Table of contents

Starting with events in Leo Esaki's student days, the evolutionary path of his research activities in Japan and the United States from 1947 to 1992 is presented in a narrative form, including the discovery of the Esaki tunnel diode at Tokyo in 1957, for which the 1973 Nobel Prize was awarded, and the invention of man-made superlattices and resonant tunnel diodes at New York in 1969, for which the 1998 Japan Prize was awarded, as well as the influence, expressed as "times cited", of his original papers.

Graphene is a two-dimensional material with a one-atom-thick layer of carbon. Since the first report of the excellent electrical properties of graphene in 2004, its unique physical properties have been attracting attention and research on the application of graphene to electronic and photonic devices has been intensively carried out. In this review, recent research trends in the application of graphene to electronic devices, particularly transistors and interconnects, and graphene formation techniques are examined. In addition, the technical issues to be addressed for its application to electronic devices and the prospects for future graphene devices are discussed.

Power semiconductor devices are key components in power conversion systems. Silicon carbide (SiC) has received increasing attention as a wide-bandgap semiconductor suitable for high-voltage and low-loss power devices. Through recent progress in the crystal growth and process technology of SiC, the production of medium-voltage (600–1700 V) SiC Schottky barrier diodes (SBDs) and power metal–oxide–semiconductor field-effect transistors (MOSFETs) has started. However, basic understanding of the material properties, defect electronics, and the reliability of SiC devices is still poor. In this review paper, the features and present status of SiC power devices are briefly described. Then, several important aspects of the material science and device physics of SiC, such as impurity doping, extended and point defects, and the impact of such defects on device performance and reliability, are reviewed. Fundamental issues regarding SiC SBDs and power MOSFETs are also discussed.

The gas phase fragmentations of perfluoro-propyl-vinyl ether (PPVE, C₅F₁₀O) are studied experimentally. Dominant fragmentations of PPVE are found to be the result of a dissociative ionization reaction, i.e., CF₃⁺ via direct bond cleavage, and C₂F₃O⁻ and C₃F₇O⁻ via electron attachment. Regardless of the appearance energy of around 14.5 eV for the dissociative ionization of CF₃⁺, the observed ion efficiency for the CF₃⁺ ion was extremely large the order of 10⁻²⁰ cm⁻², compared with only 10⁻²¹ cm⁻² for the other channels. PPVE characteristically generated CF₃⁺ as the largest abundant ion are advantageous for use of feedstock gases in plasma etching processes.

One of the key problems in the synthesis of group III nitrides is the development of an effective source of reactive nitrogen, which is necessary for incorporation into the crystal lattice. In this paper, we describe a source of reactive nitrogen based on ECR discharge plasma, which is sustained by technological gyrotron radiation. Measurements of atomic nitrogen flux were conducted by mass-spectrometric analysis of the products of the reaction of nitrogen monoxide with atomic nitrogen. The maximal atomic nitrogen flow was 5 × 10¹⁸ atoms/s (12 sccm). The atomic source was used to grow InN films.

Etching rates of silicon nitrides (SiN), SiO₂, and poly-Si films for CH₂F₂ plasmas diluted with rare gases are presented by comparing the effects of flow rates of CH₂F₂ and dilution gases (Ar and Kr). The SiO₂ etching rate was considered to be controlled by ion fluxes of the incident CHF₂⁺ and CH₂F⁺ under the conditions for the selective etching of SiO₂ and SiN over poly-Si. Interestingly, the SiN etching rate was considerably affected by the dilution gas used. The SiN surface reaction was promoted by F-rich chemistry in the Ar-diluted CH₂F₂ plasma with a relatively high density of F atoms.

The formation mechanism of N–H-related defects in GaAsN grown by chemical beam epitaxy (CBE) is studied on the basis of the isotope effects on the local vibration modes (LVMs) originating from N–H. When deuterated monomethylhydrazine (MMHy) is used as the N source, LVM signals from the nitrogen–deuterium bond (N–D) are obtained. However, there are still N–H peaks in the IR absorption spectra, which have intensities similar to those of N–D peaks. When the film is grown with deuterated triethylgallium (TEGa), there are no N–D peaks. The peak intensity at 2952 cm⁻¹ increases with increasing tris(dimethylamino)arsenic (TDMAAs) flow rate, and that at 3098 cm⁻¹ is almost constant regardless of the flow rate. These results indicate that H atoms in the N–H-related defects originate from H directly bonded to N in MMHy and CH₃ in MMHy and/or H in TDMAAs, not from TEGa.

We studied the roles of lightly doped carbon in a series of n-GaN Schottky diode epitaxial structures on freestanding GaN substrates, and evaluated the effects of the doping on diode performances. A large variation of compensation ratio was observed for carbon doping at (1–2) × 10¹⁶ cm⁻³. A model was proposed to explain this phenomenon, in which a vulnerable balance between donor-type C_Ga and deep acceptor C_N strongly affected the free-carrier generation. Application of Norde plots and reverse biased leakage current in current–voltage measurements suggested provisional optimization for a free-carrier concentration of 8 × 10¹⁵ cm⁻³ to achieve a tradeoff between breakdown voltage and on-resistance of the n-GaN diodes.

