Triboelectric generator is attracting much attention as a power source of electronics application. Electromotive force induced by rubbing is a key for triboelectric generator. From dielectric physics point of view, there are two microscopic origins for electromotive force, i.e., electronic charge displacement and dipolar rotation. A new way for evaluating these two origins is an urgent task. We have been developing an optical second-harmonic generation (SHG) technique as a tool for probing charge displacement and dipolar alignment, selectively, by utilizing wavelength dependent response of SHG to the two origins. In this paper, an experimental way that identifies polarity of electronic charge displacement, i.e., positive charge and negative charge, is proposed. Results showed that the use of local oscillator makes it possible to identify the polarity of charges by means of SHG. As an example, positive and negative charge distribution created by rubbing polyimide surface is illustrated.

In vacuum-deposited film composed of organic polar molecules, polarization charges appear on the film surface owing to spontaneous orientation of the molecule. Because its density (σ_pol) determines an amount of accumulation charge (σ_acc) in organic light-emitting diodes and output power in polar molecular-based vibrational energy generators (VEGs), control of molecular orientation is highly required. Recently, several groups have reported that dipole-dipole interaction between polar molecules induces anti-parallel orientation which does not contribute to σ_pol. In other words, perturbation inducing the attenuation of the dipole interaction is needed to enhance σ_pol. In this study, to investigate an effect of light irradiation on σ_pol, we prepared 1,3,5-tris(1-phenyl-1H-benzimidazol-2-yl)benzene (TPBi) film under illumination during its deposition, and evaluated the σ_acc in TPBi-based bilayer device, which equals to σ_pol. We found that the σ_acc was increased by light irradiation, indicating that average orientation of TPBi is enhanced. These results suggest that light irradiation during device fabrication is promising process for organic electronic devices including polar molecule-based VEGs.

We fabricated silicon nanowires (SiNWs) using a metal-catalyzed electroless etching method, which is known to be a low-cost and simple technique. The SiNW arrays with a length of 540 nm were used as a substrate of SiNWs/PEDOT:PSS hybrid solar cell. Furthermore, gold nanoparticles (AuNPs) were used to improve the light absorption of the device due to localized surface plasmon excitation. The results show that the short-circuit current density and the power conversion efficiency increased from 22.1 mA/cm² to 26.0 mA/cm² and 6.91% to 8.56%, respectively. The advantage of a higher interface area between the organic and inorganic semiconductors was established by using SiNW arrays and higher absorption light incorporated with AuNPs for improving the performance of the developed solar cell.

The optical properties of new tricyanopyrroline (TCP)-based chromophores with a benzyloxy group bound to aminobenzene donor unit were characterized by hyper-Rayleigh scattering (HRS), absorption spectrum, and ¹H-NMR measurements, and the influence of the benzyloxy group on TCP-based chromophores was discussed based on the data. A positive effect of NLO properties was found in TCP-based NLO chromophores with a benzyloxy group compared with benchmark NLO chromophores without the benzyloxy group, suggesting an influence of intra-molecular hydrogen bond. Furthermore, we propose a formation of double intra-molecular hydrogen bonds in the TCP chromophore with monoene as the π-conjugation bridge and aminobenzene with a benzyloxy group as the donor unit.

We previously developed a new terahertz (THz) wave detection method that utilizes the effect of nonlinear optical (NLO) polymers. The new method provided us with a gapless detection, a wide detection bandwidth, and a simpler optical geometry in the THz wave detection. In this paper, polarization dependences in THz wave detection by the Stark effect were investigated. The projection model was employed to analyze the polarization dependences and the consistency with experiments was observed qualitatively, surely supporting the use of the first-order Stark effect in this method. The relations between THz wave detection by the Stark effect and Stark spectroscopy or electroabsorption spectroscopy are also discussed.

It is found that the electrical resistance-length characteristic in an electroactive supercoiled polymer artificial muscle strongly depends on the temperature. This may come from the thermal expansion of coils in the artificial muscle, which increases the contact area of neighboring coils and results in a lower electrical resistance at a higher temperature. On the other hand, the electrical resistance-length characteristic collected during electrical driving seriously deviates from those collected at constant temperatures. Inhomogeneous heating during electrical driving seems to be a key for the deviation.

