Effects of Uniaxial and Biaxial Strain on Few-Layered Terrace

Feb 16, 2016 - Xiaoxu Zhao , Zijing Ding , Jianyi Chen , Jiadong Dan , Sock Mui Poh , Wei Fu , Stephen J. Pennycook , Wu Zhou , and Kian Ping Loh...
0 downloads 0 Views 6MB Size
Effects of Uniaxial and Biaxial Strain on FewLayered Terrace Structures of MoS2 Grown by Vapor Transport Amber McCreary,†,‡,§ Rudresh Ghosh,∥ Matin Amani,†,%,& Jin Wang,⊥ Karel-Alexander N. Duerloo,# Ankit Sharma,∥ Karalee Jarvis,⊗ Evan J. Reed,# Avinash M. Dongare,⊥ Sanjay K. Banerjee,∥ Mauricio Terrones,*,§,∇,$ Raju R. Namburu,‡ and Madan Dubey*,† †

Sensors & Electron Devices Directorate, U.S. Army Research Laboratory, Adelphi, Maryland 20783, United States Computational and Information Sciences Directorate, U.S. Army Research Laboratory, Aberdeen Proving Ground, Maryland 21005, United States § Department of Physics and Center for 2-Dimensional and Layered Materials, ∇Department of Materials Science & Engineering, and $ Department of Chemistry, The Pennsylvania State University, University Park, Pennsylvania 16802, United States ∥ Microelectronics Research Center, and ⊗Texas Materials Institute, The University of Texas at Austin, Austin, Texas 78758, United States ⊥ Department of Materials Science and Engineering, and Institute of Materials Science, University of Connecticut, Storrs, Connecticut 06269, United States # Department of Materials Science and Engineering, Stanford University, Stanford, California 94305, United States ‡

S Supporting Information *

ABSTRACT: One of the most fascinating properties of molybdenum disulfide (MoS2) is its ability to be subjected to large amounts of strain without experiencing degradation. The potential of MoS2 mono- and few-layers in electronics, optoelectronics, and flexible devices requires the fundamental understanding of their properties as a function of strain. While previous reports have studied mechanically exfoliated flakes, tensile strain experiments on chemical vapor deposition (CVD)-grown few-layered MoS2 have not been examined hitherto, although CVD is a state of the art synthesis technique with clear potential for scale-up processes. In this report, we used CVD-grown terrace MoS2 layers to study how the number and size of the layers affected the physical properties under uniaxial and biaxial tensile strain. Interestingly, we observed significant shifts in both the Raman in-plane mode (as high as −5.2 cm−1) and photoluminescence (PL) energy (as high as −88 meV) for the few-layered MoS2 under ∼1.5% applied uniaxial tensile strain when compared to monolayers and few-layers of MoS2 studied previously. We also observed slippage between the layers which resulted in a hysteresis of the Raman and PL spectra during further applications of strain. Through DFT calculations, we contended that this random layer slippage was due to defects present in CVD-grown materials. This work demonstrates that CVD-grown few-layered MoS2 is a realistic, exciting material for tuning its properties under tensile strain. KEYWORDS: molybdenum disulfide, strain engineering, Raman, photoluminescence, interlayer sliding, CVD, few-layer ver since the isolation of graphene in 2004,1 twodimensional (2D) layered materials, including transition metal dichalcogenides (TMDs), have become a fast growing research field that holds many exciting opportunities. With weak van der Waals interactions between the layers, these materials can be easily thinned to single and few-layers via mechanical and chemical exfoliation, exhibiting novel properties as the thickness is reduced.2−4 Although a layer of graphene is only a single carbon atom thick,3 TMD monolayers consist of three atoms commensurate with an MX2 structure in which M

E

© 2016 American Chemical Society

is a transition metal (Mo, W, Nb, Re, etc.) and X is a chalcogen atom (S, Se, Te).5 As it is naturally occurring, molybdenum disulfide (MoS2) has been the most studied of the TMDs. Bulk MoS2 possesses a 2H polytype crystalline structure that belongs to the hexagonal space group P63/mmc.6 Each monolayer of Received: July 22, 2015 Accepted: February 16, 2016 Published: February 16, 2016 3186

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

www.acsnano.org

Article

ACS Nano

Figure 1. Terrace layers of MoS2. (a) Structures of MoS2 with layers grown on top of one another. Raman and PL characterization as a function of thickness are shown on the PET substrate in (b) and (c), respectively. The colored dots on (a) indicate where the spectra were taken in (b) and (c).

both mono- and few-layered MoS2 are affected under uniaxial and biaxial tensile strain. Raman spectroscopy and PL measurements, being nondestructive and highly sensitive to layer thickness,9,35−38 have become standard characterization techniques to study the properties of 2D materials including the effects of strain,30,31,39−41 doping/defects,42−44 etc. With an off-resonance laser wavelength and in the backscattering configuration, as was utilized in this study, bulk MoS2 (which belongs to the D6h point group) displays three prominent Raman modes:37 E22g at 32 cm−1, E12g at 383 cm−1, and A1g at 408 cm−1. The E22g mode occurs at a very low frequency and could not be detected in our measurements due to filtering of the Rayleigh scattered light. The E12g is an in-plane mode resulting from opposite vibration of the two sulfur atoms with respect to the molybdenum atom while the A1g is an out-of-plane vibration of only the sulfur atoms in opposite directions.37 When the sample is thinned down to few-layers, the labeling of these modes depends on whether there are an even or odd number of layers.6 For an odd number of layers, the vibrational modes belong to the D3h point group, which lacks inversion symmetry, and are labeled as E′ and A′1, while for an even number of layers, the modes belong to the D3d point group that does exhibit a center of inversion and the notations remain the same as for the bulk. Thus, the labeling of the Raman peaks switches between A1′ and E′ for an odd and A1g and E12g for an even number of layers.6 As this study involved varying number of layers, the modes will now be referred to as the A and E modes for convenience. Previous uniaxial strain studies of TMDs have shown that the position of the in-plane E mode is very sensitive to strain and decreases linearly as a function of strain.30−33,45 Due to the symmetry breaking caused by the uniaxial strain, it is expected that the degeneracy of the E mode will be broken and thus split into two peaks, but this is not seen in all of the previous reports, possibly due to the spectral resolution of the spectrometer used.30 The out-of-plane A mode, conversely, does not display a significant shift with applied strain.30−33,45

2H-MoS2 exhibits a trigonal prismatic coordination with the molybdenum atoms hexagonally packed between two atomic layers of sulfur atoms.7 One very interesting property of MoS2 is the evolution of the band structure as the number of layers is thinned down from bulk to few-layers and even a monolayer.8,9 In its bulk form, MoS2 is an indirect-gap semiconductor in which the lowest transition goes from the valence band maximum (VBM) at the Γ point to the conduction band minimum (CBM) in the Γ−K direction. For monolayered MoS2, however, the direct gap at the K point becomes the smallest electronic transition, thus making monolayered MoS2 a direct-gap semiconductor with optical properties different from its few-layered counterparts.8,9 The optical gap of suspended monolayered MoS2 was measured to be around 1.9 eV.9 Ample research has been done to gain a better understanding of the physiochemical properties of both mono- and few-layered MoS2, including transport,10−12 photocurrent,13−15 photoluminescence,9,16 valley polarization,17 and the valley Hall effect.18 The application of strain to modify the band structure of a material is a common strategy in electronics to tune the performance of a device. In particular, monolayers of MoS2 are extremely sensitive to external perturbations,4,19,20 and thus their properties can be continuously and easily tuned. For example, it has been predicted that at relatively small tensile strains (∼2% uniaxial strain) monolayered MoS2 will undergo a direct to indirect gap transition,21,22 while ∼10% biaxial strain will even induce a semiconductor-to-metal transition.21−23 In addition, it has been shown that large amounts of strain can be applied to MoS2 before rupturing.24,25 Although monolayered MoS2 offers the thinnest semiconducting material thus far and has shown outstanding results in flexible electronics,26 in some cases, few-layers may be more suitable for large-scale applications. Due to a smaller bandgap and higher density of states,27 few-layered MoS2 devices exhibit superior on-state performance20,28 and higher effective field-effect mobilities20,28,29 when compared to monolayered devices with only a small degradation in the ON/OFF current ratios.27 While the optical properties of monolayered MoS2 under strain have been previously studied,30−33 the effects of strain on few-layered MoS2, however, are not well understood. Furthermore, previous literature reporting tensile strain experiments focused on exfoliated material, and to the best of our knowledge, no one has reported the optical properties of few-layered CVD-grown MoS2 under tensile strain.34 In this paper, we used Raman and photoluminescence (PL) spectroscopies to shed light on how

