Tailoring the electrical and photo-electrical properties of a WS₂ field effect transistor by selective n-type chemical doping

Abstract

Here, we demonstrate a doping technique which remarkably improves the electrical and photoelectric characteristics of a WS₂ field effect transistor (FET) by chemical doping. The shift of the threshold voltage towards a negative gate voltage and the red shift of the E¹_2g and A_1g peaks in the Raman spectra confirm the n-type doping effect in WS₂ FETs. WS₂ films show an unprecedented high mobility of 255 cm² V⁻¹ s⁻¹ at room temperature. The on/off ratio of the output current is ∼10⁸ at room temperature. The mobility of a multilayer ML-WS₂ FET was found to be 425 cm² V⁻¹ s⁻¹ at 5 K. Semiconductor-to-metal transitions were also observed at V_bg > 30 V. A decrease in contact and sheet resistance was observed after potassium iodide (KI) doping. The photocurrent in WS₂ FETs was also enhanced after n-type doping. Chemical doping exhibited a very stable, effective, and easy-to-apply method to enhance the performance of a WS₂ FET.

---

1. Introduction

Due to unique physical and electrical properties,^1,2 considerable interest has developed for two-dimensional (2D) atomic layered materials.^3,4 Graphene, an exemplary 2D carbon system with fascinating properties, which makes it a promising material in a variety of future nanoelectronics applications such as field effect transistors (FETs),⁵ ultracapacitors, solar cells and especially as gas and chemical sensors.⁶ Regardless of all its advantages in nanoelectronics device applications, graphene cannot be used as an active channel material in FETs due to the absence of a bandgap.⁷ Considerable research efforts have been conducted to attain a significant bandgap in graphene like patterning into nanoribbons,⁷ chemical functionalization,⁸ and dual-gated bilayer graphene.⁹ But it always shows bandgap versus significant mobility degradation with low on/off switching ratio for FETs.¹⁰ Recently, the absence of a bandgap in graphene has encouraged researchers to search for other 2D layered materials.

2D tungsten disulphide (WS₂) is a tremendous material provoking as strong competitors for next-generation nanoelectronics and optoelectronics devices. WS₂ is of atomic-layer-thick that does have a bandgap.¹¹ It has specific interest due to transition of indirect-to-direct bandgap nature by changing number of layers and observation of photoluminescence (PL) in the visible range at room temperature.^12,13 Bulk WS₂ is an n-type semiconductor with an indirect-bandgap (1.4 eV), but turn into a direct-bandgap (2.1 eV) material when cleaved as monolayer.¹⁴ WS₂ is 2D layered semiconductor material; it has been widely studied due to its attractive electrical and optical properties in a variety of applications.¹⁵ Especially in the field of low power FET,¹⁶ photodectors,¹⁷ gas and chemical sensors,¹⁸ memory and electroluminescent devices, and integrated circuits.^11,19

Like other low-dimensional semiconductors transport properties of WS₂ are strongly degraded by the extrinsic/environmental effects, which limit their intrinsic properties and overall permanence.^20,21 It has been already exposed that the electrical and photoelectrical characteristics of a mechanically exfoliated WS₂ layer could be affected by the adsorbents in ambient air.²² The improvement in sample quality is essential for the upcoming development of WS₂ device technology.¹¹ Recently, different strategies were used to improve the device performance by limiting the Schottky barrier height (SBH) by using different contact electrode,^23,24 and with different substrate to enhance charge carrier mobility.²⁵ Different doping techniques are used to control the carrier concentrations in 2D semiconductor materials.²⁶ The controlled doping of WS₂ layers is important for controlling the carrier concentrations, which has fundamental significance for FET devices. Such efforts are extremely crucial for WS₂-based FETs and related electronic device design, to curtailed environment effect in order to improve the device performance. Chemical doping is the most effective and easy method for tuning the electrical and photoelectrical properties of semiconducting materials.²⁷ Chemical doping is necessary to tune the free charge concentration and SBH in WS₂ devices.

Here, we demonstrated airstable surface charge transfer n-type doping of exfoliated single-, multi-layer WS₂ (SL-, ML-WS₂) FET by using potassium iodide (KI) as surface charge transfer donor. Shift of threshold voltage toward negative gate voltage showed n-type doping effect in WS₂ FETs. Enhanced charge carrier density, electron field effect mobility and photocurrent were observed after KI doping in WS₂ FET. WS₂ FET shows decrease in contact and sheet resistance after doping. Semiconductor-to-metal transition was also observed at low temperature. The device performing of SL-, ML-WS₂ FET was examined by electrical-transport measurements and Raman spectroscopy before and after KI doping.

