Elsevier

Materials Today Physics

Volume 27, October 2022, 100812
Materials Today Physics

The rise of AI optoelectronic sensors: From nanomaterial synthesis, device design to practical application

https://doi.org/10.1016/j.mtphys.2022.100812Get rights and content

Highlights

  • Nanomaterials employed in optoelectronic sensing technologies are generalized systematically.

  • Several application scenarios are reviewed, demonstrating great development and prosperity of intelligent application systems.

  • The future development trends of AI optoelectronic sensing technology are toward all-weather, versatile, and intelligent systems.

Abstract

We have seen the significant influence of the information revolution brought about by optoelectronic sensing technologies on human civilization over the last few decades, especially in all aspects of big data, cloud servers, closed-loop systems, smart internet of things (IoT) gateway, artificial intelligence (AI), blockchain, and the metaverse. Undeniably, recent developments have accelerated the advent of the AI and IoT era. Herein, this review presented an overview of the progress, applications, and prospects of AI optoelectronic sensing technology. Firstly, inorganic semiconductors, organic optoelectronic materials, 2D materials, and other nanomaterials employed in optoelectronic sensing technologies are generalized systematically. Then, the development status quo of optoelectronic sensing technology is discussed deeply. Next, several application scenarios related to optoelectronic sensing technology are reviewed, demonstrating the great development and prosperity of intelligent application systems. Last, the future development trends of AI optoelectronic sensing technology are portrayed, that is, toward all-weather, versatile, and intelligent systems in the age of AI/IoT.

Introduction

Today's most strategic and disruptive technology is artificial intelligence (AI) [[1], [2], [3]]. With its strong empowerment, it ushers in a new era of scientific, technical, and industrial revolutions, having a substantial influence on economic development, social advancement, and the building of the metaverse landscape. AI focuses on solving the problem that machines can listen (machine translation, speech recognition, etc.), can see (recognition for images and words), can speak (human-machine communication, speech synthesis, etc.), and can think (human-machine game, theorem-proof) Etc.), can learn (machine/deep learning, knowledge representation, etc.), can act (robots, unmanned driving, etc.) have risen to the top of the list of the six primary challenges that AI is now being integrated into. In order to tackle these problems, optoelectronic sensing technology is commonly employed. For instances, Veeralingam et al. [4] revealed the first artificial intelligence/machine learning (AI/ML) based nanomaterial-based multifunctional sensing device with great stability and reusability at ambient temperature for simultaneous and continuous monitoring of some essential body parameters. re. Pan et al. [5] proposed a smart glove for high-precision intellisense with a multi-channel capacitive pressure sensor capable of recognizing dozens of language gestures, with a maximum test classification accuracy of 99.7% and a difference of less than 0.1% compared to conventional demodulated sensing schemes. Surprisingly, Jiang et al. [6] suggested an ultra-thin eardrum-like friction electroacoustic sensor (EFES) made of silver-plated nanofibers that uses an AI algorithm to achieve 92.64% recognition rate in real-time speech conversion.

Functional optoelectronic devices play a key role in AI optoelectronic sensing technology and are constructed of a variety of materials, including semiconductors [7], organic optoelectronic materials [8,9], 2D materials [[10], [11], [12]], and some other popular materials represented by carbon materials [13]. Because of optical storage and optical integration technology, the scope of optoelectronic information functional materials research is expanding, reaching the micro-physical level and embracing inorganic and organic, metal and semiconductor, static and non-static science and technology, and the application of computers to highly intelligent information storage, transmission, and processing has enabled information technology to advance at a rapid pace.

Any advancement in AI optoelectronic sensing technology is frequently inextricably linked to the role of various types of optoelectronic sensors [14]. AI sensor is a new type of sensor that can have the function of feeling and detecting certain information of the measured object, it can learn, reason, and judge to process the signal, and has communication and management functions. AI sensors have the ability to automatically calibrate, calibrate, compensate, collect data, and so on. Its ability to determine the intelligent sensor also has a higher accuracy and resolution, higher stability and reliability, and better adaptability, compared to traditional sensors also have very high-cost performance. Early intelligent sensor is the sensor output signal that is processed and transformed by the interface to the microprocessor for arithmetic processing. In the 1980s, intelligent sensors are mainly microprocessors as the core, the sensor signal conditioning circuit, microelectronic computer memory, and interface circuit integrated into a chip, so that the sensor has a certain AI. In the 1990s, intelligent measurement technology has further improved, so that sensors to achieve miniaturization, structural integration, array, digital, easy to use, simple operation, and has self-diagnostic function, memory and information processing functions, data storage functions, multiparameter measurement functions, networking and communication functions, logical thinking and judgment functions.

