Nanopatterning and Electrical Tuning of MoS2 Layers with a

Jul 8, 2015 - Grace G. D. Han , Kun-Hua Tu , Farnaz Niroui , Wenshuo Xu , Si Zhou , Xiaochen Wang , Vladimir Bulović , Caroline A. Ross , Jamie H. Wa...
1 downloads 13 Views 2MB Size
Page 1 of 23

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

Nanopatterning and Electrical Tuning of MoS2 Layers with a Sub-Nanometre Helium Ion Beam Daniel Fox1‡, Yangbo Zhou1‡, Pierce Maguire1, Arlene O’Neill1, Cormac Ó Coileáin1, Riley Gatensby2, Alexey M Glushenkov3,4, Tao Tao3, Georg S Duesberg2, Igor V Shvets1, Mohamed Abid5, Mourad Abid5, Han-Chun Wu6, Ying Chen3, Jonathan N Coleman1, John F Donegan1, Hongzhou Zhang1*

1

School of Physics and CRANN, Trinity College Dublin, Dublin 2, Ireland

2

School of Chemistry and CRANN, Trinity College Dublin, Dublin 2, Ireland

3

Institute for Frontier Materials, Deakin University, Waurn Ponds, VIC 3217, Australia

4

Melbourne Centre for Nanofabrication, 151 Wellington Rd, Clayton, VIC 3168, Australia

5

KSU-Aramco Center, King Saud University, Riyadh 11451, Saudi Arabia

6

School of Physics, Beijing Institute of Technology, Beijing, 100081, P. R. China

Helium Ion Beam; Nanopatterning; MoS2; Stoichiometry; Nanoribbon; Electrical Tuning

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 23

We report sub-nanometre modification enabled by an ultrafine helium ion beam. By adjusting ion dose and the beam profile, structural defects were controllably introduced in a few-layer molybdenum disulphide (MoS2) sample and its stoichiometry was modified by preferential sputtering of sulphur at a few-nanometre scale. Localised tuning of MoS2’s resistivity was demonstrated and semiconducting, metallic-like or insulating material was obtained by irradiation with different doses of He+. Amorphous MoS with metallic behaviour has been demonstrated for the first time. Fabrication of MoS2 nanostructures with 7 nm dimensions and pristine crystal structure was also achieved. The damage at the edges of these nanostructures was typically confined to within 1 nm. Nanoribbons with widths as small as 1 nm were reproducibly fabricated. This nanoscale modification technique is a generalised approach which can be applied to various two-dimensional (2D) materials to produce a new range of 2D metamaterials.

The ability to modify materials at the atomic scale is crucial to fabrication of functional nanoscale building blocks for novel nanodevices. Consistency and reproducibility of modification represent the greatest technical challenges which demand finer structuring capability than the currently available methods. The fundamental restraint for realisation of atomic-scale modification is the ability to form a sub-nanometre energetic impact and to confine the induced interaction within the desired region. The dimension of the damaged region, i.e. the ultimate modification precision, is the convolution of the impact probe and the interaction volume. A thinner specimen will have less damage extension. Therefore, two-dimensional (2D) materials, exemplified by graphene[1-3], provide an ideal platform to realise atomic scale modification. Due to their unique properties[3] more 2D materials such as single layers of

ACS Paragon Plus Environment

2

Page 3 of 23

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

transition metal dichalcogenides (TMDs) have been produced, e.g. molybdenum disulphide (MoS2)[4]. A monolayer of MoS2 has a 1.8 eV direct gap[5] in contrast to the indirect band gap of 1.2 eV of the bulk. Monolayer MoS2 may therefore find applications in optoelectronic devices[6, 7], gas sensing[8, 9] and energy storage[10]. Modification of the crystal structure by structural defects can be used to tune the electronic and optical properties of MoS2[11-13]. It has been shown that a change in the chemical composition of MoS2 can alter the electrical characteristics of the material[14-16]. This is attributed to the preferential removal of sulphur, leading to a more metallic behaviour. Alteration of the geometry of MoS2 can also result in modified electrical and magnetic behaviour[17-19]. Simulations show that MoS2 nanoribbons with specified edge orientations and widths of ~3 nm (within the quantum confinement regime) have extraordinary physical properties. For example, zigzag MoS2 nanoribbons are ferromagnetic and half-metallic whereas armchair MoS2 nanoribbons are non-magnetic and semiconducting. By tailoring the crystal structure, chemical composition and geometry of MoS2, a greater range of applications, such as spintronics and photovoltaic cells[20], could be realised. What is currently required is a method to introduce a precise density of structural defects, tune the chemical composition and pattern nanoscale geometries such as nanoribbons, less than 10 nm wide, with pristine crystallinity and edges of a specified orientation. However, a scalable method to achieve this beyond the laboratory remains elusive. To date, many chemical and physical methods have been used to produce MoS2 nanostructures with a desired geometry. These approaches produced nanoribbons which were either polycrystalline or too wide (>10 nm)[21, 22]. Direct fabrication of MoS2 nanoribbons down to 0.44 nm was demonstrated in a transmission electron microscope (TEM)[15]. The composition and phase of

