ABSTRACT A unique optoelectronic synaptic device has been developed, leveraging the negative photoconductance property of a single-crystal material system called Cs2CoCl4. This device

exhibits a simultaneous volatile resistive switching response and sensitivity to optical stimuli, positioning Cs2CoCl4 as a promising candidate for optically enhanced neuromorphic

applications. The human brain is an extraordinary computer capable of processing fault-tolerant information and making decisions based on dynamic and complex situations. Unlike conventional

computers, which struggle with processing probabilistic, noisy, and inconsistent data, the brain performs these tasks with remarkable accuracy and power efficiency. Central to its superior

recognition, and decision-making. Inspired by the cognitive prowess of the brain, a new generation of technology, namely artificial synapses and neurons, has emerged in recent years1,2.

These are used to construct artificial neural networks, as illustrated in Fig. 1b, where they replicate the behavior of biological counterparts. To create artificial synapses and neurons, an

insulating or semiconducting layer is often placed between two metal electrodes, making a resistive switching device, now frequently called a memristor, as shown in Fig. 1c. Their core

functionality lies in altering electrical resistances, which can be permanent or transient depending on the choice of materials, device geometry, and operating conditions. Many artificial

synaptic and neuronal devices are based on oxides, chalcogenides, phase-change materials, and perovskites3,4,5. They primarily rely on electrical stress to process, store, and erase

information. Specifically, synaptic functions such as adjusting synaptic weights (termed as ‘potentiation’ and “depression”), as well as mimicking the spiking behavior of neurons (see Fig.

1e, f, respectively), are often achieved by applying voltage or current pulses. With the emergence of artificial intelligence and neurorobotics, a great interest has grown in developing

devices that can also respond to optical stimuli. This dual responsivity is highly advantageous for image and pattern recognition, as well as in-sensor and near-sensor computing. It offers

cost-effectiveness and energy efficiency by eliminating the need for external photo (light) sensors to capture visual (or optical) information, thus simplifying device architecture and

reducing latency, material consumption, and processing steps. Despite these promising prospects, only a few material systems have demonstrated efficient optical response while maintaining

reliable resistive switching properties, and many of them are only sensitive to a narrow band of electromagnetic spectrum such as ultraviolet illumination. This limitation greatly hinders

the development of optoelectronic neuromorphic technology. In a conventional optoelectronic synaptic device, optical stimulation is used to trigger potentiation (a gradual increase in device

conductance), and electrical stimulation is applied to induce synaptic depression (gradually restoring the device to its previous conductance or resistance state), as depicted in the middle

panel of Fig. 1e. This is possible due to the positive-photoconductance property of the photoactive material, where the device conductance increases with optical illumination. Conversely,

achieving optically triggered synaptic depression (bottom panel of Fig. 1e) is more challenging, often requiring materials with negative-photoconductance characteristics. Unfortunately,

memristor devices offering reconfigurable resistance states and negative photoconductivity are relatively scarce, making such devices a hurdle in the research field. Addressing this

scarcity, Huifang Jiang and colleagues from Zhengzhou University and Beijing Institute of Technology have introduced an exciting material system: cesium cobalt chlorine (Cs2CoCl4) in the

single-crystal form that can be used as an optoelectronic synaptic device6. Their findings, published in Light: Science & Applications, show simultaneous volatile resistive switching and

negative photoconductance, making it a unique material system to exhibit such a response. The authors demonstrated a Cu/Cs2CoCl4/ITO device that shows synaptic functions mediated by

electrically triggered Cu migration (potentiation) and negative photoconductance of Cs2CoCl4 (depression). Notably, the device is sensitive to broad optical stimuli, ranging from 265 to 780 

nm (ultraviolet to visible light), and demonstrates a highly efficient specific detectivity on the order of 10¹² Jones, which makes it a promising candidate for processing a broad range of

optical information. Through extensive experimental and theoretical analyses, the authors identified that the negative photoconductance response arises from the trapping of photo-excited

electrons in Cs2CoCl4. The authors further demonstrated the practical application of their device by constructing a simulated artificial neural network, which successfully recognized

handwritten digital images with high accuracy. A notable feature of the reported device is its volatile threshold switching response, where the threshold voltage can be engineered using

optical illumination, as illustrated in Fig. 1d. Remarkably, this characteristic persists even when the Cu electrode is replaced with an inert Au electrode, indicating that the volatile

switching mechanism is primarily governed by the conduction in Cs2CoCl4. Note that the resistive switching in devices with Cu or Ag electrodes is mainly governed by the electric

field-induced migration of metal ions, and the dielectric material plays a passive role in the conduction process. The authors have highlighted that the negative photoconductance of Cs2CoCl4

increases the threshold voltage of the device due to the increase of overall device resistance, in contrast to the behavior of positive photoconductance-based volatile switching devices,

where the threshold voltage typically decreases7. These unique properties open new pathways for developing advanced neuromorphic devices that integrate electrical and optical

functionalities, promising more versatile, cost- and energy-efficient systems for future applications such as artificial intelligence, machine vision, and neuro-robotics8. This device can

also be used beyond neuromorphic computing, including as an access device (selector) for ReRAM-based digital memories, true random number generator, creating physical unclonable functions

for application in information security, etc.9. One foreseeable application of this device is the implementation of a light-tuneable artificial neuron that operates via frequency modulation,

as schematically shown in Fig. 1f. This neuron could be utilized for in-sensor spike encoding, which is helpful for image segmentation, pattern recognition, and motion detection7,10,11,12.