As the information age progresses, optoelectronic sensors are becoming more widely utilized in numerous devices/products. Sensors that are small, digital, and intelligent are constantly gaining popularity, and they are transforming the way we live, that is, photoelectric sensors are devices that convert an optical signal into an electrical signal. Their working principles are based on the photoelectric effect. The photoelectric effect refers to the phenomenon that when light is irradiated on certain substances, the electrons of the substances absorb the energy of photons and the corresponding electrical effect occurs. These sensors allow objects to have senses such as touch, taste, and smell and have great potential in the all-encompassing digital living space for a new social system. We classify three types of optoelectronic sensors: (1) Optical sensors [[15], [16], [17], [18]] are specific functional devices for detection based on the optical fundamentals/principles, which provide a number of benefits, including contactless and undamaged testing, almost interference-free, great velocity transmission, and can be used to precise satellite tracking and remote control, etc. (2) Electrical sensors [[19], [20], [21], [22], [23], [24]] refer to frequently used inductive sensors, piezoresistance/resistive sensors, capacitive sensors, magnetic and dielectric sensors, and eddy current sensors in electrical measurement technologies. (3) Optoelectronic integrated sensors [[25], [26], [27], [28]] are these devices based on the optoelectronic effect that transposes optical module to electrical module. When light is shined on particular substances, the optoelectronic effect occurs, in which the electrons convert the energy of photons and the associated electrical reaction happens. There are three types of optoelectronic effects listed above: external photoelectric effect [29], internal photoelectric effect [30], and photovoltaic effect [31]. Photovoltaic devices include phototubes, photomultiplier tubes, photoresistors, photodiodes, phototransistors, photocells, etc. Therefore, in a broad sense, all these optoelectronic devices are important components of optoelectronic sensors. On this basis, it is of long-term significance to improve the working efficiency and accuracy of optoelectronic sensors through the comprehensive optimization of both AI algorithms and hardware. In particular, the rapid growth of the IoT application market has made the functions and uses of sensors based on the collection and effective use of data, and sensor nodes with connectivity capabilities are becoming increasingly common.

With the integration of AI technology and functional electronics, also the expansion of practical application scenarios, the ultimate goal is to make people's lives more comfortable, convenient, intelligent and visualized. Recently, AI optoelectronic sensors have great potential in industrial production, smart homes, environmental protection, underwater communication, and other aspects, and with the gradual arrival of the intelligent era, they will become more irreplaceable. As an important link of AI information search and identification transmission, a variety of intelligent optoelectronic sensors can realize the real-time monitoring of dynamic information in the environment, and complete the automatic control and management of intelligent equipment in various industries through the combination between intelligent sensors, control systems, and master control chip, which is equivalent to the heart of the IoT.

Numerous investigations of photoelectric materials-based sensors were conducted by the researchers. Few reviews have been reported to date that comprehensively review the research on information photoelectric materials in sensing applications. Thus, the techniques of preparation, advantages and disadvantages, properties, and development status of optoelectronic materials are outlined exhaustively in this review. We also highlight the latest research advances in several types of smart optoelectronic materials-based sensors, including mainly optical, electrical, and optoelectronic integrated sensors. In addition, we systematically review the different application scenarios of optoelectronic sensors with AI algorithms and provide an outlook on the future trends of optoelectronic sensing technology.

Matter, energy, and information are the three elements that constitute the macroscopic world. Information functional materials and devices, as an important basis for the development of high-tech industries in the century's information society, are involved in all aspects of information acquisition, emission, transmission, reception, processing, storage, and display, and the following is a detailed statement of the development status and trends of the four major optoelectronic functional materials.

With the accelerated pace of information carriers from electronic to optoelectronic and photonic conversion, semiconductor materials have also experienced three-dimensional (3D) bulk materials to a thin layer, two-dimensional (2D) ultra-thin layer microstructure materials, and are moving towards the integration of materials, devices, and circuits as a functional system integrated chip materials direction (one-dimensional (1D) quantum wires and zero-dimensional (0D) quantum dot materials as the representative of nanostructured materials), details are shown in Table 1. From the materials perspective, Si and Si-based materials [32,33] will continue to be the foundation of modern microelectronics technology, while, compound semiconductor microstructure materials with their excellent optoelectronic performances in great velocity, low-noise, low-energy/power microelectronics play an increasingly important role, especially in optoelectronic devices, photoelectric integration, and photonic integration (Fig. 1a). Inorganic semiconductor materials went through four stages of development (Fig. 1b) are as follows.

