Elsevier

Acta Materialia

Volume 249, 1 May 2023, 118809
Acta Materialia

Full length article
Diffusion-assisted displacive transformation in Yttrium-doped Sb2Te3 phase change materials

https://doi.org/10.1016/j.actamat.2023.118809Get rights and content

Abstract

Phase transformations between cubic and rhombohedral states of phase-change chalcogenide materials have been considered to result in small resistance drift, fast read/write speeds, and low power consumption in phase-change memory devices. However, it is difficult to observe such dynamic processes by experiments. Especially, the ultra-fast speed of, and ultra-small area affected during, the displacive transformation involved in the phase transformations, lead to a lack of direct evidence for studying the transformation mechanism. Here, we directly observed the reversible displacive transformations in Yttrium-doped Sb2Te3 (Y-Sb2Te3) by ab initio molecular-dynamics (AIMD) simulations. The forward displacive transformation, from the cubic to the rhombohedral phase, can be spontaneously realized through a two-step transformation, the shearing, and contraction between Te-Sb-Te-Sb-Te building blocks. On the other hand, the backward transformation, from the rhombohedral to the cubic phase, requires high-temperature heating (above 900 K) and atomic diffusion. The diffusion of cation-site Sb atoms into interstitial sites, located in the interlayers of building blocks, triggers the backward transformation; thus, it is diffusion-assisted. Moreover, the enhancement of thermostability of the cubic phase by Y dopants promotes the backward transformation. This study can improve the understanding of the displacive transformations in phase-change materials and may offer new insights into developing phase-change memory devices based on Sb2Te3.

Introduction

Pseudo-binary GeTe-Sb2Te3 (GST) compounds are widely used as phase-change memory (PCM) materials [1], thermoelectric materials [2], and topological insulators [3]. Especially, in the field of phase-change memory, GST materials have been regarded as promising candidates for multi-level storage [4,5] and neuromorphic computing devices [6,7]. Usually, GST materials have three solid phases, i.e., amorphous, metastable cubic, and stable trigonal (rhombohedral) phases [8]. The fast and reversible diffusional transformation between amorphous and cubic phases in GST materials has been successfully employed to achieve non-volatile data storage [9]. However, the amorphous phase suffers from severe time-dependent resistance drift [10], hindering the implementation of GST in advanced multi-level data storage and neuromorphic devices. Fortunately, the recently investigated reversible phase transformation between two crystalline phases in GST materials [11] provides a chance to facilitate its further development in such advanced devices.

The phase transformation occurs between cubic and trigonal (rhombohedral) phases in GST materials without the participation of the amorphous phase, the resistance drift caused by spontaneous relaxation of the amorphous phase can thereby be completely avoided. Besides, significant changes in the crystal structure and corresponding physical properties occur during the transformation, giving it the potential for being used as a non-volatile phase-change memory [12,13]. It has been found that the phase transformation between two crystalline phases of GST is a two-step process, i.e., the formation of vacancy layers and the displacive transformation [11]. The displacive transformation in the phase transformation is a critical step, which is a kind of solid-state phase transition that is characterized by a cooperative movement of large numbers of atoms, such as dilation, shear, and shuffle, and frequently observed in steels, nonferrous alloys, and ceramics [[14], [15], [16]]. Such a type of structural transformation does not involve the long-range diffusion of atoms, and usually exhibits advantages in terms of speed and energy consumption over diffusional types of transformation. In GST, the forward displacive transformation, from the cubic to the trigonal (rhombohedral) phase, can be achieved by simply annealing [17], [18], [19], [20], whereas the backward displacive transformation usually needs external stimulation, such as ion irradiation [21], electron-beam irradiation [11], or UV ns-laser pulses [22]. Interestingly, a reversible displacive transformation was achieved in an Yttrium-doped Sb2Te3 (Y-Sb2Te3) device by Joule heating [23,24]. However, the speed of this type of transformation is too fast to experimentally characterize in real time. The lack of knowledge about the displacive-transformation mechanism therefore impedes the optimization of cycling endurance and hinders the implementation of this material system in memory devices.

Recent theoretical calculations have provided a new way to study displacive transformations. For example, using density-functional theory (DFT) calculations, the thermodynamic conditions for triggering displacive transformations have been investigated [25], and the energy barriers for transformation have been quantitatively calculated by combining with the nudged-elastic-band (NEB) method [11,26]. However, such remedies only obtain certain thermodynamic results from static calculations by studying artificial migration paths and cannot provide information about the atom-dynamical process of displacive transformations, neglecting the kinetics of the displacive transformations. Therefore, in-situ observation and dynamic reproduction of displacive transformations are urgently needed.

