Flexible chemical sensors based on hybrid layer consisting ofmolybdenum disulphide nanosheets and carbon nanotubes
Sungho Kim a, b, 1, Jinkyu Han a, 1, Min-A. Kang a, Wooseok Song a, Sung Myung a, *,Sang-Woo Kim b, Sun Sook Lee a, Jongsun Lim a, Ki-Seok An a
a Thin Film Materials Research Center, Korea Research Institute of Chemical Technology, Yuseong Post Office Box 107, Daejeon 305-600, Republic of Koreab School of Advanced Materials Science & Engineering, Sungkyunkwan University, 2066 Seobu-ro, Jangan-gu, Suwon, Gyeonggi-do, 440-746, Republic ofKorea
a r t i c l e i n f o
Article history:Received 21 September 2017Received in revised form25 November 2017Accepted 17 December 2017Available online 20 December 2017
A facile synthesis method was developed for preparing hybrid 2-dimensional (2D) films, based on large-scale molybdenum disulfide (MoS2) nanosheets and single-walled carbon nanotubes (SWCNTs), forflexible sensors. Here, 1-dimensional (1D) SWCNTs were combined with MoS2 nanosheets during theMoS2 synthesis process for improving the flexibility and stability of the 2D MoS2 nanosheet. UniformMoS2 nanosheets were successfully synthesized via chemical vapor deposition (CVD) using a porphyrin-type organic promoter. This study demonstrates the high performance and enhanced sensitivity of thechemical gas sensors that were fabricated using hybrid MoS2-SWCNT layers; the enhancement is due tothe sensitive gas adsorption by SWCNTs in the MoS2 nanosheets. In addition, the hybrid MoS2-SWCNTfilms, transferred on a flexible polyethylene terephthalate (PET) substrate, were employed for theanalysis of physical properties of chemical sensors as a function of the number of bending cycles. Thehybrid MoS2-SWCNT-based sensors showed stable sensing performance after 105 bending cycles,whereas the resistance of MoS2-based sensors increased to approximately 300% under the same bendingprocess.
Transition metal dichalcogenides (TMDs) have attractedconsiderable attention in recent years, owing to their potentialapplication in various fields including microelectronics, flexibledevices, and chemical sensors [1e6]. Among two-dimensional (2D)TMDs, molybdenum disulfide (MoS2) is one of the most promisingand efficient semiconductors because of its inherent and thickness-dependent band gap. For example, bulk MoS2 is a semiconductorwith a narrow band gap of approximately 1.3 eV, while the MoS2single layer possesses a large band gap of 1.8 eV because of thetransition from indirect band gap to direct band gap. Synthesis oflarge-scale MoS2 nanosheets by chemical vapor deposition (CVD) issuitable for the mass-fabrication of MoS2, wherein MoS2 nano-sheets have generally been obtained by the sulfurization reaction ofdeposited molybdenum or molybdenum oxide [7e10]. Various
organic seeding promoters have been employed for large-scaleMoS2 synthesis, but a facile synthesis method for MoS2 nano-sheets with a uniform thickness remains a challenge [11]. Recently,many studies have also focused on hybrid nanostructures of TMDsand carbon-based nanomaterials to enhance the flexibility andadsorption sensitivity of gas molecules compared to carbon-basedmaterials [12,13]. Carbon nanomaterials of these hybrid nano-structures have been mainly used with reduced graphene oxide orgraphene [14e16]. Although they exhibit high gas sensing proper-ties, studies on durability for wearable devices and measurementson transparent properties have not been reported yet. To apply towearable devices, it is necessary to improve durability byimproving flexibility and to make transparent materials areessential. SWCNT has excellent flexibility applicable to wearabledevices and a transparent specification [17]. The TMD material isknown to have excellent selectivity and sensitivity, so synergisticeffects can be expected when it is combined with SWCNT. In gen-eral, SWCNTs are known to affect charge transfer, electrostaticenvironment, surface area, andmolecular adsorption in gas sensingproperties of composites [18]. Improvements in gas sensing in
composites with polymers [19], SnO2 [20], andWO2 [21] films havebeen reported using these properties.
