Recent advances in 2D organic−inorganic heterostructures for electronics and optoelectronics

Abstract

Two‐dimensional (2D) materials show outstanding properties such as dangling bond‐free surfaces, strong in‐plane while weak out‐of‐plane bonding, layer‐dependent electronic structures, and tunable electronic and optoelectronic properties, making them promising for numerous applications. Integrating 2D inorganics with organic materials to make van der Waals heterostructures at the 2D thickness limit has created new platforms for fabricating on‐demand multifunctional devices. To further broaden the limited choices of 2D inorganic‐based heterostructures, a wide range of available 2D organic materials with tunable properties have opened new opportunities for designing large numbers of heterostructures with 2D inorganic materials. This review aims to attract the attention of researchers toward this emerging 2D organic−inorganic field. We first highlight recent progress in organic−inorganic heterostructures and their synthesis and then discuss their potential applications, such as field‐effect transistors, photodetectors, solar cells, and neuromorphic computing devices. In the end, we present a summary of challenges and opportunities in this field.

INTRODUCTION

Two-dimensional (2D) materials have opened a new window with their structural, thermal, and mechanical properties for a wide range of advanced applications in electronics, optoelectronics, electrocatalysis, and several other fields.^1–10 Despite having great premises, graphene manifests zero bandgaps, large dark currents, and a low on/off ratio, limiting its application in field-effect transistors (FETs).¹¹ Analogous to graphene, transition metal dichalcogenides (TMDCs), transition metal carbide (MXene), and hexagonal boron nitride (h-BN) were discovered to complement graphene.^12–14 The 2D material family covers a broad spectrum of metal, semimetal, semiconductor, and insulators showing remarkable properties such as high surface areas, quantum confinement effects, and spin-orbit coupling, making them highly demanded to meet the requirement of multifunctional device applications.^15–19 Regardless of the extraordinary response of 2D materials, critical limitations and pending challenges remain to be addressed, including low optical absorption, the limited number of p-type semiconductors, defect-free growth, and scalable synthesis.^20–23 Unlike inorganic materials, organic materials consisting of small molecules have abundant p-type semiconductors that can be grown by low-cost syntheses.^24–28 The excellent light absorption makes them attractive for solar cells, photodetectors, and organic light-emitting diodes (OLEDs).^29–32 Nevertheless, having unique properties, organic materials still suffer from the limitations of low carrier mobility, inferior thermodynamic stability, disordered growth on conventional substrates like SiO_2, and low carrier injection.^33–36

Since both organic and inorganic materials have their pros and cons; however, the tendency of organic−inorganic heterostructures makes them suitable for futuristic flexible, and wearable devices. Interestingly this may lead to the broadening of the device's capabilities, endowing them with superior properties by taking the benefits of both materials. For example, inorganic materials, with their ultra-flat dangling bond-free surfaces and high carrier mobility, could serve as template substrates for organic materials, assisting the high-quality growth of organic materials with a sharp interface providing excellent charge transportation and paving the way for the fabrication of high-performance electronic and optoelectronic devices. In addition, organic materials with their strong tunable optical absorption have boosted the performance of 2D TMDCs (MoS₂, WSe₂, WS₂, MoSe₂) based photodetectors by extending their spectral detectivity range with high responsivity and ultrafast charge dissociation.^37–40 Furthermore, they are also helpful in tuning the electronic properties of inorganic materials such as n-type and p-type dopants.⁴¹ These materials have shown tremendous progress for defects healing of sulfur-containing TMDCs such as MoS₂, resulting in improved carrier mobility for FETs,^42,43 ultra-fast charge carrier dissociation and slow recombination in organic−inorganic heterostructure provide an excellent choice for solar cell applications.^44,45 2D organic materials are also used to improve perovskites-based devices by enhancing their spectral range from ultraviolet (UV)-Visible to NIR.^46,47 Organic−inorganic heterostructures have shown potential applications in optoelectronics,^48,49 p-n junctions,^50,51 catalysis,⁵² photovoltaics,⁵³ and neuromorphic computing devices.^54,55

Therefore, combining organic and inorganic materials creates challenges and exciting opportunities for researchers to improve existing devices with new fabrication methods. It seems that individual materials cannot accomplish all the properties to meet the fast-growing demands of multifunctional devices. Their heterostructure delivers a platform by taking advantage of both materials and integration in a single device with additional amended properties rather than an individual material. Compared with typical 2D-2D inorganic heterostructures, organic−inorganic-based heterostructures offer several advantages. First, compared with a limited number of available 2D inorganic-based heterostructures, large p-type organic materials can easily be combined with underlying 2D inorganic materials with controlled and tunable properties to fabricate novel devices at the molecular level. Second, by taking advantage of both organic and inorganic material properties, more advanced and multifunctional devices can be fabricated. Third, the underlying inorganic material's properties can easily be modulated by functionalizing those using organic materials, paving the way to design devices with desired properties. Therefore, organic−inorganic heterostructures have the potential to revolutionize further the existing materials systems, device architectures, and fabrication methods. This review aims to appeal to researchers in this newly emerging field by highlighting recent progress offered by organic–inorganic heterostructures. A comprehensive study of organic−inorganic heterostructures is presented by combining sufficient knowledge of materials, hybrid structures, synthesis methods, and their potential applications in the field of electronics and optoelectronics, as shown in Figure 1. Finally, we summarize the 2D layered organic−inorganic heterostructure, current limitations, and prospects.

Figure 1. The organic−inorganic-based potential applications include electronic devices such as FETs and optoelectronic devices such as photodetectors, solar cells, and neuromorphic synaptic devices. FET, field-effect transistor

STRUCTURE AND PROPERTIES OF 2D INORGANIC AND ORGANIC MATERIALS Overview of inorganic materials

2D inorganic materials are considered attractive for researchers due to their unique electronic, optoelectronic, and bandgap tunability, covering the broadband spectrum with various materials, semimetal, semiconductors, and insulators. They are single-layer materials exfoliated chemically or mechanically from their bulk part due to weak intermolecular forces present in stacked layers.^{1,8,15,56–62} The strong in-plane covalent bonding offers extraordinary mechanical strength while weak out-of-plane forces among stacked layers help to extract a single layer called 2D materials from their bulk counterparts. Depending on the composition, they can have a single-element plane like graphene and h-BN or multiatomic plane like MoS₂ with three atomic layers. The mechanical exfoliation of single-layer graphene from graphite opened a window for exfoliation of other 2D layered materials like TMDCs, black phosphorus (BP), h-BN, and MXenes with tremendous advancement in electronic and optoelectronic applications. We will discuss here a short introduction to these materials and their properties.

The zero bandgap of graphene led researchers to open bandgaps in graphene using different approaches such as stress, electric field, nanoribbons, and chemical functionalization. Meanwhile, it was realized to find other 2D layered materials exhibiting intrinsic bandgaps to overcome these challenges in next-generation electronic devices.^63–66 In contrast to graphene TMDCs, layered materials display intrinsic layer-dependent bandgap, high mobility, air stability, and mechanically robust due to quantum confinements and surface effects, providing more exciting opportunities for advanced electronic and optoelectronic applications with a large variety of available materials having suitable bandgaps of 1–2 eV.^67–69 Among TMDCs, MoS₂ is the most studied material due to its tunable layer-dependent bandgap, 1.2 eV indirect bandgap for bulk, and 1.8 eV direct bandgap for monolayer.⁷⁰ The high theoretical predicted carrier mobility (410 cm²/(V·s)) and large optical absorption due to strongly bounded excitons make it an excellent choice for optoelectronic applications such as high-performance photodetectors. Although WSe₂, MoS_2, and WS₂ show similar electronic properties having a monolayer direct bandgap suitable for optoelectronic applications, the MoS₂ and WS₂ are n-type, and WSe₂ is a p-type ambipolar. Depending on the composition of TMDCs materials, they represent metallic NbS₂,⁷¹ semiconductor MoS₂,⁷² superconductor NbSe₂,⁷³ and ferroelectric VSe₂⁷⁴ materials.

BP is another type of layered 2D material with high electron mobility around 1000 cm²/(V·s) with a narrow bandgap ranging from 0.3 eV for bulk to 1.5 eV by thinning down to monolayer.^75,76 BPs’ composition and electronic properties can be continuously tuned by arsenic doping to reach the long-wavelength infrared regime.^77,78 Though BP demonstrates extraordinary properties, it has to be used in an inert environment such as a vacuum or extra protection to prevent environmental degradation.^79,80 Silicene, germanene, and stanene belong to the 2D elemental family called Xenes.⁸¹ Apart from this, there are other 2D materials such as monochalcogenides SnS, GeSe, trichalcogenides TiS₃ or HfS₃, metal halides such as PbI₂ or Crl_3, and highly conducting metallic MXenes.^82–85 Further boron nitride (BN) is analogous to graphene, having a hexagonal structure, but a wide bandgap of approximately 6 eV shows insulating behavior. It is used as dielectric materials or as an encapsulation layer and as a template for the growth of organic materials to enhance their device performance.^86–88 Recently, Wu et al.^89,90 demonstrated a new 2D material, Bi₂O₂Se, with extraordinary properties, high electron mobility >20,000 cm²/(V·s), narrow bandgap 0.8 eV, near-ideal subthreshold swing around 65 mV/dec and most importantly, it shows high air stability, which created many exciting opportunities for electronic and optoelectronic applications specially ultra-fast photodetectors. The properties of 2D inorganic materials as explained above show that these materials can be used for the fabrication of a wide range of potential applications.

