Advances in Two-Dimensional Materials for Optoelectronics Applications

Abstract

The past one and a half decades have witnessed the tremendous progress of two-dimensional (2D) crystals, including graphene, transition-metal dichalcogenides, black phosphorus, MXenes, hexagonal boron nitride, etc., in a variety of fields. The key to their success is their unique structural, electrical, mechanical and optical properties. Herein, this paper gives a comprehensive summary on the recent advances in 2D materials for optoelectronic approaches with the emphasis on the morphology and structure, optical properties, synthesis methods, as well as detailed optoelectronic applications. Additionally, the challenges and perspectives in the current development of 2D materials are also summarized and indicated. Therefore, this review can provide a reference for further explorations and innovations of 2D material-based optoelectronics devices.

1. Introduction

Graphene, as a representative two-dimensional (2D) material, has attracted considerable attention since it was mechanically exfoliated from bulk graphite in 2004 [1]. Its special two-dimensional honeycomb lattice endows it with numerous excellent properties such as high carrier mobility, high mechanical strength, excellent optical transparency, and a wide absorption spectrum [2,3], which makes it exhibit great application potential in many areas [4]. The discovery and development of graphene has further stimulated research interests in other 2D materials.

Two-dimensional materials are layered crystalline solids that have strong bonding in the crystal plane, while the forces between the adjacent atomic layers are weak van-der-Waals forces [5,6,7]. The special structure of 2D materials endows them with many remarkable abilities. First, quantum confinement in the direction perpendicular to the 2D plane introduces some novel optical and electronic performances to 2D materials, which are distinct from their bulk parent materials [8]. Second, the surfaces of these materials are naturally passivated and they have no dangling bond, which facilitates the integration of 2D materials with photonic structures [8,9]. In addition, using diverse 2D materials can build vertical heterostructures, in which the different layers are bonded by van der Waals forces, averting the issue of lattice mismatch. Third, a host of 2D materials can interact strongly with light [10]. For instance, monolayer MoS₂ can absorb nearly 10% of perpendicular incident light at excitonic resonances [11]. Finally, 2D materials can cover a broad range of electromagnetic spectra due to their different electronic properties [12]. Due to their special properties, 2D materials show great application potential in various fields, especially in optoelectronics applications [13,14,15]. For example, graphene has become a potential candidate in the applications of light modulation, detection and manipulation because it can interact with light in a wide range of spectra. Some transition metal dichalcogenides (TMDCs) such as MoS₂ and WS₂ have sizable bandgaps and possess illustrious light-emitting performances, ensuring their applications in photodetectors, transistors, and other optoelectronics devices [8,16]. Hexagonal boron nitride (hBN) is also a significant 2D material and it is often used as an outstanding dielectric on account of its large bandgap. Furthermore, it can be incorporated into heterostructures to act as the electrostatic gating of other 2D materials. In this paper, we give a comprehensive review of the structure, properties, and optical applications of 2D materials. We also make an outlook of the prospects of 2D materials.

2. Graphene

2.1. Structure and Optical Property

2.1.1. Morphology and Structure

Graphene is made up of a single layer of carbon atoms which is arranged in a honeycomb lattice. The real-space images of the atomic lattice of monolayer graphene were obtained and used to characterize the thicknesses and nanoscale corrugation of a clean graphene sheet devoid of any impurities in 2007 [17,18]. The sp² hybridization between one s orbital and two p orbitals leads to a planar structure with the formation of a bond between carbon atoms that are separated by 1.42 Å. Due to the Pauli principle, these bands have a filled shell and, hence, form a deep valence band. The unaffected p orbital, which is perpendicular to the planar structure, can bind covalently with neighboring carbon atoms, leading to the formation of a π band. Since each p orbital has one extra electron, the π band is half filled.

As is well-known, graphene is a zero bandgap semi-metal material, because its conduction band and valence band intersect at the Dirac point. The energy band structure satisfies the Dirac equation, unlike traditional metals or semiconductors that satisfy the Schrödinger equation. At the K point, the energy of the electron is linearly related to its momentum, the static effective mass of the electron is 0, and the behavior is similar to that of a photon [19].

2.1.2. Optical Properties

Graphene has excellent light transmittance, with a light transmittance of 97.7% for light with a wavelength of 400–800 nm. Electrons in graphene behave as massless two-dimensional particles, which leads to a significant wavelength-independent absorption (πa < 2.3% in the infrared limit, where a is the fine structure constant) for normal incident light below about 3 eV [20]. The number of graphene layers is linearly related with the light transmittance [21]. As the number of layers increases, the light transmittance decreases, making it possible to roughly judge the number of graphene layers by an optical microscope [8]. Mono- and bi-layer graphene becomes completely transparent when the optical energy is smaller than that which is double the Fermi level, owing to Pauli blocking [22], and these properties would suit many controllable photonic devices.

The high-frequency conductivity for Dirac fermions in graphene has been considered as a constant (πe²/2h) [22,23], from the infrared to the visible range of the spectrum. Graphene has good anti-reflection properties, and its reflectivity is less than 0.1% when the light is incident perpendicular to the graphene surface. In addition, graphene also has good nonlinear optical properties. When the electric field generated by the incident light resonates with the outermost electrons of the carbon atoms in graphene, the electron cloud in graphene shifts and generates polarization, so graphene can exhibit nonlinear optical properties such as saturable absorption [24], the Kerr effect [25,26], and two-photon absorption [27]. Therefore, graphene has application potential in many optical fields such as ultrafast photonics [28].

2.2. Production Methods

The integration of novel materials can bring a new dimension to future technologies; therefore, large-scale and cost-effective production methods are necessary. Up to now, the main approaches to produce graphene are shown in Figure 1. These preparation methods can be roughly divided into two categories: one is the “top-down” synthesis method—that is, the desired nanoscale graphene is prepared from large-scale graphite by various techniques, such as mechanical exfoliation, the liquid-phase ultrasonic exfoliation method and the arc method [13]; the other is the “bottom-up” synthesis method—that is, small carbon-containing structural units are self-assembled into graphene through interaction, such as chemical vapor deposition [29].