RESULTS AND DISCUSSION In this study, few-layered MoS2 was synthesized via CVD at 950 °C, as shown in Figure 1a and in the Supporting Information (Figures S1 and S2). Different thicknesses of MoS2 domains grew as a terrace structure on a Si substrate with a 285 nm thick SiO2 layer. Raman spectroscopy (Figure 1b), PL spectroscopy (Figure 1c), atomic force microscopy (AFM) (Figure S3), and differential reflectance contrast (Figure S4) 3187

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano

Figure 2. Strain study on terrace MoS2 layers. The two structures studied are shown on PET in (a) and (b) and labeled T1 and T2, respectively. T1 and T2 had very similar structural characteristics and Pts 1−5 are labeled on both. The yellow arrow shows the approximate direction of the strain that was applied by bending the PET as shown in (c). T1 and T2 were located at the top of the arc to ensure maximum strain. The position of E Raman mode is shown for T1/T2 in (d)/(f) while the PL A-exciton energy for T1/T2 is shown in (e)/(g), respectively.

In order to study the effects of uniaxial strain on the different thicknesses of MoS2, the samples were transferred from the growth SiO2/Si substrate to a flexible substrate (polyethylene terephthalate (PET), 175 μm thick) via a wet-etch of the SiO2 using buffered oxide etching to release the MoS2.46 The two terrace structures with very similar characteristics chosen for the strain study are shown on PET in Figure 2a,b and are referred to hereafter as triangle 1 (T1) and triangle 2 (T2), respectively. Points (Pts) 1−5 are also labeled on T1 and T2,

were carried out on the different MoS2 islands. In the PL spectra, both the A- and B-excitons9 could be observed, yet the A-exciton intensity dominated for the monolayer, while for thicker MoS2 layers they were similar in intensity. Because the different thicknesses were easy to distinguish, these few-layered MoS2 terrace structures provided the perfect opportunity to study the effects of strain for different layers of MoS2 using Raman and PL as a function of applied uniaxial tensile strain. 3188

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano with Pt1 being a single layer and increasing in thickness up to Pt5, which corresponded to six layers for T2 and slightly thicker for T1 (see the Supporting Information). To apply uniaxial strain to the MoS2 structures, the PET substrate was bent as shown in Figure 2c using aluminum plates and optical micrometers. By measuring the distance between the plates and the angle (θ) of the tangent at the minimum strain point (see Figure S5), the strain was estimated with the equation ε=

τ sin θ d

(1)

where ε is the strain, τ is the thickness of the PET (175 μm), and d is the distance between the two plates.31,40,41,47 The approximate direction of strain is also indicated in Figure 2a,b by the yellow arrows. At each value of strain, Raman and direct-gap PL were measured for points 1−5 on T1 and T2. The frequencies of the in-plane E mode (ωE) and the A-exciton energy (referred to as PL(A)) as a function of the applied strain are shown in Figure 2d,e for T1 and Figure 2f,g for T2, respectively (the B-exciton energy is shown in the Supporting Information). Both ωE and PL(A) displayed a monotonic decrease as a function of strain for all thicknesses, as was previously reported.30,31,33,45 However, while the E mode for the monolayer only changed by about 1 wavenumber after approximately 1.5% uniaxial strain, the change was much more dramatic for few-layers, whose frequencies shifted by as much as 5 wavenumbers. Another interesting observation was the apparent slipping of Pt4 for T1 and Pt3 for T2 (circled in Figures 2d−g) that could be observed in both ωE and the PL(A) energy returning toward their original values. Although the T1-Pt4 and T1-Pt3 layers displayed slipping, the layers above them (Pt5 for T1 and Pt4− 5 for T2) were not affected, as seen by their continued linear decrease. This suggested that the interaction with the layers above kept the area directly underneath of the upper layers from slipping while the exposed area was more free to relax. In T1, for example, although the fourth layer slipped, the area of the fourth layer that had another layer on top was prevented from slipping due to those layer−layer interactions. From this, we concluded that the observed slipping was a local, random slipping between two layers. A possible mechanism for this sliding will be discussed below. We believe that T1-Pt2 did not display the same magnitude of shifts as the other multilayered points because, upon closer inspection of Figure 2a, the bilayer area contained some tears in it, and thus, the strain was not uniformly transferred. Figure 3 shows the change in the (3a) Raman and (3b) PL spectra for T1-Pt3 (as an example of a layer that did not slip) as a function of applied uniaxial strain. It is clear from Figure 3a that the position of the A mode was not as affected by strain as the E mode. Although we were not able to observe a splitting of the E mode into two distinct peaks for any of the layer thicknesses studied here, there was a clear, steady increase in the full width at half-maximum (FWHM) of the E mode by as much as 1.9 wavenumbers (see Figure S6), which suggested peak splitting that could not be resolved by the spectrometer used (see Figure S7 for more details). Ab initio calculations based on density functional theory (DFT) were performed to compare our experimental results with the predicted shifts for both the frequency of the Raman E mode and direct-gap energy under uniaxial strain for 1−6 layers of MoS2. Both the experimental and the calculated slopes (with respect to strain) for the different layer thicknesses are shown

Figure 3. Change in (a) Raman and (b) PL spectra as a function of strain for Pt3 on T1. The arrow indicates the direction of increasing strain. The A mode does not change as much as the E mode with the application of strain.

in Figure 4. Figure 4a shows the E mode, and Figure 4b shows the slopes in PL(A) energy (experimental) and DFT direct gaps (calculated). The error bars on the experimental values represent ± the standard error in the linear fitting. There were very small differences in the calculated slopes for 1−6 layers of MoS2 for both the E mode and the direct gap energy and a small variation in the slopes based on the direction of applied strain. As shown in Figure 4a, the uniaxial strain was predicted to break the degeneracy of the E mode, resulting in an upper and lower branch with different slopes. Although we did not observe the splitting of the in-plane E mode, our experimental results agreed well with the calculated values. Because the splitting could not be resolved, the experimental Raman slopes were in-between the upper and the lower branch (closer to the lower branch). Concerning the PL in Figure 4b, as was mentioned previously, all layer thicknesses of MoS2 except the monolayer are indirect-gap semiconductors.9 Although the lowest transition for N > 1 layers (N is the number of layers) is the indirect-gap transition from the VBM at the Γ point to the CBM in the Γ − K direction, these layers still exhibit luminescence from the two direct transitions at the K point the Brillouin zone,9 which correspond to the PL(A) and PL(B) 3189

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano

Figure 4. Slope of Raman and PL due to uniaxial tensile strain. (a) Experimental (pink and blue) and calculated (orange and green) slopes of E mode (Δω/ε) for 1−7 layers of MoS2. The calculated slopes show both the upper and lower branch. (b) Experimental PL A-exciton (pink and blue) and calculated direct gap (orange and green) slopes (change in energy per percent strain) for 1−7 layers of MoS2. All of the calculated slopes are ab initio based on DFT for uniaxial tensile strain applied in either the zigzag or armchair direction.

