Electrical Doping Effect of Vacancies on Monolayer MoS2 - The

Jan 11, 2019 - Electrical Doping Effect of Vacancies on Monolayer MoS2. Jing Yang*† , Hiroyo Kawai† , Calvin Pei Yu Wong† , and Kuan Eng Johnson...
2 downloads 0 Views 975KB Size
Subscriber access provided by Iowa State University | Library

C: Surfaces, Interfaces, Porous Materials, and Catalysis 2

Electrical Doping Effect of Vacancies in Monolayer MoS Jing Yang, Hiroyo Kawai, Calvin Pei Yu Wong, and Kuan Eng Johnson Goh

J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b10496 • Publication Date (Web): 11 Jan 2019 Downloaded from http://pubs.acs.org on January 17, 2019

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 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 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.

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 24 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

The Journal of Physical Chemistry

Electrical Doping Effect of Vacancies in Monolayer MoS2 Jing Yang,†* Hiroyo Kawai,† Calvin Pei Yu Wong† and Kuan Eng Johnson Goh†,‡* †Institute

of Materials Research and Engineering, A*STAR (Agency for Science, Technology and Research), 2 Fusionopolis Way, Innovis, #08-03, 138634 Singapore Department of Physics, Faculty of Science, National University of Singapore, 2 Science Drive 3, 117551 Singapore ‡

*Corresponding Authors’ Emails: [email protected]; [email protected]; [email protected]

ABSTRACT: Doping of transition metal dichalcogenides (TMDCs) is an effective way to tune the Fermi level to facilitate the band engineering required for different types of devices. For TMDCs, controversy abounds with regards to the doping role played by vacancytype defects. Here, we report a detailed study based on first principle calculations proposing that the native sulfur vacancies (VS) can significantly alter the electrical doping level in MoS2, and tune the material to exhibit conventional n- or p-type semiconductor characteristics. In particular, we reveal that lower concentration of the single VS (2.8% and 6.3%) yields p-type characteristics, whereas higher concentration of the single VS or a cluster of VS (12.5%, 18.8% and 25.0%) yield n-type characteristics. The trend is consistent with previous XPS and STM results. Employing this method of tuning the electron doping level, we modeled the commonly used metal-semiconductor interface in order to demonstrate both n- and p-type Schottky contact behaviours. 1

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Interestingly, we found that the defect configuration could also tune the doping and hence the contact. Simulation of the electric current at the interface as a function of bias voltage provides a reference for how the electrical characteristics would shift based on the change in vacancy concentration. Our study reveals that the VS of monolayer MoS2 at the MoS2-metal interface play an important role in engineering its electrical behavior, and suggests that developing methods to control or engineer such defects for controlling the electron doping level could be a viable alternative to conventional doping with foreign atoms.

2

ACS Paragon Plus Environment

Page 2 of 24

Page 3 of 24 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

The Journal of Physical Chemistry

1. INTRODUCTION The excellent gate electrostatics and the purported absence of dangling bond makes 2D materials, such as TMDCs, promising candidates for high performance field effect transistor (FET).1-4 Radisavljevic et al. have demonstrated the MoS2-based devices with moderate subthreshold swing approaching 60 mV/dec.5 Bao et al. have fabricated MoS2-based devices with on/off current ratio around 1010 and they also reported the room temperature electron and hole mobilities of 470 and 480 cm2V−1s−1.6 Even though TMDCs FETs exhibit these promising properties; there are still practical and fundamental challenges. For example, the effective masses at the valence band (VB) and conduction band (CB) edges have been theoretically predicted to be approximately symmetric, but MoS2 FETs usually exhibit n-type characteristics. There is still no consensus on the real reasons that account for the n-type behavior.4 The dopants such as the native defects and heteroatoms may be the possible causes, and Fermi-level pinning near the CB edge is another possible reason.7-8 In particular, a fundamental understanding of how doping can be affected by vacancies is lacking. For contacting TMDCs, various metals have been explored with both experimental and theoretical approaches. Das et al. have reported extremely low Schottky barriers on MoS2 which is 0.03 and 0.05 eV for low work function metal scandium (Φ = 3.5 eV) and titanium (Φ = 4.1 eV).9 Low electron Schottky barriers are also formed for high work function metal on MoS2 such as Pd (Φ = 5.6 eV) and Au (Φ = 5.1 eV) whose electron Schottky barrier has been determined as 0.40 and 0.13 eV, respectively.10 Christopher et al. have studied the Au, Ir, Cr and Cs contact interface 3

