Introduction

Since Mead1 performed the first trial to mimic the biological neural networks (BNNs) of the brain in the 1980s, extensive effort has been made to emulate BNNs by utilizing various synaptic devices2,3,4,5,6,7,8,9,10,11,12,13. In order to properly perform signal processing in BNNs, synaptic plasticity and its timing-dependent learning algorithm are two key computational parameters14. However, the scope of initial research on synaptic devices has been mainly focused on simply mimicking synaptic dynamics, such as long-term potentiation/depression (LTP/LTD)15, short-term plasticity16, and spike-timing-dependent plasticity (STDP), with filament-forming switching devices2,3 or phase change memory devices4 in array crossbar structures. Such synaptic devices benefit from superior integration capability and energy efficiency compared to the synaptic devices with lateral transistor structure. However, they suffer from nonlinear potentiation/depression characteristics17, small differences between conduction states18, and insufficient conductance states19,20. Organic5 and carbon nanotube6 transistors have recently shown their suitability for emulating of synaptic dynamics with better linear potentiation/depression characteristics and a larger number of usable conductance states (>500 distinct states), in addition to high energy efficiency (<10 pJ switching energy)21.

Beyond this device-level emulation of synaptic dynamics, synaptic devices have also been used to build artificial neural networks (ANNs). Subsequently, pattern recognition tasks2,3,4,5,6,7,8 have been verified by these ANNs, where winner-take-all6 and perceptron networks3 are usually applied. However, none of the previous studies have tried to functionalize devices beyond their synaptic functions, for example, by merging them with biometric sensing elements such as vision, auditory, and olfactory sensors. Current works have only demonstrated signal processing in the cerebral cortex with either binary2,3 or grayscale MNIST (Modified National Institute of Standards and Technology) datasets6. The functional integration of synaptic devices with biometric sensing elements is expected to provide new opportunities for the implementation of neural networks that comprise neural sensing and training functions, thereby enabling power-efficient pattern recognition task for complex (e.g., color-mixed and voice-mixed) patterns. In this article, we demonstrate an optic-neural synaptic (ONS) device that features synaptic and optical-sensing functions. This ONS device was fabricated on a van der Waals (vdW) heterostructure (h-BN/WSe2)22,23, which does not have interfacial defects and thereby allows modulating a number of interfacial traps for achieving the synaptic functionalities. Through an optic-neural network (ONN) formed by these ONS devices, the colored and color-mixed pattern recognition capability of the human vision system is emulated. In particular, our synaptic device is investigated and compared with other devices reported heretofore, in terms of weight update linearity, number of usable conduction states, stability of each state, and energy efficiency (see Supplementary Table 1).

