DKDongmin Kim

Dongmin Kim

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Dongmin Kim has a total of 32 co-authors. Here are the top co-authors by number of shared publications:

  • Hyunsang Hwang

    Center For Single Atom‐Based Semiconductor Device and the Department of Materials Science and Engineering Pohang University of Science and Technology Pohang Republic of Korea

    14 shared publications
  • Yoori Seo

    Department of Material Science and Engineering, Center for Single Atom-based Semiconductor Device, Pohang University of Science and Technology (POSTECH), Pohang, South Korea

    9 shared publications
  • Jangseop Lee

    Department of Material Science and Engineering, Center for Single Atom-based Semiconductor Device, Pohang University of Science and Technology (POSTECH), Pohang, South Korea

    6 shared publications
  • Geonhui Han

    Department of Material Science and Engineering, Center for Single Atom-based Semiconductor Device, Pohang University of Science and Technology (POSTECH), Pohang, South Korea

    4 shared publications
  • Kyumin Lee

    Department of Material Science and Engineering, Center for Single Atom-based Semiconductor Device, Pohang University of Science and Technology (POSTECH), Pohang, South Korea

    4 shared publications
  • Ohhyuk Kwon

    Department of Material Science and Engineering, Center for Single Atom-based Semiconductor Device, Pohang University of Science and Technology (POSTECH), Pohang, South Korea

    4 shared publications

Latest Publications

  • Open Access

    Extreme Bond Ionicity in Mg─Te Chalcogenides for Ultrathin Low Voltage Selector‐Only Memory beyond the Leakage Scaling Limit

    by Yoori Seo, Dongmin Kim, Yu Bin Park, Jangseop Lee, Tae Hoon Lee, Hyunsang Hwang

    Conventional ovonic threshold switching (OTS) chalcogenides face a fundamental scaling limit for selector‐only memory (SOM), because aggressive thickness scaling increases leakage current and hinders reliable low voltage operation. Here, this trade‐off can be overcome by exploiting the unique materials characteristics of the Mg─Te chalcogenide system. Guided by the bonding ionicity map and supported by density functional theory calculations, Mg─Te is identified as an optimal telluride material whose highly ionic bonding is associated with deeper trap levels, a large memory window of 1.75 V, and suppressed leakage in 20 nm devices. Structural analyses and multiscale simulations suggest that the spontaneous phase separation in Mg 1 Te 3 forms MgTe‐like ordered nanodomains and Te‐rich amorphous regions, providing structural partitioning that may constrain the effective amorphous switching network and reduce stochastic switching variability. With a thin Hf interlayer, the 5 nm Mg 1 Te 3 device achieves narrow SET/RESET switching uniformity of σ = 15/28 mV at ± 2.5 V operating voltage, representative 10 ns programming speed, and write endurance exceeding 10 10 cycles in the best‐performing device. These results highlight Mg─Te as a promising basis for highly scaled ultralow voltage SOM through the combined roles of ionic bonding, structural partitioning, and interfacial engineering.

  • Oxygen-Stoichiometry-Controlled Linearity and Uniformity of W/WO x /W Fixed-Resistance Arrays for Edge-AI Inference

    by Hoesung Ha, Dongmin Kim, Ohhyuk Kwon, Yoori Seo, Hyunsang Hwang

    We present W/WOx/W fixed-resistance synaptic devices that enable inference-only weight mapping without iterative programming. By establishing an oxygen-stoichiometry-dependent operating window in WOx that provides transport characteristics suitable for inference, near-ohmic read behavior (linearity factor ≤ 1.01) and low device-to-device variation (~ 0.01) are simultaneously achieved. Within this window, conductance is determined by the defined device area, enabling 5-bit weight representation. These characteristics are preserved in a packaged 32 × 32 array, where EMG-based neural inference reaches 85.4% accuracy. This work establishes an inference-suitable material regime for fixed-resistance WOx synapses and provides a practical pathway toward low-variation edge-AI hardware.

  • Open Access

    DFT-Informed Multiscale Modeling of Switching Mechanisms and Design Strategies for Selector-Only Memory Devices

    by Dongmin Kim, Yoori Seo, Ohhyuk Kwon, Yubin Park, Taehoon Lee, Hyunsang Hwang

    Chalcogenide-based selector-only memory (SOM) has emerged as a promising candidate for scalable memory expansion in Compute Express Link (CXL)-based and high-density computing systems. However, most prior studies investigate SOM devices at a single scale, either at the atomic or device level, limiting a comprehensive understanding of SOM operation. In this work, we present a density functional theory (DFT)-informed stochastically induced defect percolation (SIDP) framework that explicitly bridges atomic bonding statistics and device-scale switching behavior. The SIDP framework captures how microscopic bonding rearrangements and defect activation collectively evolve into conductive percolation paths, enabling quantitative analysis and prediction of SOM characteristics while naturally reproducing statistical switching variability. The framework is validated using fabricated In–Te-based SOM devices, revealing that Te–Te homopolar bonds govern both memory performance and switching variability, thereby giving rise to an intrinsic tradeoff in SOM performance. Leveraging this insight, we employ the SIDP model to demonstrate the feasibility of two optimization strategies, multilayer structure and device scaling. By combining these two approaches, we demonstrate that memory window (MW) up to 4.34 V and read window margins (RWMs) of 2.83 V are achievable, highlighting the optimization potential enabled by physics-informed design.

