Search

Antireflection (AR) films are useful for preventing the reflection of light from flat-panel displays, solar-cell panels, optical lenses, and similar structures. We previously reported the successful high-throughput fabrication of an AR structure (ARS) with less than 0.1% reflectivity by means of ultraviolet nanoimprint lithography (UV-NIL). However, this process tended to produce a thick residual layer that reduced transmittance and produced interfacial reflections at surfaces of lenses coated with the ARS. To reduce interfacial reflections, the thickness of the residual layer must be less than 100nm or so (roughly a quarter of the wavelength of visible light). Liquid transfer imprint lithography (LTIL) is an effective technique for reducing the thickness of the residual layer. In the LTIL process, residual resin is eliminated by separating the mold and substrate in the liquid phase. This process is also effective for curved surfaces, and it has been reported an ARS with a thin residual layer can be obtained by LTIL. However, the method that was used required a vacuum process and was time consuming. To obtain a thin residual layer at atmospheric pressure, we devised an LTIL process involving a roll-press method that does not require the presence of a vacuum. The technique permitted the fabrication of a high-performance ARS with improved reflectance and transmittance thickness, in which the thickness of the residual layer was less than a quarter of the wavelength of visible light. Furthermore, the technique could be used to produce a homogeneous ARS with a residual layer less than 100nm thick on a curved surface, such as that of a lens, in a reproducible and stable manner. The process is therefore suitable for fabricating ARSs that show a high performance with a thin residual layer on curved surfaces such as convex lenses. Display Omitted We devised an LTIL with roll-press method that does not require a vacuum process.An ARS with a residual layer less than 100-nm thick was homogeneously formed over the curved surface.The reflectance and transmittance was improved by fabricating an ARS at lens surface.

Carbon-based materials, such as carbon nanotubes and graphene nanoribbons, have been studied as interconnect materials because of their lower resistivity, higher thermal conductivity, and intrinsically higher current carrying capacity compared to Cu. We have succeeded in growing networked-nanographite (NNG) films on dielectric layers without any metal catalysts, and fabricated NNG wires as the first step toward multi-layer graphene interconnects. In this study, we investigated the electrical properties and reliability, such as current tolerance and thermal stability, of these NNG wires. In addition, we investigated the mechanism of resistivity decreases in NNG. Our results show that the NNG wires can sustain a higher current density than Cu wires. However, their resistivity gradually decreases due to thermal stress. Raman and X-ray photoelectron spectroscopy revealed that the decrease in resistivity can be attributed to an increase in sp2 bonding corresponding to the formation of a graphene sheet. It is important to clarify the mechanism of the decrease in resistivity, not only to improve thermal stability but also to obtain lower resistivity in carbon interconnects.

Selective synthesis of carbon nanotubes (CNTs) and graphene was achieved by hot-filament chemical vapor deposition (CVD) using iron films as a catalyst at a temperature of 620 °C. The selective synthesis was performed in a very simple way, just by changing the catalyst thickness under the same CVD conditions. Only CNTs were obtained with Fe films of 5 nm and less, while multi-layer graphene was formed with the catalyst films of 20 nm and more. The thickness of multi-layer graphene can also be controlled by controlling the thickness of the catalyst further.

Multiwalled carbon nanotubes (MWCNTs) were synthesized at 390 °C in a hydrogen-free atmosphere by graphite antenna chemical vapor deposition, which provides carbon radicals by the etching of the antenna itself in a He- or Ar-based noble gas plasma. An increase in the number of defects in the graphite layers of the CNTs was observed with increasing hydrogen partial pressure. The results of X-ray photoelectron spectroscopy analysis suggested that these defects were caused by the transformation from an sp 2 to an sp 3 structure in the graphite layers of CNTs due to hydrogen radicals.

Bimetallic particles of Ti and Co have been used to grow multi-walled carbon nanotubes (MWNTs) to study the effects of Ti on the growth. The particles were generated by laser ablation and size-classified with a differential mobility analyzer. Diameter-controlled MWNTs were grown on a silicon substrate from 5-nm particles by thermal chemical vapor deposition using acetylene at 510 and 610 °C. Titanium has been found to enhance the MWNT growth significantly. There is an optimum Ti fraction for the growth, as confirmed by evaluation of the growth probabilities of MWNTs. A possible mechanism of growth enhancement by Ti is discussed.