2. Experimental section

2.1. Sample preparation

The exfoliated SL-, ML-WS₂ films were mechanical exfoliated from naturally occurring bulk crystals of tungsten disulphide using standard scotch tape peel off method which were consequently transfer on Si/SiO₂ substrate. SL-, ML-WS₂ films were characterized with optical microscopy, Raman spectroscopy, and confirmed by AFM. The Raman microspectrometer with laser wavelength 514 nm was used, and the power was kept below 1.0 mW to avoid laser-induced heating. The laser spot size of Raman spectroscopy was 0.7 μm.

2.2. Device fabrication and characterization

SL-, ML-WS₂ devices were fabricated by using photolithography, e-beam lithography, and O₂ plasma etching. Large electrode patterns with a Cr/Au (6/30 nm) film were deposited using a thermal evaporator after standard photolithography. E-beam lithography was then applied to pattern the source and drain electrodes, where the film was deposited through the evaporation of Cr/Au (10/80 nm). After device fabrication, all devices were annealed in a tube furnace at a temperature of 300 °C, in a flow of Ar/H₂ gas for 2 hours. Electrical transport measurements were performed using a Keithley 2400 source meter and Keithley 6485 picometer at room temperature in a vacuum. KI power was mixed in deionized water to make KI solution with 0.01 M concentration. WS₂ FETs device was soaked in KI solution at room temperature for certain period of time. Transport characteristics and Raman spectroscopy were also performed for WS₂ FETs before and after KI doping. Electrical transport properties were also measured at low temperature. For photocurrent, WS₂ samples were exposure to deep ultraviolet (DUV) with a dominant wavelength of λ = 220 nm and an average intensity of 11 mW cm⁻² before and after KI doping under vacuum at room temperature.

3. Results and discussion

3.1. Characterization of WS₂ layers by scanning electron and atomic force microscopy

Fig. 1a shows the schematic of a WS₂ FET device. Fig. 1b shows the scanning electron microscopic (SEM) image of mechanically exfoliated SL-, ML-WS₂ films with Cr/Au electrical contacts. Optical images of SL-, ML-WS₂ films with Cr/Au electrical contacts are presented in Fig. S1.† The optical image of WS₂ flake on the Si/SiO₂ substrate displays different contrasts depending on the thickness of flake (shown in Fig. S1†). The SL-, ML-WS₂ flakes were further confirmed by using atomic force microscopy (AFM) and Raman spectra. Fig. 1c represents the surface morphology of the SL-, ML-WS₂ films obtained by using AFM. The AFM image was obtained in tapping mode under ambient conditions and the scan area was kept at 20 × 20 μm. The thickness profiles represent that surface of WS₂ was uniform with extremely low roughness. The thickness of the SL-WS₂ film was measured as 0.84 nm while the thickness of ML-WS₂ film was measured as 5.21 nm, which corresponds to seven-layer of WS₂ flakes.

Fig. 1 (a) Schematic of a WS₂ FET. (b) Scanning electron microscopic image of mechanically exfoliated SL-, ML-WS₂ films on Si/SiO₂ substrate with source–drain contacts. (c) Atomic force microscopy (AFM) image of SL-, ML-WS₂ flake. (d) Height profile of the SL-, ML-WS₂ along the green line in AFM image. The 0.84 nm thickness indicates one layer of WS₂ film. The 5.21 nm thickness indicates seven layers of WS₂ film.

