Contrasting Structural Reconstructions, Electronic Properties, and

Subscriber access provided by University of Newcastle, Australia

Communication

Contrasting Structural Reconstructions, Electronic Properties, and Magnetic Orderings along Different Edges of Zigzag Transition Metal Dichalcogenide Nanoribbons Ping Cui, Jin-Ho Choi, Wei Chen, Jiang Zeng, Chih-Kang Shih, Zhenyu Li, and Zhenyu Zhang Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.6b04638 • Publication Date (Web): 28 Dec 2016 Downloaded from http://pubs.acs.org on December 28, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Nano Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Contrasting Structural Reconstructions, Electronic Properties, and Magnetic Orderings along Different Edges of Zigzag Transition Metal Dichalcogenide Nanoribbons Ping Cui,† Jin-Ho Choi,†, ‡ Wei Chen,†,§ Jiang Zeng,† Chih-Kang Shih,‖ Zhenyu Li,*,† and Zhenyu Zhang*,† †

International Center for Quantum Design of Functional Materials (ICQD), Hefei National

Laboratory for Physical Sciences at the Microscale, and Synergetic Innovation Center of Quantum Information and Quantum Physics, University of Science and Technology of China, Hefei, Anhui 230026, China ‡

Department of Physics, Hanyang University, 17 Haengdang-Dong, Seongdong-Ku, Seoul 133-

791, Korea §

Department of Physics and School of Engineering and Applied Sciences, Harvard University,

Cambridge, Massachusetts 02138, USA ‖

Department of Physics, The University of Texas at Austin, Austin, Texas 78712, USA

ACS Paragon Plus Environment

1

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 21

ABSTRACT

Two-dimensional transition metal dichalcogenides represent an emerging class of layered materials exhibiting various intriguing properties, and integration of such materials for potential device applications will necessarily invoke further reduction of their dimensionality. Using firstprinciples approaches, here we investigate the structural, electronic, and magnetic properties along the two different edges of zigzag MX2 (M=Mo, W; X=S, Se) nanoribbons. Along the M edges, we reveal a previously unrecognized but energetically strongly preferred (2x1) reconstruction pattern, which is universally operative for all the four systems (and possibly more), characterized by an elegant self-passivation mechanism through place exchanges of the outmost X and M edge atoms. In contrast, the X edges undergo a much milder (2x1) or (3x1) reconstruction for MoX2 or WX2, respectively. These contrasting structural preferences of the edges can be exploited for controlled fabrication of properly tailored TMD nanoribbons under nonequilibrium growth conditions. We further use the zigzag MoX2 nanoribbons to demonstrate that the Mo and X edges possess distinctly different electronic and magnetic properties, which are significant for catalytic and spintronic applications.

KEYWORDS: first-principles calculations, transition metal dichalcogenides, nanoribbons, edge reconstruction, electronic and magnetic properties


2

Page 3 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Two-dimensional (2D) transition metal dichalcogenides (TMDs) represent an important new class of layered materials with immense application potentials in electronic, optical, spintronic, and catalytic devices.1-7 Integration of such 2D materials into actual devices will necessarily invoke further reduction of their dimensionality or creation of properly tailored boundaries, which often exhibit structural and physical properties distinctly different from their 2D bulk counterparts. For example, depending on their specific terminations, the edges of the widely studied MoS2 nanoribbons have been shown to exhibit metallic or semiconducting, ferromagnetic or nonmagnetic behaviors.8-12 Such intriguing properties along the structurally different edges, in turn, can be exploited for desirable technological developments. Among different possibilities of TMD edges, the most commonly encountered types are the zigzag and armchair edges, which also have their close counterparts in graphene.13-16 The main difference between them is that graphene has only one type of zigzag edge, while the TMD materials, more like hexagonal boron nitride,17 possess two types of zigzag edges: metalterminated (10ī0) edge (thereafter labeled as M edge) and chalcogen-terminated (ī010) edge (X edge). Many novel properties associated with the edges are the consequences of the specific type of termination and the exact bonding configurations due to structural reconstruction along the edges, which can be drawn close analogies to surface reconstructions of 3D compound semiconductors such GaAs.18-20 For 2D-bulk-terminated edges, each termination would be accompanied by chemically unpaired electrons along the edge known as dangling bonds, leaving the system to be energetically unfavorable. Edge reconstruction, or place readjustment of the edge atoms, is a common remedy to remove such dangling bonds, as recently established for graphene.21-23 It is naturally desirable to uncover the likely reconstruction patterns along the