In this study, we developed a facile method to synthesize MoS2-SWCNT hybrid films for flexible chemical sensors. The MoS2 layerwas synthesized via thermal CVD (TCVD) on a SWCNT network toincrease the contact force between MoS2 nanosheets and SWCNTs.SWCNTs were first spin-coated on a SiO2 substrate, and an organicpromoter with nanometer-scale thickness was deposited via ther-mal evaporation for the growth of uniformMoS2 nanosheets over alarge area. Conventional TCVD was subsequently carried out forMoS2 synthesis on the promoter layer. Finally, the MoS2-SWCNThybrid layer was transferred onto a flexible and transparent PETsubstrate using a poly (methyl methacrylate)-based (PMMA-based)transfer method, and high-performance flexible chemical sensorsbased on MoS2-SWCNT hybrid films were fabricated. This methodallowed us to improve the physical flexibility and sensing sensi-tivity of chemical sensors compared to previous devices based onpristine 2D materials.
2. Experimental
2.1. Preparation of SWCNTs on a target substrate
A SWCNT suspension was prepared by dispersing 0.2mg ofpurified SWCNTs (Carbon Nano-material Technology Co, Ltd.) in100ml of 1,2-dichlorobenzene with ultrasonic vibration for 20min.The as-prepared SWCNT solution (0.6ml) was spin-coated onto aUV-treated SiO2 substrate at 3000 rpm for 30 s. After spin-coating,the SWCNT/SiO2 was annealed at 150 �C for 1min to remove 1,2-dichlorobenzene.
2.2. CVD growth of MoS2 nanosheet on promoter layers
5, 10, 15, 20-tetraphenylporphyrin (H2TPP) thin films, used as apromoter layer, were formed on the SWCNT-coated SiO2 usingthermal evaporation. Mo solutionwas prepared by dissolving 0.1Mammonium heptamolybdate (Fluca, 99%) in 10ml of distilled water.The solution was subsequently coated onto UV-treated SiO2(300 nm) substrates by spin-coating at 2000 rpm for 30 s. Sulfurpowder amounting to 0.1 g (SAMCHUN, 98.0%), used as the sulfursource, was placed upstream in the reactor. MoS2-SWCNT nano-sheets were synthesized at 900 �C under ~1 Torr pressure whileintroducing Ar (100 sccm) for 5min.
2.3. Transferring of MoS2 and MoS2-SWCNT films to flexiblesubstrates
PMMAwas spin-coated on CVD-grown MoS2 and MoS2-SWCNThybrid films at 2000 rpm for 30 s, and subsequently, the PMMA-coated substrate was placed in 4M NaOH in 20mL of distilled-deionized water to remove the SiO2 layer. After the SiO2 layerwas completely etched away, the PMMA-coated hybrid nanosheetwas transferred to PET substrates. Finally, the PMMA layer wasremoved with acetone, and the hybrid nanosheet was rinsed withdistilled-deionized water.
3. Results and discussion
A SWCNT suspension was prepared by dispersing 0.2mg ofpurified SWCNTs (Carbon Nano-material Technology Co, Ltd.) in100ml of 1,2-dichlorobenzene with ultrasonic vibration for 20min.The as-prepared SWCNT solution (0.6ml) was spin-coated onto aSiO2 substrate at 500e3000 rpm for 30 s. The 5, 10, 15, 20-tetrakis(4-hydroxyphenyl)-21H, 23H-porphyrin (p-THPP) layer,which acts as an organic source, was formed on the SWCNT
network by thermal evaporation, and conventional TCVD was car-ried out for MoS2 growth on the organic promoter. Ammoniumheptamolybdate, used as a Mo source (1.3 g), was dissolved indistilled-deionized water (10ml), and this solution was subse-quently spin-coated onto an SiO2 substrate. Ammonium hepta-molybdate and sulfur powder (sulfur source) were locatedupstream in the reactor. The MoS2 layer was grown on the SWCNT-coated target substrate at 900 �C for 5min while simultaneouslyintroducing Ar gas, where the promoter is removed completely(Figs. S1 and S2). The as-grown MoS2-SWCNT hybrid films on theSiO2 substrate were then cooled to room temperature. After MoS2synthesis, the films were transferred onto other SiO2 surfaces usinga transfer method discussed in previous works [22]. The hybridfilms were patterned by oxygen plasma etching using Al patterns asmasks for device channels in Fig. S3 in Supporting Information.Finally, MoS2-SWCNT-based devices were fabricated by the depo-sition of source and drain electrodes (7 nm Cr/70 nm Au) through ashadow mask (Fig. 1(a)).
