lable at ScienceDirect

Current Applied Physics 14 (2014) 264e268

Contents lists avai

Current Applied Physics

journal homepage: www.elsevier .com/locate/cap

Controlled exfoliation of molybdenum disulfide for developing thinfilm humidity sensor

Shao-Lin Zhang a, Hyang-Hee Choi b, Hong-Yan Yue c, Woo-Chul Yang a,*

aDepartment of Physics, Dongguk University, Seoul 100-715, Republic of Koreab Institute of Nanoscience and Nanotechnology, Yonsei University, Seoul 120-749, Republic of Koreac School of Material Science and Engineering, Harbin University of Science and Technology, Harbin 150-040, China

a r t i c l e i n f o

Article history:Received 15 August 2013Received in revised form7 November 2013Accepted 21 November 2013Available online 2 December 2013

Keywords:Molybdenum disulfideLiquid exfoliationUltrasonic powerDynamic light scatteringHumidity sensor

* Corresponding author. Tel.: þ82 2 2290 1397; faxE-mail address: wyang@dongguk.edu (W.-C. Yang

1567-1739/$ e see front matter � 2013 Elsevier B.V.http://dx.doi.org/10.1016/j.cap.2013.11.031

a b s t r a c t

We report a facile, size-controllable exfoliation process using an ultrasound-assisted liquid method tofabricate few-layer molybdenum disulfide (MoS2) nanosheets. The morphology, structure and size dis-tribution of the nanosheets processed with different ultrasonic powers were examined by atomic forcemicroscopy, Raman spectroscopy and dynamic light scattering. It was revealed that the size of nano-sheets reduces and final yield increases with elevating ultrasonic power. Bulk and exfoliated MoS2 basedthin film sensors are fabricated by a simple drop casting method on alumina substrates. Our sensorsexhibit excellent sensitivity with very quick response and recovery speed to humidity gas. Comparativestudies are carried out to draw up the size or ultrasonic power dependent sensing behavior.

� 2013 Elsevier B.V. All rights reserved.

1. Introduction

Over the last few years, two-dimensional (2D) nanomaterials,such as graphene, tungsten disulfide (WS2), and molybdenumdisulfide (MoS2), have become the hottest research issue, becausethey exhibit unusual electronic, optical, thermal, and mechanicalproperties, which are introduced by the quantum size effectassociated with their ultra-thin structure [1e4]. These exceptionalphysical properties enable them to be the crucial elements for arange of applications, such as monolithic graphene circuits [5],high-performance MoS2-based FET devices [6], and photo-transistors [7]. Two-dimensional (2D) nanomaterials also possessinherently high surface-to-volume ratio, which make them be adominant candidate for gas sensor applications [8,9]. The 2D na-ture of these materials allows a total exposure of all their atoms,and the electron transport through these exposed surface atoms ishighly sensitive to the absorbed molecular species. Moreover, theintrinsic low electric noise, due to the quality of their crystal latticeand 2D nature, tends to screen charge fluctuations more than thatof other nanomaterials [10]. These distinct characteristics of 2Dnanomaterials strongly suggest their potential for chemical cata-lyst and gas sensor utilization [11]. A recent study showed that a

: þ82 2 2260 8713.).

All rights reserved.

graphene-based gas sensor exhibited an excellent sensing prop-erty that a detection limit down to the single-molecule level wasachieved [12].

As the analog of graphene, MoS2 with semiconducting character,together with lower background carrier densities, is thus expectedto display a comparative or preferable sensing performance. Severalprevious researches have demonstrated the promising sensingproperties of mono- and multi-layer MoS2 nanosheets towardvarious gases at room temperature. F. Perkins et al. fabricated singlemonolayer MoS2 by the mechanical cleavage method, and theirsensor exhibited highly selective reactivity to triethylamine amonga range of analytes [13]. In another group, Hai Li et al. developedmono- and multi-layer MoS2 film based field-effect transistorsusing a similar exfoliation method. Their MoS2 nanosheets havesuccessfully been used for sensing NO gas at room temperature,with rapid and dramatic response [14]. These results fully provedthat MoS2 nanosheets have potential for gas detection. Unfortu-nately, the fabrication methods used in these previous studiesinvolve manual mechanical exfoliation, which is unsuited to massproduction, and decreases the reproducibility of the electronicdevice. Moreover, the complicated and expensive e-beam lithog-raphy process utilized in these experiments also hinders theirpractical application. A facile method to efficiently exfoliate MoS2,and economical process to fabricate sensor devices with high reli-ability are urgently required.