2.3. Optical Applications

Due to the unique electrical and optical properties, including ultra-high carrier mobility, ultra-strong ballistic transport effect, ultra-fast optical response time, and ultra-wide spectral response range, graphene has considerable potential for optoelectronics applications. In this section, we review the optical applications of graphene in photodetectors, organic solar cells, electromagnetic optical nanoantennas, and so on, which are summarized in

Table 1. Optoelectronic applications based on graphene.

Applications	Materials	Photoresponsivity		Ref.
Photodetectors	Graphene/WS2/graphene	121 A W−1 (at 532 nm)		[31]
	Carbon nanotube/graphene	1.48 × 105 A W−1 (at 1550 nm)		[32]
	Graphene/PbS quantum dots	107 A W−1 (at 635 nm)		[33]
	Materials	Power conversion efficiency	Optical transmittance	Ref.
Organic solar cells	Polyimide/graphene	15.2%	92%	[34]
	AZO/graphene/Cu/AZO	4.63%	82%	[35]
	Materials	Operating frequency	Gain	Ref.
Electromagnetic optical nanoantennas	Graphene-based nano-antenna	55 THz	5.47 dB	[36]
	Graphene based slotted bowtie optical plasmonic nanoantenna	193.5 THz	7.38 dB	[37]
	Materials	Modulation depth	Bandwidth	Ref.
Optoelectronic modulators	Double-layer graphene/double rectangle metal nanoribbons	3.12 dB μm−1	380.23 GHz	[38]
	Graphene/hBN	14–20 dB μm−1	̶	[39]

.

2.3.1. Photodetectors

Photodetectors refer to devices that can convert optical signals into electrical signals, involving three physical processes, namely light capture, exciton separation, and charge transfer to electrodes. The zero bandgap electronic structure of graphene gives the graphene photodetectors broadband detection properties and an extremely fast response speed compared with other detectors [40]. However, this characteristic of graphene also makes the exciton recombination rate too fast and is not conducive to exciton separation, resulting in low detector responsivity, which is limited to about 6.1 mA W⁻¹ based on single-layer undoped graphene [41]. As a result, to overcome the above limitation, it is necessary to dope graphene with other materials. In addition, the graphene photodetector exhibits almost no attenuation in photoresponse in the modulation frequency range of 40 GHz, and the predicted theoretical bandwidth can be increased to 500 GHz, with a responsivity of 0.5 mA W⁻¹ at 80 V bias. Using nanostructured plasmas to enhance the local optical fields and combining with waveguides to increase the length of the light–graphene interaction can improve the sensitivity of graphene photodetectors. Ding et al. fabricated waveguide-coupled integrated graphene plasmonic photodetectors on a Si platform with a bandwidth over 110 GHz and a high intrinsic responsivity of 360 mA W⁻¹, which was attributed to the plasmon-enhanced graphene–light interaction and subwavelength confinement of the optical energy [42]. Guo and co-workers proposed and realized high-performance waveguide photodetectors based on bolometric/photoconductive effects by introducing an ultrathin wide silicon−graphene hybrid plasmonic waveguide, which enabled efficient light absorption in graphene at 1.55 μm and beyond [43].

2.3.2. Organic Solar Cells

Organic solar cells have attracted increasing attention due to their light weight, easy fabrication, low cost, large-area processability, low in-plane resistivity, and excellent light transmittance [44]. Due to its excellent properties, graphene shows great application potential in organic solar cells and it is expected to be an ideal material to replace indium tin oxide [45]. Li et al. reported a highly efficient graphene–Si solar cell. the corresponding schematic illustration of this graphene–Si Schottky junction solar cell is shown in Figure 2 [46]. Liu et al. [47] prepared a package-free organic photovoltaic device using highly doped multilayer CVD graphene as a top transparent electrode. The device with a double-layer graphene electrode exhibits the maximum power conversion efficiency of 3.2% and it has excellent bending stability. In addition, two or more layers of graphene top electrodes can well protect organic photovoltaic devices from air pollution because the multilayer graphene films are impermeable to air, which may simplify the device fabrication and decrease the cost of devices. Ricciardulli et al. [44] reported a cost-effective method for the fabrication of a transparent conductive electrode in an organic solar cell using solution-processed high-quality graphene. The obtained transparent electrode shows low sheet resistance and a high-power conversion efficiency value (4.23%) for an organic solar cell. In addition to graphene, graphene derivates such as graphene oxide (GO) also play an important role in enhancing the performance of organic solar cells. In organic electronics, poly(3,4-ethylenedioxythiophene): poly (styrene sulfonate) (PEDOT: PSS) is the most widely used hole transport or anode interfacial material. Although it has many merits such as high optical transparency and solution processability, it still meets some problems; for example, its high acidic and hygroscopic characteristics can damage the transparent electrode. In order to solve this problem, Hilal et al. fabricated GO/PEDOT:PSS as the hole transport layer of organic solar cells [48]. The device, based on GO/PEDOT:PSS, exhibits a maximum power conversion efficiency of 4.82%, and it retains over 30% of its initial power conversion efficiency without encapsulation after 15 days under atmospheric conditions. These excellent performances are attributed to the bond formation at the interface of the GO/PEDOT:PSS hole transport layer.

2.3.3. Graphene-Based Electromagnetic Optical Nanoantennas

Optical nanoantennas functioning at terahertz (THz) band, infrared and optical frequencies can control the energy conversion of light into localized energy and optical radiation at the subwavelength scale [27]. Therefore, they play an important role in the emerging field of photonics. The fundamental capabilities of nanoantennas can be utilized in practices including high-speed communication with high (gigabit/s) data rates in nano-networks, inter-chip communication, THz detection, optical light emission, energy harvesting, and optoelectronic devices [49]. At the most progressive principle level, optical antenna devices can be separated into passive and active classes [50]. Passive devices are characteristically linear, for instance, plane optics and light emission, whereas active devices are nonlinear, converting optical light signals into electrical current and enhancing optical signals, i.e., photodetection and light energy harvesting [51].

Graphene doping with certain chemicals and by electric/magnetic biasing portrays a remarkable tenability [52]. Highly doped graphene has recently emerged as an appealing platform for plasmonics due to its unique optoelectronic properties, which give rise to relatively long-lived, highly confined, and actively tunable plasmon resonances that mainly appear in the infrared and THz frequency regimes [53,54]. The unique properties of graphene are exploited in nanoantennas such as reflectarray antennas [55], frequency reconfigurable antennas [56] and leaky-wave antennas [57,58], resonating at much lower frequencies compared with their metallic counterparts [59].