Organic materials

Organic semiconductors have attracted incredible attention among researchers because they offer outstanding opportunities due to their low cost, flexibility, large optical absorption coefficients, and large numbers of available materials with desired tunable properties.^91,92 In 2D organic semiconductors, the π–π stacking helps in carrier conductivity. Also, the unit photoluminance exhibited by organic materials at the nano and sub-micro-level makes them highly desirable materials for flexible electronic and optoelectronic devices.^20,93 Over the last few decades, researchers have made remarkable progress in improving the carrier mobility of organic semiconductors over 10 cm²/(V·s); even the mobility of about 50 cm²/(V·s) is also achieved with fluorescence efficiency reaching up to 90%, which is extremely necessary for optoelectronic devices.⁹⁴ Organic photovoltaic devices (OPVs) have also accomplished power conversion efficiency (PCE) of more than 14%.⁹⁵ Unlike 2D inorganic materials, organic semiconductors can be synthesized using a low-cost molecular design with desirable and tunable properties within a large area, making them attractive for low costs, flexible, and lightweight devices like photodetectors organic field-effect transistor (OFET), OLED, organic light-emitting transistors, and electrically pumped lasers.^30,96,97

All the achievements mentioned above possessed by organic semiconductors are due to focused research made by scientists over the last few decades by improving synthesis protocols, crystal quality, interfacial quality studies, and understanding of essential and novel physics made by organic semiconductors. In organic semiconductors, the optical light absorption and emission are done by electron and hole separation and recombination through electronic transfer from HOMO, which is known as the highest occupied orbital level, to the LUMO, known as the partially filled lowest level, which strongly depends on the energy band alignments between HOMO and LUMO. Similarly, the charge carriers, i.e., electrons and holes, move through organic semiconductors by electronic cloud coupling among neighboring molecules. They are further classified into two categories, i.e., small organic molecules such as rubrene, pentacene, C₆₀, CuPc, dioctylbenzothienobenzothiophene (C8-BTBT), AIq₃, and organic polymers such as polystyrene, polypropylene, and proteins (Figure 2).⁹⁸

Figure 2. Overview of the most frequently studied 2D inorganic and organic materials with their crystalline and molecular structures. Reproduced with permission: Copyright 2019, Keai98

Like 2D inorganic materials, organic semiconductors exhibit layer-dependent carrier transport properties, especially in pentacene and C8-BTBT. More recently, an emerging 2D organic material, 2,6-diphenylanthracene (DPA), was synthesized using a vacuum deposition method, showing a high-quality crystalline structure. To investigate the optical response, a phototransistor was fabricated, which exhibited excellent results such as low dark current (10⁻¹² A), high photosensitivity, and photoresponsivity (1.34 × 10⁵ A/W).⁹⁹ Similarly, C8-BTBT is another common 2D organic material with outstanding properties such as high carrier mobility (30 cm²/(V·s)) and a band-like transport even at 150 K. 2D C8-BTBT exhibits excellent performance for OFET because of its ultrathin 2D nature, enabling it to be direct and nondestructive contact with electrodes, unlike bulk organic materials.¹⁰⁰

In another work, Heeger et al.¹⁰¹ studied the single-crystalline organic (2,6-diphenylanthracene, DPA) having high carrier mobility of about 34 cm²/(V·s) with strong emission and PLQY of about 42%. Further, the OLED showed blue emission having brightness up to 66,627 cd/m² with turn and on voltage about 2.8 V. 2D metal-organic framework (2D MOF) is another newly developed and highly tailorable electrically conducting material showing potential applications in optoelectronics. Fe₃(THT)₂(NH₄)₃ (THT: 2,3,6,7,10,11-triphenylenehexathiol) is one of the newly emerging 2D MOFs exhibiting high carrier mobility, band-like charge transport, and narrow bandgap (0.45 eV). To explore optoelectronic properties, a photodetector was fabricated operating in photoconductive mode. The photodetector showed excellent results, such as spectral detection ranging from UV to NIR (400–1575 nm) and high specific detectivity of 7 × 10⁸ cmHz^1/2/W at a wavelength of 785 nm.¹⁰² Overall, these results show that 2D organic material can be used for various potential applications with low cost and easy fabrication. However, their performance is still far behind in achieving many practical applications.

SYNTHESIS OF INORGANIC AND ORGANIC MATERIALS

The controllable synthesis of 2D materials with excellent quality and efficiency is essential for large-scale applications. Various strategies have been employed for the growth of 2D inorganic materials. The most prevalent are mechanical exfoliation,^103–105 chemical exfoliation,^106–108 liquid-phase exfoliation (LPE),^109–111 and chemical vapor deposition (CVD).^112–114 The mechanical exfoliation method is a low-cost and straightforward method with the privilege of exfoliating 2D material, for instance, graphene from graphite using scotch tape. Although it is used to produce high-quality single or few-layer 2D material, which helps explore their intrinsic properties, the lack of controllability and scalability inhibit its practical applications.^115–117 Conversely, the chemical exfoliation method can produce scalable 2D materials with simple and low-cost methods.¹¹⁸ Nevertheless, the phase transformation of 2D materials using this method hinders its applications for high-performance electronic devices at a practical level.¹¹⁹ To a certain extent, the LPE technique can produce 2D materials with mass production without phase transformation of 2D materials.¹²⁰ The sample obtained using this method suffers uniformity issues with a variable thickness that requires different post-synthesis processes as ultracentrifugation, making it a complex and high-cost method.¹²¹ In comparison to above mention top-down methods, the CVD method is used to grow high-quality, low-cost, and large-area 2D materials with controllability and scalable synthesis.^112,122

On the other hand, organic semiconductors can be processed with a low-cost, versatile methodology and scalable ways with tunable desired properties compared to inorganic materials. Several growing techniques can be adopted for organic materials, such as floating-coffee-ring driven assembly,¹²³ layer-by-layer deposition,¹²⁴ spin-coating,¹²⁵ self-assembly,¹²⁶ Langmuir Blodgett,¹²⁷ patterned growth,¹²⁸ and physical vapor deposition (PVD) methods¹²⁹ as illustrated in Figure 3. Among all mentioned methods, PVD is prototypical and highly recommended to deposit organic vapors on the preferred substrate or 2D materials. By controlling temperature, evaporation rate, and speed parameters, one can quickly deposit almost every small organic molecule with desired thickness and suitable crystalline surfaces, which is crucial for high-performance devices.

Figure 3. The synthesis routes for organic materials include different approaches such as floating coffee ring, spin coating, Langmuir Blodgett, LBL deposition, self-assembly, and vapor deposition

SYNTHESIS OF ORGANIC–INORGANIC HETEROSTRUCTURES

Organic materials are easily hybridized with inorganic materials showing high crystallinity and sharp interfaces due to atomically flat and dangling bond-free surfaces of 2D materials, without considering the lattice matching required to deal with 2D–2D inorganic heterostructures.^2,115,130 The organic−inorganic heterostructures are classified into two types based on interactions at interfaces. (1) Weak intermolecular van der Waals (vdW) and electrostatic interactions. (2) Covalent interactions. The unique atomically flat and dangling bond-free surfaces of 2D materials allow a variety of organic materials to self-assemble via weak vdW interactions with atomically sharp interfaces. The intermolecular forces among organic molecules and weak vdW forces in the organic−inorganic interface, facilitate the self-assembly of organic molecules over inorganic materials. These weak vdW forces at the interface further facilitate highly ordered molecules with desired molecular orientations and stacking with a variety of inorganic materials to form van der Waal's heterostructures (vdWHs). Apart from weak vdW forces, the electrostatic interactions take place between the organic−inorganic interface due to charge transfer phenomena. Mostly, these organic molecules are donors or acceptors. Their type of nature decides the self-assembly of organic molecules, which can be used for altering the charge carrier concentration of underlying inorganic materials and the formation of p-n homojunction which is highly desirable for various potential electronic and optoelectronic devices. The second type of interaction between organic−inorganic heterostructure involves sharing of electrons at the interface, resulting in covalent interactions.¹³¹ The synthesis of two materials with different thermodynamic stability is quite challenging. Therefore, the most straightforward way is to deposit less stable organic material over presynthesized inorganic materials.¹³²

Three main strategies are employed to construct organic−inorganic heterostructures: mechanical exfoliation, vapor phase growth, and solution processing. The mechanical exfoliation method is considered the most straightforward approach for organic−inorganic heterostructures. In this process, the inorganic materials are typically exfoliated from their bulk part using a simple scotch tape method on the targeted substrate. Later, the organic material is deposited over exfoliated 2D material to form organic−inorganic heterostructure, e.g., Rubrene/MoS₂, Rubrene/Bi₂Se₃, DAP/graphene heterojunctions. The transferred organic material resides on the targeted inorganic material due to the strong adhesive force between the transferred organic material and the underlying targeted inorganic material. The main advantage of this strategy is that it offers the growth of atomically sharp and clean heterointerface. The obtained atomically sharp and neat interface is helpful to understand the interfacial studies such as the charge transfer mechanism which is very crucial for the suitable selection of organic−inorganic heterostructures. Although the mechanically exfoliated approach can produce high-quality single crystalline material, the lack of thickness controllability and scalability hinders its application for high-performance electronic devices at the commercial level.

Alternatively, PVD techniques such as thermal evaporation, physical vapor transport (PVT), and molecular beam epitaxy are considered the most effective tools for preparing organic−inorganic heterostructures with precise control of organic thickness. In the PVD method, organic materials are heated to their melting points, producing organic vapors in vacuum chambers and then condensing on targeted, pre-synthesized inorganic material to form heterostructures. This approach has certain advantages: organic material can be grown with preferred thickness, high crystalline quality, and inexpensive deposition over inorganic materials. The optimized thickness of organic material is controlled by tuning parameters such as pressure, temperature, and evaporation.¹³³ Further, this process also facilitates the evaluation of the thickness-dependent properties of organic−inorganic heterostructures. Wei et al.¹³⁴ studied the CuPc-MoS_2-based heterostructures for high-performance photodetectors. CuPc was thermally evaporated over CVD-grown MoS₂ with precise control over organic thickness. The as-fabricated photodetector device exhibited high responsivity, detectivity, ultra-fast response time, and external quantum efficiency (EQE) compared to pristine MoS₂-based photodetector only. This method allows the precisely controlled growth of organic materials over desired inorganic materials with good crystallinity. Further, this method is also very important as it allows for the investigation of thickness-dependent growth of organic materials and their structural effects on inorganic material. The literature study suggests that this method is the most widely used as almost every type of organic material can easily be deposited due to its low melting temperature.

