Nanoscale

COMMUNICATION

Cite this: DOI: 10.1039/c4nr03817g

Received 8th July 2014,Accepted 8th August 2014

DOI: 10.1039/c4nr03817g

www.rsc.org/nanoscale

Etching-free patterning method for electricalcharacterization of atomically thin MoSe2 filmsgrown by chemical vapor deposition†

M. Iqbal Bakti Utama,‡a Xin Lu,‡a Da Zhan,a Son Tung Ha,a Yanwen Yuan,a

Zexiang Shena and Qihua Xiong*a,b

Patterning two-dimensional materials into specific spatial arrange-

ments and geometries is essential for both fundamental studies of

materials and practical applications in electronics. However, the

currently available patterning methods generally require etching

steps that rely on complicated and expensive procedures. We

report here a facile patterning method for atomically thin MoSe2films using stripping with an SU-8 negative resist layer exposed to

electron beam lithography. Additional steps of chemical and physi-

cal etching were not necessary in this SU-8 patterning method.

The SU-8 patterning was used to define a ribbon channel from a

field effect transistor of MoSe2 film, which was grown by chemical

vapor deposition. The narrowing of the conduction channel area

with SU-8 patterning was crucial in suppressing the leakage

current within the device, thereby allowing a more accurate

interpretation of the electrical characterization results from the

sample. An electrical transport study, enabled by the SU-8 pattern-

ing, showed a variable range hopping behavior at high

temperatures.

1. Introduction

Layered two-dimensional (2D) materials such as graphene andthe transition metal dichalcogenides (TMDs, e.g. MoS2, WS2and WSe2) have attracted significant interest as a result oftheir interesting electronic and optical properties.1–3 To makeuse of these unique material properties in practical appli-cations, we need to be able to produce these materials in ascalable manner. Chemical vapor deposition (CVD)4–7 andliquid exfoliation8 have been reported to produce layered TMD

films with promisingly large areas and at high productionrates. The next crucial step in the manufacture of these 2Dmaterials into devices and systems for various electricalstudies and applications is the patterning of the films intospecific spatial arrangements and geometries. However, theabsence of dangling bonds on the surface of 2D TMDs resultsin an inert (0001) cleavage plane, making these materialsdifficult to remove using conventional wet etchants (e.g. aquaregia, HF or KOH) at room temperature (Fig. S1 in ESI†).

Several experimental procedures have been used to patternTMDs and most of these have been demonstrated usingMoS2.

9–13 At present, the patterning of TMDs is usually carriedout via lithography and a dry etching process with reactiveions (e.g. SF6 and XeF4).

9,14 However, reactive ion etchingrequires multi-step procedures (including both lithographyand the etching process itself ) and involve highly specializedequipment. Plasma etching that sputters the 2D materialsusing ion bombardment has also been reported,10 but sampleremoval may not be thorough using this technique becausethe process does not produce volatile by-products and is proneto redeposition. A thinning method using a laser beam hasalso been introduced,11 but this method may not be scalablebecause direct laser writing is required. Other thinningmethods, such as thermal annealing, are ineffective for pat-terning that is targeted to a confined area.15,16 Top-down pat-terning has also been reported; this incorporates stampingapproaches12 and, although an ordered pattern can beachieved, the sample may have a large variability in layer thick-ness, unlike CVD synthesis which can produce very thin layersin a controlled manner.

We report here a patterning method on a layered 2D filmusing SU-8 negative resist as a stripping layer. The 2D filmis first spin-coated with SU-8 and then the region of SU-8exposed by electron beam lithography (EBL) can be used tostrip the film beneath it. The patterning method with SU-8requires neither chemical nor physical etching steps. The SU-8patterning is therefore simpler than the currently available pat-terning methods for 2D materials, incurring potential econo-mic advantages in industrial-scale fabrication. Our approach

†Electronic supplementary information (ESI) available: Further experiments onpatterning and additional electrical characterizations data. See DOI: 10.1039/c4nr03817g‡These authors contributed equally to this work.

aDivision of Physics and Applied Physics, School of Physical and Mathematical

Sciences, Nanyang Technological University, Singapore 637371, Singapore.

E-mail: Qihua@ntu.edu.sgbNOVITAS, Nanoelectronics Centre of Excellence, School of Electrical and Electronic

Engineering, Nanyang Technological University, Singapore 639798, Singapore

This journal is © The Royal Society of Chemistry 2014 Nanoscale

Publ

ishe

d on

11

Aug

ust 2

014.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 27

/09/

2014

09:

16:3

5.

