Optoelectronic Properties of Heterostructures

Optoelectronic Properties of Heterostructures The most recent developments based on graphene and transition-metal dichalcogenides.

T

The extremely high chargecarrier mobility in graphene has motivated the exploration of unique properties in other two-dimensional (2-D) materials, such as hexagonal boron nitride (h-BN), phosphorene, silicine, and the transition-metal dichalcogenides (TMDs). The latter compounds have generated great interest due to their potential for optoelectronic applications, since their band gaps are tunable as a function of the number of layers, even as the compounds remain flexible and nearly translucent when composed of a few atomic layers. Although nearly all the layered 2-D materials reveal a potential for optoelectronic applications, their heterostructures show remarkable properties that may not occur in their individual constituent layers. This review article intends to highlight some of the most recent

developments on the optoelectronic properties of heterostructures based on graphene and TMDs.

GRAPHENE AND RELATED MATERIALS The discovery of the electronic properties of graphene [1], [2] motivated the exploration of other potentially unique 2-D bonded materials. The planar structure of graphene, i.e., a single atomic layer of graphite, displays unique optoelectronic properties that have generated considerable interest for over a decade. The linear dispersion of the electronic band structure of graphene leads to its extremely high charge-carrier mobility of ~105 cm2/Vs for a monoatomic layer at room temperature. Therefore, the charge carriers are described in terms of nearly massless Dirac fermions [1], [3], [4]. Despite graphene’s high charge-carrier mobility, however, the zero

band gap of pristine graphene makes it unsuitable for optoelectronic applications. Still, graphene and its remarkable characteristics have motivated the exploration of other 2-D layered materials, which are shown in Figure 1. TMDs with the general formula MX2 display semiconducting, semimetallic, and superconducting behavior at low temperatures and are at the forefront of current research [5]–[12]. In the general formula, M could be such elements as molybdenum (Mo), tungsten (W), rhenium, or niobium, and X such elements as sulfur (S), selenium (Se), or tellurium.

NIHAR R. PRADHAN, SAIKAT TALAPATRA, MAURICIO TERRONES, PULICKEL M. AJAYAN, AND LUIS BALICAS Digital Object Identifier 10.1109/MNANO.2017.2676185 Date of publication: 3 April 2017

18 |

IEEE nanotechnology magazine | june 2017

1932-4510/17©2017IEEE

Background—©iStockphoto/chin graph, Heterostructure— courtesy of the American Chemical Society

Semiconducting TMDs have natural band gaps ranging from 0.8 to 3 eV, which stimulates interest in evaluating their potential for the next generation of optoelectronic applications including field-effect transistors (FETs), photodiodes, solar cells, and logic elements, among others. Phosphorene [13] is yet another layered material, which possesses a band gap of 0.3 eV and displays great potential for optoelectronics, given its high charge-carrier mobility. Transition metal monochalcogenides (TMMs) [14] are yet another class of layered materials. They contain a single chalcogen atom and have the general formula MX, with M being such possible elements as gallium, tin, or indium, and X such elements as S, Se, or Te. These

compounds are semiconducting, and we are currently witnessing the beginning of the exploration of their optoelectronic properties. Recently, a new class of layered ma terials, the phosphorus-based trichalcogenides, i.e., metal thiophosphates such as manganese thiophosphate and nickel thiophosphate, have attracted attention for their semiconducting and magnetic properties, which might reveal a potential for spintronics [16]. These materials are also semiconductors, but with band gaps ranging from ~0.3 to ~3 eV [16]. Transition metal trichalcogenides (TMTs) is another class of layered materials belonging to the semiconductor family. It is in the beginning of exploration for optoelectronics application.

Finally, h-BN is an insulator with a band gap of ~6.4 eV [17], which displays a perfect 2-D structural analog of graphene and acts as an excellent dielectric substrate. That is the reason why it is widely used to fabricate high-quality heterostructures in combination with other layered materials [17], [18]. Recently, many of these layered compounds were independently explored, several of them revealing a promising potential for optoelectronic applications. However, heterojunctions based on the stacking of these layered materials show remarkable properties not found in the constituent compounds.

HETEROSTRUCTURE FABRICATION Heterostructures based on 2-D materials are at the forefront of materials science

june 2017 | IEEE nanotechnology magazine |

19

h-BN is an insulator with a band gap of ~6.4 eV, which displays a perfect 2-D structural analog of graphene and acts as an excellent dielectric substrate. research due to their novel functionalities, which complement those of the constituent crystals. The basic principle of fabrication is to stack one type of 2-D layer on top of another, while keeping the interfaces clean and with a minimum of disorder or defects. See Figure 2 for a description of the method.

THE HOFSTADTER BUTTERFLY IN GRAPHENE HETEROSTRUCTURES Vertical heterostructures are produced by mechanically stacking different layers

on top of one another or by depositing one type of layer on top of the other via chemical vapor deposition (CVD) [17] or molecular beam epitaxy (MBE) [18]. CVD and MBE processes can grow large-area heterostructures. However, mechanical exfoliation of bulk single crystals and the use of the dry transfer method were recently adopted to produce high-quality heterostructures. Dean and coworkers initially produced graphene-based heterostructures on h-BN using a polymer transfer method [19].

Semiconductor

Semiconductor

TMMs InSe, GaSe, GaS, GaTe, SnSe, etc.

TMDs MoS2, MoSe2, WS2, WSe2, MoTe2, ReS2, ReSe2, MoTe2, etc. Graphene

Thiophosphate MnPS3, NiPS3, FePS3, FePSe3, ZnPS3, etc.

Semiconductor

TMTs TiS3, ZrS3, Bi2Se3 TaSe3, NbS3, etc.

Semiconductor and Magnetic

Insulator

Semimetal/ Superconductor TMDs

h-BN

WTe2, β-MoTe2, NbSe2, NbS2, etc.

FIGURE 1 A schematic representation of the class of 2-D layered materials that has been at the forefront of current research interest. The materials are divided into different categories according to their electrical properties and crystal structures.

20 |


After the development of this technique, heterostructures based on 2-D materials received a prodigious level of attention from researchers [20]–[23]. Initially, a polymer layer of polymethyl methacrylate is deposited on top of the flakes of 2-D materials that were previously exfoliated onto a silicon/silicon dioxide (Si/SiO2) substrate. Subsequently, the f lakes embedded in the polymer layer were detached from the substrate by etching the SiO2 layer and then transferred onto other exfoliated 2-D layered materials to create a stacked heterostructure. Soon thereafter, the same group reported a polymerfree stacking method that most likely produces cleaner interfaces between the stacked 2-D layers. In this method, one can pick up one type of 2-D-layered material, previously exfoliated onto the Si/SiO2 substrate, with another layered material via van der Waals attraction, as illustrated by Figure 2(a) [24]. This polymer-free method yields high-quality heterostructures, e.g., composed of graphene inserted between two h-BN layers, where one of the exfoliated h-BN layers is used to pick up the graphene layer [see Figure 2(a)]. The h-BN/ graphene stack is then transferred onto another exfoliated h-BN on an Si/SiO2 substrate, producing an h-BN/graphene/ h-BN heterostructure, as shown in Figure 2(a). In this transfer method, the graphene layer is in direct contact with the h-BN without any polymer residue, thus keeping the graphene pristine. An electron microscopy image of this heterostructure is shown in Figure 2(d), confirming the high quality of the interface between the graphene and h-BN layers. It was Dean et al. who first observed the Hofstadter butterfly in graphene [25], [26], which had been theoretically predicted by Hofstadter in 1976 [27]. The Hofstadter butterfly describes the behavior of electrons simultaneously subjected to a magnetic field and a periodic potential, the periodic potential in this case being the moiré pattern created by the relative orientation between the graphene and the h-BN crystals. As seen in Figure 3, the behavior of the Hall response shows a fractal structure, as predicted by a Diophantine