S.-L. Zhang et al. / Current Applied Physics 14 (2014) 264e268 265

Bulk MoS2 consists of hexagonal layers of molybdenum (Mo)sandwiched between two layers of sulfide (S) atoms. The bondingwithin these trilayer sheets is covalent, while the adjacent sheetsare stacked viaweak van derWaals interaction. Theweak interlayerbonding allows the bulk MoS2 to be exfoliated by proper externalforce. Up to now, besides the first adopted mechanical cleavagemethod [6,15], several methods have been developed to exfoliateMoS2 bulk material, including chemical ion intercalation [16], andultrasound-assistant liquid exfoliation [17,18]. Compared to thetime-consuming and tough environment required for the ionintercalation method, the ultrasound-assistant liquid methodpossesses advantages, such as low-cost, ease of process, and po-tential for scale-up.

In this study, utilizing the ultrasound-assistant exfoliationmethod, we fabricated MoS2 nanosheets in large scale in a sol-vent mixture with low boiling point. The effects of ultrasonicpower on the size distribution and total yield of the final pro-duction were systematically examined. The successfully preparedMoS2 nanosheets were fabricated into thin film sensors by afacile drop casting method. The thin film sensors were tested tomoisture gases with relative humidity (RH) from 0 to 60%. It wasfound MoS2 nanosheets based sensors exhibited excellentsensing performance towards humidity gas with very quickresponse and recovery. Moreover, the nanosheets fabricated withhigher ultrasonic power exhibited better sensing properties thanthose with lower ultrasonic power. The size-dependent sensingperformance of MoS2 nanosheets based thin film sensors wascarefully investigated. A reasonable sensing mechanism wasproposed.

2. Experimental details

2.1. Ultrasonic power dependent exfoliation of MoS2

All chemical regents were of analytical grade and were ob-tained commercially. Typically, MoS2 powder (300 mg, 2 um,AldricheSigma) was dispersed in 100 ml of a mixture of chloro-form and acetonitrile with the ratio of 65:35. Then the dispersionwas ultrasonically treated in iced water for 1 h using a horn probesonic tip (Sonic VCX 750). Pulsed ultrasonic irradiation was per-formed for 20 s on and 10 s off to avoid damage to the ultrasonicprocessor, and reduce solvent heating and resultant degradation ofthe MoS2 nanosheets. Various ultrasonic powers ranging from350 W to 550 W were adopted to explore the ultrasonic powerdependent effect on the final production. After ultrasonic treat-ment, the dark green dispersion was centrifuged at 3000 rpm for30 min and the supernatant was collected. This process wasrepeated one time more to completely remove the un-exfoliatedpowder and aggregates. For the characterization of exfoliatedMoS2, a drop of solution containing the produced MoS2 nano-material was placed on a Si/SiO2 substrate, and then dried in airbefore it was observed by atomic force microscopy (AFM, BrukerNano N8 NEOS) and Raman spectroscopy (HORIBA HR800). Inorder to examine the size evolution of MoS2 nanosheets exfoliatedwith different ultrasonic power, as-prepared MoS2 nanosheetswere also investigated by dynamic light scattering (DLS). The DLSmeasurements were carried out by Malvern Zetasizer Nano ZSusing a 633 nm HeNe laser. Samples were tested in quartz cuvetteshaving a 10 mm path length. The DLS apparatus was operated inbackscatter mode at an angle of 173�. Before the measurement, thesamples were equilibrated to 25 �C for 120 s. Values for solventviscosity at 25 �C, as provided by the solvent suppliers, wereentered into the software. An automatic measurement durationsetting was used with automatic measurement positioning andautomatic attenuation.

