MoS2 Heterojunction Field Effect

Jun 21, 2019 - Piezoelectricity of transition metal dichalcogenides (TMDs) under mechanical strain has been theoretically and experimentally studied...
0 downloads 0 Views 4MB Size
Download PDF
www.acsnano.org

Ultrahigh Gauge Factor in Graphene/MoS2 Heterojunction Field Effect Transistor with Variable Schottky Barrier Ilmin Lee,† Won Tae Kang,†,‡ Yong Seon Shin,†,‡ Young Rae Kim,† Ui Yeon Won,† Kunnyun Kim,⊥ Dinh Loc Duong,‡ Kiyoung Lee,§ Jinseong Heo,§ Young Hee Lee,‡,∥ and Woo Jong Yu*,† Downloaded via BUFFALO STATE on July 18, 2019 at 11:26:50 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

Department of Electrical and Computer Engineering, Sungkyunkwan University, Suwon 16419, Republic of Korea Center for Integrated Nanostructure Physics, Institute for Basic Science (IBS), Suwon 16419, Republic of Korea ⊥ Korea Electronics Technology Institute, Seongnam 13509, Republic of Korea § Samsung Advanced Institute of Technology, Suwon-si, Gyeonggi-do 443−803, Republic of Korea ∥ Department of Energy Science, Sungkyunkwan University, Suwon 16419, Republic of Korea ‡

S Supporting Information *

ABSTRACT: Piezoelectricity of transition metal dichalcogenides (TMDs) under mechanical strain has been theoretically and experimentally studied. Powerful strain sensors using Schottky barrier variation in TMD/metal junctions as a result of the strain-induced lattice distortion and associated ioncharge polarization were demonstrated. However, the nearly fixed work function of metal electrodes limits the variation range of a Schottky barrier. We demonstrate a highly sensitive strain sensor using a variable Schottky barrier in a MoS2/ graphene heterostructure field effect transistor (FET). The low density of states near the Dirac point in graphene allows large modulation of the graphene Fermi level and corresponding Schottky barrier in a MoS2/graphene junction by straininduced polarized charges of MoS2. Our theoretical simulations and temperature-dependent electrical measurements show that the Schottky barrier change is maximized by placing the Fermi level of the graphene at the charge neutral (Dirac) point by applying gate voltage. As a result, the maximum Schottky barrier change (ΔΦSB) and corresponding current change ratio under 0.17% strain reach 118 meV and 978, respectively, resulting in an ultrahigh gauge factor of 575 294, which is approximately 500 times higher than that of metal/TMD junction strain sensors (1160) and 140 times higher than the conventional strain sensors (4036). The ultrahigh sensitivity of graphene/MoS2 heterostructure FETs can be developed for next-generation electronic and mechanical−electronic devices. KEYWORDS: graphene, molybdenum disulfide, van der Waals heterostructure, strain, Schottky barrier height nearly fixed work function of metal electrodes limits the variation range of a Schottky barrier and corresponding gauge factor.4,7,9 Meanwhile, van der Waals heterostructures (vdWHs) based on stacked 2D materials have been proposed for the realization of various devices with flat formats, such as planar transistors, vertical tunneling field-effect transistor (FETs), photodetectors, light-emitting diodes (LEDs), and memory devices.14−33 The weak electrostatic screening effect of a graphene electrode allows adjustment of the Schottky barrier at the interface between graphene and TMD, enabling the fabrication of vertical FETs with ultrahigh on-current densities (over 103 A

T

wo-dimensional (2D) layered materials, such as graphene, transition metal dichalcogenides (TMDs), and hexagonal boron nitride (h-BN), have attracted considerable interest as next-generation materials because of their electrical and optical properties.1−6 In particular, TMDs demonstrate piezoelectricity as a result of the strain-induced lattice distortion and associated ion-charge polarization.7−11 Specifically, monolayer TMDs show strong piezoelectricity because of the disappearance of the opposite orientations of adjacent atomic layers in the multilayer TMDs.7 The piezoelectric characteristics of TMDs have been studied for potential applications in strain sensors, transducers, and nanogenerators.7−13 In these devices, strain-induced charges in TMD modulate the Schottky barrier (ΦSB) at the TMD/ metal electrode junction, resulting in the oscillation of piezoelectric voltage and current outputs.7 However, the © XXXX American Chemical Society

Received: May 23, 2019 Accepted: June 21, 2019 Published: June 21, 2019 A

DOI: 10.1021/acsnano.9b03993 ACS Nano XXXX, XXX, XXX−XXX

Article

Cite This: ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 1. (a, b) Schematic energy band diagram of (a) metal/MoS2 junction FET and (b) graphene/MoS2 junction FET with (blue line) and without mechanical strain (black line). (c) Schematic of the fabrication process of a flexible and transparent graphene/MoS2 vdWH FET. (d) Optical image of the flexible and transparent graphene/MoS2 vdWH FET.

cm−2) and vdWH photodetectors with high external quantum efficiencies (∼55%).25 Furthermore, efficient modulation of a Schottky barrier of graphene/TMD junctions14 yields higher carrier mobility than that of metal/TMD junctions with fixed Schottky barrier. Interesting behavior of a negative Schottky barrier was also observed in MoS2/graphene contact, which results in ohmic contact for high on-current density.14 In this work, we demonstrate an ultrahigh gauge factor of a graphene/MoS2 FET using an efficiently variable Schottky barrier in graphene/MoS2 junction.14,34,35 Unlike conventional metal/TMD junction FETs with nearly fixed work function of metal electrodes, which limits the adjustment range of the Schottky barrier by strain-induced ion-charge polarization of MoS2,7−11,14,36 the Schottky barrier in our graphene/MoS2 vdWHs FET is largely adjustable because of efficient modulation of the graphene Fermi level near the low density of states of the Dirac point. By application of mechanical strain, the strain-induced piezoelectric ion charges of MoS2 move the Fermi level of graphene and corresponding Schottky barrier in the graphene/MoS2 junction. Our theoretical simulations and measurements showed that the Schottky barrier was not sufficiently changed (ΔΦSB = 13 meV) by strain-induced polarized charges at large negative and large positive gate voltages when EF of graphene was located far from the Dirac point where the density of states is relatively high. However, the Schottky barrier changed dramatically (ΔΦSB = 118 meV) at −0.15 V gate bias when EF was located near the Dirac point in graphene because of a much lower density of states. As a result, the gauge factor was 94.6 at Vg − ΔVth = 2.6 V and significantly increased to 575 294 with the decrease of Vg − ΔVth to −0.15 V. Thus, the gauge factor of our device is approximately 500 times higher than that of metal/TMD junction strain sensors (1160)9 and 140 times higher than the highest known gauge factor of conventional strain sensors (4036).50

