Since the transistor was invented at Bell Laboratories in 1947 and the concept of the integrated circuit was presented by Jack Kilby of TI in 1958, devices using silicon semiconductors have been developed with tremendous drive. Today, ultrastructural, highly dense, and high-functional ULSI devices have become a reality. Accordingly, novel, three-dimensional devices that aim at multiple functions and high performance have been proposed, and novel materials have come into existence. As Artificial Intelligence (AI) has drawn increasing attention, the concept of “Singularity,” or singular technical point, has become a focus of great attention. Singularity is a prediction put forth by American futurist Ray Kurzweil, who said, “Singularity will come in 2045, when the speed of the evolution of technology will become infinite and Artificial Intelligence will exceed human intelligence.” This prediction is said to have its roots in “Moore’s law,” formulated by Intel founder Gordon Moore, which states that “the degree of integration of transistors doubles every year and a half.” The deep learning and self-learning functions of computers can be mentioned as significant driving factors behind the dramatic development of AI studies. The processing capacity of AI has increased exponentially owing to the evolution and combination of various technologies, and the speed of development of technology now far exceeds the biological limits of humankind. As a result, it is inevitable that “Singularity” will come to pass, and the technologies behind semiconductor devices contributing to the arrival of Singularity are expected to develop much further.

In the process of such semiconductor development, silicon carbide (SiC), among other materials, came to be expected as the next-generation semiconductor in the 1950s, but it could not succeed significantly as a practical device. SiC also attracted attention as the material used in green and red light-emitting elements. In the 1990s, SiC came into the spotlight, along with gallium nitride (GaN) crystal and other materials, by being put into practical use as the material used in blue light-emitting diodes. Today, as the silicon (Si) as power devices have already approached the physical limits of the material, next-generation devices focus on semiconductor substrates such as SiC and GaN, which have performance indexes tens to thousands of times higher than the Si semiconductor. Especially, high-power devices and high-frequency devices have attracted special attention, because the use of semiconductor devices in the automotive and other fields has increased dramatically. Furthermore, the single-crystal substrate of semiconducting diamond is considered to be the ultimate semiconductor device, so this topic has been vigorously researched.

The above-mentioned next-generation devices are called green devices because they could reduce power consumption and carbon dioxide emissions tremendously, leading to the realization of a low-carbon and energy-saving society. Such devices are utilized not only as high-power semiconductors and light-emitting semiconductors but also as various sensors, including gas sensors and UV sensors, as well as MEMS devices. Further application of such devices is expected in the future.

View PDF for the rest of the abstract

SiC, GaN, and diamond are known as super-hard-to-process substrate for next-generation green devices. In this paper, we report on some breakthrough in developing highly efficient processing for such hard-to-process materials, for which we propose improvements in conventional processing, and innovative processing. As part of our project, we developed a “dilatancy pad®” that can efficiently produce high-quality surfaces as well as a high-rigidity, high-speed and high-pressure processing machine. We also designed and prototyped “plasma fusion CMP®,” which is an innovative processing technology fusing CMP (Chemical Mechanical Polishing) with P-CVM (Plasma Chemical Vaporization Machining) to machine super-hard diamond substrates that are considered indispensable for future devices. Before the advent of “singularities” by 2045, super-hard-to-process substrates and ultra-precision polishing technology will become more and more essential.

A novel abrasive-free planarization method named catalyst-referred etching (CARE) was developed. A polishing pad is coated with a catalytic material to promote chemical etching of the work substrate. During processing, the topmost areas of the work substrate, which are in contact with the catalyst surface, are selectively etched off. Atomically highly ordered surfaces are obtained for many types of work substrates. In this paper, the removal characteristics and mechanism of CARE for single crystalline 4H-SiC are reviewed.

The polishing of 4H-SiC wafer processed under ultraviolet (UV) irradiation was investigated to verify the phenomena and effectiveness of ultraviolet-ray aided machining (U-RAM). Inductively coupled plasma spectrometry (ICPS) analysis was conducted to quantitatively determine the oxidation/dissolution volume of SiC. X-ray absorption spectroscopy (XAS) and X-ray photoelectron spectroscopy (XPS) were used to qualitatively analyze the 4H-SiC surfaces. These analyses were used to clarify the compounds that are formed/removed by the decomposition of cathilon dye and water during the polishing of 4H-SiC using TiO₂-, cathilon- and TiO₂-cathilon (mixed) slurries, all of which contained diamond particles. ICPS measurements indicate that a small amount of Si dissolves in aqueous solutions of cathilon- and TiO₂-cathilon. XAS and XPS measurements indicate that SiC composes the bulk of the as-received 4H-SiC, and the surface and thin surface form an interface oxide inside SiC. The chemical-mechanical polishing of 4H-SiC using the TiO₂-cathilon slurry forms an oxide, interface oxide, oxynitride and nitride. Diamond particles easily remove these compounds by mechanical scratching. It is possible to attain smaller surface roughness and higher polishing efficiency by combination with chemical reaction of TiO₂-cathilon slurry and mechanical action of diamond particles under UV irradiation.