2.2. Fabrication and measurement of MoS2 nanosheets based thinfilm sensor

Alumina substrates (4 mm � 4 mm) with interdigital Pt elec-trodes were pre-prepared and used as the sensor substrate. Thealumina substrate were placed on a hot-plate with a temperature ofabout 60 �C. MoS2 nanosheets solutions were then drop-cast ontothe sensor substrates with micro-pipettes. The thickness of MoS2thin films was about 1 um which was controlled by the volume ofsolution in themicro-pipettes. After coating, the alumina substrateswere heated at 80 �C for 1 h to eliminate the remaining solvent, andthen sintered at 400 �C in Ar gas for 1 h to improve the adhesionand contact. The sensing performance of the MoS2 nanosheetsbased thin film sensor towards humidity was tested. A source ofhumidity gas was created by a bubbling system with water bath.The humidity gas was then mixed with synthetic dry air with amixing system equipped with mass flow controllers and mass flowmeters. The desired humidity concentration was accurately tunedby the flow ratio of the humidity gas and dry air. The total flow wasconstantly set to 1000 sccm. A commercial humidity sensor wasused to validate the actual relative humidity. The sensitivity isdefined as follows:

S ¼ RHRD

; (1)

where, RD and RH represent the resistance of the sensor uponexposure to a dry air and a humidity gas, respectively. The responseand recovery time are defined as the time taken by the sensor toachieve 90% of the total resistance change in the respective cases ofadsorption and desorption.

3. Results and discussion

3.1. Characterization of exfoliated MoS2 nanosheets

There are two main parameters affecting the quality of finalproduction in ultrasound-assistant liquid exfoliation. These aresolvent type and ultrasonic dosing. Hitherto, most of the formerresearches have focused on the selection of solvents to improvethe dispersion of MoS2 in solution, while very few studies havebeen carried out on the effect of ultrasonic dosing. It is believedthat the parameters of the ultrasonic process also play a key role inthe exfoliation process. Herein, we demonstrate the ultrasonicpower dependent morphology evolution of MoS2 nanosheets.Fig. 1(a) and (b) shows the typical AFM images of MoS2 nanosheetsobtained by ultrasonic power of 350 W and 550 W, respectively. Itwas found that the average size of nanosheets decreased as theultrasonic dosing increased. The quantitative characterizationmade from the profile line in Figs. (a0) and (b0) further confirm thistendency. The average size of nanosheets obtained by 350 W ul-trasonic treatment was about 200 nm, which is almost 2 timescompared to that obtained by 550 W process. On the other hand,the thickness of both exfoliated MoS2 nanosheets are similarranging from 6 nm to 12 nm. Given that the thickness of a MoS2monolayer is about 0.9e1.2 nm [10], it suggests that the obtainedMoS2 existed as few-layer nanomaterials. This was also evidencedby the Raman spectroscopy, as shown in Fig. 1(c). A typical few-layer MoS2 nanosheet gave bands at 386.54 and 407.69 cm�1,which are associated with the in-plane vibrational (E2g1 ), and out-of-plane vibrational (A1g) modes, respectively. Compared to thecorresponding bulk sample, a slight red-shift of the E2g1 band andblue-shift of the A1g band in few-layer MoS2 nanosheets wereobserved, which is consistent with previous reports [14,19].

Fig. 2. Intensity particle size distribution for MoS2 nanosheets exfoliated withdifferent ultrasonic power.

Fig. 1. Typical AFM images of MoS2 after ultrasonic treatment with (a) 350 W and (b)550 W; images (a0) and (b0) are the height profiles along the white lines shown in (a)and (b), respectively; and (c) typical Raman spectra of exfoliated (top) and bulk(bottom) MoS2.

S.-L. Zhang et al. / Current Applied Physics 14 (2014) 264e268266

To gain more evidence about the ultrasonic power dependentsize evolution of exfoliated MoS2 nanosheets, further character-ization was conducted. Dynamic light scattering (DLS, also knownas photon correlation spectroscopy or quasi-elastic light scat-tering) is a technique that can be used to determine the size dis-tribution profile of small particles in liquid suspension. Theparticles in a liquid suspension at a certain temperature undergoBrownian motion. The DLS instrument monitors this Brownianmotion of particles, by applying a laser light. It records the spatialintensity distribution of light scattered by a given sample as afunction of time. The intensity distribution constantly fluctuates asparticles diffuse through the suspension. By measuring the auto-correlation of the intensity distribution as a function of time, thetranslational diffusion coefficient of the particles can be calculated.For a spherical particle, the translational diffusion coefficient andits hydrodynamic diameter are related by the StokeseEinsteinequation:

D ¼ kT6phR

(2)

where, the D, k, T, h, and R represent the translational diffusioncoefficient, Boltzmann’s constant, absolute temperature, liquidviscosity, and hydrodynamic radius of the particle, respectively.