RESULTS Figure 1c,d show a schematic and optical image of a graphene/ MoS2 vdWH FET on a flexible plastic substrate. The details of fabrication steps are presented in Supporting Information S1. In brief, our device consists of patterned large-area graphene synthesized by chemical vapor deposition (CVD) for the source, drain, and gate, an exfoliated monolayer MoS2 flake (Figure S2) for the piezoelectric channel material, and an exfoliated h-BN flake for the gate insulator. Staking of twodimensional layers was performed by the well-known drytransfer method.2 To demonstrate the effect of strain application on the graphene/MoS2 vdWH FET, the entire FET was transferred onto a flexible polyethylene terephthalate (PET) substrate by poly(methyl methacrylate) (PMMA) binder, and then the underlying SiO2 sacrificial layer was removed by immersing it in the buffer oxide etcher (BOE). It is noteworthy that the MoS2 layer contacted the graphene electrodes in two ways: at its zigzag end edges, where the piezoelectric polarization charges are distributed, and at the top surface of the basal plane of MoS2, which has no polarization charges. Although the polarization charges are induced at the zigzag edge interface, they are still able to affect the electrical transport across the whole Schottky contact area formed between MoS2 and graphene (Figure S11), similar to traditional MoS2/metal or ZnO/metal junctions.7,37 We investigated the modulation of the Schottky barrier at the graphene/MoS2 contact under mechanical strain. Typical modulation of the Schottky barrier at the graphene/MoS2 contact by an external gate field is shown in Figure S3. Although the mechanically exfoliated monolayer MoS2 flake has randomly shaped edges, the large electrodes cover the polar edges and produce an efficient piezoelectric response. The carrier transport in MoS2 is largely dominated by the piezoelectric effect, in which the strain-induced polarization of MoS2 modulates the Schottky barrier.7,9 The I−V characterB

DOI: 10.1021/acsnano.9b03993 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 2. (a) Output characteristics of a graphene/MoS2 vdWH FET under mechanical strain. Inset image shows a designed apparatus to measure temperature-dependent I−V characteristics under tensile strain. (b) Transfer characteristics of a graphene/MoS2 vdWH FET under mechanical strain. (c−e) Arrhenius plots of ln(Isat/T3/2) vs q/(kBT) at various gate voltages for a graphene/MoS2 vdWH FET under (c) 0%, (d) 0.07%, and (e) 0.17% strain. (f) Variation of Schottky barrier height with gate voltage under 0%, 0.07%, and 0.17% strain. (g−i) Color maps of temperature-dependent transfer characteristics under (g) 0%, (h) 0.07%, and (i) 0.17% strain.

derived Schottky barriers at different strains. The Schottky barrier decreases with the increase of the mechanical strain, which is largely attributed to the upshift of graphene EF by positive charge polarization of strain-induced MoS2.7,9 The reduction of the Schottky barrier by applying strain also results in a threshold voltage (Vth) shift to a low gate voltage (Figure 2g−i). The derived Schottky barrier near zero gate voltage (Vg = 1 V) effectively decreases from 170 to 130 meV (ΔΦSB = 40 meV) with increasing strain from 0% to 0.17%, while the Schottky barrier at high positive voltage (Vg = 3 V) hardly changes from 35 to 20 meV (ΔΦSB = 15 meV) within the same strain variation (Figure 2f). To explain this phenomenon, we simulated the energy band diagram by solving Poisson’s equation using MATLAB (Figure 3, Supporting Information S6). In brief, we first calculate the Gr EF (source) in a Gr (gate)/h-BN/Gr (source) vertical junction and MoS2 EF in a Gr (gate)/h-BN/MoS2 vertical junction. Based on the obtained Gr EF (source) and MoS2 EF, we simulate the band depletion at the Gr (source)−MoS2 contact in the planar direction. Figure 3a shows the simulated Schottky barrier along the gate voltage under 0% (black line) and 0.17% (blue line) strain. The Schottky barrier curve under 0.17% strain is obtained from the chemical potential change in graphene ( μc = ℏvF π |n piezo| )41 by piezoelectric polarization of MoS2

istics of our graphene/MoS2 vdWHs FET on a plastic substrate show that the current consistently increased with increasing strain (Figure 2a,b). It is noteworthy that the switch of the triode to saturation in Ids−Vds (lower slope at high Vds) is caused by “pinch-off”.5,38,39 The increase of pinch-off voltage in the higher current curve can be explained by the triode 1 W saturation transition equation: Ids = 2 k′ L Vds 2 , where k′ is the transconductance parameter, W is the channel width, and L is the channel length. The equation of the Ids−Vgs curve is 1 W Ids = 2 k′ L (Vgs − Vt )2 ; therefore, Ids should be proportional to Vgs2. To estimate the variation of the Schottky barrier at the MoS2/graphene contact under the mechanical strain, we measured the temperature-dependent I−V characteristics of the vdWHs FET (Figure S5) using a designed apparatus consisting of a thermally conductive copper substrate with curvature radii Rc of 7 cm (ε = 0.07%) and 3 cm (ε = 0.17%) (inset of Figure 2a). The inset of Figure 2b shows a schematic of uniaxial stretching of the graphene/MoS2 vdWH FET when the substrate was bent. The strain was obtained by the formula Strain = 100 ×