The first stage is the 1st generation of semiconductor materials represented by silicon (Si), germanium (Ge) [34], especially Si, it is the material basis for large-scale integrated circuits (IC), analog IC, and sensors, and it is the cornerstone of Si processing technology that enables Moore's Law to be applied. Czochralski [35,36] and zone melting techniques [37] are two of the most used semiconductor wafer production techniques. The Czochralski technique is considered as an effective way to make single-crystal Si wafers, and it is also utilized to make large-scale Si wafers. The Czochralski process may be used to make semiconductor polished wafers, epitaxial wafers, Si-On-Insulator (SOI) [38], and other types of semiconductor Si wafers, which are primarily for low-power integrated circuit components in 8-inch and 12-inch sizes. The zone melting process can now generate the highest purity Si single crystal, whereas it is also more expensive, and it is most typically employed to make small-size Si wafers. Si-based chips are widely used in computers, cell phones, TVs, aerospace, various military projects, and the rapidly developing new energy [39] and photovoltaic industries [40].

The second stage is represented by compounds such as gallium arsenide (GaAs) [41], indium phosphide (InP) [42], it also contains ternary compound semiconductors such as GaAsAl [43], GaAsP [44], and some solid solution semiconductors such as Ge-Si [45], GaAs-GaP [46], glass semiconductors (also known as amorphous semiconductors) such as amorphous Si [47], and glassy oxide semiconductors [48]. In 1968, Mullin et al. [49] grew lnP single crystal material by LEC method for the first time using boron trioxide as liquid sealant, which laid the foundation for the preparation of large diameter and high-grade group III-V single crystals. However, since the dissociation pressure of phosphorus at the melting temperature of 1335 ± 7 K for InP is 27.5 atm, the synthesis of InP polycrystals and the growth of single crystals are relatively complicated. In addition, the stacking layer of lnP has low lower fault energy and is prone to twinning, which makes the preparation of high-quality lnP single crystals more difficult. Great velocity, high-frequency, high-energy/power, and light-emitting electronic systems are mainly made with 2nd-generation semiconductor materials, nonetheless, they are also ideal materials for designing and constructing microwave/millimeter-wave devices [50], which have been extensively used in communication systems like radiofrequency, photonics, lasers, optoelectronics, and satellite navigation.

The 3rd generation of semiconductor materials, primarily gallium nitride (GaN) [51], silicon carbide (SiC) [52], and other broadband semiconductor materials, may be associated with electronics, illumination, vehicle technologies, weapons, spacecraft, and other fields, and has much better performance than the 1st and 2nd generations. Take GaN as an example, the main epitaxial growth methods are metal organic chemical vapor deposition (MOCVD), hydride vapor phase epitaxy (HVPE), and molecular beam epitaxy (MBE). The 3rd generation semiconductor materials are the main materials in the 5G era, thus expecting to replace the 1st and 2nd generation of semiconductor materials well with the advent of age of AI/IoT. Compared with the traditional 1st and 2nd generation semiconductor materials Si and GaAs, the 3rd-generation semiconductors have unique properties such as large forbidden bandwidth, high breakdown electric field, and large thermal conductivity, making them show great potential in optoelectronic devices, power electronics, radio frequency (RF) microwave devices [53], etc.

The 4th generation semiconductor is mainly diamond (C) [54], gallium oxide (Ga2O3) [[55], [56], [57], [58], [59], [60]], aluminum nitride (AlN) [61] as the representative of the ultra-wide bandgap (UWBG) semiconductor materials with a special bandwidth of greater than 4 eV. Take Ga2O3 as an example, the material does not exist in nature at all and needs to be synthesized artificially. The reaction of two elements O and Ga needs to be carried out in an atmosphere close to 1000 °C. Chemical vapor deposition (CVD) [[62], [63], [64], [65], [66]] is the primary processing technology used to grow high-purity, high-performance solid Ga2O3 films. In a typical CVD process, one or more vapor source atoms or molecules are introduced into the chamber, where the chemical reaction occurs under the action of external energy and forms the desired film on the substrate surface. Compared with other semiconductor materials, the 4th generation semiconductor materials have the advantages of smaller size, lower energy consumption, and stronger functions, which can be better used in optoelectronic devices and power electronics devices under harsh environmental conditions. The main material systems that have the potential to become the 4th generation of semiconductor technology include GaSb [67] with narrow bandgap, AsIn [68] compound semiconductors, and oxide materials with ultra wide bandgap. Among them, antimonide semiconductors have unique and irreplaceable advantages in developing next-generation lightweight, undersized, low energy/power consumption, cut-price devices, and their demanding applications.

