Unveiling Three-Dimensional Stacking Sequences ... - ACS Publications

Unveiling Three-Dimensional Stacking Sequences of 1T Phase MoS2 Monolayers by Electron Diffraction Ziqian Wang,†,‡ Shoucong Ning,§ Takeshi Fujita,‡ Akihiko Hirata,‡ and Mingwei Chen*,†,‡,⊥ †

Department of Materials Science, Graduate School of Engineering, and ‡WPI Advanced Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan § Department of Mechanical and Aerospace Engineering, School of Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong SAR ⊥ CREST, JST, 4-1-8 Honcho Kawaguchi, Saitama 332-0012, Japan S Supporting Information *

ABSTRACT: The phase transition between semiconducting 1H to metallic 1T phases in monolayered transition metal dichalcogenides (TMDs) essentially involves threedimensional (3D) structure changes of asymmetric relocations of S atoms at the top and bottom of the oneunit-cell crystals. Even though the phase transition has a profound influence on properties and applications of 2D TMDs, a viable approach to experimentally characterize the stacking sequences of the vertically asymmetrical 1T phase is still not available. Here, we report an electron diffraction method based on dynamic electron scattering to characterize the stacking sequences of 1T MoS2 monolayers. This study provides an approach to unveil the 3D structure of 2D crystals and to explore the underlying mechanisms of semiconductor-to-metal transition of monolayer TMDs. KEYWORDS: two-dimensional materials, transition metal dichalcogenide, phase interface, MoS2, electron diffraction

M

appearance of inversion symmetry results in no property changes after a 60° rotation in reality, equivalent to a six-fold rotational symmetry. However, in the case of the coexistence of the 1T phase with the parent 1H phase, separated by a coherent metal/ semiconductor interface, the breaking of reflection symmetry in 1T matters. As schematically shown in Figure 1c,d, the phase transition via the top (0001) or bottom (0001̅) S layer gliding along the same direction gives rise to two 1T phases with a 60° rotation angle difference, and more importantly, the resultant two 1T/1H metal−semiconductor interfaces are chiral to each other (Figure S1, Supporting Information) and are expected to have distinct properties in spintronics and electromagnetics of 2D TMDs. Therefore, the interfacial structure and related 1H↔ 1T transition in the 2D TMDs are essentially 3D issues in terms of the stacking sequences of the three-atomic-layer crystals. Apparently, revealing the 3D structure of the 2D TMD crystals is critical for understanding the underlying mechanisms of the metal−semiconductor transition and for designing 2D TMD

onolayer crystals, such as graphene and transition metal dichalcogenides (TMDs), are usually referred to as two-dimensional (2D) materials because they are free from repeating crystal lattices in the out-of-plane direction. This feature gives rise to tremendous properties in electronics, photonics, catalysis, mechanics, etc.1−5 Recently, three-dimensional (3D) structural features of 2D materials have attracted significant attention because of a great number of functionalities from 3D heterostructures formed by stacking or stitching 2D materials3,6−11 and, more importantly, the structure transition of TMD monolayers from semiconducting 1H to metallic 1T by relocating top or bottom chalcogen atoms with vertical symmetry breaking.12−16 The equilibrium trigonal prismatic 1H phase, such as MoS2 monolayers, belongs to the D3h group and has a reflection symmetry with the mirror plane in the in-plane directions and three-fold symmetry from the invariance of the structure after in-plane rotation by 120° (Figure 1a). In contrast, the octahedral 1T phase is a member of the D3d group and has no reflection symmetry but instead has an inversion center (Figure 1b).16 Even though the transition from 1H to 1T phase leads to the breaking of reflection symmetry in the in-plane directions, this symmetry breaking has little meaning in geometry when the 1T phase is treated alone because the © 2016 American Chemical Society

Received: September 3, 2016 Accepted: October 27, 2016 Published: October 27, 2016 10308

DOI: 10.1021/acsnano.6b05958 ACS Nano 2016, 10, 10308−10316

Article

www.acsnano.org

Article

ACS Nano

Figure 1. Structure and symmetry of 1H and 1T phase MoS2 monolayer and their interfaces. Symmetry elements of 1H MoS2 are partially shown in (a) and 1T MoS2 in (b). 1H has an in-plane mirror plane, whereas 1T has an inversion center. Blue and yellow spheres indicate Mo and S atoms, respectively. (c) Glide of (0001) S and (d) glide of (0001̅) S layer in the 1H to 1T phase transition results in different interfacial structures chiral to each other. Yellow and cyan spheres indicate (0001) S and (0001̅) S atoms, respectively.

diffraction vector variation caused by the symmetry degeneration from six-fold of bulk phases to three-fold of monolayers.21,26,33 In this work, we developed an electron diffraction method with the consideration of dynamic electron scattering to unveil the stacking sequences of the 1T phase of monolayer MoS2.

electronic devices with 1T/1H interfaces as metal/semiconductor contacts.13,16−19 High-resolution transmission electron microscopy (HRTEM) and scanning transmission electron microscopy (STEM) have been widely used as powerful tools to characterize the atomic structures of 2D materials.20−28 However, either HRTEM or STEM images are basically the projection of 3D structures and have poor spatial resolution in the vertical direction parallel to electron beams. Although the recently developed atomic counting algorithms based on high-angle annular dark-field (HAADF) STEM are capable of reconstructing 3D atomic structures of crystals,29−32 and it is not practical for this method to identify the stacking sequences of the vertically asymmetric 1T phase because there are no superposed atoms along the ⟨0001⟩ direction in the phase. In fact, monolayer TMDs do not require atomic counting. Alternatively, electron diffraction involves 3D information by means of electron wave interference with singleunit-cell-thick TMDs, and extensive perspectives of 3D structures of 2D crystals can be expected. Particularly, the combination of selected area electron diffraction (SAED) with dark-field TEM is capable of revealing structure and crystal orientation of monolayer TMDs on a large scale by utilizing the