Results

h-BN/WSe2 optic-neural synaptic device

As shown in Fig. 1, beyond the typical emulation of synaptic dynamics, we have fabricated an ONS device by integrating a synaptic device with an optical-sensing device on the same h-BN/WSe2 heterostructure. The operation of the vdW synaptic device is based on the trap** or de-trap** of electrons in the weight control layer (WCL) on h-BN, which modulates the WSe2 channel conductivity (weight of the synapse). The vdW synaptic device will be discussed in detail in the following subsections (see below). Here, the h-BN/WSe2 photodetector resistance is modulated as a function of the incident wavelength (see Fig. 1a). The additional analysis on the h-BN and WSe2 flakes is described in Supplementary Figure 1. The h-BN/WSe2 interface was confirmed to be clearly formed without polymer residues24, owing to our transfer method based on adhesion energy engineering (see Supplementary Figure 2). Because more optical absorption occurs at a shorter wavelength (see Supplementary Figure 3a), a shorter wavelength of light decreases the resistance of the optical-sensing device under a constant drain bias (see Supplementary Figure 3b, c). This reduction in resistance means that more carriers are generated in WSe2 so that the density of carriers trapped in the WCL increases. This subsequently allows the adjustment of the synaptic dynamic properties of the ONS device according to light wavelength conditions. Here, the optical-sensing device was fabricated on the same h-BN/WSe2 heterostructure where the synaptic device was formed22,25. This was done in order for the optical-sensing device to have the proper series resistance, which is comparable to that of the synaptic device, and to enable changing the synaptic properties. If the optical-sensing device has too much larger or smaller resistance value compared to the synaptic device, the resistance control in optical-sensing device by adjusting illumination condition will be ineffective to change the synaptic properties of the ONS device. To identify the synaptic dynamics of the ONS device under different light wavelength conditions, we investigated the device’s synaptic plasticity, the postsynaptic current (PSC), and LTP/LTD. The optical-sensing device was exposed to red (λ = 655 nm), green (λ = 532 nm), and blue (λ = 405 nm) lights with an optical power of 6 mW cm−2, and we then confirmed the PSC characteristics after applying ±1 V of voltage pulse (Vpulse) to the synaptic device (see Fig. 1b and Supplementary Figure 4 for the measurement set-up). When exposed to a light with shorter wavelength, the magnitude of the synaptic current increased from 1.55 nA to 2.29 μA each, where the extracted conductance change (ΔG) also presented a significant increase from 0.78 nS to 0.74 μS by three orders of magnitude. As shown in Fig. 1c, the LTP/LTD curves were distributed in different conductance regions according to the light wavelength, but they maintained the curved shape related to nonlinearity. If the conductance state varies too nonlinearly, similar to a square root function, additional periphery circuits are typically required to manage the wide range of conductance state changes26,27. In addition, a nonlinear conductance response imposes difficulty in achieving a sufficient number of usable conductance states because a voltage pulse disables the induction of a sufficiently large conductance difference in the saturated region. Here, we defined effective conductance states as those in which ΔG exceeds a certain percentage of Gmax/Gmin (thresholdΔG as described in detail in Supplementary Figure 5). Based on thresholdΔG, we excluded the pulse signals that induced states with small ΔG unimportant for neuromorphic computing, thus reducing power consumption during computation. For the quantitative analysis of the LTP/LTD characteristics of our ONS device, we extracted a nonlinearity28 (see Fig. 1d and Supplementary Figure 6) and the number of effective conductance states (see Fig. 1e). For case 1 with a pulse amplitude of ±0.3 V, the nonlinearity was maintained at the levels of 1.5 (potentiation)/1.5 (depression), regardless of the different light conditions, including a no-light condition. In the remaining pulse conditions with pulse amplitudes of ±0.5 V (case 2) and ±1 V (case 3), the nonlinearity (2.6/6 for case 2 and 2.7/18 for case 3) was higher than the values of case 1; however, they were also independent of the incident light wavelength. Moreover, the number of effective conductance states did not significantly change despite the different light conditions (82–99 for case 1, 75–82 for case 2, and 74–82 for case 3), where thresholdΔG was set to 0.3%. Overall, we observed that (i) the conductance ranges in which the LTP/LTD synaptic characteristics were clearly different as a function of the illuminating wavelength; (ii) the LTP/LTD characteristics, such as the linearity and the number of effective conductance states that significantly affect the pattern recognition rate, were almost independent of the light wavelength. Because an optical power is another important information about the optical input, the ONS device eventually needs to be distinctly operated according to the power as well as the wavelength of the optical input. To account for environmental challenges in real-world applications, we exposed our devices to ambient conditions over an extended period of time (see Supplementary Figure 7). Even after 1 month under such conditions, our device proved to be unaffected and worked properly, highlighting their potential for future applications.

Fig. 1
figure 1

Integration of the h-BN/WSe2 optic-neural synaptic device. a Schematic of the human optic nerve system, the h-BN/WSe2 synaptic device integrated with h-BN/WSe2 photodetector, and the simplified electrical circuit for the ONS device. Here, the light sources were dot lasers with wavelengths of 655 nm (red), 532 nm (green), and 405 nm (blue) with a fixed power density (P) of 6 mW cm−2 for all wavelengths. b Excitatory and inhibitory postsynaptic current characteristics and extracted conductance changes of the h-BN/WSe2 ONS device under different light conditions (no light and RGB). c Long-term potentiation and depression curves under different light conditions, where the synaptic device is controlled using input pulses with an amplitude of 0.3 V. d, e Nonlinearity magnitude (d) and the number of effective conductance states (thresholdΔG = 0.3%) (e), which were extracted for different wavelengths