In this study, we devised a biofuel cell (BFC) by impregnating sheet-like cellulose nanofiber (CNF) with liquid fuel (fructose) and sandwiching it with the electrodes, making the structure simple and compact. CNF was considered as a suitable material for BFC because it is biocompatible, has a large specific surface area, and exhibits excellent properties as a catalyst and an adsorbent. In this BFC device, graphene-coated carbon fiber woven cloth (GCFC) was used as the material for preparing the electrodes, and the amount of enzyme modification on the surface of each electrode was enhanced. Further, as the distance between the electrodes was same as the thickness of the sheet-shaped CNF, it facilitated the exchange of protons between the electrodes. Moreover, the cathode, which requires an oxidation reaction, was exposed to the atmosphere to enhance the oxygen uptake. The maximum power density of the CNF-type BFC was recorded as 114.5 µW/cm² at a voltage of 293 mV. This is more than 1.5 times higher than that of the liquid-fuel-type BFC. When measured after 24 h, the maximum power density was recorded as 44.9 µW/cm² at 236 mV, and the output was maintained at 39% of that observed at the beginning of the measurement. However, it is not the case with general BFCs, where the power generation after 24 h is less than 5%. Therefore, the CNF-type BFCs have a longer lifespan and are fuel efficient.

Improvement of output and lifetime is a problem for biofuel cells. A structure was adopted in which gelation mixed with agarose and fuel (fructose) was sandwiched by electrodes made of graphene-coated carbon fiber. The cathode surface not contacting the gel was exposed to air. In addition, the anode surface not contacting the gel was in contact with fuel liquid to prevent the gel from being dry. The power density of the fuel cell was improved by increasing oxygen supply from air and the lifetime was improved by maintaining wet gel, that is, the proposed structure was a hybrid type having advantages of both fuel gel and fuel liquid. The output increased almost up to that of just using fuel gel and did not decrease significantly over time. The maximum power density in the proposed system was approximately 74.0 µW/cm², an enhancement of approximately 1.5 times that in the case of using liquid fuel. The power density after 24 h was approximately 46.1 µW/cm², which was 62% of the initial value.

In this study, two modification methods that employ graphene-coated carbon fiber woven fabric (GCFC) as an electrode and 2,2'-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid (ABTS) as a mediator were used to evaluate cathode performance. In addition, a prototype of an atmosphere-exposed ascorbic-acid enzyme biofuel cell (AAEBFC) consisting of an improved GCFC cathode and ABTS was evaluated. No modification was made in the anode region, and only the cathode region was coated with the enzyme of bilirubin oxidase (BOD). As a result of implementing an ABTS-modified cathode in the AAEBFC, an output of 721μW/cm² was obtained at 0.189V. When the gel thickness was changed, an output of 1200μW/cm² was obtained at 0.17V. To the best of our knowledge, this is currently the highest reported output for an AAEBFC fueled by ascorbic acid.

Biofuel cells (BFCs) using graphene-coated carbon fiber cloth electrodes and glucose gel fuel were fabricated and evaluated. A new structure using fuel gel, in which the anode was embedded in gel and the cathode was exposed to the atmosphere, was adopted. Air-exposed biofuel cells using gel have already been reported, however, adhesion between the anode and the gel was improved by the proposed structure. In addition, the enlargement of the gel area prevented its drying. These innovations improved the power density and lifetime of the BFCs. The anode was modified with a glucose oxidase (GOD) enzyme and a mediator (ferrocene) and the cathode was modified with a bilirubin oxidase enzyme. The power density of the proposed structure was 176.4 µW/cm² at 0.19 V, which was approximately 3.8 times higher than that of BFCs using liquid fuel (45.9 µW/cm²).

In fields such as medicine and chemistry, methods for transporting microdroplets are currently necessitated, which include the analysis of reagents, mixing, and separation. As microdroplets become finer, their movement becomes difficult to control as a result of surface tension. This has resulted in the use of an excessive amount of reagents. In this study, we evaluated the dynamic characteristics of microdroplets and the excitation force. Microdroplets were dropped onto a tilted glass substrate, and the displacement of the microdroplets was measured while changing the droplet amount, vibration frequency, and vibration direction. Moreover, the behavior of the droplet just before it started to move was observed, and the relationship between the displacement of the minute droplet and the vibration force was compared and examined.

