperature cofired ceramic (LTCC) system containing a crystallizable K₂O–CaO–SrO–BaO–B₂O₃–SiO₂ glass, alumina and titania have been investigated. The ternary composites with a high sintered density of >95% can be obtained at 875 °C. Compositions with tailor-made properties are designed according to a working model based on the assumption of the mixing rule, and validated with experimental results. The resulting lead-free composites exhibit a dielectric constant of 7.7–20.0 at 1 MHz, and a thermal expansion coefficient of (5.37–7.22)×10^-6/°C.

In this paper, we present the results of the fabrication and characterization of Li₂CO₃ doped (Ba,Sr)TiO₃ ceramics for the low temperature sintering processes. In these days, low temperature sintering process has been widely spread out for the integrated electronic modules for the communication systems such as front-end modules, antenna modules, and switching modules. Generally it is believed that low temperature sintering process can be applied to the functional materials if they can be sintered less than 900 °C. However, BaSrTiO₃ materials for the tunable microwave devices applications have relatively high sintering temperature of 1350 °C. Therefore, in this study to obtain low sintering temperature, we have added 1–5 wt % of Li₂CO₃ to BaSrTiO₃ materials to reduce the sintering temperature from 1350 to around 900 °C.

0.68Pb(Mg_1/3Nb_2/3)O₃–0.32PbTiO₃ single crystals produced by the modified Bridgman technique were poled along the [001] orientation. We discuss the low-symmetry phase domain in the morphotropic phase boundary (MPB) from a fine structural viewpoint based on the investigation by transmission electron microscopy (TEM). By using the convergent beam electron diffraction (CBED) method, we observed the crystal symmetry in each of the three orientations – [001], [101], and [111] – and determined the point and space groups. We identified the point group as m since the mirror symmetry in the [100] and [101] orientations was observed in CBED patterns obtained at the [001] and [101] incidences and no symmetry was observed in those obtained at the [111] incidence. The classification based on the kinematical extinction rule could not be applied because of the difference in the fundamental lattice between lattice systems P and C. Therefore, we compared the symmetry observed in the CBED pattern with the theoretical prediction and identified the lattice system as P. In addition, we examined the presence or absence of the glide plane and screw axis and identified the space group as Pm. We showed that the spontaneous polarization (P_s) of the monoclinic (space group Pm) domain is in the (010) plane and lies on the M_c line connecting point T that expresses a tetragonal system directed toward the [001] orientation and point O that expresses an orthorhombic system directed toward the [101] orientation. We also revealed that four P_s vectors exist equivalently with respect to the poling direction. The angle formed by any two of these vectors was almost orthogonal when observed from the [110] orientation.

Without any assumptions regarding residual impurity species and intrinsic defects in an undoped semiconductor, it is experimentally demonstrated that the densities and energy levels of impurities and defects can be precisely determined by a graphical peak analysis method based on Hall-effect measurements, referred to as free-carrier-concentration spectroscopy (FCCS). By FCCS, the number of acceptor species in p-type undoped In_0.2Ga_0.8Sb epilayers is determined, and the densities and energy levels of these acceptor species are accurately estimated. Two acceptor species, whose acceptor levels are E_V+25 meV and E_V+86 meV, are detected, where E_V is the valence band maximum. The density of the E_V+25 meV acceptor increases with Sb₄/(In+Ga) flux beam equivalent pressure (BEP) ratio, whereas the density of the E_V+86 meV acceptor decreases with increasing BEP ratio. These observations are not consistent with the conventional assumption that these acceptor species are V_Sb⁺ and V_Sb²⁺ in GaSb-based semiconductors, where V_Sb is the Sb vacancy.

The density (N_A) and energy level (E_A) of a B acceptor in B-doped p-type diamond epilayers are usually determined from the temperature dependence of hole concentration, p(T), using the Fermi–Dirac distribution function for acceptors, which does not consider the effect of the excited states of the B acceptor. However, in samples whose Fermi levels, E_F(T), are located between the valence band maximum (E_V) and E_A in a measurement temperature range, the obtained N_A is much higher than the concentration of B atoms determined by secondary ion mass spectroscopy. Because the B acceptor level in diamond is deep, the effect of the excited states of the B acceptor on p(T) should not be ignored. When E_F(T) is between E_V and E_A, the reasonable N_A and E_A are obtained by fitting a curve to p(T) using the distribution function including the effect of its excited states.