3

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 21

more complex TMD zigzag edges, which are not only frequently observed experimentally,24-26 but are also expected to have more exotic chemical and physical properties. In this Letter, we use first-principles calculations within density functional theory (DFT) to explore systematically the likely reconstruction mechanisms along the two different edges of zigzag MX2 (M=Mo, W; X=S, Se) nanoribbons. Unlike extensive earlier studies devoted to situations with variable stoichiometric ratio,11,12,24-26 here we focus on the special but vitally important case with the stoichiometric ratio of M:X=1:2 along the edges. We find that the differently terminated edges possess striking and distinctly different reconstruction patterns. Along the M edges, an energetically highly favorable place exchange or self-passivation mechanism is universally operative for all the four systems considered (and likely for other TMD systems as well), exhibiting substantial inward displacements of all the first-row M atoms and corresponding outward displacements of half of the second-row X atoms. In the ending structure, the more reactive M atoms are all effectively passivated by the X atoms. In contrast, along the X edges, only relatively milder (2x1) or (3x1) reconstructions are found for MoX2 or WX2, respectively, characterized by slight local place readjustments of the edge atoms. These contrasting structural preferences of the edges can be exploited for controlled fabrication of properly tailored TMD nanoribbons under nonequilibrium growth conditions, as achieved recently experimentally.27 We further use the zigzag MoX2 nanoribbons to demonstrate that the differently reconstructed Mo and X edges possess distinctly different electronic properties and magnetic orderings. We also present simulated scanning tunneling microscopy (STM) images along both the reconstructed Mo and X edges, and contrast the different metallic states along the two edges. The concerted non-trivial self-passivation mechanism and the resulting novel physical


4

Page 5 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

properties uncovered here are expected to stimulate substantial future research activities in this important area. The DFT calculations were performed using the projector-augmented wave method28 implemented in the Vienna ab initio simulation package (VASP),29,30 with the Perdew-BurkeErnzerhof exchange-correlation functional.31 A plane-wave basis set was adopted with an energy cutoff of 350 eV. In each supercell, the vacuum layers between two neighboring ribbons are thicker than 16 Å. The 1D Brillouin zone was sampled using 16×1×1, 12×1×1, 10×1×1, 8×1×1, and 6×1×1 Monkhorst-Pack k-point meshes for the 2×1, 3×1, 4×1, 5×1, and 6×1 supercells, respectively. Unless otherwise specified, all the atoms were fully relaxed by the conjugate gradient algorithm until the residual forces were less than 0.01 eV/Å. To minimize the interaction between the two edges, the width of the nanoribbons is chosen to be sufficiently large (~ 20 Å, containing 8 zigzag chains across the ribbon, and labeled as 8-ZZ). When comparing the stabilities of different configurations along a given edge, the atoms in the outmost two rows of the opposite edge are fixed in their respective bulk-terminated positions as we optimize the positions of all the other atoms in the nanoribbon. Spin polarization was switched on in the DFT calculations. STM images were simulated in both constant-current and constant-height modes at different bias voltages by using the Tersoff-Hamann approximation,32 and the results are consistent with each other in all the essential aspects considered. Along either the M or X edge of a given MX2, we first compare the energies of the unreconstructed and differently reconstructed patterns (Figures 1 and S1 in the Supporting Information), then identify the most favorable structure. When specified to the M edge, the (2x1) reconstruction illustrated in Figure 1 is found to be energetically more favorable for all the four MX2 systems. The energy gain for MoS2 is 0.261 eV per formula unit along the edge direction