In this work, we utilized ab initio molecular-dynamics (AIMD) simulations to investigate the displacive transformations in Y-Sb2Te3, providing evidence of a real dynamical process and a convincing transformation mechanism. Combining with DFT calculations, we explored the requirements and unraveled the key role of atomic diffusion in the displacive transformations. Moreover, the effect of Y dopants on the displacive transformations is also revealed. This study improves our understanding of the displacive transformation mechanisms. The present findings could also be extended to other systems.

Section snippets

Computational details

Models of cubic and rhombohedral Y-Sb2Te3 and undoped Sb2Te3 were constructed using orthogonal supercells, of which the lattice parameters were a = 14.59 Å, b = 16.83 Å, c = 63.07 Å for the cubic structure, and a = 15.05 Å, b = 17.39 Å, c = 61.82 Å for the rhombohedral structure. The directions of z-axis were along with the [111]cubic and [0001]rhombohedral for cubic and rhombohedral supercells, respectively. More details on the construction of the orthogonal supercells are shown in Fig. S1.

Forward displacive transformation from cubic to rhombohedral Y-Sb2Te3

Metastable cubic Y-Sb2Te3 has various configurations, according to the random distribution of the 33.3% vacancies at the cation sites. During annealing, these vacancies spontaneously migrate to the cation layers, forming ordered vacancy layers. With the gradual formation of the vacancy layers, the forward displacive transformation from cubic to rhombohedral Sb2Te3 happens. The forward transformation is mainly completed by the collective migration of building blocks of Te-Sb-Te-Sb-Te [26]. To

Conclusions

In this work, we have directly observed the reversible displacive transformations between metastable cubic and stable rhombohedral Y-Sb2Te3 by AIMD simulations. The forward displacive transformation was spontaneously realized with the displacement of building blocks, which was divided into two steps: shearing and contraction. In the shearing process, the stacking order of building blocks was changed by three equivalent shear vectors. After the shearing process, the interlayer distance between

Declaration of Competing Interest

The authors declare no competing interests.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was financially supported by the National Natural Science Foundation of China (Grant No. 51872017). The authors also acknowledge the support of the high-performance computing (HPC) resources at Beihang University.

References (61)

  • Y. Gan et al.

    Prediction of thermoelectric performance for layered IV-V-VI semiconductors by high-throughput ab initio calculations and machine learning

    npj Comput. Mater.

    (2021)
  • B. Sa et al.

    Topological insulating in GeTe/Sb2Te3 phase-change superlattice

    Phys. Rev. Lett.

    (2012)
  • K. Ding et al.

    Phase-change heterostructure enables ultralow noise and drift for memory operation

    Science

    (2019)
  • W. Zhang et al.

    Designing crystallization in phase-change materials for universal memory and neuro-inspired computing

    Nat. Rev. Mater.

    (2019)
  • S. Yu

    Neuro-inspired computing with emerging nonvolatile memories

    Proc. IEEE

    (2018)
  • M. Xu et al.

    Recent advances on neuromorphic devices based on chalcogenide phase-change materials

    Adv. Funct. Mater.

    (2020)
  • T. Siegrist et al.

    Disorder-induced localization in crystalline phase-change materials

    Nat. Mater.

    (2011)
  • M. Wuttig et al.

    Phase-change materials for rewriteable data storage

    Nat. Mater.

    (2007)
  • S. Mori et al.

    Reversible displacive transformation in MnTe polymorphic semiconductor

    Nat. Commun.

    (2020)
  • A.M. Alper et al.

    Materials science and technology

    Martensitic Transformation

    (1978)
  • R.W. Cahn et al.

    Pergamon materials series

    Classification of Phase Transformations

    (2007)
  • F. Rao et al.

    Reducing the stochasticity of crystal nucleation to enable subnanosecond memory writing

    Science

    (2017)
  • M. Zhu et al.

    One order of magnitude faster phase change at reduced power in Ti-Sb-Te

    Nat. Commun.

    (2014)
  • Y. Xu et al.

    Materials screening for disorder-controlled chalcogenide crystals for phase-change memory applications

    Adv. Mater.

    (2021)
  • Y. Zheng et al.

    Direct observation of metastable face-centered cubic Sb2Te3 crystal

    Nano Res.

    (2016)
  • S.M.S. Privitera et al.

    Structural and electronic transitions in Ge2Sb2Te5 induced by ion irradiation damage

    Phys. Rev. B

    (2016)
  • M. Behrens et al.

    Structural transitions in Ge2Sb2Te5 phase change memory thin films induced by nanosecond UV optical pulses

    Materials

    (2020)
  • B. Liu et al.

    Ultrahigh overall-performance phase-change memory by Yttrium dragging

    J. Mater. Chem. C

    (2023)
  • W. Zhang et al.

    Role of vacancies in metal-insulator transitions of crystalline phase-change materials

    Nat. Mater.

    (2012)
  • Z. Li et al.

    Yttrium-doped Sb2Te3: A promising material for phase-change memory

    ACS Appl. Mater. Interfaces

    (2016)
  • Cited by (5)

    View full text