(Thickness of PET + Thickness of Sample) , 2R c 40

where Rc is

the radius of the substrate curvature. The strain-induced Schottky barrier was calculated from the slope of the plots of ln(Isat/T3/2) vs q/(kBT) (Figure 2c−e). Figure 2f shows the C

DOI: 10.1021/acsnano.9b03993 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 3. (a) Theoretically simulated Schottky barrier as a function of the gate voltage under 0% (black line) and 0.17% strain (blue line). DP denotes the Dirac point of graphene. (b) Difference of Schottky barriers between 0% and 0.17% strain in (a). (c, d) Simulated and schematic energy band diagram of a graphene/MoS2 vdWH FET under 0% (black line) and 0.17% strain (blue line) (c) at Vgs = 3 V and (d) Vgs = −0.15 V.

Figure 4. | (a) Simulated (Ids vs Vg) and measured (Ids vs Vg− ΔVth) transfer curves of a graphene/MoS2 vdWH FET under 0% and 0.17% strain. DP is the Dirac point of graphene. (b) Simulated and measured ΔI(0.17%)/I(0%) ratio and gauge factor ([ΔI(0.17%)/I(0%)]/Δε) of a graphene/MoS2 vdWH FET. (c, d) Color map of ΔI(0.17%)/I(0%) and gauge factor of a graphene/MoS2 vdWH FET depending on Vgs and Vds.

(0.3 nm) between graphene and MoS2,42,43 respectively. In Figure 3a, the Schottky barrier hardly changes at large negative and large positive gate voltages when EF of graphene is located far from the Dirac point (DP in Figure 3a) where the density of states is relatively high. However, the Schottky barrier changes dramatically near 0 V gate bias when EF is located near the Dirac point in graphene because of a much lower density of states. Figure 3b shows the difference of Schottky barrier between 0% and 0.17% strain in Figure 3a. The Schottky

and the corresponding Schottky barrier change (Figure 3b), where the ℏ is Plank’s constant, vF is the graphene Fermi velocity, and npiezo is induced charges by piezoelectric polarization of MoS2. The npiezo is obtained from the graphene/MoS 2 capacitor with an interlayer air gap (n piezo =

εrε0Vpiezo qtgap

), where εr, ε0, Vpiezo, q, and tgap are the

relative permittivity of air, absolute permittivity, piezoelectric voltage in MoS2,7 charge of an electron, and interlayer distance D

DOI: 10.1021/acsnano.9b03993 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano barrier difference is highest (118 meV) at −0.15 V gate voltage and decreases with the increase of the negative and positive gate voltage (13 meV at VG = 3 V). Figure 3c and d show the energy band diagram at 3 and −0.15 V gate voltages, respectively. The same mechanism is responsible for the Schottky barrier modulation by strain-induced polarization; namely, the ion-charge polarization of MoS2 hardly changes the Schottky barrier at large negative and large positive gate voltages when EF of graphene is located far from the Dirac point (Figure 3c), while the ion-charge polarization of MoS2 highly changes the Schottky barrier at −0.15 V gate bias when EF is located near the Dirac point in graphene (Figure 3d). It is noted that the Schottky barrier change by downward or upward bending of the MoS2 energy band is extracted as ∼20 meV at 0.17% strain in a previous report.9 The band bending of MoS2 is not a function of gate voltage; therefore, the Schottky barrier change by band bending of MoS2 is maintained at 20 meV in overall gate voltage. At the high gate voltage region (Vg = 3 or −3 V), both band bending in MoS2 (20 meV) and Fermi level shift in graphene (21 meV) by piezoelectric ion charges effect the Schottky barrier change similarly. At the near-zero gate voltage (Vg − ΔVth = −0.15 V), however, the Fermi level shift in graphene (152 meV) more dominantly affects the Schottky barrier change than band bending in MoS2 (20 meV). We simulated the current by the equation

Table 1. Performance of MoS2-Based Strain Sensors gate voltage

|gauge factor|

ref

0V 0V 0V 0V 0V 0V 0V 0V 0V 20 −0.15 V −0.15 V

∼5 ∼200 >20 4036 14 1000 230 243 1160 40 575 294 (flake MoS2) 92 064 (CVD MoS2)

48 49 50 51 52 53 7 8 9 11 our work

measured ΔI(0.17%)/I(0%) and gauge factor are lower than those of the simulation in this state (Vg − ΔVth > 0.5 V in Figure 4b). In the off-state (Vg − ΔVth < 0.5 V in Figure 4b), low electron carrier density in MoS2 (Figure S7) does not screen the positive piezoelectric charges at the source contact, resulting in high Schottky barrier modulation and the following high current modulation. The measured ΔI(0.17%)/I(0%) and gauge factor are well matched in this state (Vg − ΔVth < 0.5 V in Figure 4b). From the CVD graphene/MoS2 vdWH FET, a similar screening is observed (Figure S8). Under the tensile strain (black curve in Figure S8), the high density of electrons in n-type MoS2 at high positive gate voltage screens positive piezoelectricity. The measured ΔI/I0 is lower than that of simulation in this state (black curve at Vg − ΔVth > 1 V in Figure S8). Under the compressive strain (red curve in Figure S8), on the other hand, the electrons are not affecting the negative piezoelectric charge at the source contact, resulting in no reduction of the Schottky barrier modulation and following current modulation in overall gate voltage. The measured ΔI/ I0 is matched with that of the simulation in overall gate voltage. Piezoelectricity arises from the broken inversion symmetry in the atomic structure by mechanical deformation of the hexagonal structure of MoS2. The change of the orientation will change the deformation direction of the hexagonal structure of MoS2, resulting in different strengths of the piezoelectricity. The piezoelectric coupling is known to be highest at 0° and 60° (armchair direction) and lowest at 30° (zigzag direction) along the Mo→S bond direction.10 To investigate the piezoelectric effect of the graphene/MoS2 vdWH FET sensor depending on the crystal orientation of MoS2, we fabricated a strain sensor using triangular CVDgrown MoS2. The triangular shape of the CVD-grown monolayer MoS2 allows controlling the crystal orientation. We carefully connected the lateral electrodes on the triangular monolayer MoS2 to create a “zigzag direction” (Mo and S in the same line) and “armchair direction” (Mo and S parallel) along the mechanical strain direction (Figure 5a and b, another device in Figure S9). Transfer characteristics were measured by applying mechanical strain (Figure 5c,d). The gauge factors of 40 172 at 0.07% strain and 92 064 at 0.17% strain were achieved with the armchair MoS2; however, relatively small gauge factors of 10 046 at 0.07% strain and 11 481 at 0.17% strain were observed with the zigzag MoS2. This result suggests that manipulation of the MoS2 atomic orientation along an armchair MoS2 significantly improves the piezoelectric effect by efficient polarization in the strained MoS2 crystal structure. On the other hand, the Mo position coincides with S atoms in