The piezoelectricity and its related characteristics of a superstructure of silica-based thin films were investigated. Thin films with a periodic structure of nanometer-ordered silica layers containing metal dopants performed like c-axis oriented piezoelectric thin films. Piezoelectricity, pyroelectricity, and an anomalous photovoltaic effect were observed in poled superstructure silica films. The polarization was stabilized by this structure to prevent aging. We found that under some deposition conditions, piezoelectricity was observed in as-deposited films that were not poled.

(001)/(100)-oriented tetragonal epitaxial Pb(Zr_0.35,Ti_0.65)O₃ [PZT] thin films with various volume fractions of 90° domains were grown on (100)_cSrRuO₃∥(100)SrTiO₃ substrates by metalorganic chemical vapor deposition. The polarization-electric field hysteresis curve and fatigue behavior of a PZT capacitor with Pt and SrRuO₃ top electrodes were systematically investigated. In a Pt/PZT/SrRuO₃ capacitor, the degradation of switching polarization occurred between 10⁴ and 10⁷ cycles. In contrast, a PZT capacitor with SrRuO₃ top electrodes revealed almost fatigue-free behavior up to 10⁹ switching cycles. It was found that the fatigue properties of PZT films is strongly dependent on the type of top electrode, irrespective of the volume fraction of the (001) orientation. Moreover, it was also found that the a-domains switching of PZT thin films is independent of the fatigue properties.

Electrical conductivity measurements and elastic recoil detection (ERD) analysis were carried our on perfluorosulfonic acid membranes that were irradiated by ⁶⁰Co γ-ray. Maximum electrical conductivity of irradiated specimen was obtained at 40–60 °C, though that of unirradiated specimen increased with increasing temperature. And 4000-fold higher conductivity was obtained in irradiated membrane at 40 °C under dry condition. ERD analysis indicated that hydrogen concentration of the irradiated membrane is 1.5-fold higher than that of the unirradiated specimen. It is considered that the γ-ray generates a lot of site that can trap hydrogen and the trapped hydrogen cannot easily move. As the results, irradiated membrane retains more hydrogen than unirradiated one and the electrical conductivity of irradiated membrane is higher than that of unirradiated membrane in any conditions. γ-Ray irradiation is effective for improving the electrical conductivity of proton-conducting membranes.

The effects of TiO₂ addition on the microwave dielectric properties of Ba_6-3x(Sm_1-y-zNd_yBi_z)_8+2xTi₁₈O₅₄ (BSNBT) ceramics, with x=0.6, y=0.17, and z=0.03, have been investigated. The microstructure of BSNBT doped with TiO₂ was analyzed using X-ray diffraction (XRD) analysis, electron probe microanalysis (EPMA) and field-emission scanning electron microscopy (FE-SEM). XRD and EPMA analysis reveals the existence of second phases. The addition of small amounts of TiO₂ (≤1.0 wt %) increased Q·f markedly and improved TC f without any influence on dielectric constant (ε_r). However, the addition of excess TiO₂ reduced the bulk density and dielectric properties of the materials. BSNBT doped with 0.5 wt % TiO₂ exhibited excellent dielectric properties with ε_r=80.5, Q·f=9009 GHz and TC f=6.5 ppm/°C when sintered at 1340 °C for 2 h. The reasons for the nonlinear behavior of TC f are also provided.

We investigated the electrooptical properties of a carbon nanotube (CNT)-doped twisted nematic (TN) liquid crystal (LC) cell. Experimental results reveal that the doped CNTs influence the elastic constant of LC–CNT dispersion. Using a small amount of CNT dopant, the field-on response time of the LC cell is nearly invariant; the threshold voltage of the cell increases due to the increase in the elastic constant of LC–CNT dispersion. At a higher CNT concentration, the marked increase in the dielectric anisotropy of LC–CNT dispersion markedly decreases the field-on response time and threshold voltage of the LC cell. The field-off response time of this cell decreases with increasing CNT concentration due to the increase in elastic constant and the slight increase in viscosity of LC–CNT dispersion. The field-on and field-off response times of the LC cell are reduced simultaneously when the LC host is doped with a moderate amount of CNT dopant.