ACS Paragon Plus Environment

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 23

these structures is currently the subject of discussion[14]. Also, TEM is an impractical fabrication approach due to its requirement of a freestanding sample, its low rate of milling, lack of scalability and its expense. Ion beam irradiation is a widely adopted method to change both the crystal structure[23] and chemical composition[24] of materials. A gallium focused ion beam (FIB) microscope can sputter surface atoms and introduce nanoscale modification effectively with its ~5 nm probe[25]. However, a non-metallic ion species[26] and sub-nm probe size are required for atomic scale modification. A helium ion microscope (HIM) has been shown to be an effective method of introducing a precise density of structural defects by delivering an appropriate He+ dose[27]. This is because the sputtering efficiency of the He+ beam is much larger than an electron beam but smaller than a Ga+ beam, resulting in more efficient and controllable milling[28]. Due to its ~0.35 nm probe size it is also ideally suited to the fabrication of nanostructures[29-31]. For example, graphene nanoribbons with dimensions below 10 nm have been fabricated by this method[32-34]. This ribbon width may be useful in order to exploit the properties of quantum confinement. However, the nanoribbons must also have well-defined edge orientations and good crystallinity. Whether this can be achieved by He+ milling has not yet been investigated due to a lack of understanding and control of the beam-sample interaction. In this work, we have shown that a highly focused beam of helium ions can be used to tune both the crystal structure and the stoichiometry of MoS2 with unprecedented spatial resolution. A nanoscale milling approach was also developed. Freestanding ribbons of MoS2 with pristine crystal structure and widths down to ~7 nm were fabricated. A new controllability has been achieved as evidenced by electrical characterisation. Semiconducting, metallic-like or insulating

ACS Paragon Plus Environment

4

Page 5 of 23

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

MoS2 can be obtained by irradiation with different doses of He+. We illustrate that this is a generalised nano-modification approach by also patterning samples of Mn2O3 and TiO2. Structural Modification The change in crystal structure was analysed by Raman spectroscopy and transmission electron microscopy (TEM). The two characteristic Raman peaks for pristine substrate-supported MoS2 are shown in Figure 1a. This sample was grown by Molecular Beam Epitaxy (MBE). The peak positions were measured to be 382.3 and 406.6 cm-1 for the E12g (the in-plane Mo-S vibration) and A1g (the out-of-plane Mo-S vibration) modes respectively. These agree well with the accepted values from the literature[35], demonstrating the high quality of our sample. A plot of the full width at half maximum (FWHM) of the E12g and A1g peaks as a function of He+ dose is shown in Figure 1b. The width of these peaks was observed to change for doses above (1.00 ± 0.02) × 1014 He+/cm2 (The dose error was estimated by the variation of beam current, see the discussion of beam stability in the supplementary information). With greater He+ doses, the FWHMs of the E12g and A1g peaks increase and decrease respectively. The increase of the E12g peak FWHM is due to the introduction of more in-plane defects during irradiation. The A1g peak FWHM decreases because the A1g peak is not only affected by the presence of defects but is also very sensitive to layer thickness[36]. This is changing in our samples by material removal with high He+ doses. This demonstrates the ability of He+ irradiation to alter the crystal structure by defect introduction and material removal. Further Raman data is presented in Figure S1. The crystal structure was directly observed by highresolution TEM imaging of a freestanding mechanically exfoliated sample. Figure 1c shows the hexagonal structure of pristine MoS2. In Figure 1d, at a He+ dose of (1.00 ± 0.02) × 1018 He+/cm2 the hexagonal crystal structure was no longer observed, the sample was amorphous. For