−qϕB

( ), based on the simulated Schottky

Isat = AA*T 3/2exp

material conventional metal single crystal silicon silver nanowire ZnO graphene CNT Pd/MoS2 Au/Cr/MoS2 Au/MoS2 Au/MoS2 graphene/MoS2

kBT

barrier under 0% and 0.17% strain (dashed line in Figure 4a) and compared it with the measured current in a graphene/ MoS2 vdWH FET (dots in Figure 4a). The current is weakly modulated at large negative and large positive gate voltages when the Schottky barrier hardly changes and is highly modulated at −0.15 V gate bias when the Schottky barrier highly changes. It should be noted that the discrepancy between the simulation and measurements below 10−14 A was caused by the resolution limit of the source-measure unit. We calculated the gauge factor of our graphene/MoS2 vdWH FET based on the simulation (dashed line in Figure 4b) and measurements (dots in Figure 4b); the gauge factor is defined as [ΔI(ε)/I(0)]/Δε, where ΔI is the current change along the mechanical strain, I(0) is the unstrained current of the device, and ε is the applied strain. The gauge factor was calculated as 94.6 at a gate voltage of Vg − ΔVth = 2.6 V with a small ΔI(0.17%)/I(0%) of 1.16. ΔI(0.17%)/I(0%) and the corresponding gauge factor dramatically increase with the decrease of the gate voltage and reach their maxima of 978 and 575 294 at Vg − ΔVth = −0.15 V, respectively. The maximum theoretical and experimental gauge factors are 645 294 and 575 294, respectively. It is expected that ΔI(0.17%)/I(0%) and the gauge factor can further increase by increasing Vds, as measured in another device (Figure 4c,d). The gauge factor of our device is approximately 500 times higher than that of metal/TMD junction strain sensors (1160) and 140 times higher than the highest known gauge factor of conventional strain sensors (4036) (Table 1). The screening effect of the piezoelectric effect depending on the gate-controlled carrier concentration in MoS2 has been shown in our device. As a result, at on-state (Vg − ΔVth > 0.5 V in Figure 4b), high electron carrier density in MoS2 (Figure S7) screens the positive piezoelectric charges at the source contact, resulting in reduction of the Schottky barrier modulation and the following current modulation. The E

DOI: 10.1021/acsnano.9b03993 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 5. (a, b) Optical image of a graphene/MoS2 vdWH FET with CVD-grown MoS2. (a) MoS2 connects graphene electrodes along the zigzag direction. (b) MoS2 connects graphene electrodes along the armchair direction. (c, d) Transfer characteristics of (c) the zigzag direction of MoS2 and (d) the armchair direction of MoS2 under various strains.

Figure 6. (a) PL spectra of monolayer MoS2 under 0%, 0.07%, and 0.17% strain. Inset shows a photograph of the PL measurement system with monolayer MoS2 on the bending apparatus. (b) First-principles electronic structure of the monolayer MoS2 at 0%, 0.15%, and 0.4% strain. (c) Variation of band gap of monolayer MoS2 with mechanical strain. (d) Transfer (I−Vgs) characteristic of the graphene/MoS2 vdWH FETs under no strain, tensile strain, and compressive strain. (e) Schematic energy band diagram of a graphene/MoS2 vdWH FET under no strain (black line), tensile strain (red line), and compressive strain (blue line). (f) Variation of transfer characteristics of the graphene/MoS2 vdWH FETs on the strain and unstrained repetitions. Inset shows the current variation of the graphene/MoS2 vdWH FETs at Vgs = −0.7 V.

because of the high crystallinity in exfoliated MoS2.44 A large number of intrinsic point defects (in particular, sulfur vacancies) in CVD-synthesized monolayer MoS2 plays a critical role in screening piezoelectric potentials generated in piezoelectric materials under mechanical deformation. As a result, high piezoelectric polarization in single-crystalline MoS2 results in a greater Vth shift of 0.37 V in transfer characteristics,

the strained zigzag MoS2, resulting in negligible polarization in the strained MoS2 crystal structure. The zigzag MoS2 still has a piezoelectric effect, which may be attributed to slight misalignment between the metal electrodes/MoS2 or bending direction/MoS2. It is noteworthy that the piezoelectric characteristic in an exfoliated single-crystalline MoS2 monolayer is higher than that of a CVD-grown MoS2 monolayer F