HETEROSTRUCTURES OF TMDs ON h-BN h-BN was proven to be an excellent dielectric substrate for high-quality graphene-based electronic devices and presumably also for other 2-D materials. For example, it has been shown that several niche physical properties

Heterostructures based on 2-D materials are at the forefront of materials science research due to their novel functionalities, which complement those of the constituent crystals. of MoS2 can be preserved using a thin BN buffer layer between the MoS2 and the SiO2 substrate [28]. The challenge is to make high-quality electronic devices of semiconducting TMDs while Schottky barriers between the metallic contacts and the TMDs obstruct their quality. In addition, dangling bonds and the roughness of the commonly used SiO2 dielectric substrate tend to trap charge carriers and limit the carrier mobility in these devices [10], [11]. Encapsulating TMDs in h-BN and then using graphene or platinum for the contacts is found to dramatically

improve carrier mobility. For instance, Hone et al. and Tutuc et al. demonstrated that high-quality contacts on monolayer or a few layers of MoS2 and WSe2 encapsulated within h-BN layers lead to the highest carrier mobilities observed so far at low temperatures, and they were able to experimentally observe Shubnikov–de Haas (SdH) oscillations in TMDs for the first time [29], [30]. Figure 4(a) shows the schematics of a heterostructure where MoS2 flakes are sandwiched between few-layered h-BN and the contacts are fabricated by placing

C

PD

PP

equation. This observation was possible because the periodicity imposed by the moiré pattern is comparable to the magnetic length. Figure 3(a) shows the device geometry where the metal electrodes connect at the edge of the graphene, acting as a one-dimensional contact. Figure 3(b) displays the STEM image showing the edge contacts in detail. The graphene crystal is encapsulated between two h-BN layers and is subsequently etched to expose its edges before depositing the metallic contacts, as shown in Figure 3(a). This process leads to a low contact resistance. This edge contact technique, with h-BN encapsulation, enables high electronic performance, including lowtemperature ballistic transport over distances beyond 15 nm , and room- temperature mobilities comparable to the theoretical phonon-scattering limit of n ~140,000 cm2/Vs [24]. Very recently, a complex sequence of fractional quantum Hall effect states was reported from these high-quality h-BN/ graphene/h-BN heterostructures [26], as shown in Figure 3. Figure 3(c) shows the resistance as a function of the gate voltage for a sample characterized by a 10-nm moiré periodicity, which displays a peak at the charge neutrality point (CNP) ^ V g = 0 V h and two additional satellite peaks at either side that are equidistant from the CNP. Figure 3(e) shows the fractal structure in the Hall conductivity map expected for the Hofstadter butterfly. Figure 3(d) shows the Wannier diagram in which the position of the most prominent quantum Hall effect states are plotted against the normalized density. Undoubtedly, high-quality h-BN/ graphene/h-BN heterostructures open up new frontiers by offering the possibility of studying the complex behavior of electrons in a magnetic field when subjected to an underlying periodic potential.

BN

Si/SiO2

MS BN/Gr/BN

Gr

BN (a)

BN

BN

Gr BN SiO2

Gr

10 µ m (b)

BN 0.5 µ m

2 µm (c)

(d)

FIGURE 2 (a) A schematic of the van der Waals technique for polymer-free assembly of layered materials. (b) An optical image of a multilayered heterostructure using the process illustrated in (a). (c) An atomic force microscopy image of a large-area encapsulated graphene layer showing that it is pristine and completely free of wrinkles or bubbles, except at its boundary. (d) A high-resolution cross-section scanning transmission electron microscopy (STEM) image of the device in (c). The BN/Gr/BN interface is found to be pristine and free of any impurities down to the atomic scale. BN indicates boron nitride and Gr graphene. (Images courtesy of [25].) PPC: polypropylene carbonate; PDMS: polydimethylsiloxane.


21

50 nm Metal Lead BN Graphene BN SiO2

BN hene Grap BN

Si/SiO 2

Metal Lead BN

Etch Mask

Graphene Metal Leads

BN

Edge Contact

C + Pd + B + Cr Vg (Volts) (b)

(a)

5

103 102

10 λ (nm)

–5

15

1

2

3

4

5

20

30 0

20

–10

0

5

0 Vg (Volts) (c)

–20 (d)

1.0 5

0.8

3

0.4

0.6 φ/φo

0.6

log (Rxx) (kΩ)

0.8 φ/φo

0

10

1.0

0.4 0.2

0.2 0

–1

15

T = 300 K T = 1.6 K

101

–2

σxy (e 2/h)

After Annealing

–4 –3

40

B (Tesla)

40 30 20 10 0

105 104

Satellite CNP

∆ (meV)

106 R (Ω)

–5

BN SiO2 G Before BN Annealing

107

1 –4

–2

0 n/no (e)

2

4

0

–4

–3

–2

–1

0 n/no

1

2

3

4

(f)

FIGURE 3 (a) A schematic of the edge-contact fabrication process. (b) A high-resolution bright-field STEM image showing details of the edge-contact geometry. The expanded region shows a magnified false-color electron energy loss spectroscopy map of the interface between the graphene edge and metal lead. (c) Zero-field resistance versus gate bias. Left inset: an h-BN encapsulated graphene device, scale bar, 15 nm . Right inset: the gap measured by thermal activation at the CNP and hole satellite peak positions across four different devices. Error bars indicate the uncertainty in the gap, deduced from a linear fit to the activated temperature response. (d) Hall conductivity plotted versus magnetic field and gate bias and (e) longitudinal resistance versus normalized field and density for the same device. (f) A simplified Wannier diagram labeling the quantum Hall effect (QHE) states identified in (d) and (e). Light gray lines indicate all possible gap trajectories according to the Diophantine equation [n/n o] = t (z/z o) + s, where it assumed that both spin and valley degeneracy may be lifted such that s and t are allowed to take any integer, and n is electron density, n o is the number of states per unit cell, and z is the magnetic flux per unit cell. Families of states are identified by color as follows: Black lines indicate fractal integer QHE states within the conventional Hofstadter spectrum, including complete lifting of spin and valley degrees of freedom. Blue lines indicate conventional fractional QHE states. Red lines indicate anomalous QHE states that do not fit either of these descriptions, exhibiting integer Hall quantization but corresponding to a fractional Bloch index. (Images courtesy of [25] and [26].)

22 |


BN

Top h

Graphene–Metal Edge Contact

SiO2/Si

Graphene Top h-BN MoS2 Metal

Bottom h-BN SiO2

MoS2 N

m h-B

Botto

10 µm

5 nm

h-BN MoS2 Gr

h-BN

(a)

1 nm

(b)

(c)

105

1.8

2.5

Six Layers

103

102

One Layer (CVD) γ One Layer (CVD) Two Layers Three Layers Four Layers Six Layers 1

1.9 2.5 2.0 2.2 2.3

10

0.0

1.5

τq = 176 4 fs –0.4 0.04 0.10 1/B (T –1)

1.2 1.0

1.0

0.5

~T –γ

0.8 0.0

100 T (K ) (d)

0.4

Rxy (k )

Three Layers

2.0

Rxx (k )

1.4

Two Layers

Rxx (k )

µHall (cm2 V –1 s –1)

1.6 Four Layers

104

300

0

5

10

15

20

25

30

B (T ) (e)

FIGURE 4 (a) A schematic of an h-BN-encapsulated MoS2 device. The expanded view shows the individual components of the stack. (b) An optical microscopy image of the fabricated device. The graphene contacts are outlined by the dashed lines. (c) A cross-sectional STEM image of the fabricated device. The zoom-in false-color image clearly shows the ultrasharp interfaces between the different layers. (d) The Hall mobility of h-BN-encapsulated MoS 2 (with different numbers of layers) as a function of the temperature. (e) An observation of the SdH oscillations in h-BNencapsulated MoS 2. Longitudinal resistance R xx (red curve) and Hall resistance R xy (blue curve) of an h-BN-encapsulated monolayer chemical vapor deposited MoS 2 as a function of the magnetic field B measured at 0.3 K and with a carrier density of 9.7 × 1012 cm−2. (Images courtesy of [29].)