Figure 5. Second cycle of the strain measurements on T1 and T2. (a, c) Position of the E mode; (b, d) PL(A) energy for T1 and T2, respectively. In this measurement, the relaxed strain was performed from around 1.6% down to 0.7% and is displayed as dashed lines. After Raman and PL were taken at the highest strain value, the sample remained at the maximum strain overnight, and the Raman and PL were measured again.

3190

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano

Table 1. Summary of the Experimental Data for the First and Second Cycles of Strain, Two Exfoliated Few-Layered Flakes of MoS2, and Other References from the Literature E mode Δω/ε 1st T1 − Pt1 (1L) T1 − Pt2 (2L) T1 − Pt3 (3L) T1 − Pt4 (4L) T1 − Pt5 (7/8L) T2 − Pt1 (1L) T2 − Pt2 (2L) T2 − Pt3 (3L) T2 − Pt4 (4L) T2 − Pt5 (6L) crystal 1−exfoliated crystal 2−exfoliated other ref (1L), E−/E+

A mode

γE 2nd

other ref (2L), E−/E+

−1.03 −1.35 −1.33 −1.40 −3.42 −3.11 −3.74 −1.51 −3.61 −3.06 −0.74 −1.20 −3.04 −3.32 −3.62 −1.97 −3.79 −3.06 −3.24 −2.62 −1.92 −2.34 −2.130 −4.5/-1.031 −2.5/-0.833 −4.6/-1.031

other ref (FL)

−1.730

βE

Δω/ε

γE

1st

2nd

1st

2nd

1st

2nd

0.40 0.52 1.33 1.45 1.41 0.29 1.18 1.41 1.47 1.26 0.75 0.91 0.6530 1.131 0.633 1.131

0.52 0.55 1.21 0.58 1.19 0.46 1.29 0.76 1.19 1.02

0.20 0.26 0.67 0.73 0.71 0.14 0.59 0.71 0.74 0.63 0.38 0.46 0.3430 0.7831 0.333 0.7831

0.26 0.27 0.61 0.29 0.60 0.23 0.65 0.38 0.60 0.51

−0.37 −0.49 −1.00 −1.23 −1.02 −0.71 −0.45 −1.06 −1.03 −1.23 −0.65 −0.73 −0.430

−0.28 −0.41 −0.83 −0.66 −0.82 −0.32 −0.78 −0.84 −1.00 −0.70

βE

PL(A) (meV/%)

1st

2nd

1st

2nd

1st

2nd

0.14 0.18 0.37 0.45 0.37 0.26 0.16 0.39 0.37 0.45 0.24 0.27 0.2130

0.10 0.15 0.30 0.24 0.30 0.12 0.29 0.31 0.36 0.26

0.07 0.09 0.18 0.23 0.19 0.13 0.08 0.19 0.19 0.23 0.12 0.13 −0.0130

0.05 0.07 0.15 0.12 0.15 0.06 0.14 0.16 0.18 0.13

−27.4 −24.9 −58.4 −63.2 −60.5 −23.2 −53.2 −61.6 −64.3 −49.3 −41.8 −41.3 −4531 −6432 4845 −5331 −7132 4645

−21.8 −30.9 −52.6 −27.1 −51.4 −25.8 −55.1 −38.4 −58.2 −44.7

0.430

the forward and reverse strain conditions. The inconsistent forward and reverse response was due to the disturbance in the relative stacking coherence between the layers when one layer slipped. Once the layer slipped, it did not return to full coherence at any point afterward, thus producing a varied response in the forward and reverse strain measurements. Interestingly, although Pt4 was clearly not as affected by the applied strain as it was in the first measurement, Pt5 remained unaffected by Pt4’s slipping and still felt the full effect of the strain. This once again suggested that the slipping was local and that layers on top of the slipped layer kept the crystallographic coherence directly underneath such that only the exposed area was subjected to slipping. T2 (Figure 5c,d) was more complicated on the second cycle of strain measurements. The bilayer (Pt2) appeared to be the only layer that did not slip at some point during the measurement, thus displaying the same trend in the E position and PL(A) for forward and release strain. T2-Pt3, which experienced slipping in the first strain cycle, was also not as affected by the strain in the second strain measurement, with large variations between the forward and release strain. Points 4 and 5 both followed similar trends to the first strain cycle until they experienced slipping around 1.3%. Afterward, the path of the reverse strain deviated from the forward strain, with Pt4 even taking a new, linear path (see Figure 5c,d). This could have been due to a local growth defect, such as a stacking dislocation, that is relaxed away by the application of strain. Using our experimental data from Figure 2d,f, Figure 5a,c, as well as results reported by Mohiuddin et al.,40 two important parameters were experimentally determined for different layers of MoS2: the Grü neisen parameter γm and the shear deformation potential βm, where m is the phonon mode. The Grüneisen parameter is defined as

transitions measured in this experiment. The predicted slopes in Figure 4b were calculated using the smallest direct-gap transition over the Brillouin zone (see the Methods), which should have corresponded to PL(A); however, excitonic effects were not taken into account. Aside from the monolayer and T1-Pt2, which we believe were degraded during the transfer process, significant differences between the Raman and PL slopes for the few-layered MoS2 were not observed, especially within the standard error of the fittings. In addition, the calculated slopes were based on pristine MoS2, whereas CVDgrown material is prone to defects. Thus, the slightly different slopes observed for the layers could have been explained by different defect densities in the areas where the spectra were taken. The role of defects in these systems will be investigated below. The same strain measurements were performed again on the sample for up to 1.6% applied uniaxial strain and are shown in Figure 5a,b for T1 and Figure 5c,d for T2. In addition, strain releases were also measured to about 0.7% and are indicated by the dashed lines in Figure 5. Since it is important to know if these materials can hold strain for a length of time, once the Raman and PL spectra were recorded at the highest strain value, the sample stayed overnight at this strain value, after which the spectra were taken again to see if the strain was still applied. Afterward, the strain release was performed. Looking at T1 (Figure 5a,b), layers that did not experience relative sliding in the first measurement, such as Pt3 and Pt5, showed similar behavior to the first strain cycle in addition to consistent peak positions when comparing the forward and release strains. There was a greater shift in the E position per percent strain of the monolayer in the second measurement, changing by around 2 cm−1 for 1.6% strain. Once again, however, the few-layered points showed a more drastic change, where the position of the E mode shifted by approximately 4.3 cm−1 for Pt3 and Pt5. Conversely, the interlayer sliding of T1-Pt4 during the first strain application inhibited the effect of the strain in subsequent measurements as well as produced an inconsistent response in

