Pohang University of Science and Technology

In heavy ion medical synchrotrons, it is crucial to rapidly accelerate ions to reduce beam loss caused by interactions with residual gas and to shorten treatment duration. Superconducting technology can reduce machine size by generating high magnetic fields, but it is difficult to apply in realization because of the heat generation due to AC loss. As a potential solution to these problems, synchrotrons based on High-Temperature Superconductors (HTS) can offer advantages in stable accelerator operation due to their intrinsic high stability margin, despite the AC loss induced heating. Our study focuses on optimizing the ramp-up operation method of a heavy ion synchrotron (with medical applications) using HTS. We initially simulate AC losses based on the ramp-up profile in a conceptual magnet model of a medical application that utilizes HTS wire characteristics. Simultaneously, we employed a virtual beam to calculate how the beam-gas interaction affects emittance. Based on the results of these two calculations, we propose a current ramp-up profile that minimizes beam loss due to multiple scattering and maximum instantaneous heat load due to AC loss in synchrotrons using high-temperature superconducting magnets.

In unilateral teleoperation systems, robots often face challenges when performing tasks with specific geometric constraints. These constraints restrict the robot's movements to certain directions, requiring accurate control of its position and orientation. If the operator's commands do not consider these constraints, excessive contact force may occur, potentially damaging the robot and its environment. Such scenarios can also trigger frequent emergency stops, even with conventional admittance control. To mitigate these issues, we propose a new teleoperation framework tailored for handling geometric constraints. This framework comprises two main components: (1) Geometric Constraint Identification: We use a straightforward line regression method based on Lie group theory to identify geometric constraints. (2) Motion Command Reshaping: The operator's motion commands are safely recalculated using a projection filter coupled with a Lie group setpoint controller. This approach ensures that the robot's movements strictly conform to the identified geometric constraints. As a result, this approach significantly reduces the interaction forces and prevents the risk of severe failures or accidents.

Single‐walled carbon nanotubes (SWCNTs) have gained significant interest for their potential in biomedicine and nanoelectronics. The functionalization of SWCNTs with single‐stranded DNA (ssDNA) enables the precise control of SWCNT alignment and the development of optical and electronic biosensors. This study addresses the current gaps in the field by employing high‐throughput systematic selection, enriching high‐affinity ssDNA sequences from a vast random library. Specific base compositions and patterns are identified that govern the binding affinity between ssDNA and SWCNTs. Molecular dynamics simulations validate the stability of ssDNA conformations on SWCNTs and reveal the pivotal role of hydrogen bonds in this interaction. Additionally, it is demonstrated that machine learning could accurately distinguish high‐affinity ssDNA sequences, providing an accessible model on a dedicated webpage (http://service.k‐medai.com/ssdna4cnt). These findings open new avenues for high‐affinity ssDNA‐SWCNT constructs for stable and sensitive molecular detection across diverse scientific disciplines.

The selection of electrode material is a critical factor that determines the selectivity of electrochemical organic reactions. However, the fundamental principles governing this relationship are still largely unexplored. Herein, we demonstrate a photoelectrocatalytic (PEC) system as a promising reaction platform for the selective radical–radical coupling reaction owing to the inherent charge-transfer properties of photoelectrocatalysis. As a model reaction, the radical trifluoromethylation of arenes is shown on hematite photoanodes without employing molecular catalysts. The PEC platform exhibited superior mono- to bis-trifluoromethylated product selectivity compared to conventional electrochemical methods utilizing conducting anodes. Electrochemical and density functional theory (DFT) computational studies revealed that controlling the kinetics of anodic oxidation of aromatic substrates is essential for increasing reaction selectivity. Only the PEC configuration could generate sufficiently high-energy charge carriers with controlled kinetics due to the generation of photovoltage and charge-carrier recombination, which are characteristic features of semiconductor photoelectrodes. This study opens a novel approach towards selective electrochemical organic reactions through understanding the intrinsic physicochemical properties of semiconducting materials.