An indium oxide transparent electrode for organic light-emitting diodes was fabricated by the inexpensive spray chemical vapor deposition method. The high work function (5.1 eV) necessary for a transparent anode and a hole-injection layer was successfully achieved with a vanadium doping concentration of 1.5 at. % V without any significant increase in resistivity and surface roughness or loss of transparency. The effect of vanadium doping on indium oxide was systematically investigated. The resistivity, average transmittance in the visible range, and surface roughness (R_a) were 1.08 × 10⁻³ Ω·cm, 84%, and 4.0 nm, respectively, for the vanadium-doped indium oxide.

The thermal stability of β-Ga₂O₃(010) substrates was investigated at atmospheric pressure between 250 and 1450 °C in a flow of either N₂ or a mixture of H₂ and N₂ using a radio-frequency induction furnace. The β-Ga₂O₃ surface was found to decompose at and above 1150 °C in N₂, while the decomposition of β-Ga₂O₃ began at only 350 °C in the presence of H₂. Heating β-Ga₂O₃ substrates in gas flows containing different molar fractions of H₂ demonstrated that the decomposition was promoted by increasing the H₂ molar fractions. Thermodynamic analysis showed that the dominant reactions are $\text{Ga}_{2}\text{O}_{3}\text{(s)} = \text{Ga}_{2}\text{O(g)} + \text{O}_{2}\text{(g)}$ in N₂ and $\text{Ga}_{2}\text{O}_{3}\text{(s)} + \text{2H}_{2}\text{(g)} = \text{Ga}_{2}\text{O(g)} + 2\text{H}_{2}\text{O(g)}$ in a mixed flow of H₂ and N₂. The second-order reaction with respect to H₂ determined for the mixed flows agrees with the experimental results for the dependence of the β-Ga₂O₃ decomposition rates on the H₂ molar fraction.

Field-effect transistors (FETs) with c-axis-aligned crystalline In–Ga–Zn–O (CAAC-IGZO) active layers have extremely low off-state leakage current. Exploiting this feature, we investigated the application of CAAC-IGZO FETs to LSI memories. A high on-state current is required for the high-speed operation of these LSI memories. The field-effect mobility μ_FE of a CAAC-IGZO FET is relatively low compared with the electron mobility of single-crystal Si (sc-Si). In this study, we measured and calculated the channel length L dependence of μ_FE for CAAC-IGZO and sc-Si FETs. For CAAC-IGZO FETs, μ_FE remains almost constant, particularly when L is longer than 0.3 µm, whereas that of sc-Si FETs decreases markedly as L shortens. Thus, the μ_FE difference between both FET types is reduced by miniaturization. This difference in μ_FE behavior is attributed to the different susceptibilities of electrons to phonon scattering. On the basis of this result and the extremely low off-state leakage current of CAAC-IGZO FETs, we expect high-speed LSI memories with low power consumption.

By using first-principles calculations, we study the formation energy and concentration of the silicon monovacancy. We use large-scale supercells containing up to 1728 atomic sites and confirm the convergence of calculational results with respect to the cell size. The formation energy is calculated to be 3.46 eV, and the vacancy concentration at the silicon melting point is estimated to be 7.4 × 10¹⁶ cm⁻³. These values are consistent with experimental results. We find that the vibrational effect significantly increases the vacancy concentration about 10⁴ times.

Electrical behaviors of amorphous carbon nitride (a-CN_x) films following exposure to various gas species were investigated. The a-CN_x film was prepared by reactive RF magnetron sputtering at 873 K. The resistivity of the film was measured in flowing He, N₂, O₂, Ar, and CO₂. Regardless of gas species, the electrical resistivity of the a-CN_x film dropped rapidly following gas injection and recovered in the evacuation process. This phenomenon is closely related to the changes in the film microstructure and electrical state. The response time in the adsorption process decreased with increasing gas pressure. The variation range of the electrical resistivity changed depending on the gas species.

The hydrogen content (C_H) and microstructure of a hydrogenated amorphous-silicon (a-Si:H) layer fabricated by plasma-enhanced chemical vapor deposition (PECVD) were analyzed to determine the effect of surface passivation on crystalline silicon (c-Si). The ratio of radio-frequency power to gas pressure (C_pp in W·Pa⁻¹) of the PECVD system is defined as a characterization parameter. It was found that C_H and the passivation of a-Si:H layers were sensitively affected by C_pp. However, C_H remained almost constant at the same C_pp even though the PECVD power and pressure were widely varied. We determined that an optimal region exists in the range of 0.75 < C_pp < 1.25, where a high implied open-circuit voltage of 732 mV was obtained from a passivated a-Si:H/c-Si/a-Si:H structure, indicating that C_pp was a useful parameter for characterizing the surface passivation effect of a a-Si:H layer on c-Si.