Similarly, it can also be used as a thermoreceptor by harnessing temperature-dependent threshold switching characteristics of the device13, which might be useful for applications such as

robotic touch and human-machine interaction. As this study advances the growing field of photo-tunable volatile memristive switching devices, it opens up several future research

opportunities. These include exploring the reliability and reproducibility of such characteristics at the nanoscale, expanding sensitivity to the near-infrared and short-wave infrared

spectrum through bandgap or defect engineering, improving the linearity of synaptic potentiation and depression through materials and interface engineering, and studying frequency tunability

in artificial neurons under both electrical and optical stimuli, to name a few. Overall, this work offers a fresh perspective on exploring new materials and strategies for developing

optoelectronic devices with cognitive abilities. By adopting this material, we are moving closer to mimicking biological neurons and synapses using a single material system, marking a

significant step forward in developing hardware-based neuromorphic computing. REFERENCES * Kumar, S., Williams, R. S. & Wang, Z. W. Third-order nanocircuit elements for neuromorphic

engineering. _Nature_ 585, 518–523 (2020). Article  ADS  MATH  Google Scholar  * Indiveri, G. et al. Integration of nanoscale memristor synapses in neuromorphic computing architectures.

_Nanotechnology_ 24, 384010 (2013). Article  MATH  Google Scholar  * Ielmini, D. Brain-inspired computing with resistive switching memory (RRAM): devices, synapses and neural networks.

_Microelectron. Eng._ 190, 44–53 (2018). Article  Google Scholar  * Zhu, J. D. et al. A comprehensive review on emerging artificial neuromorphic devices. _Appl. Phys. Rev._ 7, 011312 (2020).

Article  ADS  Google Scholar  * Wright, C. D. et al. Beyond von-Neumann computing with nanoscale phase-change memory devices. _Adv. Funct. Mater._ 23, 2248–2254 (2013). Article  MATH 

Google Scholar  * Jiang, H. et al. Simultaneous achieving negative photoconductivity response and volatile resistive switching in Cs2CoCl4 single crystals towards artificial optoelectronic

synapse. _Light Sci. Appl._ 13, 316 (2024). Article  MATH  Google Scholar  * Nath, S. K. et al. Optically tunable electrical oscillations in oxide‐based memristors for neuromorphic

computing. _Adv. Mater._ 36, 2400904 (2024). Article  MATH  Google Scholar  * Cho, S. W. et al. Progress of materials and devices for neuromorphic vision sensors. _Nano-Micro Lett._ 14, 203

(2022). Article  ADS  MATH  Google Scholar  * Wang, R. P. et al. Recent advances of volatile memristors: devices, mechanisms, and applications. _Adv. Intell. Syst._ 2, 2000055 (2020).

Article  Google Scholar  * Ahsan, R. et al. Ultralow power in-sensor neuronal computing with oscillatory retinal neurons for frequency-multiplexed, parallel machine vision. _ACS Nano_ 18,

23785–23796 (2024). Article  MATH  Google Scholar  * Li, Z. Y. et al. Crossmodal sensory neurons based on high-performance flexible memristors for human-machine in-sensor computing system.

_Nat. Commun._ 15, 7275 (2024). Article  MATH  Google Scholar  * Xi, Z. Y. et al. Retina-like neuromorphic visual sensor for sensing broad-spectrum ultraviolet light. _Adv. Opt. Mater._

2024, 2402193 (2024). Article  Google Scholar  * Bonagiri, A. et al. Biorealistic neuronal temperature-sensitive dynamics within threshold switching memristors: toward neuromorphic

thermosensation. _ACS Appl. Mater. Interfaces_ 16, 31283–31293 (2024). Article  Google Scholar  Download references AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * School of Photovoltaic and

Renewable Energy Engineering, University of New South Wales (UNSW Sydney), Kensington, NSW, Australia Shimul Kanti Nath Authors * Shimul Kanti Nath View author publications You can also

search for this author inPubMed Google Scholar CORRESPONDING AUTHOR Correspondence to Shimul Kanti Nath. RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons

Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original

author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the

article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use

is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit

http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Nath, S.K. A light-driven device for neuromorphic computing. _Light Sci Appl_ 14,

37 (2025). https://doi.org/10.1038/s41377-024-01722-9 Download citation * Published: 08 January 2025 * DOI: https://doi.org/10.1038/s41377-024-01722-9 SHARE THIS ARTICLE Anyone you share the

following link with will be able to read this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer

Nature SharedIt content-sharing initiative