23

Encapsulating TMDs in h-BN and then using graphene or platinum for the contacts is found to dramatically improve carrier mobility.

states, are observed at even filling factors of v = 4 , 6, 8, and 10. These studies con firm the high quality of the heterostructures produced by stacking 2-D-layered materials, which are used to explore novel electron physics at low temperatures and under high magnetic fields.

GRAPHENE/h-BN/TMD HETEROSTRUCTURE

Rxx (kΩ)

0.3

26

0.8

0.2 0.4

0.1 0.0

28 0

3

6 9 B (T ) (a)

4

1.2

12

0.0

p = 3.2

1012 cm–2

T = 1.5 K

3

Rxx (kΩ)

p = 7.9 1012 cm–2 T = 1.6 K

Rxy (kΩ)

0.4

These results suggest that one can achieve the quantum Hall regime in these materials by improving the quality of the device. Soon thereafter, Tutuc et al. reported SdH oscillations and evidence for quantum Hall states in monolayer and bilayer WSe2 encapsulated within h-BN in a dual-gated configuration [30]. The measured longitudinal resistance (R xx) and Hall resistance (Rxy) are depicted in Figure 5(a) for monolayered WSe2 displaying well-defined SdH oscillations at applied gate voltages VTG = −6 V (top gate) and Vbg = 0 V (back gate). The filling factors corresponding to the two lowest Landau levels probed in this measurement, v = 26 and v = 28, are indicated by the green dotted lines. The carrier density for holes (p), calculated from the slope of R xy as a function of B at low fields, is p = 7.9 × 1012 cm−2. The corresponding data of a bilayer WSe2 heterojunction measured at T = 1.5 K and under applied gate voltages VTG = −6.4 V and VBG = 60 V are shown in Figure 5(b). As seen, SdH oscillations, along with quantum Hall

6 5

6

2

4

8 2

1 0

Rxy (kΩ)

few-layered graphene. Figure 4(b) shows an optical micrograph of an h-BN/MoS2/ h-BN heterostructure configured in a Hall bar geometry. The high-resolution scanning electron microscopy image shown in Figure 4(c), reveals ultraclean interfaces between each h-BN flake and the graphene crystal in the stacked heterostructure. The sandwich of MoS2 with h-BN layers reduces the scattering from phonons from the SiO2 substrate and, while minimizing the role of the charged impurities in the substrate, results in a band-like transport that approaches the ideal limit imposed by the acoustic phonons at room temperature or ballistic phonons over distances exceeding 15 nm at low temperatures. Figure 4(d) shows the Hall mobility (n Hall) of single-layer and of few-layered MoS2 FETs, with the highest chargecarrier mobility extracted from a six-layered device being as high as 34,000 cm2/Vs at 4 K. These heterostructures also display SdH oscillations in the magnetoresistivity [R xx(H)] (red curve) as a function of applied magnetic field [see Figure 4(e)].

0

10

20 B (T ) (b)

30

0

FIGURE 5 (a) The longitudinal resistance ^R xxh and Hall resistance ^R xy h of a WSe2 monolayer and bilayer heterostructures ^h-BN/WSe 2/ h-BNh. Instead of graphene, platinum was used for the contacts with the WSe 2 layers. Monolayer WSe2: R xx (left axis) and R xy (right axis) as a function of field B measured at T = 1.6 K. The two lowest filling factors v = 26 and v = 28 are indicated. (b) A bilayer WSe 2: R xx (left axis) and R xy (right axis) as a function of B measured at T = 1.5 K. A quantized R xy plateau is observed at v = 6. (Images courtesy of [30].)

24 |


Here, we present quantum wells (QWs), which use graphene as the transparent conductive layer, h-BN as tunnel barriers, and different TMDs as the semiconducting materials for the QWs to create unique light-emitting diodes (LEDs), as shown in Figure 6 [31]. In these QWs, electron–hole pairs are injected from the two graphene electrodes into the TMD layers. The long lifetime of the charge carrier in these QWs [32] (as determined by the thickness of the neighboring layers) allows the electrons to recombine with their hole pairs, which in turn causes the QWs to emit photons. The emission wavelength can be adjusted by choosing the appropriate band gap materials, and the efficiency of the LEDs can be enhanced by using multiple QWs (MQWs). Figure 6(a) and (b) shows schematics of heterostructures composed of a single QW (SQW) and of MQWs, along with an optical image of a typical device shown in Figure 6(c). The bright-field STEM images in Figure 6(d) and (e) are from an SQW and from MQW, respectively, indicating a clean and contaminantfree interface between each constituent layer stacked through the peel/lift van der Waals technique previously mentioned. Figure 6(f) displays an optical image incorporating the electroluminescence (EL) response of the device in Figure 6(c). Figure 6(g) and (h) shows a comparison between the photoluminescence (PL) and the EL spectra measured at T = 7 K for SQWs of MoS2 and WS2, respectively. This vertical QW-like heterostructure displayed improved LED performance in several aspects, particularly 1) reduced contact resistance and 2) higher current densities, a llowing brighter EL from the whole device area. These QWs continue to yield 100% in efficiency after months of periodic measurements, clearly demonstrating the robustness of this technology.

INTERLAYER COUPLING IN AN MoS2/WSe2 HETEROSTRUCTURE Heterostructures composed solely of semiconducting TMDs are different from graphene/h-BN/TMD heterostructures in the sense that both compounds have a better lattice match at the interface when compared to those made of graphene, h-BN, and TMDs. Here, we introduce a heterostructure whose interlayer coupling bet ween monolayer WSe 2 and monolayer MoS2 shows promising, controllable optical properties. Javey et al. [33] reported the optical properties of monolayer WSe2/MoS2 heterostructures produced via mechanical exfoliation and a transfer process onto an Si/SiO2 substrate. They observed that a TMD placed

These studies confirm the high quality of the heterostructures produced by stacking 2-D-layered materials. onto a TMD bilayer leads to a lattice mismatch of only 3.8%, the angular alignment of which is controllable to form a moiré pattern very much like graphene on h-BN [17]. Single layers of MoS2 and of WSe2 tend to behave as n-type and p-type direct band-gap semiconductors, with

band gaps of 1.86 and 1.64 eV, respectively. Remarkably, a PL study on a WSe2/MoS2 bilayer shows emission at a wavelength between 1.5 and 1.55 eV, which is lower than that of the constituent single layers, as seen in Figure 7(a). To understand the behavior of the heterostructure, in Figure 7(b) the authors

h-BN GrT

WS2 GrB

h-BN (b)

WS2 Monolayer

PL EL Normalized Counts (a.u.)

GrT

MoS2

PL EL

MoS2 Monolayer

GrB 1.7

h-BN (e)

(d)

(c)

Normalized Counts (a.u.)