  • Open Access

    Boosting Stochasticity in Ovonic Threshold Switches Through Cryogenic First Firing for Fast and Reliable Entropy Generation

    by Dongmin Kim, Jangseop Lee, Yoori Seo, Ohhyuk Kwon, Pendar Azaripour Masouleh, Jisung Lee, Joonhyun Kwon, Chul‐Heung Kim, Hyunsang Hwang

    As encryption demands at the edges grow, volatile switching devices have emerged as promising candidates for entropy sources because of their inherent stochastic properties, offering fast, energy‐efficient operation and a minimal footprint. Although most studies have focused on exploiting inherent stochasticity, efforts to analyze and optimize these devices to enhance their randomness remain scarce. In this study, the stochastic switching characteristics of Ovonic Threshold Switch (OTS) devices by controlling the first firing temperatures to amplify their inherent stochasticity are explored. It is demonstrated that firing at cryogenic temperatures (77 K) induces field‐dominant firing and a considerable increase in traps within the device. These additional traps lead to a substantial enhancement in switching variability, with the switching time fluctuation increasing up to four times compared to the first firing temperature of 298 K. Furthermore, a reference‐free entropy‐harvesting method is proposed that ensures robust and stable operation even under cycling degradation. Based on this approach, the OTS devices that undergo first firing at cryogenic temperatures achieve stable entropy generation at speeds exceeding 20 Mbit s−1. This study demonstrates the potential of optimizing OTS devices to satisfy the increasing demand for fast and energy‐efficient entropy sources in advanced cryptographic systems.


  • Open Access

    Physical Reservoir Computing for Real‐Time Electrocardiogram Arrhythmia Detection Through Controlled Ion Dynamics in Electrochemical Random‐Access Memory

    by Kyumin Lee, Dongmin Kim, Jongseon Seo, Hyunsang Hwang

    From the perspective of developing low‐power mobile healthcare devices capable of real‐time electrogram diagnosis, memristor‐based physical reservoir computing (PRC) offers a promising alternative to conventional deep neural network (DNN)‐based systems. Here, real‐time electrocardiogram (ECG) monitoring and arrhythmia detection are demonstrated using electrochemical random‐access memory (ECRAM)‐based PRC. ECRAM devices provide the millisecond‐range temporal resolution required for bio‐potential signals like ECG. Through material and process engineering, it is identified that higher ionic conductivity (σion) in the electrolyte layer and lower ionic diffusivity (Dion) in the channel layer are crucial for achieving non‐linear dynamics and fading memory characteristics. In addition, LaF3/WOx‐based ECRAM exhibits low‐power operation (≈300 pW spike−1) with minimal cycle‐to‐cycle (CTC) variation (3/WOx‐based ECRAM with engineering guidelines of ion dynamics makes a significant contribution to mobile healthcare systems for electrogram diagnosis.


  • Enhancing Selector-Only Memory Reliability Through Multi-Step Write Pulse

    by Yoori Seo, Sanghyun Ban, Jangseop Lee, Dongmin Kim, Laeyong Jung, Hyunsang Hwang

    In this study, we investigated the impact of overshoot current (I over ) on the reliability of selector-only memory (SOM) devices based on an ovonic threshold switch (OTS). We found that implementing a multi-step pulse during the write operation reduced I over by 40%, significantly decreased threshold voltage (V th ) variability during the read process to ensure sufficient read window margin, and improved endurance by more than an order of magnitude, compared to the traditional square pulse. Our analysis revealed that the I over leads to increased I on variability, which in turn causes greater variability in the write pulse’s impact on subsequent read process. Furthermore, I over generates extra injection charge, which directly correlating with increased device stress and significantly impacting endurance. Through intermittent and continuous pulse measurements, we confirmed the critical role of I over in cycling stability. Our findings underscore the importance of suppressing I over to enhance SOM reliability, emphasizing the need for optimizing the multi-step pulse method.