DOI: 10.1021/acsnano.9b03993 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano while low piezoelectric polarization in CVD MoS2 shifts Vth only by 0.3 V in transfer characteristics (Figure S10 and Table S1). The strain-dependent current change in MoS2 is mainly due to two effects: Schottky barrier change by the band gap change of MoS2 and Schottky barrier change by piezoelectric polarization of MoS2. To investigate the band gap change of MoS2 under the mechanical strain, we measured the photoluminescence (PL) of MoS2 and calculated the theoretical band gap of MoS2 by the density functional theory (DFT) under tensile strain (Figure 6). The energy of PL shifted from 1.888 to 1.856 eV by inducing 0.17% strain, indicating that the band gap variation of MoS2 was only 20 meV (Figure 6a). The band gap change in the DFT calculation was 18 meV induced by 0.15% strain (Figure 6b,c), which is much smaller than the Schottky barrier variation (118 meV) in our measurement. Furthermore, the difference in the Schottky barrier change between positive gate voltage (ΔΦSB = 13 meV at Vg − ΔVth = 3 V) and near-zero gate voltage (ΔΦSB = 118 meV at Vg − ΔVth = −0.1 V) cannot be explained by the band gap change of MoS2. We measured the transfer curve of graphene/MoS2 vdwH FET under tensile and compressive strain (Figure 6d). It shows negative gate shift and current increase under tensile strain (red curves Figure 6d) and positive gate shift and current decrease under compressive strain (blue curves Figure 6d). It is because the Schottky barrier is decreased by positive polarization charges of tensile strain and increased by negative polarization charges of compressive strain (Figure 6e). Our device also shows high stability in the transfer curve and current variation under bending−releasing repetition (Figure 6f). It can be noted that the MoS2 is an n-type material; electrons rather than holes flow from the source to MoS2 and the drain. Even though Schottky barriers at the source and drain are changed together by polarization charges, the Schottky barrier at the source contact plays a dominant role in current modulation (Figure S12). Therefore, we focused on the Schottky barrier change at source contact in the energy band diagrams.

Schottky barrier to enable great improvement in strain sensors based on vdWH FETs, and it can lead to further developments in future high-performance electronic and mechanical− electronic devices.

METHODS Fabrication Steps of Flexible Graphene/MoS2 vdWH FET. We transferred a large-area graphene film grown by CVD onto a Si/ SiO2 (300 nm) substrate by the well-known wet-transfer method and then patterned it as the source and drain electrode by photolithography and reactive ion etching. The MoS2 exfoliated flakes (their Raman spectrum is shown in Figure S1a) for the semiconductor channel were transferred by the well-known dry-transfer method2 to form contact with the source−drain graphene electrode (Figure 1a). Next, h-BN flakes were transferred as the gate dielectric material using the same dry-transfer method (Figure S1b), after which an entire graphene film was transferred and patterned as the gate electrode (Figure S1c). To demonstrate the effect of strain application on the graphene/MoS2 vdWH FET, the entire fabricated FET was transferred onto a flexible PET substrate. For this process, a binder, PMMA, was spin-coated onto the vdWH FET located on the Si wafer. The PMMA-coated sample was detached from the Si wafer by removal of the underlying SiO2 sacrificial layer by immersing it in the BOE (Figure S1d). The FET structure bound on the PMMA layer was then transferred onto a flexible PET substrate, and the process was completed by removal of PMMA in an acetone solution (Figure S1e). Figure S1f shows a photograph and an optical image of the flexible graphene/MoS2 vdWHs FET. The FET boundary is hardly distinguishable, owing to the high transparency of the 2D materials. The schematic and optical images of the fabrication process are shown in Figure S1. Microscopic and Electrical Characterization. We used a confocal Raman spectrometer (ALPHA 300R, WITEC) to determine the number of layers of MoS2. The difference between the peak frequencies E12g and A1g was below 18 cm−1 for monolayer MoS2,46,47 while this difference was approximately 20 cm−1 for MoS2 (Figure S2). A cryogenic probe station (CRX-VF, Lake Shore) and a semiconductor parameter analyzer (4200, Keithley) were used to measure the electrical characteristics at low temperature. As copper has a high heat conductivity, we used a bent copper substrate to transfer heat from the cryogenic probe station to the MoS2 vdWH FET (Figure 2a).

ASSOCIATED CONTENT

CONCLUSIONS In conclusion, we have demonstrated an ultrahigh gauge factor by using a graphene/MoS2 vdWH FET with an efficiently controllable Schottky barrier. With a low density of states near the Dirac point in graphene, strain-induced polarized charges of MoS2 efficiently modulate the graphene Fermi level and corresponding Schottky barrier in the graphene/MoS 2 junction. Thus, an ultrahigh gauge factor of 575 294 was achieved, which is about 500 times higher than that of conventional metal/MoS2 junction strain sensors. The straindependent gauge factors of the devices with armchair MoS2 are higher than that of zigzag MoS2. The practical mass production of armchair-oriented MoS2 strain sensors can be achieved by using the growth of centimeter-scale monolayer MoS2 with control over lattice orientation by lattice alignment via van der Waals interaction between a sapphire substrate and MoS2.45 In practical use of the sensor, each sensor is loaded in its own sensor system with individual ROIC (readout integrated circuits). The calibration block of ROIC calibrates the signal based on its own sensor output. Therefore, even though the baselines are different between sensors, the sensor system can correctly calibrate the signals individually. Our study demonstrates a general strategy for successful control of the

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.9b03993. Additional details on experimental methods, figures for fabrication process, Raman spectrum, AFM data, I−V characteristics at different temperatures, photoluminescence, first-principles electronic structure, other data of different samples using CVD MoS2, and method for simulation of energy band diagram and carrier concentration measurement (PDF)

AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. ORCID

Young Hee Lee: 0000-0001-7403-8157 Woo Jong Yu: 0000-0002-7399-307X Author Contributions

W.J.Y., Y.H.L., and I.L. conceived the research and designed the experiment. I.L. performed most of the experiments including device fabrication, characterization, and data analysis. G

DOI: 10.1021/acsnano.9b03993 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