Structural analyses of SWCNTs on a solid substrate, MoS2nanosheets without SWCNTs, and MoS2-SWCNT hybrid films wereperformed using scanning electron microscopy (SEM) and atomicforce microscopy (AFM), as shown in Fig. 1(b) and Fig. S4 in thesupporting information. When a 0.05mg/ml SWCNT solution wasspin-coated at 3000 rpm for 30 s, the density of SWCNTs depositedon the SiO2 substrate was measured to be 68.8 mm�2 in Fig. S5. TheSEM image (Fig. 1(b, i)) of the SWCNTs assembled on the SiO2substrate shows that the SWCNTs were dispersed uniformlywithout any aggregation. The SEM images (Fig. 1(b, ii and iii)) ofMoS2 without SWCNTs andMoS2-SWCNT hybrid films confirm thatthe MoS2 nanosheets were synthesized uniformly on the SiO2substrate and that the SWCNTs remained stable without any loss ofphysical properties during the CVD process for MoS2 growth. Also,AFM images show that MoS2-SWCNT hybrid sheets increased RMSroughness by SWCNTs morphology (Fig. S4 in SupportingInformation).
Raman spectroscopy was utilized to study the crystallinestructures of theMoS2 nanosheets and SWCNTs. Raman is a suitablemethod for identifying the structure of a CNT-based hybrid struc-ture [23]. We confirmed the state of MoS2 and SWCNTs in MoS2-SWCNT hybrid films through Raman measurement (Fig. S6). In thelow-energy region, the Raman spectrum of MoS2-SWCNT hybridfilms exhibits peaks at 404.709 cm�1 and 385.566 cm�1, corre-sponding the out-of-plane vibration mode (A1g mode) of sulfuratoms and the in-plane vibration mode (E12g mode) of molybde-num and sulfur atoms, respectively (Fig. 2(a)). This result indicatesthat the MoS2 nanosheets were successfully grown on the pro-moter. In addition, the number of MoS2 layers can be determinedfrom the energy difference between the A1g and E12g modes [24].Fig. 2(b) shows the energy difference of the A1g and E12g modes forMoS2-SWCNT hybrid films as a function of the number of SWCNTcoating cycles, which indicates that bilayer MoS2 nanosheets weresynthesized on the promoter layer with SWCNTs without anynotable change in the number of MoS2 layers. Raman mapping alsoconfirmed uniformity of the MoS2-SWCNT hybrid films. Ramanmaps of A1g mode prove excellent uniformity in the number oflayers over large areas, and mapping of the RBM peak indicates thatthe SWCNTs were stable during the CVD growth process (Fig. 2(c)).As uniform and high-performance MoS2-SWCNT hybrid films weresynthesized over a large area without the loss of electrical andphysical properties of TMDs and SWCNTs, we suggest that oursynthetic method is suitable for the fabrication of advanced flexibledevices.
MoS2-SWCNT-based field-effect transistors were prepared forthe analysis of electrical transport in MoS2-SWCNT hybrid layers asdepicted in Fig. 3(a). Here, Cr (7 nm)/Au (70 nm) were used as the
Fig. 1. (a) Schematic of the fabrication process of chemical sensors based on MoS2 nanosheets and SWCNTs: (i) spin-coating of SWCNTs on the SiO2 substrate, (ii) deposition of anorganic promoter layer using thermal evaporation process, (iii) growth of MoS2 nanosheets through CVD process, (iv) transfer of SWCNT and MoS2 hybrid films onto arbitrarysubstrates, and (v) fabrication of a flexible device based on MoS2-SWCNT hybrid films. (b) SEM images of (i) SWCNTs spin-coated on the SiO2 substrate, (ii) MoS2 grown on thepromoter layer without SWCNTs, and (iii) MoS2-SWCNT hybrid film.
Fig. 2. Raman spectra of hybrid films based on SWCNTs and MoS2. (a) Raman spectra of hybrid films based on SWCNTs and MoS2 from excitation at 514 nm: (i) Overview spectra, (ii)G-band peaks, and (iii) RBM peaks of SWCNTs and MoS2-SWCNT hybrid films. (b) Raman spectra of MoS2-SWCNT hybrid nanosheets with various SWCNT densities. (c) Differencebetween the A1g and E12g Raman mode of the MoS2-SWCNT hybrid nanosheets as a function of the number of SWCNT-coating cycles. (d) Raman A1g and RBM map of the hybridnanosheets.