5

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 21

relative to the unreconstructed case, and is much larger for the other three MX2 systems (Table 1). The underlying atomistic mechanism involved is via place exchange, with substantial inward displacements of all the first-row M atoms (Mo 1 and Mo 2 in the insets of Figure 1a) and corresponding outward displacements of half of the second-row X atoms (X 3 and X 4). In the ending structure, the displaced M atoms are dimerized, and each dimer also connects with another M atom belonging to the third row to form a planar M3 trimer, with all the edge M atoms now effectively passivated by the displaced X atoms. The corresponding edge-on views are shown in Figure S2. When the M edges are constrained to (3x1) reconstruction patterns, the optimized structures (see Figure S3) are metastable, as shown in Table 1. For the X edges, the most stable structures of MoX2 are (2x1)-reconstructed, with periodically alternating Mo-X bond lengths (Figure 1a). The short and long Mo-S (Mo-Se) bond lengths are 2.36 (2.50) and 2.46 (2.57) Å, respectively, with the corresponding edge-on views shown in Figure S2a. The second-row Mo atoms are also dimerized, resulting in shortened Mo-Mo distances of 3.02 (3.12) Å from the bulk-terminated values of 3.18 (3.32) Å for MoS2 (MoSe2). The corresponding energy gain is 0.320 (0.158) eV per formula unit along the edge for MoS2 (MoSe2). The underlying driving force for the distinctly different (2x1) reconstructions along both the Mo and X edges can be uniquely attributed to the same dimerization trend of the quasi1D Mo atoms along the Mo and X edges. For the X edges of WX2, the (3x1) reconstructions are found to be the most stable, with 2/3 of the W-X bonds shortened, and the other 1/3 elongated (Figure 1b). The second-row W atoms are linearly trimerized, which may provide the underlying driving force for the overall (3x1) reconstructions. The distance between two neighboring W atoms in a linear trimer is 3.06 (3.19) Å and the gap between two neighboring W3 trimers is 3.43 (3.58) Å for WS2 (WSe2).


6

Page 7 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

For the ribbon structures of such polar materials, electric fields across the ribbon widths may also play an important role in determining their overall stabilities.12 The strengths of such polar fields can be weakened by saturating the M edges with additional X atoms, or by the reconstructions identified here. Indeed, our detailed calculations show that the dipole moment of an 8-ZZ MoS2 nanoribbon is reduced from -1.02 e⋅Å per twice the (1x1) unit cell for the unreconstructed structure to 0.30 e⋅Å per (2x1) unit cell for the reconstructed one. The contrasting reconstruction geometries at the two zigzag edges of MX2 are expected to also result in different electronic and magnetic properties, as exemplified below using the MoX2 nanoribbons. We start from the spin-unpolarized band structure of MoS2, which has two edge states crossing the Fermi level, with band I originated from the Mo edge, and band II from the S edge (Figure 2a). When spin polarization is considered, both bands have sizable spin splits (0.3 ~ 0.4 eV), much larger than the other bands. As shown in Figure 2b, the spin-up component of band I (labeled as I-u) becomes fully occupied, while the spin-down component (I-d) is still partially occupied. For band II, the spin-up channel (II-u) is partially occupied, while the spindown channel (II-d) is totally unoccupied. As a consequence, even though both edges are still metallic, the contributing densities of states around the Fermi level have been significantly reduced, resulting in an overall more stable configuration. The energy gain is 0.071 eV per (2x1) unit cell compared to the spin-unpolarized state. The spatial charge density distributions of the bands crossing the Fermi level at the Γ point are displayed in Figure 2c. Around the Fermi level, the spin-polarized I-d and II-u bands are both quite flat, which may provide an important materials-specific platform for exploiting novel physical phenomena involving collective and/or correlated electrons.33,34


7

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 21

Specifically, the partially filled flat bands around the Fermi level suggest that ferromagnetism is likely to be stabilized along either reconstructed edge of the MoS2 nanoribbon.35 Quantitative confirmations of this expectation are presented in Figure S4, displaying the dominant contributions to the non-zero total magnetic moment by the ferromagnetically coupled constituent atoms along either edge. Furthermore, we find that the total magnetic moment per (2x1) unit cell is 1.38 µB when the two edges are in ferromagnetic (FM) order and is zero in antiferromagnetic (AFM) order. Both states are energetically degenerate, because at the 8-ZZ width the edge-edge coupling is negligible. Therefore, the AFM band structure shown in Figure S5a is essentially the same as the FM band structure shown in Figure 2b, but here the two spin channels of band II are mutually exchanged, while the two spin channels of band I stay intact. On a detailed atomistic level, the magnetic moments are located mainly on the outmost S atoms and the inner Mo atoms of the reconstructed Mo3 trimers along the Mo edge (Figure S4). In contrast, along the S edge, the magnetic moments are mainly concentrated on the edge Mo atoms, and the outmost S atoms contribute much weaker and AFM-coupled moments to the total moment of the dominant Mo atoms. The electronic structures of a MoSe2 nanoribbon are very similar to that of MoS2, as shown in Figure 3. But unlike MoS2, for MoSe2 another edge state (band III) located far away from the Fermi level in the spin-unpolarized case moves closer to the Fermi level in the spin-polarized case, resulting in another metallic channel from band III-u (Figure 3b). This band is derived from the py/pz orbitals of the Se atoms contained in the shorter Mo-Se bonds along the Se edge. The spatial charge density distributions of the bands crossing the Fermi level at the Γ point are contrasted in Figure 3c. The FM state has a total magnetic moment of 1.32 µB per (2x1) unit cell,


