J. Mater. Sci. Technol. ›› 2021, Vol. 95: 203-212.DOI: 10.1016/j.jmst.2021.03.079

• Research Article • Previous Articles     Next Articles

Controlled multilevel switching and artificial synapse characteristics in transparent HfAlO-alloy based memristor with embedded TaN nanoparticles

Chandreswar Mahataa, Hassan Algadib, Muhammad Ismaila, Daewoong Kwonc, Sungjun Kima,*()   

  1. aDivision of Electronics and Electrical Engineering, Dongguk University, Seoul 04620, Republic of Korea
    bDepartment of Electrical Engineering, Faculty of Engineering, Najran University, P.O. Box 1988, Najran 11001, Saudi Arabia
    cDepartment of Electrical Engineering, Inha University, Incheon, Republic of Korea
  • Received:2021-02-25 Revised:2021-03-26 Accepted:2021-03-30 Published:2021-12-30 Online:2021-06-01
  • Contact: Sungjun Kim
  • About author:* E-mail address: sungjun@dongguk.edu (S. Kim).

Abstract:

Atomic layer deposition technique has been used to prepare tantalum nitride nanoparticles (TaN-NPs) and sandwiched between Al-doped HfO2 layers to achieve ITO/HfAlO/TaN-NP/HfAlO/ITO RRAM device. Transmission electron microscopy along with energy dispersive spectroscopy confirms the presence of TaN-NPs. X-ray photoelectron spectroscopy suggests that part of TaN converted to tantalum oxynitride (TaOxNy) which plays an important role in stable cycle-to-cycle resistive switching. Charge trapping and oxygen vacancy creation were found to be modified after the inclusion of TaN-NPs inside RRAM structure. Also, HfAlO/TaOxNy interface due to the presence TaN-NPs improves the device-to-device switching reliability by reducing the probability of random rupture/formation of conductive filaments (CFs). DC endurance of more than 103 cycles and memory data retention up to 104 s was achieved with an insignificant variation of different resistance states. Multilevel conductance was attained by controlling RESET voltage with stable data retention in multiple states. The volatile threshold switching was monitored after controlling the CF forming at 200 nA current compliance with high selectivity of ~103. Synaptic learning behavior has been demonstrated by spike-rate-dependent plasticity (SRDP). Reliable potentiation and depression processes were observed after the application of suitable negative and positive pulses which shows the capability of the TaN-NPs based RRAM device for transparent synaptic devices.

Key words: RRAM, ALD TaN-nanoparticles, Threshold switching, Spike-rate-dependent plasticity, Multilevel conductance, Synaptic properties