View Article OnlineView Journal

also minimizes the possibility of affecting the pre-existingstructure or device on the sample compared with other ionetching methods.

As a case study, we have demonstrated SU-8 patterning onan MoSe2 film grown by CVD.17 The patterning is effective forthe atomically thin samples that we produce, which can beprimarily controlled to thicknesses of 1–6 layers. We have alsofabricated devices from a patterned CVD MoSe2 ribbonchannel as a field-effect transistor (FET) to study the electricaltransport properties of MoSe2 produced by CVD. AlthoughMoS2 and MoSe2 are both layered TMD crystals, the muchlower natural abundance of MoSe2 causes the studies onMoSe2 to be less common than those on MoS2. With therecent emergence of CVD synthesis for the production of atom-ically thin MoSe2 over large areas,17–20 it is now important toinvestigate the electrical transport properties of MoSe2. SU-8patterning enables the fabrication of well-defined channels inthe MoSe2 film that minimize the leakage current, thus allow-ing the electrical characterization results to be accuratelyinterpreted.

2. Experimental2.1 Sample preparation

Large area atomically thin MoSe2 films (∼1 cm2) were syn-thesized on highly p-doped Si substrates with 100 nm thickSiO2 (University Wafers, Prime grade) using CVD.17 The fabri-cation and patterning of the MoSe2 devices were carried outusing EBL (a JEOL JSM-7001F instrument equipped with aDeben beam blanker and Nanometer Pattern GenerationSystem) at an acceleration voltage of 30 kV. The contact was50 nm of Ni in all the devices. Negative epoxy-based SU-8 2002(MicroChem) was used in all the SU-8 patterning. SU-8 wasspin-coated on the sample at 4000 rev min−1 for 40 s and thensoft-baked on a hot-plate at 65 °C for 1 min, followed byanother 1 min at 95 °C. After the EBL exposure the sample wasimmersed in SU-8 developer (MicroChem) for 1 min, bakedpost-exposure at 95 °C for 1 min, immersed in Remover PG(MicroChem) at ∼65 °C for 10–30 min to remove the SU-8pattern and then blow-dried with an N2 flow.

2.2 Characterization

The samples were characterized by scanning electronmicroscopy (SEM; JEOL JSM-7001F) and tapping mode atomicforce microscopy (AFM; Veeco Dimension V). SEM was per-formed with a 5 kV acceleration voltage in the top view bydetecting the secondary electron emission from the sample.Raman spectra were obtained on a triple-grating micro-Ramanspectrometer (Horiba-JY T64000). All the samples were excitedwith a 532 nm laser. The signal was collected through a 100×objective, dispersed with a 1800 g mm−1 grating, and detectedusing a charge-coupled device cooled by liquid N2. Ramanmapping was performed with a Renishaw InVia Raman systemusing 532 nm laser excitation. The device measurements weremade on a probe station (Lakeshore) connected to a current

pre-amplifier and controlled by a Semiconductor Device Analy-zer (Keithley). The electrical measurements were made atseveral temperatures between 350 and 83 K, with pressure<10−5 mbar.

3. Results and discussion3.1 Patterning of atomically thin MoSe2 film with SU-8

Fig. 1 illustrates schematically the process of patterning withSU-8. Briefly, after the synthesis of the large area and atomic-ally thin MoSe2 film with CVD on an SiO2/Si substrate, theSU-8 solution was spin-coated and soft-baked on top of thesample. As SU-8 is a negative resist, EBL was used to exposethe area of the SU-8 layer in a defined shape and size thatremained on the sample after the resist development step. Thedeveloped SU-8 structure carries the MoSe2 layer along beneathit on removal of the SU-8 layer with the Remover PG solvent.Thus the MoSe2 film is removed in an area, the shape of whichcorresponds to the SU-8 area exposed during the EBL process.

Fig. 2a and 2b show the fabrication process for an array ofhexagons opening from the MoSe2 film grown by CVD as seenusing optical microscopy. The color contrast between theMoSe2 film and the exposed SiO2 substrate is not immediatelydistinguishable since the film thickness is only within a fewnanometers. Fig. 2c shows the same sample area under SEM,where the difference in the yield of the secondary electronemission signal from the MoSe2 film and SiO2 substrate issufficient to produce an image with a visible difference in con-trast. The MoSe2 under the SU-8 regions not exposed by EBL isnot noticeably changed by the patterning (Fig. S2†).