Recent IoT devices require extremely low standby power consumption, while a certain performance is needed during the active time, and Coarse-Grained Reconfigurable Arrays (CGRAs) have received attention because of their high energy efficiency. For further reduction of the standby energy consumption of CGRAs, the leakage power for their configuration memory must be reduced. Although the power gating is a common technique, the lost data in flip-flops and memory must be retrieved after the wake-up. Recovering everything requires numerous state transitions and considerable overhead both on its execution time and energy. To address the problem, Non-volatile Cool Mega Array (NVCMA), a CGRA providing non-volatile flip-flops (NVFFs) with spin transfer torque type non-volatile memory (NVM) technology has been developed. However, in general, non-volatile memory technologies have problems with reliability. Some NVFFs are stacked-at-0/1, and cannot store the data in a certain possibility. To improve the chip yield, we propose a mapping algorithm to avoid faulty processing elements of the CGRA caused by the erroneous configuration data. Next, we also propose a method to add an error-correcting code (ECC) mechanism to NVFFs for the configuration and constant memory. The proposed method was applied to NVCMA to evaluate the availability rate and reduction of write time. By using both methods, the average availability ratio of 94.2% was achieved, while the average availability ratio of the nine applications was 0.056% when the probability of failure of the FF was 0.01. The energy for storing data becomes about 2.3 times because of the hardware overhead of ECC but the proposed method can save 8.6% of the writing power on average.

IoT devices operate with a battery and have embedded firmware in flash memory. If the embedded firmware is not kept up to date, there is a possibility of problems that cannot be linked with other IoT networks, so it is necessary to maintain the latest firmware with frequent updates. However, because firmware updates require developers and equipment, they consume manpower and time. Additionally, because the device must be active during the update, a low-power operation is not possible due to frequent flash memory access. In addition, if an unexpected interruption occurs during an update, the device is unavailable and requires a reliable update. Therefore, this paper aims to improve the reliability of updates and low-power operation by proposing a technique of performing firmware updates at high speed. In this paper, we propose a technique to update only a part of the firmware stored in nonvolatile flash memory without pre-processing to generate delta files. The firmware is divided into function blocks, and their addresses are collectively managed in a separate area called a function map. When updating the firmware, only the new function block to be updated is transmitted from the host downloader, and the bootloader proceeds with the update using the function block stored in the flash memory. Instead of transmitting the entire new firmware and writing it in the memory, using only function block reduces the amount of resources required for updating. Function-blocks can be called indirectly through a function map, so that the update can be completed by modifying only the function map regardless of the physical location. Our evaluation results show that the proposed technique effectively reduces the time cost, energy consumption, and additional memory usage overhead that can occur when updating firmware.

Due to the slowdown of Moore's Law, power limitation has been one of the most critical issues for current and future HPC systems. To more efficiently utilize HPC systems when power budgets or deadlines are given, it is very desirable to accurately estimate the performance or power consumption of applications before conducting their tuned production runs on any specific systems. In order to ease such estimations, we showcase a straight-forward and yet effective method, based on the enhanced power management framework and DSL we developed, to help HPC users to clarify the performance and power relationships of their applications. This method demonstrates an easy process of profiling, modeling and management on both performance and power of HPC systems and applications. In our evaluations, only a few (up to 3) profiled runs are necessary before very precise models of HPC applications can be obtained through this method (and algorithm), which has dramatically improved the efficiency of and lowered the difficulty in utilizing HPC systems under limited power budgets.

Since most sensor data depend on each other, time-series anomaly detection is one of practical applications of IoT devices. Such tasks are handled by Recurrent Neural Networks (RNNs) with a feedback structure, such as Long Short Term Memory. However, their learning phase based on Stochastic Gradient Descent (SGD) is computationally expensive for such edge devices. This issue is addressed by executing their learning on high-performance server machines, but it introduces a communication overhead and additional power consumption. On the other hand, Recursive Least-Squares Echo State Network (RLS-ESN) is a simple RNN that can be trained at low cost using the least-squares method rather than SGD. In this paper, we propose its area-efficient hardware implementation for edge devices and adapt it to human activity anomaly detection as an example of interdependent time-series sensor data. The model is implemented in Verilog HDL, synthesized with a 45 nm process technology, and evaluated in terms of the anomaly capability, hardware amount, and performance. The evaluation results demonstrate that the RLS-ESN core with a feedback structure is more robust to hyper parameters than an existing Online Sequential Extreme Learning Machine (OS-ELM) core. It consumes only 1.25 times larger hardware amount and 1.11 times longer latency than the existing OS-ELM core.

Scalability of distributed DNN training can be limited by slowdown of specific processes due to unexpected hardware failures. We propose a dynamic process exclusion technique so that training throughput is maximized. Our evaluation using 32 processes with ResNet-50 shows that our proposed technique reduces slowdown by 12.5% to 50% without accuracy loss through excluding the slow processes.