γm = − 3191

h 1 ∂ωm ωm0 ∂εh

(2) DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano where εh = εll + εtt is the hydrostatic component of the applied uniaxial strain, l is the direction parallel to the strain, t is the direction perpendicular to it, and ω0m is the frequency of phonon mode m at zero strain. The shear deformation potential βm is defined to be βm = −

s 1 ∂ωm ωm0 ∂εs

(3)

in which εs is the shear component of the strain and is defined as εs = εll − εtt. For uniaxial strain, εll = ε and εtt = −νε, where ν is the Poisson ratio. If the MoS2 had been suspended, then the proper choice for the Poisson’s ratio would have been that of MoS2. Since the MoS2 in this work was being strained on the PET substrate and good adhesion to the substrate was assumed, the Poisson’s ratio of PET33 (ν = 0.33) was used. Table 1 summarizes the experimental data, including Δω/ε, γm, and βm for the E and A Raman modes as well as ΔmeV/ε for the PL Aexciton energy for points 1−5 on T1 and T2 for the first and second cycles of strain, two exfoliated few-layered flakes of MoS2 (Figure S11) and other references from the literature.30−33,45 We also briefly examined the effects of biaxial strain on the CVD-grown few-layered MoS2 structures by studying the shifts in the E peak position and PL A-exciton energy due to the hydrostatic (biaxial) tensile strain that has been reported48,49 for MoS2 synthesized by CVD on SiO2/Si substrates. This intrinsic strain is present within MoS2 structures on their initial growth substrate and is believed to be due to the thermal coefficient of expansion mismatch between the MoS2 and the SiO2 during the cool-down process of growth. When the MoS2 is transferred to a new substrate, such as the PET used in this study, the intrinsic strain is released; the MoS2 on the new substrate is assumed to be strain free. By carefully measuring the average lengths of the triangle sides of monolayered MoS2 produced by CVD at 850 °C both before and after transfer with a scanning electron microscope (SEM), Amani et al. estimated an intrinsic tensile hydrostatic strain of 1.24% was applied to MoS2 on the growth substrate.48 Although the growth temperature for the MoS2 terrace structures in this study was higher than from the previous report (950 °C compared to 850 °C,50 respectively), we assumed that the same magnitude of biaxial strain should have been present in our materials as well. We were unable to measure this for our sample due to the poor optical contrast of MoS2 on PET. By monitoring the optical properties of T2 on the original SiO2/Si growth substrate (with intrinsic biaxial strain, referred to hereafter as εi,b) and after transfer onto the PET (no strain, ε = 0), we compared the effects of εi,b on the different layer counts of MoS2. This is illustrated in Figure 6, which shows the change in frequency of the E mode and PL(A) energy (with respect to ε = 0 on PET) due to εi,b. For εi,b, the change in the position of the E mode was relatively consistent for all layers. Nevertheless, the shift did decrease for thicker layers. The PL(A) energy also followed a similar trend, although the monolayer shift of about 90 meV was significantly higher than the shift for the few-layers. Our observed shift in the direct gap energy of ∼90 meV for approximately 1% strain was very close to DFT-predicted values of 114.1,51 86,52 and 94.553 meV/% strain for monolayered MoS2. A similar shift in the direct gap of bilayered MoS2 (113.07 meV/% strain54) under biaxial tensile strain has also been predicted, yet we observed much smaller shifts (around 30 meV) for the bilayer as well as for the few-

Figure 6. Comparing the shifts due to intrinsic biaxial strain. Shifts in the E mode peak position (purple, left axis) and PL(A) energy (orange, right axis) due to biaxial intrinsic strain on SiO2 relative to their ε = 0 values (after transfer) for different number of layers of MoS2.

layers. The drastic difference between the shifts for the monolayer versus the few-layers raised many questions about how the strain was being distributed between the layers. While it could be speculated that, due to the high temperatures reached during growth, the monolayer experienced more biaxial strain than the few-layers when on the growth substrate, the lack of significant layer-dependence of the shift in the E mode implied that all of the layers underwent approximately the same strain shift when transferred. Changes in dielectric screening could have had a greater impact on the monolayer than the fewlayers since the sample was transferred onto a different type of substrate. This implies the need for a further study on the role of the substrate, as this must be considered when studying the properties of few-layered TMDs on their original CVD-grown substrate where strain is most likely present. From the comparisons with previous literature in Table 1, it was clear that something different occurred in these CVDgrown terrace structures when compared to exfoliated materials. Other references reported that few-layered materials (including graphene) showed lower,30,55 if not very similar,31,45 shifts in the in-plane E mode than their monolayered counterparts. While it is possible the small shifts in the monolayer were due to damage during transfer, the shifts in the few-layers observed in this study were higher than those previously reported. Not only were other reports studying materials exfoliated from bulk MoS2, but also no other report has studied these terrace structures with all of the individual layers exposed and of different sizes. In order to understand if the stacking of the layers played a role in how the strain was distributed between the layers, TEM was performed on a triangle similar to T1 and T2 (see Figure 7). The inset of the left panel shows the selected area electron diffraction (SAED) pattern from the whole area, which showed the hexagonal lattice and ordered stacking of the layers. Diffraction-STEM (D-STEM) from individual spots marked on the stacked MoS2 were also taken and reconfirmed that the layers were ordered. It has been reported56 that, for bilayered MoS2, triangles that have a rotation of θ = 0° between the layers have AA stacking, whereas rotation angles of θ = 60° correspond to AB stacking. From this and the AFM/differential reflectance, the stacking 3192

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano

Figure 7. TEM of MoS2 terrace. (Left) Low magnification TEM of a MoS2 terrace structure similar to those used in the Raman/PL strain studies. The inset of the left panel shows the selected area electron diffraction (SAED) from the whole area viewed down the [001] beam direction. (Right) Diffraction − STEM (D-STEM) from individual spots marked on the stacked MoS2. The D-STEM reconfirms that the diffraction signal from the different parts are the same and shows ordered stacking. The scale bars on all of the diffraction patterns are 10 nm−1.

intrinsic defects, such as vacancies, interstitials, and grain boundaries (McCreary et al., in press73)57,58 that could weaken the interlayer bonding and thus contribute to the slipping. Thus, it was speculated that the unpredictable and random slippage of a certain layer observed in our experiments was attributed to the heterogeneous distribution of local defects in our CVD-grown samples. The local defects may have lowered the energy barrier required to slip, thus increasing the possibility of slippage. In order to validate this idea, DFT simulations were performed to compare the energy barrier along the zigzag sliding pathway between pristine bilayered MoS2 and vacancy-doped bilayered MoS2. The zigzag sliding route is AA′ − AB′ − AA′ (see Figure S15b), which was an optimal pathway found by CI-NEB searching for pristine bilayered MoS2.70 The energy profiles are presented in Figure 8, where the x coordinate represented the intermediate stacking sequences between AA′ and AB′, labeled as the sliding coordinate (SC) from 1 to 4. A sliding coordinate of SC = 0 corresponded to AA′ and SC = 5 corresponded to AB′. As shown in Figure 8, for both pristine and vacancy-doped MoS2, AA′ stacking stayed at a global minimum energy while AB′ stacking was at a local minimum. The energy barrier of 3.29 meV for Vs-doped (one sulfur vacancy) (3.25 meV for Vs2doped, two sulfur vacancies) MoS2 was significantly reduced when compared to the 3.91 meV energy barrier for pristine MoS2. Therefore, the slippage was more likely to occur locally in regions with defects.

