Bottom-up direct writing approach for controlled fabrication of WS₂/MoS₂ heterostructure systems

Abstract

The ability to construct heterostructures that consist of layered transition metal dichalcogenides materials (MX₂) in a controlled fashion provides an attractive solution for materials design and device applications. For nanotechnology applications it is important to control the shape, geometry and precise position of the grown heterostructure assemblies on a variety of substrates. In this study, we developed a “direct writing” technique to fabricate arrays of WS₂/MoS₂ and MoS₂/WS₂/MoS₂ heterostructures at predefined locations on a silicon substrate in a controlled fashion. Water based precursor inks were implemented to ensure the formation of quality heterostructure surfaces which are free of residual polymer contaminants. With this technique we demonstrate an attractive and scalable technology with unique capabilities for precise growth of dissimilar MX₂ materials, layered either in a vertical or lateral arrangement.

---

Introduction

The progress of graphene research has stimulated interest in other types of two-dimensional (2D) atomic crystals, such as hexagonal boron nitride, germanene, and layered transition metal dichalcogenides (TMDCs) or MX₂ (M: transition metal; X: chalcogen). Although MX₂ materials share similar crystalline structures, their physical properties (band gap, light matter interaction, and spin–orbit coupling strength) can vary significantly. Among these 2D materials, the applications of MX₂ in logic electronics are promising^1–8 because of its sizable bandgaps of around 1–2 eV, comparable to those of conventional groups of semiconductors (groups IV, III–V). In parallel with the study of single layer MX₂ materials, van der Waals (vdW) heterostructures that consist of dissimilar MX₂ materials, arranged in a vertical direction, have recently been gaining extensive attention. In typical semiconductor heterostructures made of two or more dissimilar materials, highly matched crystalline lattices are required to yield high quality interfaces, hence the designs for multi-compositional conventional semiconductors is complex. In the MX₂ heterostructures however, high quality interfaces can be achieved even in the mismatched systems.⁹ That is due to the nature of the weak vdW forces that are governing the interactions in layered MX₂ materials.^9–14 Novel properties/phenomena that have been revealed in MX₂ heterostructures could trigger a revolution in the design of heterostructure systems for applications such as photovoltaics, optoelectronics, and spontaneous water splitting.^10,11

Several approaches have been developed to prepare MX₂ layers such as micromechanical exfoliation, chemical/electrochemical exfoliation, and chemical vapour deposition (CVD). To date, fabrication of MX₂ heterostructures are mainly based on the combination of exfoliation and CVD methods. Huo et al.^12,13 micromechanically exfoliated MoS₂ and WS₂ flakes, and demonstrated the transfer of WS₂ layer onto MoS₂ flakes using polymethyl methacrylate (PMMA) coating and sodium hydroxide (NaOH) solution etching methods. Yu et al.⁹ prepared MoS₂ and WS₂ layer using CVD methods then lifted off and transferred the synthesized MoS₂ onto as-grown WS₂ layer manually. Chen et al.¹⁵ fabricated MoS₂ by CVD method, and WS₂ layer was obtained by the thermolysis of ammonium tungstate hydrate in an inert gas environment. Zhang et al.¹⁶ synthesized MoS₂/WS₂ heterostructures using core–shell WO_3−x/MoO_3−x nanowires as precursors. Jung et al.¹⁰ deposited and patterned tungsten/molybdenum (W/Mo) and then introduced sulfur (S) vapor for the growth of heterostructure. Gong et al.¹⁷ fabricated vertically stacked and in-plane (lateral) WS₂/MoS₂ heterostructures by controlling the CVD growth temperature and using sulphur, tungsten, and molybdenum oxide as precursors. Although exciting properties have been discovered in these heterostructure systems fabricated with different processing methods, it is challenging to control the shapes, geometry and precise positions of MX₂ heterostructure formations at preselected locations. In order to control the position of MoS₂ and WS₂ layers, additional photo/e-beam lithography is usually required, which inevitably introduces organic/polymer residues and may introduce strain to the interface between the layers and thus could degrade the heterostructure interface quality.