Besides mechanical exfoliation and PVD, solution processing is low-cost, scalable, and straightforward for fabricating organic−inorganic heterostructures. Different approaches are used for this process, such as directly mixing organic material with inorganic material to form hybrid heterostructures, which can be deposited on the targeted substrate using spin casting, drop-casting, or inject printing. The main advantage of this method is that it can grow large areas of organic materials at a low cost. This method is mostly used to functionalize organic−inorganic heterostructure, such as doping to alter the carrier concentration of underlying inorganic material and p-n homojunction. However, this method has a limitation as it requires strict surface energy with underlying materials to achieve well-wetting behavior.^135–139

Another approach is to deposit solution processible organic material on CVD grown inorganic material.^140–143 Ultra-flat and dangling bond-free surfaces with weak intermolecular forces allow solution processible organic materials to self-assemble over inorganic material. The above mentioned three methods for organic−inorganic heterostructures have advantages and disadvantages. The mechanical exfoliation method is most suitable for constructing a sharp and clean interface that can be used to understand interfacial studies. However, the lack of scalability and bulk layer growth of organic materials limits its use at the commercial level. On the other hand, vapor phase growth is useful for the controlled synthesis of organic material at a low cost and scalable level. However, the crystalline quality of the interface still needs to be improved. Solution processing is a useful method to grow organic material on the desired substrate with a large area and to functionalize inorganic material. However, the interface quality is not so clean which still requires further studies.

APPLICATIONS BASED ON ORGANIC–INORGANIC HETEROSTRUCTURES Field-effect transistor

A FET is a fundamental component of various electronic devices.^144,145 Over the last few decades, many efforts have been made to follow the fundamental Moore's law, that is, the increment in FET density will double every 2 years on integrated chips. In this regard, the conventionally Si-based FET is reaching its physical limit for further miniaturization. The performance of the FET device begins to fail due to a sharp reduction in the size of channel length as it reaches up to several nanometer regimes of 3 nm.¹⁴⁶ To address this issue, 2D materials have shown better performance for next-generation FET by overcoming the short channel effects offered at the nanometer scale because the ultra-flat, dangling bond-free surface with an atomic thickness enables efficient electrostatic control. However, the performance of 2D materials-based FET still needs to compete with silicon-based electronic devices.

Organic−inorganic hybrid structures have displayed remarkable progress in the last few years, especially in electronics. The hybrid structures-based electronic devices result in more advanced multifunctional properties than their individual materials-based devices. The ultra-flat 2D surface and in-plane lattices act as a template for the epitaxial growth of organic semiconductors on 2D materials, which is crucial for OFET device performance. Although 2D inorganic-based heterostructures exhibit extraordinary performance, the limited number of available p-type materials limits the choice of heterostructures. Alternatively, organic materials provide many choices for organic−inorganic-based heterostructures due to the sheer number of available n-type inorganic and p-type organic materials. Hence, by taking advantage of both materials, 2D organic−inorganic heterostructures have created new avenues for the fabrication of multifunctional devices.

2D materials as a template for organic semiconductors

OFET has shown remarkable progress in the last few decades due to its potential applications for drive circuits, display panels, radio frequency identification, biosensors, photodetectors, and many other fields.^147–149 One of the vital parameters to optimize the performance of OFET is its carrier mobility which is closely dependent on the degree of ordering of organic material, i.e., how well organic semiconductors are organized on the targeted substrate. The disordering of organic channel material creates scattering and reduction in charge carrier mobility which affects the performance of OFET.¹⁵⁰ For the growth of highly ordered organic semiconductors, the intermolecular forces among organic molecules and organic/substrate play an important role.

It is hard to get highly ordered organic semiconductors on an inert substrate such as SiO₂ without any template because weak intermolecular forces dominate organic molecules, which reduces ordering due to random nucleation and further degrades the performance of OFET. To tackle this problem, it is recommended to use a template for organic semiconductors growth, which supports the growth of highly ordered organic semiconductors for high-performance OFET. In this regard, the layered molecular template provides the solution for the high-order growth of organic material, which may boost the OFET performance to a promising level.³⁵ Nevertheless, the limited number of accessible molecular templates and their small domain size hinder the success of molecular template-based OFET. Alternatively, 2D materials with their in-plane lattices act as a template for organic semiconductors. The weak vdW forces among organic and inorganic interfaces help the crystalline growth of organic semiconductors, enabling them to reduce trapping density and scattering with the enhancement of OFET performance.⁸⁸

Among the various 2D materials, graphene and h-BN are the most promising materials used as a template for the crystalline growth of organic semiconductors.¹⁴⁵ Wang et al.¹⁵¹ investigated the controllable synthesis of pentacene using the PVT method on already mechanically exfoliated h-BN. The ultra-flat 2D nature of h-BN, absence of dangling bond-defect-free surface, and wide bandgap around 6 eV are prominent features that make h-BN an efficient template for the single crystalline growth of pentacene which is essential for the study of intrinsic charge transport properties of layer dependent pentacene. For structural analysis, different characterizations were performed, such as high-resolution atomic force microscopy (AFM) confirming the crystallinity of 1L pentacene, as shown in Figure 4A. To further confirm the structural analysis, transmission electron microscopy (TEM) and selected area electron diffraction (SAED) was performed, confirming the high-quality crystalline growth of pentacene layers on the h-BN template layer (Figure 4B). To evaluate the charge transport properties among different layers of pentacene, the back gate OFET was fabricated. The device with 1L pentacene exhibited p-type characteristics with an on/off ratio of about 10⁸, field-effect mobility around 1.7 cm²/(V·s) in the linear region, nearly zero threshold voltage V_th, and small threshold swing (SS) of 450 mV/dec, and negligible hysteresis. The temperature-dependent electrical measurement revealed that the charge transport properties were dominated by hooping effect in 1L pentacene (Figure 4C), but for 2L and 3L bandlike transformations were observed due to changes in the molecular packing because of interfacial vdW interactions.¹⁵¹

Figure 4. (A) High-resolution AFM image of 1L pentacene, the scale bar is 1 nm. (B) SAED pattern of few-layer pentacene, where the blue circle is marked for BN (100) directions and green circles for pentacene (110), (120), and (020) directions. (C) Ids/Vgs characteristics at different temperatures for the same device. Images (A)–(C) reproduced with permission: Copyright 2016, American Physical Society.151 (D) Schematic illustration for the contact between C8-BTBT and Au. (E) Cross-sectional image of TEM for Au/C8-BTBT interface. (F) Extrinsic and intrinsic mobility as a function of temperature. Images (D)–(F) reproduced with permission: Copyright 2017, American Association for the Advancement of Science.152 (G) Schematic of the hybrid organic−inorganic template. (H) HRTEM images and the corresponding SAED patterns to conform the epitaxial growth of C60 over a hybrid template. (I) Transfer curves of graphene–FET with and without ODTS-SAM. Images (G)–(I) reproduced with permission: Copyright 2020, Wiley.153 AFM, atomic force microscopy; FET, field-effect transistor; SAED, selected area electron diffraction; TEM, transmission electron microscopy

Conventional organic thin-film transistors are highly demanded low-cost organic electronics. However, they suffer from high contact resistance due to interfacial traps and defects, which degrade their device performance. Alternatively, 2D organic semiconductors exhibit low contact resistance due to direct nondestructive contact between the 2D organic transport layer and metal, crucial for fabricating high-performance OFET with high carrier mobility. One such example is C8-BTBT which was grown with controlled layered using a vacuumed processing method on exfoliated h-BN by Wang et al.¹⁵² Further, they studied the monolayer charge transport and contact resistance properties of C8-BTBT. The schematic illustration of contact between Au and C8-BTBT is shown in Figure 4D. The interface study between C8-BTBT/h-BN was performed using TEM, and the results are shown in Figure 4E showing a clean interface with a slight roughness. Further, the OFET based on monolayer C8-BTBT showed remarkable intrinsic mobility around (20–30 cm²/(V·s)) at room temperate due to reduced contact resistance around (100–400 Ω/cm²). On the other hand, few-layered C8-BTBT OFET exhibited mobility around 10 cm²/(V·s). The low mobility was due to high Schottky contact and higher contact resistance than that of monolayer C8-BTBT (Figure 4F).¹⁵² Similarly, in another work, Lee et al.¹⁵⁴ demonstrated another example of the highly crystalline growth of C₆₀ on h-BN. Using different characterizations such as SAED, grazing incident X-ray diffraction, and DFT calculation, studied the crystalline structure of C₆₀ on h-BN, which exhibited high-quality crystalline growth of C₆₀ on the h-BN template. Further, to evaluate the electrical properties of C₆₀ on h-BN, FET was fabricated, which showed average charge carrier mobility around 1.7 cm²/(V·s), which was 40 times higher as compared to FET based on C₆₀ grown on SiO₂ substrate.¹⁵⁴

More recently, He et al.¹⁵³ developed a new type of organic−inorganic hybrid template resulting in the high-quality growth of organic semiconductor thin films. Instead of using graphene as a primary template, they transferred the graphene over ODTS-SAM and then deposited the C₆₀ molecule layer over this graphene-ODTS-SAM hybrid template as shown in Figure 4G. Here, they found that the ODTS-SAM sandwiched between graphene and substrate, suppressed the corrugation of graphene along with the reduction in the electron-hole puddles in graphene. Further, the graphene template grown over ODTS-SAM facilitated high crystalline growth of C₆₀ molecules compared with the direct growth of C₆₀ over graphene. To confirm the crystalline growth of hybrid OSC, HRTEM was performed as shown in Figure 4H. The obtained results confirmed the epitaxial growth of graphene over ODTS-SAM, which facilitated further the crystalline growth of C₆₀ over the graphene template. In addition to examining the charge carrier properties, the as-fabricated FET as shown in Figure 4I revealed that the dry transfer of graphene over ODTS-SAM enhanced the mobility of graphene by 3.3 order, which was also greater than the mobility of graphene deposited over the h-BN template.¹⁵³ Further, other groups have also tried to grow high-quality films of small molecules such as a C₆₀ on graphene, but the resulting devices displayed a low device performance such as a low on/off ratio.¹⁵⁵ However the organic−inorganic hybrid template method has created new opportunities to grow high crystalline growth of organic molecules which can enhance the performance of optoelectronic devices.