S. Kim et al. / Carbon 129 (2018) 607e612 609
source/drain electrodes, heavily p-doped silicon was used as theback gate, and 300-nm SiO2 was used as the back gate dielectric[25] (inset of Fig. 3 a-(i)). The transfer curve (IDS-VG) of pristineMoS2 devices at VDS¼ 1 V exhibited a charge-neutral Dirac point ata near-zero gate voltage (VG) and an asymmetric hole and electronconduction, as shown in Fig. 3(a). In case of devices based on MoS2-SWCNT hybrid films, the on-state current increased and the voltageat the Dirac point showed a positive shift as the number of CNTcoating cycles increased, as seen in Fig. 3(a, ii-iv). These positiveshifts indicate that hole conduction is larger than electron con-duction and that p-type doping is formed. This suggests that theelectrical properties of hybrid films are dominant in SWCNTs as
shown in Fig. S7.MoS2-based devices with SWCNT densities of 0, 5.5, 15.5, 34.0,
and 68.8 mm�2 have resistances of 5.65� 107, 3.66� 107, 1.36� 106,and 6.45� 103U at VDS¼ 1 V and Dirac voltages of 7.2, 16.0, 24.0,and 58.4 V respectively. While the sheet conductance of MoS2-SWCNT hybrid films increased dramatically compared to pristineMoS2, it was still significantly lower than that of SWCNTs coated onthe transferred MoS2 layer; this is because of the existence ofSWCNT. XPS analysis of SWCNT, MoS2, and MoS2-SWCNT was car-ried out to identify the interface between SWCNT and MoS2 asshown in Fig. S8. In the case of C 1s, MoS2þSWCNT shifts to 0.2 eVtoward the high binging and 0.3 eV shifts to the high binging for
Fig. 3. (a) Transfer characteristics (IDS-VG) of the MoS2-SWCNT-based transistors obtained at VDS¼ 1 V: (i) pristine MoS2 nanosheets and (ii)-(iv) MoS2 hybrid film with variousSWCNT densities. (b) Resistance (i) and threshold voltage (ii) of MoS2-based transistors as a function of the number of SWCNTs coating cycles. (c) I-V curves of chemical sensorsbased on MoS2-SWCNT films under exposure to NO2 gas. (d) Gas sensing response for NO2 molecules at room temperature under various exposure conditions. I-V curve (e) and gassensing response (f) of chemical MoS2-based sensors for NH3 molecules at room temperature.
S. Kim et al. / Carbon 129 (2018) 607e612610
MoS2 compared to pristine SWCNT. This suggests that when MoS2is synthesized, electrons move from SWCNT to MoS2, and strongerbonds are formed so as not to lose electrons. Mo and S 2p peaks ofthe hybrid films showed that theMo and S peaks of the hybrid filmsshifted toward the low binding side by 0.3 eV compared with thoseof the pristine MoS2 films. This result indicated that it is n-dopingbecause it gets electrons from SWCNT.
High-sensitivity gas detection has recently become an essentialrequirement for preventing environmental pollution caused byNH3, NO2, fine dust etc., thus making it necessary to monitorpollutant emissions accompanying industrial or recycling pro-cesses. In previous reports, TMD-based layers were utilized forincreasing the sensitivity of sensors as they have a high surface-to-volume ratio and thus provide more active sites for the reactionwith specific gases [26]. In our study, MoS2-SWCNT hybrid filmswere used to improve the electrical conductance response forexposure NH3 or NO2 gases, owing to the excellent gas adsorptionability of SWCNTs. Exposure of NH3 causes the shift of the valenceband from the Fermi level of the SWCNT to lower hole depletionand conductance. In the case of NO2, the Fermi level approaching tothe valence band causes enriched hole carriers of the SWCNT andincreases the conductance of the sample [27].