In this study, we have developed a “direct writing” technique to fabricate arrays of WS₂/MoS₂ heterostructures at predefined locations on silicon substrates using precursor solutions. This technique involves two main steps: (a) the writing of ammonium tetrathiomolybdate ((NH₄)₂MoS₄) precursor inks with cantilever tips (followed by thermal annealing to form MoS₂ layer), and (b) the subsequent writing of ammonium tetrathiotungstate ((NH₄)₂WS₄) precursor (followed by thermal annealing to form WS₂ layer). The technique is simple, flexible and potentially extendable to variety of substrates and heterostructure systems. Here we demonstrate a sub-micrometer structure formation in a simple two-step table-top process. The reduction in the number of steps in the process of patterning and growth of heterostructure systems carries a significant benefit; it introduces an alternative unconventional approach to the fabrication of sophisticated engineered materials. In contrast to the commonly used conventional lithographic methods, here neither patterned masks or MX₂ layer transferring were required in our method, hence contamination between two materials can also be significantly reduced. Our technique demonstrates an attractive technology with unique capabilities for precisely aligning and growing dissimilar MX₂ layers in both vertical and lateral directions. This technique can be further extended to a large area fabrication of high quality MX₂ heterostructures.

Results and discussion

Fig. 1 depicts a schematic overview of the production protocol for creating arrays of WS₂/MoS₂ heterostructures. Fig. 1a shows (NH₄)₂MoS₄ precursor ink deposited first as an array of ribbons/lines at specific location on SiO₂/Si substrate along y-axis. Subsequently, samples with patterned (NH₄)₂MoS₄ precursor structures are processed in the CVD chamber at the required temperature (∼450 °C) to form an array of MoS₂ ribbons. Next, a different ink precursor (NH₄)₂WS₄ is patterned atop of MoS₂ ribbons (along x-axis) as shown schematically in Fig. 1b. Once again, samples with (NH₄)₂WS₄ patterned precursor structures are subsequently treated in the CVD chamber to complete the formation of the resulting WS₂/MoS₂ vertical heterostructures. This direct writing technique also provides a unique way to prepare WS₂/MoS₂ lateral heterostructures where the arrays of (NH₄)₂WS₄ ink precursor can be easily written in the same direction to align with pre-existing MoS₂ structures (Fig. 1c). Moreover, writing (NH₄)₂MoS₄ precursor multiple times as shown in Fig. 1d, will produce MoS₂/WS₂/MoS₂ lateral tri-layer heterostructures. These architectures would be challenging to fabricate with conventional lithography and would require multiple additional steps in the fabrication protocol. In contrast, such heterostructures are relatively simple to produce with the use of direct-write scanning probe based nanolithography approach, which effectively offers a nanoscale precision in position and registry with simple mask-free geometry defining capabilities.

Fig. 1 Schematic representation of the production process that shows variety of geometric designs of WS₂/MoS₂ heterostructures. (a) (NH₄)₂MoS₄ precursor ink was deposited in the form of ribbon arrays on SiO₂/Si substrate and then patterns are subsequently processed in the CVD furnace to crystallize into MoS₂; (b) (NH₄)₂WS₄ precursor ink was deposited atop of the MoS₂ ribbons to obtain vertical bilayer heterostructures; (c) and (d) show lateral heterostructures; (c) (NH₄)₂WS₄ precursor ink is shown to be deposited adjacent to the MoS₂ to obtain WS₂/MoS₂ lateral bilayer heterostructures; (d) example of MoS₂/WS₂/MoS₂ lateral tri-layer heterostructures assembly is shown.