2.3.4. Optoelectronic Modulators

Graphene has strong inter-band optical transitions due to the strong interaction with light. It has an extremely wide spectral range under different conditions, strong nonlinearity and tunable Fermi energy levels, giving it many advantages in photoelectric modulators [60,61]. Researchers have made a graphene-based optical modulator with a broad optical bandwidth (1.35–1.6 mm); small device footprint (25 mm²); and high operation speed (1.2 GHz at 3 dB) under ambient conditions, which is essential for optical interconnects for future integrated optoelectronic systems [9]. Christopher T. Phare et al. [62] reported a graphene electro-optic modulator that operates with a 30 GHz bandwidth and with a state-of-the-art modulation efficiency of 15 dB per 10 V. The graphene-based all-optical modulators (AOM) delivered the best performance among AOMs, presenting the smallest power threshold and fastest relaxation process [63]. Chi et al. fabricated a mid-infrared enhanced plasmonic modulator by integrating bilayer graphene in a reflective structure [64]. In comparison with monolayer graphene, bilayer graphene is highly doped to increase the carrier density in the modulator. The proposed modulator exhibits a modulation depth up to 21 dB and 3 dB with a bandwidth of 47.4 GHz over a wide range of wavelengths (3.17 μm to 4.4 μm). It can enable more efficient modulation with the same variation of Fermi levels than the modulator with monolayer graphene.

2.3.5. Graphene-Based Ultra-Fast Lasers

Graphene is an excellent wide-spectrum saturated absorber in the laser field because of its strong optical absorption capacity and fast carrier transport speed [65,66]. Lasers operating in the THz range can be used in a variety of applications because their beams can pass through many materials. However, these lasers have a fixed wavelength, greatly limiting their practical applications [67]. Graphene-based wavelengths can be changed in electric fields. Tunable graphene optical properties and GPs can be used to modulate the emission of a THz quantum cascade laser [68]. As illustrated in Figure 3, Marco Polini gained modulation by graphene plasmons in aperiodic lattice lasers using graphene to replace metal in a laser and placed them on the substrate in combination with aluminum gallium arsenide quantum dots and gallium arsenide wells with different thicknesses [69].

3. Transition Metal Dichalcogenides

3.1. Basic Structure and Property

TMDCs are a class of materials with the formula of MX₂, where M is the transition metal element from group IV (Ti, Zr, Hf, and so on), group V (for instance V, Nb, and Ta) or group VI (Mo, W and so on), and X is a chalcogen (S, Se or Te) [70]. The structure of TMDC is shown in Figure 4a, and the chalcogen atoms are in two hexagonal planes which are separated by a plane of metal atoms [71]. Monolayer TMDC is typically 6–7 Å thick, depending on the material [72].

MX₂ are indirect bandgap semiconductors [74,75], whereas their monolayers have direct bandgaps, which are favorable for optoelectronic applications [76]. Native bandgap provides an excellent current on/off ratio in field-effect transistors (FETs) [77]. Their bandgap can be tuned based on the layer thickness, greatly expanding their range of applications in photonic devices.

TMDCs have excellent optical properties and a wide range of application prospects in the fields of light detection and photoluminescence. MoS₂ is a commonly used MX₂ nanomaterial. Although the photoluminescence of bulk MoS₂ is not obvious, the monolayer MoS₂ has strong photoluminescence [73], so the monolayer MoS₂ can be used in solar panels, photodetectors, and photovoltaic emitters.

The photoluminescence of TMDCs is closely related to the layer thickness. As shown in Figure 4b, the photoluminescence of MoS₂ increases as the number of layers decreases, and the strongest photoluminescence is observed for a single layer [73]. Photoluminescence and Raman characterizations show that the direct bandgap can be blue-shifted for ~300 meV per 1% strain, which can effectively modulate its energy band structure [78]. In addition, the doping treatment can also significantly modulate the photoluminescence peak intensity of MoS₂ [79].

3.2. Production Methods

The preparation of large-area, high-quality 2D TMDCs is a prerequisite for electronic device applications, and there are two main types of preparation methods: top-down and bottom-up methods [80]. The top-down method is a method to obtain 2D TMDCs by chemical or mechanical exfoliation from the bulk material, while the bottom-up method is a method to prepare 2D TMDCs by CVD or thermal decomposition. Exfoliation procedures can produce 2D materials with the best accessible quality, and thus the method is widely used to prepare samples for fundamental physics and device studies [81]. However, this method is time-consuming and cannot be scaled up, which will provide significant challenges for large-scale implementation in the future. Physical and chemical processes provide another way to manufacture nanosheet materials, with the intention of achieving controllable, high-quality, and large-scale nanosheets. Among the numerous chemical synthesis methods including CVD, the hydrothermal method, laser-induced synthesis, and molecular beam epitaxy (MBE), CVD has the most potential for realizing controllable, high-quality, wafer-scale 2D TMDCs; the flowchart is shown in Figure 5 [82]. PVD is also a viable option for producing high-quality TMDCs on a large scale. PVD has been proven on TMDCs films for several decades with homogeneity across wide areas for various materials, and it is created at low processing temperatures. The most often utilized PVD-based technologies investigated for depositing TMDCs materials include sputtering and pulsed laser deposition (PLD) [81,83,84]. PLD is an alternate method for synthesizing TMDCs thin films with a wide area and homogeneity at low growing temperatures. Except for Si substrate, PLD is also suitable for various substrates such as Al₂O₃, GaN, and SiC-6H [85], on which MoS₂ shows a high degree of crystallinity and out-of-plane texture. Anti-position defects are usually observed in the samples that are prepared via the PVD method, while the vacancy defects that are usually found in the TMDCs films are prepared by the CVD method [86]. PVD approaches can obtain large-scale and uniform TMDC films for devices and circuits. However, they perform at a lower quality than chemical methods and exfoliation approaches because of small domains and numerous defects [87].

3.3. Optical Applications

Single-layer TMDCs are atomically thin and processable. They have primarily direct bandgaps; hence, they have great potential for applications in flexible and transparent optoelectronics.