A thermoelectric material, γ-Na_0.7CoO₂, with a high figure of merit, ZT>1, at high temperatures has been studied using the atomic pair distribution function (PDF) analysis of pulsed neutron powder diffraction data. We have found significant PDF peak broadening in the correlation between sodium and oxygen atoms at high temperatures. The broad peak indicates that a sodium atom frequently leaves a NaO₆ prism. In other words, sodium atoms hop in a two-dimensional layer at high temperatures. The PDF intensity of the Na–O correlation below a critical length, where a sodium atom is leaving a prism, is inversely proportional to thermal conductivity. This suggests that the short phonon lifetime originates from the frequent sodium hopping in the two-dimensional layer, leading to high-performance thermoelectric generation.

Secondary ion mass spectrometry is applied to the detection of hydrogen in individual diamond grains 1–4 µm in diameter. In the analysis, the primary Cs ion beam was fine-tuned at diameters of 50–200 nm on the surface, and then a combination of presputtering and spot analysis was carried out along a line defined in the scanned ¹²C image with a high lateral resolution. The intensity ratios of (H/¹²C) in the (111) and (100) planes were determined as having values of 1.36×10^-2 and 3.2×10^-3, respectively. The value obtained in the (100) plane is background in the analysis. The results demonstrate the potential of the analytical technique for the detection of impurities in materials science.

The control of wettability on a diamond-like carbon (DLC) surface was carried out by exposure to synchrotron radiation (SR) in the soft X-ray region under the perfluorohexane (C₆F₁₄) gas atmosphere. It was found that the contact angle of a water drop against the DLC surface increased with SR dose; in particular, it increased from 73 to 91° by the SR exposure of a 1 mA·h dose. The formation of the hydrophobic DLC surface was attributable to the fluorine-functional groups that existed on the DLC surface after SR exposure under the C₆F₁₄ gas atmosphere.

The binary compound tantalum nitride (TaN) and ternary compounds tantalum tungsten nitrides (Ta_1-xW_xN_y) exhibit interesting properties such as high melting point, high hardness, and chemical inertness. Such nitrides were deposited on a tungsten carbide (WC) die and silicon wafers by ion-beam-sputter evaporation of the respective metal under nitrogen ion-assisted deposition (IAD). The effects of N₂/Ar flux ratio, post annealing, ion-assisted deposition, deposition rate, and W doping in coating processing variables on hardness, load critical scratching, oxidation resistance, stress and surface roughness were investigated. The optimum N₂/Ar flux ratios in view of the hardness and critical load of TaN and Ta_1-xW_xN_y films were ranged from 0.9 to 1.0. Doping W into TaN to form Ta_1-xW_xN_y films led significant increases in hardness, critical load, oxidation resistance, and reduced surface roughness. The optimum doping ratio was [W/(W+Ta)]=0.85. From the deposition rate and IAD experiments, the stress in the films is mainly contributed by sputtering atoms. The lower deposition rate at a high N₂/Ar flux ratio resulted in a higher compressive stress. A high compressive residual stress accounts for a high hardness. The relatively high compressive stress was attributed primarily to peening by atoms, ions and electrons during film growth, the Ta_1-xW_xN_y films showed excellent hardness and strength against a high temperature, and sticking phenomena can essentially be avoided through their use. Ta_1-xW_xN_y films showed better performance than the TaN film in terms of mechanical properties and oxidation resistance.

In this study, a theoretical approach to determine effective nanoindentation range was conducted. The tip radius was first calculated from the experimental contact area function. Then a minimal indentation depth, at which the tip rounding effect can be negligible, was determined. As a representative system for the hard film on the soft substrate, the (Ti,Al)N film on the Si substrate was selected. The yield strengths were estimated for the evaluation of the substrate effect. The minimal indentation depth, at which the substrate effect can be negligible, was determined using the yield strength and tip radius. Finally, the effective nanoindentation range was estimated for the sample. The elastic modulus and the hardness of the film were also numerically verified.

We fabricated an electrode for electrochemical generation of ozone by sputtering. The vertical structure of the electrode was composed of silicon/titanium oxide/platinum/tantalum oxide (Si/TiO_X/Pt/TaO_X). The catalyst layer of the electrode was an insulator comprising only tantalum oxide. By using the electrode as the anode and perchloric acid as the electrolyte solution, ozone generation was observed at a current density of ca. 40 mA/cm² and above. According to the electrochemical measurement, the potential of the electrode was ca. 8 V [vs a saturated calomel electrode (SCE)] at a current density of ca. 40 mA/cm², and it was higher than the potential of the Pt electrode. The electrode was efficient for the electrochemical generation of ozone because of its high electrode potential.