order between the layers was determined, as shown in Figure S12. The stacking order between points 1−5 was the same for both triangles. It appeared that for both of the layers that slipped (T1-Pt3 and T2-Pt4) the stacking was AB. Another aspect that could have been significant for these MoS2 structures was the size of the individual layers. Unlike exfoliated flakes that have been strained previously30−33,45 where the layers that composed a few-layered flake were the same size, the layers studied here consisted of terrace MoS2 layers that were very different in sizes, with the areas of the first and second layers about seven times greater than the area of the third layer. Because of this, one question to be considered with these structures was how the strain was distributed throughout the stacked layers of different sizes. If the bottom layer was strained, did the top layer experience the same amount of strain? Were the edges or edge−edge interactions affecting the strain distribution? To elucidate this, we performed molecular dynamics simulations of micron-scale terrace structures similar to those studied experimentally with the lengths of the layers as 15.14, 15.12, 5.06, 5.06, and 2.44 μm from the bottom to top, respectively, as is shown in Figure S13a. Quasistatic deformation was then applied to this five-layered MoS2 system with uniaxial tension in the x direction applied to the bottom layer; the structure was minimized after each strain of 0.1%. The results are shown in Figure S14a,b where the “applied strain” was the strain applied to the bottom layer, the “layer strain” was the strain experienced by each individual layer, and the “layer stress” was the stress in each individual layer. It was clear that regardless of the layer size all of the strain from the bottom layer was transferred to the top, even up to 4% strain. We also studied the stress distribution in the layers (Figure S13b−f) and observed that, for these micron scale structures, it was homogeneous throughout the layer, even on the edges. In addition, the stress vs strain curves were the same for all of the layers. For the molecular dynamics simulations described above, no strain relaxation between the layers was predicted for up to 4% strain. However, the MoS2 in the simulation was completely defect f ree, whereas MoS2 grown via CVD is more prone to

CONCLUSIONS We studied applied uniaxial and intrinsic biaxial strain for CVDgrown terrace structures of MoS2. We observed significant shifts in both the in-plane E Raman mode (up to −5.2 cm−1) and PL energy (up to −88 meV) for the few-layers after approximately 1.5% applied uniaxial tensile strain, which is larger than what has been reported previously. These strong shifts indicated that the MoS2 was well adhered to the PET substrate after transfer. We performed both DFT and molecular dynamics simulations to study how the strain affected the different number of layers in the structure and found good 3193

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano

Transfer of MoS2 onto PET. To transfer the MoS2 (both CVDgrown terrace structures and exfoliated flakes discussed in the Supporting Information), we utilized a wet etch of the SiO2 similar to what was reported previously by Elias et al.46 The substrates were spin-coated with PMMA (MW 495000, A6) at 2000 rpm for 60 s and cured at 40 °C for at least 4 h. Afterward, the samples were placed in a buffered oxide etch (6:1) for a few hours to etch the SiO2 until the PMMA/MoS2 was released from the substrate. The PMMA/MoS2 was cleaned in three DI water baths before being transferred onto the PET. Once dried, acetone was used to remove the PMMA, leaving the MoS2 on the PET. Characterization/Optical Measurements. The Raman/PL measurements were performed on a Witec Alpha 300RA system using a 532 nm laser line of a frequency-doped Nd:YAG laser with the power kept much lower than 1 mW to avoid sample heating. The spectra were measured in the backscattering configuration using a long working distance 100× objective and either a 600 (PL) or 1800 (Raman) grooves/mm grating. AFM measurements were performed on a Veeco AFM system in tapping mode. For the differential reflectance measurements, the reflectance contrast is defined as 1 − Rs/RPET, where Rs is the reflected light intensity from the MoS2 on PET and RPET is the reflected light intensity from the bare PET substrate. A supercontinuum laser (Fianium) was used as a light source and focused onto the sample with a 100× objective. Reflected light spectra were collected by a spectrometer equipped with a charge coupled device (CCD) (Princeton Instruments, Acton SpectraPro 2300i). Laser power on the sample was less than 10 μW and typical integration time for each spectrum was 1 s. All measurements were done at room temperature. The uniaxial strain was applied using the bending method (described in the Supporting Information). HRTEM imaging and D-STEM59 were performed on a JEOL Model 2010F TEM System. DFT for Raman and Direct Gap Calculations as a Function of Strain. DFT calculations were performed using the ProjectorAugmented Wave60,61 (PAW) pseudopotential implementation of the Vienna Ab Initio Simulation Package62 (VASP), version 5.3.5. Electron exchange and correlation effects were described by the generalized gradient approximation (GGA) functional of Perdew, Burke, and Ernzerhof.63 Wave functions were expanded in a planewave basis set with a cutoff of 350 eV. The unstrained DFT computational cell was hexagonal, defined by a = b = 90°, γ = 120°, a = b = 3.18 Å. The out-of-plane lattice parameter c of the computational cell varied with the number of MoS2 layers N following c = N14.0 Å. Strain was applied in increments of 0.01 from zero to 0.02. Electronic Brillouin zone integration was carried out on a 32 × 32 × 1 Monkhorst−Pack64 k-point grid. Atomic geometries were initially relaxed using the DFT-D3 force field65 to approximately account for dispersion forces. The convergence thresholds were 10−6 and 10−5 eV for electronic and ionic relaxations, respectively. Subsequent linearresponse phonon calculations used an electronic threshold of 10−7 eV and an additional support grid for the evaluation of the augmentation charges (the ADDGRID tag). Raman mode intensities could be calculated using the Python tool by Fonari and Stauffer.66 Molecular Dynamics Simulation Details for Layer Strain Distribution. Large-scale MD simulations were performed using LAMMPS software.67 A five-layer MoS2 sheet used for the uniaxial tension simulation was composed of 7,297,641 atoms with a length in the zigzag direction (zz) of 15.14, 15.12, 5.06, 5.06, and 2.44 μm from the bottom to top, respectively. A periodic boundary was applied in the armchair direction (ac). The interatomic potential for Mo−S systems combined a many-body reactive empirical bond-order (REBO) potential68,69 with a two-body Lennard-Jones (LJ) potential. The time step was chosen as 1 fs. The Nose-Hoover method was adopted in temperature and pressure rescaling. The sample was first equilibrated for 10 ps in NPT ensemble at zero temperature and ambient pressure, then subjected to a uniaxial tensile strain in the zz direction. DFT Simulation Details for Sliding Energy Barrier Calculation. DFT calculations were carried out using projector-augmented wave pseudopotentials60 as implemented in the VASP code.62 The

Figure 8. Energy profile of pristine (black), Vs vacancy-doped (red), and Vs2 vacancy-doped (blue) bilayered MoS2 along the zigzag sliding pathway from AA′ to AB′ and then back to AA′. Four intermediate stacking sequences between AA and AB′ were created and labeled as sliding coordinates (SC) from 1 to 4. SC = 0 corresponded to AA′ and SC = 5 corresponded to AB′. The sliding pathway is indicated by the red arrow in Figure S12b.

agreement with experimental results. The difference in shifts between the applied uniaxial and intrinsic biaxial strain is still not well understood, but additional studies will need to be carried out in order to understand the role of the substrate. We also observed a relative sliding between the layers which resulted in a hysteresis of the in-plane peak position and PL energy upon further application of strain that has not been observed before in these types of structures. We contended that this interlayer sliding was due to defects present in the sample, as was shown through DFT energy barrier calculations for MoS 2 with and without sulfur vacancies. This report demonstrated that the properties of CVD-grown few-layered MoS2 studied here can be tuned under strain as well as, if not better than, its exfoliated monolayered counterpart.