To corroborate the success of the SU-8 patterning, we com-pared the Raman spectra collected from the area from whichMoSe2 was removed with that from the layer where MoSe2remained (Fig. 2d). Raman spectroscopy is a reliable tool tocharacterize the presence of atomically thin TMDs as phononfrequencies are not only sensitive to the materials beingstudied, but are also rich in information with which to probeinterlayer coupling and layer thickness.21,22 The micro-Ramansetup also allows signal collection from only the local areawhere the MoSe2 film has been removed. The Raman signa-tures of MoSe2 were identified from the remaining layer, inwhich the signal from the out-of-plane A1g mode was particu-

Fig. 1 Schematic diagram of the patterning process.

Communication Nanoscale

Nanoscale This journal is © The Royal Society of Chemistry 2014

Publ

ishe

d on

11

Aug

ust 2

014.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 27

/09/

2014

09:

16:3

5.

View Article Online

larly intense. These Raman modes were not seen in the spec-trum of the area from which MoSe2 had been removed. Themapping of the Raman intensity of the A1g mode (241 cm−1;Fig. 2e) shows that the MoSe2 layer was removed by the SU-8patterning method (AFM; Fig. 2f) at the boundary of theremoved area on a similarly patterned sample, which verifiesthat the patterning can result in the clean removal of 2Dmaterials with a relatively sharp edge. Interestingly, we did notobserve redeposition issues with SU-8 patterning, unlike withthe etching methods that rely on purely physical means(Fig. S3†).

We attribute the mechanism behind the removal of MoSe2by SU-8 patterning to the adhesion between the exposed SU-8and the MoSe2 film. As long as the adhesion between theMoSe2 film and SU-8 is sufficiently stronger than the adhesionbetween the MoSe2 film and the SiO2 substrate, the MoSe2 willbe removed when the SU-8 is dissolved. In addition, theadhesion of the films to the SiO2 layer is generally difficult; toachieve good adhesion, an intermediate layer that can beoxidized (e.g. Cr, Ti) is needed to block the transport of oxygenfrom within the SiO2 to the interface. As MoSe2 does notoxidize rapidly in air, the adhesion between MoSe2 and SiO2 isrelatively weak and therefore the SU-8 layer can be used to peelthe MoSe2 film off from the SiO2.

23

Similar to most other negative resists, exposure to EBL andpost-exposure baking cause the SU-8 polymer to be highlycross-linked and more rigid. The rigidity of the exposed SU-8structure helps to improve the conformal contact with the areaof MoSe2 below the SU-8, ensuring that the MoSe2 layer isuniformly stripped from the substrate as a single unit – as

opposed to being dissolved in smaller patches – for the com-plete removal of the MoSe2 along with the SU-8 stripping.Given the inertness of the (0001) plane of MoSe2, we do notassume that there are any unique chemical bonds betweenSU-8 and MoSe2 that allow the patterning. Consistent with ourassumption, we observed that the patterning approach is notunique to SU-8; we found that patterning can be achievedusing a metal sacrificial layer which strips the MoSe2 filmwhen the metal is etched (Fig. S1†). SU-8 patterning can alsobe used to remove MoS2 flakes produced on SiO2 by theexfoliation method (Fig. S4†). It is conceivable that similarmethods could be developed using other lithography resistpolymers. However, the inability of other EBL resists, such aspolymethyl methacrylate (PMMA), to be used as the strippinglayer for patterning was attributed to the poorer adhesionbetween PMMA and MoSe2.

With respect to the adhesion mechanism of the removal,we believe that the SU-8 patterning method will be especiallyuseful for layered material samples with small grains. Whenthe grain size of MoSe2 is smaller than the feature size of theSU-8 pattern, the presence of the grains will help to ensurethat the film will break following the shape of the SU-8. Thuswe believe that the SU-8 patterning method will have impor-tant applications in layered material films produced from pro-cesses such as liquid phase exfoliation8 and certain CVDsynthesis approaches.6,17

3.2 Device applications

An immediate application of the SU-8 patterning is to define aconduction channel for electrical characterization.Wepatterned

Fig. 2 Characterization of patterning on MoSe2. Optical micrographs of MoSe2 samples (a) with developed SU-8 layer and (b) after removal of theSU-8 layer, thus also removing MoSe2 in an array with a hexagonal area. (c) SEM image of the sample after patterning. (d) Raman spectra collected inthe area of the MoSe2 and the area where the MoSe2 layer was removed by SU-8. The wavenumber of the identified modes is given in the plot. (e)Raman mapping of the sample at the A1g mode. (f ) Typical AFM image at the boundary of a patterned area. Inset: height profile from the image.