Ripplon spectroscopy was used to observe the propagation modes of surface waves on a layered structure. We experimentally prepared a water surface covered with an oil layer of µm-order thickness, on which two surface modes, bending and squeezing, were theoretically predicted to stand independently. Thermally excited capillary waves, called ripplons, propagating on the layered surface were observed with an optical beating light-scattering system developed by us and two modes were clearly observed as the doublet peaks of the Stokes and anti-Stokes components, respectively. The dispersion and the damping constant of ripplons of the two modes as well as their intensity ratio agreed well with the theoretical calculations. The dispersion of the mechanically excited surface waves, which is attributed to the bending mode, was also measured to examine the mode assignment in the light-scattering experiment.

A compact, novel photonic crystal cavity aimed at applications with strict area limitations is presented. Optimization shows that the gentle confinement method previously used for line-defect cavities can be applied to more limited geometries. It also shows that it is paramount to consider the boundary region to minimize in-plane losses. The investigation show that a near optimum boundary thickness can be found by considering the boundary region as a Fabry–Pérot resonator. This optimization strategy is shown to be deterministic in terms of resonance wavelength. For an optimized air-clad, silicon cavity, finite-difference time-domain simulations give Q-values as high as 75,000 which is comparable to other photonic crystal cavities of similar size.

We investigated the shape effects of GeSbTe nanodots on the near-field interaction with a silver triangle antenna using the three-dimensional finite-difference time-domain method, avoiding the difficulty of detecting near-field signals from a single dot that occurs in current measurements. The surface plasmon resonance of silver strengthens the near-field around nanodots made of GeSbTe, commonly used in phase-change recording. Using GeSbTe spheres and pillar dots with various top plane shapes, we investigated the relationship between the inner electric field concentration of GeSbTe nanodots and the radius of curvature of the corners facing the antenna tip. Reducing the radius of curvature strengthens the inner electric field of the dots, enhancing the near-field difference in intensity for the GeSbTe phase change. GeSbTe diamond pillars with a radius of curvature of 1 nm exhibit a near-field difference in intensity of 28% for the phase change. Using the antenna and the GeSbTe nanodot array, optical write-once recording is realized. The preliminary research in this study is expected to realize future optical disk storage using GeSbTe nanodots with diameters below 10 nm.

We investigate inductively coupled plasma (ICP) deep dry etching of Al_0.8Ga_0.2As for photonic crystal (PC) fabrication using a Cl₂/BCl₃/CH₄ gas mixture. On the basis of our previous report [Y. Kitabayashi et al., Jpn. J. Appl. Phys. 52, 04CG07 (2013)], we obtained a PC structure having air holes deeper than 1.5 µm and a diameter of 120 nm by adjusting the gas flow rate and increasing the process pressure. In this study, silicon nitride (SiN_x) and SiO₂ were both used as the mask layer. Furthermore, samples with SiN_x and SiO₂ masks for ICP deep dry etching were also fabricated and compared. The vertical profile of the PC structure with the SiN_x mask layer displayed a rounded shape that was caused by the charge up in the mask layer. Then, a thinner mask layer was used to ease the effects of mask retardation and charge up. As a result, a PC structure with a SiN_x mask layer having air holes deeper than 1.7 µm and a diameter of 190 nm was successfully fabricated.

A solution-processing method for obtaining graphite oxide (GO), which is used for the organic light-emitting diodes (OLEDs), is presented herein. The use of poly(arylene ether)s as the solid carbon source is demonstrated for the first time. Raman spectroscopy, Fourier transform infrared (FTIR), X-ray photoelectron spectroscopy (XPS), luminance, and electroluminescence (EL) reveal the essential properties. The respective percentages of the peak areas of sp² C and sp³ C in the GO are calculated to be 54.47 and 37.32%. The GO had a sheet resistance of 97 Ω/square determined by the four-probe method and a conductivity of 5.538 × 10¹ Ω⁻¹ cm⁻¹ determined by the Hall-effect method.

We investigated the effect of step-down indium content in InGaN quantum wells (QWs) on the output efficiency of fully packaged InGaN-based light-emitting diodes (LEDs). Both the reference and step-down LED chips give maximum external quantum efficiencies (EQE) of 54.65 and 54.99% at a current density of 4.17 A/cm², respectively. Step-down LEDs show a lower efficiency droop than reference LEDs. The step-down LEDs exhibit a 5.3% higher EQE at 83.3 A/cm² than the reference LEDs. As the current density increases from 1.39 to 9.03 A/cm², the electroluminescence (EL) intensity peaks of the step-down LEDs are slightly more shifted towards the larger energy side than those of the reference LEDs. The polarization field is estimated to be 1.34 and 1.41 MV/cm for the reference and step-down LEDS, respectively. The simulated internal quantum efficiency results of the reference and step-down LEDs are in agreement with the experimental results. The simulation results show that the step-down LEDs have higher hole injection efficiency than the reference LEDs. On the basis of the simulation results, the blue-shift behavior of the reference and step-down LEDs is described and discussed.