(a)

(f)

1.8 1.9 Energy (eV) (g)

2.0

1.8

1.9

2.0

2.1

Energy (eV) (h)

FIGURE 6 (a) A schematic of a single QW heterostructure composed of h-BN/GrB /2h-BN/WS2/2h-BN/GrT /h-BN. (b) A schematic of a multiple QW composed of h-BN/GrB /2h-BN/MoS2/2h-BN/MoS2/2h-BN/MoS2/2h-BN/MoS2/2h-BN/GrT /h-BN. (c) An optical image of an operational QW (h-BN/GrB / 3h-BN/MoS2/3h-BN/GrT /h-BN). 2h-BN and 3h-BN stand for bi- and trilayer h-BN, respectively, and GrT and GrB indicate the graphene top layer and graphene bottom layer, respectively. The scale bar is 10 nm . (d) and (e) Cross-sectional bright-field STEM images of the heterostructures presented in (a) and (b), respectively. The scale bars are 5 nm in (d) and (e). (f) An optical image with EL from the device in (c) at Vb = 2.5 V, T = 300 K. (g) and (h) PL and EL spectra of monolayer MoS 2 and WS 2, respectively. (Images courtesy of [31].) a.u.: arbitrary units.


25

signal is observed in heterobilayers, suggesting that the majority of the photoexcited carriers recombine at the interface. Therefore, the highest luminescent signal would result from an indirect recombination process among the spatially separated carriers. Interestingly, one can tune the interaction between the charge carriers in the WSe2 and in the MoS2 layers by placing a thin layer of h-BN between them, as depicted in Figure 7(d). Figure 7(e) shows the normalized PL of monolayers of WSe2 and of MoS2 and of heterobilayers using either one or three separation layers of h-BN between them. The interaction between carriers in MoS2 and in WSe2 is greatly reduced by the insertion of three atomic layers of h-BN (3h-BN) between them. This can be seen from the PL response of the MoS2/3h-BN/WSe2 heterostructure, which is identical to that of monolayer

Single layers of MoS2 and of WSe2 tend to behave as n-type and p-type direct band-gap semiconductors. excitons relax at the interface, where electrons from the WSe2 conduction band can transfer to the MoS2 conduction band, and holes from the MoS2 valence band can transfer to the WSe2 valence band because of their band offset. The 100-meV energy shift in this heterobilayer is due to the balance between the conduction bands offset between both monolayers versus the diminished exciton binding energy associated with their spatial distribution. A weak luminescent

Normalized PL

oS 2

M e 2/ WS

1.4

SL WSe2 SL MoS2

SL MoS2

1.6 1.8 Photon Energy (eV) (a)

WSe2/MoS2

Normalized Absorbance

W SL

PL Intensity (a.u.)

Se

2

PL Abs

1.4

e–

(2)

EC (3)

(1)

(1) EV h+

SL MoS2

1.6 1.8 Photon Energy (eV) (b)

(2) SL WSe2

E z

(c) N=3

WSe2

h-BN

N Layers

Normalized PL

performed absorption measurements and compared them to the PL spectrum. When comparing the absorption spectra with respect to the normalized PL spectra, the bilayers exhibit a striking shift in absorption, i.e., by ~100 meV. This large Stokes shift in a WSe2/MoS2 heterostructure is consistent with a type-II band alignment [33]. The optical transition presented in Figure 7(c) suggests exciton generation within each layer, and then photoexcited

N=1 N=0 SL WSe2 SL MoS2

MoS2

1.4 (d)

1.6 1.8 Photon Energy (eV) (e)

FIGURE 7 (a) The PL spectra of single-layer (SL) WSe 2 and MoS2 and of the corresponding heterobilayer. (b) The normalized PL (solid lines) and absorbance (dashed lines) spectra of SL WSe 2, SL MoS 2, and the corresponding heterobilayer, with the spectra normalized with respect to the height of the strongest PL/absorbance peak. (c) A band diagram of a WSe 2 /MoS 2 heterobilayer under photoexcitation, depicting (1) absorption and exciton generation in the WSe 2 and MoS 2 SLs, (2) relaxation of excitons at the MoS 2 / WSe 2 interface, driven by the band offset, and (3) r adiative recombination of spatially indirect excitons, respectively. (d) An illustration of the heterostructure composed of SL WSe 2 on SL MoS 2, with few-layer h-BN acting as a spacer. (e) Normalized PL spectra from an SL WSe 2 /MoS 2 heterostructure with N layers of h-BN (N = 0, 1, and 3). (Images courtesy of [33].)

26 |


transistors for low-power applications. Palacios and coworkers [35] reported negative differential resistance (NDR) in FETs composed of multilayers of WSe2/ MoS2. Figure 8(a) shows the schematics of multilayers of WSe2/MoS2 fabricated on a 300-nm SiO2-on-Si substrate. Figure 8(b) displays the FET characteristics of the MoS2 monolayer (blue curve), which shows electron-doped behavior, while the WSe2 monolayer (red curve) displays an ambipolar (n- and p-type conduction) response. The FET characteristics of the MoS2/ WSe2 heterostructure are depicted in Figure 8(c) and reflect the unique behavior measured across the junction region.

WSe2. However, a single atomic layer of h-BN between both layers does not suppress the interlayer interaction between the carriers. These results demonstrate that the interlayer coupling is controllable via the intercalation of dielectric layers, thus providing yet another degree of control over the properties of van der Waals heterostructures.

MoS2/WSe2 HETEROJUNCTION TRANSISTOR Apart from the tunable optical properties of heterostructure-based single layers of WSe2 and of MoS2, a very promising field of application for functional heterostructures is as interband tunneling

10 µ

MoS2 FET

1µ WSe2

In this configuration, an electron moves from the MoS2 layer and has to tunnel toward the WSe2 layer in the overlapping region to be collected by the drain electrode. Three regions characterize these types of WSe2/MoS2 heterostructures with distinctive FET characteristics, as shown in Figure 8(c). The depleted MoS2 layers dominate region I, which corresponds to the off state of the hetero-FET, where the electron path is blocked. However, in region II (−53 V < Vg < −30 V), the current increases and shows a peak at Vb , - 42 V. Thus, region II corresponds to the subthreshold voltage regime for the stack of both materials, where hole conduction increases exponentially while

WSe2 FET

100 n Id (A/µm)

10 n 1n 100 p

WSe2

10 p MoS2

1p MoS2

Vd = 1 V

100 f

–60 –45 –30 –15 (a) 10 µ

MoS2/WSe2

1µ

30

45

60

75

90

Series Mode Vd = 1 V II

III

1n 100 p 10 p

60

1p

30 0 –60

100 f 0

γ = 62 V–1 1n

–30 –90

–60 –45 –30 –15

S = 75 mV/dec Id /Vd (S)

I

10 n

90 gm (pS/µm)

Id (A/µm)

15 (b)

100 n

Hetero-FET

100 n 10 n

0

15

Vg (V) (c)

100 p –60 –45 –30 –15 Vg (V)

0

30

90

45

60

75

Vg = 0.15 V 10 p

–0.8 –0.4 0.0 0.4 Vd (V)

0.8

1.2

(d)

FIGURE 8 (a) A schematic of a WSe2-on-MoS 2 heterostructure. (b) The characteristics of a few-layered WSe2-on-MoS 2 FET under Vd = 1V. The inset shows the optical image of a hetero-FET device. (c) The FET characteristics of the WSe2 /MoS 2 hetero-FET. The inset shows the transconductance of the device, displaying negative differential resistance. (d) The I d as a function of Vd characteristics of the heterojunction. (Images courtesy of [35].)