8

Page 9 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

which is energetically degenerate with the state where the two edges are in AFM order (Figure S5b). Both states are 0.069 eV per (2x1) unit cell lower than the spin-unpolarized state. Recent studies have shown the importance of spin-orbit coupling (SOC) in such TMD systems.2,5 As a cross check, we have tested the SOC effects on the edge states of the MoX2 ribbons. The results are presented in Figure S6, showing that the spin-resolved electronic structures along the edges stay qualitatively intact in the presence of the SOC. The underlying reason lies in the substantially larger exchange splittings due to the coupling of different spin states along each edge. The distinctly different reconstruction patterns and electronic structures should provide an important basis for differentiating such edges once formed experimentally. For this purpose, we have simulated the STM images of the 8-ZZ MoX2 nanoribbons at different bias voltages (Vb), as shown in Figure 4. For MoS2, at small biases (Vb = ± 0.005 V), the STM images highlights the contributions from the states very close to the Fermi level. We identify a quasi-1D metallic state along either the Mo or S edge (Figure 4a,b). These two metallic edge states correspond to the two spin-polarized bands (I-d and II-u in Figure 2b) crossing the Fermi level. At higher biases (Vb = ± 0.1 V), the whole bands crossing the Fermi level (I-d and II-u in Figure 2b) are captured, and the STM images still have different brightness along the two edges (Figure 4c,d). It is interesting to note that, at all the bias voltages considered, the brightest location along the S edge is not on the outmost atoms, but mainly on the second-row S atoms with dimerized images. Furthermore, because these bright images are connected through the bright first-row S atoms, metallic states can still be sustained along the S edge, showing the quasi-1D nature of the conducting channel.


9

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 21

To a large extent, the electronic structures along the Mo and Se edges of MoSe2 are similar to that of MoS2. On the other hand, because of the existence of the third band right below the Fermi level (III-u in Figure 3b), the simulated STM images of the MoSe2 nanoribbons show different bias dependence from that of the MoS2 nanoribbons. In particular, as shown in Figure 4e-h, the Se edge displays bright images along the second-row Se atoms at the positive biases, but the first-row Se atoms become brighter at the negative biases. Furthermore, either along the S or Se edge of MoX2, the STM images show clear differences from that of the corresponding Mo edge. When the ribbons are narrower, the two edges will interact with each other. To quantitatively examine the inter-edge interactions, the structural and magnetic properties of n-ZZ MoX2 nanoribbons with n = 3,4,5,6,7 are systematically studied. It turns out that the contrasting edge reconstructions are largely preserved even for ribbons as narrow as 3-ZZ. Furthermore, the FM and AFM states stay degenerate within an accuracy of 0.001 eV per (2x1) unit cell even for ribbons as narrow as 4-ZZ, which is consistent with the generally much weaker nature of magnetic interaction when compared with structural or electronic interactions. The universal (2x1) reconstruction pattern at the zigzag M edge, characterized by the novel place exchanges of the outmost M and X atoms, has no precedence, which calls for experimental validations. The MX2 edges that preserve the M:X = 1:2 ratio can be obtained physically realistically using different means. For example, a recent experiment on epitaxial growth of MoSe2 nanoribbons27 provides convincing evidence that it is possible to tune the edge decorations by controlling the Se:Mo flux ratio, e.g., the Mo atoms at the edges are not passivated by extra Se atoms at low Se:Mo flux ratios, thus maximizing the chance for the Mo edge to execute the (2x1) reconstruction via place exchanges. The selection of the (2x1) reconstructed Mo edge, in turn, favors the nanoribbon structure rather than the commonly


10

Page 11 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

observed triangular or hexagonal clusters.27,36 These findings further justify the physical basis of the M:X = 1:2 regime emphasized in the present study, in contrast to earlier emphases on variable stoichiometric ratio.11,12,24-26 In conclusion, we have systematically investigated the edge properties of zigzag MX2 (M=Mo, W; X=S, Se) nanoribbons. Although distinctly different at the two opposite zigzag edges, reconstructions at both edges are energetically favorable. Along the M edges, the reconstruction is universally characterized by an intriguing self-passivation mechanism, which is likely to be valid for other TMD systems as well. Reconstructions at the X edges are milder, characterized by varying M-X bond lengths. These reconstructed edges are also characterized by robust edge states and exotic magnetic orderings. Collectively, the novel structural and physical properties uncovered along the different edges of the MX2 materials are expected to have important nanoelectronic, spintronic, optical, and catalytic applications.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: XXXX. Other reconstruction patterns along the two zigzag edges of MX2 nanoribbons and edge-on views of the most stable structures, atom-resolved magnetic moment distributions, and spin-polarized band structures with antiferromagnetic orderings between two opposite edges or with ferromagnetic orderings and spin-orbit coupling (PDF) AUTHOR INFORMATION