A MoS2-SWCNT junction, 500 mm wide and 50 mm long, wasprepared between the source and drain electrodes for monitoring
the conductance change under exposure to specific gases. Fig. 3(c)and (e) show the I-V behavior of MoS2-SWCNT-based chemicalsensors under exposure to NO2 and NH3 gas, respectively. Theelectrical conductance of MoS2-SWCNT-based devices under NO2exposure was approximately twice that obtained with pristineMoS2 nanosheets. In case of 40 ppm NO2 detection, the electricalconductance increased by 156% (Fig. 3(c)), while the electricalconductance decreased by 23% in case of 40 ppm NH3 detection(Fig. 3(e)). The rate of change in NO2 is about 7 times greater thanthat of NH3. The relatively large rate of change in NO2 compared toNH3 is related to the binding energy of SWCNTs. According totheoretical numerical modeling, NO2, and NH3 molecules can easilybound to the surface of SWCNT or MoS2, and defects such as thecarboxylic group of SWCNT play a major role during binding pro-cess [28]. Adsorption energy for various gas molecules on SWCNThas been also reported [29]. Here, the NO2 adsorption energy onSWCNT surface is about 7 times larger than that of NH3 (NO2:797meV, NH3: 149meV). In case of NO2, the adsorption energy onMoS2 surface is about 2 times larger than that of NH3 [30]. Sensingbehavior of our devices based on MoS2 and SWCNTs was resultedfrom the difference in the electrical conductance by molecule ab-sorption on SWCNTs. This is consistent with the result of thetransfer curve (IDS -VG) in Fig. 3(a).
Fig. 3(d) and (f) show the gas-sensing response for NH3 and NO2
S. Kim et al. / Carbon 129 (2018) 607e612 611
molecules on MoS2-SWCNT films at room temperature, respec-tively. The gas sensitivity was calculated according to the followingequation:
DR/Ra ¼ (Rg� Ra)/Ra (1)
Where Ra and Rg are resistance values before and after the gas re-action, respectively. As a result, the gas-sensing response of theMoS2-SWCNT-based sensors was higher than that obtained withthe MoS2 layer without SWCNTs. On exposure to NO2 at a con-centration of 1.5 ppm, a gas response with negative sensitivity wasobserved, which increased as the NO2 concentration was increasedto 40 ppm. Further, on exposure to NH3 at a concentration of1.8 ppm, a gas response with positive sensitivity was observed,which decreased as the NH3 concentration was gradually increasedto 40 ppm. As a results, when gas sensors with MoS2 nanosheetswere replaced by those with MoS2-SWCNT films, the gas responseimproved by approximately 54 and 34.5% during exposure to40 ppm NO2 and NH3 gas, respectively. In case of exposure to40 ppm NO2, the gas response of the MoS2 filmwas 29% and that ofMoS2-SWCNT films was 54%. In case of exposure to 40 ppm NH3,the gas response of MoS2 film was 24%, and that of MoS2-SWCNTfilms was 34%. That is, owing to the presence of SWCNTs, there is adifference of approximately 88% in NO2, but approximately 40% inNH3. MoS2-SWCNT films showmore activity in the presence of NO2
than NH3. According to first-principles density functional theorycalculations, both NO2 and NH3 have negative adsorption energy.However, the adsorption energy of NO2 is about 0.02 eV higher thanthat of NH3, which seems to be the cause of the high gas response inNO2 [31]. In previous studies on the gas sensing by using MoS2nanosheets, the rate-changes of 2% (50 ppm) [27] and 25% (50 ppm)[31] of NO2 were reported, and we found that the rate of change ofour MoS2eSWCNT device by NO2 absorption was larger or similarwhen compared with similar experiments. Therefore, gas sensingproperties of our devices are believed to be improved due to thepresence of SWCNTs.
Fig. 4. (a) Optical images and bending test of MoS2-SWCNT hybrid films transferred on PETMoS2 films. (c) The resistance variation according to the number of bending cycles (bendinSWCNT hybrid films and MoS2 films after 105 bending cycles under various conditions of N
We also fabricated MoS2-SWCNT-based flexible chemical sen-sors using a polymethyl methacrylate (PMMA)-assisted wettransfer method [32]. Fig. 4(a) shows a photograph of MoS2-SWCNTnanosheets transferred onto a polyethylene terephthalate (PET)substrate. First, PMMA was spin-coated onto MoS2-SWCNT/SiO2,and the PMMA-coated substrate was placed in a 0.1M NaOHaqueous solution. After the SiO2 layer was completely etched away,the PMMA-coated hybrid nanosheet was transferred to a PET sub-strate. Finally, the PMMA layer was removed with acetone, and thehybrid nanosheet was rinsed with DI water. An ultraviolet-visiblespectrometer (UV-2501PC) was used (Fig. 4(b)) for measuring theoptical transmittance of MoS2-SWCNT hybrid films on PET. Thetransmittance of MoS2-SWCNT hybrid films is 92.83% in the visiblerange, similar to that of the CVD-grownMoS2 films. The opacity of asingle MoS2 layer is 96.06%, which indicates that our MoS2 nano-sheet is almost a monolayer. In future, the approach developed inthis study for fabricating MoS2 nanosheets could prove to be a keytechnology for building thin and flexible devices based on two-dimensional TMD nanosheets. The structure of the MoS2-SWCNThybrid films transferred to PET was almost similar to those grownon SiO2 substrates (Fig. S9).