(13) Kim, S. K.; Bhatia, R.; Kim, T.-H.; Seol, D.; Kim, J. H.; Kim, H.; Seung, W.; Kim, Y.; Lee, Y. H.; Kim, S.-W. Directional Dependent Piezoelectric Effect in CVD Grown Monolayer MoS2 for Flexible Piezoelectric Nanogenerators. Nano Energy 2016, 22, 483. (14) Qiu, D.; Kim, E. K. Electrically Tunable and Negative Schottky Barriers in Multi-layered Graphene/MoS2 Heterostructured Transistors. Sci. Rep. 2015, 5, 13743. (15) Tian, H.; Tan, Z.; Wu, C.; Wang, X.; Mohammad, M. A.; Xie, D.; Yang, Y.; Wang, J.; Li, L.; Xu, J.; Ren, T. Novel Field-Effect Schottky Barrier Transistors Based on Graphene-MoS2 Heterojunctions. Sci. Rep. 2015, 4, 5951. (16) Sarkar, D.; Xie, X.; Liu, W.; Cao, W.; Kang, J.; Gong, Y.; Kraemer, S.; Ajayan, P. M.; Banerjee, K. A Subthermionic Tunnel Field-Effect Transistor with an Atomically Thin Channel. Nature 2015, 526, 91. (17) Roy, T.; Tosun, M.; Cao, X.; Fang, H.; Lien, D.-H.; Zhao, P.; Chen, Y.-Z.; Chueh, Y.-L.; Guo, J.; Javey, A. Dual-Gated MoS2/WSe2 Van der Waals Tunnel Diodes and Transistors. ACS Nano 2015, 9, 2071. (18) Shin, Y. S.; Lee, K.; Kim, Y. R.; Lee, H.; Lee, I. M.; Kang, W. T.; Lee, B. H.; Kim, K.; Heo, J.; Park, S.; Lee, Y. H.; Yu, W. J. Mobility Engineering in Vertical Field Effect Transistors based on Van der Waals Heterostructures. Adv. Mater. 2018, 30, 1704435. (19) Zan, W.; Geng, W.; Liu, H.; Yao, X. Electric-Field and StrainTunable Electronic Properties of MoS2/h-BN/Graphene Vertical Heterostructures. Phys. Chem. Chem. Phys. 2016, 18, 3159. (20) Huo, N.; Kang, J.; Wei, Z.; Li, S. S.; Li, J.; Wei, S. H. Novel and Enhanced Optoelectronic Performances of Multilayer MoS2−WS2 Heterostructure Transistors. Adv. Funct. Mater. 2014, 24, 7025. (21) Vu, Q. A.; Lee, J. H.; Nguyen, V. L.; Shin, Y. S.; Lim, S. C.; Lee, K.; Heo, J.; Park, S.; Kim, K.; Lee, Y. H.; Yu, W. J. Tuning Carrier Tunneling in Van der Waals Heterostructures for Ultrahigh Detectivity. Nano Lett. 2017, 17, 453. (22) Yu, W. J.; Vu, Q. A.; Oh, H.; Nam, H. G.; Zhou, H.; Cha, S.; Kim, J. Y.; Carvalho, A.; Jeong, M.; Choi, H.; Castro Neto, A. H.; Lee, Y. H.; Duan, X. Unusually Efficient Photocurrent Extraction in Monolayer Van der Waals Heterostructure by Tunnelling through Discretized Barriers. Nat. Commun. 2016, 7, 13278. (23) Xue, Y.; Zhang, Y.; Liu, Y.; Liu, H.; Song, J.; Sophia, J.; Liu, J.; Xu, Z.; Xu, Q.; Wang, Z.; Zheng, J.; Liu, Y.; Li, S.; Bao, Q. Scalable Production of a Few-Layer MoS2/WS2 Vertical Heterojunction Array and Its Application for Photodetectors. ACS Nano 2016, 10, 573. (24) Ye, L.; Li, H.; Chen, Z.; Xu, J. Near-Infrared Photodetector Based on MoS2/Black Phosphorus Heterojunction. ACS Photonics 2016, 3, 692. (25) Yu, W. J.; Liu, Y.; Zhou, H.; Yin, A.; Li, Z.; Huang, Y.; Duan, X. Highly Efficient Gate-Tunable Photocurrent Generation in Vertical Heterostructures of Layered Materials. Nat. Nanotechnol. 2013, 8, 952. (26) Sundaram, R. S.; Engel, M.; Lombardo, A.; Krupke, R.; Ferrari, A. C.; Avouris, P.; Steiner, M. Electroluminescence in Single Layer MoS2. Nano Lett. 2013, 13, 1416. (27) Ye, Y.; Ye, Z.; Gharghi, M.; Zhu, H.; Zhao, M.; Yin, X.; Zhang, X. Exciton-Dominant Electroluminescence from a Diode of Monolayer MoS2. Appl. Phys. Lett. 2014, 104, 193508. (28) Ross, J. S.; Klement, P.; Jones, A. M.; Ghimire, N. J.; Yan, J.; Mandrus, D. G.; Taniguchi, T.; Watanabe, K.; Kitamura, K.; Yao, W.; Cobden, D. H.; Xu, X. Electrically Tunable Excitonic Light-Emitting Diodes based on Monolayer WSe2 p−n Junctions. Nat. Nanotechnol. 2014, 9, 268. (29) Withers, F.; Del Pozo-Zamudio, O.; Schwarz, S.; Dufferwiel, S.; Walker, P. M.; Godde, T.; Rooney, A. P.; Gholinia, A.; Woods, C. R.; Blake, P.; Haigh, S. J.; Watanabe, K.; Taniguchi, T.; Aleiner, I. L.; Geim, A. K.; Fal’ko, V. I.; Tartakovskii, A. I.; Novoselov, K. S. WSe2 Light-Emitting Tunneling Transistors with Enhanced Brightness at Room Temperature. Nano Lett. 2015, 15, 8223. (30) Bertolazzi, S.; Krasnozhon, D.; Kis, A. Nonvolatile Memory Cells Based on MoS2/Graphene Heterostructures. ACS Nano 2013, 7, 3246.