RESULTS AND DISCUSSION Reciprocal lattice configurations of 1H and 1T MoS2 monolayers are shown schematically in Figure S2. To be specific, both reciprocal lattices have a repeating unit indicated by the green rhombus area in Figure S2, which can be explained by the kinematic diffraction theory. Accordingly, the diffraction intensity is given by |G2D(K)|2 × |F(K)|2, where G2D(K) is the Laue function and F(K) is the structural factor. Since 1H and 1T monolayers have the same 2D Laue function, G2D(K), the term determining the diffraction intensity is the absolute square of the structural factor of each phase F1H or F1T. When the lattice vectors of 1H and 1T are defined as shown in Figure 1a,b, the absolute squares of structural factors of the (hkil) diffraction are derived to be

⎧ ′ + 2f S′ )2 + (f Mo ″ + 2f S″ )2 (f Mo h − k ≡ 0 (mod 3) ⎪ ⎪ ′ − f S′ + 3 f S″ )2 + (f Mo ″ − 3 f S′ − f S″ )2 h − k ≡ 1 (mod 3) |F1H|2 ≈ ⎨(f Mo ⎪ ⎪(f ′ − f ′ − 3 f ″ )2 + (f ″ + 3 f ′ − f ″ )2 h − k ≡ 2 (mod 3) ⎩ Mo S S Mo S S

⎧(f ′ + 2f ′ )2 + (f ″ + S Mo ⎪ Mo ⎪ 2 2 ′ − f S′ ) + (f Mo ″ − |F1T| ≈ ⎨ (f Mo ⎪ ⎪ (f ′ − f ′ )2 + (f ″ − ⎩ Mo S Mo

10309

(1)

2f S″ )2 h − k ≡ 0 (mod 3) f S″ )2

h − k ≡ 1 (mod 3)

f S″ )2

h − k ≡ 2 (mod 3)

(2)

DOI: 10.1021/acsnano.6b05958 ACS Nano 2016, 10, 10308−10316

Article

ACS Nano

Figure 2. Illustration of the origin of non-centrosymmetry in the diffraction space of centrosymmetric 1T monolayer MoS2. (a) 1T monolayer solely has an inversion center. (b) Inversion center disappears when the 1T monolayer is viewed with an electron beam (the diffraction optics) due to the asymmetric boundary conditions on incident and exit sides. (c) Breaking of inversion symmetry by the diffraction optics leads to the non-centrosymmetry in diffraction patterns even though the 1T monolayer itself is centrosymmetric.

in 1H and 1T structure, respectively. Herein, fatom = fatom ′ + ifatom ″ is the atomic shape factor containing real and imaginary parts, and “a  b (mod c)” means that b is the remainder of the division of a by c. Distribution of the spots defined by 0, 1, or 2 (mod 3) in the reciprocal space is shown in Figure S2. However, electron diffraction is not a sample-only issue as simplified by the kinematic theory above but is a dynamic process where multiple Bloch waves inside the sample are excited during an electron beam penetrating a crystal, or in an equivalent real space description, electrons might be multiply scattered in the sample due to its particle nature.34 Deducing the diffraction contrast of each individual diffraction beams has to resort to solving the Schrodinger equation of the electron optics from incident to exit. However, the boundary conditions at the incident and exit sides of the monolayer crystals are resultantly different because of only one incident beam but multiple exit diffraction beams, which put the two S layers in different conditions with regard to the electron

beam and could give rise to the possibility of resolving the difference between two S layers as vertical information on 2D crystals. Specifically, the inversion symmetry of the 1T monolayer disappears (Figure 2) from the viewpoint of the whole diffraction optics containing not only the 1T monolayer sample but also the incident and exit electron beams. This symmetry breaking could cause noticeable inaccuracy of the diffraction intensities predicted by the conventional kinematic theory. Therefore, the dynamic theory of electron diffraction is required. In this study, we applied “effective atomic scattering factors” to discriminate the top and bottom S layers in monolayer MoS2 by coercively combining the kinematic and dynamic theories. The effective atomic scattering factors of S at incident and exit sides f Sin = f S′ in + if S″in and f Sex = f S′ ex + if S″ex are not equivalent, and eq 1 and eq 2 become as follows:

⎧ ′ + Σf S′ )2 + (f Mo ″ + Σf S″ )2 (f Mo h − k ≡ 0 (mod 3) ⎪ ⎪ ⎛ ⎞2 ⎛ ⎞2 ⎪ ⎪ ⎜f ′ − 1 Σf ′ + 3 Σf ″ ⎟ + ⎜f ″ − 3 Σf ′ − 1 Σf ″ ⎟ h − k ≡ 1 (mod 3) 2 S S |F1H| ≈ ⎨ ⎝ Mo ⎠ ⎝ Mo 2 S 2 2 2 S⎠ ⎪ ⎪⎛ ⎞2 ⎛ ⎞2 1 3 3 1 ⎪ ⎜f ′ − Σ f ′ − ″ ″ ′ ″ Σ f + f + Σ f − Σ f ⎟ ⎜ ⎟ h − k ≡ 2 (mod 3) S S ⎪ ⎝ Mo ⎠ ⎝ Mo 2 S 2 2 2 S⎠ ⎩

(3)

⎧ ′ + Σf S′ )2 + (f Mo ″ + Σf S″ )2 (f Mo h − k ≡ 0 (mod 3) ⎪ ⎪ ⎛ ⎞2 ⎛ ⎞2 1 3 3 1 ⎪ ⎪ ″ − Δf S″ ⎟ + ⎜f Mo Δf S′ − Σf S″ ⎟ h − k ≡ 1 (mod 3) ⎜f ′ − Σf S′ + |F1T|2 ≈ ⎨ ⎝ Mo ⎠ ⎝ ⎠ 2 2 2 2 ⎪ 2 ⎪⎛ ⎞ ⎛ ⎞2 1 3 3 1 ⎪ ⎜f ′ − Σf ′ − ″ + Δf S″ ⎟ + ⎜f Mo Δf S′ − Σf S″ ⎟ h − k ≡ 2 (mod 3) ⎪ ⎝ Mo ⎠ ⎝ ⎠ 2 S 2 2 2 ⎩