In this study, the fundamental sputtering effects of gas cluster ion beams (GCIBs), especially for surface planarization, are reported. Because gas cluster ions are aggregates of thousands of gas atoms, the collision process for a GCIB, with dense and multiple collisions, differs from that of atomic ions via collision cascading; thus, GCIBs have many unique irradiation effects. Among them, the low-damage and surface smoothing effects are beneficial for the planarization of wide-bandgap semiconductor wafers. The planarization of SiC, diamond, and GaN has been demonstrated using GCIB irradiation.

To achieve a 50% worldwide reduction of CO₂ by the middle of this century, development of energy saving power device technology using wide bandgap materials is urgently needed. Diamond is receiving increasing attention as a next generation material for wide bandgap semiconductors owing to its extreme characteristics. Research studies investigating large wafers, low resistivity, and low dislocation have accelerated. This study targets the use of wafers for power electronics applications, and the required machining technologies for diamond, including wafer shaping, slicing, and surface finishing, are introduced.

GaN-based light emitting diodes (LEDs) were epitaxially grown on patterned sapphire substrates (PSSs) to investigate the effectiveness of PSSs for improving the internal quantum efficiency (IQE) and light extraction efficiency (LEE) of the LEDs. Using X-ray diffraction (XRD) and light output measurements, it was observed that the PSSs improved the crystal quality of the LED films and enhanced the LED light intensity. Based on these experimental results, we discuss whether the enhanced light intensity can be attributed to improvements in the IQE or the LEE. The contribution of the IQE improvement to the light intensity was estimated through a comparison of the calculated light emitting area of the LED chip and the measured light output. As a result, it was revealed that the IQE improvement is not the main cause of the increase in the light intensity, indicating that the PSSs mainly improve the LEE. A comparison of the calculated number of bumps on the PSSs and the measured light output of the LEDs suggests that an increase in the number of bumps could affect the improvement in the LEE.

4H-SiC substrate was ablated by linearly polarized femtosecond (fs) laser in three direct write methods at different parameters, such as repetition rate, scanning velocity and fluence, etc. Two processing modes, transverse scanning mode (TSM) and cross irradiation mode (CIM), were introduced. The surface morphologies were observed by scanning electron microscopy (SEM) for detailed investigation. It was found that the surface morphologies differed remarkably at different processing parameters. Firstly, the shapes of micro craters fabricated at different repetition rates and ablation time duration were respectively investigated. The shape of fs laser spot was demonstrated to play an important role for the generation of micro craters. Secondly, the effect of scanning velocity on the formation of nanoripples and micro grooves were investigated. It was found that the spatial ripples could be refabricated during repeated fs laser ablation; periodic ripples and micro grooves depended on fs laser scanning velocity. Agglomerative substance was fabricated especially at slow scanning velocity. Furthermore, rippled surfaces induced at different fluence were achieved and exhibited. Regular and uniform surfaces with periodic ripples were fabricated at the fluence of 0.31∼0.38 J/cm². Finally, CMP was carried out to study the effect of fs laser ablation on polishing.

In this paper, an optimized method of measuring the geometric motion errors of a coordinate measuring machine (CMM) is proposed. The method is based on an improved double ball bar (DBB) that acquires the motion and link errors of the CMM and its actual rotation angles through simultaneous circular tests. The improved DBB has embedded a ring encoder system to the bottom of a commercial DBB on an auxiliary platform. In addition, an improved motion and link error separation algorithm is established by considering the difference angle Δθ between the actual rotation angle and the theoretical rotation angle of the DBB. Both influential factors of the center offset of the DBB and Δθ are discussed through simulations. When geometric motion errors are compensated for and measured on a 400 mm × 400 mm × 150 mm CMM, the standard deviations of the roundness errors decrease to 1.9 μm and 1.5 μm on the XY and ZX planes, respectively.