The pricing dynamics of sports cars are influenced by a complex interplay of technical specifications and market perceptions. This paper presents a comprehensive analysis using machine learning models to uncover the relationships between various car features and their impact on pricing. We employed linear regression, decision trees, and random forests to predict sports car prices based on features such as torque, horsepower, engine size, acceleration metrics, and model year. The study revealed that while technical features like torque and horsepower significantly affect prices, non-technical factors such as brand prestige and the allure of vintage models also play crucial roles. Our results demonstrate that tree-based models, specifically decision trees and random forests, provide high predictive accuracy, capturing complex non-linear relationships better than linear models. These models effectively highlighted the predominant influence of performance-related features while also suggesting the significant impact of intangible factors like brand and historical value. This study opens the door for future research to integrate broader variables, including consumer behavior and economic conditions, to refine the understanding of pricing strategies in the sports car market. By leveraging advanced machine learning technique.( Abstract )

Modulating the electronic structure of transition metal dichalcogenides (TMDs) via alloying is challenging despite the additional potential applications. In this study, a solvothermal reaction is used to synthesize composition‐tuned ReS2–VS2 (Re1‐xVxS2) alloy nanosheets featuring an expanded interlayer distance. Increasing x induces a phase transition from the semiconducting 1T″ phase ReS2 to the metallic 3R‐stacking 1T phase VS2. Alloying via homogeneous atomic mixing renders the nanosheets more metallic and with less oxidation than VS2. First‐principles calculations consistently predict the 1T″–1T phase transition of the atomically mixed alloy structures. The calculation also suggests that intercalation drives the 3R stacking of 1T phase VS2. The Re1‐xVxS2 nanosheets at x = 0.3–0.8 exhibit enhanced electrocatalytic activity toward water‐splitting hydrogen evolution reaction (HER) in an acid electrolyte. In situ X‐ray absorption fine structure measurements reveal that the metallicity of the alloys is fully retained during HER. The density of states and Gibbs free energy calculations show that alloying increases the metallicity and thus effectively activates the basal S atoms toward the HER, supporting the observed increased HER performance of the alloy nanosheets.

Thermoelectric (TE) charge transport in organic TE nanocomposite systems is a critical consideration in designing high‐performance TE materials. Here, the relationship between the TE properties and energy structure of conducting polymer/quantum dot (QD) nanocomposites is systematically investigated by developing a potential wall or potential well in poly(3,4‐ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) with CdTe QDs. The added QDs are primarily distributed within the electrically insulating PSS shell and act as stepping stones for charge transport between PEDOT‐rich grains. The embedded QDs generate an energy‐filtering effect, which is induced by both potential wall and potential well states established by the QDs in the PEDOT:PSS films. The induced energy‐filtering effect increases the Seebeck coefficient S with limited loss of electrical conductivity σ, thereby overcoming the TE trade‐off relation S ∝ σ −1/4. The energy‐filtering effect is optimized by carefully controlling the QD size. The PEDOT:PSS/QD nanocomposite containing the smallest QDs exhibits a power factor of 173.8 µW m⁻¹ K⁻², which is 80% larger than the value for the pristine PEDOT:PSS film. This work suggests a strategy for designing TE nanocomposites with improved TE performance and emphasizes the importance of fine‐tuning the interfacial energy gap to achieve an effective energy‐filtering effect.

We present a study on the superconducting properties of 500 nm thick NbTiN films grown by reactive co-sputtering on silicon substrates at room temperature. The samples exhibit a chemical composition with Nb (50 at.%) and Ti (50 at.%), revealing a polycrystalline structure characterized by columnar growth and an average lateral grain size of approximately 40 nm. The superconducting critical temperature (Tc) was measured at 13.8 K, and the upper critical field extrapolated to zero temperature reached 22 T, resulting in a coherence length (ξ) of 3.8 nm. The penetration depth (λ) was determined through local magnetic force microscopy measurements conducted at temperatures of 4.25 and 6 K. The obtained values were 400 (15) nm at 4.25 K and 430 (15) nm at 6 K. Extrapolating these measurements to zero temperature, we obtained an estimated value of 380 nm. A comparison was made with samples that underwent thermal annealing at 700 °C, resulting in a reduction of disorder at the nanoscale and an increase in Tc to 14.2 K. Despite this enhancement, the coherence length ξ (0) remained at approximately 3.8 nm, with no appreciable changes in the λ values. Our findings contribute to understanding fundamental superconducting parameters in nitride thin films, with potential applications ranging from resonant accelerator cavities to Josephson junctions and radiation detectors.