3.2. Electrical transport measurement in the pristine state

To check the electrical characteristics of the device, electrical transport measurement was investigated at room temperature under vacuum. Fig. 2a shows the transfer characteristics (I_ds–V_bg) of the SL-WS₂ FET at a fixed source–drain voltage, V_ds = 0.5 V. The black curve in the graph is plotted in the logarithmic scale for the I_ds–V_bg curve. Output current on/off ratio of the SL-WS₂ FET was ≈10⁶, and threshold voltage (V_th) approximately was at 28 V.^28,29 The field-effect mobility (μ) of pristine SL-WS₂ FET is 5.8 cm² V⁻¹ s⁻¹. Fig. 2b shows output characteristic (I_ds–V_ds) curves at various gate voltages ranging from 10 V to 60 V in steps of 10 V for the SL-WS₂ FET. The nonlinear behavior in output (I_ds–V_ds) characteristics represents the existence of Schottky barriers between Cr/Au metal contact and WS₂ at the source and drain parts. Similarly, Fig. 2c represents the transfer characteristics (I_ds–V_bg) for the ML-WS₂ FET. The output current on/off ratio of the ML-WS₂ FET was ≈10⁷. Hysteresis in threshold voltage of WS₂ FETs was measured around 15–18 V in the pristine samples, less hysteresis in WS₂ FETs was measured after KI doping which indicates less charge impurities scattering after the KI doping results are shown Fig. S2.† We have measure three different concentrations (0.01 M, 0.1 M, and 1.0 M) to optimize the controlled doping effect in ML-WS₂. Doping is well controlled with optimized time for concentration 0.01 M but with higher concentration like 1.0 M solution. It reaches at saturation point even after 1 min soaking as shown in Fig. S3;† further soaking does not change its characteristics. So different concentration give the similar doping effect but it is time depended. With lower concentration 0.01 M, we have well controlled doping with time. Output (I_ds–V_ds) characteristic curves at various gate voltages ranging from −10 V to +60 V for the ML-WS₂ FET are shown in Fig. 2d.

Fig. 2 (a) Transfer characteristics (I_ds–V_bg) of the pristine SL-WS₂ FET. On/off ratio of the device is ∼10⁶. (b) Output characteristics (I_ds–V_ds) of SL-WS₂ FET at different back gate voltages ranging from 10 V to 60 V in steps of 10 V. (c) Transfer characteristics (I_ds–V_bg) of the pristine ML-WS₂ FET. On/off ratio of the device is ∼10⁷. (d) Output characteristics (I_ds–V_ds) of ML-WS₂ FET at different back gate voltages ranging from −10 V to +60 V in steps of 10 V. All measurements were done in vacuum at T = 300 K.

3.3. Electrical transport measurement after KI doping

After measuring transport properties at pristine state, we introduced a simple tuning technique to enhance the charge carrier density and mobility. WS₂ FETs were treated in KI solution with molar concentration 0.01 M for different periods of time. The effect of KI solution on WS₂ FET indicates n-type doping behaviour. Fig. 3a and b show the respective transfer characteristics (I_ds–V_bg) of SL-, ML-WS₂ FETs after KI doping for different periods. The threshold voltage shifted toward negative gate voltages after KI treatment, which revealed n-type doping for all WS₂ FETs. The output current I_ds of all WS₂ FETs was increased after KI treatment. Fig. 3c represent the change in V_th as a function of treatment time on SL-, ML-WS₂ FETs. After the 30 min KI treatment, V_th reached −71, and −88 V for SL-, ML-WS₂ FETs respectively. We have also calculated the changes in mobility before and after the chemical doping. Fig. 3d shows the field-effect mobilities of SL-, ML-WS₂ FETs at room temperature as a function of reaction time.

Fig. 3 Doping effect on WS₂ layers by KI treatments. (a) Transfer characteristics (I_ds–V_bg) of SL-WS₂ FET for different times of treatment. (b) Transfer characteristics (I_ds–V_bg) of ML-WS₂ FET for different treatment times. (c) Shift of threshold voltage as a function of reaction time by KI for SL-, ML-WS₂ FET. (d) Electron mobility as a function of reaction time for SL-, ML-WS₂ FET. All measurements were performed in vacuum at T = 300 K.

Field-effect mobility was calculated by using the equation ^30–32 where L is the channel length, W is the channel width, is the slope of transfer characteristic of the device with V_ds = 0.5 V, and C_g is the gate capacitance of ∼115 aF μm⁻² for the Si/SiO₂ substrate.³³ The value of is obtained by fitting the linear regime of the transfer characteristic curves of WS₂ FETs. After 30 min KI treatment, the field-effect mobilities were enhanced to 30 cm² V⁻¹ s⁻¹ and 255 cm² V⁻¹ s⁻¹ with on/off ratio ≈10⁷ and ≈10⁸ (shown in Fig. S4†) for SL-, ML-WS₂ FETs respectively. While it was measured around 6 cm² V⁻¹ s⁻¹ and 73 cm² V⁻¹ s⁻¹ for SL-, ML-WS₂ FETs in pristine devices respectively.

