Introduction

The successful isolation of two-dimensional (2D) graphene gave birth to a new era of atomic solid-state electronics, which ultimately led to the search for other 2D vdW materials. The structural and electronic variety in vdW-bonded 2D materials opens the possibility of 2D/2D heterogeneous integration at the atomic scale for fundamental scientific studies and applied device designs.1 These all-2D vdW heterostructures show many unique electronic, optical, thermal, and mechanical properties, imbuing them with new functionalities.2,3,4,5,6 However, to date, there still exists a challenge to produce all-2D materials and heterostructures over large areas with high quality. Furthermore, carrier concentrations, doping type, and stoichiometry remain big challenges for 2D materials and vdW heterostructures, limiting their practical applications. The vdW interactions are not limited to 2D materials. Indeed, any 2D materials can be integrated with a variety of materials of differing dimensionality to form mixed-dimensional vdW heterostructures (2D + nD, n = 0, 1, and 3).7 Such combinations show characteristically advantageous optical, electronic, and/or electrochemical properties of each class of materials.8

Compared to three-dimensional (3D) bulk materials, the electronic wave function of low-dimensional materials (zero-dimensional [0D], one-dimensional [1D], and 2D) is spatially restricted by a physical boundary due to quantum confinement.9,10 The physical properties of these materials can be highly tunable as a function of geometry and are sensitive to external stimuli because of the large surface-to-volume ratios.11 As shown in Figure 1a, the materials can be categorized by the number of dimensions along which charge carriers are free to move. The degrees of freedom for a charge carrier dictate its density of states and the degree to which its k-space is filled with electrons. Figure 1b shows the probability curves for two atoms that are vdW-bonded to each other. The low-dimensional materials are usually more likely to reach the vdW distance and activate the vdW interaction.12 The wave functions of the electrons of the two atoms overlap, which means that the two electrons will interact. This interaction or perturbation results in the quantized energy level splitting into two discrete energy levels. The splitting of the discrete state into two states is consistent with the Pauli exclusion principle. In mixed-dimensional vdW heterostructures, the periodic arrangement of atoms with initial quantized energy level will split into a band of discrete energy levels.

Figure 1
figure 1

Schematic diagram of quantum confinement and density of states. (a) Schematic diagram of 0D, 1D, 2D, and 3D materials and their corresponding density of states. The 0D materials have no degrees of freedom for charge carriers and thus available states exist only at discrete orbital energies. In 1D/2D materials, there are one/two degrees of freedom for charge carriers allowing for the existence of states in between discrete orbital energies, whereas for the 3D bulk materials, the charge carriers have three degrees of freedom for transport, and thus they show quasi-continuous density of states. (b) Schematic illustration of the van der Waals (vdW) distance and vdW gap in a two-atom vdW system, and the overlapping probability density functions of the two adjacent atoms.

Full size image

Multifunctional devices from vdW heterostructures

Over the past few years, several functional devices have been realized as a result of creating or forming novel mixed-dimensional van der Waals heterostructures. These device applications range from simple controlled doping of the 2D channel layer by 0D molecular materials to low-power logic transistors, memory devices as well as photodetectors. Some selected examples illustrating this are detailed next.

As shown in Figure 2a, the mixed-dimensional vdW heterostructures not only retain the exceptional tunable properties of 2D materials, but also incorporate controllable doping, sharp interfaces, and regulated growth characteristics from 0D, 1D, and 3D materials.

Figure 2
figure 2

Multifunctional devices of mixed-dimensional van der Waals (vdW) heterostructures. (a) A schematic diagram of a mixed-dimensional vdW heterojunction. (b) A schematic diagram of a 0D-benzyl viologen molecules/2D graphene dual gate transistor. (c) Optical image of InSe/Si tunnel triodes; the average of threshold slopes is as low as 34 mV·decade−1. (d) Fast photoresponse time of ~15 μs for 1D-single-walled carbon nanotubes (SWCNTs)/2D graphene photodetector. Inset: the photocurrent mapping of the 1D/2D photodiode. (e) A 2D-3D of graphene/mercury–cadmium–telluride (MCT) vdW heterojunction photodetector, showing the 3 dB cutoff frequency of 77 MHz. (f) A schematic diagram of the n-MoS2/p-pentacene gate-tunable photodetector. (g) Structure of MoS2/Si light-emitting heterojunctions. Electrons are injected from n-type MoS2, while holes are injected from the p-Si substrate. (h) A schematic of the integrated complementary metal oxide semiconductor readout circuit and graphene sensor junctions. (i) A schematic of the MoS2/AlScN ferroelectric field-effect transistor (FE-FET) and scanning electron microscopy image of the FE-FET array. (b–i) Reprinted with permission from References 13, 18,19,20, 26, 27, 33, and 34. © 2011, 2014, 2016 American Chemical Society, 2017, 2022, 2023 Springer Nature, 2013 National Academy of Sciences, 2022 Wiley group. FET, field-effect transistor; HJ, Heterojunction; TT: tunneling triode; NC-FET, negative capacitance field-effect transistor; BP FET, black phosphorus field-effect transistor.