11

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 21

Corresponding Author *[email protected] *[email protected] Notes The authors declare no competing financial interests. ACKNOWLEDGMENTS This work was supported in part by the NSFC (Grant Nos. 11634011, 61434002, 11374273, 21222304, and 21421063), NKBRPC (Grant No. 2014CB921103), and by the US NSF (Grant No. EFMA-1542747). The calculations were carried out at the National Supercomputer Center in Tianjin.

REFERENCES (1) Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A. Nat. Nanotechnol. 2011, 6, 147-150. (2) Xiao, D.; Liu, G. B.; Feng, W. X.; Xu, X. D.; Yao, W. Phys. Rev. Lett. 2012, 108, 196802. (3) Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.; Strano, M. S. Nat. Nanotechnol. 2012, 7, 699-712. (4) Gong, Z. R.; Liu, G. B.; Yu, H. Y.; Xiao, D.; Cui, X. D.; Xu, X. D.; Yao, W. Nat. Commun. 2013, 4, 2053. (5) Xu, X. D.; Yao, W.; Xiao, D.; Heinz, T. F. Nat. Phys. 2014, 10, 343-350.


12

Page 13 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

(6) Jariwala, D.; Sangwan, V. K.; Lauhon, L. J.; Marks, T. J.; Hersam, M. C. ACS Nano 2014, 8, 1102-1120. (7) Kappera, R.; Voiry, D.; Yalcin, S. E.; Branch, B.; Gupta, G.; Mohite, A. D.; Chhowalla, M. Nat. Mater. 2014, 13, 1128-1134. (8) Li, Y. F.; Zhou, Z.; Zhang, S. B.; Chen, Z. F. J. Am. Chem. Soc. 2008, 130, 16739-16744. (9) Botello-Méndez, A. R.; López-Urías, F.; Terrones, M.; Terrones, H. Nanotechnology 2009, 20, 325703. (10) Ataca, C.; Şahin, H.; Aktürk, E.; Ciraci, S. J. Phys. Chem. C 2011, 115, 3934-3941. (11) Lucking, M. C.; Bang, J.; Terrones, H.; Sun, Y.-Y.; Zhang, S. B. Chem. Mater. 2015, 27, 3326-3331. (12) Gibertini, M.; Marzari, N. Nano Lett. 2015, 15, 6229-6238. (13) Jia, X. T.; Hofmann, M.; Meunier, V.; Sumpter, B. G.; Campos-Delgado, J.; Romo-Herrera, J. M.; Son, H.; Hsieh, Y.-P.; Reina, A.; Kong, J.; Terrones, M.; Dresselhaus, M. S. Science 2009, 323, 1701-1705. (14) Girit, Ç. Ö.; Meyer, J. C.; Erni, R.; Rossell, M. D.; Kisielowski, C.; Yang, L.; Park, C.-H.; Crommie, M. F.; Cohen, M. L.; Louie, S. G.; Zettl, A. Science 2009, 323, 1705-1708. (15) Son, Y.-W.; Cohen, M. L.; Louie, S. G. Nature 2006, 444, 347-349. (16) Magda, G. Z.; Jin, X. Z.; Hagymási, I.; Vancsó, P.; Osváth, Z.; Nemes-Incze, P.; Hwang, C. Y.; Biró, L. P.; Tapasztó, L. Nature 2014, 514, 608-611.