Organic−inorganic based p-n junctions

The p-n junction is one of the oldest and major constituents of various electronic and optoelectronic devices such as FET, solar cells, photodetectors, and LEDs.^156–158 To construct the p-n junction, p-type materials having holes as majority charge carriers are combined with n-type materials having electrons as majority charge carriers. The resulting p-n junction creates a built-in potential at the interface that can rectify current and separate the photogenerated charge carriers (electrons-holes), essential for high-performance electronic and optoelectronic devices. However, 2D materials-based p-n junctions have been created successfully for advanced multifunctional flexible devices, but the strict conditions such as lattices match and lattices constant limit the selection of materials.

Alternatively, organic materials with a large number of availabilities with desirable properties can easily be combined with inorganic materials due to ultra-flat dangling bond-free surfaces of inorganic materials and weak intermolecular interactions among organic and inorganic materials without any constrained lattices match. Organic−inorganic-based heterostructures exhibit advanced properties compared to inorganic-based ones, e.g., high optical absorbance coefficients play an important role, such as ultra-fast charge dissociation at organic−inorganic heterointerface, which is essential for many photovoltaic devices. Similarly, organic−inorganic p-n junction shows considerable recombination time of separated charge carriers at the interface, making them highly desirable for high-performance flexible optoelectronic devices such as solar cells, photodetectors, and photodiodes. Further, the organic−inorganic-based heterostructures exhibit a charge recombination time 20–60 times longer than TMDCs-based heterostructures.

Wang et al.¹⁵⁹ studied the p-type C8-BTBT and n-type MoS₂-based heterostructures C8-BTBT/MoS₂ to study their electrical transport and photovoltaic properties. The p-type C8-BTBT was epitaxially grown on n-type MoS₂ with a controllable thickness (Figure 5A). The device showed diode-like behavior with a rectifying ratio of about 10⁵ at room temperature (Figure 5B). Furthermore, the device exhibited photoresponsivity up to 22 mA/W and photovoltaic responses (Figure 5C) with PCE of about 0.31%. All these findings suggested that organic−inorganic heterostructures can be used for optoelectronic devices.¹⁵⁹ Liu et al.¹⁶⁰ successfully fabricated ambipolar FET based on p-type rubrene and n-type MoS₂ (rubrene/MoS₂) (Figure 5D). The ambipolar FET exhibited balanced hole and electron mobility of 0.36 and 1.27 cm²/(V·s), respectively (Figure 5E). The complementary metal-oxide semiconductor (CMOS) inverter was also designed based on rubrene/MoS₂, which exhibited a gain of 2.3 at a switching threshold voltage of −26 V (Figure 5F).¹⁶⁰

Figure 5. (A) Microscope image of MoS2/C8-BTBT heterojunction and illustration for heterojunction device structure. (B) Room temperature Jds-Vg characteristics of MoS2/C8-BTBT heterojunction. (C) Ids/Vds characteristics of heterojunction MoS2/C8-BTBT under white light. Images (A)–(C) reproduced with permission: Copyright 2015, Applied Physics Letters.159 (D) Schematic illustration of ambipolar FET based on MoS2/rubrene heterostructure. (E) Transfer curve of ambipolar MoS2/rubrene FET. (F) Voltage gain of CMOS inverter in different Vdd conditions. Images (D)–(F) reproduced with permission: Copyright 2016, Wiley.160 (G) Device configuration of pentacene-MoS2 based p-n junction. (H) The transconductance of the device under VD = 10 V. (I) Transfer curves of the device. Images (G)–(I) reproduced with permission: Copyright 2017, Royal Society of Chemistry.161 FET, field-effect transistor

In another work, Jiang et al.¹⁶¹ designed another organic−inorganic p-n vdWH based on single crystalline pentacene and MoS₂. A single crystalline pentacene was grown using the PVT method on a few layers of mechanically exfoliated MoS₂ (Figure 5G). The pentacene/MoS₂ p-n junction exhibited large current density and anti-ambipolar characteristics with the highest trans-conductance of 211 nS (Figure 5H). Further electrical characteristics were also demonstrated based on configuration dependence, and the results show the presence of a space charge zone at vertical p-n heterojunction (Figure 5I).¹⁶¹ Similarly, other research groups used different organic materials like ZnPc, CuPc, and C₆₀ to form organic−inorganic p-n junctions. Organic−inorganic p-n vdWHs are also employed for high-performance solar cells having large open-circuit voltage V_oc even higher than 2D–2D inorganic-based heterostructures.

For example, pentacene/MoS₂ p-n vdWHs showed photovoltaic effects under optical illumination with V_oc up to 0.3 V and short circuit current I_sc 3 nA showing potential applications of organic−inorganic p-n junctions for solar cells.³⁷ These results show that the well-defined interface between 2D organic−inorganic heterostructures has opened new innovative opportunities to fabricate more advanced electronic and optoelectronic devices.

Organic doping for the construction of p-n homojunction

Apart from combining p-type organic material and n-type inorganic material for p-n heterojunction, chemical doping of 2D materials using organic materials offers an alternative route for homojunction. Organic molecules can tune the electronic properties of inorganic materials by modulating the carrier concentration of 2D material surface through physisorbed or chemisorbed, which helps fabricating lateral p-n homojunctions.^162–166 Using this strategy, homojunction of graphene, TMDCs, and BP is successfully fabricated.^167–170

Li et al.¹⁷¹ explored the vertical p-n homojunction in which organic dopant was employed to tune the carrier concentration. To investigate the electronic properties, the back gate FET was fabricated in which mechanically exfoliated MoS₂ with a thickness of around 11 nm was used as a channel layer. To make p-n vertical homojunction, the underlying bottom surface of MoS₂ was n-doped by using benzyl viologen, while the top surface of MoS₂ was p-doped by employing AuCl₃, as shown in Figure 6A. The as-fabricated p-n homojunction-based FET showed rectifying behavior with a current rectifying ratio of around 100, as shown in Figure 6B. Ambipolar carrier transport characteristics were also observed with a hysteresis window of 60 V due to the n-type and p-type doping of MoS_2, as shown in Figure 6C.¹⁷¹ Recently, Hiroki et al.¹⁷² reported chemical doping of WSe₂ to convert its ambipolar to p-type and n-type for low power consumption CMOS inverters (Figure 6D,E). By optimizing the concentration of p-type 4-NBD and n-type diethylenetriamine (DETA) the carrier mobility improved four times compared to pristine WSe₂. With this chemically doped WSe_2, CMOS inverters were made with a low power consumption of around 170 pW with V_gain of 10 V (Figure 6F).¹⁷² Homojunction formation through chemical doping is a low-cost, simple and effective strategy to modulate the charge carrier concentration of 2D materials by forming p-n junctions for flexible electronic and optoelectronic applications.

Figure 6. (A) The schematic diagram illustrates the fabrication of vertical p-n homogeneous homojunction of few-layer MoS2 flake by organic molecule doping. (B) Output characteristics (ID−VD) at various VG from −60 to 60 V. (C) The transfer characteristics and photoresponse of the device with and without laser illumination during forward and reverse sweeps. Images (A)–(C) reproduced with permission: Copyright 2015, Springer Nature.170 (D, E) Chemical doping of WSe2 using 4-NBD and DETA molecules for n-type and p-type doping. (F) CMOS inverter based on p-type and n-type doped WSe2. Images (D)–(F) reproduced with permission: Copyright 2019, Wiley172

Photodetectors

Photodetectors are optoelectronic devices' most pivotal and significant family members that convert optical light into electrical signals.^90,173 Nowadays, photodetectors are used in many advanced applications such as infrared thermal imaging, environmental monitoring, ray measurement, infrared remote sensing, and military.^174–177 The primary working mechanism of photodetectors is when the incident light has photon energy more significant than the underlying semiconductor bandgap, it creates electrons and holes, and these separated carriers are collected at electrodes which create an electric current. Previously conventional silicon-based photodetectors were used, which shows a fundamental limitation for NIR to IR detection due to its indirect bandgap of 1.1 eV. In addition, they cannot be used as flexible and wearable photodetectors.¹⁷⁸ Later, InGaAs-based photodetectors displayed photodetection in the IR region due to their low bandgap. However, they also exhibited some drawbacks, e.g., high cost and unsuitable for flexible photodetectors.

2D materials show tremendous progress in addressing the fundamental limitations of conventional bulk materials due to their strong light-matter interaction and strong in-plane interaction bonding, thickness-dependent bandgap and indirect to direct and high carrier mobility. Graphene proved a strong candidate for high-speed and broadband photodetectors due to its high carrier mobility 20,000 cm²/(V·s) and comprehensive photon detection. However, due to its zero bandgap and low light absorption of around 2.3%, it is unsuitable for photodetectors with high responsivity; further, the large on/off ratio limits its efficiency. Later in the discovery of new layered materials, TMDCs got their attention due to the wide range of suitable bandgap and high carrier mobility, making them an appropriate choice for photodetector applications. Apart from the tremendous qualities of TMDCs, they show limitations like wide-bandgap unsuitable for NIR and IR photodetection, defect-free synthesis, and the limited number of p-type TMDCs that limit the formation of p-n heterostructures. The discovery of low bandgap BP addressed the limitation of TMDCs because of their IR photodetection, but BP is not air-stable.

On the other hand, organic materials are potential candidates for photodetectors applications due to their strong light absorption coefficients. Nevertheless, the device performance of organic photodetectors is limited due to the low carrier mobility of organic materials. By comparing the properties of organic and inorganic materials, we can conclude that by combining organic materials with inorganic materials, we can accomplish photodetector with superior performance, which cannot be achieved with individual materials, e.g., organic materials can be used as light absorption material and inorganic materials as high carrier mobility material.