27

One can tune the interaction between the charge carriers in the WSe2 and in the MoS2 layers by placing a thin layer of h-BN between them. the electron conduction decreases. The transconductance shown in the inset rapidly changes from positive to negative, which occurs when one of the semiconducting layers is in the depletion regime. This phenomenon reflects the resonant tunneling effect, which occurs at matched carrier densities in tunneling FETs [36]. Figure 8(d) shows the absolute value of the conductance (Id/Vd) as a function of the drain voltage Vd for an MoS2/ WSe2 hetero-FET. In the forward bias region, the NDR displays a maximum peak-to-valley ratio of 1.6 as observed at Vg = 0.15 V. The NDR effect has many potential applications, e.g., in electronics as amplifiers converting dc to ac signals in an oscillator and in applications in the microwave spectrum. They can also display the hysteresis and bistable behavior used in switching and in memory circuits [37]. This tunneling diode also shows an average conductance, as extracted from the slope, of S = 75 mV/decade over two decades, very close to the theoretical value of 60 mV/decade, and a maximum cur vature coeff icient of c = ^d 2 I/dV 2 h / ^dI/dV h = 62.2 V−1. The curvature coefficient is an important figure-of-merit parameter when designing high-performance tunneling diodes for applications such as high-frequency detectors [38]. Since the operation of these diodes relies on band-to-band tunneling at the edge of the WSe2/MoS2 junction, thermionic emission does limit the I–V characteristics.

2-D HETEROJUNCTION PHOTOVOLTAICS Generating sufficient energy is one of the major challenges faced by the scientific community as the world’s population increases. Thus, it is imperative to look for alternative sources of energy to 28 |


meet the global demand. Solar energy is at the forefront of this search. Many materials are being explored as alternatives to Si, which is the basis of the solar cell technology currently available on the market. Organic perovskites are one of the most promising fields of research in this area because of the high efficiency observed in solar cells based on these materials [39]. The 2-D layered materials with sem iconduct ing proper t ies ca n be f lexible and nearly transparent when composed of just a few atomic layers. Therefore, they are promising candidates for niche solar cell applications. Their tunable band gap as a function of the number of layers lies in the visible region of the solar spectrum. These compounds can be deposited over large areas through a CVD method or a liquid-phase exfoliation technique from bulk materials, making them potentially viable for solar cell technology. Here, we discuss a few recent developments in solar cells built as heterostructures based on 2-D compounds. Mueller and coworkers [40] presented a detailed study of the photovoltaic effect from p–n junctions built by stacking WSe2 and MoS2 monolayers, for which they observed a maximum photovoltaic power conversion efficiency of 0.2%. In this WSe2/MoS2 heterostructure, photons absorbed in the junction area create electron–hole pairs in both monolayers. The relaxation of photogenerated carriers (electrons) occurs between the conduction band (for electrons) and the valence band (for holes) of the WSe2 and MoS2 layers, driven by type-II band offsets, where electrons and holes remain confined within the MoS2 and the WSe2 layers, respectively. Since the lowest energy electron and hole states are spatially separated, the charge transfer across the

junction occurs between the WSe2 and the MoS2 layers. These carriers then diffuse laterally toward the contacts to generate the photocurrent. A higher photovoltaic power conversion efficiency ( h ) was recently reported for ambipolar MoSe2/h-BN p–n junctions by Memaran and coworkers [41]. The p–n junction consists of ten atomic layers of MoSe 2 transferred onto the h-BN, which is itself placed on a prepatterned dual, bottom-gated configuration. Figure 9(a) shows the optical image of the MoSe2/h-BN heterostructure, with both metallic bottom gates. F igure 9(b) displays the schematic of the dual electrostatic p–n junction solar cell. Figure 9(c) and (d) correspond to data collected under the spectrum of a mercury lamp, while Figure 9(d) and (f) correspond to data collected under a solar simulator using AM1.5 light. Figure 9(c) and (e) shows the I–V characteristics of the p–n junction under several values of the optical power density. Figure 9(d) and (f) includes the electrical output power generated from the p–n junction as extracted from Figure 9(c) and (e), respectively. The maximum electrical power obtained from this ten-layer MoSe2-on-h-BN p–n junction is P el,m ~ 2 nW, for a maximum solar cell efficiency of h ~ 14% [Figure 9(i)]. For comparison, the efficiency of Si solar modules currently available on the market is ~20%. Thus, few-layered TMDs are seen to display a remarkable potential for transparent and flexible solar cells. The current challenge is to translate this efficiency into nonelectrostatically doped vertical p–n junctions.

WSe2/GRAPHENE/MoS2 BROADBAND PHOTOVOLTAIC DETECTORS By using graphene between two TMDs in a heterostructure assembly, one can increase the range of the absorbed wavelengths, i.e., from ultraviolet to infrared. Miao and coworkers [42] reported a high-sensitivity photodetection techn ique ba sed on WSe 2 /g raphene/ MoS2 (p-type/graphene/n-type or PGN) heterostructures, where MoS2 and WSe2 act as the electron and the hole-doped

Vbg2

Vbg1

MoSe2

Vds

h-BN MoSe2 Ti/Au

S

h-BN SiO2

D

5 µm d

Vbg2 S

D

Si

Vbg1 (a) 0.0 W/m2 6.1 kW/m2 41 kW/m2

(b) 0.0 kW/m2 0.6 kW/m2 1 kW/m2

1.0 kW/m2 20 kW/m2 2

5

0.5 kW/m2 0.8 kW/m2

1.0 Hg Lamp

Hg Lamp

0.3

AM-1.5

AM-1.5

1

Pel (nW)

0

Ids (nA)

Pel (nW)

Ids (nA)

0.5 0.0

0.1

–0.5 –5

Vj = 4 V

–1

0 Vds (V)

0 0.0

1

–1.0

0.5 Vds (V)

(c)

0.2

Vj = 4 V –1

0 Vds (V)

(d)

0.0 0.0

1

0.5 Vds (V) (f)

(e)

0.7

0.1

103

104 p (W/m2) (g)

105

Linear Fit

0.1

103

104 p (W/m2) (h)

105

0.1

10 0.5

1

Hg Lamp AM-1.5

103

104 p (W/m2) (i)

FF

Hg Lamp AM-1.5

η (%)

(nW)

1

max

0.8

Pel

1

Voc (V)

Isc (nA)

0.9 Hg Lamp AM-1.5

0.0

105

FIGURE 9 (a) An optical micrograph of an MoSe2 /h-BN p–n junction with dual gates Vbg1 and Vbg2 . (b) Schematics of the p–n junction, indicating the configuration of the measurements. (c) and (d) Drain-to-source current Ids as a function of the bias voltage Vds under several illumination power densities p, under the spectrum of a mercury (Hg) lamp and of a solar simulator (AM-1.5), respectively. (e) and (f) From the curves in (c) and (d), the concomitant photogenerated electrical power Pel = I ds Vds as a function of Vds . Here, and for each curve, Pel was calculated by subtracting the p = 0.0W/m 2 data (black line). The red markers indicate the maximum values of the photogenerated electrical power P elmax. (g) Log–log plot of the short circuit current I sc (brown markers) and semilog plot of the open-circuit voltage Voc (cyan markers) as functions of p. The red line is a linear fit of I ds (p), while the violet line corresponds to a semilogarithmic fit of Voc (p) . (h), P elmax as a function of p, from the red markers in (e) and (f). The red line is a linear fit. (i) The photovoltaic efficiency h (orange markers) and fill factor (FF) (blue markers) as functions of p. For (g), (h), and (i), the solid and open markers indicate values measured under an Hg lamp and under AM-1.5 irradiation, respectively. (Images courtesy of [41].) S: source contacts; D: drain contacts.

materials, respectively. A schematic of the vertical PGN heterostructure is shown in Figure 10(a). Graphene, a gapless material, can absorb photons over a wide spectral range, where the p–n junction formed between WSe 2 and

MoS2 can separate the photogenerated electron–hole pairs in the depletion region through the built-in electric f ield. Figure 10(b) shows the o ptical image of a WSe2/graphene/MoS2 heterostructure with palladium/gold con-

tacts. Figure 10(c) illustrates its Ids as a function of Vds characteristics in the absence of illumination and under several applied gate voltages. As can be seen, it clearly shows a diode-like response or current rectification.