3.3.1. Light-Emitting Diodes

A light-emitting diode (LED) is an optical source in which photons are generated by the electroluminescence (EL) effect [88,89]. Electrons and holes are permitted to recombine in a p-n junction so that the recombination energy can be released as photons in response to injected electrical bias currents. This is because of the availability of a large variety of direct bandgap TMDC monolayers with excellent photoluminescence quantum yield at sub-nanometer thicknesses [90,91,92]. TMDC materials have recently been investigated for their usage in LED fabrication. Due to several thermally aided processes resulting from impact ionization across a Schottky junction or a p-n junction [93,94,95], the EL is observed in various MoS₂ monolayer based devices [95]. However, because of the low optical quality of MoS₂ and the inadequate electrode connections, poor EL efficiency and considerable linewidth widening are found [96]. Due to the decreased contact resistance, increased current density, and the effective current injection, 2D TMDC-based heterostructure LEDs exhibit a high EL efficiency [96]. Electrically adjustable excitonic LEDs with p-n junctions based on WSe₂ monolayers may effectively inject electrons and holes; the corresponding structure is shown in Figure 6.

3.3.2. Solar Cells

As clean and renewable energy sources are widely studied, the photovoltaics (PV) effect has attracted the attention of many because it shows great application potential in energy harvesting. Due to the enormous surface area, lack of dangling chemical bonds on surfaces, and their potential as sunlight absorbers, 2D TMDC materials are regarded as promising prospects for solar applications. Two-dimensional TMDCs are applied in solar cells by forming a Schottky or p-n junction that works as an interface for the separation of charge carriers. TMDCs (MoS₂, MoSe₂, and WS₂) can absorb up to 5–10% incident sunlight in a monolayer (thickness less than 1 nm). The absorption of sunlight is one order of magnitude more than that of typical semiconductors such as GaAs and Si. The efficiency of solar cells based on ultrathin TMDCs is limited by the loss of absorption under the thickness limitation [97]. Calculations suggest that a monolayer TMDC could absorb as much sunlight as 50 nm of Si and generate electrical currents as high as 4.5 mA cm⁻² [98]. There are three ways to improve the performances of TMDC solar devices [99]: (1) careful control over the doping levels of TMDCs and Si, which would reduce the series resistance of the device; (2) utilization of large-area grown or deposited materials; (3) incorporating additional 2D semiconducting layers such as WSe₂ with complementary absorption spectra. In recent years, ultrathin solar cells have drawn much attention due to the possible reduction in cost and semiconductor material consumption, as well as their suitability for flexible and ultralight photovoltaics. A 120 nm thick MoS₂ p-n junction is presented [100]. Researchers have developed a straightforward method to fabricate ohmic contacts to both n and p MoS₂ and added an h-BN layer on top of the semiconductor to minimize the reflectance of the front surface. The homojunction device exhibits (3.8 ± 0.2)% efficiency and a 57% fill factor under AM1.5G illumination, which contributes to the maturity of the emerging technology of TMDC-based ultrathin solar cells. The incorporation of WS₂ as a photovoltaic material was presented [101]. With this optimized deposition parameter, WS₂ thin film was successfully fabricated for the very first time as a window layer in CdTe solar cells. The new device (ITO/WS₂/CdTe/CuC/Ag) exhibited V_oc = 0.39 V, J_sc = 10.45 mA cm⁻², fill factor = 29.42%, and efficiency = 1.2%. Shin et al. fabricated an organic solar cell (TFSA-GR/MoS₂/P3HT:PCBM/Al) by using MoS₂ as a hole transport layer, bis(trifluoromethanesulfonyl)-amide-doped graphene (TFSA-GR) as a transparent conductive electrode, and a GR quantum dots (GQDs)-added active layer [102]. They investigated the effect of the number of layers (L_n) of MoS₂ on the power conversion efficiency of organic solar cells. When L_n = 2, the fabricated solar cell shows a maximum power conversion efficiency of 4.23%. When L_n = 1, the power conversion efficiency is small because of the low absorption in the visible region. While L_n > 2, the power conversion efficiency is also lowered due to the lower amount of sunlight reaching the active layer. As a result, the selection of a suitable L_n of MoS₂ plays an important role in the performance of solar cells.

3.3.3. Photodetectors

Photodetectors have been the most extensively researched optoelectronic device for TMDCs, with research ranging from novel device topologies to utilizing unique physical features to strategies to improve performance [103,104,105,106]. A 3D schematic of supported and suspended channel FET architectures with ReS₂ as the channel material is shown in Figure 7 [107]. ReS₂ has been shown to have an anisotropic band structure resulting in two major excitons with unequal binding energy [108,109,110]. This allows control over polarized light absorption, and the consecutive polarization of a sensitive photodetector. The appropriate calibration of such photodetectors permits the detection of both light intensity and polarization, allowing for the detection of polarization-encoded messages. Some reports on photodetectors consisting of TMDC materials are listed in

Table 2. Photodetectors based on TMDC materials.

Material	Laser λ (nm)	Intensity (mW cm−2)	Detectivity (Jones)	Response Time (s)	Responsivity (A W−1)	Ref.
MoS2	635	0.1	7.7 × 1011	8.0 × 10−3	5 × 104	[111]
ReS2	532	-	1.2 × 1012	100	8.86 × 104	[112]
ReS2	589	0.12	-	3.2	654	[113]
MoTe2	1064	-	-	2.4 × 10−5	0.11	[114]
WSe2/ZnS2	520	100	4.7 × 1010	5.0 × 10−4	0.1087	[115]
ReS2	630	140	-	2.0 × 10−6	4	[107]
WSe2/ZnS2	550	3.77	1.3 × 1013	0.25	11.5	[116]
WS2	1310	-	3.0 × 1011	3.2 × 10−2	510	[117]
MoS2	1310	-	1.0 × 1011	0.2	103	[117]
MoS2-ZnCdSe	450	-	1.0 × 1012	1.2	3.7 × 10−4	[118]
MoS2	532	1	1.0 × 1012	0.5	150	[119]
MoS2	442	0.6	5.0 × 1011	-	186	[120]
AgBiBr6/WS2	455	0.071	1.5 × 1013	5.23 × 10−5	0.52	[121]

.