Combining organic and inorganic materials results in advantages in making high-performance photodetectors by modulating spectral range from visible to NIR with high responsivity, low dark current, and high gain. Photodetectors are classified further as phototransistors, photoconductors, photodiodes, and bolometers (temperature dependent). The significant merits of photodetectors are photoresponsivity, response time, EQE, photocurrent/dark current ratio I_p/I_d, and detectivity. These key parameters are charge carrier generation, dissociation of charge carriers, collection of carriers, and transport. All the parameters mentioned above can be accomplished by making organic−inorganic heterostructures, e.g., significant light absorption coefficients of organics can help in charge carrier generation while inorganics provide high charge mobility, long carrier diffusion length, and stability. We will discuss recent advancements in light-sensitive photodetectors. We will further present the progress of organic−inorganic-based photodetectors as organic-graphene, organic-TMDCs, organic-topological insulators (TIs) and organic-perovskites based photodetectors.

Graphene-organic based photodetectors

Graphene has proved itself a strong candidate for advanced optoelectronic devices due to its high carrier mobility, flexibility, and wide range of optical light absorption. Despite all these extraordinary properties, the lack of intrinsic bandgap limits its application for photodetectors, such as low light absorption, large dark current, and high off current.¹⁷⁹ Many efforts, such as doping, quantum dots, and nanoribbons, are made to create bandgap and modulate electronic properties. Among all these efforts, combining graphene with organic semiconductors is an efficient, easy, and low-cost strategy to improve the photodetection of graphene-based photodetectors.^180–183 Considerable progress has been made in the last few years by combining organic materials with graphene to improve the graphene-based phototransistors. Organic semiconductors serve as a light-absorbing layer, while the ultra-flat 2D surface of raphene offers an ideal platform to grow organic semiconductors over graphene, which is critical to optimize phototransistor performance.^184,185 Russo et al.¹⁸⁶ examined the single crystalline rubrene deposited using PVT on graphene for phototransistors (Figure 7A). They found that the long-range order of single-crystalline rubrene over graphene facilitated the charge creation and transfer from rubrene to graphene. Graphene and rubrene interface helped achieving photodetection for visible range with responsivity 10⁷ A/W for sub-femto watt incident signals, which is comparable to graphene-quantum dots based phototransistors with relatively fast rise time of 100 ms and slow fall time (Figure 7B).¹⁸⁶ Wang et al.¹⁸⁷ investigated the thickness-dependent epitaxial growth of C8-BTBT over CVD-grown graphene to optimize the C8-BTBT-graphene heterostructure-based phototransistor (Figure 7C). It was further explored that performance of phototransistors was extraordinarily improved even for the monolayer C8-BTBT on graphene. The device showed superior performance, such as monolayer C8-BTBT on graphene exhibited high responsivity about 10⁴ A/W with a response time of 25 ms and photoconductive gain of around 10⁸. It was also revealed that the thicker C8-BTBT layers showed high responsivity and EQE but compromised bandwidth due to slow charge transport in C8-BTBT thicker layers.¹⁸⁷

Figure 7. (A, B) Rubrene-graphene-based heterostructure with high detectivity and fast response. Images (A) and (B) reproduced with permission: Copyright 2017, Wiley.186 (C) Thickness-dependent photoresponse of C8-BTBT-graphene-based photodetector. Image (C) reproduced with permission: Copyright 2016, Wiley.187 (D, E) Schematic illustration of phototransistor based on pentacene-PTCDA-graphene with photocurrent measurement at 550 nm. Images (D) and (E) reproduced with permission: Copyright 2017, American Chemical Society.188 (F, G) Schematic illustration of C60-pentacene- graphene-based photodetector with the bi-directional response. Images (F) and (G) reproduced with permission: Copyright 2018, Wiley.188 (H, I) Schematic illustration for TTF-CA/graphene-based photometer for SWIR detection and high responsivity. Images (H) and (I) reproduced with permission: Copyright 2020, Wiley.189 (J, K) Rhodamine- graphene-based photodetector having wide-bandgap photodetection. Images (J) and (K) reproduced with permission: Copyright 2016, Wiley190

Similarly, in another report Wang et al.¹⁹¹ studied graphene as a template for the growth of C₆₀ for phototransistors. Thermally evaporated C₆₀ was deposited using the thermal evaporation method over CVD-grown graphene. Well-aligned C₆₀ over graphene resulted in charge dissociation at interface and UV light absorption, which helped achieve high gain and photoresponsivity around 10⁷ A/W; the response time was not evident due to charge recombination at the heterointerface.¹⁹¹ To further improve the device performance of photodetectors such as EQE and bandwidth, Wang and his team used PTCDA and pentacene heterostructure grown over graphene, which facilitated fast charge dissociation at the interface and low charge recombination due to the built-in electric field at the interface of PTCDA-pentacene (Figure 7D). With this device structure, the achieved responsivity and response time was 1 to 2 order magnitude higher than single organic layers on graphene for phototransistors with wideband photodetection (400–700 nm) and the highest response time gate tunable around 30 µs for graphene-based phototransistors (Figure 7E).¹⁸⁸

One of the key limitations for graphene-organic based phototransistors is the spectral detectivity in a narrow range, i.e., in the visible range, to further extend the spectral responsivity. In another report, Wang et al.¹⁸⁸ designed the C₆₀/PTCDA heterostructure on graphene (Figure 7F). This device's architecture achieved broadband photodetection (405–1550 nm) with a high gain of 5.2 × 10⁵ and a fast response time of 275 µs. The photoresponsivity for the visible range was about 9127 A/W at 650 nm and for NIR 1800 A/W at 808 nm. The unique band alignment for this device structure resulted in bidirectional photoresponse for the first time due to opposite charge transfer of photoexcited charge carriers at a different wavelength (Figure 7G). Graphene/C₆₀/PTCDA-based phototransistor has created new opportunities for functional large-scale optical imaging and infrared focal plane array detection.¹⁸⁸ Furthermore, Xie et al.¹⁸⁹ explored tetrathiafulvalene–chloranil (TTF–CA) complex and graphene-based hybrid SWIR photodetector. TTF-CA complex serves as SWIR absorbing layer because of charge transfer electronic transition at 0.5 eV, while graphene was used as a high carrier mobility channel (Figure 7H). The photodetector achieved ultra-high photoresponsivity up to 10⁵ A/W (Figure 7I), high specific detectivity of about 10¹³ Jones, and a fast response time of 2–12 ms at 0.55–2.25 µm wavelength. Overall, they found that organic complex for SWIR light absorption is a good choice for organic-graphene based photodetector for SWIR detection also because of its low cost, scalability and room-temperature operation.¹⁸⁹

Lu et al.¹⁹² recently demonstrated the graphene-COF (2D covalent organic framework) for photodetector applications. 2D COF is a next-generation layered organic material with excellent optoelectronic properties due to π-electronic skeletons and high-ordered topological structures, which can also tune optoelectronic properties by selecting specific monomers. Graphene-COF-based heterostructure was successfully fabricated using a low-cost and straightforward in situ growth method. COF_ETBC-TAPT/graphene-based heterostructure photodetectors showed excellent photoresponsivity about 3.3 × 10⁷ A/W at a wavelength of 437 nm with a fast response time of 1.14 ms. They also explored that the high surface area and polarity selectivity of COF enable us to achieve specific photo-sensing using targeted molecules. Some strategies were also used to improve the graphene-organic-based photodetectors, such as using h-BN layers between graphene and organic layers, improved the charge carrier injections, and further the dielectric interface reduces the traps and scattering centers.¹⁹²

In another work, Rhodamine-based dye molecules were deposited on graphene using the simple dip-coting method with different color selectivities (Figure 7J). The organic dye on graphene resulted in spectral detectivity around 10¹⁰ Jones and responsivity about 10³ A/W for spectral color selectivity (Figure 7K).¹⁹⁰ The above mentioned results show that 2D organic-graphene-based photodetectors exhibits high performance compared with their individual material-based photodetector devices. Here the ultra-flat graphene provides more crystalline growth of organic material, and high optical light absorption enhances the charge carrier at the interface to improve the responsivity of photodetector.

TI-organic-based photodetectors

Due to their potential applications, photodetectors with a spectral range from NIR to IR are highly demanded.¹⁹³ To achieve NIR to IR spectral range, materials with zero bandgap or narrow bandgap are considered the most suitable choice.^194–197 Although graphene exhibits an intrinsic zero bandgap, it has a low absorption light of 2.3%, limiting its device performance as a photodetector, such as low photo carrier accumulation and charge carrier separation, which results in low absorption light responsivity.^198–200 On the other hand, TIs are emerging materials owing to their narrow bandgap, high carrier mobility, and high optical absorption since they have unique Dirac-like stable surfaces.^201–203 Similar to graphene, TIs possess high electron mobility, limiting their device performance for infrared photodetectors due to large dark current.

To overcome this issue, Yang et al.²⁰⁴ demonstrated n-type Bi₂Te₃ and p-type pentacene heterostructure for broadband photodetection (450–3500 nm) (Figure 8A). From the energy band diagram as shown in Figure 8B, IT can be seen that pentacene serves as a visible light absorption layer (400–700 nm) due to a wide bandgap around 1.9 eV, while for photon energy less than 1.9 eV Bi₂Te₃ serves as a mid-infrared photodetecting material for larger wavelength around 3500 nm. The device showed responsivity for the visible region around 14.89 and 1.55 A/W at a wavelength of 3500 nm, which is several order magnitudes compared to other 3D TIs-based photodetectors (Figure 8C). The response time for the Bi₂Te₃/pentacene-based photodetector was about 1.89 ms which was one order magnitude compared to Bi₂Te₃-based photodetectors.²⁰⁴

Figure 8. (A) Schematic illustration of pentacene/Bi2Te3 heterostructure-based Photodetector. (B) Energy band alignment illustration between Bi2Te3 and pentacene. (C) Photocurrent measurement of the responsivity of the device. Reproduced with permission: Copyright 2019, American Chemical Society204

Similarly, in another work, Zhang et al.²⁰⁵ investigated the comparison analysis of the photoelectric performance of Cd₃As₂ and Cd₃As₂-CuPc heterostructure based photodetectors. The Cd₃As₂-CuPc heterostructure exhibited high performance compared to individual Cd₃As_2-based photodetectors due to the built-in potential created at the interface of Cd₃As₂-CuPc, which resulted in fast charge dissociation at the interface. The as-fabricated device showed broadband photodetection from visible (450 nm) to NIR (980 nm) with high 142.5 A/W as compared to Cd₃As₂-based photometers (R = 1.46 A/W) with high detectivity D* up to 7.83 × 10¹⁰ Jones. The result suggests that high-performance photodetectors can be fabricated by employing narrow bandgap and high mobility 3D TIs with organics. Further research is needed to improve the 3D TIs-organic-based heterostructure for ultra-fast response time with high responsivity and detectivity by combining other organic materials with 3D TIs.²⁰⁵ These results show that 2D organic-topological insulator-based heterostructure has opened a window to explore new NIR to IR photodetectors that are highly desirable for many advanced applications.