The average energies required to produce one scintillation photon W_s were determined for 662 keV gamma rays in NaI(Tl) and CsI(Tl) crystals to be 15.0±1.3 and 13.3±1.1 eV, respectively, from the absolute numbers of photoelectrons measured for several combinations of a crystal and a photomultiplier tube (PMT) used as a vacuum photodiode. The numbers of scintillation photons were obtained by calculating the collection efficiency of scintillation photons at the photocathode using Monte Carlo simulations and by determining experimentally the photon-to-photoelectron conversion efficiency at PMT photocathode. The values of W_s determined in the present study are in good agreement with the theoretical values presented recently. The factors affecting energy resolutions were also examined. The calculated resolution agrees well with that obtained experimentally.

Channel-dependent fission barriers of n+²³⁵U were obtained from the analysis on the data of fission product yields using the selective channel scission (SCS) model. This analysis showed good agreement with the change in fission product yield with the incident neutron energy. Also, the degrees of the deformation of the nuclei obtained from the SCS model analysis were compared with those at scission configurations associated with mass-symmetric and mass-asymmetric fission modes.

Bottom-contact polymerized 10,12-tricosadiynoic acid (PTDA) field-effect transistors (FETs) were fabricated successfully by the Langmuir–Blodgett (LB) method. The surface morphology of PTDA LB films deposited on a FET substrate was imaged by atomic force microscopy (AFM). The current–voltage (I–V) characteristics of the bottom-contact PTDA FETs were measured in detail. The equation of the FET characteristics was derived with consideration of the carrier transport mechanism of holes conveyed along the channel, and the effective mobility was evaluated. To clarify the carrier injection and succeeding carrier accumulation and carrier transport from a source electrode to a drain electrode across PTDA, a Maxwell–Wagner model was employed to analyze the I–V characteristics of the bottom-contact PTDA FETs. Finally, to further clarify the carrier injection, we measured the capacitance–frequency (C–F) characteristics of the bottom-contact PTDA FETs and the surface potential formed across PTDA LB films on a Au electrode.

Cinnamate groups are well-known for a dimerization reaction upon exposure to ultraviolet irradiation and a thermal reaction after being heated. In this study, to verify the thermal reaction of the cinnamate group in detail, we investigated the thermal crosslinking of cinnamate oligomers. The thermal reaction of cinnamate oligomers of low molecular weight is induced more readily by thermal energy compared with that of cinnamate polymers. This reaction is attributed to a radical reaction involving the carbon-carbon double bond in the cinnamate group. The orientation of the liquid crystal depended on the length of the spacers in the cinnamate oligomers.

Striae configurations in TiO₂–SiO₂ ultra-low-expansion glasses caused by variations in TiO₂ concentration were investigated through measurement of leaky surface acoustic wave (LSAW) velocities by the line-focus-beam (LFB) ultrasonic material characterization system. LSAW velocity distributions were measured on the surface of a specimen substrate prepared from a large circular plate glass ingot produced by the direct method using a flame hydrolysis process. A subsequent procedure that included polishing the surface of the specimen to reduce the specimen thickness by 40 µm and measuring the LSAW velocity distributions on the surface was repeated five times in order to examine at least one period of the striae. TiO₂ concentration variations were clearly observed in the deposit direction of the glass ingot and in its radial direction. The maximum LSAW velocity difference over the whole examined region was 11.7 m/s, corresponding to 0.70 wt % TiO₂ concentration. The three-dimensional striae structure revealed that the striae plane in the examined region was almost parallel to the substrate surface but gently curved down in the radial direction. The plane had a slightly convex-shaped cross section layered with a striae periodicity of about 0.16 mm and a curvature radius of about 440 mm, and also existed in a circular form with a curvature radius of about 450 mm. The ultrasonic method will contribute to improvement of characteristics and homogeneity of glass associated with production-process conditions.

An ellipsometer with nanosecond time resolution has been proposed for the investigation of the phase change behavior of Ge₂Sb₂Te₅ heated by electrical pulses of 20–100 ns in real time. This passive single-wavelength ellipsometer has a division-of-amplitude photopolarimeter (DOAP) configuration for polarization state detection to collect ellipsometric data in nanoseconds and consists of a microfocusing lens system to achieve a spot size of ∼15 µm.