The intensity particle size distributions obtained by DLS forMoS2 nanosheets prepared with different ultrasonic power aregiven in Fig. 2. The typical distribution is characterized by a bell-shaped curve centered on a single peak value. It was observedthat the primary peak position of the three samples had a slightshift. As shown in the inset of Fig. 2, with the increase of ultrasonic

power, the peak position shifted from 68 nm to 51 nm. It is impliedthat elevating the ultrasonic dosing resulted in the size reduction ofthe products. It is worthwhile to note that the peak size here rep-resents the size of a sphere with volume equal to the volume of theparticles. Considering the actual sheet-shape of our products,where the thickness is relatively small, a slight shift on peak posi-tion may imply a huge change in nanosheet diameter. It is alsonoted that the distribution of the sample obtained with ultrasonicpower of 350 W displayed a double-peak. In addition to the pri-mary peak, a small peak at around 6 um was observed. This couldbe attributed to the small dust particles or air bubbles in thesuspension.

Ultrasound brings two effects into the exfoliation process: thecavitation effect andmicro-jet effect [20]. Cavitation generates highenergy in the solution via the process of acoustic cavitation: theformation, growth and implosive collapse of bubbles. The local hightemperature and pressure created by the collapse of bubble is themain driving force for the exfoliation of MoS2. When the cavitationoccurs at the interface of liquid and solid, for example the interfaceof solvent and MoS2 nanosheet, the collapse of bubble turns tomicro-jet towards solid surface, leading to a tearing effect to theMoS2 nanosheet. The exfoliation effect and tearing effect aresimultaneously improved when the ultrasonic dosing is increased,resulting in a reduction of nanosheets size.

The yields of MoS2 nanosheets fabricated with different ultra-sonic powerwere estimated. As shown in Fig. 3(a), it was found thatthe production of MoS2 nanosheets increased with the increase ofultrasonic power. The highest production reached about 0.43 mg/ml obtained by the highest ultrasonic dosing, which is approxi-mately 20 times higher than that in previous research [18]. Fig. 3(b)shows a red column appeared when the laser light passed throughthe diluted MoS2 solution. This phenomenon is known as theTyndall effect. The well-defined Tyndall effect indicated the pres-ence of ultrathin MoS2 nanomaterial in the solution. This MoS2nanosheet solution was also highly stable, as shown in Fig. 3(c), inthat no precipitation was observed upon standing for more thanone week under ambient condition.

3.2. Humidity gas sensing properties of the MoS2 based thin filmsensor

Researches of MoS2 nanosheets as a gas sensor application havemostly focused on the detection of toxic gas or volatile organic gas[13,14]. Recently, it was reported that transition metal dichalcoge-nides (TMDs), such as vanadium disulfide (VS2) nanosheets,

Fig. 3. (a) Production of MoS2 nanosheets fabricated with different ultrasonic power; (b) Tyndall effect appears in diluted MoS2 nanosheets solution; and (c) photographs of MoS2nanosheets after standing for more than two weeks under ambient condition.

Fig. 4. Typical single-cycle response of the bulk and exfoliated MoS2 based thin filmsensors to humidity gas with RH value about 60%.

Fig. 5. Dynamic response of the bulk and exfoliated MoS2 based thin film sensors tohumidity gases with alternating RH value from 10% to 60%.