The n-type doping effect for WS₂ FETs by KI treatment can be described as: tungsten (W) has valence electronic configuration 6s² 5d⁴ retain an electropositive behavior, presenting capacity to accept electrons. Furthermore, the KI could be dissociated in electropositive K⁺ (electron acceptor) and electronegative I⁻ (electron donor) ions. When I⁻ atoms are combined in WS₂, they occupy the location of sulfur vacancies. Separate impurity energy levels in WS₂ would emerge into the conduction band resulting in the bandgap narrowing upon I⁻ ions doping. Thus n-type doping effect in WS₂ can attain by the donation of an extra electron from I⁻ ions when they occupy the location of sulfur vacancies.

3.4. Temperature-dependent electrical transport properties of ML-WS₂ FETs

Temperature-dependent transport properties of ML-WS₂ FETs were studied after doping. Fig. 4a represent the transfer characteristics (I_ds–V_bg) of the ML-WS₂ FET after 30 min chemical doping at fixed V_ds = 0.5 V at different temperatures. ML-WS₂ FET shows a traditional semiconductor behavior as conductance decreasing with temperature decrease at low gate voltage V_bg < 30 V.^34,35 whereas conductance increases as temperature decreased at high gate voltage V_bg > 30 V.³⁶ Semiconductor-to-metal transition at V_bg 30 V suggests that a degenerately doped state is realized in ML-WS₂ film.^37,38 Fig. 4b shows the temperature dependence of I_ds for different values of V_bg. Here, we can clearly understand the critical V_bg of 30 V, at which I_ds remains almost independent of temperature. However, the I_ds of ML-WS₂ FET increases with decreasing temperature for V_bg > 30 V, indicating metallic behavior. For V_bg < 30 V, I_ds decreases with decreasing temperature, indicating a semiconducting behavior. Explanations of a similar semiconductor-to-metal transition were described for other transition metal dichalcogenide (TMDCs) materials.³⁹ We have further investigated the electron filed effect mobility of SL-WS₂ FET and ML-WS₂ on SiO₂ at various temperatures. The temperature dependence of μ of SL-, ML-WS₂ FETs is compared in Fig. 4c. The electron field-effect motilities of SL-, ML-WS₂ FETs on SiO₂ were 30 and 255 cm² V⁻¹ s⁻¹ respectively at T = 300 K and 65 and 425 cm² V⁻¹ s⁻¹ respectively at T = 5 K.

Fig. 4 (a) Transfer characteristics (I_ds–V_bg) of the mechanically exfoliated ML-WS₂ FET on SiO₂ at different temperatures. (b) Output current as function of temperature for different values of the back-gate voltage. (c) Electron field-effect mobility of SL-, ML-WS₂ FETs on SiO₂ at various temperatures.

3.5. Photo-electrical measurement under deep-ultraviolet light

Due to a well-developed bandgap, WS₂ FETs can also be used as optoelectronic devices. The optoelectronic response of SL-, ML-WS₂ FETs in the vacuum by a 220 nm dominant wavelength of deep ultraviolet (DUV) light was measured. A constant bias voltage of 0.5 V was applied between the source and the drain of the WS₂ FET, and I_ds was monitored for optoelectronic response. Photocurrent (I_ph) can be described as difference of I_ds with and without light illumination environment.^40–42

I_ph = I_illuminated − I_dark

Fig. 5a shows the photocurrent for SL-WS₂ FET before and after KI treatment. The photocurrent of SL-WS₂ FET was increased about 3 times after chemical doping. The increase in I_ds current was due to the improvement in electron mobility by the chemical doping as shown in Fig. 3. A large increase in photocurrent was also observed in ML-WS₂ FETs after KI doping as shown in Fig. 5b.

Fig. 5 Photocurrent of the WS₂ FET by 220 nm wavelength dominant DUV illumination before and after KI treatment. A constant bias voltage of 0.5 V was applied between the source and the drain of the WS₂ FET. (a) DUV-driven I_ds of SL-WS₂ FET before (black line) and after (red line) KI treatment. (b) DUV-driven I_ds of ML-WS₂ FET before (black line) and after (red line) KI treatment.