Nanoscale Communication

This journal is © The Royal Society of Chemistry 2014 Nanoscale

Publ

ishe

d on

11

Aug

ust 2

014.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 27

/09/

2014

09:

16:3

5.

View Article Online

a ribbon channel from an MoSe2 film in a back-gated FETdevice (Fig. 3a and 3b). Two 80 × 80 μm2 Ni pads were used asthe device electrodes. It is possible to avoid any defect struc-tures existing on the substrate that might hamper the perform-ance of the device by constraining the current flow into thechannel. For example, we noticed a significant amount of gateleakage current (∼nA; Fig. 3c) in an MoSe2 film FET before pat-terning. A typical signature of gate leakage was shown, i.e. thecurrent direction was inverted when the gate voltage polaritywas switched, which implies that there was conductionbetween the source–drain electrodes and the back gate. Bynarrowing the area of the MoSe2 conduction channel into aribbon, the gate leakage level was suppressed by two orders ofmagnitude (Fig. 3c, inset) and became nearly independent ofthe gate voltage sweep. As a direct benefit of such gate leakagesuppression, the performance properties of the device, such asthe I–V characteristics (Fig. S5a†) and electrical transport pro-perties (Section 3.3), could be measured more accurately.

Our hypothesis for the mechanism of gate leakagereduction with device patterning is explained schematically inFig. 3d. Defects on the substrate, such as pinholes in the gateoxide, may form due to imperfections in the production of the

substrate, the high temperature during sample synthesis (asalso observed for the MoS2 FET grown by CVD24) or during thedevice fabrication and processing steps. The creation of oxidedefects is particularly problematic as it generally createsreliability issues, such as a short circuit of the back gate withsource/drain electrodes. Before patterning, the MoSe2 film mayhave an electrical connection with these defects, which resultsin appreciable gate leakage. The current leakage issue isexacerbated as a result of the large area (∼1 cm2) of the as-grown film,17 which causes a connection between the deviceand existing oxide defects somewhere on the substrate. Byreducing the area of the active conduction channel by SU-8patterning it is possible to avoid electrical connection betweenthe MoSe2 layer and such oxide defects, thus eliminating thegate leakage current.

Consistent with our hypothesis that SU-8 patterning maybreak the electrical connection between a device and any oxidedefects, we showed that SU-8 patterning can also be used toelectrically isolate the device. By performing SU-8 patterningon the conduction channel (Fig. 3e), the channel can be cut toproduce an open circuit. Before the channel cut, the draincurrent flows in response to the application of a bias voltage(Fig. 3f). However, the drain current becomes insensitive tothe bias voltage after the channel cut, decreasing to about thepA level, which corresponds to the noise level of the measure-ment (Fig. 3f, inset). This result confirms that the removal ofMoSe2 from the channel was complete.

3.3 Electrical transport measurements

The SU-8 patterning method enables a more in-depth charac-terization of the transport properties of the MoSe2 grown byCVD. Temperature-dependent electrical characterizations wereperformed on the sample (Fig. 4a). The MoSe2 channel had ahigh resistance of tens of MΩ at room temperature, multipleorders of magnitude higher than that of the contact resistancebetween MoSe2 and Ni, which reaches only hundreds of kΩ ata 35 V gate overdrive.25 The drain current was also stronglydependent on temperature, showing an insulating behaviorwhere the conductance decreased with decreasing tempera-ture. The current level at temperatures <80 K was beyond ourmeasurable noise level. This suggests that the MoSe2 grown byCVD is highly disordered and that electron localization andthermally activated hopping conduction may play importantroles in the electronic transport within the MoSe2 grown byCVD. We also noted that the temperature-dependent conduc-tance data show non-linearity when graphed in an Arrheniusplot (Fig. 4b). The non-linearity of the graph indicates that thesimple thermally activated model of transport, where σ ∝exp(−Ea/kBT ), does not explain the temperature dependence ofour sample. Thus we attempt here to explain the transportbehavior in terms of the variable range hopping (VRH)mechanism.