The inks developed in this study ((NH₄)₂MoS₄ and (NH₄)₂WS₄ precursors) are water based. Since water is a neutral solvent, it is therefore reasonable to conclude that no chemical reactions occur at room temperature under ambient conditions during the entire writing process. The direct writing technique essentially has two main steps for precursor deposition: “inking” and “writing” as shown in Fig. S1.† In the inking step, the tips of an atomic force microscope (AFM) cantilever are dipped into an “ink”, and then, the ink is transferred onto the selected substrate during the writing step. Either single or multi-pen cantilevers can be employed. The non-interacting chemical nature of the developed inks allows us to simplify and understand the kinetic model of the writing steps as follows: (a) the steps of the ink transferring from tip to substrate with the assistance of water meniscus, (b) the ink lateral diffusion on substrate and (c) the stopping of diffusion by surface tension of the substrate. We can characterize our direct writing method as scanning probe based nanolithography, a cousin technique to dip pen nanolithography and other related methods.^18–23 In order to obtain large area patterned arrays with high throughput, as required by the semiconductor industry, multi-pen cantilevers had been utilized here for parallel writing. Piezo driven scanners with mounted multi-pen cantilever chip can repeat multiple automated sequences for the production of large area patterns. The use of commercially available custom inkwells with multiple channels enables a convenient solution for improving the patterning efficiency (Fig. S1†).

This mask free approach can potentially reduce the amount of residue between the layers of MoS₂ and WS₂ since no polymer resists were required in the process of writing. An additional advantage of this approach is that there is no need for MX₂ materials transfer from growth surfaces to the desirable substrates which has been commonly used and reported in the MX₂ heterostructure preparations.^12–14 The solvent chosen for the precursor inks was water, which is normally removed at low annealing temperatures (∼200 °C), prior to the formation of MX₂ materials (see the discussion of MoS₂ and WS₂ formation below). Heat-treatment of patterned (NH₄)₂MoS₄ and (NH₄)₂WS₄ structures with the presence of hydrogen gas (H₂) in the CVD furnace has shown to lower the required temperature^24,25 at which MoS₂ and WS₂ crystalline structures are formed (∼800 °C to ∼450 °C) as described in the eqn (1) and (2) in the Experimental section.

It is possible to generate more complex structures with the combination of two basic patterns, i.e. dots and ribbons/lines using the software protocol sequencing. Arrays of dots were produced by holding the inked cantilever in contact with the substrate so that inks diffuse out in a radial direction to form a circular dot pattern. Then the tip was moved to the next position and the process was repeated as shown in Fig. S2.† By controlling the humidity in the working chamber (environmental cell), the tip moving speed and the dwell time, various shapes of MoS₂ and WS₂ structures (dot arrays and L-shaped ribbons) were fabricated on SiO₂/Si substrates (Fig. S2†). Other substrates (not shown in this report) have also been successfully tested.

The fabrication of MoS₂ and WS₂ ribbons/lines is more challenging as compared to the dot patterns where the AFM tip must continuously move along the sample as shown schematically on the Fig. 1a. Therefore, a dynamic balance needs to be maintained between the ink transfer (from cantilever to substrate), the ink lateral diffusion and stopping of the ink diffusion process in order to modulate the formation of ribbons/lines on a substrate. Although the “writing” of desired ribbons/lines might also be adjusted by modulating the environmental humidity and temperature, all studies of the patterns in this work were performed at a fixed humidity (50%) and temperature (23 °C) for the consistency of the analyses. The writing of MoS₂ and WS₂ ribbons was studied on SiO₂/Si substrates. Surface of SiO₂/Si substrates is easy to modify and ensures that the key parameters of the ribbons, i.e. width and thickness, could also be precisely controlled and optimized. Once parameters were optimized for individual ribbons, more complex systems such as WS₂/MoS₂ bilayers and MoS₂/WS₂/MoS₂ tri-layer heterostructures were created.

It was established that single and multiple MX₂ ribbons with controlled parameters can be fabricated in this direct writing fashion. Multiple parallel ribbons with selected widths and thicknesses were prepared by employing multi-pen cantilever as shown in the optical image (Fig. 2a). In the case where one specific cantilever tip was inked, individual ribbons with fixed width and thickness were written directly at pre-selected locations as shown in Fig. 2b. Due to the fact that the processes of ink transfer (from cantilever to substrate) and diffusing (lateral diffusion on substrate) occur simultaneously when tip is in contact with the substrate, it was found that control of the tip movement speed was the most optimal way to determine the width of the resulting MoS₂ and WS₂ ribbons. With relatively faster tip speed (in the range of 1–5 μm s⁻¹), the process of lateral diffusion of ink on the substrate can be controlled hence narrow ribbons can be obtained. For example, the width of WS₂ ribbons were decreased from 3 μm to 1.2 μm when tip moving speed increased from 2 μm s⁻¹ to 5 μm s⁻¹, as shown in the Fig. 2c. However, it should be noted that at significantly slower tip moving speeds such as 0.1 μm s⁻¹ (the slowest speed achievable with this tool), the ink meniscus will have more time to diffuse laterally and also produce thicker vertical deposits on the substrate. Table S1† in the ESI shows the relationship between the tip speed and the width of the MoS₂ ribbons.