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 24

with MoS2 and claimed that not only the chemistry present at the contact metal−MoS2 interface but also the deposition chamber ambient can affect the contact metal chemistry.11 Kang et al. have reported a systematic theoretical study comparing In, Ti, Au and Pd contacts to monolayer MoS2 and WSe2 as well as Mo-MoS2 and W-WSe2 with both top and edge contact.12 They have evaluated the contact by calculating the tunnel barrier, Schottky barrier and orbital overlap with density functional theory (DFT) and concluded that Mo and Ti is the best candidates of n-type top-contact metals for monolayer intrinsic MoS2. For monolayer intrinsic WSe2, W and Pd is the best n- and p-type top-contact metal, respectively. Ti-MoS2 and Au-MoS2 top contacts,13 Sc, Ni, and Au contacts to multilayer MoS2,9 Pd-WSe2 contact,2 In-, Al-, and Ag-WSe2 contact,14 and 2D compound metal contacts to MoS215 have also been studied based on the ideal structure. However, DFT models of such interfaces only consider pristine TMDC layers, which does not account for the impact of vacancies which are prevalent in real materials. Some level of vacancies appears unavoidable during the growth of monolayer MoS2, and it is known that these vacancies have considerable effects on electronic properties. However, whether the monolayer MoS2 with VS results in n-type or p-type semiconductor is still controversial. For example, Kim et al. have varied the stoichiometry of monolayer MoS2 during chemical vapor deposition (CVD) via controlled sulfurization and reported the VS as n-type doping.16 Qiu et al. have combined

variable-temperature

transport

measurements,

aberration-corrected

transmission electron microscopy (TEM), DFT and tight-binding calculations and 4

ACS Paragon Plus Environment

Page 5 of 24 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

The Journal of Physical Chemistry

observed VS on the surface of MoS2 acting as electron donors (n-type doping) inducing localized states in the bandgap.17 On the other hand, others claimed that the VS were ptype dopants. Zhou et al. have reported direct observation of various VS in CVD MoS2 with a combination of direct atomic resolution imaging and first-principles calculations.18 They found VS to be detrimental for n-type conductivity of MoS2 as they create deep trap states for electrons and made the MoS2 p-type. DFT calculations by Noh et al. showed that VS acts as electron trap centers (p-type) for the positively charged states of VS are not stable in a single-layer MoS2, and thus are unlikely to be a source of electron carriers.19 Wallace et al. have reported that in general, the stoichiometry of n-type MoS2 appears to be about 1.8:1 (S/Mo) as determined by XPS, whereas in the p-type region, the stoichiometry is 2.3:1.8 However, the electrical doping effect of VS on MoS2 is still unclear. In this study, we consider models which more realistically represent the contact metal−MoS2 interface by including defects in MoS2, in order to obtain an in-depth understanding of the interface interaction and the role of the VS in its conductivity. In addition, we systematically study the electronic structure of isolated MoS2 and MoS2/Au metal contact, with and without different types of VS. By calculating the charge transfer between MoS2 and the Au substrate, we show that the electronic doping effect of VS on MoS2 is sensitive to its concentration. At lower concentration, such as 0.36 and 0.82×1014cm-2 (corresponding to 2.8% and 6.3%, respectively), the VS provide p-type doping while n-type doping is possible when the concentration is higher, such as 1.64, 2.46 and 3.28×1014cm-2 (corresponding to 12.5%, 18.8% and 25.0% 5