Mixed dimensional heterostructures and plasmonic-, organic material- or quantum dot-enhanced structures have been studied for performance enhancement [113,122,123]. According to preliminary research, photodetectors based on MoS₂ monolayers have better photoresponsivity than graphene-based devices [124,125]. MoS₂/SnSe₂ heterostructure-based photodetectors exhibit a high responsivity of up to 9.1 × 10³ A W⁻¹, which is significantly greater than MoS₂ only film-based photodetectors [126]. The photoexcited electron-hole pairs of GaTe-MoS₂ p-n heterojunction phototransistors are separated by large built-in potential, formed at the GaTe-MoS₂ interface efficiently to generate a self-driven photocurrent within <10 ms [127]. Two-dimensional TMDCs heterostructures can also be explored for possible THz detection at room temperature [128].

4. Other 2D Materials

4.1. Black Phosphorus

4.1.1. Morphology and Structure

Black phosphorus (BP) is a stable layered structure similar to 2D structures such as graphene and TMDCs [129]. Like bulk graphite, BP has a layered substance in which individual atomic layers are stacked together by van der Waals interactions and can be similarly isolated from black phosphorus by the mechanical exfoliation method. BP is not planer but puckered due to the sp³ hybridization. Each phosphorus atom is covalently bound with three adjacent phosphorus atoms inside a single layer to form a puckered honeycomb structure (Figure 8) [130].

4.1.2. Optical Properties

The direct bandgap (≈0.3 eV) of BP thin film (eight layers, or thickness > 4 nm) can link the energy gap between the zero bandgap of graphene and the comparatively large bandgaps of numerous TMDCs (1.5–2.5 eV) [16,132,133,134]. Unlike graphene and TMDCs, the bandgap of few-layer BP can be tuned by interlayer interactions and is strongly dependent on the number of layers.

Due to the “puckered” crystal lattice, BP shows strong band dispersion anisotropy [135]. The anisotropic absorption was identified with a maximum absorption value of 28% in the mid-infrared region when the incident light was polarized along the armchair direction of BP [136]. Thin-film BP has also been widely applied for mid-infrared ultrafast photonics as saturable absorbers due to the decent optical absorption. BP exhibits broadband absorption from the visible to the mid-infrared spectral region, which is also manifested by the broadband nonlinear optical response. Multilayer BP flakes of 3–20 nm thickness dispersed in liquid exhibit strong saturable absorption at both 400 nm and 800 nm, as measured by the z-scan technique and shown in Figure 9a,b [137]. In multilayer BP, one photon provides sufficient energy to excite an electron from the valence band to the conduction band under femtosecond laser excitation of 400 nm (3.1 eV) and 800 nm (1.55 eV) wavelengths, whereas in monolayer BP, two photons must arrive simultaneously, as shown in the inset of Figure 9c [137].

The robust optical conductivity of BP thin film in the wavelength range from 1 to 5 μm suggests BP as an attractive contender for near- and mid-infrared optoelectronics applications such as modulators, photodetectors, ultrafast optical switches, and, possibly, light-generation devices such as LEDs and ultrashort pulsed lasers [138].

4.1.3. Optical Applications

Numerous studies have also been performed on the applications of BP in photodetectors [139,140,141], polarization-sensitive detectors [142,143], saturable absorbers [144,145] and emitters [146], and photovoltaics devices [147,148].

Because of the direct bandgap and high carrier mobility, BP is promising in building broadband phototransistors for imaging and photodetection, especially in the near-infrared and mid-infrared wavelengths [149]. The response wavelength range of BP-based photodetectors can be increased by an external vertical electric field because the optical bandgap of few-layer BP is highly adjustable by the quantum-confined Stark effect [150,151,152]. Different portions of a single BP flake can be adjusted into opposing doping regimes to create an artificial PN junction that can generate a photocurrent via the photovoltaic effect. Because of the versatility and compatibility of 2D BP, it can be combined with other nanomaterials to improve the performances of photodetectors [153,154,155]. BP can be combined with other low-dimensional nanomaterials such as nanowires and quantum dot [156,157], allowing for the creation of unique heterostructures based on materials with different/multiple dimensions. BP is also an excellent midinfrared electro-optic material for modulation applications due to its solid electro-optical response and relatively small bandgap.

4.2. 2D Transition Metal Carbides MXenes

Transition metal carbides, carbonitrides, and nitrides were first discovered and termed MXenes (MX) by Barsoum et al. in 2011 [158]. The common formulation of MXs is M_n+1X_nT_x, where M embodies the transition metals (such as Ti, Ta, Mo, and Cr); X signifies nitrogen or carbon, n may differ from 1 to 4; and T_x denotes surface ends on the outer M layers (such as -O, -OH, -F, and -Cl) [159]. M¹ and M² are two different Ms. In case two different Ms have in-plane arranging and form rotating chains of M¹ and M² molecules inside the similar M layer, the ensuing MX configuration is known as i-MX with a general formula of $(M_{4/3}^1{M}_{2/3}^2)$XT_x, where the percentage of every compound is mentioned as a decimal value. The M¹ and M² molecules positioned in distinct atomic planes holding an out-of-plane arrangement are known as o-MXs, in which M² molecules compose the internal layers and M¹ molecules are in the exterior surface. o-MXs are represented by two chemical formulations as $\left(M_2^1{M}^2\right)X_2{T}_x$Tx and $\left(M_2^1{M}_2^2\right)X_3{T}_x$. MXs with V, Ti, Nb, Cr, Mo, Zr, Sc, Hf, Ta, Y, and W in the M spot were investigated. W, Sc, Y, and Cr have only been registered as elements of i-MXs, o-MXs, or in combination with the other metals listed above [160].

4.2.1. Morphology and Structure

MXenes consist of a transition metal-made (M) hexagonal close-packed crystal symmetry, and the X molecules occupy an octahedral spot between the adjacent M layers [159]. MXenes can be classified as ordered mono-M, ordered double-M and solid-solution M elements based on their atomic lattices and composition, as shown in Figure 10 [161].

4.2.2. Optical Properties

MXenes have unique optical properties and a wide range of tunability, making them ideal materials for a variety of optical applications. Because of their large density of states at the Fermi level, MXenes have intriguing transport features, while TMDCs have low carrier mobility. MXene films have been found to exhibit a broadband optical transmittance of more than 90% [162,163], with a transmission valley of around 750–800 nm [164], which has been attributed to the surface plasmon resonance (around 780 nm) and inherent out-plane interband transitions (around 800 nm). When compared with the pristine MXene, the oxidized sample has a higher absorption in the visible region, while fluorinated and hydroxylated MXenes have a lower absorption. In the ultraviolet optical region, all functional groups lead to an enhancement of both the absorption and reflectivity of MXenes [165]. M_n+1X_n with a smaller n is expected to be more transparent than MXenes with a larger n due to the lower density of states [165,166].

