In this work, we propose three types of resistive switching behaviors by controlling operation conditions. We confirmed well-known filamentary switching in Al2O3-based resistive switching memory using the conventional device working operation with a forming process. Here, filamentary switching can be classified into two types depending on the compliance current. On top of that, the homogeneous switching is obtained by using a negative differential resistance effect before the forming or setting process in a negative bias. The variations of the low-resistance and high-resistance states in the homogeneous switching are comparable to the filamentary switching cases. However, the drift characteristics of the low-resistance and high-resistance states in the homogeneous switching are unstable with time. Therefore, the short-term plasticity effects, such as the current decay in repeated pulses and paired pulses facilitation, are demonstrated when using the resistance drift characteristics. Finally, the conductance can be increased and decreased by 50 consecutive potentiation pulses and 50 consecutive depression pulses, respectively. The linear conductance update in homogeneous switching is achieved compared to the filamentary switching, which ensures the high pattern-recognition accuracy.

Here, we propose a Pt/HfO2/TaOx/TiN artificial synaptic device that is an excellent candidate for artificial synapses. First, XPS analysis is conducted to provide the dielectric (HfO2/TaOx/TiN) information deposited by DC sputtering and atomic layer deposition (ALD). The self-rectifying resistive switching characteristics are achieved by the asymmetric device stack, which is an advantage of the current suppression in the crossbar array structure. The results show that the programmed data are lost over time and that the decay rate, which is verified from the retention test, can be adjusted by controlling the compliance current (CC). Based on these properties, we emulate bio-synaptic characteristics, such as short-term plasticity (STP), long-term plasticity (LTP), and paired-pulse facilitation (PPF), in the self-rectifying I–V characteristics of the Pt/HfO2/TaOx/TiN bilayer memristor device. The PPF characteristics are mimicked by replacing the bio-stimulation with the interval time of paired pulse inputs. The typical potentiation and depression are also implemented by optimizing the set and reset pulse. Finally, we demonstrate the natural depression by varying the interval time between pulse inputs.

No abstract

In this study, we fabricate and characterize a Ti/TiO2/Si device with different dopant concentrations on a silicon surface for neuromorphic systems. We verify the device stack using transmission electron microscopy (TEM). The Ti/TiO2/p++Si device exhibits interface-type bipolar resistive switching with long-term memory. The potentiation and depression by the pulses of various amplitudes are demonstrated using gradual resistive switching. Moreover, pattern-recognition accuracy (>85%) is obtained in the neuromorphic system simulation when conductance is used as the weight in the network. Next, we investigate the short-term memory characteristics of the Ti/TiO2/p+Si device. The dynamic range is well-controlled by the pulse amplitude, and the conductance decay depends on the interval between the pulses. Finally, we build a reservoir computing system using the short-term effect of the Ti/TiO2/p+Si device, in which 4 bits (16 states) are differentiated by various pulse streams through the device that can be used for pattern recognition.

Memristor-type synaptic devices that can effectively emulate synaptic plasticity open up new directions for neuromorphic hardware systems. Here, a double high-k oxide structured memristor device (TaOx/HfO2) was fabricated, and its synaptic applications were characterized. Device deposition was confirmed through TEM imaging and EDS analysis. During the forming and set processes, switching of the memristor device can be divided into three types by compliance current and cycling control. Filamentary switching has strengths in terms of endurance and retention, but conductance is low. On the other hand, for interface-type switching, conductance is increased, but at the cost of endurance and retention. In order to overcome this dilemma, we proposed pseudo interface-type switching, and obtained excellent retention, decent endurance, and a variety of conductance levels that can be modulated by pulse response. The recognition rate calculated by the neural network simulation using the Fashion Modified National Institute of Standards and Technology database (MNIST) dataset, and the measured conductance values show that pseudo interface-type switching produces results that are similar to those of an interface-type device.