ACS Paragon Plus Environment

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 23

a freestanding sample, a He+ dose which is one order of magnitude greater is required to create the same irradiation effects as a sample on a substrate[27]. It is clear from the Raman and TEM results that the crystal structure of the MoS2 was controllably amorphised by the He+ irradiation process. Stoichiometry Tuning To gain a better understanding of the He+ irradiation process, chemical composition analysis was performed using energy-dispersive X-ray spectroscopy (EDX). Doses from (1.00 ± 0.02) × 1016 to (1.60 ± 0.03) × 10 He+/cm2 were used to irradiate a mechanically exfoliated, freestanding, few-layer MoS2 sample. The irradiated region of MoS2 is displayed in the highangle annular dark-field (HAADF) scanning TEM (STEM) image in Figure 2a. Each region (labelled from 1 to 12) was irradiated with twice the dose of the previous region, with a starting dose of (1.00 ± 0.02) × 1016 He+/cm2 at region 1. At doses of (2.56 ± 0.05) × 10 He+/cm2 (region 9) and above milling was observed. EDX spectra from each of the regions labelled 1–8 were recorded. The evolution of the stoichiometry of the material with respect to the He+ dose is shown in Figure 2b. It was observed that as the He+ dose increased to 10 He+/cm2 (where the sample is predominantly amorphous), preferential sputtering of sulphur from the sample occurred. Intuitively it is clear why sulphur should be subjected to preferential erosion compared to molybdenum. The mass of sulphur atoms is three times less than that of molybdenum. Upon collision, He+ transfers more energy to sulphur than to molybdenum. As in the collision of spherical objects in classical mechanics, the closer the masses of the balls, the more efficient the exchange of energy is. The helium ion is much lighter than both sulphur and molybdenum, but is still much closer in mass to sulphur than molybdenum. After the collision sulphur atoms will have a lot more energy than molybdenum atoms. This preferential sputtering of sulphur has been

ACS Paragon Plus Environment

6

Page 7 of 23

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 for both electron and argon ion beams in the past[14, 37]. However, our technique enables localised stoichiometry modification and areas with a range of stoichiometries were produced within a region of just a few tens of nm. At doses above (2.56 ± 0.05) × 10 He+/cm2 complete removal of material was observed for this sample. Nanostructure Fabrication For few-layer samples of MoS2 the interaction volume can be approximated to an interaction area (see supplementary information for an investigation of sample thickness effect). In this case, the He+ probe size plays a crucial role on the lateral damage extension and crystal structure of milled nanosystems. An experiment was designed in order to better understand the relationship between the probe size and lateral damage extension in the material. This enables control of the width of the amorphous edge region. As an example, defocus was used to adjust the probe size to 12 ±1, 5.9 ±0.7 and 1.7 ±0.2 nm (details on the probe size measurement can be found in the supporting information). In Figure 3, three different edges were fabricated on a mechanically exfoliated, freestanding MoS2 flake. Each edge was fabricated with a different probe size. Each probe delivered the same dose of (1.90 ± 0.04) × 10 He+/cm2. The structure of each edge was imaged in high resolution TEM. Figures 3a, 3b and 3c correspond to the ~12, ~5.9 and ~1.7 nm probe sizes respectively. The lattice structure was not affected at a distance (white line) from the edge which is approximately equal to the probe size (red circle). We simulated the distribution of He+ ions for the three different probe sizes used in the experiment. During imaging and milling, the He+ beam is moved from pixel to pixel in a raster scan. An array of Gaussian distributions was simulated in order to account for beam overlap during scanning (Figure S4). The milled MoS2 edge was assumed to be found where the dose drops below (1.30 ± 0.03) × 10 He+/cm2, a dose below this value was found experimentally

ACS Paragon Plus Environment

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 23

to produce extensive damage but not complete milling. The point at which the crystal structure was assumed to remain intact was where the ion dose drops below (1.00 ± 0.02) × 10 He+/cm2, as determined by Raman spectroscopy. Our simulation predicts an amorphous region which extends 7, 3.5 and 1.0 nm for the 12 ±1, 5.9 ±0.7 and 1.7 ±0.2 nm probes respectively. The simulation data corresponds well with the probe radii (half the FWHM). Further analysis of the extension of the amorphous region from the milled edges was done using local FFTs from the TEM images. This data can be found in the supporting information (Figure S5). The strong agreement between the experimental results and simulation indicates that for this sample (which is thin and was irradiated with a sufficient dose for milling) control of the edge damage range can be achieved by adjusting the size of the He+ probe. With the probe size (1.0 ± 0.5 nm) and He+ dose ((3.10 ± 0.06) × 10 He+/cm2) optimised, a few-layer MoS2 flake was selected and nanoribbons were fabricated. In Figure 4a a 9 nm wide ribbon is shown. The well-defined crystal lattice fringes span across the whole ribbon, demonstrating that crystal structure of the MoS2 remains entirely intact. In Figure 4b a 5 nm wide ribbon with an amorphous structure can be seen. The width of the narrowest structure that can be reproduced consistently is less than 1 nm, while the structure is no longer crystalline MoS2 (see Figure 1c). The narrowest crystalline nanoribbon was 7 nm wide as shown in Figure 4d. MoS2 ribbons are theoretically stable down to widths of ~1 nm[19]. We speculate that a heating effect may be occurring during milling[38], which might destabilise ribbons narrower than 7 nm. Our future work will aim to address this question. The edges of these crystalline ribbons exhibit well-defined lattice configuration, indicating a subnanometre precision of structuring. More importantly, this fabrication technique is not limited to MoS2 or graphene. A 9 nm wide Mn2O3 nanoribbon is shown in Figure 4e with intact crystal