The thermal conductivities of molten bismuth and tin were measured using the hot-disk method. A hot-disk sensor was made of molybdenum foil (thickness, 20 µm; radius of sensor region, 3.05 mm) cut in a conducting pattern and placed between two aluminum nitride plates (plate thickness, 0.05 mm). Aluminum nitride did not corrode because of its contact with a molten metal and thus the molybdenum foil was protected from the molten metal. The thermal conductivities of molten bismuth and tin were measured during a short-duration of microgravity using a 10 m drop tower, to confirm the thermal convection effect during the measurement. The thermal conductivities measured in normal gravity were found to be approximately equal to those measured during the microgravity (during microgravity, thermal convection is suppressed), up to 977 K. Moreover, at 1083 K, normal gravity results were found to be higher than microgravity results.

After the successful results obtained these last years, electron beam direct write (EBDW) lithography use for integrated circuit manufacturing has now been demonstrated. However, throughput and resolution capabilities need to be improved to push its interest for fast cycle production and advanced research and development applications. In this way, the process development needs good patterns dimensional accuracy, i.e., a better control of the proximity effects caused by back scattering electrons and others phenomenon. It exists several methods to provide a correction for these effects and the most commonly used is based on dose modulation. However it has been observed that this correction is not perfect and can significantly fail to accurately correct the smallest and most dense structures encountered in sub 65 nm design features. To continue reducing feature sizes a method to provide a complementary correction to the dose modulation solution is proposed. Based upon detailed characterization of the observed effects a rules correction scheme has been developed not dissimilar to the rule based corrections used in optical proximity correction (OPC). This rule based electron beam proximity correction, or R-EBPC, provides good results down to 40 nm, with improvements in critical dimension (CD) linearity, the isolated dense bias, line end shortening and the energy latitude. All of which leads to an improvement in the overall accuracy of the design, and furthermore an improvement in the process window.

We studied the characteristics of Mott scattering systematically by constructing a special chamber, where target film thickness and the position of electron detectors can be controlled by an engineering workstation using stepping motors. Using this instrument, key parameters such as effective Sherman function and intensity were measured as functions of the scattered electron energy, target film thickness, and electron scattering angle. In the range of the electron energies we studied, i.e., 60–120 keV, a lower energy is preferable to achieve a large figure of merit. The results are compared with previously reported simulations.

Droplet ejection by an electrospray using a prototype inkjet head equipped with a ring-shaped gate electrode was demonstrated. The observation of droplet ejection with a high-speed camera revealed that ejection behavior can be categorized into two modes: a spindle-mode ejection and a multispindle-mode ejection. The former operation makes it more possible than the latter operation for the droplet to traverse the gate and travel toward the substrate where it produces a pattern. It was also found that the mode of ejection could be determined by the frequency of the applied voltage and the gap between the inkjet head and the gate electrode. Furthermore, droplet ejection orbit could be precisely controlled by a double-gate operation. A second gate electrode was important for realizing this. The ejection using a needle-type inkjet head was also investigated.

The time dependence of airborne particle count in connected clean units with a feedback loop in a clean-unit system platform (CUSP) is studied. We have estimated the mutual conductance of air flow in connected CUSP units, and the conductance is confirmed to be irrespective of the sizes of airborne particles. The connected CUSP units can maintain a cleanliness as good as ISO class 1–2 even when they are set in a typical office room. The emission rate of physically adsorbed particles from the inner walls of our CUSP unit is found to be 0.012 m^-2 s^-1 for particles with diameters of 0.3 µm. The airborne particle count in a deactivated CUSP unit, under the condition that the neighboring CUSP unit is activated, remains only ∼5 times larger than that in the activated CUSP units even for particles with diameters of 0.1 µm. The CUSP can serve as a platform for cross-disciplinary fields such as nanostructure physics as well as nanobiotechnologies.

We present experimental results that show the importance of the direct measurement of the irradiance of 185 nm radiation from low-pressure mercury (Hg) lamps in vacuum ultraviolet (VUV)/O₃ surface dry cleaning. It was found that the contact angles of water droplets on glass substrates can be correlated with the accumulated irradiance of the 185 nm radiation, not that from the 254 nm radiation that is conventionally monitored for the dry cleaning process control. Because the diamond-based VUV sensors are sensitive only to the 185 nm radiation, they could be a useful monitoring tool for controlling the VUV dry cleaning process in which low-pressure Hg lamps are used.