K.W.T. and Y.S.S. prepared the samples. W.J.Y. and D.L.D. performed electrostatic simulation. K.K., K.L., and J.H. helped with the data analysis. W.J.Y., Y.H.L., and I.L wrote the paper. All authors discussed the results and commented on the manuscript. Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS W.J.Y. acknowledges Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (NRF2015R1C1A1A02037387 and NRF-2018R1A2B2008069), the R&D program of MOTIE/KEIT (10064078), MultiMinistry Collaborative R&D Program (Development of Techniques for Identification and Analysis of Gas Molecules to Protect Against Toxic Substances) through the National Research Foundation of Korea (NRF) funded by KNPA, MSIT, MOTIE, ME, NFA (2017M3D9A1073539), and the framework of the International Cooperation Program managed by the National Research Foundation of Korea (2018K2A9A2A06017491). Y.H.L. wishes to acknowledge the Institute for Basic Science (IBS-R011-D1). REFERENCES (1) Novoselov, K. S.; Jiang, D.; Schedin, F.; Booth, T. J.; Khotkevich, V. V.; Morozov, S. V.; Geim, A. K. Two-Dimensional Atomic Crystals. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 10451. (2) Dean, C. R.; Young, A. F.; Meric, I.; Lee, C.; Wang, L.; Sorgenfrei, S.; Watanabe, K.; Taniguchi, T.; Kim, P.; Shepard, K. L.; Hone, J. Boron Nitride Substrates for High-Quality Graphene Electronics. Nat. Nanotechnol. 2010, 5, 722. (3) Lee, G.-H.; Yu, Y.-J.; Lee, C.; Dean, C.; Shepard, K. L.; Kim, P.; Hone, J. Electron Tunneling through Atomically Flat and Ultrathin Hexagonal Boron Nitride. Appl. Phys. Lett. 2011, 99, 24. (4) Lee, G.-H.; Yu, Y.-J.; Cui, X.; Petrone, N.; Lee, C.-H.; Choi, M. S.; Lee, D.-Y.; Lee, C.; Yoo, W. J.; Watanabe, K.; Taniguchi, T.; Nuckolls, C.; Kim, P.; Hone, J. Flexible and Transparent MoS2 FieldEffect Transistors on Hexagonal Boron Nitride-Graphene Heterostructures. ACS Nano 2013, 7, 7931. (5) Yoon, Y.; Ganapathi, K.; Salahuddin, S. How Good Can Monolayer MoS2 Transistors Be? Nano Lett. 2011, 11, 3768−3773. (6) Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A. Single-Layer MoS2 Transistors. Nat. Nanotechnol. 2011, 6, 147−150. (7) Wu, W. Z.; Wang, L.; Li, Y. L.; Zhang, F.; Lin, L.; Niu, S. M.; Chenet, D.; Zhang, X.; Hao, Y. F.; Heinz, T. F.; Hone, J.; Wang, Z. L. Piezoelectricity of Single-Atomic-Layer MoS2 for Energy Conversion and Piezotronics. Nature 2014, 514, 470. (8) Manzeli, S.; Allain, A.; Ghadimi, A.; Kis, A. Piezoresistivity and Strain-induced Band Gap Tuning in Atomically Thin MoS2. Nano Lett. 2015, 15, 5330. (9) Qi, J.; Lan, Y.-W.; Stieg, A. Z.; Chen, J.-H.; Zhong, Y.-L.; Li, L.J.; Chen, C.-D.; Zhang, Y.; Wang, K. L. Piezoelectric Effect in Chemical Vapour Deposition-Grown Atomic-Monolayer Triangular Molybdenum Disulfide Piezotronics. Nat. Commun. 2015, 6, 7430. (10) Zhu, H. Y.; Wang, Y.; Xiao, J.; Liu, M.; Xiong, S. M.; Wong, Z. J.; Ye, Z. L.; Ye, Y.; Yin, X. B.; Zhang, X. Observation of Piezoelectricity in Free-Standing Monolayer MoS2. Nat. Nanotechnol. 2015, 10, 151. (11) Tsai, M.-Y.; Tarasov, A.; Hesabi, Z. R.; Taghinejad, H.; Campbell, P. M.; Joiner, C. A.; Adibi, A.; Vogel, E. M. Flexible MoS2 Field-Effect Transistors for Gate-Tunable Piezoresistive Strain Sensors. ACS Appl. Mater. Interfaces 2015, 7, 12850. (12) Lee, J.; Wang, Z.; He, K.; Shan, J.; Feng, P. X.-L. High Frequency MoS2 Nanomechanical Resonators. ACS Nano 2013, 7, 6086. H