ACS Paragon Plus Environment

8

Page 9 of 23

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

structure. In Figure 4f a TiO2 sample with minimal edge damage is shown. An array of 10 nm wide nanoribbons was fabricated in Mn2O3 (Figure 4g). The scalability of this method means that these ribbon structures could be fabricated over a millimetre region. Electrical Characterisation To investigate the effects of structure, stoichiometry and geometry changes on the transport properties of an MoS2 flake, we performed electrical measurements. Mechanically exfoliated bilayer MoS2 on a 300 nm SiO2/Si substrate was used for these experiments. In Figure 5a the resistivity evolution of mechanically exfoliated bilayer MoS2 as a function of He+ dose was plotted, which shows four distinct stages. In the first stage, the resistivity decreased with the increasing dose and dropped by 28 % at a dose of (1.00 ± 0.02) × 1014 He+/cm2. The second stage shows a dramatic increase in resistivity (by five orders of magnitude) as the dose increased from 1014 to 1015 He+/cm2. As the dose further increased to 1017 He+/cm2, the resistivity drastically decreased, recovered its initial value of the intact sample (106 Ω), and decreased to a lower value (105 Ω, one order of magnitude smaller than the initial value). In the last stage, the resistivity increased sharply to 1011 Ω, as the dose increased to 1018 He+/cm2. To understand the change in the resistivity, I-V characteristics were measured (Figure 5b). The pristine MoS2 flake exhibited a nonlinear current-voltage response (red I-V response curve in Figure 5b), revealing the semiconducting nature of the pristine MoS2. The semiconducting behaviour of the irradiated MoS2 flake was well maintained in the first stage of beam irradiation, while the high resistivity indicates that the MoS2 flake has been tuned from a semiconductor to an insulator (blue I-V response curve in Figure 5b) in the second stage. For irradiation doses higher than 1015 He+/cm2 (the third stage) a linear current-voltage response (green curve in Figure 5b) was observed. Gate bias measurements were conducted in order to further investigate

ACS Paragon Plus Environment

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 23

this behaviour. Pristine MoS2 showed a semiconducting behaviour with electron doping. The resistivity was tuned over 3 orders of magnitude by applying a back gate from -40 V to +40 V. The resistivity of MoS2 irradiated with a dose of 1017 He+/cm2 could only be tuned ~3 % when applying a back gate from -40 V to +40 V. The linear I-V characteristic, together with the greatly reduced and gate-independent resistivity, indicates that the MoS2 irradiated with 1017 He+/cm2 was modified to metallic-like behaviour. The decrease in resistance of just one order of magnitude observed between the semiconducting to metallic-like MoS2 was also reported in the semiconductor to metal phase transition process that occurs in other binary compound transition metals, such as MoTe2[39] and FeSi2[40]. Doses above 1017 He+/cm2 (the fourth stage) increased the flake resistivity while maintaining its metallic behaviour. At a very high irradiation dose of 1018 He+/cm2, the flake behaved as an insulator again (resistivity >1011 Ω). The morphology modification introduced in previous sections can help to clarify the observed phase transitions. An order of magnitude greater He+ dose is required to create the same irradiation effects as a sample on a substrate[27]. For example, as the EDX was done on a freestanding sample it would be expected that the change in stoichiometry observed at 1018 He+/cm2 would occur at a dose closer to 1017 He+/cm2 for a sample on a substrate. The first stage of irradiation (Figure 5a) introduced structural defects, while keeping the crystal structure. The structural defects act as dopants to the MoS2 lattice and cause the decrease in the resistance[41]. This semiconductor-to-insulator transition in the second stage can be explained as the formation of disordered domains in the amorphised MoS2 flake[11]. The metallic-like behaviour of this sample in the third stage is most likely due to the preferential removal of sulphur and increased proportion of molybdenum. It has been reported that the existence of these sulphur vacancies can result in the appearance of a mid-gap state in the band gap of MoS2[42, 43], forming an