TMDC-organic based photodetectors

Layered TMDCs have shown tremendous progress in optoelectronic devices due to their unique electronic, optoelectronic, and tunable bandgap depending on the thickness of TMDC layers, direct and indirect bandgap transitions, and a variety of available bandgaps materials covering the broadband spectrum.^67,206 Regardless of showing extraordinary properties, TMDCs possess key limitations for photodetectors. For example, wideband gaps limit their photodetection from UV to visible region, trap states that affect photodetector's response speed, and defects during synthesis may degrade device performance.^207,208 MoS₂ is the most studied TMDCs due to its high carrier mobility, air stability, tunability of bandgap, and 1.8 eV direct bandgap for monolayer. MoS₂-based photodetectors also show spectral limitations due to wide bandgap and UV-Visible detection. To overcome above mentioned limitations, many strategies have been applied, among them combing organic with MoS₂ has shown remarkable progress in enhancing the spectral range of MoS₂ and other critical parameters for photodetector application.

Lee et al.²⁰⁹ studied the thickness dependence performance of CuPc layers, thermally deposited on MoS₂ for photodetector. The charge transfer phenomena at CuPc/MoS₂ interface enable the threshold voltage of MoS₂ FET by controlling the thickness of CuPc on MoS₂. CuPc/MoS₂ hybrid devices improved performance as photodetectors than their MoS₂ or CuPc-based devices because CuPc facilitated charge separation at the p-n interface. They also investigated that the CuPc layer about 2 nm on MoS₂ shows the best performance as a photodetector with a high responsivity of 1.98 A/W, detectivity of 6.11 × 10¹⁰ Jones and EQE about 12.57%.²⁰⁹

Similarly, Zhai et al.²¹⁰ further improved the photodetector performance of MoS₂ by immersing MoS₂ in the ZnPc solution (Figure 9A). The ZnPc assembled easily on MoS₂ and acted as electron doping by withdrawing electrons from MoS_2, which assisted in stopping electron-hole recombination upon light illumination and driving holes into ZnPc and resulted in a fast response time <8 ms, which is more than 3 orders as compared to bared MoS₂ (20–40 s). Also, with Al₂O₃ encapsulation on the device, the responsivity reached up to 430 A/W (Figure 9B).²¹⁰ More recently, Wei and coworkers¹³⁴ demonstrated CuPc/MoS₂ based heterostructure to improve MoS₂-based photodetectors by combining CuPc using the thermal vaporization method (Figure 9C). CuPc layers over MoS₂ facilitated ultra-fast charge transfer at an interface of about 16 ps resulting in fast charge dissociation for achieving photoresponse time 436 µs which is four orders of magnitude compared to pristine monolayer MoS₂ based photodetectors. CuPc/MoS₂ based heterostructure showed responsivity about 3.0 × 10³ A/W detectivity about 2.0 × 10¹⁰ Jones (Figure 9D) and EQE 483%.¹³⁴

Figure 9. (A, B) Schematic illustration of ZnPc-MoS2 based heterostructure for photodetector with high detectivity and responsivity. Images (A) and (B) reproduced with permission: Copyright 2018, American Chemical Society.210 (C, D) Device architecture illustration for CuPc-MoS2 based photodetector with enhanced responsivity and detectivity. Images (C) and (D) reproduced with permission: Copyright 2020, Elsevier.134 (E, F) Schematic illustration of PDPP3T-MoS2 basedsssss flexible photodetector with fast response time. Images (E) and (F) reproduced with permission: Copyright 2018, Wiley.40 (G, H) Schematic illustration of p-MSB-WSe2 based high-performance photodetector with enhanced photoelectric properties. Images (G) and (H) reproduced with permission: Copyright 2018, Springer Nature.211 (I, J) Rhodamine-MoS2 based photodetector exhibiting high spectral range compared to MoS2 based photodetector. Images (I) and (J) reproduced with permission: Copyright 2014, American Chemical Society.39 (K, L) Device structure illustration of P(VDF-TrFE)-MoS2 heterostructure resulted in self-powered behavior even at zero bias voltage. Images (K) and (L) reproduced with permission: Copyright 2015, Wiley212

Although previously reported photodetectors based on MoS₂ have shown high responsivity and photoresponse speed, its spectral detectivity is limited to the UV-Visible range due to its wide bandgap. To overcome this limitation, Zhang et al.⁴⁰ extended the spectral range of MoS₂ from UV-Visible to NIR range by using a low-cost and straightforward spin coating method by depositing PDPP3T over MoS₂ (Figure 9E). In PDPP3T/MoS₂ heterostructure-based photodetector, PDPP3T acts as a photon absorbing layer in the NIR spectrum while MoS₂ acts as UV-Visible detecting material. The device exhibited a fast response time of 4 ms and decay time of 40 ms with a broadband photoresponse from UV-Visible to NIR range (Figure 9F). The asymmetric electrode and PDPP3T/MoS₂ heterostructure were also fabricated on PET substrate for wearable photodetectors.⁴⁰ Wei et al.²¹¹ demonstrated epitaxial growth of p-MSB crystal over WSe₂ for high-performance photodetectors (Figure 9G). The p-MSB/WSe₂ interface showed electric-gate tenability, which helps in the interfacial charge trapping process by gate voltage V_g. In a gate voltage of −60 V, the device exhibited responsivity about 3.6 × 10⁶ A/W and detectivity of 8.6 × 10¹⁴ Jones, increased by 25-fold and 3-fold, which was the highest among all TMDCs based heterostructures (Figure 9H).²¹¹

Some other strategies were also employed to improve the performance of the TMDCs-based photodetector, e.g., Alshareef and his team members²¹³ reported a low-cost, simple solution processible method for fabricating g-C₃N₄/MoS₂ hybrid thin-film based photodetectors. The device displayed high performance for UV and visible light detection with 5:5 hybrid films. Further, the device exhibited high responsivity, detectivity, and photoswitching properties with mechanical flexibility and environmental stability.²¹³ Similarly, in another work, Cho et al.³⁹ demonstrated the use of low-cost drop-casting of 6G(R6G) organic dye on pristine MoS₂ to increase the photoresponsivity of pristine MoS₂ (Figure 9I). Upon light illumination, photoinduced charge carrier transfer from the organic dye to MoS₂ resulted in high-performance photodetectors with responsivity 1.17 A/W, detectivity 1.5 × 10⁷ Jones (Figure 9J), EQE 280% at a wavelength of 520 nm, and broadband spectral range (405–980 nm).³⁹ In another work, Wang et al.²¹² explored the use of P(VDF-TrFE) copolymer as a dielectric for MoS₂ transistor to improve the performance of the MoS₂ photodetector (Figure 9K). The P(VDF-TrFE) polarization helps producing a local electrostatic field of about 10⁹ V/m; this significant electrostatic effect creates a depleted state in MoS₂ channel layers, achieving high photosensitivity and reduced dark current even at zero voltage. The photodetector exhibited high photoresponsivity about 2570 A/W and detectivity around 2.2 × 10¹² Jones with wideband spectral range from visible to NIR (0.85–1.55 µm) (Figure 9L).²¹² Above mention results show that a tradeoff between the speed, responsivity and bandwidth can be overcome by taking advantage of large charge carrier generation due to organic materials along with excellent charge dissociation at 2D organic−inorganic heterojunction.

Perovskites-organic based photodetectors

In the last few years, lots of research has been done on new emerging organometallic halide perovskites (OHPs) materials for advanced optoelectronic applications due to their unique properties such as large light absorption coefficient, low cost, simple fabrication methods, high mobility and long carrier diffusion length.^{46,47,214–217} Despite having all extraordinary properties, perovskites-based photodetectors show fundamental limitations, such as the wide bandgap limiting their spectral range to UV-Visible light and device degradation due to chemical reaction between H₂O and Iodine of perovskites.^218,219 Narrow bandgap organic semiconductors help overcome these limitations by combing them with OHPs. Shi et al.²²⁰ demonstrated photodetectors by combining bilayer MAPbI₃ with light-sensitive narrow bandgap organic PDPP3T. By employing PDPP3T with MAPbI₃, the spectral range increased from UV-Visible to NIR. The photodetector also exhibited strong photoresponse as compared to individual MAPbI₃-based photodetectors. The results confirmed that the spectral range could be extended from UV-Visible to NIR by combining narrow bandgap organic material with perovskites.²²⁰

Similarly, in another work, Peng et al.²²¹ investigated the lateral photoconductors of CH₃NH₃PbI_3-1/PbPc (Figure 10A). The as-fabricated devices showed high photoresponsivity 10⁴ A/W at low intensity of light 10 mW/cm², spectral response in NIR, with ultra-low dark current in pA (Figure 10B,C). The obtained results provided low cost, simple processing, flexible, and highly sensitive UV-Visible to NIR.²²¹ The above mentioned literature results suggest new innovative ideas for extending the photodetection performance of perovskite materials by combining them with organic materials using a simple and low-cost method.