13

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 21

(17) Liu, Y. Y.; Bhowmick, S.; Yakobson, B. I. Nano Lett. 2011, 11, 3113-3116. (18) Northrup, J. E.; Froyen, S. Phys. Rev. B 1994, 50, 2015-2018. (19) Hashizume, T.; Xue, Q. K.; Zhou, J.; Ichimiya, A.; Sakurai, T. Phys. Rev. Lett. 1994, 73, 2208-2211. (20) Zhang, S. B.; Zunger, A. Phys. Rev. B 1996, 53, 1343-1356. (21) Koskinen, P.; Malola, S.; Häkkinen, H. Phys. Rev. Lett. 2008, 101, 115502. (22) Koskinen, P.; Malola, S.; Häkkinen, H. Phys. Rev. B 2009, 80, 073401. (23) Kim, K.; Coh, S.; Kisielowski, C.; Crommie, M. F.; Louie, S. G.; Cohen, M. L.; Zettl, A. Nat. Commun. 2013, 4, 2723. (24) Hansen, L. P.; Ramasse, Q. M.; Kisielowski, C.; Brorson, M.; Johnson, E.; Topsøe, H.; Helveg, S. Angew. Chem. Int. Ed. 2001, 50, 10153-10156. (25) Lauritsen, J. V.; Kibsgaard, J.; Helveg, S.; Topsøe, H.; Clausen, B. S.; Lægsgaard, E.; Besenbacher, F. Nat. Nanotechnol. 2007, 2, 53-58. (26) Zhou, W.; Zou, X. L.; Najmaei, S.; Liu, Z.; Shi, Y. M.; Kong, J.; Lou, J.; Ajayan, P. M.; Yakobson, B. I.; Idrobo, J.-C. Nano Lett. 2013, 13, 2615-2622. (27) Y. Chen, P. Cui, X. Ren, C. Zhang, C. Jin, Z. Zhang, C. K. Shih, “Controlled fabrication of MoSe2 nanoribbons via an unusual morphological phase transition”, Nat. Comm. (under revision). (28) Blöchl, P. E. Phys. Rev. B 1994, 50, 17953-17979.


14

Page 15 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

(29) Kresse, G.; Furthmüller, J. Phys. Rev. B 1996, 54, 11169-11186. (30) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758-1775. (31) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865-3868. (32) Tersoff, J.; Hamann, D. R. Phys. Rev. Lett. 1983, 50, 1998-2001. (33) Wu, C. J.; Bergman, D.; Balents, L.; Das Sarma, S. Phys. Rev. Lett. 2007, 99, 070401. (34) Liu, Z.; Liu, F.; Wu, Y.-S. Chin. Phys. B 2014, 23, 077308. (35) Marcus, P. M.; Moruzzi, V. L. Phys. Rev. B 1988, 38, 6949-6953. (36) Jaramillo, T. F.; Jørgensen, K. P.; Bonde, J.; Nielsen, J. H.; Horch, S.; Chorkendorff, I. Science 2007, 317, 100-102.


15

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 21

FIGURES AND FIGURE CAPTIONS

Figure 1. Side (left) and top (right) views of optimized atomic structures of 8-ZZ (a) MoX2 and (b) WX2 (X=S, Se) nanoribbons. The reconstructed edge structures are indicated by the dotted rectangles. The insets in the middle display the place exchange mechanism of the (2x1) reconstruction of the Mo edge.


16

Page 17 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 2. (a) Spin-unpolarized and (b) spin-polarized band structures of a (2x1) reconstructed 8ZZ MoS2 nanoribbon. (c) Spatial charge density distributions of the spin-polarized bands crossing the Fermi energy at the Γ point. The red and blue colors denote spin-up and spin-down states, respectively.


17

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 21

Figure 3. (a) Spin-unpolarized and (b) spin-polarized band structures of a (2x1) reconstructed 8ZZ MoSe2 nanoribbon. (c) Spatial charge density distributions of the spin-polarized bands crossing the Fermi energy at the Γ point.


18

Page 19 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 4. Simulated constant-current STM images of 8-ZZ (a)-(d) MoS2 and (e)-(h) MoSe2 nanoribbons at different bias voltages. A ball-and-stick model of ribbon structure is superposed onto the simulated STM image. The MoS2 system displays contrasting metallic states along the Mo and S edges.


19

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 21

TABLE Table 1. Energy change (∆ ∆E) due to edge reconstructiona

MoS2

a

MoSe2

WS2

WSe2

2x1

3x1

2x1

3x1

2x1

3x1

2x1

3x1

M edge

0.261

-0.014

0.823

0.515

0.807

0.531

1.012

0.690

X edge

0.320

0.259

0.158

0.115

0.120

0.172

0.009

0.046

The energy change is defined by ∆E = Eunr - Erec, where Eunr and Erec are respectively the total

energies of a given system before and after a specific edge reconstruction, and measured in eV per formula unit (MX2) along the edge direction. The values in bold highlight the preferred reconstruction patterns.


20

Page 21 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

TOC GRAPHIC


21

Contrasting Structural Reconstructions, Electronic Properties, and

Recommend Documents