Optimizing exchange–correlation functionals for both core/valence ionization potentials (cIPs/vIPs) and valence excitation energies (VEEs) at the same time in the framework of MRSF-TDDFT is self-contradictory. To overcome the challenge, within the previous “adaptive exact exchange” or double-tuning strategy on Coulomb-attenuating XC functionals (CAM), a new XC functional specifically for cIPs and vIPs was first developed by enhancing exact exchange to both short- and long-range regions. The resulting DTCAM-XI functional achieved remarkably high accuracy in its predictions with errors of less than half eV. An additional concept of “valence attenuation”, where the amount of exact exchange for the frontier orbital regions is selectively suppressed, was introduced to consistently predict both VEEs and IPs at the same time. The second functional, DTCAM-XIV, exhibits consistent overall prediction accuracy at ∼0.64 eV. By preferentially optimizing VEEs within the same “valence attenuation” concept, a third functional, DTCAM-VAEE, was obtained, which exhibits improved performance as compared to that of the previous DTCAM-VEE and DTCAM-AEE in the prediction of VEEs, making it an attractive alternative to BH&HLYP. As the combination of “adaptive exchange” and “valence attenuation” is operative, it would be exciting to explore its potential with a more tunable framework in the future.

Neurodevelopmental disorders (NDD) in offspring are associated with a complex combination of pre-and postnatal factors. This study uses machine learning and population data to evaluate the association between prepregnancy or perinatal risk factors and the NDD of offspring. Population-based retrospective cohort data were obtained from Korea National Health Insurance Service claims data for 209,424 singleton offspring and their mothers who gave birth for the first time in 2007. The dependent variables were motor development disorder (MDD), cognitive development disorder (CDD) and combined overall neurodevelopmental disorder (NDD) from offspring. Seventeen independent variables from 2002 to 2007 were included. Random forest variable importance and Shapley Additive Explanation (SHAP) values were calculated to analyze the directions of its associations with the predictors. The random forest with oversampling registered much higher areas under the receiver-operating-characteristic curves than the logistic regression of interaction and non-linearity terms, 79% versus 50% (MDD), 82% versus 52% (CDD) and 74% versus 50% (NDD). Based on random forest variable importance, low socioeconomic status and age at birth were highly ranked. In SHAP values, there was a positive association between NDD and pre- or perinatal outcomes, especially, fetal male sex with growth restriction associated the development of NDD in offspring.

The fabrication of environmentally benign, solvent‐processed, efficient, organic photovoltaic sub‐modules remains challenging due to the rapid aggregation of the current high performance non‐fullerene acceptors (NFAs). In this regard, design of new NFAs capable of achieving optimal aggregation in large‐area organic photovoltaic modules has not been realized. Here, an NFA named BTA‐HD‐Rh is synthesized with longer (hexyl‐decyl) side chains that exhibit good solubility and optimal aggregation. Interestingly, integrating a minute amount of new NFA (BTA‐HD‐Rh) into the PM6:L8‐BO system enables the improved solubility in halogen‐free solvents (o‐xylene:carbon disulfide (O‐XY:CS2)) with controlled aggregation is found. Then solar sub‐modules are fabricated at ambient condition (temperature at 25 ± 3 °C and humidity: 30–45%). Ultimately, the champion 55 cm² sub‐modules achieve exciting efficiency of >16% in O‐XY:CS2 solvents, which is the highest PCE reported for sub‐modules. Notably, the highest efficiency of BTA‐HD‐Rh doped PM6:L8‐BO is very well correlated with high miscibility with low Flory‐Huggins parameter (0.372), well‐defined nanoscale morphology, and high charge transport. This study demonstrates that a careful choice of side chain engineering for an NFA offers fascinating features that control the overall aggregation of active layer, which results in superior sub‐module performance with environmental‐friendly solvents.