Figure 10. (A) Schematic device illustration CH3NH3PbI3-1/PbPc based heterostructure for NIR photodetectors. (B) Responsivity under NIR illumination. (C) Ultra-low dark current. Reproduced with permission: Copyright 2017, Springer Nature221

Solar cells

Considering the energy crisis and environmental concerns, lots of efforts are carried out to find environmentally friendly, low cost and renewable sources of energy, among them solar cells, has fulfilled the desired requirement to meet the energy demands by following above mentioned parameters. The solar cell is one of the renewed members belonging to energy harvesting devices that convert optical energy into electrical energy using photovoltaic (PV) effects.²²² The working mechanism of solar cell is quite similar to that of photodetectors, except solar cell does not require any external bias as it works as an electric power source. The most critical parameters to optimize the performance of solar cells are high open-circuit voltage (V_oc), increased short circuit current density (J_sc), PCE and fill factor (FF) value. Different strategies are used to fabricate solar cell devices, such as silicon-based solar cells, perovskites solar cells (PSCs), dye synthesized solar cells, quantum dots solar cells, and many other approaches. Organic solar cells are considered the most efficient PV device due to their low cost, flexibility, easy processing, scalability, and many available organic semiconductors with tunable and desired properties.^223,224 Although remarkable progress is shown by OSCs such as PCE has reached more than 18%, a lot to improve OSCs device performance, e.g., environmental stability, development in device architecture, and scalability. Combining 2D materials with OSCs can overcome such challenges to enhance the device performance of OSCs by increasing their PCE.

2D material as a transparent electrode for OSCs

Indium titanium oxide (ITO) has been used as a high transparent electrode for solar cells because of its high conductivity and low sheet resistivity. Although it shows extraordinary properties as a transparent electrode, ITO displays several limitations such as its high cost, not being stable for high pH, and its brittleness. Alternatively, highly transparent graphene has strong mechanical strength, and low-cost synthesis processing can replace ITO as a transparent electrode for flexible organic solar cell applications. In 2008 for the first time, Peumans and co-workers²²⁵ demonstrated that solution processible graphene could be used as a transparent electrode for organic solar cells. Still, the device performance was low compared to ITO-based OSCs due to the high sheet resistivity of graphene compared to ITO.²²⁵ Later, many efforts were made to improve the OSC's device performance using graphene as a transparent electrode. More recently, a flexible OSC with 15% efficiency was fabricated using polyimide integrated with graphene (PI-Gr) as an electrode. Graphene was synthesized using the CVD method, with a large area of around 25 µm on Cu foil which exhibited high transparency and conductivity transferred on PI flexible conducting substrate (Figure 11A). The PI-Gr-based electrode exhibits high transparency around 92% with a low sheet resistance of about 83 Ω/sq (Figure 11B). The as-fabricated OSC with PM6:Y6 as an active channel layer and PI-Gr electrode exhibited high performance, such as PCE around 15.2%, J_sc 25.8 mA/cm², V_oc of 0.84 V, and FF 70%, which is the highest reported efficiency for Gr electrode-based OSC (Figure 11C,D). Further, the integration of PI with graphene resulted in the enhancement of device stability and mechanical flexibility.²²⁶

Figure 11. (A) Schematic illustration for the two-step fabrication process of PI-Gr as a flexible electrode. (B) Optical transmission of PI-GR compared to graphene and PI. (C) J-V characteristics and (D) EQE spectra with an integrated current density of OSC. Images (A)–(D) reproduced with permission: Copyright 2020, Elsevier.226 (E) Schematic illustration for the fabrication of MXene as a flexible electrode for OSC. (F) Optical transmission spectra of MXene/AgNWs-PU. (G) R/R0 of MXene/AgNWs-PU film as a function of bending and unbending cycles. (H) Normalized PCE of OSC with flexible MXene/AgNWs-PU electrode. Images (E)–(H) reproduced with permission: Copyright 2019, American Chemical Society227

Apart from graphene, researchers have also tried to utilize highly conductive MXene as a transparent electrode for OSCs. Chen et al.²²⁷ demonstrated about Ag/MXene network as a transparent, conducting, and flexible electrode for OSCs (Figure 11E). The hybrid Ag/MXene was prepared using a simple, low-cost solution processible method and showed excellent conductivity, mechanical stability, and transparency (Figure 11F). To optimize the elastic properties of OPV, Ag/MXene hybrid was used as a transparent flexible electrode for ternary structure PBDB-T:ITIC:PC₇₁BM. The results from the OPV device exhibited PCE up to 8.3% with excellent mechanical flexibility by keeping its original PCE value up to 86% after 1000 bending and unbending with a radius of 5 mm, which shows that in the future, MXene can be used as a transparent conducting electrode for flexible solar cell (Figure 11G,H).²²⁷

2D materials as hole transport layer (HTL) for OSCs

The HTL is the most essential component of OSCs to improve performance, such as PCE and device stability. HTL can be further divided into two categories, i.e., organic HTL, such as PEDOT:PSS, the most widely used conducting polymer for OSCs, and inorganic HTL. Although PEDOT:PSS is the most widely used HTL for OSCs, it offers some limitations, such as its high acidity, which creates ITO corrosion and is not stable for a more extended period. Metal oxides such as V₂O₅, MoO₃, and NiO have displayed considerable progress as efficient HTL, but their high-cost synthesis methods limit their further use at the commercial level. Alternatively, 2D materials with suitable bandgaps and work functions have presented remarkable progress in replacing both PEDOT:PSS and metallic oxide for HTL. They can be synthesized using inexpensive methods. Graphene derivatives such as graphene oxide (GO) are the potential candidates for an efficient HTL for OSCs due to their unique electronic structure, high stability, high optical transparency, mechanical flexibility and low-cost production at a larger scale. GO was used as an HTL for (poly(3-hexylthiophene) (P3HT): PCBM) based OSC, and the as-fabricated OSC exhibited an improved PCE of 3.5 ± 0.3%.²²⁸ Functionalized graphene has also shown remarkable progress as an HTL for OSCs, e.g., sulfated GO containing the functionalized group –OSO₃H was employed as an HTL which exhibited a high PCE of about 4.37% due to a well-matched work function of HTL with active layer organic material used for OSC.²²⁹ Similarly, GO as an HTL used for PTB7:PC₇₁BM organic blend improved the OSC performance by achieving efficiency of 7.4%.²³⁰ Apart from graphene and its derivatives, 2D TMDCs have been used as HTL for OSCs to improve the contact interface of the active organic layer and MoS₂ HTL and reduce the surface defects of MoS₂. It is reported that the UV-ozone treated MoS₂ HTL was employed along PEDOT:PSS for OSC to increase the device performance such as stability along with an increment in efficiency up to 2.81%.²³¹ Similarly, Yang's group²³² demonstrated the use of UV-ozone treated MoS₂ as an HTL for OSC. The as-fabricated device exhibited PCE up to 7.64%, which was 53% higher than MoS₂ HTL without UV-ozone treatment.²³² Hybrid MoS₂ and gold nanoparticle-based HTL used for OSC exhibited up to 7.25% PCE.²³³

2D material as an electron transport layer (ETL) for OSCs

The ETL is also considered one of the essential components to improve the device performance of OSCs such as PCE. Graphene derivatives GO and rGO can be used as an ETL for OSCs. It was further revealed that GO-grafted polyethyleneimine ethoxylate-based HTL used for OSCs with PTB7:PC₇₁BM as an active layer exhibited an efficiency of around 8.21%.²³⁴ GO/metallic oxide hybrid ETL exhibits better performance than conventionally used metallic oxide HTL such as ZnO and TiO_x. The rGO-TiO_x based HTL for OSCs improved efficiency by 8% compared to HTL with only TiO_x based, with PCE around 2.7%.²³⁵ MoS₂-ZnO hybrid ETL was used for OSC with an active layer of PTB7-Th:PC₇₁BM exhibited PCE around 10.1%, which was more than 15% higher than without MoS₂-based OSC.²³⁶ Few layer BP deposited on ZnO was employed as an ETL for PTB7/PC₇₁BM bulk heterojunction OPVs for inverted OSC (Figure 12A). The device improved the performance with PCE up to 8.18% due to the excellent alignment of work function between BP/ZnO ETL and the organic channel layer (Figure 12B,C).²³⁷

Figure 12. (A) Schematic illustration for inverted OSC based on BP as ETL. (B) Energy band alignment of OPV between ETL BP and PTB7/PC71BM. (C) J-V characteristics curve for inverted OPV based on BP incorporated with PTB7/PC71BM. Reproduced with permission: Copyright 2015, Wiley.237 BP, black phosphorus; ETL, electron transport layer

2D materials as additives for OSCs

Besides serving 2D material for OSCs such as HTL and ETL, they can also enhance their performance using an additive with organic bulk heterojunctions. BP quantum dots (BPQDs) with variable sizes were utilized for organic bulk heterojunction material. Furthermore, the hybrid BPQDs-based structure exhibited a cascaded band structure that enhanced the interface's carrier separation (Figure 13A). The scattering sites for organic bulk heterojunction enhanced device performance, such as PTB7/PC₇₁BM and PBDTTT-EFT/PC₇₁BM (Figure 13B). BPQDs offered the role of light-harvesting with 0.55 wt%, resulting in a PCE increment of around 10% (Figure 13C).²³⁸ Kakavelakis et al.²³⁹ demonstrated, few-layer WSe₂ obtained by the liquid exfoliation method, an efficient additive for binary PTB7/PC₇₁BM blend base inverted bulk heterojunction OSC (Figure 13D). The energy band alignment between few-layer WSe₂ and PTB7/PC₇₁BM improved charge transport (Figure 13E). Using few-layer WSe₂ as an additive for inverted PTB7/WSe₂/PC₇₁BM ternary PV exhibited a high PCE of about 9.3%, 15% more than binary PTB7/PC₇₁BM based device without WSe₂ (PCE = 8.7%) (Figure 13F).²³⁹

Figure 13. (A) Schematic illustration for BPQDs-based OSC. (B) Energy band alignment illustration of BPQDs-based heterostructure solar cells. (C) J-V characteristics for different BPQDs heterostructure-based solar cells. Images (A)–(C) reproduced with permission: Copyright 2017, Wiley.238 (D) Schematic illustration of WSe2-based additive ternary inverted solar cell. (E) Energy band alignment illustration for PTB7/WSe2/PC71BM ternary inverted solar cell. (F) J-V characteristic curve for WSe2-based additive with different lateral sizes. Images (D)–(F) reproduced with permission: Copyright 2017, American Chemical Society.239 BP, black phosphorus; BPQDs, BP quantum dot

2D MXene (Ti₃C₂T_x) was recently utilized as an additive for PBDB-T:ITIC and PM6:Y6 organic bulk heterojunctions. The additive of 2D Ti₃C₂T_x enhanced the photovoltaic properties of OSCs such as PCE for PBDB-T:ITIC-based OSC up to 10.72% and 10.25% for PM6:Y6 with increment of short current density and charge dissociation.²⁴⁰ The results demonstrate that high-performance solar cells can be designed after carefully selecting 2D organic and inorganic combinations that allow interface engineering using low-cost methodologies.