Full size image

The chemical doping effect of 0D benzyl viologen molecules can modulate the Fermi level in graphene, enabling tunable Dirac points in graphene transistors, as shown in Figure 2b.13 The 1D single-walled carbon nanotubes (SWCNTs) serve as a gate electrode for transistors, breaking free from the constraints of traditional gate electrode dimensions and no damage is inflicted upon the 2D channel, resulting in exceptionally high charge carrier mobility.14,15 This makes it suitable for high-speed, high-frequency electronic applications. Considering the controllable doping and stability of 3D materials, their combination with 2D materials leads to high-performance transistors.16,17 In Figure 2c, 2D InSe/3D Si tunnel triodes with a sharp vdW interface achieve the average subthreshold of 34 mV·decade−1 (red star symbol in Figure 2c), breaking the thermionic limit of 60 mV decade−1.18

Furthermore, the performance of photonic and energy devices, such as photodetectors, solar cells, and light-emitting diodes, can be significantly enhanced by using the mixed-dimensional vdW heterostructures. In Figure 2d, the combination of 1D single-walled carbon nanotubes (SWCNTs) and 2D graphene can achieve response time less than 15 µs.19 In Figure 2e, given the ultrahigh mobility of 2D graphene and high quality of 3D HgCdTe, the graphene/HgCdTe heterojunction devices show a response time of 13 ns and a peak detectivity of 2 × 1010 cm Hz1/2 W−1 at room temperature for blackbody irradiation.20 The mixed-dimensional vdW heterostructures also enabled dual-color and polarization-sensitive photodetection, for example, the bP/MoS2/Si dual-color photodetectors,21 and HgCdTe/black phosphorus polarization-sensitive photodetection.22

For energy devices, the graphene/semiconductor heterojunction solar cells garnered tremendous attention owing to their favorable energy-conversion efficiency.23,24 Using the gate-tunable structure in graphene/GaAs heterojunctions, researchers achieved a remarkable peak energy-conversion efficiency of 18.5 percent.25 Note that solar cells combining organic small molecules of pentacene, and MoS2, along with gate control, managed to attain an open-circuit voltage exceeding 0.5 V, demonstrating the potential mixed-dimensional vdW heterostructures for solar cells,26 as shown in Figure 2f. The 2D/3D heterostructures can serve as light-emitting diodes, for example, the p-Si/MoS2 heterojunction with the recombination of injected holes and electrons in MoS2 leads to strong light emission,27 as shown in Figure 2g. The 2D graphene and h-BN play an important role in device photoluminescence; where they can assist in transferring light-emitting devices onto various substrates,28,29 and enhance the device cooling.30 A 0D organic quantum dot (NSOD)/2D g-C3N4 vdW heterojunction effectively promotes charge separation and electron transfer, enabling efficient photocatalytic hydrogen production processes.31

The 3D silicon-based materials possess well-established growth and device fabrication processes, and the graphene and silicon dominate this field, encompassing graphene/Si-based CMOS integrated circuits32 and gas sensors.33 As shown in Figure 2h, the integration of monolayer graphene with the CMOS platform, with superior gas sensitivity of graphene and the low power and cost attributes of CMOS, enables the realization of high-performance radio-frequency gas sensors.33 In Figure 2i, the hetero-integration of 2D MoS2 with ferroelectric 3D AlScN ferroelectric field-effect transistors (FE-FETs) has enabled a storage window exceeding 7.8 V, with 10-year retention characteristics and 4-bit pulse-programmable memory functionality, opening the door to BEOL-compatible large-scale 2D/3D transistor memory technology.34