ACS Paragon Plus Environment

10

Page 11 of 23

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

electronic channel and improving the conductivity of MoS2. The increase of resistance in the last stage can be attributed to the removal of MoS2 due to excessive milling effects. We note that while other methods have been used to produce insulating MoS2[44, 45], the observed transition from insulator to metal-like behaviour has not been reported. It is possible that other atomic species, e.g. carbon, were incorporated into our sample to passivate the sulphur vacancies, resulting in the metallic-like behaviour observed. More experiments are required to further characterise the metallic-like behaviour observed here (such as temperature dependence of the resistance) and will be conducted in order to further develop our model in a separate paper. The substrate supported MoS2 can be milled into nanoscale geometries such as a ribbon with a width below 20 nm (inset of Figure 5b). The fabricated MoS2 nanoribbon had a nonlinear current-voltage response (orange curve in Figure 5b), it therefore still exhibited a semiconducting behaviour. In order to control the electronic and magnetic properties of MoS2 which occur due to quantum confinement, the ribbon width must be reduced by another order of magnitude [18]. In summary, this He+ based patterning technique provides sub-nanometre scale precision milling for 2D materials with a scalability not previously practical in electron or ion based milling. It has been shown that the structure, composition, geometry and material properties of MoS2 can be manipulated and tailored using He+ irradiation. This was achieved by controlling the ion dose and the irradiated area. With a ~1 nm He+ probe, pristine nanoribbons of MoS2 and Mn2O3 with sub-10 nm widths were fabricated. This technique can be applied to a range of 2D material systems such as those produced by Coleman et al.[46] in order to fabricate nanostructures whose shape, orientation, crystal structure and stoichiometry can be modified. This will result in the

ACS Paragon Plus Environment

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 23

production of novel nanostructures with tuneable properties for a wide range of device applications. Methods Sample Preparation Chemically exfoliated MoS2 flakes: MoS2 flakes with few-layer thickness were prepared by sonication of MoS2 powder in Nmethyl-pyrrolidone as by O’Neill et al.[47]. The MoS2 dispersion was dropped directly onto a TEM grid. The grid was left to dry in air overnight. The sample was then baked at 80 °C for 3 hrs. MBE MoS2 sample: A 2.5 nm thick Mo layer was grown on MgO with a substrate temperature of 600 °C during deposition (prior to growth the substrate was annealed at 500 °C in UHV for half an hour followed by 2 hrs at 500 °C in 2 × 10 Torr O2). The MoOx sample was inserted into a tube furnace and sulphurised at 700 °C for 1 to 2 minutes. The result was a layer of high quality MoS2. Mechanical exfoliation of MoS2: MoS2 flakes were exfoliated from bulk MoS2 crystals (SPI Supplies) on Si substrates with a 285 nm oxide layer by the Scotch-tape micromechanical cleavage technique. The layer thickness was identified using optical contrast and atomic force microscopy (AFM). Few layer samples (monolayer to trilayer) were selected for the experiment. Mn2O3 and TiO2: Two-dimensional nanomaterials of Mn2O3 and TiO2 were obtained as follows. Commercially available MnO2 and rutile TiO2 powders were used as starting materials (Sigma Aldrich). The

ACS Paragon Plus Environment

12

Page 13 of 23

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

powders were ball milled in a magneto-ball mill for a 150 hrs under argon atmosphere. The as produced ball milled powders were hydrothermally treated in 2 M NaOH aqueous solutions at 120 °C. To recover the structures in the solid precipitate, the suspensions produced in each experiment were filtered, and the obtained solid samples were washed and dried.

Helium Ion Microscopy Freestanding samples: A holey carbon coated TEM grid was used to support freestanding MoS2. The grid was inserted into an SEM stub designed to hold TEM grids and loaded into the HIM. The surface beneath the TEM grid (several mm below) was coated with a non-conductive carbon layer. The carbon layer has a very low SE yield due to the charging effect; the positively charged surface inhibits the escape of negative SEs. The yield of backscattered ions is also very low for carbon, less than 0.2 % as calculated by a SRIM simulation[48]. This ensures that the SE contrast between the flake and the surrounding hole is maximised. It also ensures that the milling is done by the primary ion beam, and not the less localised backscattered ions, resulting in the highest milling resolution. Once a thin, free-standing flake was located the He+ beam was used to remove a region of material. A beam current of 1.00 pA with a variation of 0.02 pA (see the discussion in the supplementary information), a beam energy of 30 keV, a dwell time of 3 - 5 ms per pixel and a pixel spacing of 1 nm were used for milling. Supported sample: For Raman spectroscopy the MBE grown MoS2 sample was used. A mark was made on the sample with a sharp blade, this ensured that the irradiated areas could be relocated after transfer from the HIM to the optical microscope. An array of 5 × 5 μm areas were irradiated with a