We also evaluated the durability, electrical resistance, andsensing performance of the MoS2-based and MoS2-SWCNT-basedchemical sensors as a function of the number of bending cycles. Inour work, a bending tester (JUNIL TECH Co., LTD-JIBT-610-RadiusBending), with a bending radius of 5mm, was used (Fig. 4(a)).Fig. 4(c) shows that the resistance of MoS2-based sensors increasedto approximately 300% after 105 bending cycles, while the resis-tance of the MoS2-SWCNT sensors remained stable under the samebending process; this was presumably due to the superior me-chanical flexibility of SWCNTs as SWCNTs in the hybrid film play arole in minimizing the cracks or damage in the MoS2 layer. Inaddition, we carried out real-time sensor monitoring under variousgas exposures after 105 bending cycles to investigate the stability ofMoS2-based sensors. Interestingly, after 105 bending cycles, MoS2-SWCNT sensors exhibited a reduction of 18% and 14% in gas sensing
by PMMA-assisted method. (b) Optical transmittance of MoS2-SWCNT hybrid films andg radius¼ 5mm). (d) Time-dependent bias current, measured at VDS¼ 0.1 V, of MoS2-O2 and NH3 exposure.
S. Kim et al. / Carbon 129 (2018) 607e612612
during the monitoring of NO2 and NH3 respectively, which is lowersensing in comparison to the reduction observed for devices basedon MoS2 without SWCNTs (Fig. 4(d)). These results show that thestability of the sheet resistance and gas sensor performance ofMoS2-SWCNT films are significantly better compared to MoS2 filmswithout SWCNTs; this suggests that the reduction of cracks anddamage is an important factor to overcome the extreme situationsof repeated bending. As a significant outcome, this approach maypave the way toward the fabrication of flexible sensors using TMD-based sheets combined with SWCNTs, owing to the excellent flex-ibility of TMDs sheets and SWCNTs.
4. Conclusion
In summary, a carbon material (SWCNT) and a TMD nanosheet(MoS2 layer) were combined for fabricating high-performanceflexible chemical sensors. SWCNTs increased the contact bindingbetween MoS2 nanosheets and SWCNTs, and TCVD was subse-quently carried out for the synthesis of large-areaMoS2 nanosheetsusing an organic promoter layer. As proof of the theoretical con-cepts, we demonstrated high-performance flexible chemical sen-sors based onMoS2-SWCNT hybrid films. Notably, chemical sensorsbased on the hybrid filmswere stable and exhibited a low reductionin gas sensing after 105 bending cycles.
Acknowledgments
This research was supported by a grant (2011-0031636) fromthe Center for Advanced Soft Electronics under the Global FrontierResearch Program of the Ministry of Science, ICT and Future Plan-ning, Korea.
Appendix A. Supplementary data
Supplementary data related to this article can be found athttps://doi.org/10.1016/j.carbon.2017.12.065.
References
[1] L.F. Mattheis, Phys. Rev. B 8 (1973) 3719.[2] M. Chhowalla, H.S. Shin, G. Eda, L.-J. Li, K.P. Loh, H. Zhang, Nat. Chem. 5 (2013)
263.[3] 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, V. Nicolosi, Science331 (2011) 568.
[4] M. Park, Y.J. Park, X. Chen, Y.-K. Park, M.S. Kim, J.-H. Ahn, Adv. Mater. 28(2016) 2556.
[5] B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti, A. Kis, Nat. Nanotechnol.6 (2011) 147.
[6] D.j. Late, Y.-K. Huang, B. Liu, J. Acharya, S.N. Shirodkar, J. Luo, A. Yan, D. Charles,U.V. Waghmare, V.P. Dravid, C.N.R. Rao, ACS Nano 7 (2013) 4879.
[7] W. Wu, L. Wang, Y. Li, F. Zhang, L. Lin, D. Chenet, X. Zhang, Y. Hao, T.F. Heinz,J. Hone, Z.L. Wang, Nature 514 (2014) 470.
[8] 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, L.-J. Li, Nano Lett. 12 (2012) 1538.