The correlation dimension of edge turbulence associated with the density fluctuation in a mirror plasma has been analyzed. A low embedding dimension indicating a strange attractor in the phase space is found. A high correlation dimension is also found in the same plasma discharge, but it could not be determined for the embedding dimension below 15.

In this paper, we describe a longitudinally coupled double-mode surface acoustic wave (SAW) filter on a 45° X–Z Li₂B₄O₇ (LBO) substrate for 1 GHz radio frequency identification (RF-ID) systems in Japan. Since LBO has a low dielectric constant, the aperture of an interdigital transducer (IDT) on LBO tends to be wide, resulting in the input and output impedances of 50 Ω. Higher order transverse modes existing in such a wide-aperture transducer generate ripple responses in the pass band of the filter. Generally, to suppress such modes, IDTs are apodized. However, it appears that a conventional apodized structure generates a notch response in the pass band of the filter due to longitudinal modes. We analyze the longitudinal mode response theoretically using a longitudinal mode separation technique and propose a novel-shaped apodized structure, which we call the "wave-shaped apodized structure." This apodized structure completely suppresses both the notch response due to longitudinal modes and higher order transverse modes. By applying this structure, we successfully develop a practical 953 MHz SAW filter for the above-mentioned RF-ID systems.

We present a direct comparison between numerical simulation of wave propagation, performed through 28 volumes of trabecular bone, and the corresponding experimental data obtained on the same specimens. The volumes were reconstructed from high resolution synchrotron microtomography experiments and were used as the input geometry in a three-dimensional (3D) finite-difference simulation tool developed in our laboratory. The version of the simulation algorithm that was used accounts for propagation in both the saturating fluid and bone, and does not take absorption into account. This algorithm has been validated in a previous paper [Bossy et al.: Med. Biol. 50 (2005) 5545] for simulation of wave propagation through trabecular bone. Two quantitative ultrasound parameters were studied at 1 MHz for both simulated and experimental signals: the normalized slope of the frequency dependent attenuation coefficient (also called normalized broadband ultrasound attenuation (nBUA) in the medical field), and the phase velocity at the center frequency. We show that the simulated and experimental nBUA are in close agreement, especially for the high porosity specimens. For specimens with a low porosity (or a high solid volume fraction), the simulation systematically underestimate the experimentally observed nBUA. This result suggests that the relative contribution of scattering and absorption to nBUA may vary with the bone volume fraction. A linear relationship is found between experimental and simulated phase velocity. Simulated phase velocity is found to be slightly higher than the experimental one, but this may be explained by the choice of material properties used for the simulation.

Laser irradiation using a Nd³⁺:YAG laser (532 nm, 8 ns, 10 Hz) on an ethanol suspension of micrometer-sized pentacene crystals induced absorption spectral changes in the solution, as a result of the formation of pentacene nanoparticles. The nanoparticles were platelike crystals. The nanoparticles formed by irradiation with a fluence of 50 mJ/cm² for 30 min were 4–13 nm in height and 10–70 nm in width. The threshold laser fluence for the formation was determined to be ca. 20 mJ/cm², and irradiation with fluences above 80 mJ/cm² induced the decomposition of the pentacene nanoparticles. The nanoparticle size decreased with an increase in the laser fluence, which was reflected in the longest-wavelength peak position of the absorption bands of the pentacene nanoparticles. On the basis of these results, the mechanism of pentacene nanoparticle formation by laser irradiation is discussed in connection with those of vanadyl phthalocyanine and quinacridone systems reported by Masuhara and coworkers [J. Phys. Chem. A 106 (2002) 2135; Jpn. J. Appl. Phys. 42 (2003) 2725; Jpn. J. Appl. Phys. 45 (2006) 384].

Single-walled carbon nanotubes (SWNTs) were prepared by ethanol chemical vapor deposition (CVD), using Co and Mo as metal catalysts deposited on Si/SiO₂ substrates with and without an Al underlayer. The effects of Mo addition, catalyst particle size and the Al underlayer were discussed on the basis of the results of scanning electron microscopy (SEM), atomic force microscopy (AFM) and Raman measurement. In substrates with Si/SiO₂/Co/Mo structures, a Co layer with a thickness of 1 nm produced smaller catalyst particles and SWNTs with a thinner diameter and at a higher yield than a Co layer with a thickness of 2 nm. The addition of Mo to Co films in substrates was shown to have the effect of suppressing Co aggregation and keeping catalyst particles small, even with high-temperature CVD growth. It was also found that the introduction of an Al underlayer promoted SWNT growth because of the formation of Al₂O₃ clusters after heat treatment.