The electrolysis of aromatic molecules is a useful method for the synthesis and deposition of conducting polymers. However, this method cannot be applied to diverse large‐area electronic devices because films grow vertically on the surface of an electrically connected working electrode. Herein, the remote‐controllable lateral electropolymerization of 3,4‐ethylenedioxythiophene using direct‐current voltage superimposed alternating‐current voltage (ADC)‐bipolar electrochemistry is reported. The use of shape‐designed dual bipolar electrodes and systematic optimization of the process parameters led to the fabrication of uniform poly(3,4‐ethylenedioxythiophene) (PEDOT) films on 2‐inch glass wafers and flexible plastic substrates. The oxidation levels and microstructures of ADC‐electropolymerized PEDOTs with various supporting electrolytes are investigated and correlated with their thermoelectric properties. A soft thermocouple and resistive‐type gas sensor based on an ADC‐electrodeposited PEDOT are demonstrated to monitor cerebral temperature in a brain replica and to sense nitrogen dioxide gas, respectively.

We present the fabrication of 4 K-scale electrochemical random-access memory (ECRAM) cross-point arrays for analog neural network training accelerator and an electrical characteristic of an 8 × 8 ECRAM array with a 100% yield, showing excellent switching characteristics, low cycle-to-cycle, and device-to-device variations. Leveraging the advances of the ECRAM array, we showcase its efficacy in neural network training using the Tiki-Taka version 2 algorithm (TTv2) tailored for non-ideal analog memory devices. Through an experimental study using ECRAM devices, we investigate the influence of retention characteristics on the training performance of TTv2, revealing that the relative location of the retention convergence point critically determines the available weight range and, consequently, affects the training accuracy. We propose a retention-aware zero-shifting technique designed to optimize neural network training performance, particularly in scenarios involving cross-point devices with limited retention times. This technique ensures robust and efficient analog neural network training despite the practical constraints posed by analog cross-point devices.

As the incidence of extreme precipitation events attributable to global climate change increases, providing policymakers with accurate model predictions is of the utmost importance. However, model projections have inherent uncertainties. The present study attempted to distinguish the sources of the uncertainty of the mean and extreme precipitation projections in the East Asia region using the mean boreal summer precipitation, simple precipitation intensity index (SDII), maximum cumulative 5 day precipitation, and annual maximum daily precipitation (Rx1d). The results show that while the mean precipitation was projected to change very little regardless of the scenario, more extreme indices were projected to increase considerably by the end of the century, particularly in the high-emissions scenarios. On average, model uncertainty accounted for the largest part of the uncertainty. However, for Rx1d in the 2030s, as well as mean and SDII in some regions until the 2060s, the internal variability was the largest contributor. In addition, whilst scenario uncertainty accounted for a negligible proportion of average precipitation variability, for the more extreme the precipitation indices, scenario uncertainty contribution to total variability by the end of the century was significant; namely, the scenario uncertainty contribution was overall highest for the maximum one-day precipitation. Additionally, comparatively wetter regions had greater overall projection uncertainties, especially uncertainty arising from internal variability, likely due to the influence of interannual variability from the EA summer monsoon.

Neuromorphic technologies typically employ a point neuron model, neglecting the spatiotemporal nature of neuronal computation. Dendritic morphology and synaptic organization are structurally tailored for spatiotemporal information processing, such as visual perception. Here we report a neuromorphic computational model that integrates synaptic organization with dendritic tree-like morphology. Based on the physics of multigate silicon nanowire transistors with ion-doped sol–gel films, our model—termed dendristor—performs dendritic computation at the device and neural-circuit level. The dendristor offers the bioplausible nonlinear integration of excitatory/inhibitory synaptic inputs and silent synapses with diverse spatial distribution dependency, emulating direction selectivity, which is the feature that reacts to signal direction on the dendrite. We also develop a neuromorphic dendritic neural circuit—a network of interconnected dendritic neurons—that serves as a building block for the design of a multilayer network system that emulates three-dimensional spatial motion perception in the retina.