ACS Paragon Plus Environment

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 23

range of He+ doses. A 1.00 ± 0.02 pA beam current and 10 nm pixel spacing were used. The beam dwell time at each pixel was varied to achieve the desired dose. The HIM chamber pressure was 35 µPa. After HIM irradiation all samples were exposed to air before any of the following characterisation techniques were used. Raman Spectroscopy Raman spectroscopy was carried out at atmospheric pressure with a Horiba Jobin-Yvon 633 nm laser equipped with 1,200 grooves/mm and a CCD camera. Acquisition time was fixed at 50 seconds. A 100 × objective lens was used. The laser spot size was ~1.05 µm. An array of 5 × 5 µm areas were irradiated with helium ion doses ranging from 1013 to 1017 He+/cm2. This range of doses was selected given that a typical HIM imaging dose is ~1014 He+/cm2 and the milling dose of a monolayer of MoS2 is ~1018 He+/cm2. Electrical Measurements MoS2 transistor devices were fabricated by the electron beam lithography process with the film deposition of a 5 nm Ti and 45 nm Au. The electrical measurements were carried by an Agilent B2912A dual channel sourcemeter with a nano-manipulator system (Imina technologies miBots) for electrical contacts. The He+ irradiated devices were measured in ambient conditions while the ribbon devices were measured in a SEM vacuum chamber (base pressure of 300 µPa). Transmission Electron Microscopy The TEM used was an FEI Titan 80-300 operated at 300 kV. The TEM chamber pressure was 40 µPa. For EDX mapping the system was operated in STEM mode. The STEM beam had a probe size of 0.5 nm and a probe current of 0.5 nA. A pixel spacing of 1.7 nm and an acquisition time of 1 s per pixel were used during mapping.

ACS Paragon Plus Environment

14

Page 15 of 23

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 1| Structural modification of MoS2 using He+. a, Raman spectrum of the E12g and A1g modes of MgO substrate supported pristine ~6 layer MoS2 at 382.3 and 406.6 cm-1 respectively. b, FWHM of the E12g and A1g modes as a function of He+ irradiation dose. The error bars were obtained by the peak fittings in Figure 1a. c, High-resolution TEM image of pristine MoS2. d, High resolution-TEM image of MoS2 after irradiation with (1.00 ± 0.02) × 1018 He+/cm2. Scale bars are 2 nm.

ACS Paragon Plus Environment

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 23

Figure 2| Stoichiometry alteration of an MoS2 flake by He+ irradiation. a, STEM image of a freestanding MoS2 flake irradiated with 15 different He+ doses. Each region was irradiated with twice the dose of the previous region, starting with a dose of (1.00 ± 0.02) × 1016 He+/cm2 at region 1. Regions 9 ((2.56 ± 0.05) × 10 He+/cm2) and above were milled. b, Concentration of Mo and S for different He+ doses as extracted from EDX analysis. The error bars were obtained by the fitting of EDX data.

Figure 3| HRTEM images of the edges of three regions in the same freestanding MoS2 flake milled with three different probe sizes. a, b and c were milled with 12 ±1, 5.9 ±0.7 and 1.7 ±0.2 nm He+ probes respectively. The probe size is indicated by the red dashed circles overlaid on each image. The white dashed lines approximately show the edge of the amorphous region and agree well with the simulated values for the damage extension.

ACS Paragon Plus Environment

16

Page 17 of 23

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 | He+ fabricated freestanding nanoribbons in 2D materials. a, TEM image of a 9 nm wide crystalline MoS2 nanoribbon. b, A 5 nm wide amorphous MoS2 nanoribbon. c, An amorphous nanoribbon with a minimum width of less than 1 nm. d, STEM image of a 7 nm wide crystalline MoS2 nanoribbon. e, A 9 nm wide crystalline Mn2O3 nanoribbon milled with HIM. f, TiO2 edge milled with HIM. g, An array of 10 nm wide Mn2O3 nanoribbons milled with HIM.

ACS Paragon Plus Environment

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 23

Figure 5| Electrical characterisation of He+ modified devices. a, Double log plot of electrical resistivity versus He+ dose for a substrate supported, mechanically exfoliated bilayer MoS2 flake. The letters S, I and M correspond to regions with semiconducting, insulating and metallic-like behaviour respectively. Inset: Resistivity measurement as a function of gate bias. The resistivity of the pristine sample (dashed red line) was widely tuneable. The sample irradiated with 1017 He+/cm2 showed little gate response (solid green line). b, A 20 nm wide, electrically contacted, MoS2 nanoribbon. The ribbon is shown with red false colouring indicating the measured region, the gold electrodes are also coloured. The blue dashed line shows the isolating cut made with the He+ beam. Scale bar is 500 nm. c, Current-voltage plots for the indicated doses as well as for the 20 nm wide MoS2 nanoribbon. All of the results were measured at 0 V back gate. ASSOCIATED CONTENT Supporting Information. Raman spectra details; sample thickness effect investigation; probe size measurement method; beam stability; ion dose distribution simulation; milling dose calculation. This material is available free of charge via the Internet at http://pubs.acs.org.

ACS Paragon Plus Environment

18

Page 19 of 23

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

AUTHOR INFORMATION Corresponding Author *Corresponding author: [email protected] Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. ‡These authors contributed equally. The authors declare no competing financial interest. ACKNOWLEDGMENT We would like to thank the staff at the Advanced Microscopy Laboratory (AML), CRANN, Trinity College Dublin. We would like to acknowledge support from the following funding bodies: Science Foundation Ireland [grant numbers: 12/RC/2278, 11/PI/1105, 07/SK/I1220a, and 08/CE/I1432] and the Irish Research Council [grant number: EPSG/2011/239]. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.