The fast growth of semiconductors has given the foundation for our technological boom during the last half-century. Moore's Law appears to be over when 10 nm chips start commercializing, and 7 nm and 5 nm processes are approaching their limits as well. No one really knows when the next semiconductor tipping point will occur, but the current state of quantum technology looks to be pointing us on the right path. Quantum chips come to the forefront of the introduction of quantum communication, quantum entanglement, and other new terms into our field of vision, and the traditional chip with 0 and 1 for computing processing was replaced by quantum chips with multiple quantum bits, instead of only 0 and 1 again two logical states, allowing the chip's computing power to grow exponentially, thus having to break the Moore's law.

Organic photovoltaic materials are a class of organic materials with photoelectric activity, which are widely used in organic light-emitting diodes (OLED), organic transistors, organic solar cells (OSC), organic sensors, organic memories, and other fields. Organic photoelectric materials are generally carbon-rich organic compounds with a broad π-conjugation system, which are split into two types: small molecules [117] and polymeric materials (Fig. 2a and b), details are shown in Table 2. Compared with inorganic materials, organic photoelectric materials can be synthesized using the solution approach for massive preparation and flexible sensors construction. Moreover, organic materials have diverse structural compositions and a wide range of performance tuning, allowing the molecular design to obtain the desired properties and bottom-up device assembly such as self-assembly to prepare nanodevices and molecular devices.

Due to the advantages of organic molecules compared with inorganic materials, such as low density, low price, and easy structural modification, organic photoelectric functional materials have recently attracted the attention of academic and industrial research and commercialization. In the meantime, the design and synthesis of special organic π-conjugated systems can be explored for their applications in the field of optoelectronic functions.

Research work on organic electroluminescence began in the 1960s, but it was not until 1987 that Kodak company [143] used a multilayer film structure to obtain the first OLED with high quantum efficiency, high luminescence efficiency, high brightness, and low drive voltage (Fig. 3). Compared with traditional light-emitting and display technologies, OLED have the advantages of the low driving voltage, small size, lightweight, and a great deal of materials to choose from, and is easy to achieve large-area preparation, wet preparation, and preparation of flexible devices.

In recent years, OLED technology has developed rapidly. A novel blended EIL [144] was created to perform all of the required functionalities by fusing the benefits of organic and inorganic EILs. The PFN-OX/ZnO hybrid electron injection layer not just outperformed than the organic PFN-OX EIL in terms of electron injection, but also preserved the inorganic ZnO EIL's solvent resistance. Chen et al. [145] used TPE-TAPBI pure films as blue-emitting components to obtain warm white OLED with two colors that are quite efficient, which provided ideal consumption, total current, and EQE. TPAATPE as the blue emitter was paired with the TADF molecule, PTZMes 2B as the green emitter and phosphor component to adjust the longwave emission to create a bunch of extremely effective mixed white OLED [146]. Three-color white OLED, in particular, obtained pure white emission with proper CIE values and a highest front-view EQE of 25.3%, as well as an amazingly CRI of 92. Xiao et al. [147] has developed an in situ method to map the parameters of triplet exciton and related TTA processes in OLED so that useful triplet-triplet annihilation (TTA) materials can be identified and the design of device structures can be optimized.

Small molecules (Fig. 4a, c, 4e, and 4f) are a class of organic molecules with a large π-conjugation system rich in carbon atoms, take PDPP-MH and Cl8PTCDI as examples, their crystal structures as shown in Fig. 4b and g. They can be divided into p-type and n-type semiconductors according to the type of carrier charge transported. In 2006, Bao et al. [149] successfully prepared single-crystal arrays of pentacene and red fluorene, and assembled transistor devices on this basis. They first prepared a layer of patterned octadecylchlorosilane (OTS) on Si/SiO2 substrate by stamping method, and then constructed organic semiconductors such as pentacene, red fluorene, and C60 on the substrate by vacuum vapor deposition. The carrier mobility of pentacene is 0.2 cm2 V−1 s−1 and the switching current ratio is 106, while the carrier mobility of red fluorene is 2.4 cm2 V−1 s−1 and the switching current ratio is 106. Behrman et al. [150] described the relationship between contact overlap, the protective effect of increased overlap, the electric field in the dielectric, and its effect on electrostatic discharges (ESD) resilience. When designing OFET devices, the negative overlap length should be carefully considered as a trade-off between reduced performance and higher ESD resilience. Zou et al. [151] proposed a possible mechanism for molecular stacking in different temperature ranges in response to the challenge that higher preparation temperatures in the liquid crystal or isotropic liquid range cause many defects during crystal formation and lead to a sharp decrease in FET mobility due to cell tilting. This strategy is expected to provide new opportunities for the cost-efficient fabrication of OFET.