(4)

Here, ∑f ′S(″) = f ′S(in″) + f ′S(ex″) and Δf ′S(″) = f ′S(in″) − f ′S(ex″). Note that Δf S′(″) is nonzero since two S layers are not effectively equivalent in a real diffraction. Consequently, neither of diffraction peaks of h − k  1, 2 or 3 (mod 3) in 1H or 1T shows the same intensity, and this directly leads to the non-centrosymmetry of real diffraction patterns in either case. Especially for the case of 1T MoS2 monolayers, contrary to the centrosymmetric diffraction

patterns deduced by the kinematic theory in eq 2, noncentrosymmetric diffraction patterns should be observed and h − k  1 (mod 3) diffractions have different intensity from h − k  2 (mod 3) diffractions with the same distance from the [0000] peak. The symmetry-breaking electron optics of 2D crystals in the electron diffraction mode could be utilized as a tool to resolve

10310

DOI: 10.1021/acsnano.6b05958 ACS Nano 2016, 10, 10308−10316

Article

ACS Nano

Figure 3. Quantitative analyses of SAED of 1H phase MoS2. (a) SAED pattern taken from a 1H region (200 nm in diameter) with ⟨101̅0⟩ and ⟨1120̅ ⟩ diffraction peaks noted on the image. Integrated intensity of each peak over the corresponding square regions is shown by the columns in (b) and is normalized by the average intensity of the ⟨112̅0⟩ diffraction peaks. Black scatters and lines in (b) show the simulated values of each diffraction peak assuming the sample is ideally clean and flat. Integrated intensity of ⟨101̅0⟩ diffraction peaks shows two non-equivalent groups: the h − k  1 (mod 3) group (101̅0, 01̅10, and 1̅100), which has higher intensity, and the h − k  2 (mod 3) group (11̅00, 1̅010, and 011̅0), which has lower intensity. Almost no observable difference in the intensities of six ⟨112̅0⟩ diffraction peaks. Associated with corresponding HAADFSTEM image (c), directions of brighter (blue solid lines and arrows) and dimmer (red dot lines and arrows) peaks are noted in the atomic model in (d). Atomic model in the right panel of (d) is the side view of the atomic model in the left panel of (d) viewed along the direction indicated by the black dotted line and glasses symbol.

Figure 4. Quantitative analyses of SAED of 1T phase MoS2. Notations and color stipulations are the same as that in Figure 3. Integrated intensity of each peak in (b) is also the normalized value by the average intensity of six ⟨1120̅ ⟩ diffraction peaks. Black scatters and lines in (b) show the simulated values of each diffraction peak assuming the sample is ideally clean and flat. Brighter group h − k  2 (mod 3) and dimmer group h − k  1 (mod 3) are inversed in comparison with the case of 1H shown in Figure 3.

structural information in the out-of plane direction of 2D crystals involved in the diffraction patterns.

As a model system, two-phase MoS2 monolayers, prepared by chemical vapor deposition (CVD),35 were placed on holey 10311

DOI: 10.1021/acsnano.6b05958 ACS Nano 2016, 10, 10308−10316

Article

ACS Nano

Figure 5. Bright-field and dark-field TEM imaging taken from the co-lattice 1H/1T region. (a) Schematic illustration of the reciprocal space of monolayer 1H/1T MoS2. Red point in the middle indicates the center, and the green shaded area shows a representative repeating unit containing four non-equivalent points of 11̅ 00, 0110̅ , 12̅ 10̅ , and 1120̅ . Bright-field and dark-field TEM micrographs of the same region imaged by selecting (b) 0000, (c) 1̅100, (d) 011̅0, (e) 1̅21̅0, and (f) 112̅0 diffraction vectors shown in the insets of the images (scale bar: 0.2 μm). Images by 0000, 12̅ 10̅ , and 1120̅ show no difference in 1H and 1T regions corresponding to (b), (e) and (f), whereas images by 11̅ 00 and 0110̅ exhibit visible contrast difference, i.e., brighter 1H in (c) and brighter 1T in (d). (g) Profiles across the phase interface along the yellow arrows shown in (c−f). White arrows in (g) indicate the position of the phase interface, and a small hill appearing next to the interface on the 1H side is indicated by the black arrows.

intensity higher than that for the h − k  2 (mod 3) group ([11̅00], [1̅010], and [011̅0]) in 1H (red bars in Figure 3b), and the opposite contrast relation is observed in 1T (red bars in Figure 4b). This splitting of the ⟨101̅0⟩ peak intensity is consistent with previous observations in 1H TMD monolayers, whereas no such reports on the 1T monolayer are found.21,26,33 According to our experiments, the integrated intensity ratio of brighter and dimmer ⟨101̅0⟩ peaks is 1.22 for 1H and 1.35 for 1T phase of free-standing MoS2 monolayers. Thus, the intensity splitting of 1T MoS2 is more evident compared with that for 1H MoS2. We simulated the diffraction patterns based on multislice algorithm resorting to the dynamic diffraction theory by numerically solving the Schrodinger equation. Simulated results assuming the sample is ideally clean and flat are plotted as black dots in Figures 3b and 4b, which confirm the experimental observations of the intensity splitting in the 2D crystals. The brighter/dimmer ratios of the ⟨101̅0⟩ peaks predicted by the simulations are 1.08 for 1H and 1.15 for 1T at room temperature. Discrepancy in the normalized intensity values between experiments and simulations in 1T ⟨101̅0⟩ diffraction peaks (Figure 4b) could be due to the surface rippling of the freestanding MoS2 (S4, Supporting Information).21,36,37 According to the experimental and simulated results, there is a well-defined