Both MXenes quantum dots and nanosheets exhibit strong photoluminescence due to the direct bandgap, which allows for radiative electronic transitions [167,168]. MXene quantum dots typically have three absorption peaks at 260 nm, 310 nm, and 350 nm, depending on the particle size and composition. The excitation wavelengths have a big impact on the photoluminescence spectral band. For example, as shown in Figure 11, the photoluminescence spectra of Ti₃C₂ quantum dots range from 400 to 600 nm when the excitation wavelengths vary from 340 to 440 nm [167].

4.2.3. Optical Applications

The MXenes such as Ti₃C₂T_x slim layers are well-suited for optoelectronic applications which require adaptive translucent conductive electrodes due to their high transparency and low layer resistance [169]. Translucent conductive electrodes based on MXenes have been used to provide appropriate volumetric capacitance in translucent solid-state supercapacitors [162]. Except translucent conductive electrodes, MXenes also show great potential applications in other optical applications. For example, Ti₃C₂T_x nanostructures have been widely used to realize ultrafast pulses in different lasers. In 2017, Jhon et al. reported MXene-based SA-activated ultrafast pulses, which achieved mode-locking at 1.5 m-band and q-switching pulses at 2 m-band, respectively [170]. MXenes exhibit excellent broadband saturable absorption capability in the near-mid-infrared region [171]. In 2019, Yi et al. reported the broadband nonlinear properties of MXenes in the three-wavelength band, in which mode-locked pulses operating at 1051 and 1565 nm and q-switched pulses operating at 2798 nm were realized using the same Ti₂CT_x-based SA [172]. A total of 1.2 nm of spin-coated Ti₃C₂T MXene film on glass, quartz and polyetherimide substrates achieves an ultra-low optical attenuation of 3% in the visible region [173]. It is noteworthy that the prepared Ti₃C₂T MXene films have a transmittance of up to 98% and can be used for flexible and conductive high-capacity capacitors [162,174]. In addition, Ti₃C₂T_x nanosheets can be easily oxidized to TiO₂ during the delamination process. If stored in an oxygen-rich atmosphere, Ti₃C₂T_x-TiO₂ nanocomposites can be obtained in situ, further expanding their applications in electrochemical and optoelectronic devices. In 2008, Mochalin et al. prepared Ti₃C₂T_x-TiO₂ nanocomposites by the in situ oxidization of Ti₃C₂T_x for UV photodetection [175]. Ti₃C₂T_x MXenes can also serve as electron and hole collection layers in solar cells [176], accelerating charge extraction and improving photovoltaic efficiency. Meanwhile, doping Ti₃C₂T_x MXenes into the electron transport layer, hole transport layer or activation layer can change the energy level and improve the electrical conductivity and carrier mobility. For example, Agresti et al. used Ti₃C₂T_x with different termination groups (T_x) to tune the work function of the perovskite absorber and the TiO₂ electron transport layer, as well as to engineer the perovskite/electron transport layer interface [177]. The combined effect of work function tuning and interface engineering can significantly improve the performance of MXene-modified perovskite solar cells. Compared with the reference cell without MXene, the power conversion efficiency was improved by 26% and the hysteresis was reduced.

4.3. Hexagonal Boron Nitride

4.3.1. Morphology and Structure

hBN is one of the key 2D materials composed of alternating B and N atoms bonded by sp² covalent bonds in its hexagonal lattices. hBN crystal contains alternating B and N atoms in a base plane and, thus, has great variations in growth kinetics due to its different edge terminations. As a result, the morphologies of h-BN domains are varied with its B or N terminations (Figure 12) [178].

4.3.2. Optical Properties

Despite the fact that hBN is an indirect bandgap semiconductor (with an energy gap of about 6 eV) [179], it has a very high internal quantum efficiency for deep UV emission (up to 40%). The frequency of the normal lattice vibrational modes (optic phonons) is similarly very anisotropic due to the highly anisotropic crystal structure of hBN, with two different optic phonon branches. When monolayer hBN was placed on different substrates and under THz irradiation, the carrier density could be enhanced considerably, especially on SiO₂/Si substrate, which suggested that the SiO₂/Si substrate was more appropriate for the construction of monolayer hBN-based electronic and optoelectronic devices when compared with quartz, PET, and sapphire [180]. The wavelength responses of S-doped hBN monolayer films on molten Au substrates are extended to 280 nm, and the photocurrent and responsivity for light irradiation with a wavelength of 280 nm are ~50 times higher than that of pristine hBN, as shown in Figure 13a,b [181].

4.3.3. Optical Applications

h-BN shows great potential applications in many optical areas and some applications are summarized in

Table 3. Optoelectronic devices based on hBN.

Applications	Materials	Intensity	Detectivity (Jones)	Response Time	Responsivity	Ref.
Photodetectors	hBN/Cu	9.937 μW cm−2	6.1 × 1012	0.2 s	5.022 A W−1	[185]
	graphene/hBN/n-AlGaN	16.1 μW mm−2	1.76 × 1012 (monolayer hBN)	1.42 s	3.63 A W−1 (280 nm)	[186]
			2.05 × 1012 (bilayer hBN)	1.14 s	3.54 A W−1
			9.88 × 1010 (trilayer hBN)	1.13 s	0.24 A W−1
	hBN/black arsenic phosphorus/hBN				190 mA W−1 (3.4 μm)	[187]
					16 mA W−1 (5.0 μm)
					1.2 mA W−1 (7.7 μm)
	S-doped h-BN on Au substrate	0.2 μW mm−2			~0.018 mA W−1 (280 nm)	[181]
					~1 mA W−1 (230 nm)
	Materials	Power conversion efficiency	Voc	Jsc	FF
Solar cells	ITO/I-INP/P-INP/5 ML hBN	17.2%	0.78 V	29.4 mA cm−2	75.2%	[184]
	ITO/I-INP/P-INP/7 ML hBN	15.7%	0.80 V	27.1 mA cm−2	72.1%
	MoS2/WSe2/h-BN surface passivation layer	Enhance 74%	0.38 V	1.69 mA cm−2	-	[188]