S.-L. Zhang et al. / Current Applied Physics 14 (2014) 264e268 267

exhibited an excellent sensing response toward humidity gasowing to their large specific surface area [21]. The study on MoS2,which is the homologue of VS2, as a humidity gas sensor is therebyinspired. Fig. 4 displays a typical single-cycle response of the bulkand exfoliated MoS2 based thin film sensors to humidity gases withRH value of about 60%. It was observed that the resistance of theMoS2 sensor increased rapidly once upon exposure of humidity gas.When the humidity gas was shut off, the resistance of the sensorrecovered 100% to the original point. The sharp shift of the plotcurve throughout the sensing test suggested that the MoS2 sensorexhibited a very sensitive response toward humidity. The responseand recovery time were estimated to be about 9 s and 17 s,respectively. Compared to the previously reported SnO2 nanowireand VS2 nanosheets based humidity sensors [21,22], our sensorsmarkedly exhibit a faster response. The excellent response speed ofthe MoS2 sensor to humidity gas is partially attributed to itsintrinsic hydrophobic property [23,24], which accelerates thedesorption process of water molecules on the surface. It was alsoobserved that the response of exfoliated MoS2 sensors well sur-passed that of the bulk MoS2 sensor. This can be explained by theincreased surface area of theMoS2 nanosheets after exfoliation. Thegreater surface area leading to more absorption of water moleculesbrought a larger mutation in the electrical conduction of the MoS2sensor. It is worth emphasizing that the responses of exfoliatedMoS2 nanosheets based sensors showed an ultrasonic powerdependent tendency, in that the responses increased as the ultra-sonic power increased.

It is well known that MoS2 is composed by stacking covalentlybonded SeMoeS layers by Van der Waals interactions. In eachlayer, the atoms can be divided into basal plane sites and edge sites.The basal plane terminated by sulfur atoms with lone-pair elec-trons, which lacks dangling bonds, presents an inactive property[25]. On the other hand, the edge sites terminated by either mo-lybdenum or sulfur atoms missing coordination bonds tends to bemore active [11]. When the MoS2 is exfoliated into few-layernanosheets, it gives rise to a large number of low-coordinationstep-edges, kinks and corner atoms. These exposed edge sitesplay a dominant role in the gas sensing behavior.

Generally, there are two ways to increase the density of activeedge sites in nanomaterials. One is shape control. Nanomaterialswith a special three-dimensional network, for example, foam-likenanostructure, possess a large number of edge sites favoringsensing performance [26]. Another method is size control. Thedensity of edge sites increases as the size decreases. In our

experiment, we controlled the size of MoS2 nanosheets by varyingthe ultrasonic power. Rather than the low sonication power(w285 W) adopted in other reports [27], relatively high sonicationpower (350e550 W) was used in our experiment. The high ultra-sonic dosing effectively boosts the exfoliation effect and tearingeffect during the exfoliation process. Furthermore, the higher

Fig. 6. Repeated response of the bulk and exfoliated MoS2 based thin film sensors tohumidity gases with RH value of (a) 10%, and (b) 60%.

S.-L. Zhang et al. / Current Applied Physics 14 (2014) 264e268268

ultrasonic power results in a smaller size of MoS2 nanosheets. Thus,the sensing performance of exfoliated MoS2 with higher ultrasonicpower surpassed that obtained with lower power.

Fig. 5 shows the dynamic response of bulk and exfoliated MoS2based thin film sensors toward humidity gases at concentrationsranging from 0% to 60% RH. The sensors responded quickly uponexposure to humidity gas. After removing the humidity gas, theresponse of the sensors recovered 100%. The sensors respondedwell to the reiterative changes of humidity concentration. Thereproducibility of bulk and exfoliated MoS2 based thin film sensorswas also investigated. Fig. 6 shows the repeated response of thesensors toward humidity gases with RH of 10% and 60%. The testswere repeated over five times. No variance in response wasobserved, revealing that the MoS2 nanosheet based thin film sen-sors have excellent reproducibility.

4. Conclusion

In summary, we have successfully fabricated few-layer MoS2nanosheets in large scale by a simple but effective ultrasound-assisted liquid method. It was found that the quality of MoS2nanosheets is highly dependent on the ultrasonic power. As theultrasonic power increased, the average size of nanosheetsdecreased while the yield of products increased. The highestproduction reached 0.43 mg/ml. The MoS2 nanosheets based thinfilm gas sensors were fabricated by a facile drop casting method.The MoS2 nanosheets based thin film sensor exhibited a veryquick response (w9 s) and recovery (w17 s) which well sur-passed recently reported humidity sensors. The sensors showed a

size-dependent performance, in that the nanosheets with smallersize exhibited a better response toward humidity gas. Theexcellent sensing performance of MoS2 nanosheets was attrib-uted to the high density of edge sites, which were effectivelyinduced by enhanced ultrasonic treatment.

Acknowledgment

This research was supported by Leading Foreign ResearchInstitute Recruitment Program through the National ResearchFoundation of Korea (NRF) funded by the Ministry of Education,Science and Technology (MEST) (no. 2012-00109).