This paper investigates the effects of geometrical features of tailored textures on friction reduction of a surface with a lubricant-film thickness of several to several-tenths micrometers, using the Navier-Stokes equations and the orthogonal experimental design. The results indicate that the surface textured with the selected sawtooth riblets in lubricant can have up to 93.83% less friction than an untextured surface. The thickness of the lubricant film plays the most important role in friction reduction; the height and the ridge angle of the riblets are the secondary factors. The results and principles obtained can potentially be used in the designing of low-friction surfaces in precision machines with transmission parts.

This work studies a control strategy of screw motion to improve the plasticizing precision for an all-electric injection molding machine (AIMM) based on a biaxial simultaneous motion system. In the standard plasticizing process, a screw retraction before or after the metering phase is proposed to link up with the traditional metering process for reducing the residual pressure at the end of holding and metering, respectively. The modified incomplete plasticizing process is then applied to prevent the potential over-travel of the screw. The incomplete plasticizing process uses a control strategy to obtain a fixed start point for the screw at injection in each cycle. Experimental results showed that both the standard and incomplete plasticizing process improve the visual quality and molded precision of parts.

This paper discusses the topographic features of wheel working surfaces and the grinding force distributions in wheel-work contact zones of creep feed grinding. Grain cutting edge wear is observed by a Scanning Electron Microscope (SEM) and quantitatively evaluated in terms of attritious wear flat percentage, which is able to characterize the wear behavior. By measuring the normal and tangential grinding force distribution in the grinding zone, the distribution form of grinding forces can be approximated to be triangular and the grinding forces increased rapidly due to workpiece burn. It is shown that the variation of the grinding force and the distribution are closely related to cutting edge wear characteristics.

This study proposes an identification and compensation method for the geometric errors of the rotary axes in five-axis machining centers, based on the on-machine measurement results of the machined workpiece. Geometric errors can be identified from the shape geometry of the workpiece machined by five-axis motions because the influence of the errors appears on the shape geometry. An observation equation can be obtained based on the geometric error model and machined shape. The actual geometric errors can be identified by the least square matching of the measured and simulated machined shapes. In order to confirm the effectiveness of the proposed method, an actual cutting test and a simulation are performed. Based on their results, it is confirmed that the proposed method can successfully identify the geometric errors in the simulation. However, these errors cannot be identified in the experiments because a few of them do not have sufficient influences onto the machined shape. On the other hand, although the geometric errors cannot be correctly identified, it is confirmed that the they can be adequately compensated for based on the identified errors in both the simulation and experiment.

When the minor diameter of an end-mill is 1.0 mm or less, handling of tools becomes difficult because of the influence of the characteristic size effect and bending of the cutting edge. Furthermore, it is hard for engineers to derive the cutting conditions that can serve as indexes in the early stage of micro end-milling. In this study, a system that can make instantaneous decisions was developed, on the basis of workpiece material-characteristics and tool shape parameters, by applying data mining techniques together with non-hierarchical and hierarchical clustering methods on micro end-mill catalog data. Slotting experiments using cemented carbide square micro end-mill were carried out to investigate the practicability of derived mining conditions under slotting of A7075 (JIS). We found that catalog mining can be used to derive the guidelines for deciding the micro end-milling conditions.

This study investigated the process evaluation of robot development projects. We propose a quantitative method to evaluate whether the projects advanced to the social experiment phase and analyze their causes of failure. With the proposal analysis method, proposals for future similar developments were submitted across six development projects by staff members in accordance with their actual experience. By analyzing the amounts of proposals classified into phases, categories, and factors, the issues that occurred during each process could be identified. These issues consisted of delays, unclear specifications, and confusion with regard to the organizational structure. The proposed method was suitable for robot development projects with very short time spans such as support robots for disaster victims and guide robots for guests of the Olympic Games.

When designing a large-sized cast product for a multi-spindled machine tool, such as a multi-turret type multi-tasking machine, we must first determine its thermal deformations by the finite element method (FEM) in order to ensure that the structure is designed with high thermal rigidity. Casting technology has progressed so much that we can now form more complicated internal structures and produce much thinner and lighter cast products. Moreover, since designing cast products has become much faster, higher-efficiency analytical techniques are also required. Such analytical techniques are operated by designers where analyses and design are executed interchangeably. In this study, we compared the experimental results to the analytical results in order to evaluate the employed analytical technique based on an actual analysis of a multi-tasking machine bed with a few different rib structures.