Neuromorphic computing devices

The rapidly growing progress showed by artificial intelligence, machine learning, and the internet of things has increased the demand for high-power computing that can process massive data transfers using efficient energy consumption and high speed to overcome the limitations of conventional von Neumann computing devices.^241–243 Among all computing devices, neuromorphic-based devices that are bioinspired synaptic devices emulating the human brain are considered the most efficient, with unique properties such as high energy efficiency, parallelism, and ability to perform cognitive tasks such as recognition, learning, and adaptations. Synapses are known as the fundamental unit of the human brain that is used for computing and memory.^244,245 The human brain consists of 10¹¹ neurons with 10¹⁵ synapses responsible for neural signal transmission and power consumption of only 20 W power with operation consumption for each stimulus around 1–100 fJ.²⁴⁶ Short-term plasticity (STP) and long-term plasticity (LTP) are considered important parameters to optimize the performance of neuromorphic computing devices. Considerable efforts have been done to imitate synaptic functions using traditional devices such as FET,^247,248 memristors,^249,250 OFET,^251,252 ferroelectric memory,^253,254 floating gate memory,^255,256 resistance switching²⁵⁷ etc. Recently organic−inorganic-based neuromorphic devices have been reported. The idea behind combining organic−inorganic-based neuromorphic computing devices is to utilize the biocompatibility of organic materials with tunable properties and high carrier mobility of inorganic materials for high-speed neuromorphic computing devices. The charge trapping properties of organic−inorganic hybrid materials make them attractive for fabricating neuromorphic computing devices with improved performance.

Recently, Wang's group²⁵⁸ investigated MoS₂/PTCDA hybrid heterostructure for artificial neural synapse transistor, which exhibited neuromorphic synaptic function exhibiting STP and LTP with mimic behavior of biological neurons as shown in Figure 14A. Using this hybrid structure for the first, both electrical and optical modulations were achieved due to the unique type II band alignment of MoS₂/PTCDA. The charge carriers were electrically and optically injected into MoS₂, which are further captured and separated at the heterojunction interface, resulting in gradually biological synaptic functions because of gate tunable band alignments of type II heterojunction. The AFM characterization was used to analyze the interface between MoS₂/PTCDA (Figure 14B) showing a clean interface between MoS₂ and PTCDA. The inhibition and excitation were simultaneously achieved using gate voltage tuning. The flexible tenability in STP and LTP with synaptic weight changes up to 60 shows neuromorphic performance. Using optical and electrical modulation, the synaptic device also exhibited excitatory postsynaptic current EPSC as shown in Figure 14C device based on MoS₂/PTCDA compared to previously reported devices based on only optical modulation.²⁵⁸

Figure 14. (A) Synaptic device structure based on MoS2/PTCDA based heterojunction modulated by optical and electrical spikes. (B) Illustration of MoS2/PTCDA interface by AFM. (C) EPSC of MoS2/PTCD heterostructure under optical modulation, i.e., triggered by pair of laser pulses. Images (A)–(C) reproduced with permission: Copyright 2018, Wiley.258 (D) Schematic illustration of the synaptic device based on graphene quantum dots-PEDOT:PSS. (E) Current/voltage characteristic of graphene-quantum dot-PEDOT:PSS synaptic device under positive voltage sweep. Images (D) and (E) reproduced with permission: Copyright 2017, Springer Nature259

Graphene quantum dot-PEDOT:PSS-based hybrid structure was also employed to explore the synaptic functional-based device (Figure 14D). From current-voltage characteristics, it was found that under dual positive voltage sweep, the device showed memory behavior due to a gradual increment of hysteresis, which also exhibits the memristive synaptic behavior of the device (Figure 14E).²⁵⁹ Yang et al.²⁶⁰ studied the graphene-perovskite-based hybrid structure for optical synaptic devices for artificial devices. The as-fabricated graphene/perovskite heterostructure-based device exhibited high photoresponsivity up to 740 A/W showing enormous stability for 74 days. The optical synaptic-based device also revealed light-evoked excitatory/inhibitory functions used for image recognition with an accuracy of 80%.²⁶⁰ The above mentioned 2D organic−inorganic-based neuromorphic devices inspire unprecedented new ideas for designing future high-performance synaptic electronic devices.

SUMMARY AND OUTLOOK

This review has discussed the recent progress in the 2D organic−inorganic heterostructures, including their structure, synthesis, and potential applications in electronic and optoelectronics. Organic−inorganic heterostructures have displayed remarkable properties compared to their individual components. Integrating organic and inorganic materials has created many opportunities for fabricating multifunctional devices by taking advantage of both materials. Although considerable progress has been made by organic−inorganic hybrid structures in the synthesis process, device fabrication, interface studies, and applications in electronics and optoelectronics, there is still a lot to do to overcome some key challenges in this field to get significant output from them.

First, as the potential applications based on organic−inorganic heterostructures mainly depend on the charge transport at the interface, which requires further research to explain the complex interactions in organic and inorganic heterostructure. There is still no solid theory or model that can completely describe the charge transfer mechanism at the organic−inorganic interface. For example, whether the interface is governed by dipole-dipole interactions or covalent bonding. Second, 2D inorganic materials with their flat surfaces can act as a template for the crystalline growth of organic materials, which is significant for high-performance device applications, so it is essential to develop synthesis methods to grow large-scale, low-defect, low-cost, and large crystalline inorganic domains so that organic materials can grow efficiently with a high crystalline structure. Similarly, there is also an urgent need to develop a synthesis mechanism to grow large-area organic−inorganic heterostructure, i.e., wafer-scale synthesis, for highly demanded practical application at the commercial level. Third, most organic semiconductors have wide bandgaps, which limits their application domains. New organic materials should be processed with narrow bandgaps that are demanded to fabricate photodetectors for NIR to IR detection to overcome this problem. Fourth, one of the primary differences between organic material and inorganic material is their chemical and thermal stabilities. Organic materials are thermodynamically less stable; although it helps for low-temperature growth, their low melting points can degrade device performance due to thermodynamically instability in ambient environmental conditions. To deal with these issues, much effort is required to find new organic materials that are highly stable, or some other strategies may be applied, such as encapsulation layers on organic materials to reduce environmental effects. Fifth, the material selection for heterostructure is crucial to optimize the performance of optoelectronic devices such as photodetectors, solar cells and p-n junctions for electronic devices; for that, it is essential to study the energy band alignment of organic−inorganic heterostructures carefully. Interface studies, such as charge transfer at the interface and charge dissociation and recombination studies, are important because they can help selecting the most suitable materials for heterostructures showing excellent properties.

We think that significant breakthroughs are possible by combining organic materials with new 2D inorganic materials such as highly conductive MXene for advanced electronic and optoelectronic devices, which are still rarely explored. Similarly, apart from MoS₂, many other TMDCs are still not combined with organic materials, suggesting exciting opportunities for applications, e.g., for flexible electronics. Another type of organic material not mentioned in this review is called photochromatic molecules, which are functional and reversibly converted into two distinct isomers upon specific light wavelength illumination. Combining photochromatic molecules with 2D materials may result in stimulus changes such as electronic and photonic due to photoinduced doping effects, enabling organic−inorganic-based molecular switch applications. Photodetectors having spectral detectivity in NIR to IR are highly demanded potential applications such as optical communication, thermal imaging, and medical fields. There are few reports about combining organic with narrow bandgap topological insulators with the spectral range in NIR to IR. Further research is needed to use 3D topological insulators to make their heterostructures with organic materials to find applications in ultrafast flexible photodetectors with NIR to IR spectral range. Although the research on organic−inorganic heterostructures is at its early stage with challenges to be addressed, we believe this newly emerging field will bring many exciting opportunities by overcoming these challenges.

In general, to satisfy the requirement of potential applications at commercial level, heterostructure plays a vital role. Although considerable progress has been made by recently emerging 2D–2D inorganic heterostructures, however, their low optical absorption, defect free synthesis and limited available choices of p-n heterostructures are key challenges. The recent progress in material synthesis and fabrication methods has open new opportunities to combine organic materials with inorganic materials to overcome above mention challenges. The large available variety of organic materials will further expand the library of heterostructures family with freedom in materials choices and without considering lattices match requirements. The tunable properties of organic materials can alter the solid-state properties which is quite unique to fabricate potential multifunctional electronic and optoelectronic devices. By further development in the synthesis and fabrication techniques, it would be possible in the near future to fabricate more advanced organic−inorganic heterostructures for applications such as optoelectronics, electronics, sensors, neuromorphic devices, and solar cells.

ACKNOWLEDGMENTS

This study was financially supported by the National Science Fund for Distinguished Young Scholars (No. 52125309), the National Natural Science Foundation of China (Nos. 51991343, 52188101, and 51991340), the National Key R&D Program (No. 2018YFA0307300), Guangdong Innovative and Entrepreneurial Research Team Program (No. 2017ZT07C341), and the Shenzhen Basic Research Project (No. JCYJ20200109144616617).

CONFLICTS OF INTEREST

The authors declare no conflicts of interest.