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

respectively) . At a more subtle level, we show that even for the same lateral VS concentration, the degree of charge transfer can be tuned by the atomic configuration of the VS. This possibility to tune the doping type by engineering the VS density and/or configuration could be attractive compared to conventional doping with foreign atomic species, especially those carried by molecular precursors which may leave behind unwanted atomic species that may adversely alter the desirable properties of the host material. These findings should find relevance not only in engineering contact interfaces with 2D semiconductors for electronic devices, but also for efforts seeking to exploit the reactivities of such VS for sensing or catalytic applications.20-21 As an example, this paper focuses on the role of such vacancies in the Au/MoS2 interface without considering the complications of surface contaminants such as molecules adsorbed on vacancies. However, we note that such an 'ideal' interface could be approximated in practice if the contacts are deposited on synthesized MoS2 without air exposure.

2. MODELS AND METHODS All calculations were performed by using DFT-based Vienna ab initio simulation package (VASP).22-23 The Perdew−Burke−Ernzerhof (PBE) functional and the projector augmented wave (PAW) potential were used.24-25 While hybrid functionals (e.g. HSE, PBE0) may be used, they have higher computational overheads. Since other works have shown that such hybrid functionals do not affect the general 6

ACS Paragon Plus Environment

Page 6 of 24

Page 7 of 24 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

The Journal of Physical Chemistry

trend,26-27 we have chosen to use the PBE functional in this study for computational efficiency. The electronic wave functions were expanded in plane-wave basis with a cutoff energy of 500 eV. 4×4 and 6×6 supercells were constructed for studying how the interaction of Au and MoS2 is influenced by various VS defects (as shown in Figure 1). The first Brillouin zone of isolated MoS2 and Au-MoS2 stack was sampled by Γcentered 10×10×1 and 2×2×1 K-point meshes, respectively. Sensitivity test with the Kpoint mesh is summarized in Table S1. To minimize the interaction between periodic images in the slabs, a vacuum layer of 25 Å was inserted normal to the MoS2 surface for all interface structures. Grimme’s DFT-D2 method was applied to correct for the van der Waals interactions in DFT.28 All structures were optimized until the Hellmann−Feynman force on each atom was smaller than 0.05 eV/Å. The charge transfer at the interface was studied by the electron density difference map which is defined by the difference in charge density before and after the contact between Au and MoS2: ∆𝜌 = 𝜌𝐴𝑢 ― 𝑀𝑜𝑆2 ― 𝜌𝐴𝑢 ― 𝜌𝑀𝑜𝑆2 The charge transfer amount was evaluated by Bader charge analysis.

7

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Figure 1. Top view of a 4×4 pristine MoS2 unit cell (a) and the unit cell with V1S (b), V2S (c), V3S (d) and V4S (e). The VS concentration is 0.82, 1.64, 2.46 and 3.28 ×1014cm-2 for geometry b, c, d and e, respectively. Yellow and purple spheres represent S and Mo atoms. The pink dots show the position of the VS.

3. Results and discussion Firstly, we carried out detailed study of electronic structure on monolayer MoS2 with VS. Subsequently, the contact between MoS2 and Au were evaluated by physical separation, band structure, electron density difference and Bader charge calculations. Finally, the junction current in an inhomogeneous contact was simulated. 3.1 Electronic structure of MoS2 with VS Native defects are inevitably introduced during the material growth, which occurs often at relatively high temperatures. As any other material, TMDCs have defects, which have significant effects in their electrical29, magnetic30, and optical properties31. Therefore, it is important to study the electronic structure of the MoS2 with VS. The Density of states (DOS) results of the pristine monolayer MoS2 and with VS is shown in Figure 2. The VS introduce defect states near the band edges, and they overlap with 8

ACS Paragon Plus Environment

Page 8 of 24

Page 9 of 24 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

The Journal of Physical Chemistry

both the valence and conduction bands. The Fermi level is located in between these defect states, which agrees with the previous DFT calculations and scanning tunneling spectroscopy results.19, 31-32 The number of defect states increases with the increasing defect concentration. These defect states are expected to play an important role during its contact with Au.