29

3.0 Dark –60 V –40 V –20 V 0V 20 V 40 V 60 V

2.0 +

+

+ e–

–

MoS2

–

–

Graphene

Pd/Au

–

SiO2 Si

(a)

100 80 60 40 20 0 0 5 10 15 20 25 P1 (µ W)

0.6 0.2

–0.2 –1.0

–0.5

5.0 W Dark 0.5 0.4 0.3 0.2 0.1 0.0 –0.1

–1.0

–0.5

0.5 0.0 Vds (V) (c)

1014 104 1012 102 1010

100

0.5

0.0 Vds (V)

10–2

1.0

500

1,000 1,500 Wavelength (nm) (e)

(d) MoS2 (n) Graphene WSe2 (p)

e– e–

Eg 1

e– hv

h+

1.0

106

R (AW–1)

1.0

10 µ W 0.5 µ W

Isc (mA)

Ids (µ A)

1.4

0.0

(b)

Voc (V)

25 µ W 2.5 µ W

1.8

WSe2 Pd/Au

h+

–

1.0

e–

e–

h+ h +

MoS2 (n)

h+

2,000

108 2,500

Graphene WSe2 (p)

e–

e– hv E g2

hv

D∗ (Jones)

+

Ids (nA)

+

h+

hv

h+

e– e–

e–

hv Eg 1

Eg 2

hv

h+

h+ h+

h+ hv > Eg 1 > Eg 2

hv > Eg 2 > Eg 1 (f)

FIGURE 10 (a) A schematic drawing of the PGN heterostructure for photodetection. (b) Top half: an optical image of a fabricated WSe 2 /graphene/ MoS2 heterostructure. Scale bar, 5 nm . Bottom half: a side view of the heterostructure, showing the palladium/gold (Pd/Au) contacts. (c) The I ds −Vds characteristics without illumination, measured under various Vg (from −60 to 60 V). (d) Output I ds −Vds curves measured with exposure to a focused 488-nm laser beam under various laser powers and under Vg = 0 V. The inset shows the extracted short circuit current Î sch and open-circuit voltage ^Voc h versus the incident power intensity. (e) The broadband photoresponsivity (R) (red) and specific detectivity D* (blue) for a typical device under wavelengths ranging from 400 to 2,400 nm. The device was tested under ambient conditions at Vds = 1 V and Vg = 0 V. (f) Left side: schematic band diagrams and the light absorption process for the PGN heterostructure in the ultraviolet and visible regions, in which the photon energy is larger than the band gaps of both MoS 2 Ê g 1h and WSe 2 Ê g 2h. The combination of TMDs and graphene responds to light, producing photogenerated free carriers. Right side: the situation when the wavelength is approaching the infrared region and the photon energy becomes smaller than E g 2. Only graphene responds to light, producing photogenerated free carriers. (Images courtesy of [42].)

Figure 10(d) shows the Ids–Vds cha racterist ics under illumination by a m = 488-nm laser source, and for several values of the optical power under 30 |


Vg = 0 V. The short circuit current (Isc) increases linearly as a function of the applied laser power, as indicated in the inset, whereas the open-circuit voltage

remains nearly constant at Voc ~ 0.23 V. One of the most compelling features of having a PGN heterostructure is the potential for realizing broadband

photodetection by using graphene as the gapless material between the WSe2 and MoS2. The high responsivity occurs within the visible region, where the photon energy is larger than the band gap of both WSe2 and MoS2. Figure 10(f ) (left), presents schematics describing the light absorption process occurring at the PGN heterostructure in the visible region, where photon energy is larger than the band gaps of both WSe 2 and MoS2 . When the wavelength is near the infrared region, or when the energy of the photon energy is smaller than the size of the smaller gap Eg2, interband absorption for each MoS2 and WSe2 monolayer is forbidden. At this wavelength, the graphene layers become the only absorbing material g enerating electron– hole pairs. The charge transfer occurs between graphene and the WSe2/MoS2 layers, as shown in Figure 10(f) (right). This process yields a relatively smaller photoresponse in the infrared region because of the generation of a lower number of electron–hole pairs. Nevertheless, the ultrathin vertical PGN heterojunction plays a crucial role in inducing an eff icient photogain over the entire spectrum. The photoexcited electron–hole pairs in the strong builtin electric f ield separate rapidly and spontaneously. The interlayer inelastic tunneling process occurs because of a lateral momentum mismatch in randomly stacked interfaces. This is beneficial for reducing the interlayer carrier recombination process and ultimately increasing the carrier lifetime and allowing the photogenerated carriers to move to the source/drain contacts, leading to high gain.

CONCLUSION Heterostructures based on the stacking of individual 2-D layered compounds display a remarkable potential for applications as FETs, tunneling transistors, diodes, p–n junctions, and photovoltaic cells, or as complementary logic elements. Despite this potential, remarkable challenges lie ahead to translate this potential into practical applications because of the current lack of high-quality, low-cost, large-area material that is the

basis of any fabrication process. As discussed in this article, most of the recent developments in this area rely on atomic layers mechanically exfoliated from bulk crystals and on an arduous layer-by-layer stacking process, which clearly is not practical for the development of largescale technologies. However, the rapid development of CVD, sputter deposition [43], and MBE methods should allow us to circumvent these difficulties. Another aspect that will require a great focus is the development of better electrodes for contacting 2-D layers, since their inherent lack of dangling bonds makes it quite challenging to achieve ohmic contacts. Current attempts to transform the structural/electronic phase via a chemical method, from semiconducting to a metallic phase at the edges of the semiconducting channel, might open a route to overcome this difficulty. In any case, the very strong interaction with light displayed by certain compounds such as the TMDs, along with our ability to achieve the ultimate miniaturization, i.e., transistors based on a single monolayer that can be effectively turned off, justifies a continued effort in this research area.

ACKNOWLEDGMENTS This work was supported by the U.S. Army Research Office through MURI Grant W911NF-11-1-0362. Luis Balicas acknowledges support from DOEBES through award DE-SC0002613 and Office of Naval Research DURIP grant #11997003. Nihar R. Pradhan also acknowledges the support of the NHMFL, which is supported by the NSF through grant NSF-DMR-1157490 and the state of Florida.

ABOUT THE AUTHORS Nihar R. Pradhan (pradhan@magnet .fsu.edu) earned his M.S. and Ph.D. degrees in physics from Worcester Polytechnic Institute, Massachusetts. He is currently an assistant professor in physics at Jackson State University, Mississippi. He was postdoctoral researcher at National High Magnetic Field Laboratory, Florida State University, Tallahassee. His research interests are in the electrical, optical, thermal, and magnetic properties of two-dimensional

materials, heterostructures, and other low-dimensional compounds; magnetic nanostructures for data storage; and spintronic devices. Saikat Talapatra (stalapatra@physics .siu.edu) earned his Ph.D. degree in engineering science from Southern Illinois University, Carbondale, where he is currently a professor in the Department of Physics. His expertise includes low-temperature and condensed-matter physics, and his present research interests include a broad variety of nanomaterial synthesis and characterization pertaining to advanced energy solutions. Mauricio Terrones (mut11@ psu .edu) earned his Ph.D. degree in chemical physics from the University of Sussex, Falmer, United K ingdom. He is currently a professor of physics, chemistry, and materials science and engineering at Pennsylvania State University, University Park. He is a pioneer in work on carbon nanotubes, graphene, and other two-dimensional materials beyond graphene. He is the director of the Materials Research Institute, Pennsylvania State University. Pulickel M. Ajayan ([email protected]) earned his Ph.D. degree in materials science and engineering from Northwestern University, Evanston, Illinois. He is the Benjamin M. and Mary Greenwood Anderson Professor of Engineering and the founding chair of the Department of Materials Science and NanoEngineering, Rice University, Houston, Texas. He holds joint appointments in the Department of Chemistry and Department of Chemical and Biomolecular Engineering, Rice University. Luis Balicas ([email protected]) earned his master’s and Ph.D. degrees from the University of Paris XI, Orsay, France. He is a research professor in the Physics Department and a Distinguished University Scholar at the National High Magnetic Field Laboratory, Florida State University, Tallahassee. He has spent the last 18 years as a scientist at the National High Magnetic Field Laboratory, working on a range of subjects, from unconventional superconductivity and frustrated/ quantum magnetism to two-dimensional and topological materials. He is a fellow of the American Physical Society and is


31

supported by the U.S. Army Research Office and the U.S. Office of Naval Research for work on two-dimensional materials and their heterostructures.