VRH transport has been observed in both monolayer andfew-layered 2D materials, including MoS2.

26–30 Likewise, VRHhas also been reported in bulk MoSe2.

31 However, the trans-port behavior in ultra-thin MoSe2 remains elusive. Previous

Fig. 3 SU-8 patterning for device application. All scale bars in theimage correspond to 10 μm. (a, b) Typical example of an optical micro-graph of a MoSe2 back-gated FET (a) before and (b) after patterning intoa ribbon channel. (c) Gate leakage current measured for an MoSe2device before and after patterning. Inset: the same data plotted on asemi-log scale. (d) Schematic explanation of the reduction in theleakage current with device patterning. (e) Process of cutting thechannel to create an open circuit. Inset: SEM image of the cut channel.(f ) I–V plot of an MoSe2 device before (red) and after (blue) channelcutting. Inset: plot of cut channel on a semi-log scale.

Communication Nanoscale

Nanoscale This journal is © The Royal Society of Chemistry 2014

Publ

ishe

d on

11

Aug

ust 2

014.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 27

/09/

2014

09:

16:3

5.

View Article Online

experimental data on few-layered MoSe2 exfoliated frompowders seemed to indicate phonon-limited transport, wherethe device mobility decreases with increasing temperature.25

In both exfoliated MoS2 and bulk MoSe2, the VRH behavioronly started at temperatures <200 K. In contrast, our MoSe2deposited by CVD devices show VRH at high temperatures(350 K), similar to that seen in previous work on reducedgraphene oxide.32

In VRH, the electrons occupy localized states which have adistribution of energy and positions in real space. Hoppingoccurs due to the mutual overlap of the eigen-functions ofneighbouring states. The conductance of electrons is pro-portional to the probability per unit time of hopping occur-ring, which is given by P(r) ∝ exp[−2r/ξ − ΔE(r)/kBT], where r isthe hopping distance, ξ is the localization length and ΔE is theactivation energy.33 For Mott-VRH in 2D materials it isassumed that the density of states N(E) is constant near theFermi level. The most likely hopping maximizes P(r) by circum-scribing dP/dr = 0. This results in the VRH conductivity, whichcan be written as

σðTÞ ¼ σ0 expð�T0=TÞ1=3; ð1Þ

where T0 is a fitting parameter.33 We note that there have beendifferent accounts of the temperature dependence of the expo-nential prefactor, σ0, in 2D Mott-VRH. Ghatak et al.26 reportedthat σ0 ∝ T0.8, as determined in a theoretical study.34 Other

previous experimental and theoretical works have reported avalue of σ0 that is independent of temperature,32,33,35 whereasother workers have reported different types of temperaturedependence.36,37 For our experiments, we note that the temp-erature dependence of σ0 would not dramatically affect ourconclusion and hence we assume in the following discussionthat σ0 is constant.

We extracted the temperature-dependent conductance ofthe sample at the quasi-Ohmic regime near Vds = 0 V. Fig. 4cshows that the data can be fitted adequately with the Mott-VRH model from eqn (1), giving a coefficient of determination(R2) > 0.995 for all values of Vg. The inset shows the T0 fittingparameter with respect to the carrier concentration, n =Cg(Vg − Vth), where Cg = 3.45 × 10−4 F m−2 is the gate capaci-tance and Vth is the threshold voltage. In support of our con-clusion about hopping transport, we find that the mobility ofthe sample increases with temperature (Fig. 4d). This obser-vation is in contrast with the band-like diffusion transportobserved in MoS2 and MoSe2 flakes from exfoliation.25,29 More-over, the mobility also follows an exp(−T0*/T )

1/3 trend, inagreement with a Mott-VRH in 2D samples.26,37

In MoS2, Qiu et al.27 attributed the carrier localizationresponsible for VRH to the sulfur vacancies, which are also theelectron donors for the sample. In an analogous manner, wealso observed that there are selenium vacancies in our sample,as verified from scanning transmission electron microscopy.17

Thus it is reasonable to suggest that a similar mechanism also

Fig. 4 Electrical transport properties of patterned MoSe2 grown by CVD ribbon channel. (a) I–V characteristics of two-probe device at varioustemperatures; channel length 0.5 μm. Inset: optical micrograph of a typical patterned ribbon device. Scale bars: 100 μm (left) and 10 μm (right). (b)Arrhenius plot of the width-normalized conductance data. Black, broken straight line is a guide to the eye. (c) Temperature dependence of width-normalized conductance on a semi-log scale plotted with respect to T−1/3. Inset: fitting parameter T0 as a function of carrier density. (d) Temperaturedependence of the field-effect mobility of electrons. Inset: simplified band diagram showing the position of localized states.