carbon-coated Cu grids for structure characterization. The TEM and STEM images of the region in vicinity of a 1H/1T interface are shown in Figure S3. SAED patterns were taken from a 200 nm diameter region containing exclusively the 1H or 1T phase, as illustrated in Figures 3a and 4a, respectively. The diffraction peaks in Figures 3a and 4a are indexed according to the lattice vectors shown in their corresponding HAADF-STEM images in Figures 3c and 4c and structure models in Figures 3d and 4d. Considering the inevitable influences on the intensity profiles of diffraction peaks by actual acquisition conditions, such as focus parameters and charge-coupled device performances, we evaluated the integrated intensity of each diffraction peak rather than peak height or peak width, and the analytical results are shown in Figures 3b and 4b. The peak intensity values are integrations over the areas enclosed by the squares with a constant area marked in Figures 3a and 4a and normalized by the average value of six ⟨112̅0⟩ diffraction peaks in either the 1H or the 1T case. The six ⟨112̅0⟩ peaks in both diffraction patterns have a nearly identical intensity (green bars in Figures 3b or 4b), verifying that the incident electron beam perfectly aligns with the [0001] zone orientation of the 2D crystals. In contrast to the ⟨112̅0⟩ peaks, the intensities of six ⟨101̅0⟩ diffraction peaks in both 1H and 1T show two non-equivalent groups: the h − k  1 (mod 3) group, including [101̅0], [01̅10], and [1̅100], has an 10312

DOI: 10.1021/acsnano.6b05958 ACS Nano 2016, 10, 10308−10316

Article

ACS Nano

Figure 6. Illustration on the identification of stacking sequences of 1T phase by electron diffraction. Electron diffraction pattern appears upon an electron beam penetrating the MoS2 monolayer perpendicularly to the basal plane. Brighter and dimmer groups of diffraction peaks become signatures of 3D crystallography of both 1H and 1T phases. Inversion of the brightness of the two groups indicates the occurrence of relative glide of the bottom (0001̅) S layer (from AbA to AbC stack, middle and left panel), whereas identical brightness distribution indicates the relative glide of the top (0001) S layer (from AbA to CbA stack, middle and right panel).

correspondence between diffraction vectors and the atomic configuration, as shown in the enlarged STEM images and the atomic models (Figures 3c,d and 4c,d), in which brighter and dimmer directions are noted. Accordingly, the brighter/dimmer relation is dependent solely on the atomic configuration of the Mo and exit side (0001)̅ S layers in 1H or 1T structure; that is, brighter diffraction vectors are parallel to the in-plane projection of (0001̅) S−Mo bonds toward the Mo direction and dimmer vectors toward S, and the relative position of (0001) S on the incident side only modifies the intensities very slightly. Taking advantage of distinct intensity variations in ⟨1010̅ ⟩ diffraction peaks in 1H and 1T phases, the two phases and corresponding interfaces can be visualized by dark-field (DF) TEM through the selection of one specific diffraction vector for imaging. Figure 5a shows a typical repeating unit in the reciprocal lattice (green rhombus), which involves four diffraction spots: 11̅ 00, 0110̅ , 12̅ 10̅ , and 1120̅ in either the 1H or the 1T phase. A bright-field (BF) TEM image using the direct beam (0000) is shown in Figure 5b, and DF-TEM images by (1̅100), (011̅0), (1̅21̅0), and (112̅0) are shown in Figure 5c−f with their corresponding diffraction vectors marked by yellow circles in the inseted SAED patterns. The same contrast of 1H and 1T regions in Figure 5b,e,f manifests the equivalence of the center and all ⟨112̅0⟩ diffraction peaks. Visible contrast difference in Figure 5c,d confirms the intensity variation of ⟨101̅0⟩ diffraction peaks between 1H and 1T; that is, 1H shows higher intensity over 1T for peak 1̅100, while the 011̅0 peak gives an opposite contrast (Figures 3 and 4). Due to phase identification by DF-TEM, 1T/ 1H interfaces can be readily identified at the microscopic scale (Figure S3c,d). Figure 5g shows the intensity profiles across the phase interface, as indicated by the yellow square in Figure 5c−f for (1̅100), (011̅0), (1̅21̅0), and (112̅0) DF-TEM images. The xaxis in Figure 5g indicates the distance along the yellow arrows, and the y-axis is the integrated intensity along the dotted line direction. A slight dip at the interface, marked by white arrowheads, can be seen in all of the profiles, which may be caused by diffuse scattering from interface defects and thereby

gives a lower intensity in the diffraction images. Moreover, small hills can always be observed next to the interface at 1H sides, as indicated by the black arrows. The additional elastic scattering from the semiconductor side has not been well understood. At the 1H/1T phase interfaces of monolayer MoS2, the stacking order changes from AbA stacked 1H to 1T by two possible paths with different stacking sequences: gliding the top (0001) S layer to produce CbA stacked 1T or gliding the bottom (0001̅) S layer to produce AbC stacked 1T. The two stacking structures of 1T MoS2 coincide with each other after a relative rotation of 60°, as described in Figure 1c,d. Figure 6 schematically illustrates the diffraction pattern-forming process as the electron beam penetrates the MoS2 monolayer perpendicularly to its basal plane in the TEM. After the incident beam interacts with crystal lattices, brighter and dimmer groups of diffraction peaks arise and become signatures of 3D crystallographic orientations of both 1H and 1T phases. In the case of the (0001) S glide (right and middle panel of Figure 6), brighter and dimmer peak directions are identical to that of AbA stacked 1H phase and CbA stacked 1T phase. On the contrary, in the case of the (0001̅) S glide (left and middle panel of Figure 6), the brighter/dimmer direction inversion is observable upon the phase transition between AbA stacked 1H and CbA stacked 1T. Hence, by simply recording and analyzing changes of diffraction patterns across a phase interface, the stacking sequence change in the resultant 1T phase and the chirality of the 1T/1H interface can be extracted. By using this method, the 3D structure of the 1T phase shown in Figure 5 can be determined to be CbA stacking. Thus, the transition from 1H to 1T in this domain takes place by the gliding of the (0001̅) S layer, which directly contacts the glass substrate during CVD growth. The same stacking sequences were observed from five 1T phase domains that were found from about 50 grains in a greater than 100 × 100 μm2 sampling area without exception. This observation indicates that the 1H to 1T phase transition in the CVD-grown TMD monolayers may be associated with the weak interaction between the films and substrates during growth and cooling. 10313