Fig. 2 Selectively fabricated MoS₂ and WS₂ ribbons on SiO₂/Si substrate. (a) The optical view of the array of (NH₄)₂MoS₄ precursor ribbons with selected width/thickness (∼10 μm/∼160 nm), prepared with multi-pen cantilevers; (b) the optical view of a single ribbon (NH₄)₂WS₄ written on pre-selected location crossing three parallel ribbons; (c) the AFM image of WS₂ ribbons written by controlling the tip moving speed (left: 5 μm s⁻¹, right: 2 μm s⁻¹), an overplayed line profile is also included and shows proportionally increased width of the ribbons with decreased speed of writing; (d) the AFM image of the WS₂ ribbon were obtained with the controlled ink concentration, overplayed line profile shows thickness of the ribbon to be 1.6 nm, which is approximately equivalent to double layer WS₂ structure. Average width of the ribbon in (d) is approximately 700 nm.

Higher ink concentration in this study implies the amount of reactant ammonium tetrathiomolybdate ((NH₄)₂MoS₄) or ammonium tetrathiotungstate ((NH₄)₂WS₄), used for “writing” on the substrate was increased. In order to prepare thinner/thicker structures the amount of reactant was correspondingly decreased/increased by adjusting as-prepared ink concentration (see more details in Experimental section and ESI†). Therefore, to optimize the process it was concluded that the most convenient approach to determine the thicknesses of the resulting MoS₂ and WS₂ ribbons is to adjust precursor ink composition or concentration and perform the writing at tip speeds in the range between 2 μm s⁻¹ to 5 μm s⁻¹. The influence of ink concentration is clearly observed in the AFM images (Fig. 2d), where double layers WS₂ ribbon (∼1.6 nm thick and ∼0.7 μm wide) was produced by controlling the ink concentration. Table S2† in the ESI shows the relationship between the ink concentrations and the thicknesses of the MoS₂ ribbons.

This direct writing approach provides additional flexibility for precise patterning of MoS₂ and WS₂ ribbons aligned to prefabricated structures on the substrate. Fig. 3a shows AFM images of the WS₂ ribbon structures formed between prefabricated Al/Ti electrodes on SiO₂/Si substrate. The connection between Al/Ti electrodes and WS₂ ribbons is continuous and robust and can be observed in the enlarged AFM image (Fig. 3b). An AFM image of the surface of WS₂ ribbon is shown in Fig. 3c and the height profile measurement indicates an approximately 1 nm average surface roughness (inset of Fig. 3c). The three dimensional AFM representation of the same area is shown in Fig. 3d. The robust connection between WS₂ and Al/Ti electrode has been further confirmed with the transport measurement at room temperature in vacuum at approximate pressure of 10⁻⁴ Torr. Linear current–voltage behaviour was recorded, as shown in Fig. 3e. The resistance obtained here is comparable to the devices reported elsewhere, where electrodes are usually deposited on top of the grown WS₂ structures.^16,26 These results demonstrate that with the proposed bottom-up fabrication method robust connection between MX₂ ribbons and prefabricated electrodes is achievable. Moreover, the surface quality of MX₂ ribbons is uniform and the surface roughness of ∼1 nm is better than typical roughness for the materials derived from the solution based methods. The high-resolution transmission electron microscopy (HRTEM) image in Fig. 3f clearly reveals the periodic atom arrangement of the MoS₂ film at a selected location. HRTEM characterization also demonstrates that heat-treatment of patterned (NH₄)₂MoS₄ structures with the presence of argon and hydrogen gas atmosphere would indeed lower the required MoS₂ crystallization temperature (see ESI† for more details on TEM and X-ray diffraction (XRD) data).