3.6. Contact resistance measurement before and after doping

We have measured the contact resistance of ML-WS₂ by using transfer length method (TLM) before and after doping. The specific resistance (ρ = RW) normalized to the sample width was measured due to dependence of resistance (R) on the sample channel width (W). Fig. 6a shows the specific resistance as a function of distance between the contacts. The specific contact resistance (ρ_c = R_cW) of the ML-WS₂ device after doping was 3.75 kΩ μm, whereas it was 36.15 kΩ μm for undoped devices. Remarkably decrease in contact resistant after doping leads to increase the drain–source current and improves the device performance. Fig. 6b shows the specific contact resistance of the ML-WS₂ film for freshly doped and after 2 weeks to check stability of doping. The specific contact resistance was slightly increased from 3.75 to 4.2 kΩ μm, which confirm that chemical doping is stainable technique.

Fig. 6 Contact resistance measurement by transfer length method. (a) Specific contact resistant (ρ_c = R_cW) for the ML-WS₂ device before and after doping. Contact resistant is 3.75 kΩ μm after doping, whereas it was 36.15 kΩ μm before doping. (b) Specific contact resistant (ρ_c = R_cW) for the ML-WS₂ after two weeks of doping.

3.7. Raman spectra measurement for different number of WS₂ layers

Raman spectroscopy is fast, nondestructive, high resolution technique to study the fundamental properties and structural characterizations of various 2D materials.⁴³ Raman spectroscopy was performed to identify the number of layer and doping effect on SL-WS₂, ML-WS₂ films. Fig. 7a represents Raman spectra taken for pristine SL-, ML-WS₂ films used in our study. Raman spectra of the SL-, ML-WS₂ films showed strong signals of in-plane E¹_2g, out-of-plane A_1g, and vibration second-order 2LA(M) modes.⁴⁴ The first-order E¹_2g and A_1g optical modes was used to explain the properties of 2D material.⁴⁵ The frequency difference between A_1g and E¹_2g modes was used to explain difference between SL-, ML-WS₂.⁴⁴ The frequency difference between A_1g and E¹_2g modes (Δ = A_1g − E¹_2g) is about 62.4 cm⁻¹ for SL-WS₂, and for ML-WS₂ it is Δ = 65.7 cm⁻¹, indicating multiple layers of WS₂ film.⁴⁵ Although the 2LA(M) mode overlapped with the first-order E¹_2g mode, the multi-peak Lorentzian fitting can clarify their individual contributions as seen in Fig. 7b. Chemical doping effect was also examined through Raman spectroscopy. Fig. 7c shows the Raman spectra of the SL-, ML-WS₂ films before and after 30 min doping. Red shift was observed in A_1g and E¹_2g peaks after doping. A_1g peak shifted from 416.3 cm⁻¹ to 412.5 cm⁻¹ for SL-WS₂ and from 420.7 cm⁻¹ to 418.2 cm⁻¹ for ML-WS₂. Red shift in E¹_2g mode was also observed: 1.9 cm⁻¹ for SL-WS₂ and 1.3 cm⁻¹ for ML-WS₂. These observations indicate that the A_1g vibration mode is more sensitive to doping as compared to E¹_2g. The strong coupling of A_1g phonon mode with electron was already reported.³² The A_1g Raman peak softens and broadens with the increase of electron doping, whereas the E¹_2g Raman peak weakly depends on the doping. Furthermore, FWHMs of the A_1g and E¹_2g peaks were increased for all WS₂ films after the 30 min treatment.

Fig. 7 Raman spectra for SL-, ML-WS₂ films using a laser with 514 nm wavelength. (a) Raman spectra for SL-, ML-WS₂ in pristine state. (b) Lorentzian fitting for E_2g and 2LA(M) peaks. Red circles represent experimental data, while blue, black, and green lines represent E_2g, 2LA(M), and combined peak fitting, respectively. (c) Raman spectra of SL-, ML-WS₂ films after 30 min KI treatment. Red shift in A_1g and E¹_2g modes was observed after doing.