METHODS CVD Growth of MoS2. Our MoS2 atomic layer structures were grown by a vapor-transfer process similar to those reported in the literature.57,58 A schematic of the growth setup is shown in Figure S1a. The inner diameter of the quartz tube is 22 mm. The starting materials were MoO3 (15 mg) and sulfur (1 g) powders that were loaded in alumina crucibles and placed inside the tube. The substrates (285 nm of thermally grown SiO2 on Si(100)) were placed with the polished side facing the MoO3 powder. The tube was then sealed and evacuated to a base pressure of 10 mTorr. After the base pressure was reached, the vacuum pump was shut off and N2 gas was introduced into the tube at a flow rate of 200 sccm. Once the atmospheric pressure was reached, the N2 gas flow was reduced to 10 sccm. At this point, the temperature of the furnace was raised to the growth temperature at a rate of 50 °C/min. Sulfur was evaporated using a secondary furnace held at 200 °C. The secondary furnace was turned on once the temperature inside the main furnace had reached 600 °C. The main furnace was held for 5 min at different growth temperatures (T = 750, 850, 950 °C), after which the heater in the furnace was turned off and the N2 flow rate was increased to 200 sccm for cooling down. It was found that the growth temperature had a significant impact on the MoS2 structures involved, as shown in Figure S1. 3194

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano cutoff energy was 500 eV for the plane-wave expansions, and the MonkhorstPack k-point mesh is 3 × 3 × 1. The atomic positions were optimized until all components of the forces on each atom were reduced to values below 0.02 eV/ Å. The exchange correlation functional was treated with the PerdewBurkeErnzerhof (PBE) generalized gradient approximations (GGA).63 The conventional DFT energy was supplemented with a pairwise interatomic vdW potential which was determined by Tkatchenko and Scheffler (TS-vdW) from nonempirical mean-field electronic structure calculations.71 Two types of intrinsic MoS2 defects72 with monosulfur vacancy (Vs) and disulfur vacancy (V2s) were considered, as shown in Figure S15a. Bilayered MoS2 supercells were created in our simulation with a dimension of 18.96 Å in the zigzag direction and 16.41 Å in the armchair direction, comprised of 216 atoms for pristine MoS2, 215 atoms for Vs-doped MoS2, and 214 atoms for Vs2-doped MoS2. A vacuum space of 20 Å thick was used along the vertical direction to prevent unphysical interactions between adjacent images along this direction. The zigzag pathway started from AA′ stacking, passed through AB′ stacking, and returned to AA′ stacking, which is indicated by the red arrow lines in Figure S15b. Four intermediate stacking sequences between AA′ and AB′ were created and labeled as sliding coordinate (SC) from 1 to 4. In addition, SC = 0 corresponded to AA′ and SC = 5 corresponded to AB′.

reflectance measurements. J.W. and A.D. acknowledge Professor Douglas E. Spearot for giving us the implemented parameters of Mo−S in LAMMPS. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the Army Research Laboratory or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein.

REFERENCES (1) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric Field Effect in Atomically Thin Carbon Films. Science 2004, 306, 666−669. (2) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Katsnelson, M. I.; Grigorieva, I. V.; Dubonos, S. V.; Firsov, A. A. TwoDimensional Gas of Massless Dirac Fermions in Graphene. Nature 2005, 438, 197−200. (3) Geim, A. K.; Novoselov, K. S. The Rise of Graphene. Nat. Mater. 2007, 6, 183−191. (4) Butler, S. Z.; Hollen, S. M.; Cao, L.; Cui, Y.; Gupta, J. A.; Gutierrez, H. R.; Heinz, T. F.; Hong, S. S.; Huang, J.; Ismach, A. F.; Johnston-Halperin, E.; Kuno, M.; Plashnitsa, V. V.; Robinson, R. D.; Ruoff, R. S.; Salahuddin, S.; Shan, J.; Shi, L.; Spencer, M. G.; Terrones, M.; et al. Progress, Challenges, and Opportunities in Two-Dimensional Materials Beyond Graphene. ACS Nano 2013, 7, 2898−2926. (5) Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.; Strano, M. S. Electronics and Optoelectronics of Two-Dimensional Transition Metal Dichalcogenides. Nat. Nanotechnol. 2012, 7, 699−712. (6) Terrones, H.; Del Corro, E.; Feng, S.; Poumirol, J. M.; Rhodes, D.; Smirnov, D.; Pradhan, N. R.; Lin, Z.; Nguyen, M. A. T.; Elias, A. L.; Mallouk, T. E.; Balicas, L.; Pimenta, M. A.; Terrones, M. New First Order Raman-active Modes in Few Layered Transition Metal Dichalcogenides. Sci. Rep. 2014, 4, 4215. (7) Chhowalla, M.; Shin, H. S.; Eda, G.; Li, L. J.; Loh, K. P.; Zhang, H. The Chemistry of Two-Dimensional Layered Transition Metal Dichalcogenide Nanosheets. Nat. Chem. 2013, 5, 263−275. (8) Splendiani, A.; Sun, L.; Zhang, Y. B.; Li, T. S.; Kim, J.; Chim, C. Y.; Galli, G.; Wang, F. Emerging Photoluminescence in Monolayer MoS2. Nano Lett. 2010, 10, 1271−1275. (9) 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. (10) Baugher, B. W. H.; Churchill, H. O. H.; Yang, Y.; JarilloHerrero, P. Intrinsic Electronic Transport Properties of High-Quality Monolayer and Bilayer MoS2. Nano Lett. 2013, 13, 4212−4216. (11) Schmidt, H.; Wang, S.; Chu, L.; Toh, M.; Kumar, R.; Zhao, W.; Neto, A. H. C.; Martin, J.; Adam, S.; Oezyilmaz, B.; Eda, G. Transport Properties of Monolayer MoS2 Grown by Chemical Vapor Deposition. Nano Lett. 2014, 14, 1909−1913. (12) Amani, M.; Chin, M. L.; Birdwell, A. G.; O’Regan, T. P.; Najmaei, S.; Liu, Z.; Ajayan, P. M.; Lou, J.; Dubey, M. Electrical Performance of Monolayer MoS2 Field-Effect Transistors Prepared by Chemical Vapor Deposition. Appl. Phys. Lett. 2013, 102, 193107. (13) Perea-López, N.; Lin, Z.; Pradhan, N.; Iñiguez-Rábago, A.; Elías, A. L.; McCreary, A.; Lou, J.; Ajayan, P. M.; Terrones, H.; Balicas, L.; Terrones, M. CVD-grown Monolayered MoS2 as an Effective Photosensor Operating at Low-Voltage. 2D Mater. 2014, 1, 011004. (14) Wu, C.-C.; Jariwala, D.; Sangwan, V. K.; Marks, T. J.; Hersam, M. C.; Lauhon, L. J. Elucidating the Photoresponse of Ultrathin MoS2 Field-Effect Transistors by Scanning Photocurrent Microscopy. J. Phys. Chem. Lett. 2013, 4, 2508−2513. (15) Britnell, L.; Ribeiro, R. M.; Eckmann, A.; Jalil, R.; Belle, B. D.; Mishchenko, A.; Kim, Y. J.; Gorbachev, R. V.; Georgiou, T.; Morozov, S. V.; Grigorenko, A. N.; Geim, A. K.; Casiraghi, C.; Castro Neto, A. H.; Novoselov, K. S. Strong Light-Matter Interactions in Heterostructures of Atomically Thin Films. Science 2013, 340, 1311−1314.