Figure 2. DOS of pristine monolayer MoS2 (a) and monolayer MoS2 with V1S (b), V2S (c), V3S (d) and V4S (e). Black line shows the total DOS; Red and blue line correspond to the partial DOS on Mo d electron and S p electrons, respectively. The band structures projected on Mo atoms are summarized in Figure 3. As shown in Figure 3(a), the pristine monolayer MoS2 has a direct band gap of about 1.84 eV, which is in line with previous DFT calculations.19, 32 Due to the presence of VS, defect states are created at the band gap regions (as shown in Figure 3b-e), and the number of the defect states increase as the vacancy concentration increases. The relative 9

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Fermi energy level (EF) moves towards to the conduction band, which makes the system with higher VS concentration to have more n-type characteristic. Since the band structure is not sufficient to predict a semiconductor as n or p type, here the band structure results are used only to evaluate a general trend. More detailed calculation, such as the formation energy of the defects at various charge states, are necessary for a more accurate prediction of the EF position.

Figure 3. Projected band structure on Mo of pristine MoS2 (a) and MoS2 with V1S (b), V2S (c), V3S (d) and V4S (e). The EF is shown as the red dot line. The pink shadow represents the original band gap area. The red arrow and the number show the energy level distance between EF and CB bottom for each band structure. 3.2 Physical separation of Au and MoS2 The optimized structures of contacted Au-MoS2 systems with and without VS are shown in Figure 4. d is defined as the physical separation, which is the distance between the bottom-most Au layer and the top-most S layer, as shown in Figure 4(a). 10

ACS Paragon Plus Environment

Page 10 of 24

Page 11 of 24 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

The Journal of Physical Chemistry

A smaller d value indicates a closer contact of the metal and MoS2 layers, which enhances the overlap of the electronic orbitals and the formation of covalent bond. The DFT-calculated d for Au-MoS2 without any defects is 2.38 Å, which is slightly smaller than the literature reported value of 2.62 Å.13 This interlayer distance is 0.12 Å longer than the sum of the S and Au covalent radii, and it is large enough to suppress efficient wave function overlap which results in very weak contact between Au and MoS2. This is consistent with the literature, which also claims that Au adheres very weakly to MoS2 and the interface interaction is characterized by van der Waals force.12 The optimized structure of Au-MoS2-V1S is shown in Figure 3(b). There are slight rearrangements of the Au atoms upon the creation of the VS, and the interaction pushes the Au atom slightly higher and away from the VS. For Au-MoS2 with V2S, V3S and V4S vacancies, the Au atoms above the vacancies have a strong tendency to fill in the vacancy site, which enhances the orbital overlap and increase their interaction.

11

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Figure 4. The optimized structures of Au-MoS2 (a), Au-MoS2-V1S (b), Au-MoS2V2S (c), Au-MoS2-V3S (d) and Au-MoS2-V4S (e) contacts. Grey, yellow and purple spheres represent Au, S and Mo atoms. The interface thickness is defined as the distance between the bottom Au layer and the top most S layer, as shown in (a).

3.3 Charge transfer at the interface Detailed charge transfer at the interface is studied by the electron density difference map, which directly shows the charge rearrangement associated with the interaction. Figure 5(a) shows the electron density difference map associated with the interaction between Au and pristine MoS2. There is hardly any charge redistribution upon their contact, which is in line with the literature claiming that the interaction between Au and MoS2 is weak.12 Bader charge calculations reveal a similar trend: the charge transfer upon the contact between Au and pristine MoS2 is almost zero electrons. Figure 5(b) shows the electron density difference map associated with the contact 12

ACS Paragon Plus Environment

Page 12 of 24

Page 13 of 24 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

The Journal of Physical Chemistry

between Au and MoS2-V1S. The electrons tend to move from Au layers to MoS2, and the Bader charge shows the charge transfer of 0.28 e-. The electron transfer from Au to MoS2-V1S suggests that the MoS2 with 1S vacancy is an electron acceptor, and the 1S vacancy provides p-type doping. This is in line with the literature which claims that although the semiconductor with anion defect is normally electron rich, the deep levels below the CB minimum for MoS2 with native VS make them act as an electron trap center.18-19

Figure 5. Electron density difference map associated with the contact between Au and pristine MoS2 (a), MoS2-V1S (b), MoS2-V2S (c), MoS2-V3S (d), MoS2-V4S (e). The grey, yellow and purple spheres represent Au, S and Mo atoms. The blue and red zones correspond to electron deduction and accumulation areas, respectively. The arrow and the number show the charge transfer direction and amount predicted by Bader charge, respectively.