Fig. 3 Direct write of WS₂ ribbon at predefined locations of the selected substrate. (a) AFM images of the WS₂ ribbon structures formed between the prefabricated Al/Ti electrodes on SiO₂/Si substrate; (b) enlargement of AFM images of connection between Al/Ti electrodes and WS₂ ribbon; (c) enlargement of AFM images and the height profile measurement (inset); (d) AFM 3D image of the area shown in (a); (e) transport measurement of WS₂ ribbon fabricated by direct write technique atop of metal electrodes (inset: AFM images of the measured device; a horizontal ribbon of WS₂ between two electrode is shown, the scale bar is 5 μm). (f) High-resolution TEM image for the MoS₂ polycrystalline film (see ESI† for more details).

The success in controlled production of MX₂ ribbons with specific thickness and width further enables the fabrication of more complex MX₂ structures directly on existing devices. The arrays of WS₂/MoS₂ heterostructures (in vertical and lateral geometries) have been written at predefined locations between pre-deposited electrodes. As shown in the Fig. 4a and b, nine vertical heterostructure regions of WS₂/MoS₂ were prepared by cross-patterning of the arrays of MoS₂ ribbons (x-axis direction) and the array of WS₂ ribbons (y-axis direction) in the 3 × 3 arrangement. Representative AFM image (3D view) of the cross-pattered WS₂/MoS₂ vertical heterostructures is shown in Fig. 4c, where spacing between heterostructures regions is approximately ∼20 μm (see more details in ESI in Fig. S3†). In addition, the lateral heterostructures of WS₂/MoS₂ were also prepared by connecting patterns of MoS₂ ribbons (x-axis direction) with WS₂ ribbons (y-axis direction), in 3 × 3 L-shaped junction arrangement, as shown in the Fig. 4d and e. Furthermore, we have shown that our approach has proven to be a versatile and enables the fabrication of more complicated MX₂ tri-layer heterostructures, i.e., MoS₂/WS₂/MoS₂ alternating structures directly patterned on prefabricated devices. See ESI† for representative examples of vertical tri-layer structure of MoS₂/WS₂/MoS₂ shown in Fig. S4a and b,† and the lateral tri-layer structure of MoS₂/WS₂/MoS₂ shown in Fig. S4c and d.†

Fig. 4 Direct write of various WS₂/MoS₂ heterostructures on pre-fabricated devices. (a) The production scheme for the arrays of WS₂/MoS₂ bilayer vertical heterostructures; (b) optical image of WS₂/MoS₂ bilayer vertical heterostructures; (c) representative AFM 3D image of patterned cross-bar WS₂/MoS₂ vertical heterostructures (see more details in ESI in Fig. S3†); (d) the production scheme for the arrays of WS₂/MoS₂ bilayer lateral heterostructures; (e) optical image of WS₂/MoS₂ bilayer lateral heterostructures; (f) representative resonant Raman spectroscopy of the fabricated ribbons of MoS₂, WS₂ and WS₂/MoS₂ vertical heterostructures.

Resonant Raman spectroscopy was also utilized to characterize the fabricated ribbon arrays of MoS₂, WS₂ and WS₂/MoS₂ vertical heterostructures (Fig. 4f). Raman spectra acquired from the pure MoS₂ region which exhibited two peaks located at 382 cm⁻¹ and 407 cm⁻¹, corresponding to E¹_2g and A_1g modes. Raman spectra from the pure WS₂ region show two peaks at 352 cm⁻¹ and 419 cm⁻¹, also typical signals of E¹_2g and A_1g modes. The spectra collected at these regions show the characteristic peaks A_1g and E¹_2g with a wave-number difference of Δ ∼ 25 cm⁻¹ and Δ ∼ 67 cm⁻¹ which corresponds to the multilayer values for MoS₂ and WS₂ as reported previously.¹¹ In contrast, the Raman spectra collected at the vertical heterostructure region show four discrete peaks, which match well to the above E¹_2g and A_1g modes of multilayer MoS₂ and WS₂. The relative Raman intensities of E¹_2g and A_1g modes of MoS₂ are maintained without considerable changes before and after the formation of WS₂, demonstrating the feasibility for the application of direct write lithography for the production of complex MX₂ heterostructures. It was suggested that Raman spectral signatures from vertical heterostructures should not be different from the Raman plots of the lateral ones, although if a strain/stress is present at the lateral heterostructure interface, it may potentially alter Raman spectral characteristics.¹⁵ One can further improve on the quality (crystallinity) of such polycrystalline materials by increasing the base temperature of the thermal treatment. This is a common approach that could be applied to most polycrystalline systems. The relatively low temperatures were chosen in this study to allow for testing the concept of direct fabrication of precursor inks on metallic devices and even flexible substrates (not shown in this report). We note that high temperature thermal treatment of MX₂ thin films may induce the sulfur deficiently, which might subsequently degrade their quality. Liu et al.²⁷ suggested that introducing sulfur vapor during the thermal treatment (∼1000 °C) would significantly improve MX₂ material quality.