Geim, A.K. and K.S. Novoselov, The Rise of Graphene. Nature Materials, 2007. 6(3): p. 183-191. Geim, A.K., Graphene: Status and Prospects. Science, 2009. 324(5934): p. 1530-1534. Novoselov, K.S., et al., A Roadmap for Graphene. Nature, 2012. 490(7419): p. 192-200. Joensen, P., R.F. Frindt, and S.R. Morrison, Single-Layer MoS2. Materials Research Bulletin, 1986. 21(4): p. 457-461. Mak, K.F., et al., Atomically Thin MoS2: A New Direct-Gap Semiconductor. Physical Review Letters, 2010. 105(13): p. 136805. Lopez-Sanchez, O., et al., Ultrasensitive Photodetectors Based on Monolayer MoS2. Nature Nanotechnology, 2013. 8(7): p. 497-501. Newaz, A.K.M., et al., Electrical Control of Optical Properties of Monolayer MoS2. Solid State Communications, 2013. 155(1): p. 49-52. Yue, Q., et al., Adsorption of Gas Molecules on Monolayer MoS2 and Effect of Applied Electric Field. Nanoscale Research Letters, 2013. 8(1): p. 425. Lee, K., et al., High-Performance Sensors Based on Molybdenum Disulfide Thin Films. Advanced Materials, 2013. 25(46): p. 6699-6702.

ACS Paragon Plus Environment

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

10.

11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

21.

22. 23. 24. 25. 26. 27. 28. 29. 30.

Page 20 of 23

Peng, Q. and S. De, Outstanding Mechanical Properties of Monolayer MoS2 and its Application in Elastic Energy Storage. Physical Chemistry Chemical Physics, 2013. 15(44): p. 19427-19437. Ghorbani-Asl, M., et al., Defect-Induced Conductivity Anisotropy in MoS2 Monolayers. Physical Review B, 2013. 88(24): p. 245440. Zhou, W., et al., Intrinsic Structural Defects in Monolayer Molybdenum Disulfide. Nano Letters, 2013. 13(6): p. 2615-2622. Kang, N., et al., Photoluminescence Quenching in Single-Layer MoS2 via Oxygen Plasma Treatment. The Journal of Physical Chemistry C, 2014. 118(36): p. 21258-21263. Liu, X., et al., Top–Down Fabrication of Sub-Nanometre Semiconducting Nanoribbons Derived from Molybdenum Disulfide Sheets. Nature Communications, 2013. 4: p. 1776. Lin, J., et al., Flexible metallic nanowires with self-adaptive contacts to semiconducting transition-metal dichalcogenide monolayers. Nat Nano, 2014. 9(6): p. 436-442. Kim, I.S., et al., Influence of Stoichiometry on the Optical and Electrical Properties of Chemical Vapor Deposition Derived MoS2. ACS Nano, 2014. 8(10): p. 10551-10558. Botello-Méndez, A.R., et al., Metallic and Ferromagnetic Edges in Molybdenum Disulfide Nanoribbons. Nanotechnology, 2009. 20(32): p. 325703. Li, Y., et al., MoS2 Nanoribbons: High Stability and Unusual Electronic and Magnetic Properties. Journal of the American Chemical Society, 2008. 130(49): p. 16739-16744. Ataca, C., et al., Mechanical and Electronic Properties of MoS2 Nanoribbons and Their Defects. The Journal of Physical Chemistry C, 2011. 115(10): p. 3934-3941. Pan, H. and Y.W. Zhang, Tuning the Electronic and Magnetic Properties of MoS2 Nanoribbons by Strain Engineering. The Journal of Physical Chemistry C, 2012. 116(21): p. 11752-11757. Li, Q., et al., Molybdenum Disulfide Nanowires and Nanoribbons by Electrochemical/Chemical Synthesis. The Journal of Physical Chemistry B, 2005. 109(8): p. 3169-3182. Ghosh, S.K., et al., Simple Formation of Nanostructured Molybdenum Disulfide Thin Films by Electrodeposition. International Journal of Electrochemistry, 2013. 2013: p. 7. Wahl, K.J., D.N. Dunn, and I.L. Singer, Effects of ion implantation on microstructure, endurance and wear behavior of IBAD MoS2. Wear, 2000. 237(1): p. 1-11. Quan, M., et al., Controlled argon beam-induced desulfurization of monolayer molybdenum disulfide. Journal of Physics: Condensed Matter, 2013. 25(25): p. 252201. Orloff, J. and L.W. Swanson, Field Ionisation Sources, in High Resolution Focused Ion Beams: FIB and Its Applications. 2003, Springer. p. 5-20. Kiener, D., et al., FIB damage of Cu and possible consequences for miniaturized mechanical tests. Materials Science and Engineering: A, 2007. 459(1–2): p. 262-272. Fox, D., et al., Helium ion microscopy of graphene: beam damage, image quality and edge contrast. Nanotechnology, 2013. 24(33). Yang, J., et al., Rapid and precise scanning helium ion microscope milling of solid-state nanopores for biomolecule detection. Nanotechnology, 2011. 22(28): p. 285310. Fox, D., et al., Nano-structuring, surface and bulk modification with a focused helium ion beam. Beilstein Journal of Nanotechnology, 2012. 3: p. 579-585. Lemme, M.C., et al., Etching of Graphene Devices with a Helium Ion Beam. ACS Nano, 2009. 3(9): p. 2674-2676.