  • Highly Scalable Vertical Bypass RRAM Using Interface-Type Resistive Switching Mechanism for V-Nand memory Applications

    by Geonhui Han, Jongseon Seo, Kyumin Lee, Dongmin Kim, Yoori Seo, Jinwoo Lee, Jinmyung Choi, Dongho Ahn, Sechung Oh, Hyunsang Hwang

    In this study, we present a vertical bypass resistive random access memory (VB-RRAM) that integrates interface-type resistive switching RRAM with excellent memory characteristics for V-NAND applications. The VB-RRAM combines a resistivity switching (RS) layer, serving as memory, with a transistor selector to enable V-NAND operations. The device operates through bypass reading between the transistor and the RS layer. Therefore, through materials engineering, we optimized the resistance ratio among components by varying the V o concentration in the IGZO transistor, adjusting the stoichiometry of the WO x RS layer, and modifying the ion conduction barrier in HfO x electrolyte layer. Furthermore, we demonstrated that bypass reading is dependent on the RS states and memory characteristics are attributed to the RS layer with electrochemical reaction, resulting in low-voltage operation (4 bits), and reliable retention characteristics (~107 s). In addition, we addressed the key vertical integration issues, such as isolation layer (IL) thickness and read disturbance issue to guarantee reliable V-NAND operations. By integrating a vertical structure with a 30-nm channel length, we achieved the excellent memory performance, including a high ON/OFF ratio (>105), stable endurance (>107 cycles), fast switching speed (<50 ), and low energy consumption (<5 fJ), demonstrating the remarkable scalability of bypass RRAM.


  • Open Access

    Volatile threshold switching devices for hardware security primitives: Exploiting intrinsic variability as an entropy source

    by Wooseok Choi, Ohhyuk Kwon, Jangseop Lee, Seungyeol Oh, Seongjae Heo, Sanghyun Ban, Yoori Seo, Dongmin Kim, Hyunsang Hwang

    In the age of the Internet of Things, the proliferation of edge devices has resulted in a significant increase in personal information that is susceptible to theft and counterfeiting at various stages of data communication. As a result, substantial attention has been focused on hardware (HW) security elements, such as the true random number generator and physical unclonable function. With the recent surge in research and development of emerging memristors, which exploit the inherent variability of these devices, there has been a notable increase in studies on HW security. Particularly, volatile threshold switch (TS) devices, which exhibit insulator/metal characteristics below/above a certain threshold voltage, show great promise as security devices due to their lower power consumption and higher cycling endurance compared to nonvolatile memory devices. Despite the promising attributes and increasing demand for TS devices for HW security, there remains a lack of a comprehensive overview covering various TS devices and their potential contributions to HW privacy. To address this gap, this review provides an encompassing analysis of different types of TS devices and their performance in HW security literature, providing insight into current limitations and the future prospects of HW security primitives based on TS devices.


  • 3D Stackable Vertical‐Sensing Electrochemical Random‐Access Memory Using Ion‐Permeable WS2 Electrode for High‐Density Neuromorphic Systems

    by Kyumin Lee, Seungkwon Hwang, Dongmin Kim, Jongwon Yoon, Jung‐Dae Kwon, Yonghun Kim, Hyunsang Hwang

    Ion‐based electrochemical random‐access memory (ECRAM) is proposed for synaptic applications owing to its promising characteristics that have the potential to accelerate data processing through neuromorphic systems. However, attaining ideal synaptic functionalities and constructing high‐density vertical synapse arrays are challenging due to issues related to uncontrolled ion migration and constraints in 3D multi‐stacking. Here, a breakthrough using 3D stackable Li ion‐based vertical‐sensing ECRAM (VS‐ECRAM) is presented with an ion‐permeable ultrathin WS2 electrode synthesized through low‐temperature (200 °C) atmospheric‐pressure plasma‐enhanced chemical vapor deposition (AP‐PECVD). The direct AP‐PECVD of the WOx channel layer induces WS2 formation in the surface region, which exhibits sufficient electrical conductivity to function as an electrode. By utilizing the WS2 electrode as an ion‐barrier layer in the VS‐ECRAM synapse, excellent weight update linearity and cycling variability are achieved due to the finely controlled ion migration. Furthermore, a two‐layer stacked 3D VS‐ECRAM is successfully fabricated through the vertical WS2 formation, and independent weight updates without any disturbance are confirmed. Finally, a high pattern recognition accuracy of 95.22% is obtained using a multi‐layer perceptron‐based neural network. Therefore, the proposed 3D stackable WS2‐based VS‐ECRAM exhibits a strong potential for application in high‐density neuromorphic devices with excellent synaptic performances.


  • Open Access

    Publisher's Note: “Improving the selector characteristics of ovonic threshold switch via UV treatment process” [Appl. Phys. Lett. 123, 242103 (2023)]

    by Yoori Seo, Jangseop Lee, Sanghyun Ban, Dongmin Kim, Geonhui Han, Hyunsang Hwang

    This article was originally published online on 13 December 2023 with errors on pages 4 and 6. On page 4 above Eq. (1), the number 3 is not a reference citation