In this work, an aqueous solution method that entails processing at low temperatures is utilized to deposit a ZnO interlayer in poly(3-hexylthiophene-2,5-diyl):[6,6]-phenyl C₆₁ butyric acid methyl ester-based inverted polymer solar cells (PSCs). The effect of ZnO annealing temperature from 50 to 150 °C on PSC performance is systemically studied and it is found that the transition point is approximately 80 °C. When the ZnO annealing temperature is higher than 80 °C, PSCs show similar current density–voltage (J–V) characteristics and achieve a power conversion efficiency higher than 3.5%. Transmittance spectrum, PL spectrum, and surface morphology studies show that an annealing temperature above 80 °C is sufficient for ZnO to achieve a relatively good quality, and that a higher temperature only slightly improves ZnO quality, which is confirmed from statistical results. Furthermore, flexible PSCs based on PET substrates show a comparable power conversion efficiency and good flexibility.

Cu₂ZnSn(S,Se)₄ (CZTSSe) solar cells are fabricated with CdS and (Cd,Zn)S buffer layers of different thicknesses to investigate sputtering damage on the absorber surfaces during ZnO:Al/ZnO layer deposition. In this work, the sputtering damage is scrutinized by photoluminescence (PL) measurement. The damage (i.e., non-radiative recombination centers) near the absorber surfaces investigated on the basis of PL peak intensity decreases with increasing thickness of the buffer layer. Furthermore, the intensity in the solar cells with the CdS buffer is higher than that with the (Cd,Zn)S buffer layer, suggesting that the CdS buffer layer demonstrates better capability of preventing sputtering damage near the CZTSSe surface than the (Cd,Zn)S buffer. PL ratio defined as the ratio of the PL peak intensity after sputtering to the PL peak intensity before sputtering is utilized to quantify sputtering damage. The solar cell performance increases with increasing PL ratio up to 0.5, followed by saturation at a ratio higher than 0.5. Taken together, PL ratio is proposed as a tool for monitoring sputtering damage for improving cell performance.

Dual relaxation processes have been observed in a straightforward way by photon correlation spectroscopy (PCS) for polarization clusters in barium titanate during paraelectric–ferroelecric phase transition. The observed fast relaxation process for polarization clusters covers a broad temperature region and exhibits an anomaly near the Curie temperature (T_C). The slow relaxation process occurs at 4 K above T_C and covers a limited temperature region, where the relaxation time exhibits an approximately linearly decreasing behavior. This study provides deep insight into the complex phase transition of barium titanate.

The aluminum concentration dependence of the energies of the direct and indirect bandgaps arising from the Γ and X conduction bands are measured at 1.7 K in the semiconductor alloy Al_xGa_1−xAs. The composition at which the bands cross is determined from photoluminescence of samples grown by molecular-beam epitaxy very close to crossover at x ≈ 0.4. The use of resonant laser excitation and the improved sample linewidth allows excitation intensities as low as 10⁻² W/cm², giving a precise determination of the bound exciton transition energies and their Γ and X crossover. Photoluminescence excitation spectroscopy is then used to measure the binding energies of the donor-bound excitons and the Γ free exciton binding energy. After correcting for the Γ- and X-dependence of these quantities, the crossover of the bandgap is determined to be at x = 0.401 and E = 2.086 eV.

Manipulation of the optical property of fluorescent probes has been a powerful strategy to establish super-resolution microscopy. We describe a new strategy to realize a probe with a nonlinear fluorescence response by using photoinduced charge separation. In this scheme, the first photon is used for the generation of the charge-separated state and the second photon is for fluorescence excitation. This stepwise two-photon absorption was confirmed by detection of a second-order nonlinear fluorescence response. Transient absorption spectra studies and simulation indicate that fluorescence is emitted through the photophysical pathways we proposed. Fluorescence imaging of biological cells showed marked improvements in image contrast and resolution, demonstrating the usefulness of the fluorescent probe in laser scanning confocal microscopy.

We present a single-shot approach for separating optical direct and global components from an object. The former component is caused by direct illumination that travels from a light source to a point on the object and goes back to a camera directly. The latter one is caused by indirect illumination that travels from the light source to a point on the object through other points and goes back to the camera, such as multi-path reflection, diffusion, and scattering, or from another unintended light source, such as ambient illumination. In this method, the direct component is modulated by a single coded pattern from a projector. The modulated direct and un-modulated global components are integrated on an image sensor, which captures a single image. These two components are separated from the single captured image with a numerical algorithm employing a sparsity constraint. Ambient light separation and descattering based on the proposed scheme are experimentally demonstrated.