Wafer-scale growth of vdW heterostructures

Currently, mixed-dimensional vdW heterostructures have been synthesized through various methods, including chemical vapor deposition (CVD), metal–organic chemical vapor deposition (MOCVD), and mechanical transfer.35,36,37,38,39 Figure 3a shows the 2D/0D vdW heterostructures of graphene/WS2 quantum dots.40 The 0D WS2 quantum dots synthesized by a Li-intercalation method are spin-coated onto CVD-grown 2D graphene. The transmission electron microscopy (TEM) reveals that WS2 quantum dots have uniform dimensions with a diameter of ∼2.75 nm. In addition, the Raman spectrum of 2D graphene/0D WS2 quantum dots vdW heterostructure shows that the 2D peak shifts to a higher frequency, and both the G peak and 2D peak intensities significantly increase, indicating photogenerated holes are injected from 0D WS2 quantum dots into 2D graphene. The photoluminescence (PL) spectrum of the 2D graphene/0D WS2 quantum dots vdW heterostructure further supports the occurrence of charge transfer at the interface. Figure 3b shows the vdW heterostructure of 2D graphene/1D InxGa1−xAs.41 Due to the near-lattice-matched registry between InAs and graphene, and graphene’s lack of dangling bonds, InGaAs cannot achieve strain sharing through elastic deformation with graphene. Therefore, InGaAs self-organizes into InAs core and InGaAs shell segments. Using the Au-assisted vapor–liquid–solid (VLS) mechanism, InGaAs nanowire can be epitaxially grown on 2D graphene. A TEM image of InAs nanowires is grown on a large-area (1 × 1 cm2) defect-free graphene substrate, where the height and diameter of the InAs nanowires can be precisely controlled. Raman studies indicate that the graphene maintains excellent crystalline quality after the growth of InAs nanowires. Figure 3c shows the 2D/2D vdW heterostructures of 2D MCs (metal chalcogenides)/2D TMD (transition-metal dichalcogenides).42 To grow the MC/TMD heterostructures, first, metal iodides are epitaxially grown on CVD-grown TMDs substrates. Subsequently, sulfur atoms replace the iodine atoms in a low-temperature epitaxial substitution process, leading to the formation of 2D MC/2D TMD vdW heterostructures. The TEM characterizations reveal the nearly perfect SnS2/WS2 vdW interface. Utilizing the low-temperature epitaxial substitution approach, wafer-level integration of various 2D MC/2D TMD materials is achieved. Figure 3d shows the 2D/3D van der Waals heterostructure of 2D MoS2/3D GaN.43 A 3D GaN thin film is epitaxially grown on a sapphire (c-plane) substrate using the MOCVD method. Subsequently, a monolayer of MoS2 nanosheets is grown on the surface of the GaN thin film using the CVD method. Raman spectroscopy characterization confirms the high crystalline quality of the 2D MoS2 grown epitaxially on the 3D GaN thin film. Further analysis using Kelvin force microscopy images indicates excellent van der Waals contact and uniformity in the 2D MoS2/3D GaN heterostructure.

Figure 3
figure 3

Wafer-scale growth of van der Waals (vdW) heterostructures. (a) Graphene/WS2 quantum dots 2D/0D vdW heterostructure. Transmission electron microscopy image, Raman spectra, and photoluminescence spectra of pristine WS2 quantum dots (QDs) and the graphene/WS2 QDs heterostructure. (b) Two-dimensional graphene/1D-InxGa1−xAs vdW heterostructure. Large-area InAs nanowires (1 × 1 cm2) growth on defect-free graphene substrate. (c) Wafer-scale growth of 2D-metal chalcogenides/2D-transition-metal dichalcogenides  vdW heterostructure array. (d) High-quality large-scale MoS2/GaN 2D/3D vdW heterostructure. SMLs, Single monolayers; MLs, monolayers. (a–d) Reprinted with permission from References 40,41,42,43. © 2013, 2016, 2022 American Chemical Society, 2023 Springer Nature.

Full size image

Outlook

Despite the excellent physical properties and many exciting demonstrations of proof-of-concept devices from mixed-dimensional van der Waals heterostructures previously discussed, the challenges of turning them into practical technologies are still abundant. For example, for most vdW interfaces, the uniformity, wrinkles, size, and interface/surface contamination will severely limit the device yield and performance, whereas these issues become even worse with growing integration complexity. Another challenge is the stability and reliability of the heterostructures. The weakly bonded vdW interactions and thermal properties of individual components could lead to interfacial strain and sliding. With atomically thin geometry, highly selective etching methods, metallic contacts, and dielectric integration for developing high-performance devices have not yet been fully explored.

In this vein, this issue of MRS Bulletin contains four excellent invited contributions on various aspects and applications of mixed-dimensional van der Waals heterostructures. First, the article by Utama et al.44 focuses on various mixed-dimensional heterostructures for quantum photonic applications. This is a particularly exciting space in which recent progress has been made by engineering strain and defects in 2D quantum materials with the help of 0D and 1D nanostructures. This article accurately and thoroughly covers these topics with an emphasis on quantum emitters and opportunities to scale them. The article by Dong et al.45 specifically focuses on another important area of photonic technology, which is infrared photodetection. In this case, the authors highlight several recent important works and advances in integration of 2D materials with bulk 3D semiconductors and metals, and how that has led to improvements and novel functionalities in various modalities of infrared photodetection both at a device and system level. Besides quantum photonics and infrared photodetection, there are two other articles that are germane to the growing body of literature on mixed-dimensional heterostructures. One of them focuses on a key technological question for the future of this field, which is wafer-scale epitaxy of 2D materials on 3D substrates by Hakami et al.46 By thoroughly discussing various atomic-scale consideration for epitaxial nucleation and templating as well as defect formation, the authors have presented an invaluable account on scaling growth of high-quality single crystalline 2D semiconductors and 1D ribbons of the same on 3D substrates. Finally, mixed-dimensional heterostructures such as between 1D/2D and 0D/2D materials not only provide an increased surface area but also electrical connectivity for charge storage and facilitating chemical reactions. This makes them uniquely suited for both electrochemical storage as well as electrocatalysis applications in which a large body of work has emerged using mixed-dimensional heterostructures, expertly reviewed in the article by Chen et al.47

These discussions and specifically the four invited contributions in this issue are a testament to the growing activity in this field. With further advances in scalable growth, assembly, fabrication, and discovery of new physical properties, research in this field is expected to grow for many years to come.