13

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

For the contact between Au and MoS2 with V2S, V3S and V4S vacancies, the electron density difference maps (as shown in figure 5(c-e)) suggest that the electrons have tendency to move from MoS2 to Au, and the amount of charge transfer is larger when the VS concentration is higher. This suggests that V2S, V3S and V4S cause unsaturated electrons in the surrounding Mo atoms, which acts as electron donors, to make the MoS2 electron-rich and serve as the source of electron carriers. This agrees with the projected band structure calculation shown in Figure 3 where the defect state is dominated by Mo. This is also in line with the work by Kim et al. where they tuned the stoichiometry of MoS2 and showed that the VS are sufficiently shallow to act as an electron donor in n-type monolayer MoS2.16 In addition, Qiu et al. provided direct theoretical evidence that VS exist in MoS2, introducing localized donor states inside the bandgap.17 To achieve a lower VS concentration, we also studied a 1S vacancy in a 6×6 unit cell as shown in Figure S1(a). The concentration is 0.36×1014cm-2 and we found the electrons tend to flow from Au to MoS2 suggesting that this defected MoS2 is p-type. To check whether VS configuration could affect the doping, another geometry which has the same lateral concentration of V4S but with the different configuration shown in Figure S1(b) was also studied. As shown in Figure S1(b), the 4 missing S atoms are evenly distributed at the top layer MoS2. However, the amount of charge transfer for this configuration is about 0.80 e-, which is smaller than the one shown in Figure 5(e) with the same VS concentration. Thus, although these two scenarios have 14

ACS Paragon Plus Environment

Page 14 of 24

Page 15 of 24 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

The Journal of Physical Chemistry

the same VS concentration, the different position of the VS leads to a different extent of charge transfer. Overall, the p- to n-type doping transition may be ascribed to the competition between two effects: (i) excess electrons on the Mo atoms left by missing S atoms leading to n-type or donor-like doping16-17, and (ii) the whole surrounding of the defect acting like an electron trap center resulting in p-type or acceptor-like doping18-19.

Figure 6. The amount of charge transfer (per surface Au atom) as a function of the VS concentration. A positive charge transfer amount corresponds to the charge flow from MoS2 to Au, suggesting an n-type contact, and a negative charge transfer denotes an opposite direction, indicating a p-type contact.

Figure 6 summarizes the charge transfer associated with the interaction of Au and MoS2 with different VS concentration. When the VS concentration is lower (1S vacancy within a 6×6 and 4×4 unit cell), the MoS2 serves as an electron acceptor, 15

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

whereas for higher VS (V2S, V3S and V4S within a 4×4 unit cell), interface charge transfer is enhanced and the MoS2 play the role of electron donor. The trend is in line with XPS and STM results which show that for MoSx, the n-type region has the ratio of S/Mo around 1.8:1 and the p-type region has the ratio of 2.3:1.8 We note that Mo vacancies may also be present in real MoS2 monolayers. To provide comparison, we also provide a summary of the same DFT and Bader charge transfer analysis for the 1Mo vacancy and 2Mo vacancies cases in Table S2. Our results suggest that they are p-type doping, in agreement with reported experiments.8 While these findings are consistent with our DFT predictions, they are specific to the Au-MoS2 contact system. However, we do expect similar possibilities to tune the semiconductor-metal contact via vacancy doping for other systems.