Compared with the optoelectronic conversion efficiency of inorganic Si-based solar cells, the photoelectric conversion efficiency of OSC still stayed at a relatively low level. Therefore, the core of OSC research is to improve the photoelectric conversion efficiency of the cell. By designing a reasonable device structure (Fig. 5a-5d), improving the interface morphology, and increasing the degree of polymer crystallization, the photoelectric conversion efficiency of OSC has been greatly improved [[153], [154], [155], [156]]. A brightly colored ST-OSC [157] was successfully developed by using a fixed active mixture and high quality Fabry-Pérot electrodes in a rationally designed device layout. With the help of optical simulations, a brightly colored ST-OSC with a power conversion efficiency >14% and a maximum transmittance up to 31% was obtained. Duan et al. [158] summarized the factors limiting OSC stability, and surveyed recent advances in strategies to improve OSC stability, such as material design, device engineering of active layers, utility of inverted geometries, optimization of buffer layers, and design of stable electrodes and packaging materials. Potential research strategies to achieve the desired device stability and efficiency were highlighted, thus providing possible avenues to promote the viable commercialization of OSC.

Organic sensors based on organic transistors can be widely used in chemical and biological fields for the detection of chemical substances and biological macromolecules. Compared with conventional sensors, organic transistor sensors are of the advantages of small size, easy array implementation, mobility, and low cost. In addition, the response signal of organic transistor sensors is usually a current signal, which is easy to test. Compared with other chemical sensors, organic transistor sensors based on various polymer materials like hydrogel [160], ionogel [161], polymer composites [162], and conductive polymers [163] (Fig. 6a-6d) also have the advantage of providing more electrical information, such as the conductivity, field-effect conductivity, threshold voltage, and field-effect mobility of organic films. Rullyani et al. [164] presented an OTFT-based temperature sensor with a broad sensing range, thus covering human body temperature. The flexible OTFT fabricated on a polyethylene terephthalate substrate has a hole mobility of up to 0.39 ± 0.01 cm2 V−1 s−1, a Vth value of −18.6 ± 0.45, an on/off ratio of 102, and can undergo more than 100 bending cycles. The future development of organic transistor sensors is to further improve the response speed, detection limit, and stability of the devices. With the development of organic transistor technology, especially the continuous progress of flexibility, array, and patterning technology, organic transistor sensors are expected to realize flexible sensors and simultaneous online analysis of multiple samples, becoming a veritable “electronic nose".

For a particular material, the device may change from an insulating state (0) to a conducting state (1) when the field strength reaches a certain value by applying a voltage to both sides of the film. By some stimulus (such as reverse electric field, current pulse, light or heat, etc.), the device can be restored from the 1 state to the 0 state, such devices are called switching devices. When the applied electric field disappears, the 0 or 1 state can exist stably, which means that it has memory characteristics and becomes a memory device. Organic memory provides a number of benefits over traditional silicon memory [[165], [166], [167], [168]], characteristics like simple manufacturing, cheap cost, and a high surface area, flexible devices, and three-dimensional storage (high storage capacity), as shown in Fig. 7a-7d.

On a flexible PET substrate, an organic resistive switching memory device with Ag/Leaves/Ti/PET structure was fabricated [169], which exhibited unique resistive switching memory properties with a large shifting resistance ratio and stable cycling outcomes at RT. Gao et al. [170] reported a heterostructured vertical organic storage transistor that used a p/n semiconductor bulk heterojunction as the semiconductor layer without any additional charge capture layer. Thanks to the formation of the p/n semiconductor interface and nanometer-scale transport length, the device achieved a storage window of 58 V, a high storage ratio of 105, and retention characteristics of more than 10 years under visible light stimulation, superior to most reported optical organic memory devices.

Compared with the three-pole organic memory based on transistor structure, the two-pole memory is with the advantages of simple structure, easy integration, and can fully utilize the characteristics of organic materials, hence, the two-pole organic memory will likely become the mainstream of future development. Another development trend of organic memory is to combine with nanotechnology to realize the assembly of nano-devices and even molecular devices to improve the storage density.

In order to compensate for the shortcomings of zero-bandgap graphene, more and more 2D materials with certain bandgaps (such as transition metal-sulfur compounds (TMDS, MX2) [171], black phosphorus [172], and InSe materials [173]) have been successfully synthesized, and substantial progress has been done in related physical properties and device studies, thus providing greater hope for the continuation of Moore's law.