In conclusion, we report on a unique approach which allows for an easy formation of high quality WS₂/MoS₂ heterostructures with controllable lengths, widths and thicknesses. An important aspect of this study was the demonstration of the selective growth of heterostructures and the formation of various architectures on the substrate, which is presently not easily achievable by other fabrication methods that are commonly used. In this report we have demonstrated several heterostructure geometries and hybrid device architectures. Our approach can potentially be extrapolated to produce an entire new range of heterostructured systems. This versatile and scalable mask free technology combines the simplicity of solution based processing and the precision of scanning probe based nanolithography which makes it a promising tool for nanoscale materials engineering future applications.

Experimental

Preparation of (NH₄)₂MoS₄ precursor

Ammonium thiomolybdates ((NH₄)₂MoS₄) powder (Alfa Aesar, purity of 99.95%; 0.28 g) was added into 60 mL of deionized (D.I.) water. The obtained precursors were sonicated for 30 min then filtered with 0.2 μm PTFE membranes to get highly dispersed clear solutions. The obtained (NH₄)₂MoS₄ solution was used as ink precursor for the formation of MoS₂ structures.

Preparation of (NH₄)₂WS₄ precursor

Ammonium tetrathiotungstate ((NH₄)₂WS₄) powder (Sigma-Aldrich, purity of 99.9%; 0.026 g) was added into 60 mL of D.I. water. The obtained precursors were sonicated for 30 min then filtered with 0.2 μm PTFE membranes to get highly dispersed clear solutions. The obtained (NH₄)₂WS₄ solution was used as ink for the formation of WS₂ structures. The diluted inks with a volume ratio of 1 : 9 (as-prepared ink : deionized water) was used to produce double layer WS₂ ribbon.

Direct write patterning process

The patterning of (NH₄)₂MoS₄ and ((NH₄)₂WS₄) precursor inks were performed with a custom made patterning platform with motorized piezo-stages with a resolution of approximately 20 nm for all three xyz axes. The sample holding stage is also equipped with tilt correction capabilities. The writing tool is essentially a cantilever array of twelve tips with approximately 60 μm inter-tip spacing mounted on the tip holder. The tips and the inkwells with matching pitch were purchased from Advanced Creative Solutions Technology LLC. Si substrates (with 300 nm SiO₂ layer) were cleaned by 10 min sonication of acetone, IPA and D.I. water respectively. Cantilevers and substrates were additionally cleaned in ozone cleaner to render them hydrophilic. Alphabetical markers on Si/SiO₂ substrates were deposited via standard e-beam lithography and lift-off process (Pt/Ti metal with thickness of 70 nm/5 nm respectively was used for metallization). Samples with the metalized alphabetical markers were additionally cleaned in the furnace with the mixture of Ar/H₂ at 450 °C to remove any possible polymer residuals.