Nanoscale Communication

This journal is © The Royal Society of Chemistry 2014 Nanoscale

Publ

ishe

d on

11

Aug

ust 2

014.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 27

/09/

2014

09:

16:3

5.

View Article Online

contributes to the VRH in MoSe2. We also observed the fittingparameter T0 in MoSe2 to be one order of magnitude higherthan that in MoS2. For 2D Mott-VRH, T0 is given by

T0 ¼ 13:8kBNðEFÞξ2

; ð2Þ

where N(EF) is the density of states near the Fermi level.38

Given that eqn (1) in Mott-VRH is derived by assuminga nearly constant N(EF) within the range of several kBT, then∂n/∂T = N(EF)kB at temperatures approaching 0 K. As a roughorder of magnitude approximation35 we can take ∂n/∂T ≈n(100 K)/100 K to calculate N(EF), which is about 3 × 1014 eV−1

cm−2. Thus, using eqn (2), we obtain ξ ≈ 1 Å, lower than thatassumed in MoS2 exfoliated flakes (6 Å) in previously pub-lished work.27

A shorter ξ suggests that it is more difficult for charge car-riers to hop across localized states distantly separated in space.This is consistent with the relatively low conductance andmobility of the MoSe2 film grown by CVD (an electron mobi-lity17 reaching only 0.02 cm2 V−1 s−1 at 300 K). Although theMoSe2 film mobility seems low, the mobility value is withinthe typical values obtained for previously reported MoS2films grown by CVD, which also contain numerous grains(∼0.003–0.8 cm2 V−1 s−1).4–6,39 The occurrence of VRH behaviorat room temperature can also explain the performance of theMoSe2 FET, which has an on–off ratio of only 102 at a 37.5 Vgate overdrive. Considering that carrier localization inducesstates at the band gap (see Fig. 4d, inset), it is still possible forthe charge carriers to contribute towards the electrical conduc-tion across the device, although the gate voltage was tuned toposition the Fermi level within the band gap of MoS2. Thedecrease in T0 at higher carrier concentrations is attributed tothe increased screening effect, thus extending the localizationlength.

4. Conclusions

In conclusion, we have shown the patterning of 2D materialswith SU-8 on MoSe2 grown by CVD. The success of the pattern-ing was characterized optically and electrically. The patterningenables further study and the development of applications ofMoSe2 grown by CVD in electronics, where the narrowing ofthe conduction channel area with SU-8 patterning is beneficialin significantly reducing the leakage current. We found thatthe film showed a variable range hopping behavior.

The SU-8 patterning method can be used as an alternativeto an etching process with a pre-existing structure on thesample. For instance, it is possible to use the patterningmethod to create an array of openings on the sample forfurther fabrication steps, such as another growth step of the2D material to stitch additional material into the patternedarea. In this way, lateral heterostructures with more complexarrangements can be realized.40,41

Although EBL was used in this study, we believe that it isalso possible to use UV photolithography for SU-8 patterning

because this method could also create similar cross-linking toinduce stronger adhesion of the SU-8 to the 2D film. The useof photolithography will also help to improve the throughputof the patterning.

With respect to the materials used in the patterning, it islikely that other lithography resists could be used for a similarpatterning approach, subject to adequate adhesion with thematerial to be patterned. However, considering that the pat-terning depends only on the strong adhesion of SU-8 to MoSe2after exposure with EBL, with neither unique chemical reactionnor binding assumed, we also think that the method shouldbe general and therefore also applicable to other 2D materials,including graphene, h-BN, MoS2, and WSe2.

Acknowledgements

Q.X. thanks the Singapore National Research Foundation for afellowship grant (NRF-RF2009-06), the Ministry of Educationfor an AcRF tier2 grant (MOE2012-T2-2-086) and NanyangTechnological University for a start-up grant (M58110061)supports.