After the discovery of graphene, many physical properties intrinsic to 2D materials have been probed on mechanically solved 2D materials. However, mechanical deconvolution techniques have long been difficult to solve bottlenecks in terms of yield and size, thus affecting the progress of research on 2D materials. Scientists have developed novel oxygen plasma-enhanced and gold film-assisted pervasive solvation techniques to efficiently obtain more than 40 large-area high-quality 2D materials, including graphene, MoS2, WSe2, WTe2, MnBi2Te4, etc., with sample sizes ranging from millimeter to subcentimeter scale. In particular, the gold film-assisted deconvolution technique can achieve substrate conductivity and insulation by controlling the thickness of the metal film, providing diverse options for different research tools. This new deconvolution technique can also be used to prepare suspended two-dimensional materials, providing the most ideal samples for studying their intrinsic optical and electrical properties, etc. By optimizing the desorption process, controllable preparation of some 2D materials with special structures can be achieved, such as pleated and bubble structures, providing ideal models for studying the behavior of 2D materials under different strain environments. With the help of large size is graphene and MoS2 bubble structures, we observed Raman and fluorescence oscillation loops and observed the splitting of multilayer MoS2 under strain fields. In the pleated structure, the presence of a strain field reduces the energy required in the chemical reaction process, making the chemical reaction activity in the pleated region significantly higher than that in the flat 2D material region. Therefore, the novel deconvolution technique after overcoming the limitations of preparation efficiency and sample size will continue to play a unique advantage in future studies.

In recent years, the application of 2D layered compounds in various optoelectronic devices has received a lot of attention from researchers, mainly due to the following advantages of two-dimensional materials: (1) Atomic-level thickness, so that the charge carrier concentration and photoelectric properties in them can be effectively regulated by means of local electric fields. (2) band gap distribution of 0–6 eV, including insulators (e.g. BN), semiconductors (e.g. MoS2, PtSe2, black phosphorus, tellurium), semimetals (e.g. MoTe2), topological insulators (e.g. Bi2Se3), metals (e.g. 1T-TaS2), can achieve a wide range of electromagnetic spectral responses. (3) Layer-dependent electronic energy band structures, such as the bulk and monolayer states of black phosphorus materials exhibit direct band gaps with forbidden band widths of ∼0.3 eV and ∼2 eV, respectively (Fig. 8a-8d), details are shown in Table 3. The interlayer is a weaker van der Waals force, which does not need to consider the lattice matching limitation, and thus a variety of different van der Waals heterostructures can be prepared to meet different requirements and device applications.

In addition to the above-mentioned materials, carbon-based materials [[192], [193], [194]], liquid metals [195], and various types of conducting precious metal nanowires [196] in Fig. 9a-9c have also shown exceptional brilliance. Among them, carbon-based materials are abundant in nature and are the only single-element system with isotopic configurations covering 0D to 3D. The successive discoveries of fullerenes, carbon nanotubes, and graphene have made carbon-family materials be a long-term research hotspot in the fields of materials science and nanotechnology. Particularly, 1D carbon nanotubes and 2D graphene have shown great applications in the field of electronics and optoelectronics. To date, a series of research results have been achieved in the preparation of all-carbon heterogeneous materials, optoelectronic devices, and optoelectronic physical properties research. To address the current problems of low responsiveness of graphene-based photodetectors and slow response of carbon nanotube thin-film detectors, Liu et al. [197] achieved a photoconductive gain of 105 using the carbon nanotube photogatting effect, and achieved a synergistic optimization of superior device responsiveness (>100 A/W) and high response speed while maintaining a broad spectral response (400–1550 nm). This study also extended van der Waals heterostructures based on 2D materials to 1D/2D composite structures for the first time, providing a completely new platform for studying novel physical phenomena based on quantum confinement effects in van der Waals heterostructures. Immediately after, Zhang's team [198] investigated the carrier dynamics at the graphene/carbon nanotube heterojunction interface by photocurrent characterization combined with Raman spectroscopy statistical analysis, and confirmed that the carrier dynamics at the all-carbon heterojunction interface was closely related to the chirality of carbon nanotubes. It is found that photogenerated electrons of semiconductor-type carbon nanotube/graphene heterojunctions were transferred to graphene applied the action of the built-in electric field under light illumination to form photocurrents, while for metal-type carbon nanotube/graphene heterojunctions, photocurrents were generated due to the thermal effect of graphene under light illumination. The extraordinary mechanical properties of the all-carbon system make it a great application in the field of flexible electronics. Liu et al. [199] used graphene/carbon nanotube heterojunction films to build transistors on top of flexible substrate PET, which demonstrated the devices can still maintain a superior optoelectronic response (responsiveness ∼50 A/W, response speed ∼40 ms) under different stress curvature. This work verified that the all-carbon material system was with good mechanical flexibility and the advantage of large area integration, which had potential applications in the field of wearable devices. The future will be “carbon-based materials + photoelectric process” (or “carbon + light”) of the era, graphene, carbon-based nanomaterials, organic polymer materials, as well as laser and optical communications, optical storage, optical display, etc. The “carbon-based materials + photoelectric process” has given birth to the flexible electronics and flexible electronics industry, and opened up a very broad space for its development, which will profoundly change the way of human production, lifestyle, and thinking.