DOI: 10.1021/acsnano.6b05958 ACS Nano 2016, 10, 10308−10316

Article

ACS Nano The “2D materials” is an ideal platform to investigate fundamental physical phenomena, including electron diffraction, of crystals. Since MoS2 monolayers are only three atomic layers thick and the 1H or 1T phase only differs by the stacking order of the three atomic layers, the 1H/1T two-phase MoS2 is a good system for studying the scattering of atomic layers with highenergy electron waves. As described in Figure 6, the changes of stacking order of only one atomic layer from 1H to 1T lead to the reversion of diffraction contrast. Moreover, the intensity variation of different ⟨101̅0⟩ vectors from the monolayer thick 1T phase crystals indicates that the approximation of the conventional kinematic theory of electron diffraction cannot stand due to multiple scattering even down to the sample thickness of the three-atom-layer regime. HRTEM is a commonly used technique for structural analysis of TMDs in various reports.10,21,24,38−50 Distinguishing 1H and 1T regions remains difficult because the phase contrast nature of the HRTEM image leads to the nonintuitive structure determination for 1H and 1T phases, and contrast inversion occurs easily due to small variations in the defocus conditions. As an important perspective in understanding phase transition in monolayer TMDs, Cs-corrected STEM enables the straightforward and clear view of the atomic structure or even atomic motion during the 1H-to-1T phase transition.12,13,15 However, the basic 3D information cannot be extracted directly from the atomically resolved STEM images. For an instance, STEM nearly ignores the difference between (0001) and (0001̅) S layers, whereas this 3D information is preserved by electron wave interference in SAED patterns. Different from STEM and HRTEM, electron diffraction has no competency in resolving individual atomic positions in real space but has the irreplaceable advantage in crystal symmetry and orientation analyses. As evidenced in this work, by resolving the intensity variation of ⟨101̅0⟩ diffraction peaks, 3D crystallographic orientation and stacking sequences of 1H and 1T phases in TMD monolayers can be experimentally determined. The discrimination of different S layers in the three-atomic-layer TMDs provides the opportunity to unveil the structure and properties of the 2D crystals and the underlying mechanisms of the semiconductorto-metal transition. We have noticed that all of the 1T phase domains in CVD-grown MoS2 monolayers are formed by the glide of the S layers that contact with the glass substrates, implying that the structure transition in the as-grown CVD TMD monolayers is possibly associated with the weak confinement from the glass substrate.35

in designing and fabricating TMD-based electronic devices by utilizing 1T/1H interfaces as the metal/semiconductor contacts. This work also underlines the robustness of commonly used electron diffraction techniques in exploring the structure of 2D materials as an important supplement for HRTEM and STEMHAADF.

METHODS Monolayer MoS2 Crystal Growth. The MoS2 samples are synthesized using a low-pressure chemical vapor deposition method.35 MoO3 (Sigma-Aldrich, purity >99.5%) and S (Wako Pure Chemical Industry, purity 99%) powders are used as precursors and microslide glass (Matsunami Glass, S1214) as growth substrates. Growth chamber is pumped down to a base pressure of 5 × 10−2 mbar before growth. Then, the growth is performed at 700 °C for 1 h with ultrapure Ar as carrier gas under pumping, followed by rapid cooling to room temperature in 5 min. Microstructure Characterization. TEM and STEM images and SAED patterns are captured using a JEOL JEM-2100F TEM equipped with double spherical aberration (Cs) correctors (CEOS) for imaging and probing lenses. Observations are performed at the acceleration voltage of 200 kV, and the spatial resolution of the STEM is ∼0.1 nm. The collecting angle for HAADF-STEM is between 100 and 267 mrad. The exposure time for SAED is 0.5 s, and all calibrations of acquisition conditions are performed on adjacent areas prior to final acquisition on the target area; detectable sample damage by the electron beam cannot be seen. Simulations of Dynamic Electron Diffraction. The multislice method is used to simulate electron diffraction with the consideration of the dynamic effect. This is a more generalized method for the simulations of electron propagation in an arbitrary model compared with the Bloch wave method.51 By separating the samples into thin slices in real space, the wave function at each layer is solved in an analytical way from top to bottom. In our simulations, the accelerating voltage is set as 200 kV, and 35 × 20 unit cells with 300 phonon configurations under ambient temperature are used. Kirkland52 and Lobato53 potentials are both adopted in the construction of the real space potential field, albeit with negligible difference.

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.6b05958. Experimental results and discussion, Figures S1−S4, and Table S1 (PDF)

AUTHOR INFORMATION CONCLUSION In summary, we developed an approach to irreplaceably characterize 1T and 1H phases and their stacking sequences of 2D TMD crystals on the basis of the dynamic effect of electron diffraction. The 3D stacking sequences of the 2D crystals revealed by the diffraction method provide a perspective for understanding the underlying mechanisms of semiconductor-tometal transition of the of TMD monolayers. The finding that the phase transition takes place by the glide of the S layers contacting with the glass substrate during CVD growth unveils the possible role of substrate confinement in the structure phase transition. The chirality of the 1T/1H interfaces, characterized by the electron diffraction method, may lead to properties in spintronics and electromagnetics of 2D TMDs for future exploration. Technically, imaging 1T and 1H phase domains and corresponding interfaces at the tens of micrometers scale by the dynamic electron diffraction method could be very important