. h-BN can be used as a wide bandgap semiconductor for deep-UV emitters and detectors because of its large bandgap. To further improve the response rate and detection rate, it is necessary to modulate the optical and electrical properties of hBN. Tan et al. successfully prepared S-doped hBN monolayers on molten Au substrates by atmospheric CVD, and they systematically studied their optical and electrical properties [181]. The band gap is reduced from 5.83 eV to 5.69 eV, and the conductivity is reduced by 140 meV. Moreover, the response range of the photodetector is extended to a 280-nm wavelength; the power density of the light source is 16.1 μW mm⁻² (280 nm); and the responsiveness of the device with single, double and three-layer hBN is 3.63, 3.54 and 0.24 A W⁻¹, respectively. In addition, numerous studies have shown that hBN is also an excellent insulator that can be used as a substrate material for graphene-based devices [182,183]. Wu et al. demonstrated a deep ultraviolet photodetector based on a graphene/hBN/n-AlGaN heterostructure. The device not only exhibits high responsivity under UV illumination, but also exhibits a high UV/Vis rejection ratio. Several nanolayered graphene-hBN heterostructures are used to enhance the performance of photodetectors, successfully solving the strain problem between graphene and traditional bulk insulators [25]. Two-dimensional hBN film can provide almost ideal passivation due to its wide bandgap, no dangling bond and high dielectric constant. Raj et al. reported the passivation properties of monolayer hBN that was grown by metal-organic CVD [184]. Using an ITO/I-INP/P-INP solar cell structure with a small amount of hBN monolayer as the passivation layer, the solar cell performance is significantly improved. The maximum efficiency of five-monolayer (ML) hBN is 17.2%, V_oc is 0.78 V, J_sc is 29.4 mA cm⁻², and FF is 75.2%. The INP solar cell with 7  ML hBN has a measured efficiency of 15.7%, V_oc, J_sc and FF are 0.8 V, 27.1 mA cm⁻² and 72.1%, respectively [184].

4.4. 2D Metal Oxide

4.4.1. Basic Structure and Optical Properties

Two-dimensional metal oxides, formed by the O element and the metallic element, are gradually attracting more attention [189,190]. This is because they are not only rich in material species, but also contain a variety of structures including a nonlayered structure, a layered structure such as MoS₂, and inorganic molecular crystal. The special structure of 2D metal oxides endow these materials with numerous outstanding properties. Compared with typical 2D materials such as graphene and BP, 2D metal oxides possess many special characteristics. For example, they have excellent air stability due to the involvement of O and the relatively stable valence state of the elements that are contained in 2D metal oxides [191]. Second, 2D metal oxides generally have low requirements for a synthetic environment. Some 2D metal materials can even be obtained in an atmospheric environment. Third, most reported 2D metal oxides have wide bandgaps and exhibit excellent ultraviolet detection performances. As a result, they show great application potential in optoelectronics.

4.4.2. Optical Applications

Due to the unique performances of 2D metal oxides, they have been widely used in the fabrication of photodetectors, and some applications are summarized in

Table 4. Photodetectors based on 2D metal oxides.

Materials	Wavelength	Responsivity	Detectivity	Ref.
ZnO	254 nm	2.0 × 104 A W−1	6.83 × 1014 Jones	[192]
Fe3O4	10,600 nm	562.1 A W−1	7.42 × 108 Jones	[194]
Bi2O3	365 nm	400 A W−1	1.1 × 1013 Jones	[193]
α-MoO3	254 nm	67.9 A W−1	-	[195]
Β-Ga2O3	254 nm	335 A W−1	-	[196]
MgO	150 nm	1.86 A W−1	1.8 × 1010 Jones	[197]

. Recently, Yu et al. fabricated a photodetector based on ZnO nanosheets [192]. At a visible-blind wavelength of 254 nm, this photodetector exhibits a high responsivity up to 2.0 × 10⁴ A W⁻¹ and high detectivity (6.83 × 10¹⁴ Jones). Messalea et al. reported an ultra violet photodetector based on ultrathin bismuth oxide (Bi₂O₃) nanosheets that were synthesized by a liquid metal facilitated approach [193]. The photodetector shows a high responsivity of about 400 A W⁻¹ and a fast response time of ~70 μs at the wavelength of 365 nm. In addition to the application of photodetection in ultraviolet, 2D metal oxides can also be used in the long-wave-length infrared (LWIR) detecting devices. Yin et al. synthesized air-stable nonlayered ultrathin Fe₃O₄ nanosheets by a space-confined CVD approach [194]. The prepared Fe₃O₄ nanosheets can be used to fabricate high-performance ultrabroadband photodetectors with a detection range from ultraviolet to LWIR. Due to the synergistic mechanisms of the photoconductive effect and bolometric effect, the photodetectors exhibit an ultrahigh photoresponsivity of 562.1 A W⁻¹, a detectivity of 7.42 × 10⁸ Jones, and an external quantum efficiency of 6.6 × 10³% at the laser wavelength of 10.6 μm. Some 2D materials have a low symmetric crystal structure, leading to in-plane optical anisotropy [191]. Zhong et al. synthesized large-size 2D α-MoO₃ single crystals with structural in-plane anisotropy, which showed a remarkable ultraviolet photoresponse and electron transport anisotropies [195]. The photodetectors based on α-MoO₃ demonstrate a photoresponsivity of 67.9 A W⁻¹ and an external quantum efficiency of 3.3 × 10⁴% under solar-blind ultraviolet light (254 nm). Importantly, the photodetectors exhibit strong in-plane anisotropy in the optoelectronic response and transport properties. Although some photodetectors based on 2D metal oxides show excellent performances, there is still a long way to go for them to achieve practical applications. Exploring methods to produce high-quality 2D metal oxides is one of the effective ways to drive the further development of high-performance photodetectors. We believe that with the deepening of research, more and more photodetectors with outstanding performances will be fabricated.