3.4 I-V curve simulation Current-Voltage (I-V) simulations have been performed using the thermionic emission model to provide a reference for how device behavior changes with respect to the amount of vacancy doping in the system. Employing the Schottky-Mott model, we assumed the pristine Schottky barrier height of 1 eV for n-type and 0.8 eV for p-type MoS2-Au devices (details in the Supporting Information). For our simulations, we estimated an effective Schottky barrier lowering of 0.2 eV (p-type) for V1S and 0.2 eV (n-type) for V2S and used the relative differences in the conduction band shifts between 16

ACS Paragon Plus Environment

Page 16 of 24

Page 17 of 24 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

The Journal of Physical Chemistry

V2S, V3S and V4S (Figure 3) to obtain respective effective Schottky barrier lowering of 0.38 eV for V3S and 0.48 eV for V4S. The magnitude of Schottky barrier lowering used in our simulation is consistent with a report by McDonnell et al. where a Fermi level shift of up to 1 eV was measured across a single sample.8 Figure 7 shows the simulated I-V characteristic of the MoS2-Au interface for the different vacancy concentrations from which we can conclude that the several orders of current density increase can be effected. The simulated I-V characteristics also protrays visually the switch from p-type at the lower VS concentrations to n-type at the higher VS concentrations. This highlights the possibility of using defect engineering or defect passivation to tune the MoS2-Au contact properties. We would like to emphasize that our I-V simulations are derived from a first order approximation of the effective Schottky barrier height at the MoS2-Au interface based on the relative band shifts from our DFT calculations. We also modelled the effect of an inhomogeneous interface at the atomic level in our simulation using the parallel conduction model8 and found a similar trend in the current densities as our effective barrier lowering model. Our DFT calculations assume a homogeneous distribution of the vacancy through the periodic repetition of unit cells and hence an effective doping model would be more appropriate. In addition, the potential pinch-off effect for Schottky barriers would render atomic size defects invisible, but their electronic doping effect should remain.33-35

17

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Figure 7. Simulated junction currents as a function of Vs for the MoS2-Au interface. (a) V1S junction which correspond to 1S vacancy in a 4×4 unit cell and (b) V2S, V3S and V4S which correspond to 2S, 3S and 4S vacancies in a 4×4 unit cell respectively. The values of the Schottky barrier heights used are shown in the legend and the series resistance is arbitrarily modeled as 25 Ω.

4. CONCLUSION We have carried out a detailed study on the electronic structure of MoS2 with different VS and how these vacancies influence the contact with Au. We find that the effective doping type is highly sensitive to the concentration of VS which provides a possible clarification of the different doping types reported in intrinsic MoS2 due to the presence of VS. We demonstrate the effect of using such VS doping to achieve different contact types with Au as the contact metal. Under lower VS concentration (0.36 and 0.82×1014cm-2), the MoS2 serves as an electron acceptor resulting in a p-type Schottky contact. For higher VS concentrations (1.64, 2.46 and 3.28×1014cm-2), the contact changes to an n-type Schottky behavior due to the VS serving as electron donor. By comparing the charge transfer amount between 2 different configurations of V4S, we show that the configuration is also able to tune the contact. Our study reveals the 18

ACS Paragon Plus Environment

Page 18 of 24

Page 19 of 24 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

The Journal of Physical Chemistry

important role of interface defects of monolayer MoS2 in tailoring the doping in MoS2 without the need of foreign atomic species, and highlights the potential of exploiting such defect engineering for achieving different n- or p-type regions just by varying the defect concentration/configuration. We note here that real 2D materials often incorporate contaminants from the processing environment, and the dangling bonds from VS may also enhance reactivity with various molecular species resulting in other defect complexes that can alter the electronic characteristics of the 2D materials. The study of these is certainly relevant extension from the foundational insights into the first order effects of sulfur vacancies in doping the monolayer MoS2 presented here.

19

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

ACKNOWLEDGMENTS This work was funded by A*STAR Pharos Grant No. 1527000016 and 1527000017.