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS This work is sponsored by JST-CREST “Phase Interface Science for Highly Efficient Energy Utilization”, JST, Japan; World Premier International (WPI) Research Center Initiative for Atoms, Molecules and Materials, MEXT, Japan, and partially supported by Japan Society for the Promotion of Science (JSPS) Grant-in-Aid for Scientific Research on Innovative Areas “Science of Atomic Layers” (Grant No. 26107504). S.N. is grateful for the financial support by the General Research Fund (Project No. 622813) from the Hong Kong Research Grants Council. 10314

DOI: 10.1021/acsnano.6b05958 ACS Nano 2016, 10, 10308−10316

Article

ACS Nano

Carrier Dynamics in Phase-Engineered Monolayer MoS2. ACS Nano 2015, 9, 840−849. (20) Najmaei, S.; Liu, Z.; Zhou, W.; Zou, X.; Shi, G.; Lei, S.; 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. (21) Brivio, J.; Alexander, D. T. L.; Kis, A. Ripples and Layers in Ultrathin MoS2 Membranes. Nano Lett. 2011, 11, 5148−5153. (22) 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. (23) Lin, Y.-C.; Björkman, T.; Komsa, H.-P.; Teng, P.-Y.; Yeh, C.-H.; Huang, F.-S.; Lin, K.-H.; Jadczak, J.; Huang, Y.-S.; Chiu, P.-W.; Krasheninnikov, A. V.; Suenaga, K. Three-Fold Rotational Defects in Two-Dimensional Transition Metal Dichalcogenides. Nat. Commun. 2015, 6, 6736. (24) Rasool, H. I.; Ophus, C.; Zettl, A. Atomic Defects in Two Dimensional Materials. Adv. Mater. 2015, 27, 5771−5777. (25) Hong, J.; Hu, Z.; Probert, M.; Li, K.; Lv, D.; Yang, X.; Gu, L.; Mao, N.; Feng, Q.; Xie, L.; Zhang, J.; Wu, D.; Zhang, Z.; Jin, C.; Ji, W.; Zhang, X.; Yuan, J.; Zhang, Z. Exploring Atomic Defects in Molybdenum Disulphide Monolayers. Nat. Commun. 2015, 6, 6293. (26) van der Zande, A. M.; Huang, P. Y.; Chenet, D. A.; Berkelbach, T. C.; You, Y.; 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. (27) Azizi, A.; Zou, X.; Ercius, P.; Zhang, Z.; Elías, A. L.; Perea-López, N.; Stone, G.; Terrones, M.; Yakobson, B. I.; Alem, N. Dislocation Motion and Grain Boundary Migration in Two-Dimensional Tungsten Disulphide. Nat. Commun. 2014, 5, 4867. (28) Ly, T. H.; Perello, D. J.; Zhao, J.; Deng, Q.; Kim, H.; Han, G. H.; Chae, S. H.; Jeong, H. Y.; Lee, Y. H. Misorientation-Angle-Dependent Electrical Transport across Molybdenum Disulfide Grain Boundaries. Nat. Commun. 2016, 7, 10426. (29) Bals, S.; Goris, B.; Altantzis, T.; Heidari, H.; Van Aert, S.; Van Tendeloo, G. Seeing and Measuring in 3D with Electrons. C. R. Phys. 2014, 15, 140−150. (30) Martinez, G. T.; Rosenauer, A.; De Backer, A.; Verbeeck, J.; Van Aert, S. Quantitative Composition Determination at the Atomic Level Using Model-Based High-Angle Annular Dark Field Scanning Transmission Electron Microscopy. Ultramicroscopy 2014, 137, 12−19. (31) Van Aert, S.; Batenburg, K. J.; Rossell, M. D.; Erni, R.; Van Tendeloo, G. Three-Dimensional Atomic Imaging of Crystalline Nanoparticles. Nature 2011, 470, 374−377. (32) De Backer, A.; Martinez, G. T.; Rosenauer, A.; Van Aert, S. Atom Counting in HAADF STEM Using a Statistical Model-Based Approach: Methodology, Possibilities, and Inherent Limitations. Ultramicroscopy 2013, 134, 23−33. (33) Shi, J.; Ma, D.; Han, G.; Zhang, Y.; Ji, Q.; Sun, J.; Song, X.; Li, C.; Zhang, Y.; Lang, X.; Zhang, Y.; Liu, Z. Controllable Growth and Transfer of Monolayer MoS2 on Au Foils and Its Potential Application In Hydrogen Evolution Reaction. ACS Nano 2014, 8, 10196−10204. (34) Wang, Z. L. Elastic and Inelastic Scattering in Electron Diffraction and Imaging; Springer: Boston, MA, 1995; pp 23−94. (35) Wang, Z.; Liu, P.; Ito, Y.; Ning, S.; Tan, Y.; Fujita, T.; Hirata, A.; Chen, M. W. Chemical Vapor Deposition of Monolayer Mo1−xWxS2 Crystals with Tunable Band Gaps. Sci. Rep. 2016, 6, 21536. (36) Miro, P.; Ghorbani-Asl, M.; Heine, T. Spontaneous Ripple Formation in MoS2 Monolayers: Electronic Structure and Transport Effects. Adv. Mater. 2013, 25, 5473−5475. (37) Mannebach, E. M.; Li, R.; Duerloo, K. A.; Nyby, C.; Zalden, P.; Vecchione, T.; Ernst, F.; Reid, A. H.; Chase, T.; Shen, X.; Weathersby, S.; Hast, C.; Hettel, R.; Coffee, R.; Hartmann, N.; Fry, A. R.; Yu, Y.; Cao, L.; Heinz, T. F.; Reed, E. J.; et al. Dynamic Structural Response and Deformations of Monolayer MoS2 Visualized by Femtosecond Electron Diffraction. Nano Lett. 2015, 15, 6889−6895. (38) Zhan, Y.; Liu, Z.; Najmaei, S.; Ajayan, P. M.; Lou, J. Large-Area Vapor-Phase Growth and Characterization of MoS2 Atomic Layers on a SiO2 Substrate. Small 2012, 8, 966−971.