4. Conclusion

In summary, a reliable doping method was developed to considerably enhance the electrical and photoelectric properties of SL-, ML-WS₂ FETs. Enhancements in charge carrier mobility and photocurrent response in WS₂ FETs after KI doping was confirmed by using electrical transport measurement and Raman spectroscopy. Shift of threshold voltage toward the negative gate voltage and the red shift of the E¹_2g and A_1g peaks in the Raman spectra confirm the n-type doping effect in WS₂ FETs. Semiconductor-to-metal transition was also observed when V_bg was increased over 30 V. WS₂ films show unprecedented high mobility of 255 cm² V⁻¹ s⁻¹ at room temperature. The on/off ratio of output current is ∼10⁸ at room temperature. The mobility of a ML-WS₂ FET was found to be 425 cm² V⁻¹ s⁻¹ at 5 K. The specific contact resistance was reduced up to 9 times after KI doping. n-Type doping also enhanced the photocurrent in WS₂ FETs. Chemical doping exhibited a very stable, effective, and easy-to-apply method to enhance the performance of a WS₂ FET.

Acknowledgements

The authors thank the Deanship of Scientific Research, Majmaah University under project No. 37-1-2 for funding this work for financial support. This research was supported by Nano-Material Technology Development Program (2012M3A7B4049888) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning. This research was also supported by Priority Research Center Program (2010-0020207) and the Basic Science Research Program (2013R1A1A2061396) through NRF funded by the Ministry of Education.