REFERENCES

[1] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva, and A. A. Firsov, “Electric field effect in atomically thin carbon films,” Science, vol. 306, no. 5696, pp. 666–669, 2004. [2] K. S. Novoselov, “Nobel lecture: Graphene: Materials in the flatland,” Rev. Mod. Phys., vol. 83, no. 3, pp. 837–849, 2011. [3] Y. Zhang, Y. W. Tan, H. L. Stormer, and P. Kim, “Experimental observation of the quantum Hall effect and Berry’s phase in graphene,” Nature, vol. 438, no. 10, pp. 201–204, 2005. [4] L. Wang, I. Meric, P. Y. Huang, Q. Gao, Y. Gao, H. Tran, T. Taniguchi, K. Watanabe, L. M. Campos, D. A. Muller, J. Guo, P. Kim, J. Hone, K. L. Shepard, and C. R. Dean, “One-dimensional electrical contact to a two-dimensional material,” Science, vol. 342, no. 6158, pp. 614–617, 2013. [5] B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti, and A. Kis, “Single-layer MoS2 transistors,” Nature Nanotechnol., vol. 6, no. 3, pp. 147–150, 2011. [6] N. R. Pradhan, D. Rhodes, Q. Zhang, S. Talapatra, M. Terrones, P. M. Ajayan, and L. Balicas, “Intrinsic carrier mobility of multi-layered MoS2 field-effect transistors on SiO2,” Appl. Phys. Lett., vol. 012, no. 12, pp. 123105–123109, 2013. [7] M. Chhowalla, H. S. Shin, G. Eda, L. Li, K. P. Loh, and H. Zhang, “The chemistry of twodimensional layered transition metal dichalcogenide nanosheets,” Nature Chem., vol. 5, no. 4, pp. 263–275, 2013. [8] N. R. Pradhan, D. Rhodes, Y. Xin, S. Memaran, L. Bhaskaran, M. Shiddiq, S. Hill, P. M. Ajayan, and L. Balicas, “Ambipolar molybdenum diselenide field-effect transistors: Field-effect and Hall mobilities,” ACS Nano, vol. 8, no. 8, pp. 7923–7929, 2014. [9] A. L. Elías, N. P. López, A. C. Beltrán, A. Berkdemir, R. Lv, S. Feng, A. D. Long, T. Hayashi, Y. A. Kim, M. Endo, H. R. Gutiérrez, N. R. Pradhan, L. Balicas, T. E. Mallouk, F. López-Urías, H. Terrones, and M. Terrones, “Controlled synthesis and transfer of large-area WS2 sheets: From single layer to few layers,” ACS Nano, vol. 7, no. 6, pp. 5235–5242, 2013. [10] N. R. Pradhan, D. Rhodes, S. Memaran, J. M. Poumirol, D. Smirnov, S. Talapatra, S. Feng, N. Perea Lopez, A. L. Elias, M. Terrones, P. M. Ajayan, and L. Balicas, “Hall and field-effect mobilities in few layered p-WSe2 field effect transistors,” Scientific Reports, vol. 5, no. 8979, 2015, pp. 8979-1 to 8979-8. [11] N. R. Pradhan, A. McCreary, D. Rhodes, Z. Lu, S. Feng, E. Manousakis, D. Smirnov, R. Namburu, M. Dubey, A. R. H. Walker, H. Terrones, M. Terrones, V. Dobrosavljevic, and L. Balicas, “Metal to insulator quantum phase transition in few-layered ReS2,” Nano Lett., vol. 15, no. 12, pp. 8377–8384, 2015. [12] X. Xi, L. Zhao, Z. Wang, H. Berger, L. Forró, J. Shan, and K. F. Mak, “Strongly enhanced chargedensity-wave order in monolayer NbSe2,” Nature Nanotechnol., vol. 10, no. 9, pp. 765–770, 2015. [13] H. Liu, A. T. Neal, Z. Zhu, Z. Luo, X. Xu, D. Tománek, and P. D. Ye, “Phosphorene: An unexplored 2D semiconductor with a high hole mobility,” ACS Nano, vol. 8, no. 4, pp. 4033–4041, 2014. [14] H. L. Zhuang and R. G. Hennig, “Single-layer group-III monochalcogenide photocatalysts for water splitting,” Chem. Materials, vol. 25, no. 15, pp. 3232–3238, 2013.