[9] S.J. Kim, M.-A. Kang, S.H. Kim, Y. Lee, W. Song, S. Myung, S.S. Lee, J. Lim, K.-S. An, Sci. Rep. 6 (2016) 24054.
[10] Y.R. Lim, W. Song, J.K. Han, Y.B. Lee, S.J. Kim, S. Myung, S.S. Lee, K.-S. An,C.J. Choi, J. Lim, Adv. Mater. 28 (2016) 5025.
[11] X. Ling, Y.-H. Lee, Y. Lin, W. Fang, L. Yu, M.S. Dresselhaus, J. Kong, Nano Lett. 14(2014) 464.
[12] Y. Niu, R. Wang, W. Jiao, G. Ding, L. Hao, F. Yang, X. He, Carbon 95 (2015) 34.[13] S.S. Varghese, S. Lonkar, K.K. Singh, S. Swaminathan, A. Abdala, Sensor.
Actuator. B Chem. 218 (2015) 160.[14] Y. Niu, W. Jiao, R. Wang, G. Ding, Y. Huang, J. Mater. Chem. A 4 (2016) 8198.[15] H. Long, A. Harley-Trochimczyk, T. Pham, Z. Tang, T. Shi, A. Zettl, C. Carraro,
M.A. Worsley, R. Maboudian, Adv. Funct. Mater. 26 (2016) 5158.[16] Q. He, Z. Zeng, Z. Yin, H. Li, S. Wu, X. Huang, H. Zhang, Small 19 (2012) 2994.[17] N.L.W. Septiani, B. Yuliarto, J. Electrochem. Soc. 163 (2016) B97.[18] B.-Y. Wei, M.-C. Hsu, P.-G. Su, H.-M. Lin, R.-J. Wu, H.-J. Lai, Sens. Actuators B
101 (2004) 81.[19] K.H. An, S.Y. Jeong, H.R. Hwang, Y.H. Lee, Adv. Mater. 12 (2004) 1005.[20] F. Mendoza, D.M. Hernandez, V. Makarov, E. Febus, B.R. Weiner, G. morell,
Sens. Actuators B 190 (2014) 227.[21] C. Bittencourt, A. Felten, E.H. Espinosa, R. Ionescu, E. Llobet, X. Correig, J.-
J. Pireaux, Sens. Actuators B 115 (2006) 33.[22] Y. Lee, J. Lee, H. Bark, I.-K. Oh, G.H. Ryu, Z. Lee, H. Kim, J.H. Cho, J.-H. Ahn,
C. Lee, Nanoscale 6 (2014) 2821.[23] S.H. Kim, W. Song, M.W. Jung, M.-A. Kang, K. Kim, S.-J. Chang, S.S. Lee, J. Lim,
J. Hwang, S. Myung, K.-S. An, Adv. Mater. 26 (2014) 4247.[24] Q.H. Wang, K. Kalantar-Zadeh, A. Kis, J.N. Coleman, M.S. Strano, Nat. Nano-
(2012) 5635.[26] S.-Y. Cho, S.J. Kim, Y. Lee, J.-S. Kim, W.-B. Jung, H.-W. Yoo, J. Kim, H.-T. Jumg,
ACS Nano 9 (2015) 9314.[27] J. Kong, N.R. Franklin, C. Zhou, M.G. Chapline, S. Peng, K. Cho, H. Dai, Science
287 (2000) 622.[28] J.A. Robinson, E.S. Snow, S.C. Badescu, T.L. Reinecke, F.K. Perkins, Nano Lett. 6
(2006) 747.[29] G. Alcala, P. Skeldon, G.E. Thompson, A.B. Mann, H. Habazaki, K. Shimizu,
Nanotechnology 13 (2002) 451.[30] L. Kou, A. Du, C. Chen, T. Frauenheim, Nanoscale 6 (2014) 5156.[31] B. Cho, M.G. Hahm, M. Choi, J. Yoon, A.R. Kim, Y.-J. Lee, S.G. Park, J.-D. Kwon,
C.S. Kim, M. Song, Y. Jeong, K.-S. Nam, S. Lee, T.J. Yoo, C.G. Kang, B.H. Lee,H.C. Ko, P.M. Aiayan, D.H. Kim, Sci. Rep. 5 (2015) 8052.
[32] X. Li, Y. Zhu, W. Cai, M. Borysiak, B. Han, D. Chen, R.D. Piner, L. Colombo,R.S. Ruoff, Nano Lett. 9 (2009) 4359.