Section snippets

The development status of AI optoelectronic sensing technology under the background of AI and IoT era

The sensors for AI optoelectronic sensing technology, as one of the critical sensing devices for the IoT perception layer, have also progressed at a fast pace in the context of the industry's rapid growth. The IoT relies on it to acquire information and exercise object control, conversely, its performance has a substantial influence on the IoT. Optoelectronic sensors are currently evolving in the direction of intelligence, downsizing, and multi-functionality.

Application scenarios

Owing to its quick reaction time, non-contact measurement, high accuracy, high resolution, and dependability, along with their compact size, lightweight, low power consumption, and simplicity of integration, optoelectronic sensors are one of the most productive and commonly used sensors. They have extensively applied in quantities of domains involving industry, intelligent medicine, communications, water quality monitoring, defense and civil facilities construction, daily household, climate

Prospects

Entering the era of AI, modern science and technology have also entered an explosive period. For the material research, with the continuous extension of its application fields, other disciplines have also gained unexpected gains in the combination with AI, and new materials are one of them. Since this century, with the further improvement of high-end manufacturing, new materials have become an important means of scientific and technological progress around the development path of

Declaration of competing interest

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.

Acknowledgements

The research was supported by the Key Project of Department of Education of Guangdong Province (No.2018KCXTD026), Science and Technology Innovation Commission of Shenzhen(No.JCYJ20170818093453105, No. GJHZ20180928160209731), Shenzhen Science and Technology R and D and Innovation Foundation (Grant Nos. JCYJ20200109105608771), and the authors are grateful for financial support from the National Key R&D Program of China (2019YFB2204500) and the Shenzhen Science and Technology Program (

References (239)

  • Z. Zhang et al.

    High-performance Si/nano-Cu/CNTs/C anode derived from photovoltaic silicon waste: a potential photovoltaic-energy storage strategy

    Mater. Today Energy

    (2021)
  • J.B. Mullin et al.

    J. Cryst. Growth

    (1968)
  • C. Tan et al.

    Duplex strengthening via SiC addition and in-situ precipitation in additively manufactured composite materials

    Compos. B Eng.

    (2022)
  • V.S. Sedov et al.

    Deposition of diamond films on Si by microwave plasma CVD in varied CH4-H2 mixtures: reverse nanocrystalline-to-microcrystalline structure transition at very high methane concentrations

    Diam. Relat. Mater.

    (2020)
  • D. Guo et al.

    Review of Ga2O3-based optoelectronic devices

    Mater. Today Phys.

    (2019)
  • H. Nishinaka et al.

    Rapid homoepitaxial growth of (010) β-Ga2O3 thin films via mist chemical vapor deposition

    Mater. Sci. Semicond. Process.

    (2021)
  • A. Mondal et al.

    Transition from thin film to nanostructure in low pressure chemical vapor deposition growth of β-Ga2O3: impact of metal gallium source

    Thin Solid Films

    (2020)
  • T.F. Weng et al.

    VLS growth of pure and Au decorated β-Ga2O3 nanowires for room temperature CO gas sensor and resistive memory applications

    Appl. Surf. Sci.

    (2020)
  • I.J. Park et al.

    A three-terminal monolithic perovskite/Si tandem solar cell characterization platform

    Joule

    (2019)
  • Z. Qiu et al.

    Monolithic perovskite/Si tandem solar cells exceeding 22% efficiency via optimizing top cell absorber

    Nano Energy

    (2018)
  • S. Srivastava et al.

    Bulk and interface defects analysis of n-CdS/p-Si heterojunction solar cell

    Opt. Mater.

    (2021)
  • X. Zhang et al.

    Emitter ballasting resistor design of InGaP/GaAs power HBT for high linearity based on Volterra Series distortion analysis

    Microelectron. J.

    (2018)
  • N. Li et al.

    A low temperature formaldehyde gas sensor based on hierarchical SnO/SnO2 nano-flowers assembled from ultrathin nanosheets: synthesis, sensing performance and mechanism

    Sensor. Actuat. B-Chem.

    (2019)
  • Z. Wang et al.