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.5b04550. Additional information on the growth of terrace MoS2 structures, AFM characterization, differential reflectance measurements, strain estimation, FWHM of Raman E mode during strain, spectrometer resolution, the Bexciton energy as a function of strain, strain on exfoliated flakes, stacking orders of T1 and T2 layers, molecular dynamics simulation details, and the energy sliding barrier for bilayered MoS2 DFT calculation details (PDF)

AUTHOR INFORMATION Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Present Addresses %

(M.A.) Electrical Engineering and Computer Sciences, University of California, Berkeley, CA 94720. & (M.A.) Materials Science Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720. Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS This research was sponsored by the Army Research Laboratory. A.M., M.A., R.N., and M.D. acknowledge the support of the U.S. Army Research Laboratory (ARL) Director’s Strategic Initiative (DSI) program on interfaces in stacked 2D atomic layered materials. A.M. and M.T. acknowledge the support of the U.S. Army Research Office MURI (Grant No. W911NF-111-0362). R.G. and S.K.B. thank the Army Research Office for partial support of this work under STTR Award No. W911NF14-P-0030. J.W. and A.D. acknowledge that this research was accomplished under Cooperative Agreement No. W911NF-142-0059 with the Army Research Laboratory. A.M. thanks Zefang Wang from the Physics Department at The Pennsylvania State University for her help with the differential 3195

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano

(37) 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, 2695−2700. (38) Berkdemir, A.; Gutierrez, H. R.; Botello-Mendez, A. R.; PereaLopez, N.; Elias, A. L.; Chia, C.-I.; Wang, B.; Crespi, V. H.; LopezUrias, F.; Charlier, J.-C.; Terrones, H.; Terrones, M. Identification of Individual and Few Layers of WS2 Using Raman Spectroscopy. Sci. Rep. 2013, 3, 1755. (39) Ni, Z. H.; Yu, T.; Lu, Y. H.; Wang, Y. Y.; Feng, Y. P.; Shen, Z. X. Uniaxial Strain on Graphene: Raman Spectroscopy Study and BandGap Opening. ACS Nano 2008, 2, 2301−2305. (40) Mohiuddin, T. M. G.; Lombardo, A.; Nair, R. R.; Bonetti, A.; Savini, G.; Jalil, R.; Bonini, N.; Basko, D. M.; Galiotis, C.; Marzari, N.; Novoselov, K. S.; Geim, A. K.; Ferrari, A. C. Uniaxial Strain in Graphene by Raman Spectroscopy: G Peak Splitting, Gruneisen Parameters, and Sample Orientation. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 79, 205433. (41) Desai, S. B.; Seol, G.; Kang, J. S.; Fang, H.; Battaglia, C.; Kapadia, R.; Ager, J. W.; Guo, J.; Javey, A. Strain-Induced Indirect to Direct Bandgap Transition in Multi layer WSe2. Nano Lett. 2014, 14, 4592−4597. (42) Das, A.; Pisana, S.; Chakraborty, B.; Piscanec, S.; Saha, S. K.; Waghmare, U. V.; Novoselov, K. S.; Krishnamurthy, H. R.; Geim, A. K.; Ferrari, A. C.; Sood, A. K. Monitoring Dopants by Raman Scattering in an Electrochemically Top-Gated Graphene Transistor. Nat. Nanotechnol. 2008, 3, 210−215. (43) Ferrari, A. C. Raman Spectroscopy of Graphene and Graphite: Disorder, Electron-Phonon Coupling, Doping and Nonadiabatic Effects. Solid State Commun. 2007, 143, 47−57. (44) Mak, K. F.; He, K. L.; Lee, C.; Lee, G. H.; Hone, J.; Heinz, T. F.; Shan, J. Tightly Bound Trions in Monolayer MoS2. Nat. Mater. 2012, 12, 207−211. (45) 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. (46) Elias, A. L.; Perea-Lopez, N.; Castro-Beltran, A.; Berkdemir, A.; Lv, R. T.; Feng, S. M.; Long, A. D.; Hayashi, T.; Kim, Y. A.; Endo, M.; Gutierrez, H. R.; Pradhan, N. R.; Balicas, L.; Mallouk, T. E.; LopezUrias, F.; Terrones, H.; Terrones, M. Controlled Synthesis and Transfer of Large-Area WS2 Sheets: From Single Layer to Few Layers. ACS Nano 2013, 7, 5235−5242. (47) Matthewson, M. J.; Kurkjian, C. R.; Gulati, S. T. Strength Measurement of Optical Fibers By Bending. J. Am. Ceram. Soc. 1986, 69, 815−821. (48) Amani, M.; Chin, M. L.; Mazzoni, A. L.; Burke, R. A.; Najmaei, S.; Ajayan, P. M.; Lou, J.; Dubey, M. Growth-Substrate Induced Performance Degradation in Chemically Synthesized Monolayer MoS2 Field Effect Transistors. Appl. Phys. Lett. 2014, 104, 203506. (49) Liu, Z.; Amani, M.; Najmaei, S.; Xu, Q.; Zou, X.; Zhou, W.; Yu, T.; Qiu, C.; Birdwell, A. G.; Crowne, F. J.; Vajtai, R.; Yakobson, B. I.; Xia, Z.; Dubey, M.; Ajayan, P. M.; Lou, J. Strain and structure heterogeneity in MoS2 atomic layers grown by chemical vapour deposition. Nat. Commun. 2014, 5, 5246. (50) Najmaei, S.; Liu, Z.; Zhou, W.; Zou, X. L.; Shi, G.; Lei, S. D.; Yakobson, B. I.; Idrobo, J. C.; Ajayan, P. M.; Lou, J. Vapour Phase Growth and Grain Boundary Structure of Molybdenum Disulphide Atomic Layers. Nat. Mater. 2013, 12, 754−759. (51) Dong, L.; Namburu, R. R.; O’Regan, T. P.; Dubey, M.; Dongare, A. M. Theoretical Study on Strain-Induced Variations in Electronic Properties of Monolayer MoS2. J. Mater. Sci. 2014, 49, 6762−6771. (52) Scalise, E.; Houssa, M.; Pourtois, G.; Afanas’ev, V. V.; Stesmans, A. First-Principles Study of Strained 2D MoS2. Phys. E (Amsterdam, Neth.) 2014, 56, 416−421. (53) Yun, W. S.; Han, S. W.; Hong, S. C.; Kim, I. G.; Lee, J. D. Thickness and Strain Effects on Electronic Structures of Transition Metal Dichalcogenides: 2H-MX2 Semiconductors (M = Mo, W; X = S, Se, Te). Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 85, 033305.