Preparation of MX₂ heterostructures

Ribbons of (NH₄)₂MoS₄ precursor ink were patterned on cleaned SiO₂/Si substrate and then transferred into CVD system for crystallization of MoS₂. The formation of MoS₂ was performed in ambient condition, annealed in a mixture of Ar/H₂ with the respective flow rates of 400 sccm/100 sccm. The annealing at lower temperatures (∼200 °C) can efficiently remove the residual D.I. water. The subsequently higher temperature annealing (∼450 °C) orders crystallinity of MoS₂. Ribbons of (NH₄)₂WS₄ precursor ink were patterned atop of MoS₂ ribbons (to form vertical bilayer heterostructures) and adjacent to MoS₂ ribbons (to form lateral bilayer heterostructures). The formation of WS₂/MoS₂ heterostructures was also performed in ambient condition and annealed in a mixture of Ar/H₂ with the respective flow rates of 400 sccm/100 sccm. A slightly lower temperature of 400 °C was used for crystallization of WS₂ layer. Same protocol was used for the multi-layered patterned structures. Eqn (1) and (2) show the resections in presence of H₂ gas.

(NH₄)₂MoS₄ + H₂ → 2NH₃ + 2H₂S + MoS₂ (1)

(NH₄)₂WS₄ + H₂ → 2NH₃ + 2H₂S + WS₂ (2)

Device characterization

AFM topographic images were acquired in non-contact mode with a Park NX10 system. Raman spectroscopy was obtained with a Renishaw InVia Raman Spectrometer with the laser excitation wavelength of 532 nm. The Si peak at 520 cm⁻¹ was used as reference for wavenumber calibration in all Raman spectral data. A JEOL-2100F system working at 200 kV was employed for the HRTEM microstructure characterization.

Acknowledgements

Use of the Center for Nanoscale Materials was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02 06CH11357. I. K. acknowledges support of NSF MRI program (Award No. 1338021), and the Saint Louis University seed funds.