Full size image

Characteristics of h-BN/WSe2 synaptic device

The synaptic plasticity of the ONS device derives from the operation of the vdW synaptic device with the WCL. The first step toward achieving this synaptic device was to create a charge trap** layer on top of h-BN for adjustment of the WSe2 channel conductivity29. By forming the WCL on h-BN with an O2 plasma treatment, we implemented the vdW synaptic device, as shown in Fig. 2a. The hysteresis characteristic was observed in the current between the presynaptic and postsynaptic terminals, which was dependent on the voltage applied to the synaptic cleft terminal (VSCT). This occurs because charges trapped in the WCL partially screen VSCT and thereby influence the current flow through the synaptic device (see Supplementary Figure 8). Figure 2b shows cross-sectional transmission electron microscopy (X-TEM) images taken near the WCL of the synaptic device. After the O2 plasma treatment was performed for 5 min, an 11.3 nm-thick WCL was observed on the h-BN surface region. In addition, as seen in Fig. 2c, high-resolution electron energy loss spectroscopy (EELS) and energy-dispersive X-ray spectroscopy (EDS) map** analyses were performed to investigate the atomic compositions of the WSe2/WCL/h-BN region. In the WCL, signals related to O and B elements clearly appeared, but N, W, and Se signals were not present, indicating that the WCL consisted of the oxidized boron transformed from h-BN. As expected, this WCL was not observed in the EELS and EDS map** images for the control sample that was not subjected to the O2 plasma treatment (refer to Supplementary Figure 9). To analyze the dynamic nature of the conductivity (see Fig. 2e), voltage pulses with 0.1 and 1 V amplitudes were applied to the synaptic device. In the case where the voltage pulse was 0.1 V, the transient electrons trapped in fast traps recovered, and the abruptly increased current returned to its initial value after 8 s, resulting in no persistent ΔG. On the other hand, when a pulse with 1 V amplitude was applied, a positive ΔG occurred because electrons were captured and held in the traps deeper inside the WCL. Calculation of the trap density in the WCL gave a value of approximately 7.4 × 10−17 cm−3. When extrapolated to current memory cell sizes, the number of trapped electrons was on the same order of electrons stored on a floating gate of a flash memory cell (see Supplementary Figure 10) (http://www.itrs2.net/). The Fig. 2e shows the ratio of the remaining charges in fast traps and slow traps at 1 s after the pulse, according to the pulse amplitude. When a pulse with higher amplitude was applied, the contribution of slow traps to the increase in PSC increased noticeably, consequently exceeding that of fast traps. As shown in Fig. 2f, by applying a pulse with higher amplitude, more charges are expected to be trapped in slow traps that will not recover at room temperature, which subsequently changes the synaptic conductance of WSe2. We also note that the contribution of fast traps to the conductance change diminishes as time elapses after the pulse (see Supplementary Figure 11). The magnitude of ΔG increased until the O2 process time reached the 5 min point and then saturated beyond that time, where the |ΔG| values were similar in both excitatory and inhibitory synaptic responses. We also estimated the switching energy of the synaptic device as a function of the O2 plasma process time (see Fig. 2g). Regardless of the process time related to the amount of traps in the WCL, the switching energy was approximately 66 fJ (at 0.3 V of presynaptic pulse) and it increased to 532 fJ (at 1 V of pulse), which is comparable to those of previously reported synaptic devices (see Supplementary Table 1b). ΔG is a function of the amplitude and duration of the presynaptic pulse for the weight update, such that the dissipated energy per event can be determined by P = I × V × tduration5.

Fig. 2
figure 2

Structure and operating mechanism of the h-BN/WSe2 synaptic device. a Functional/structural/architectural comparison of biological synapse with our synthetic WSe2/WCL/h-BN synaptic device. b X-TEM image of the WSe2/WCL/h-BN structure, and the high-resolution images corresponding to the WSe2/WCL and WCL/h-BN interfaces. c, d EELS (c) and EDS (d) map** images obtained on the cross-section of the WSe2/WCL/h-BN structure. e Current relaxation curves after pulse amplitudes of 0.1 V and 1 V, and contribution ratio from unrecovered electrons in fast traps and slow traps at 1 s after the pulse. f Illustration of energy band diagrams after pulse and after de-trap** of carriers in fast traps. g Change in postsynaptic conductance and the switching energy measured as a function of O2 plasma process time. Here, all the Vpulse were applied with a duration of 10 ms

Full size image

For neuromorphic computing based on deep neural networks (DNNs) with the well-studied back-propagation learning algorithm30, as mentioned above, synaptic devices are required to have the following characteristics: linear conductance responses17,19,Full size image