For the realization of on-chip optical interconnects, light sources enabling ultralow power consumption and high-efficiency operation are required. With this aim, we fabricated lateral-current-injection-type membrane Fabry–Perot lasers with a threshold current of 3.5 mA and an external differential quantum efficiency of 11% under a room temperature-continuous wave (RT-CW) condition. To the best of our knowledge, we experimentally evaluated the thermal properties of a membrane laser for the first time. From the measurement, we obtained a thermal resistance of 330 K/W, which well agreed with the theoretical value of 340 K/W. From the theoretical analysis, it was found that a reduction of the benzocyclobutene thickness was effective for reducing the thermal resistance of the membrane laser. Finally, we determined that the increase in thermal resistance for short cavity (less than 50 µm) devices is not a problem because self-heating is small for low operation current.

A high power, 6 kHz, single-mode Ti:sapphire laser operating at 904 nm has been developed to produce a 193 nm light source. The output power was above 10 W with a bandwidth of 160 MHz. The Hänsch–Couillaud locking scheme was successfully applied to stabilize the frequency of the pulse laser. The thermal lens in the Ti:sapphire crystal having a focal length down to 10 mm along with strong astigmatism was compensated by distributing thermal load to three amplifiers with an even number of passes, resulting in a nearly diffraction limited beam. This Ti:sapphire laser contributed to the generation of 193 nm light with an output power above 200 mW.

We carry out first principles calculations to study the magnetic properties of a multilayer structure consisting of one monolayer (ML) of Co (Co_1ML) on top of two ML of Ni (Ni_2ML) deposited on top of Cu(111), i.e., Co_1ML/Ni_2ML/Cu(111). We found that the magnetic anisotropy of the Co_1ML/Ni_2ML multilayer does not change much with variation in the in-plane and inter-layer structures of the Co/Ni multilayer. On the other hand, interaction with the Cu(111) plays an important role. As a result of the electron transfer from the Cu(111) substrate to the first deposited Ni-layer, i.e., Ni(1), there is a reduction in the magnetic moment of Ni(1) and an increase in the magnetic crystalline anisotropy energy of Co_1ML/Ni_2ML/Cu(111) multilayer structure.

The magnetic field generated by cryogenic PdNi ferromagnetic patterns was controlled at 4.2 K to fabricate superconducting phase-shift elements (PSEs) that would lead to production of energy-efficient devices. Cryogenic ferromagnetic patterns of Pd_0.89Ni_0.11 were magnetized by field cooling (FC) involving the application of external magnetic fields during sample cooling. The magnetic field from the PdNi patterns was evaluated by examining rectangular Josephson junctions in the vicinity of the patterns. Our experimental results show that the strength of the field induced by the PdNi patterns was nearly proportional to the applied field strength during FC. The field strength also depended on the end temperature of FC.

We have investigated a Te/Se substitution effect on the field dependence of Hall resistivity in FeSe_1−xTe_x (x = 0.5–0.7) thin films grown on LaAlO₃ and CaF₂ substrates. By observing the magnetic field dependence of Hall resistivity, the crossover from hole- to electron-dominant regions is observed to occur between x = 0.5 and 0.6 in the films on LaAlO₃, while no such crossover is observed in those on CaF₂. The results indicate that the substitution of Te for Se effectively acts as electron doping, while the lattice strain also has an additional effect on the balance of hole and electron densities. These two factors can be independently used to optimize superconducting transition temperature so as to tune the doping level near the boundary where the hole density becomes equal to the electron density.

Threshold voltage drift under gate bias stress was investigated in gate-recessed enhancement mode (E-mode) GaN MOSFET and depletion mode (D-mode) GaN MOS high-electron-mobility transistor (MOSHEMT) with Al₂O₃ gate dielectric layer. Besides the positive shift of threshold voltage in both devices under positive gate stress, it is also found that positive shift could also exist in E-mode GaN MOSFET under negative gate bias stress, while negative shift is observed in D-mode MOSHEMT. A three-step trapping and detrapping process was observed in the drain current transient of the device after negative gate bias stress. It was suggested that gate electron injection and the following trapping in the "damaged" gate recessed GaN channel layer is the dominant mechanism for the positive shift of the threshold voltage under negative gate bias in the enhancement mode GaN MOSFET.

The ideal drain electrode assumed in the analysis of MOSFETs accepts carriers from the channel without reflection. The realistic drain, however, backscatters some of the carriers into the channel once again, and the backscattering is intensified if elastic scattering is dominant within the drain. The backscattering strongly affects the ballistic performance of MOSFETs. The channel length dependence of the ballistic performance of MOSFETs is explored in consideration of the backscattering. It is concluded that the ballistic limit of MOSFETs is not realized in the short-channel limit of the device, but, on the contrary, in the comparatively long-channel device in which minimization of the elastic scattering is pursued. The tendency may possibly explain the observed result that the performance of a 0.1 µm MOSFET reported a quarter-century ago showed a good agreement with the ballistic MOSFET characteristics.