Notes and references

1. M. Chhowalla, H. S. Shin, G. Eda, L.-J. Li, K. P. Loh and H. Zhang, Nat. Chem., 2013, 5, 263–275.
2. S. Das, R. Gulotty, A. V. Sumant and A. Roelofs, Nano Lett., 2014, 14, 2861–2866.
3. G. Fiori, F. Bonaccorso, G. Iannaccone, T. Palacios, D. Neumaier, A. Seabaugh, S. K. Banerjee and L. Colombo, Nat. Nanotechnol., 2014, 9, 768–779.
4. F. Schwierz, J. Pezoldt and R. Granzner, Nanoscale, 2015, 7, 8261–8283.
5. F. H. L. Koppens, T. Mueller, P. Avouris, A. C. Ferrari, M. S. Vitiello and M. Polini, Nat. Nanotechnol., 2014, 9, 780–793.
6. R. Dong, Z. Guo, J. Palmer, Y. Hu, M. Ruan, J. Hankinson, J. Kunc, S. K. Bhattacharya, C. Berger and W. A. d. Heer, J. Phys. D: Appl. Phys., 2014, 47, 094001.
7. A. C. Ferrari, F. Bonaccorso, V. Fal'ko, K. S. Novoselov, S. Roche, P. Boggild, S. Borini, F. H. L. Koppens, V. Palermo, N. Pugno, J. A. Garrido, R. Sordan, A. Bianco, L. Ballerini, M. Prato, E. Lidorikis, J. Kivioja, C. Marinelli, T. Ryhanen, A. Morpurgo, J. N. Coleman, V. Nicolosi, L. Colombo, A. Fert, M. Garcia-Hernandez, A. Bachtold, G. F. Schneider, F. Guinea, C. Dekker, M. Barbone, Z. Sun, C. Galiotis, A. N. Grigorenko, G. Konstantatos, A. Kis, M. Katsnelson, L. Vandersypen, A. Loiseau, V. Morandi, D. Neumaier, E. Treossi, V. Pellegrini, M. Polini, A. Tredicucci, G. M. Williams, B. Hee Hong, J.-H. Ahn, J. Min Kim, H. Zirath, B. J. van Wees, H. van der Zant, L. Occhipinti, A. Di Matteo, I. A. Kinloch, T. Seyller, E. Quesnel, X. Feng, K. Teo, N. Rupesinghe, P. Hakonen, S. R. T. Neil, Q. Tannock, T. Lofwander and J. Kinaret, Nanoscale, 2015, 7, 4598–4810.
8. B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti and A. Kis, Nat. Nanotechnol., 2011, 6, 147–150.
9. Y. Yu, S. Hu, L. Su, L. Huang, Y. Liu, Z. Jin, A. A. Purezky, D. B. Geohegan, K. W. Kim, Y. Zhang and L. Cao, Nano Lett., 2015, 15, 486–491.
10. Y. Jung, J. Shen, Y. Sun and J. J. Cha, ACS Nano, 2014, 8, 9550–9557.
11. S. Tongay, W. Fan, J. Kang, J. Park, U. Koldemir, J. Suh, D. S. Narang, K. Liu, J. Ji, J. Li, R. Sinclair and J. Wu, Nano Lett., 2014, 14, 3185–3190.
12. N. Huo, J. Kang, Z. Wei, S.-S. Li, J. Li and S.-H. Wei, Adv. Funct. Mater., 2014, 24, 7025–7031.
13. N. Huo, Z. Wei, X. Meng, J. Kang, F. Wu, S.-S. Li, S.-H. Wei and J. Li, J. Mater. Chem. C, 2015, 3, 5467–5473.
14. G. R. Bhimanapati, Z. Lin, V. Meunier, Y. Jung, J. Cha, S. Das, D. Xiao, Y. Son, M. S. Strano, V. R. Cooper, L. Liang, S. G. Louie, E. Ringe, W. Zhou, S. S. Kim, R. R. Naik, B. G. Sumpter, H. Terrones, F. Xia, Y. Wang, J. Zhu, D. Akinwande, N. Alem, J. A. Schuller, R. E. Schaak, M. Terrones and J. A. Robinson, ACS Nano, 2015, 9, 11509–11539.
15. K. Chen, X. Wan, J. Wen, W. Xie, Z. Kang, X. Zeng, H. Chen and J.-B. Xu, ACS Nano, 2015, 9, 9868–9876.
16. Q. Zhang, X. Xiao, R. Zhao, D. Lv, G. Xu, Z. Lu, L. Sun, S. Lin, X. Gao, J. Zhou, C. Jin, F. Ding and L. Jiao, Angew. Chem., Int. Ed., 2015, 54, 8957–8960.
17. Y. Gong, J. Lin, X. Wang, G. Shi, S. Lei, Z. Lin, X. Zou, G. Ye, R. Vajtai, B. I. Yakobson, H. Terrones, M. Terrones, K. T. Beng, J. Lou, S. T. Pantelides, Z. Liu, W. Zhou and P. M. Ajayan, Nat. Mater., 2014, 13, 1135–1142.
18. I. Kuljanishvili, D. A. Dikin, S. Rozhok, S. Mayle and V. Chandrasekhar, Small, 2009, 5, 2523–2527.
19. R. D. Piner, J. Zhu, F. Xu, S. Hong and C. A. Mirkin, Science, 1999, 283, 661–663.
20. B. Wu, A. Ho, N. Moldovan and H. D. Espinosa, Langmuir, 2007, 23, 9120–9123.
21. S. Rozhok, R. Piner and C. A. Mirkin, J. Phys. Chem. B, 2003, 107, 751–757.
22. J. Li, C. Lu, B. Maynor, S. Huang and J. Liu, Chem. Mater., 2004, 16, 1633–1636.
23. B. W. Maynor, Y. Li and J. Liu, Langmuir, 2001, 17, 2575–2578.
24. G. Alonso, M. Del Valle, J. Cruz, A. Licea-Claverie, V. Petranovskii and S. Fuentes, Catal. Lett., 1998, 52, 55–61.
25. J. L. Brito, M. Ilija and P. Hernández, Thermochim. Acta, 1995, 256, 325–338.
26. N. Huo, S. Yang, Z. Wei, S.-S. Li, J.-B. Xia and J. Li, Sci. Rep., 2014, 4, 5209.
27. K.-K. Liu, W. Zhang, Y.-H. Lee, Y.-C. Lin, M.-T. Chang, C.-Y. Su, C.-S. Chang, H. Li, Y. Shi, H. Zhang, C.-S. Lai and L.-J. Li, Nano Lett., 2012, 12, 1538–1544.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra12576j

---