REFERENCES (1) Akinwande, D.; Petrone, N.; Hone, J. Two-Dimensional Flexible Nanoelectronics. Nat. Commun. 2014, 5, 5678. (2) Bhimanapati, G. R.; Lin, Z.; Meunier, V.; Jung, Y.; Cha, J.; Das, S.; Xiao, D.; Son, Y.; Strano, M. S.; Cooper, V. R.; Liang, L.; Louie, S. G.; Ringe, E.; Zhou, W.; Kim, S. S.; Naik, R. R.; Sumpter, B. G.; Terrones, H.; Xia, F.; Wang, Y.; et al. Recent Advances in Two-Dimensional Materials beyond Graphene. ACS Nano 2015, 9, 11509−11539. (3) Deng, D.; Novoselov, K. S.; Fu, Q.; Zheng, N.; Tian, Z.; Bao, X. Catalysis with Two-Dimensional Materials and Their Heterostructures. Nat. Nanotechnol. 2016, 11, 218−230. (4) Fiori, G.; Bonaccorso, F.; Iannaccone, G.; Palacios, T.; Neumaier, D.; Seabaugh, A.; Banerjee, S. K.; Colombo, L. Electronics Based on Two-Dimensional Materials. Nat. Nanotechnol. 2014, 9, 768−779. (5) Xia, F.; Wang, H.; Xiao, D.; Dubey, M.; Ramasubramaniam, A. Two-Dimensional Material Nanophotonics. Nat. Photonics 2014, 8, 899−907. (6) Gibney, E. The Super Materials that Could Trump Graphene. Nature 2015, 522, 274−276. (7) Gong, Y.; Lin, J.; Wang, X.; Shi, G.; Lei, S.; Lin, Z.; Zou, X.; Ye, G.; Vajtai, R.; Yakobson, B. I.; Terrones, H.; Terrones, M.; Tay, B. K.; Lou, J.; Pantelides, S. T.; Liu, Z.; Zhou, W.; Ajayan, P. M. Vertical and inPlane Heterostructures from WS2/MoS2 Monolayers. Nat. Mater. 2014, 13, 1135−1142. (8) Ito, Y.; Tanabe, Y.; Qiu, H. J.; Sugawara, K.; Heguri, S.; Tu, N. H.; Huynh, K. K.; Fujita, T.; Takahashi, T.; Tanigaki, K.; Chen, M. W. HighQuality Three-Dimensional Nanoporous Graphene. Angew. Chem., Int. Ed. 2014, 53, 4822−4826. (9) Li, M.-Y.; Shi, Y.; Cheng, C.-C.; Lu, L.-S.; Lin, Y.-C.; Tang, H.-L.; Tsai, M.-L.; Chu, C.-W.; Wei, K.-H.; He, J.-H.; Chang, W.-H.; Suenaga, K.; Li, L.-J. Epitaxial Growth of a Monolayer WSe2-MoS2 Lateral P-N Junction with an Atomically Sharp Interface. Science (Washington, DC, U. S.) 2015, 349, 524−528. (10) Ling, X.; Lin, Y.; Ma, Q.; Wang, Z.; Song, Y.; Yu, L.; Huang, S.; Fang, W.; Zhang, X.; Hsu, A. L.; Bie, Y.; Lee, Y.-H.; Zhu, Y.; Wu, L.; Li, J.; Jarillo-Herrero, P.; Dresselhaus, M.; Palacios, T.; Kong, J. Parallel Stitching of 2D Materials. Adv. Mater. 2016, 28, 2322−2329. (11) Tan, Y.; Liu, P.; Chen, L.; Cong, W.; Ito, Y.; Han, J.; Guo, X.; Tang, Z.; Fujita, T.; Hirata, A.; Chen, M. W. Monolayer MoS2 Films Supported by 3D Nanoporous Metals for High-Efficiency Electrocatalytic Hydrogen Production. Adv. Mater. 2014, 26, 8023−8028. (12) Eda, G.; Fujita, T.; Yamaguchi, H.; Voiry, D.; Chen, M. W.; Chhowalla, M. Coherent Atomic and Electronic Heterostructures of Single-Layer MoS2. ACS Nano 2012, 6, 7311−7317. (13) Kappera, R.; Voiry, D.; Yalcin, S. E.; Branch, B.; Gupta, G.; Mohite, A. D.; Chhowalla, M. Phase-Engineered Low-Resistance Contacts for Ultrathin MoS2 Transistors. Nat. Mater. 2014, 13, 1128− 1134. (14) Eda, G.; Yamaguchi, H.; Voiry, D.; Fujita, T.; Chen, M. W.; Chhowalla, M. Photoluminescence from Chemically Exfoliated MoS2. Nano Lett. 2011, 11, 5111−5116. (15) Lin, Y.-C.; Dumcenco, D. O.; Huang, Y.; Suenaga, K. Atomic Mechanism of the Semiconducting-to-Metallic Phase Transition in Single-Layered MoS2. Nat. Nanotechnol. 2014, 9, 391−396. (16) Voiry, D.; Mohite, A.; Chhowalla, M. Phase Engineering of Transition Metal Dichalcogenides. Chem. Soc. Rev. 2015, 44, 2702− 2712. (17) Katagiri, Y.; Nakamura, T.; Ishii, A.; Ohata, C.; Hasegawa, M.; Katsumoto, S.; Cusati, T.; Fortunelli, A.; Iannaccone, G.; Fiori, G.; Roche, S.; Haruyama, J. Gate-Tunable Atomically Thin Lateral MoS2 Schottky Junction Patterned by Electron Beam. Nano Lett. 2016, 16, 3788−3794. (18) Cho, S.; Kim, S.; Kim, J. H.; Zhao, J.; Seok, J.; Keum, D. H.; Baik, J.; Choe, D.-H.; Chang, K. J.; Suenaga, K.; Kim, S. W.; Lee, Y. H.; Yang, H. Phase Patterning for Ohmic Homojunction Contact in MoTe2. Science (Washington, DC, U. S.) 2015, 349, 625−628. (19) Yamaguchi, H.; Blancon, J.-C.; Kappera, R.; Lei, S.; Najmaei, S.; Mangum, B. D.; Gupta, G.; Ajayan, P. M.; Lou, J.; Chhowalla, M.; Crochet, J. J.; Mohite, A. D. Spatially Resolved Photoexcited Charge10315