Notes and references

1 Q. H. Wang, K. Kalantar-Zadeh, A. Kis, J. N. Coleman andM. S. Strano, Nat. Nanotechnol., 2012, 7, 699–712.

2 S. Z. Butler, S. M. Hollen, L. Cao, Y. Cui, J. A. Gupta,H. R. Gutiérrez, T. F. Heinz, S. S. Hong, J. Huang,A. F. Ismach, E. Johnston-Halperin, M. Kuno,V. V. Plashnitsa, R. D. Robinson, R. S. Ruoff, S. Salahuddin,J. Shan, L. Shi, M. G. Spencer, M. Terrones, W. Windl andJ. E. Goldberger, ACS Nano, 2013, 7, 2898–2926.

3 A. K. Geim and I. V. Grigorieva, Nature, 2013, 499, 419–425.4 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 andL. J. Li, Nano Lett., 2012, 12, 1538–1544.

5 Y. H. Lee, X. Q. Zhang, W. Zhang, M. T. Chang, C. T. Lin,K. D. Chang, Y. C. Yu, J. T. Wang, C. S. Chang, L. J. Li andT. W. Lin, Adv. Mater., 2012, 24, 2320–2325.

6 Y. Zhan, Z. Liu, S. Najmaei, P. M. Ajayan and J. Lou, Small,2012, 8, 966–971.

7 Y. Zhang, Q. Ji, J. Ju, H. Yuan, J. Shi, T. Gao, D. Ma, M. Liu,Y. Chen, X. Song, H. Y. Hwang, Y. Cui and Z. Liu, ACSNano, 2013, 7, 8963.

8 J. N. Coleman, M. Lotya, A. O’Neill, S. D. Bergin, P. J. King,U. Khan, K. Young, A. Gaucher, S. De, R. J. Smith,I. V. Shvets, S. K. Arora, G. Stanton, H. Y. Kim, K. Lee,G. T. Kim, G. S. Duesberg, T. Hallam, J. J. Boland,J. J. Wang, J. F. Donegan, J. C. Grunlan, G. Moriarty,A. Shmeliov, R. J. Nicholls, J. M. Perkins, E. M. Grieveson,K. Theuwissen, D. W. McComb, P. D. Nellist andV. Nicolosi, Science, 2011, 331, 568–571.

Communication Nanoscale

Nanoscale This journal is © The Royal Society of Chemistry 2014

Publ

ishe

d on

11

Aug

ust 2

014.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 27

/09/

2014

09:

16:3

5.

View Article Online

9 Y. Huang, J. Wu, X. Xu, Y. Ho, G. Ni, Q. Zou, G. Koon,W. Zhao, A. H. Castro Neto, G. Eda, C. Shen andB. Özyilmaz, Nano Res., 2013, 6, 200–207.

10 Y. L. Liu, H. Y. Nan, X. Wu, W. Pan, W. H. Wang, J. Bai,W. W. Zhao, L. T. Sun, X. R. Wang and Z. H. Ni, ACS Nano,2013, 7, 4202–4209.

11 A. Castellanos-Gomez, M. Barkelid, A. M. Goossens,V. E. Calado, H. S. van der Zant and G. A. Steele, Nano Lett.,2012, 12, 3187–3192.

12 H. Nam, S. Wi, H. Rokni, M. Chen, G. Priessnitz, W. Luand X. Liang, ACS Nano, 2013, 7, 5870–5881.

13 F. Bonaccorso, A. Lombardo, T. Hasan, Z. Sun, L. Colomboand A. C. Ferrari, Mater. Today, 2012, 15, 564–589.

14 V. Rousset, S. Itoua, C. Joachim, F. Tsobnang, B. Roussetand H. H. Pham, Nanotechnology, 1996, 7, 144.

15 X. Lu, M. I. Utama, J. Zhang, Y. Zhao and Q. H. Xiong,Nanoscale, 2013, 5, 8904–8908.

16 J. Wu, H. Li, Z. Yin, J. Liu, X. Cao, Q. Zhang and H. Zhang,Small, 2013, 9, 3314.

17 X. Lu, M. I. B. Utama, J. Lin, X. Gong, J. Zhang, Y. Zhao,S. T. Pantelides, J. Wang, Z. Dong, Z. Liu, W. Zhou andQ. H. Xiong, Nano Lett., 2014, 14, 2419–2425.

18 J. Shaw, H. Zhou, Y. Chen, N. Weiss, Y. Liu, Y. Huang andX. Duan, Nano Res., 2014, 7, 1–7.

19 X. Wang, Y. Gong, G. Shi, W. L. Chow, K. Keyshar, G. Ye,R. Vajtai, J. Lou, Z. Liu, E. Ringe, B. K. Tay andP. M. Ajayan, ACS Nano, 2014, 8, 5125–5131.