We systematically study the channel size dependence of drain current noise amplitude in tri-gate silicon nanowire transistors (NW Tr.) by measuring a large number of samples with various parameters such as gate length (L_g), NW width (W_NW), and NW height (H_NW) down to 10 nm. The noise of NW Tr. with W_NW < 20 nm rapidly increases with 1/W_NW because an electron in a single trap on the NW channel blocks the entire current flow (bottleneck effects). The strength of bottleneck effects is determined solely by W_NW not by H_NW. The noise increase at narrow W_NW is smaller in short-L_g Tr. (down to 15 nm).

High-resolution silicon X-ray detectors with a large active area are required for effectively detecting traces of hazardous elements in food and soil through the measurement of the energies and counts of X-ray fluorescence photons radially emitted from these elements. The thicknesses and areas of commercial silicon drift detectors (SDDs) are up to 0.5 mm and 1.5 cm², respectively. We describe 1.5-mm-thick gated SDDs (GSDDs) that can detect photons with energies up to 50 keV. We simulated the electric potential distributions in GSDDs with a Si thickness of 1.5 mm and areas from 0.18 to 168 cm² at a single high reverse bias. The area of a GSDD could be enlarged simply by increasing all the gate widths by the same multiple, and the capacitance of the GSDD remained small and its X-ray count rate remained high.

Semiconducting, octadecylamine (ODA)-capped, rod-like FeS₂ nanocrystals (NCs) are prepared and their properties are characterized. Transmission electron microscopy studies revealed the lattice fringes were clearly visible, confirming that the rod-like FeS₂ NCs obtained in this study were highly crystalline. X-ray diffraction (XRD) studies indicated that the structure of FeS₂ NCs is in cubic phase. Optical properties showed that the FeS₂ NCs can be a promising candidate for the electron acceptor material in the bulk hetero-junction (BHJ) solar cell applications. The power conversion efficiency of the BHJ solar cells using polymer and the rod-like FeS₂ NCs with a structure of glass/ITO/PEDOT:PSS/(FeS₂+polymer)/Al was found to be ∼0.45%.

Pulsed wire discharge (PWD) is one of the many nanoparticle preparation methods known for its high efficiency and high production rate. Particle size is controlled by gas pressure and input energy. However, the effect of ambient gas species on particle size is not well known. In this study, single-phase palladium (Pd) nanoparticles were prepared in N₂, Ar, and He ambient gasses by PWD. The mean diameter of the prepared particles in Ar was smaller than those of the other nanoparticles because of the low ionization energy required to form an alternate current path and the partial wire evaporation. A novel equation for estimating the average grain size was proposed. By comparing the estimated and measured mean diameters, the validity of the equation was shown.

We have successfully improved the weight density of a 40-µm-long carbon nanotube (CNT) to 0.27 g/cm³ by improving the slope control of temperature profile (STEP) growth. It was found that the CNT growth is reaction-limited in the early stage of STEP growth and supply-limited in the late stage. We succeeded in doubling the CNT density over the conventional method by optimizing the raw material supply and temperature slope during the reaction-limited and supply-limited periods. A CNT thermal interface material was fabricated using a CNT film prepared by this method. Thermal resistance was 11% below a conventional indium thermal interface material. Owing to the above reasons, CNTs show promise as heat dissipation materials.

An amorphous TiO₂ film (180 nm) was deposited as a protective layer on the surface of a triple-layer thin-film switchable mirror (Pd/Ti/Mg₄Ni deposited on glass) by a sol–gel coating process, and its optical switching behavior and switching lifetime under 4% hydrogen gas loading were evaluated. The use of a TiO₂ coating extended the switching durability to about 1600 cycles, which is a fourfold increase compared with that of uncoated mirrors. The switching response of the Pd/Ti/Mg₄Ni thin film was not affected by the presence of the TiO₂ film, with hydrogenation and dehydrogenation speeds being almost the same as those of uncoated mirrors. The optical properties of the TiO₂-coated mirrors were improved in the hydrogenated state, and a diffuse reflection phenomenon was observed in the dehydrogenated state.

We designed a zinc oxide (ZnO) growth device using a cylindrical piezoelectric lead zirconate titanate (PZT) vibrator. This growth device can be employed to realize a simpler and more efficient synthesis method using focused ultrasound. To confirm the effectiveness of the device, the distribution and nanostructure size were investigated under different ultrasonic powers and precursor concentrations. The size of the synthesized nanostructure was proportional to the precursor concentration and ultrasound intensity. The chemical effect of the ultrasound is more dominant than the mechanical one when the ultrasonic cavitation is generated. According to the results of the X-ray diffraction pattern and distribution, the synthesized nanostructure showed a high crystallinity and a populous distribution at the target area.