References

1. M. Iqbal, A. K. Singh, M. Iqbal and J. Eom, J. Phys.: Condens. Matter, 2012, 24, 335301.
2. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191.
3. M. Iqbal, A. K. Singh, M. Iqbal, S. Seo and J. Eom, J. Appl. Phys., 2012, 111, 084307.
4. M. Iqbal, S. Siddique, M. Iqbal and J. Eom, J. Mater. Chem. C, 2013, 1, 3078–3083.
5. M. Iqbal, M. F. Khan, M. W. Iqbal and J. Eom, J. Mater. Chem. C, 2014, 2, 5404–5410.
6. C. Zhou, W. Yang and H. Zhu, J. Chem. Phys., 2015, 142, 214704.
7. M. Y. Han, B. Özyilmaz, Y. Zhang and P. Kim, Phys. Rev. Lett., 2007, 98, 206805.
8. A. J. Samuels and J. D. Carey, ACS Nano, 2013, 7, 2790–2799.
9. J. B. Oostinga, H. B. Heersche, X. Liu, A. F. Morpurgo and L. M. Vandersypen, Nat. Mater., 2008, 7, 151–157.
10. X. Li, X. Wang, L. Zhang, S. Lee and H. Dai, Science, 2008, 319, 1229–1232.
11. T. Georgiou, R. Jalil, B. D. Belle, L. Britnell, R. V. Gorbachev, S. V. Morozov, Y.-J. Kim, A. Gholinia, S. J. Haigh and O. Makarovsky, Nat. Nanotechnol., 2013, 8, 100–103.
12. W. Zhao, Z. Ghorannevis, L. Chu, M. Toh, C. Kloc, P.-H. Tan and G. Eda, ACS Nano, 2012, 7, 791–797.
13. N. Peimyoo, W. Yang, J. Shang, X. Shen, Y. Wang and T. Yu, ACS Nano, 2014, 8, 11320–11329.
14. T. Georgiou, H. Yang, R. Jalil, J. Chapman, K. S. Novoselov and A. Mishchenko, Dalton Trans., 2014, 43, 10388–10391.
15. S. Jo, N. Ubrig, H. Berger, A. B. Kuzmenko and A. F. Morpurgo, Nano Lett., 2014, 14, 2019–2025.
16. W. S. Hwang, M. Remskar, R. Yan, V. Protasenko, K. Tahy, S. D. Chae, P. Zhao, A. Konar, H. G. Xing and A. Seabaugh, Appl. Phys. Lett., 2012, 101, 013107.
17. N. Perea-López, A. L. Elías, A. Berkdemir, A. Castro-Beltran, H. R. Gutiérrez, S. Feng, R. Lv, T. Hayashi, F. López-Urías and S. Ghosh, Adv. Funct. Mater., 2013, 23, 5511–5517.
18. N. Huo, S. Yang, Z. Wei, S.-S. Li, J.-B. Xia and J. Li, Sci. Rep., 2014, 4, 5209.
19. S. Bertolazzi, D. Krasnozhon and A. Kis, ACS Nano, 2013, 7, 3246–3252.
20. D. Ovchinnikov, A. Allain, Y.-S. Huang, D. Dumcenco and A. Kis, ACS Nano, 2014, 8, 8174–8181.
21. F. Withers, T. H. Bointon, D. C. Hudson, M. F. Craciun and S. Russo, Sci. Rep., 2014, 4, 4967.
22. D. J. Late, B. Liu, H. R. Matte, V. P. Dravid and C. Rao, ACS Nano, 2012, 6, 5635–5641.
23. W. Liu, J. Kang, D. Sarkar, Y. Khatami, D. Jena and K. Banerjee, Nano Lett., 2013, 13, 1983–1990.
24. S. Das, H.-Y. Chen, A. V. Penumatcha and J. Appenzeller, Nano Lett., 2012, 13, 100–105.
25. M. P. Levendorf, C.-J. Kim, L. Brown, P. Y. Huang, R. W. Havener, D. A. Muller and J. Park, Nature, 2012, 488, 627–632.
26. P. Yen, Y. Huang and K. Tiong, J. Phys.: Condens. Matter, 2004, 16, 2171.
27. S. Mouri, Y. Miyauchi and K. Matsuda, Nano Lett., 2013, 13, 5944–5948.
28. D. Kiriya, M. Tosun, P. Zhao, J. S. Kang and A. Javey, J. Am. Chem. Soc., 2014, 136, 7853–7856.
29. M. W. Iqbal, M. Z. Iqbal, M. F. Khan, M. A. Shehzad, Y. Seo and J. Eom, Nanoscale, 2015, 7, 747–757.
30. Y. Zhan, Z. Liu, S. Najmaei, P. M. Ajayan and J. Lou, Small, 2012, 8, 966–971.
31. Y.-H. Lee, L. Yu, H. Wang, W. Fang, X. Ling, Y. Shi, C.-T. Lin, J.-K. Huang, M.-T. Chang and C.-S. Chang, Nano Lett., 2013, 13, 1852–1857.
32. J. D. Lin, C. Han, F. Wang, R. Wang, D. Xiang, S. Qin, X.-A. Zhang, L. Wang, H. Zhang and A. T. S. Wee, ACS Nano, 2014, 8, 5323–5329.
33. M. W. Iqbal, M. Z. Iqbal, X. Jin, J. Eom and C. Hwang, J. Mater. Chem. C, 2014, 2, 7776–7784.
34. S. Kim, A. Konar, W.-S. Hwang, J. H. Lee, J. Lee, J. Yang, C. Jung, H. Kim, J.-B. Yoo and J.-Y. Choi, Nat. Commun., 2012, 3, 1011.
35. S. El-Mahalawy and B. Evans, Phys. Status Solidi A, 1977, 79, 713–722.
36. V. Podzorov, M. Gershenson, C. Kloc, R. Zeis and E. Bucher, Appl. Phys. Lett., 2004, 84, 3301–3303.
37. B. Chamlagain, Q. Li, N. J. Ghimire, H.-J. Chuang, M. M. Perera, H. Tu, Y. Xu, M. Pan, D. Xaio and J. Yan, ACS Nano, 2014, 8, 5079–5088.
38. B. Radisavljevic and A. Kis, Nat. Mater., 2013, 12, 815–820.
39. B. Chakraborty, A. Bera, D. Muthu, S. Bhowmick, U. V. Waghmare and A. Sood, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 85, 161403.
40. M. Buscema, D. J. Groenendijk, S. I. Blanter, G. A. Steele, H. S. van der Zant and A. Castellanos-Gomez, Nano Lett., 2014, 14, 3347–3352.
41. O. Lopez-Sanchez, D. Lembke, M. Kayci, A. Radenovic and A. Kis, Nat. Nanotechnol., 2013, 8, 497–501.
42. C. Lan, C. Li, Y. Yin and Y. Liu, Nanoscale, 2015, 7, 5974–5980.
43. M. W. Iqbal, M. Z. Iqbal, X. Jin, C. Hwang and J. Eom, ACS Appl. Mater. Interfaces, 2014, 6, 4207–4213.
44. A. Berkdemir, H. R. Gutiérrez, A. R. Botello-Méndez, N. Perea-López, A. L. Elías, C.-I. Chia, B. Wang, V. H. Crespi, F. López-Urías and J.-C. Charlier, Sci. Rep., 2013, 3, 1755.
45. H. Zeng, G.-B. Liu, J. Dai, Y. Yan, B. Zhu, R. He, L. Xie, S. Xu, X. Chen and W. Yao, Sci. Rep., 2013, 3, 1608.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra02390h

---