Skip to main navigation Skip to search Skip to main content

Design and characterization of two-terminal thyristor dynamic random-access memory with a localized lightly-doped base insertion for improving data retention

  • Jaeung Ryu
  • , Minjae Kim
  • , Jueun Kim
  • , Minju Hwang
  • , Woojoo Lee
  • , Yeji Lee
  • , Seongjae Cho
  • , Il Hwan Cho

Research output: Contribution to journalArticlepeer-review

Abstract

Scaling challenges in dynamic random-access memory (DRAM) have driven the development of one-transistor capacitor-less (1T) DRAM architectures, such as Thyristor RAM (TRAM). However, silicon-based two-terminal TRAM (2-T TRAM) suffers from short retention times, limiting its practical application. This study proposes a localized lightly doped base (LLDB) structure to address this issue. By optimizing the LLDB doping concentration to 1.0 × 1017 cm−3, the retention time is improved by 138.49% to 601 ms, while energy consumption is reduced by 3.76% to 149.15 pJ. Simulation results confirm that the LLDB structure effectively suppresses Shockley-Read-Hall recombination, thereby enhancing retention characteristics and energy efficiency. These improvements make the LLDB-enhanced 2-T TRAM a promising candidate for high-performance, compact memory applications.

Original languageEnglish
Article number084001
JournalJapanese Journal of Applied Physics
Volume64
Issue number8
DOIs
StatePublished - 1 Aug 2025

Bibliographical note

Publisher Copyright:
© 2025 The Japan Society of Applied Physics. All rights, including for text and data mining, AI training, and similar technologies, are reserved.

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

  1. SDG 7 - Affordable and Clean Energy
    SDG 7 Affordable and Clean Energy

Keywords

  • DRAM
  • local doping
  • reliability
  • thyristor

Fingerprint

Dive into the research topics of 'Design and characterization of two-terminal thyristor dynamic random-access memory with a localized lightly-doped base insertion for improving data retention'. Together they form a unique fingerprint.

Cite this