20 J. Xia, X. Huang, L.-Z. Liu, M. Wang, L. Wang, B. Huang,D.-D. Zhu, J.-J. Li, C.-Z. Gu and X.-M. Meng, Nanoscale,2014, 6, 8949–8955.

21 Y. Zhao, X. Luo, H. Li, J. Zhang, P. T. Araujo, C. K. Gan,J. Wu, H. Zhang, S. Y. Quek, M. S. Dresselhaus andQ. H. Xiong, Nano Lett., 2013, 13, 1007–1015.

22 X. Luo, Y. Zhao, J. Zhang, Q. H. Xiong and S. Y. Quek, Phys.Rev. B: Condens. Matter, 2013, 88, 075320.

23 H. Nagata, T. Shinriki, K. Shima, M. Tamai and E. MinHaga, J. Vac. Sci. Technol., A, 1999, 17, 1018–1023.

24 H. Schmidt, S. Wang, L. Chu, M. Toh, R. Kumar, W. Zhao,A. H. Castro Neto, J. Martin, S. Adam, B. Özyilmaz andG. Eda, Nano Lett., 2014, 14, 1909–1913.

25 S. Larentis, B. Fallahazad and E. Tutuc, Appl. Phys. Lett.,2012, 101, 223104.

26 S. Ghatak, A. N. Pal and A. Ghosh, ACS Nano, 2011, 5,7707–7712.

27 H. Qiu, T. Xu, Z. Wang, W. Ren, H. Nan, Z. Ni, Q. Chen,S. Yuan, F. Miao, F. Song, G. Long, Y. Shi, L. Sun, J. Wangand X. Wang, Nat. Commun., 2013, 4, 2642.

28 B. Radisavljevic and A. Kis, Nat. Mater., 2013, 12, 815–820.29 D. Jariwala, V. K. Sangwan, D. J. Late, J. E. Johns,

V. P. Dravid, T. J. Marks, L. J. Lauhon and M. C. Hersam,Appl. Phys. Lett., 2013, 102, 173107.

30 M. Ahmad, M. A. Rafiq, Z. Imran, K. Rasool, R. N. Shahid,Y. Javed and M. M. Hasan, J. Appl. Phys., 2013, 114, 043710.

31 J. C. Bernègde, A. Mallouky and J. Pouzet, Phys. StatusSolidi A, 1989, 111, 181–191.

32 C. Gómez-Navarro, R. T. Weitz, A. M. Bittner, M. Scolari,A. Mews, M. Burghard and K. Kern, Nano Lett., 2007, 7,3499–3503.

33 N. F. Mott, Electronic Processes in Non-Crystalline Materials,Clarendon Press, Oxford, UK, 1979.

34 F. W. Van Keuls, X. L. Hu, H. W. Jiang and A. J. Dahm,Phys. Rev. B: Condens. Matter, 1997, 56, 1161–1169.

35 G. Eda, C. Mattevi, H. Yamaguchi, H. Kim andM. Chhowalla, J. Phys. Chem. C, 2009, 113, 15768–15771.

36 D. N. Tsigankov and A. L. Efros, Phys. Rev. Lett., 2002, 88,176602.

37 G. Paasch, T. Lindner and S. Scheinert, Synth. Met., 2002,132, 97–104.

38 B. I. Shklovskii and A. L. Efros, Electronic Properties ofDoped Semiconductor, Springer, Berlin, 1984.

39 Y. C. Lin, W. Zhang, J. K. Huang, K. K. Liu, Y. H. Lee,C. T. Liang, C. W. Chu and L. J. Li, Nanoscale, 2012, 4,6637–6641.

40 Z. Liu, L. Ma, G. Shi, W. Zhou, Y. Gong, S. Lei, X. Yang,J. Zhang, J. Yu, K. P. Hackenberg, A. Babakhani,J.-C. Idrobo, R. Vajtai, J. Lou and P. M. Ajayan, Nat. Nano-technol., 2013, 8, 119–124.

41 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.

Nanoscale Communication

This journal is © The Royal Society of Chemistry 2014 Nanoscale

Publ

ishe

d on

11

Aug

ust 2

014.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 27

/09/

2014

09:

16:3

5.

View Article Online