DOI: 10.1021/acsnano.9b03993 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano (31) Zhang, E.; Wang, W.; Zhang, C.; Jin, Y.; Zhu, G.; Sun, Q.-Q.; Zhang, D. W.; Zhou, P.; Xiu, F. Tunable Charge-Trap Memory Based on Few-Layer MoS2. ACS Nano 2015, 9, 612. (32) Wang, J.; Zou, X.; Xiao, X.; Xu, L.; Wang, C.; Jiang, C.; Ho, J. C.; Wang, T.; Li, J.; Liao, L. Floating Gate Memory-based Monolayer MoS2 Transistor with Metal Nanocrystals Embedded in the Gate Dielectrics. Small 2015, 11, 208. (33) Vu, Q. A.; Shin, Y. S.; Kim, Y. R.; Nguyen, V. L.; Kang, W. T.; Kim, H.; Luong, D. H.; Lee, I. M.; Lee, K.; Ko, D.-S.; Park, S.; Lee, Y. H.; Yu, W. J. Two-Terminal Floating-Gate Memory with Van der Waals Heterostructures for Ultrahigh On/Off Ratio. Nat. Commun. 2016, 7, 12725. (34) Lin, P.; Zhu, L. P.; Li, D.; Xu, L.; Pan, C. F.; Wang, Z. L. PiezoPhototronic Effect for Enhanced Flexible MoS2/WSe2 van der Waals Photodiodes. Adv. Funct. Mater. 2018, 28, 1802849. (35) Lin, P.; Chen, X.; Yan, X.; Zhang, Z.; Yuan, H.; Li, P.; Zhao, Y.; Zhang, Y. Enhanced Photoresponse of Cu2O/ZnO Heterojunction with Piezo-modulated Interface Engineering. Nano Res. 2014, 7, 860− 868. (36) Quereda, J.; Palacios, J. J.; Agräit, N.; Castellanos-Gomez, A.; Rubio-Bollinger, G. Strain Engineering of Schottky Barriers in Singleand Few-Layer MoS2 Vertical Devices. 2D Mater. 2017, 4, No. 021006. (37) Hu, Y.; Klein, B. D.; Sun, Y.; Niu, S.; Liu, Y.; Wang, Z. L. Temperature Dependence of the Piezotronic Effect in ZnO Nanowires. Nano Lett. 2013, 13, 5026. (38) Anderson, B.; Anderson, R. Fundamentals of Semiconductor Devices; McGraw-Hill, Inc.: New York, 2004. (39) Kim, S.; Konar, A.; Hwang, W.; Lee, J. H.; Lee, J.; Yang, J.; Jung, C.; Kim, H.; Yoo, J.; Choi, J.; Jin, Y. W.; Lee, S. Y.; Jena, D.; Choi, W.; Kim, K. High-Mobility and Low-Power Thin-Film Transistors Based on Multilayer MoS2 Crystals. Nat. Commun. 2012, 3, 1011. (40) Zhu, C. R.; Wang, G.; Liu, B. L.; Marie, X.; Qiao, X. F.; Zhang, X.; Wu, X. X.; Fan, H.; Tan, P. H.; Amand, T.; Urbaszek, B. Strain Tuning of Optical Emission Energy and Polarization in Monolayer and Bilayer MoS2. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 88, 121301. (41) Chen, C.; Deshpande, V. V.; Koshino, M.; Lee, S.; Gondarenko, A.; MacDonald, A. H.; Kim, P.; Hone, J. Modulation of Mechanical Resonance by Chemical Potential Oscillation in Graphene. Nat. Phys. 2016, 12, 240−244. (42) Mak, K. F.; Lee, C.; Hone, J.; Shan, J.; Heinz, T. F. Atomically Thin MoS2: A New Direct-Gap Semiconductor. Phys. Rev. Lett. 2010, 105, 136805. (43) Ma, Y.; Dai, Y.; Guo, M.; Niu, C.; Huang, B. Graphene Adhesion on MoS2 Monolayer: an AB initio Study. Nanoscale 2011, 3, 3883−3887. (44) Han, S. A.; Kim, T. H.; Kim, S. K.; Lee, K. H.; Park, H. J.; Lee, J. H.; Kim, S. W. Point-Defect-Passivated MoS2 Nanosheet-Based High Performance Piezoelectric Nanogenerator. Adv. Mater. 2018, 30, 1800342. (45) Dumcenco, D.; Ovchinnikov, D.; Marinov, K.; Lazić, P.; Gibertini, M.; Marzari, N.; Sanchez, O. L.; Kung, Y.-C.; Krasnozhon, D.; Chen, M.-W; Bertolazzi, S.; Gillet, P.; Morral, A. F.; Radenovic, A.; Kis, A. Large-Area Epitaxial Monolayer MoS2. ACS Nano 2015, 9 (4), 4611−4620. (46) Lee, C.; Yan, H.; Brus, L. E.; Heinz, T. F.; Hone, J.; Ryu, S. Anomalous Lattice Vibrations of Single- and Few-Layer MoS2. ACS Nano 2010, 4, 2695. (47) Li, H.; Zhang, Q.; Yap, C. C. R.; Tay, B. K.; Edwin, T. H. T.; Olivier, A.; Baillargeat, D. From Bulk to Monolayer MoS2: Evolution of Raman Scattering. Adv. Funct. Mater. 2012, 22, 1385. (48) Dobie, W. B., Isaac, P. C. G. Electric Resistance Strain Gauges; English Universities Press Limited: Madison, WI, USA, 1948. (49) Beeby, S. Mems Mechanical Sensors; Artech House: Boston, 2004. (50) Kim, K. K.; Hong, S.; Cho, H. M.; Lee, J.; Suh, Y. D.; Ham, J.; Ko, S. H. Highly Sensitive and Stretchable Multidimensional Strain

Sensor with Prestrained Anisotropic Metal Nanowire Percolation Networks. Nano Lett. 2015, 15, 5240. (51) Zhang, Z.; Liao, Q.; Zhang, X.; Zhang, G.; Li, P.; Lu, S.; Liu, S.; Zhang, Y. Highly Efficient Piezotronic Strain Sensors with Symmetrical Schottky Contacts on the Monopolar Surface of ZnO Nanobelts. Nanoscale 2015, 7, 1796. (52) Bae, S.-H.; Lee, Y.; Sharma, B. K.; Lee, H.-J.; Kim, J.-H.; Ahn, J.-H. Graphene-Based Transparent Strain Sensor. Carbon 2013, 51, 236. (53) Cao, J.; Wang, Q.; Dai, H. Electromechanical Properties of Metallic, Quasimetallic, and Semiconducting Carbon Nanotubes under Stretching. Phys. Rev. Lett. 2003, 90, 157601.

I

DOI: 10.1021/acsnano.9b03993 ACS Nano XXXX, XXX, XXX−XXX