(16) Eda, G.; Yamaguchi, H.; Voiry, D.; Fujita, T.; Chen, M. W.; Chhowalla, M. Photoluminescence from Chemically Exfoliated MoS2. Nano Lett. 2011, 11, 5111−5116. (17) Mak, K. F.; He, K. L.; Shan, J.; Heinz, T. F. Control of Valley Polarization in Monolayer MoS2 by Optical Helicity. Nat. Nanotechnol. 2012, 7, 494−498. (18) Mak, K. F.; McGill, K. L.; Park, J.; McEuen, P. L. The Valley Hall Effect in MoS2 Transistors. Science 2014, 344, 1489−1492. (19) Xu, M. S.; Liang, T.; Shi, M. M.; Chen, H. Z. Graphene-Like Two-Dimensional Materials. Chem. Rev. 2013, 113, 3766−3798. (20) 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, 7931−7936. (21) Li, T. S. Ideal Strength and Phonon Instability in Single-Layer MoS2. Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 85, 5. (22) Johari, P.; Shenoy, V. B. Tuning the Electronic Properties of Semiconducting Transition Metal Dichalcogenides by Applying Mechanical Strains. ACS Nano 2012, 6, 5449−5456. (23) Scalise, E.; Houssa, M.; Pourtois, G.; Afanas’ev, V. V.; Stesmans, A. Strain-Induced Semiconductor to Metal Transition in the TwoDimensional Honeycomb Structure of MoS2. Nano Res. 2012, 5, 43− 48. (24) Bertolazzi, S.; Brivio, J.; Kis, A. Stretching and Breaking of Ultrathin MoS2. ACS Nano 2011, 5, 9703−9709. (25) Castellanos-Gomez, A.; Poot, M.; Steele, G. A.; van der Zant, H. S. J.; Agrait, N.; Rubio-Bollinger, G. Elastic Properties of Freely Suspended MoS2 Nanosheets. Adv. Mater. 2012, 24, 772−775. (26) Amani, M.; Burke, R. A.; Proie, R. M.; Dubey, M. Flexible Integrated Circuits and Multifunctional Electronics Based on Single Atomic Layers of MoS2 and Graphene. Nanotechnology 2015, 26, 115202. (27) Kim, S.; Konar, A.; Hwang, W.-S.; Lee, J. H.; Lee, J.; Yang, J.; Jung, C.; Kim, H.; Yoo, J.-B.; Choi, J.-Y.; Wan Jin, Y.; Yoon Lee, S.; Jena, D.; Choi, W.; Kim, K. High-Mobility and Low-Power Thin-Film Transistors Based on Multilayer MoS2 Crystals. Nat. Commun. 2012, 3, 1011. (28) Kang, J.; Liu, W.; Banerjee, K. High-Performance MoS2 Transistors with Low-Resistance Molybdenum Contacts. Appl. Phys. Lett. 2014, 104, 093106. (29) Das, S.; Appenzeller, J. Screening and Interlayer Coupling in Multilayer MoS2. Phys. Status Solidi RRL 2013, 7, 268−273. (30) Rice, C.; Young, R. J.; Zan, R.; Bangert, U.; Wolverson, D.; Georgiou, T.; Jalil, R.; Novoselov, K. S. Raman-Scattering Measurements and First-Principles Calculations of Strain-Induced Phonon Shifts in Monolayer MoS2. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 081307. (31) Conley, H. J.; Wang, B.; Ziegler, J. I.; Haglund, R. F., Jr.; Pantelides, S. T.; Bolotin, K. I. Bandgap Engineering of Strained Monolayer and Bilayer MoS2. Nano Lett. 2013, 13, 3626−3630. (32) He, K.; Poole, C.; Mak, K. F.; Shan, J. Experimental Demonstration of Continuous Electronic Structure Tuning via Strain in Atomically Thin MoS2. Nano Lett. 2013, 13, 2931−2936. (33) Wang, Y.; Cong, C.; Qiu, C.; Yu, T. Raman Spectroscopy Study of Lattice Vibration and Crystallographic Orientation of Monolayer MoS2 under Uniaxial Strain. Small 2013, 9, 2857−2861. (34) Hui, Y. Y.; Liu, X.; Jie, W.; Chan, N. Y.; Hao, J.; Hsu, Y.-T.; Li, L.-J.; Guo, W.; Lau, S. P. Exceptional Tunability of Band Energy in a Compressively Strained Trilayer MoS2 Sheet. ACS Nano 2013, 7, 7126−7131. (35) Ferrari, A. C.; Meyer, J. C.; Scardaci, V.; Casiraghi, C.; Lazzeri, M.; Mauri, F.; Piscanec, S.; Jiang, D.; Novoselov, K. S.; Roth, S.; Geim, A. K. Raman Spectrum of Graphene and Graphene Layers. Phys. Rev. Lett. 2006, 97, 187401. (36) 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, 1385−1390. 3196

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197

Article

ACS Nano (54) Dong, L.; Dongare, A. M.; Namburu, R. R.; O’Regan, T. P.; Dubey, M. Theoretical Study on Strain Induced Variations in Electronic Properties of 2H-MoS2 Bilayer Sheets. Appl. Phys. Lett. 2014, 104, 053107. (55) Gong, L.; Young, R. J.; Kinloch, I. A.; Riaz, I.; Jalil, R.; Novoselov, K. S. Optimizing the Reinforcement of Polymer-Based Nanocomposites by Graphene. ACS Nano 2012, 6, 2086−2095. (56) Liu, K. H.; Zhang, L. M.; Cao, T.; Jin, C. H.; Qiu, D. A.; Zhou, Q.; Zettl, A.; Yang, P. D.; Louie, S. G.; Wang, F. Evolution of Interlayer Coupling in Twisted Molybdenum Disulfide Bilayers. Nat. Commun. 2014, 5, 4966. (57) van der Zande, A. M.; Huang, P. Y.; Chenet, D. A.; Berkelbach, T. C.; You, Y. M.; Lee, G. H.; Heinz, T. F.; Reichman, D. R.; Muller, D. A.; Hone, J. C. Grains and Grain Boundaries in Highly Crystalline Monolayer Molybdenum Disulphide. Nat. Mater. 2013, 12, 554−561. (58) Liu, Y.; Ghosh, R.; Wu, D.; Ismach, A.; Ruoff, R.; Lai, K. Mesoscale Imperfections in MoS2 Atomic Layers Grown by a Vapor Transport Technique. Nano Lett. 2014, 14, 4682−4686. (59) Ganesh, K. J.; Kawasaki, M.; Zhou, J. P.; Ferreira, P. J. D-STEM: A Parallel Electron Diffraction Technique Applied to Nanomaterials. Microsc. Microanal. 2010, 16, 614−621. (60) Blochl, P. E. Projector Augmented-Wave Method. Phys. Rev. B: Condens. Matter Mater. Phys. 1994, 50, 17953−17979. (61) Kresse, G.; Joubert, D. From Ultrasoft Pseudopotentials to the Projector Augmented-Wave Method. Phys. Rev. B: Condens. Matter Mater. Phys. 1999, 59, 1758−1775. (62) Kresse, G.; Furthmuller, J. Efficient Iterative Schemes for ab initio Total-Energy Calculations Using a Plane-Wave Basis Set. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 11169−11186. (63) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (64) Monkhorst, H. J.; Pack, J. D. Special Points for Brillouin-Zone Integrations. Phys. Rev. B 1976, 13, 5188−5192. (65) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A Consistent and Accurate ab initio Parametrization of Density Functional Dispersion Correction (DFT-D) for the 94 Elements H-Pu. J. Chem. Phys. 2010, 132, 154104. (66) Fonari, A.; Stauffer, S. vasp_raman.py; https://github.com/ raman-sc/VASP/ (accessed Apr 7, 2015). (67) Plimpton, S. Fast Parallel Algorithms for Short-Range Molecular Dynamics. J. Comput. Phys. 1995, 117, 1−19. (68) Liang, T.; Phillpot, S. R.; Sinnott, S. B. Parametrization of a Reactive Many-Body Potential for Mo-S Systems. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 79, 14. (69) Stewart, J. A.; Spearot, D. E. Atomistic Simulations of Nanoindentation on the Basal Plane of Crystalline Molybdenum Disulfide (MoS2). Modell. Simul. Mater. Sci. Eng. 2013, 21, 045003. (70) Tao, P.; Guo, H.-H.; Yang, T.; Zhang, Z.-D. Stacking Stability of MoS2 Bilayer: An ab initio Study. Chin. Phys. B 2014, 23, 106801. (71) Tkatchenko, A.; Scheffler, M. Accurate Molecular Van Der Waals Interactions from Ground-State Electron Density and FreeAtom Reference Data. Phys. Rev. Lett. 2009, 102, 073005. (72) Zhou, W.; Zou, X.; Najmaei, S.; Liu, Z.; Shi, Y.; Kong, J.; Lou, J.; Ajayan, P. M.; Yakobson, B. I.; Idrobo, J.-C. Intrinsic Structural Defects in Monolayer Molybdenum Disulfide. Nano Lett. 2013, 13, 2615− 2622. (73) McCreary, A.; Berkdemir, A.; Wang, J.; Nguyen, M. A.; Elías, A. L.; Pérea-López, N.; Fujisawa, K.; Kabius, B.; Carozo, V.; Cullen, D. A.; Mallouk, T. E.; Zhu, J. Distinct Photoluminescence and Raman Spectroscopy Signatures for Identifying Highly Crystalline WS2.J. Mater. Sci. 2016, in press.

3197

DOI: 10.1021/acsnano.5b04550 ACS Nano 2016, 10, 3186−3197