ACS Paragon Plus Environment

20

Page 21 of 23

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

31. 32.

33.

34.

35. 36. 37.

38.

39. 40. 41.

42. 43.

44. 45. 46. 47.

48.

Bell, D.C., et al., Precision cutting and patterning of graphene with helium ions. Nanotechnology, 2009. 20(45). Kalhor, N., S.A. Boden, and H. Mizuta, Sub-10nm patterning by focused He-ion beam milling for fabrication of downscaled graphene nano devices. Microelectronic Engineering, 2014. 114(0): p. 70-77. Pickard, D. and L. Scipioni. Graphene Nano-Ribbon Patterning in the Orion Plus. 2009 [cited 2015 16/05]; Application Note]. Available from: http://www.tcd.ie/Physics/ultramicroscopy/teaching/PY5019/HIM/AN_ORION_PLUS_ Graphene.pdf. Abbas, A.N., et al., Patterning, Characterization, and Chemical Sensing Applications of Graphene Nanoribbon Arrays Down to 5 nm Using Helium Ion Beam Lithography. ACS Nano, 2014. 8(2): p. 1538-1546. Lee, C., et al., Anomalous Lattice Vibrations of Single- and Few-Layer MoS2. ACS Nano, 2010. 4(5): p. 2695-2700. Li, H., et al., From Bulk to Monolayer MoS2: Evolution of Raman Scattering. Advanced Functional Materials, 2012. 22(7): p. 1385-1390. Baker, M.A., et al., XPS Investigation of Preferential Sputtering of S from MoS2 and Determination of MoSX Stoichiometry from Mo and S Peak Positions. Applied Surface Science, 1999. 150(1–4): p. 255-262. Nastasi, M., J.W. Mayer, and J.K. Hirvonen, Ion-Solid Simulations and Diffusion, in IonSolid Interactions: Fundamentals and Applications. 1996, Cambridge University Press. p. 209-216. Vellinga, M.B., R. de Jonge, and C. Haas, Semiconductor to metal transition in MoTe2. Journal of Solid State Chemistry, 1970. 2(2): p. 299-302. Nishida, I., Study of Semiconductor-to-Metal Transition in Mn-Doped FeSi2. Physical Review B, 1973. 7(6): p. 2710-2713. Komsa, H.P., et al., Two-Dimensional Transition Metal Dichalcogenides under Electron Irradiation: Defect Production and Doping. Physical Review Letters, 2012. 109(3): p. 035503. Nagisa, K., et al., Electronic States of Sulfur Vacancies Formed on a MoS2 Surface. Japanese Journal of Applied Physics, 2010. 49(8S3): p. 08LB01. Kim, B.H., et al., Effect of sulphur vacancy on geometric and electronic structure of MoS2 induced by molecular hydrogen treatment at room temperature. RSC Advances, 2013. 3(40): p. 18424-18429. Islam, M.R., et al., Tuning the Electrical Property via Defect Engineering of Single Layer MoS2 by Oxygen Plasma. Nanoscale, 2014. 6(17): p. 10033-10039. Kim, T.Y., et al., Irradiation Effects of High-Energy Proton Beams on MoS2 Field Effect Transistors. ACS Nano, 2014. 8(3): p. 2774-2781. Coleman, J.N., et al., Two-Dimensional Nanosheets Produced by Liquid Exfoliation of Layered Materials. Science, 2011. 331(6017): p. 568-571. O’Neill, A., U. Khan, and J.N. Coleman, Preparation of High Concentration Dispersions of Exfoliated MoS2 with Increased Flake Size. Chemistry of Materials, 2012. 24(12): p. 2414-2421. Ziegler, J. Stopping and Range of Ions in Matter (SRIM). [cited 2015 26/05]; Available from: http://www.srim.org/.

ACS Paragon Plus Environment

21

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 22 of 23

Table of contents graphic.

ACS Paragon Plus Environment

22

Page 23 of 23

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

85x45mm (300 x 300 DPI)

ACS Paragon Plus Environment