    SnO-SnO2 modified two-dimensional MXene Ti3C2Tx for acetone gas sensor working at room temperature

    J. Mater. Sci. Technol.

    (2021)
  • M.S. Barbosa et al.

    Investigation of electronic and chemical sensitization effects promoted by Pt and Pd nanoparticles on single-crystalline SnO nanobelt-based gas sensors

    Sensor. Actuat. B-Chem.

    (2019)
  • M.S. Barbosa et al.

    Gas sensor properties of Ag-and Pd-decorated SnO micro-disks to NO2, H2 and CO: catalyst enhanced sensor response and selectivity

    Sensor. Actuat. B-Chem.

    (2017)
  • R. Barras et al.

    Porous PDMS conformable coating for high power output carbon fibers/ZnO nanorod-based triboelectric energy harvesters

    Nano Energy

    (2021)
  • X. Pu et al.

    Flexible PVDF/nylon-11 electrospun fibrous membranes with aligned ZnO nanowires as potential triboelectric nanogenerators

    Chem. Eng. J.

    (2020)
  • P. Supraja et al.

    A simple and low-cost triboelectric nanogenerator based on two dimensional ZnO nanosheets and its application in portable electronics

    Sensor. Actuat. A-Phys.

    (2022)
  • D. Li et al.

    Electron-transfer mechanisms for confirmation of contact-electrification in ZnO/polyimide-based triboelectric nanogenerators

    Nano Energy

    (2020)
  • J. He et al.

    The practical implementation of artificial intelligence technologies in medicine

    Nat. Med.

    (2019)
  • A. Hosny et al.

    Artificial intelligence in radiology

    Nat. Rev. Cancer

    (2018)
  • S. Veeralingam et al.

    AI/ML-enabled 2-D-RuS2 nanomaterial-based multifunctional, low Cost, wearable sensor platform for non-invasive point of care diagnostics

    IEEE Sensor. J.

    (2020)
  • J. Pan et al.

    A Wireless multi-channel capacitive sensor system for efficient glove-based gesture recognition with AI at the edge

    IEEE T. Circuits-II

    (2020)
  • Y. Jiang et al.

    Ultrathin eardrum-inspired self-powered acoustic sensor for vocal synchronization recognition with the assistance of machine learning

    Small

    (2022)
  • X. Huang et al.

    Recent progress in organic phototransistors: semiconductor materials, device structures and optoelectronic applications

    ChemPhotoChem

    (2020)
  • Q. Li et al.

    Molecular packing: another key point for the performance of organic and polymeric optoelectronic materials

    Accounts Chem. Res.

    (2020)
  • W. Li et al.

    Phase transitions in 2D materials

    Nat. Rev. Mater.

    (2021)
  • X. Liu et al.

    2D materials for quantum information science

    Nat. Rev. Mater.

    (2019)
  • Z. Dai et al.

    Strain engineering of 2D materials: issues and opportunities at the interface

    Adv. Mater.

    (2019)
  • J. Xiao et al.

    Dimensionality, function and performance of carbon materials in energy storage devices

    Adv. Energy Mater.

    (2022)
  • Q.B. Zhu et al.

    A flexible ultrasensitive optoelectronic sensor array for neuromorphic vision systems

    Nat. Commun.

    (2021)
  • A.K. Ghosh et al.

    Exploring Plasmonic Resonances toward “Large-scale” flexible optical sensors with deformation stability

    Adv. Funct. Mater.

    (2021)
  • M. Rossi et al.

    A comparison of the signal from diverse optical sensors for monitoring alpine grassland dynamics

    Rem. Sens.

    (2019)
  • V.S. Kumar et al.

    Development of electrical transduction based wearable tactile sensors for human vital signs monitor: fundamentals, methodologies and applications

    Sensor. Actuat. A-Phys.

    (2021)
  • M.A. Squillaci et al.

    3D hybrid networks of gold nanoparticles: mechanoresponsive electrical humidity sensors with on-demand performances

    Nanoscale

    (2019)
  • F. Mokhtari et al.

    Highly stretchable self-powered wearable electrical energy generator and sensors

    Adv. Mater. Technol.

    (2021)
  • T. Xie et al.

    Understanding and optimization of graphene gas sensors

    Appl. Phys. Lett.

    (2021)
  • N. Schmidová et al.

    Impact damage detection of a glass fabric composite using carbon fiber sensors with regard to mechanical loading

    Appl. Sci.

    (2022)
  • Q. Yu et al.

    Highly sensitive strain sensors based on piezotronic tunneling junction

    Nat. Commun.

    (2022)
  • Cited by (0)

    View full text