Diameter-controlled multiwalled carbon nanotubes (MWNTs) have been grown by hot-filament chemical vapor deposition using nickel nanoparticles as catalyst. The nanoparticles were generated by laser ablation, classified with a differential mobility analyzer, and deposited onto silicon substrate. The particle size is tunable down to 2–3 nm, and particles with a geometric mean diameter of 5.1 nm (geometric standard deviation: 1.1) were used for carbon-nanotube growth. MWNTs were grown on the substrate using acetylene at 550 °C. The particles did not coalesce during growth, and the MWNTs had outer diameters matching the particle sizes, indicating that the current method can produce diameter-controlled MWNTs.

We have succeeded in making peapod structures with vertically aligned multi-walled carbon nanotubes (MWNTs), grown on a silicon substrate using hot filament chemical vapor deposition. Opening the tips of the MWNTs was performed at 600 ∘C using oxygen and then C60 molecules were deposited on them to fill their inner spaces. We confirmed that the MWNTs were open at the tips and filled with C60 molecules using transmission electron microscopy. In addition, it has been determined that MWNTs can be filled with other materials, such as molybdenum oxide.

We demonstrated that the thinning of InGaP electron supply layers is very effective in reducing short channel effects, even for 0.065 μm gate InGaP/InGaAs/GaAs pseudomorphic HEMTs. By thinning the InGaP electron supply layer to 11.5 nm with a high doped density of 5 × 10 18 cm −3 , we were able to minimize the threshold voltage shift to less than 0.15 V compared to 1.65 μm gate length HEMTs. We could also increase the extrinsic transconductance ( g m ) by thinning the InGaP electron supply layer. We observed no degradation of gate leakage characteristics, regardless of such a highly-doped, thin InGaP layer. The 0.065 μm gate length HEMT exhibits an extrinsic transconductance ( g m ) of 585 mS mm −1 , a current gain cutoff frequency ( f τ ) of 110 GHz and a maximum oscillation frequency ( f max ) of 150 GHz.

During annealing in vacuum and dry O 2 , the amount of suboxide, Θ s , in the 1000 A SiO 2 films on Si(111) was measured as a function of the annealing time by angle-resolved X-ray photoelectron spectroscopy (ARXPS). From ARXPS spectra of Si 2p taken at two polar angles of 0° (bulk sensitive) and 75° (surface sensitive), it was clarified that during annealing in vacuum, Θ s increases with time underneath the surface more largely than on the surface, but annealing in O 2 atmosphere at 0.5 Torr reduces Θ s on the surface more significantly than underneath the surface. These indicate that SiO molecules outdiffusing from the SiO 2 /Si interface are oxidized by permeating O 2 in the SiO 2 films.

The thermal oxidation kinetics of Si with dry oxygen was investigated by X-ray and ultraviolet photoelectron spectroscopy. The SiO desorption was found to be maintained for a while even after the oxide growth during initial oxidation. By the annealing of thick oxides in O 2 , the oxidation of SiO molecules in SiO 2 was verified more straightforwardly. Further, it was observed that the decomposition kinetics are closely concerned with the SiO 2 /Si strained interface at 730°C, suggesting that the point defect produced by the strain is responsible for the SiO formation. Based on the observations, a unified thermal oxidation model is proposed, in which oxidation and decomposition always take place simultaneously by vacancies and Si self-interstitials due to point defect generation at the interface, respectively, in addition to the oxidation of SiO by O 2 permeating into SiO 2 . It is pointed out that SiO molecules are so unstable that they diffuse through the repeated exchange reaction of SiO with SiO 2 . Although SiO can be oxidized under O 2 in SiO 2 , SiO would desorb from the SiO 2 surface when the O 2 pressure is too low and in vacuum, leading to the removal of oxide. By the use of the present unified model, one can interpret successfully the 18 O profile measured by SIMS for the oxidi reoxidized with 18 O 2 isotope [J.A. Costello and R.E. Tressler, J. Electrochem. Soc. 131 (1984) 1944] and the initial rapid oxidation below 300 A.