Indium tin oxide (ITO) thin films have been deposited by pulsed laser deposition on m-plane (100) and r-plane (012) sapphire substrates. For both substrates, the films were grown with their [110] direction perpendicular to the substrate planes under the conditions of high growth temperature and high oxygen pressure. Their in-plane epitaxial relations with the substrates were identified to be ITO[001] ∥ Al₂O₃[020] and $\text{ITO}[1\bar{1}0]\parallel \text{Al}_{2}\text{O}_{3}[001]$ for the m-plane substrate. For the r-plane substrate, two types of lattice matching were observed: one being $\text{ITO}[001]\parallel \text{Al}_{2}\text{O}_{3}[2,1, - 1/2]$ and $\text{ITO}[1\bar{1}0]\parallel \text{Al}_{2}\text{O}_{3}[4/3, - 4/3,2/3]$, the other being $\text{ITO}[001]\parallel \text{Al}_{2}\text{O}_{3}[1, - 1,1/2]$ and $\text{ITO}[1\bar{1}0]/\text{Al}_{2}\text{O}_{3}[8/3,4/3, - 2/3]$. The electrical properties were measured by the Hall effect and van der Pauw methods at room temperature. All of the samples have low electrical resistivity on the order of 3.0 × 10⁻⁴ Ω cm, high carrier concentration of about 2.5 × 10²⁰ cm⁻³, and mobility ranging from 70 to 90 cm² V⁻¹ s⁻¹.

A superhydrophobic miniature boat was fabricated with aluminum alloy plates treated with atmospheric-pressure helium (He)/octafluorocyclobutane (C₄F₈) plasma using 13.56 MHz rf power. When only 0.13% C₄F₈ was added to He gas, the contact angle of the surface increased to 140° and the surface showed superhydrophobic properties. On the basis of chemical and morphological analyses, fluorinated functional groups (CF, CF₂, and CF₃) and nano-/micro-sized particles were detected on the Al surface. These features brought about superhydrophobicity similar to the lotus effect. While the miniature boat, assembled with plasma-treated plates, was immersed in water, a layer of air (i.e., a plastron) surrounded the superhydrophobic surfaces. This effect contributed to the development of a 4.7% increase in buoyancy. In addition, the superhydrophobic properties lasted for two months under the submerged condition. These results demonstrate that treatment with atmospheric-pressure He/C₄F₈ plasma is a promising method of improving the load capacity and antifouling properties, and reducing the friction of marine ships through a fast and low-cost superhydrophobic treatment process.

We report a new GaN etching technique with high anisotropy involving a thermal decomposition reaction in a low-pressure H₂ environment. A GaN microridge stripe structure (5 µm in width and 1.2 µm in height) with extremely smooth sidewalls was fabricated at 1,050 °C and a H₂ pressure of 10 Pa for 15 min using a SiO₂ mask. The activation energy of the vertical etching was calculated to be 62–77 kcal/mol. In the GaN nanoridge stripe structure, the side etching under the SiO₂ mask was less than 5 nm in depth and showed top width and height of ∼40 and ∼180 nm, respectively. The sidewall was extremely smooth and tilted by ∼15° from the m-plane along the a-axis, while being slightly rough and tilted by ∼30° from the a-plane along the m-axis. The $\{ n\bar{n}02\}$ ($n = 4,5,6,7$) planes were relatively stable in this etching technique.

We investigated a device isolation process for MoS₂-based devices and fabricated high-k/metal-gate MoS₂ MOSFETs. An Ar-ion etching process was utilized for the device isolation process. It circumvents damage in the device channel, as confirmed by Raman spectroscopy. A top-gate MoS₂ MOSFET was fabricated with a HfO₂ thin film 16 nm thick as the gate insulator. Utilizing capacitance–voltage (C–V) measurements, the capacitance equivalent thickness (CET) was estimated to be 5.36 nm, which indicates that a gate stack with the sufficiently thin insulator was successfully realized. The device exhibited a mobility of 25.3 cm²/(V·s), a subthreshold swing (SS) of 86.0 mV/decade, and an ON/OFF ratio of 10⁷. This satisfactory device performance demonstrates the feasibility of the proposed device isolation process.

The bending curvature, stresses, and stress relaxation of various multi-layered structures with different adhesive layers pertaining to the polarizer in a thin-film transistor liquid-crystal display (TFT-LCD) have been successfully characterized by using bending beam technique under reliability test. To be more specific, three different types of pressure-sensitive adhesive (hard-, middle-, and soft-type) and various poly(vinyl alcohol) (PVA) stretched directions are devised to examine to key stress contributors and correlations with light leakage. The shrinkage stress in stretched PVA film and stress relaxation ability of pressure-sensitive adhesives (PSA) layers are found to be the key factors determining the stress distribution and out-of-plane displacement of a polarizer stack. For hard-type PSA, its polarizer stack generates the highest bending curvature with maximum out-of-plane displacement but minimum in-plane displacement, leading to anisotropic stress distribution with high stress around the edges. On the other hand, polarizer stack with soft-type PSA yields the maximum in-plane displacement but the minimum out-of-plane displacement, resulting in isotropic stress distribution.