References

[1] T. Yu, B. Lim, Y. Xia, Angew. Chem. Int. Ed. 49 (2010) 4484.[2] C. Wang, Y. Zhou, M. Ge, X. Xu, Z. Zhang, J.Z. Jiang, J. Am. Chem. Soc. 132

(2009) 46.[3] T. Yu, B. Lim, Y. Xia, Angew. Chem. 122 (2010) 4586.[4] M. Osada, T. Sasaki, J. Mater. Chem. 19 (2009) 2503.[5] Y. Lin, A. Valdes-Garcia, S. Han, D.B. Farmer, I. Meric, Y. Sun, Y. Wu,

C. Dimitrakopoulos, A. Grill, P. Avouris, K.A. Jenkins, Science 332 (2011)1294.

[6] B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti, A. Kis, Nat. Nanotechnol.6 (2011) 147.

[7] Z. Yin, H. Li, H. Li, L. Jiang, Y. Shi, Y. Sun, G. Lu, Q. Zhang, X. Chen, H. Zhang, ACSNano 6 (2012) 74.

[8] G. Ko, H.Y. Kim, J. Ahn, Y.M. Park, K.Y. Lee, J. Kim, Curr. Appl. Phys. 10 (2010)1002.

[9] S. Hwang, J. Lim, H. Park, W. Kim, D. Kim, I. Song, J. Kim, S. Lee, D. Woo, S. Jun,Curr. Appl. Phys. 12 (2012) 1017.

[10] E. Massera, V. Ferrara, M. Migietta, T. Polichetti, I. Nasti, G. Francia, Chem.Today 29 (2011) 39.

[11] M. Chhowalla, H.S. Shin, G. Eda, L. Li, K. Loh, H. Zhang, Nat. Chem. 5 (2013)263.

[12] F. Schedin, A.K. Geim, S.V. Morozov, E.W. Hill, P. Blake, M.I. Katsnelson,K.S. Novoselov, Nat. Mater. 6 (2007) 652.

[13] F.K. Perkins, A.L. Friedman, E. Cobas, P.M. Campbell, G.G. Jernigan, B.T. Jonker,Nano Lett. 13 (2013) 668.

[14] H. Li, Z. Yin, Q. He, H. Li, X. Huang, G. Lu, D. Wen, H. Fam, A. Iing, Y. Tok,Q. Zhang, H. Zhang, Small 8 (2012) 63.

[15] A.K. Geim, K.S. Novoselov, Nat. Mater. 6 (2007) 183.[16] A. O’Neill, U. Khan, J. Coleman, Chem. Mater. 24 (2012) 2414.[17] M. Yi, Z. Shen, X. Zhang, S. Ma, J. Phys. D 46 (2013) 025301.[18] K. Zhou, N. Mao, H. Wang, Y. Peng, H. Zhang, Angew. Chem. Int. Ed. 50 (2011)

10839.[19] C. Lee, H. Yan, L.E. Brus, T.F. Heinz, J. Hone, S. Ryu, ACS Nano 4 (2010) 2695.[20] K.S. Suslick, Science 247 (1990) 1439.[21] J. Feng, L. Peng, C. Wu, X. Sun, S. Hu, C. Lin, J. Dai, J. Yang, Y. Xie, Adv. Mater. 24

(2012) 1969.[22] Q. Kuang, C. Lao, Z.L. Wang, Z. Xie, L. Zheng, J. Am. Chem. Soc. 129 (2007) 6070.[23] M. Tahir, N. Zink, M. Eberhardt, H. Therese, U. Kolb, P. Theato, W. Tremel,

Angew. Chem. Int. Ed. 45 (2006) 4809.[24] H. Seo, Y. Lee, J. Kor. Phys. Soc. 59 (2011) 730.[25] Q. Wang, K. Kalantar-Zadeh, A. Kis, J.N. Coleman, M.S. Strano, Nat. Nano-

technol. 7 (2012) 699.[26] F. Yavari, Z. Chen, A.V. Thomas, W. Ren, H. Cheng, N. Koratkar, Sci. Rep. 1

(2011) 1.[27] J.N. Coleman, M. Lotya, A. O’Neill, S.D. Bergin, P.J. King, U. Khan, K. Young, et

al., Science 331 (2011) 568.