5. 2D Heterojunction Materials

5.1. Morphology and Structure

Atomically thin 2D materials with a wide range of properties can be manufactured and engineered separately and then stacked together to form van der Waals heterostructures (such as graphene-hBN, graphene-BP, TMD-hBN, graphene-TMDs, and TMD–TMD) [198], resulting in surprising novel optical properties for functional devices [199]. Different atomically thin 2D materials can be simply vertically stacked together to produce 2D heterostructures, which not only avoids the traditional issue of lattice mismatch, but can also introduce many outstanding physical or optical properties. The most widely utilized 2D material in heterostructures is hBN, primarily owing to its role as the ‘ideal substrate’ for graphene [200,201].

5.2. Optical Properties

Three types of heterojunctions (types I, II, and III) can be easily built by employing chosen materials based on various combinations of band alignments, depending on the energy difference between the conduction and valence-band extrema in the constituent layers (Figure 14) [202]. The valence band maximum and conduction band minimum of two separate components are positioned on the same side of the heterointerface in type-I (straddling gap) heterojunctions, and radiative recombination is enhanced as both electrons and holes reside in one material. Type-I heterojunctions have a quantum well structure, which makes them suitable for optoelectronic devices such as LEDs. The conduction band minimum and valence band maximum in type-II (staggered gap) heterojunctions belong to two independent components with different work functions, providing chances to regulate the interlayer transition energy and to generate charge spatial separation. When the valence band maximum of one semiconductor is higher than the conduction band minimum of the other semiconductor, it consists of type-III (broken gap) heterojunction, and this unique electronic structure is very useful for tunnelling devices such as negative differential resistance devices.

The MoSe₂/BP and WSe₂/BP vdW heterostructures possess a direct bandgap, which is characterized by type-II band alignment, and demonstrate powerful built-in electric field across the interface, which can effectively separate the photogenerated charges. Figure 15 shows the light absorption capacity of MoS₂/BP, MoSe₂/BP, WS₂/BP, and WSe₂/BP vdW heterostructures, which share a similar good ability to absorb the visible and near-infrared light, enabling them to be used in photocatalytic, photovoltaic, and optical devices [203]. The WS₂/GaN vdW heterobilayer structure exhibits significant optical absorption for the visible and near-ultraviolet light, which can be further improved by in-plane biaxial tensile and vertical uniaxial strain [204].

5.3. Optical Applications

The unique electronic and optoelectronic properties of heterostructures bring promising perspectives for the potential applications in the next generation of atomically thin electronic and optoelectronic devices such as LEDs [205], photodetectors [206,207,208,209], photovoltaics [129,133], and other innovative functional devices [202]. Recently, a vdW-integrated LED-based on monolayer WSe₂−CdS nanoribbon hybrid structure was demonstrated, and its structure is shown in Figure 16 [205]. Room temperature electrically driven light emission from monolayer WSe₂ was successfully achieved, with a turn-on voltage of ~2.0 V. By taking advantage of the CdS nanoribbon waveguide, the efficient optical routing of WSe₂ photoluminescence and electroluminescence emission was realized, showing great potential to interconnect with other functional optoelectronic units. A sub-band-gap photodetection in the MoS₂/2D Ruddlesden−Popper perovskite vdW heterojunction was obtained with a strong interlayer charge transfer caused by the type II energy band alignment (Figure 17) [210]. The reduced energy interval facilitates the photodetection ability in the near-infrared region, with the maximum photoresponsivity of 121 A W⁻¹, a detectivity of 4.3 × 10¹⁴ Jones at λ = 860 nm, and a rapid rise/decay (8.2/28.6 μs) time. The BP/MoSi₂P₄ heterostructure has promising prospects in optoelectronic devices such as solar cells due to its high stability and a typical type-II band alignment with a direct band gap, which can effectively facilitate the separation of photogenerated electron−hole pairs [211]. Based on the findings, the predicted photoelectric conversion efficiency for the BP/MoSi₂P₄ bilayers can reach 22.2%, which is larger than many other existing heterostructures. Two-dimensional materials-based heterostructures offer potential avenues for the continued development of novel photodetectors with broad spectral ranges. The various reported photodetectors based on different materials are summarized in

Table 5. Comparison of the performances of photodetectors based on different materials.

Materials	Properties	Ref.
GaTe/MoS2 p-n heterojunctions	Photoresponsivity: 1.36 A W−1 (at 633 nm); external quantum efficiency: 266%	[127]
Graphene/GaSe/WS2/Graphene heterojunction	Photoresponsivity: 149 A W−1 (at 410 nm)	[212]
WS2/Si heterojunction	Photoresponsivity: 224 mA W−1 (at 200–3043 nm)	[213]
InSe/Se heterojunction	Photoresponsivity: 110 mA W−1 (at 460 nm); external quantum efficiency: 8.7%	[206]
PdSe2/MoSe2 heterojunction	Photoresponsivity: 651 mA W−1 (at 532 nm)	[214]
InGaAs/SiO2/graphene heterostructure	Photoresponsivity: 103 A W−1 (under weak light irradiation)	[215]
VO2/MoTe2 heterostructure	The heterostructure can realize 3 different functional modes: p-n junction mode, Schottky junction mode, and bolometer mode. Detection range: from visible light to longwave infrared radiation	[216]

.

6. Conclusions and Outlook

The discovery and development of 2D materials provides many exciting new chances for the exploration of high-performance optical devices. In this review, we introduced the properties as well as the fabrication methods of 2D materials and summarized their applications in the optoelectronics applications such as modulators, photodetectors, solar cells, and ultra-fast lasers. Hybrid integration with 2D materials can improve the performances of traditional optoelectronics devices, further expanding the application range of 2D materials; however, 2D materials-based optoelectronic devices have already achieved some milestone achievements. To realize industrial production, many problems still need to be considered and solved. First, convenient and cost-effective approaches should be explored to develop high-quality 2D material films. Second, many novel properties of 2D materials such as mechanical and magnetic properties remain to be further explored. Third, some high-performance 2D heterojunction materials or hybrid materials by the combination of the 2D materials and nanomaterials need further exploration and research. Fourth, the long-term operation stability of some optoelectronic devices in air needs to be enhanced as well, which plays a determining role for practical applications. Finally, researchers should delve into the mechanism of optoelectronic devices. The comprehensive understanding of optoelectronic devices will benefit the fabrication of high-performance devices. With continuous investigation, we believe that more and more practical optoelectronic devices with outstanding properties will emerge in the near future.