SUPORTING INFORMATION: Electron density difference map of V1S within a 6×6 unite cell and evenly distributed V4S within a 4×4 unite cell; Charge transfer amount calculated with different K-points mesh; Charge transfer amount calculated for different vacancy types; I-V curve simulation details; Complete references for those with more than 10 authors

20

ACS Paragon Plus Environment

Page 20 of 24

Page 21 of 24 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

The Journal of Physical Chemistry

Reference: 1. Larentis, S.; Fallahazad, B.; Tutuc, E., Field-Effect Transistors and Intrinsic Mobility in Ultra-Thin MoSe2 Layers. Appl. Phys. Lett. 2012, 101, 223104. 2. Fang, H.; Chuang, S.; Chang, T. C.; Takei, K.; Takahashi, T.; Javey, A., HighPerformance Single Layered WSe2 p-FETs with Chemically Doped Contacts. Nano Lett. 2012, 12, 3788-3792. 3. Fuhrer, M. S.; Hone, J.; Measurement of Mobility in Dual-Gated MoS2 Transistors. Nat. Nanotech. 2013, 8, 146. 4. Chhowalla, M.; Jena, D., Zhang, H., Two-Dimensional Semiconductors for Transistors. Nat. Rev. Mater. 2016, 1, 16052. 5. Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A., Single-Layer MoS2 Transistors. Nat. Nanotech. 2011, 6, 147-150. 6. Bao, W.; Cai, X.; Kim, D.; Sridhara, K.; Fuhrer, M. S., High Mobility Ambipolar MoS2 Field-Effect Transistors: Substrate and Dielectric Effects. Appl. Phys. Lett. 2013, 102, 042104. 7. Enyashin, A.; Seifert, G., Electronic Properties of MoS2 Monolayer and Related Structures. Nanosystems:Phys. Chem., Math. 2014, 5, 517-539. 8. McDonnell, S.; Addou, R.; Buie, C.; Wallace, R. M.; Hinkle, C. L., DefectDominated Doping and Contact Resistance in Mos2. ACS Nano 2014, 8, 2880-2888. 9. Das, S.; Chen, H.-Y.; Penumatcha, A. V.; Appenzeller, J., High Performance Multilayer MoS2 Transistors with Scandium Contacts. Nano Lett. 2013, 13, 100-105. 10. Kaushik, N.; Nipane, A.; Basheer, F.; Dubey, S.; Grover, S.; Deshmukh, M. M.; Lodha, S., Schottky Barrier Heights for Au and Pd Contacts to MoS2. Appl. Phys. Lett. 2014, 105. 113505 11. Smyth, C. M.; Addou, R.; McDonnell, S.; Hinkle, C. L.; Wallace, R. M., Contact Metal–Mos2 Interfacial Reactions and Potential Implications on MoS2-Based Device Performance. J. Phys. Chem. C 2016, 120, 14719-14729. 12. Kang, J.; Liu, W.; Sarkar, D.; Jena, D.; Banerjee, K., Computational Study of Metal Contacts to Monolayer Transition-Metal Dichalcogenide Semiconductors. Phys. Rev. X 2014, 4, 031005. 13. Popov, I.; Seifert, G.; Tománek, D., Designing Electrical Contacts to MoS2 Monolayers: A Computational Study. Phys. Rev.Lett. 2012, 108, 156802. 14. Liu, W.; Kang, J.; Sarkar, D.; Khatami, Y.; Jena, D.; Banerjee, K., Role of Metal Contacts in Designing High-Performance Monolayer n-Type WSe2 Field Effect Transistors. Nano Lett. 2013, 13, 1983-1990. 15. Gan, L.Y.; Zhao, Y. J.; Huang, D.; Schwingenschlögl, U., First-Principles Analysis of MoS2/Ti2C and MoS2/Ti2CY2 (Y=F and OH) All-2D Semiconductor/Metal Contacts. Phys. Rev.B 2013, 87, 245307. 16. Kim, I. S. et al., Influence of Stoichiometry on the Optical and Electrical Properties of Chemical Vapor Deposition Derived MoS2. ACS Nano 2014, 8, 10551-10558. 17. Qiu, H., et al., Hopping Transport through Defect-Induced Localized States in Molybdenum Disulphide. Nat. Comm. 2013, 4, 2642. 21