A double-line laser source as an irradiation pattern for the generation of ultrasonic waves is proposed. Simulated results show that the shear-vertical (SV) waves generated by the double-line laser source superimpose and produce a convergence effect at the center axis with a fixed angle. The amplitude of the converging SV wave is significantly improved compared to the ultrasonic field generated by a single-line source under the same condition of the thermoelastic regime. Simulations show good agreement with experimental results, which demonstrate that the SV waves in a steel plate subjected to the double-line laser source converge efficiently at the center axis. Then, a method using the converging SV to detect surface defect on a steel plate is proposed. The characterization of the surface defect has been quantitatively measured with high accuracy by the well-designed double-line source.

The theory of best focus determination in focus wedge mark (FWM) metrology is extended from coherent illumination to partially coherent illumination. Through a theoretical analysis, we propose a focus curve fitting function suitable for the FWM, which is expressed by the sinc function. The experimental results for best focus estimations using two types of fitting functions, which are a conventional quadratic polynomial and the proposed sinc function, are reported for various defocus regions to be fitted. When the first inflection points of the sinc function are included in the fitting defocus region, we can achieve stable best focus estimation within 3 nm in our i-line exposure experiments.

The electrochemical properties of lithium–oxygen batteries are improved by coating a tungsten carbide layer onto an oxygen cathode. Tungsten carbide (WC) coating is conducted by physical vapor deposition. The uniform deposition of WC onto the cathode is confirmed by scanning electron microscopy, field-emission transmission electron microscopy and electron probe microanalysis. The discharge-recharge voltage gap of the cell with a WC-coated electrode is estimated to be 0.88 V, which is 700 mV smaller than that of an electrode without the WC coating at 100 mA g⁻¹_carbon. The overpotentials of the WC-coated electrode remain unchanged even after the 10th cycles, while that of the uncoated electrode gradually increases. In addition, the WC-coated electrode exhibits a stable voltage profile, even at a current density of 200 mA g⁻¹_carbon. The observed improvement in cell performance can be attributed to the catalytic property and high electrical conductivity of WC. The enhanced electrochemical properties of cells are examined by impedance analysis and SEM.

In this work, a lift-off process with bi-layer photoresist patterns was applied to the formation of hydrophobic/hydrophilic micropatterns on practical polymer substrates used in healthcare diagnostic commercial products. The bi-layer photoresist patterns with undercut structures made it possible to peel the conformal-coated silicon oxide (SiO_x) films from substrates. SiO_x and silicon carbide (SiC_x) layers were deposited by pulsed plasma chemical vapor deposition (PPCVD) method which can form roughened surfaces to enhance hydrophilicity of SiO_x and hydrophobicity of SiC_x. Microfluidic applications using hydrophobic/hydrophilic patterns were also demonstrated on low-cost substrates such as poly(ethylene terephthalate) (PET) and paper films.

A high-density through-wafer vertical copper via array in insulating glass interposer is demonstrated. The glass reflow and bottom-up filling copper electroplating process enables fabrication of a vertical through-wafer copper via array with high aspect ratio and high density. The minimum diameter of the copper vias and the gaps in between are 20 and 10 µm, respectively. Three failures among one hundred measurement points were detected within a 1 cm² area of the via array, and the resistance of the 20-µm-diameter copper via was measured to be 153 ± 23 mΩ using the four-probe method. The optical transmittance and RF performance of the reflowed glass substrate were compared with those of bare glass. The reliability of the copper via in harsh environments was evaluated through thermal shock and pressure cooker tests.

We report magnetization of a non-centrosymmetric superconductor, LaPt₃Si below 0.6 K up to 200 Oe. The home-made SQUID and Hall sensor magnetometers that operate below 1 K were constructed for this purpose. Although the SQUID magnetometer is more sensitive than the Hall sensor's one, it was found not to work correctly for the rapid magnetization change of LaPt₃Si below 0.4 K. The Hall sensor magnetometer, in contrast, can properly detect magnetization jumps in the M–H curve of the superconducting state. The observed flux jumps are probably related to the interfusion of the mixed state of the LaPt₃Si that is observed in the μSR measurements.

RF plasmas in pure water (≃0.2 mS/m) can be generated within a hole in an insulating plate, apart from the electrodes. When the diameter of the hole is 3–10 mm, the plasmas can be maintained stably. The existence of bubbles in the hole is important for the generation of plasmas. In the results of spectroscopic measurements, only the lines from the atoms (O, H) and molecule (OH) derived from H₂O molecules were observed.