DOI: 10.1021/acsnano.6b05958 ACS Nano 2016, 10, 10308−10316

Article

ACS Nano (39) Ji, Q.; Zhang, Y.; Gao, T.; Zhang, Y.; Ma, D.; Liu, M.; Chen, Y.; Qiao, X.; Tan, P.-H.; Kan, M.; Feng, J.; Sun, Q.; Liu, Z. Epitaxial Monolayer MoS2 on Mica with Novel Photoluminescence. Nano Lett. 2013, 13, 3870−3877. (40) Tongay, S.; Fan, W.; Kang, J.; Park, J.; Koldemir, U.; Suh, J.; Narang, D. S.; Liu, K.; Ji, J.; Li, J.; Sinclair, R.; Wu, J. Tuning Interlayer Coupling in Large-Area Heterostructures with CVD-Grown MoS2 and WS2 Monolayers. Nano Lett. 2014, 14, 3185−3190. (41) Gutiérrez, H. R.; Perea-López, N.; Elías, A. L.; Berkdemir, A.; Wang, B.; Lv, R.; López-Urías, F.; Crespi, V. H.; Terrones, H.; Terrones, M. Extraordinary Room-Temperature Photoluminescence in Triangular WS2 Monolayers. Nano Lett. 2013, 13, 3447−3454. (42) Zhang, Y.; Zhang, Y.; Ji, Q.; Ju, J.; Yuan, H.; Shi, J.; Gao, T.; Ma, D.; Liu, M.; Chen, Y.; Song, X.; Hwang, H. Y.; Cui, Y.; Liu, Z. Controlled Growth of High-Quality Monolayer WS2 Layers on Sapphire. ACS Nano 2013, 7, 8963−8971. (43) Song, J.-G.; Park, J.; Lee, W.; Choi, T.; Jung, H.; Lee, C. W.; Hwang, S.-H.; Myoung, J. M.; Jung, J.-H.; Kim, S.-H.; Lansalot-Matras, C.; Kim, H. Layer-Controlled, Wafer-Scale, and Conformal Synthesis of Tungsten Disulfide Nanosheets Using Atomic Layer Deposition. ACS Nano 2013, 7, 11333−11340. (44) Zhang, L.; Liu, K.; Wong, A. B.; Kim, J.; Hong, X.; Liu, C.; Cao, T.; Louie, S. G.; Wang, F.; Yang, P. Three-Dimensional Spirals of Atomic Layered MoS2. Nano Lett. 2014, 14, 6418−6423. (45) Cong, C.; Shang, J.; Wu, X.; Cao, B.; Peimyoo, N.; Qiu, C.; Sun, L.; Yu, T. Synthesis and Optical Properties of Large-Area SingleCrystalline 2D Semiconductor WS2 Monolayer from Chemical Vapor Deposition. Adv. Opt. Mater. 2014, 2, 131−136. (46) Huang, J.-K.; Pu, J.; Hsu, C.-L.; Chiu, M.-H.; Juang, Z.-Y.; Chang, Y.-H.; Chang, W.-H.; Iwasa, Y.; Takenobu, T.; Li, L.-J. Large-Area Synthesis of Highly Crystalline WSe2 Monolayers and Device Applications. ACS Nano 2014, 8, 923−930. (47) Zhang, J.; Yu, H.; Chen, W.; Tian, X.; Liu, D.; Cheng, M.; Xie, G.; Yang, W.; Yang, R.; Bai, X.; Shi, D.; Zhang, G. Scalable Growth of HighQuality Polycrystalline MoS2 Monolayers on SiO2 with Tunable Grain Sizes. ACS Nano 2014, 8, 6024−6030. (48) Xu, Z.-Q.; Zhang, Y.; Lin, S.; Zheng, C.; Zhong, Y. L.; Xia, X.; Li, Z.; Sophia, P. J.; Fuhrer, M. S.; Cheng, Y.-B.; Bao, Q. Synthesis and Transfer of Large-Area Monolayer WS2 Crystals: Moving Toward the Recyclable Use of Sapphire. ACS Nano 2015, 9, 6178−6187. (49) Lehtinen, O.; Komsa, H.; Pulkin, A.; Whitwick, M. B.; Chen, M.; Lehnert, T.; Mohn, M. J.; Yazyev, O. V.; Kis, A.; Kaiser, U.; Krasheninnikov, A. V. Atomic Scale Microstructure and Properties of Se-Deficient. ACS Nano 2015, 9, 3274−3283. (50) Fu, L.; Sun, Y.; Wu, N.; Mendes, R. G.; Chen, L.; Xu, Z.; Zhang, T.; Rümmeli, M. H.; Rellinghaus, B.; Pohl, D.; Zhuang, L.; Fu, L. Direct Growth of MoS2/h-BN Heterostructures via a Sulfide-Resistant Alloy. ACS Nano 2016, 10, 2063−2070. (51) Stadelmann, P. A. EMS - a Software Package for Electron Diffraction Analysis and HREM Image Simulation in Materials Science. Ultramicroscopy 1987, 21, 131−145. (52) Kirkland, E. J. Advanced Computing in Electron Microscopy; Springer: Boston, MA, 2010; pp 77−112. (53) Lobato, I.; Van Dyck, D. MULTEM: A New Multislice Program to Perform Accurate and Fast Electron Diffraction and Imaging Simulations Using Graphics Processing Units with CUDA. Ultramicroscopy 2015, 156, 9−17.

10316

DOI: 10.1021/acsnano.6b05958 ACS Nano 2016, 10, 10308−10316