Barium zirconate titanate (BZT) is considered to be an important dielectric material for the fabrication of multilayer ceramic capacitors (MLCCs). The synthesis of nanocrystalline powders of high purity is the key to improving the performance of BZT-based ceramics. In this study, we developed Ba(Zr_0.15Ti_0.85)O₃ nano powders with a perovskite structure and a homogeneous granularity through the oxalate precipitation method. The average particle size of the powders can be controlled from 30 to 100 nm at different calcining temperatures. The sintering property of the powders is superior to that of powders prepared by the conventional solid-state method, such that the more compact ceramics with a smaller grain size can be obtained at a sintering temperature of 1250 °C. The dielectric property of the ferroelectric-powder-based ceramics was so excellent for the development of dielectric materials for base-metal electrode multilayer ceramic capacitors (BME MLCCs) with a large capacitance.

Using only a simplified hot-filament chemical vapor deposition (SHFCVD) that has been improved and modified by drawing upon Matsumoto's hot-filament chemical vapor deposition (HFCVD) technique, the growth regions of carbonaceous product, such as carbon nanofibers or nanotubes and simple carbon films were examined consistently using of identical catalysts and carbon sources under different growth conditions. Ethanol (C₂H₅OH) and Ni metal were chosen and used as the carbon source and catalyst, respectively. This was considered to be the first examination, although other growth methods to synthesize carbonaceous products exist. Experimental results showed that the products could possibly be used to cover all grown regions by only changing substrate temperature (reaction temperature) and flow rate of hydrogen gas for bubbling in a carbon source. A comparison with products obtained using other equipment was performed.

We have examined the device characteristics of solution-processed single-walled carbon nanotube (SWNT) transistors. By using an electrical breakdown, SWNT-field-effect transistors (FETs) exhibited an on/off ratio (I_on/I_off) of 10⁴ and a field-effect mobility of 3.6 cm² V^-1 s^-1 in air, which are comparable to those of other organic FETs. We investigated the detailed mechanism of carrier injection from electrode metals into SWNTs. From the temperature dependence of source–drain current, we evaluated the effective Schottky barrier height for holes to be 170 meV.

We have investigated a method of magnetically orienting FePt nanoparticles. Uniaxial magnetic crystals such as chemically ordered FePt can be oriented by an external magnetic field. We therefore attempted to prepare colloidal chemically ordered FePt nanoparticles, using a surface reduction and heat treatment. The heat treatment changed the crystal structure of the nanoparticles from disordered to partially ordered, sustaining the colloidal status. A film of the partially ordered nanoparticles was then annealed in a magnetic field. As a result, the annealed film became magnetically anisotropic; that is, both coercivity and magnetic squareness along the direction of the applied field during annealing were >30% larger than those along the perpendicular direction. Although the observed magnetic anisotropy differed from the ideal, we analyzed the resultant magnetic orientation status by applying switching field distributions (SFDs). The analysis implied that annealing in a magnetic field has great potential for orienting uniaxial magnetic nanoparticles.

Stability and atomic geometry of mono-, di-, and trivacancies in graphene sheets are studied by using first-principles calculations. We find that the atomic relaxation substantially contributes to the stability of the vacancies. The monovacancy is found to have a nonplanar structure, i.e., its symmetry is C_1h, while the ideal monovacancy has D_3h symmetry. The divacancy is found to have a 5-8-5 membered ring structure. The trivacancy is also found to have two five membered rings. The energetics of these vacancies are not explained by the conventional dangling-bond counting model, which does not include lattice relaxation. Our calculations show that the divacancy is very stable and is thus expected to be detected under some experimental conditions.

The sterilization of materials exposed to a pulsed laser through an optical fiber has been evaluated. The results show that microorganisms can be killed by multiple laser shots of 266 nm wavelength radiation but not by multiple laser shots of 355 or 532 nm wavelength radiation, indicating that there is a threshold lethal wavelength and dosage. An ecologically friendly and inexpensive method of UV sterilization is expected to be developed as a result.