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

18. 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. 19. Noh, J. Y.; Kim, H.; Kim, Y. S., Stability and Electronic Structures of Native Defects in Single-Layer MoS2. Phys. Rev. B 2014, 89, 205417. 20. Li, G., et al., All the Catalytic Active Sites of MoS2 for Hydrogen Evolution. J. Am. Chem. Soc. 2016, 138, 16632-16638 21. Liu, Y.; Xie, Y.; Liu, L.; Jiao, J., Sulfur Vacancy Induced High Performance for Photocatalytic H2 Production over 1T@2H Phase MoS2 Nanolayers. Catal. Sci. Technol. 2017, 7, 5635-5643. 22. Kresse, G., Ab Initio Molecular Dynamics for Liquid Metals. J. Non-Crystal. Solids 1995, 192-193, 222-229. 23. Kresse, G.; Hafner, J., Ab Initio Molecular Dynamics for Open-Shell Transition Metals. Phys. Rev.B 1993, 48, 13115-13118. 24. Perdew, J. P.; Burke, K.; Ernzerhof, M., Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865-3868. 25. Blöchl, P. E., Projector Augmented-Wave Method. Phys. Rev.B 1994, 50, 1795317979. 26. Pilania, G.; Gubernatis, J. E.; Lookman, T., Multi-Fidelity Machine Learning Models for Accurate Bandgap Predictions of Solids. Comp. Mater. Sci. 2017, 129, 156163. 27. Ida, S., et al., A Cocatalyst That Stabilizes a Hydride Intermediate During Photocatalytic Hydrogen Evolution over a Rhodium-Doped TiO2 Nanosheet. Angew. Chem. Int. Ed. 2018, 57, 9073-9077. 28. Wu, X.; Vargas, M. C.; Nayak, S.; Lotrich, V.; Scoles, G., Towards Extending the Applicability of Density Functional Theory to Weakly Bound Systems. J. Phys. Chem. C 2001, 115, 8748-8757. 29. Jena, D.; Konar, A., Enhancement of Carrier Mobility in Semiconductor Nanostructures by Dielectric Engineering. Phys. Rev. Lett. 2007, 98, 136805. 30. Han, S. W., et al., Controlling Ferromagnetic Easy Axis in a Layered MoS2 Single Crystal. Phys. Rev. Lett. 2013, 110, 247201. 31. Korn, T.; Heydrich, S.; Hirmer, M.; Schmutzler, J.; Schüller, C., Low-Temperature Photocarrier Dynamics in Monolayer MoS2. Appl. Phys. Lett. 2011, 99, 102109. 32. Haldar, S.; Vovusha, H.; Yadav, M. K.; Eriksson, O.; Sanyal, B., Systematic Study of Structural, Electronic, and Optical Properties of Atomic-Scale Defects in the TwoDimensional Transition Metal Dichalcogenides MX2 (M = Mo, W; X = S, Se, Te). Phys. Rev. B 2015, 92, 235408. 33. Tung, R. T., The Physics and Chemistry of the Schottky Barrier Height. Appl. Phys. Rev. 2014, 1, 11304. 34. Olberich, A.; Vancea, J.; Kreupl. F.; Hoffmann, H., Potential Pinch-off Effect in Inhomogeneous Au/Co/GaAs67P33(100)-Schottky Contacts. Appl. Phys. Lett. 1998, 70, 2559. 22

ACS Paragon Plus Environment

Page 22 of 24

Page 23 of 24 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

The Journal of Physical Chemistry

35. Wong, C. P. Y, Studying the Metal/Layered Semiconductror Contacts using Temperatre Dependent Current-Voltage Measurements and Ballistic Electron Emission Microscopy, PhD Thesis, National University of Singapore 2018.

23

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Table of Contents Graphic:

The native sulfur vacancies can significantly alter the electrical doping level in MoS2, and tune the material to exhibit conventional n- or p-type semiconductor characteristics.

24

ACS Paragon Plus Environment

Page 24 of 24