32 |


[15] J. Dai, M. Li, and X. C. Zeng, “Group IVB transition metal trichalcogenides: A new class of 2D layered materials beyond graphene,” WIREs Computational Mol. Sci., vol. 6, no. 2, pp. 211–222, 2016. [16] K. Du, X. Wang, Y. Liu, P. Hu, M. I. Bakti Utama, C. K. Gan, Q. Xiong, and C. K loc, “Weak van der Waals stacking, wide-range band gap, and Raman study on ultrathin layers of metal phosphorus trichalcogenides,” ACS Nano, vol. 10, no. 2, pp. 1738–1743, 2016. [17] C. R. Dean, A. F. Young, I. Meric, C. Lee, L. Wang, S. Sorgenfrei, K. Watanabe, T. Taniguchi, P. Kim, K. L. Shepard, and J. Hone, “Boron nitride substrates for high-quality graphene electronics,” Nature Nanotechnol., vol. 5, no. 10, pp. 722–726, 2010. [18] Y. Gong, S. Lei, G. Ye, B. Li, Y. He, K. Keyshar, X. Zhang, Q. Wang, J. Lou, Z. Liu, R. Vajtai, W. Zhou, and P. M. Ajayan, “Two-step growth of twodimensional WSe2/MoSe2 heterostructures,” Nano Lett., vol. 15, no. 9, pp. 6135–6141, 2015. [19] X. Yuan, L. Tang, S. Liu, P. Wang, Z. Chen, C. Zhang, Y. Liu, W. Wang, Y. Zou, C. Liu, N. Guo, J. Zou, P. Zhou, W. Hu, and F. Xiu, “Arrayed van der Waals vertical heterostructures based on 2D GaSe grown by molecular beam epitaxy,” Nano Lett., vol. 15, no. 5, pp. 3571–3577, 2015. [20] L. Britnell, R. V. Gorbachev, R. Jalil, B. D. Belle, F. Schedin, A. Mischenko, T. Georgiou, M. I. Katsnelson, L. Eaves, S. V. Morozov, N. M. R. Peres, J. Leist, A. K. Geim, K. S. Novoselov, and L. A. Ponomarenko, “Field-effect tunneling transistor based on vertical graphene heterostructures,” Science, vol. 335, no. 6071, pp. 947–950, 2012. [21] A. K. Geim and I. V. Grigorieva, “Van der Waals heterostructures,” Nature, vol. 499, no. 7459, pp. 419–425, 2013. [22] T. Georgiou, R. Jalil, B. D. Belle, L. Britnell, R. V. Gorbachev, S. V. Morozov, Y.-J. Kim, A. Gholinia, S. J. Haigh, O. Makarovsky, L. Eaves, L. A. Ponomarenko, A. K. Geim, K. S. Novoselov, and A. Mishchenko, “Vertical field-effect transistor based on graphene–WS2 heterostructures for flexible and transparent electronics,” Nature Nanotechnol., vol. 8, no. 2, pp. 100–103, 2013. [23] H. Huang, “Ultrathin two dimensional nanomaterials,” ACS Nano, vol. 9, no. 10, pp. 9451– 9469, 2015. [24] L. Wang, I. Meric, P. Y. Huang, Q. Gao, Y. Gao, H. Tran, T. Taniguchi, K. Watanabe, L. M. Campos, D. A. Muller, J. Guo, P. Kim, J. Hone, K. L. Shepard, and C. R. Dean, “One-dimensional electrical contact to a two-dimensional material,” Science, vol. 342, no. 6158, pp. 614–617, 2010. [25] C. R. Dean, L. Wang, P. Maher, C. Forsythe, F. Ghahari, Y. Gao, J. Katoch, M. Ishigami, P. Moon, M. Koshino, T. Taniguchi, K. Watanabe, K. L. Shepard, J. Hone, and P. K im, “Hofstadter’s butterfly and the fractal quantum Hall effect in moiré superlattices,” Nature, vol. 497, no. 7451, pp. 598–602, 2013. [26] L. Wang, Y. Gao, B. Wen, Z. Han, T. Taniguchi, K. Watanabe, M. Koshino, J. Hone, and C. R. Dean, “Evidence for a fractional fractal quantum Hall effect in graphene superlattices,” Science, vol. 350, no. 6265, pp. 1231–1234, 2015. [27] D. Hofstadter, “Energy levels and wave functions of Bloch electrons in rational and irrational magnetic fields,” Phys. Rev. B, vol. 14, no. 6, pp. 2239–2249, 1976. [28] M. K. L. Man, S. D. Jones, A. Winchester, G. Shi, G. Gupta, A. D. Mohite, S. Kar, E. Kioupakis, S. Talapatra, and K. M. Dani, “Protecting the properties of monolayer MoS2 on silicon based substrates, with an atomically thin buffer,” Scientific Reports, vol. 6, pp. 20890-1 to 20890-9 p. 20890, Feb. 2016. [29] X. Cui, G. H. Lee, Y. D. Kim, G. Arefe, P. Y. Huang, C. H. Lee, D. A. Chenet, X. Zhang, L. Wang, F. Ye, F. Pizzocchero, B. S. Jessen, K. Watanabe, T. Taniguchi, D. A. Muller, T. Low, P. Kim,

and J. Hone, “Multi-terminal transport measurements of MoS2 using a van der Waals heterostructure device platform,” Nature Nanotechnol., vol. 10, no. 6, pp. 534–540, 2015. [30] B. Fallahazad, H. C. P. Movva, K. Kim, S. Larentis, T. Taniguchi, K. Watanabe, S. K. Banerjee, and E. Tutuc, “Shubnikov–de Haas oscillations of high-mobility holes in monolayer and bilayer WSe2: Landau level degeneracy, effective mass, and negative compressibility,” PRL, vol. 116, no. 8, p. 086601, 2016. [31] F. Withers, O. D. Pozo-Zamudio, A. Mishchenko, A. P. Rooney, A. Gholinia, K. Watanabe, T. Taniguchi, S. J. Haigh, A. K. Geim, A. I. Tartakovskii, and K. S. Novoselov, “Light-emitting diodes by band-structure engineering in van derWaals heterostructures,” Nature Materials, vol. 14, no. 3, pp. 301–306, 2015. [32] H. Wang, C. Zhang, and F. Rana, “Surface recombination limited lifetime of photoexcited carriers in few-layered transition metal dichalcogenide MoS2,” Nano Lett., vol. 15, no. 12, pp. 8204–8210, 2015. [33] H. Fang, C. Battaglia, C. Carraro, S. Nemsak, B. Ozdol, J. S. Kang, H. A. Bechtel, S. B. Desai, F. Kronasth, A. A. Unalh, G. Conti, C. Conlon, G. K. Palsson, M. C. Marting, A. M. Minore, C. S. Fadley, E. Yablonovitcha, R. Maboudian, and A. Javey, “Strong interlayer coupling in van der Waals heterostructures built from single-layer chalcogenides,” Proc. Nat. Acad. Sci., vol. 111, no. 17, pp. 6198–6220, 2014. [34] B. A. Wilson, “Carrier dynamics and recombination mechanisms in staggered alignment heterostructures,” IEEE J. Quantum Electron., vol. 24, no. 8, pp. 1763–1777, 1988. [35] A. Nourbakhsh, A. Zubair, M. S. Dresselhaus, and T. Palacios, “Transport properties of a MoS2/ WSe2 heterojunction transistor and its potential for application,” Nano Lett., vol. 16, no. 2, pp. 1359–1366, 2016. [36] F. Capasso, S. Sen, and A. Y. Cho, “Negative transconductance resonant tunneling f ieldeffect transistor,” Appl. Phys. Lett., vol. 51, no. 7, pp. 526–528, 1987. [37] B. Beneking, High Speed Semiconductor Devices: Circuit Aspects and Fundamental Behaviour. Berlin: Springer-Verlag, 1994, pp. 114–117. [38] S. Agarwal and E. Yablonovitch, “Band-edge steepness obtained from Esaki/backward diode current-voltage characteristics,” IEEE Trans. Electron Devices, vol. 61, no. 5, pp. 1488−1493, 2014. [39] M. Liu, M. B. Johnston, and H. J. Snaith, “Efficient planar heterojunction perovskite solar cells by vapour deposition,” Nature, vol. 501, no. 7467, pp. 395–402, 2013. [40] M. M. Furchi, A. Pospischil, F. Libisch, J. Burgdörfer, and T. Mueller, “Photovoltaic effect in an electrically tunable van der Waals heterojunction,” Nano Lett., vol. 14, no. 8, pp. 4785–4791, 2014. [41] S. Memaran, N. R. Pradhan, Z. Lu, D. Rhodes, J. Ludwig, Q. Zhou, O. Ogunsolu, P. M. Ajayan, D. Smirnov, A. I. Fernández-Domínguez, F. J. G. Vidal, and L. Balicas, “Pronounced photovoltaic response from multilayered transition-metal dichalcogenides PN-junctions,” Nano Lett., vol. 15, no. 11, pp. 7532–7538, 2015. [42] M. Long, E. Liu, P. Wang, A. Gao, H. Xia, W. Luo, B. Wang, J. Zeng, Y. Fu, K. Xu, W. Zhou, Y. Lv, S. Yao, M. Lu, Y. Chen, Z. Ni, Y. You, X. Zhang, S. Qin, Y. Shi, W. Hu, D. Xing, and F. Miao, “Broadband photovoltaic detectors based on an atomically thin heterostructure,” Nano Lett., vol. 16, no. 4, pp. 2254–2259, 2016. [43] H. Samassekou, A. Alkabsh, E. Miller, M. Wasala, A. Walker, S. Talapatra, T. Jayasekara, and D. Mazumdar. (2016, May 24). Viable route towards large-area two dimensional MoS2 using magnetron sputtering. Cornell Univ. Library. [Online]. Available: https://arxiv.org/abs/1605.07568#