# Fileset

[Jain_2018_Nanotechnology_29_265203.pdf](https://mdr.nims.go.jp/filesets/2b5c661b-a7da-4083-a9ff-d60c7fc78cca/download)

## Creator

Achint Jain, Palash Bharadwaj, Sebastian Heeg, Markus Parzefall, [Takashi Taniguchi](https://orcid.org/0000-0002-1467-3105), [Kenji Watanabe](https://orcid.org/0000-0003-3701-8119), Lukas Novotny

## Rights



## Other metadata

[Minimizing residues and strain in 2D materials transferred from PDMS](https://mdr.nims.go.jp/datasets/3880e115-c3e7-4b6f-9162-cb49711d82c6)

## Fulltext

Minimizing residues and strain in 2D materials transferred from PDMSNanotechnologyPAPER • OPEN ACCESSMinimizing residues and strain in 2D materialstransferred from PDMSTo cite this article: Achint Jain et al 2018 Nanotechnology 29 265203 View the article online for updates and enhancements.Related contentSimple fabrication of air-stable blackphosphorus heterostructures with large-area hBN sheets grown by chemical vapordeposition methodSapna Sinha, Yuya Takabayashi, HisanoriShinohara et al.-Enhanced dielectric deposition on single-layer MoS2 with low damage using remoteN2 plasma treatmentQingkai Qian, Zhaofu Zhang, MengyuanHua et al.-Effective characterization of polymerresidues on two-dimensional materials byRaman spectroscopyJi-Hoon Park, Soo Ho Choi, Won Uk Chaeet al.-This content was downloaded from IP address 144.213.253.16 on 25/06/2018 at 10:51https://doi.org/10.1088/1361-6528/aabd90http://iopscience.iop.org/article/10.1088/2053-1583/3/3/035010http://iopscience.iop.org/article/10.1088/2053-1583/3/3/035010http://iopscience.iop.org/article/10.1088/2053-1583/3/3/035010http://iopscience.iop.org/article/10.1088/2053-1583/3/3/035010http://iopscience.iop.org/article/10.1088/1361-6528/aa6756http://iopscience.iop.org/article/10.1088/1361-6528/aa6756http://iopscience.iop.org/article/10.1088/1361-6528/aa6756http://iopscience.iop.org/article/10.1088/0957-4484/26/48/485701http://iopscience.iop.org/article/10.1088/0957-4484/26/48/485701http://iopscience.iop.org/article/10.1088/0957-4484/26/48/485701http://oas.iop.org/5c/iopscience.iop.org/509731966/Middle/IOPP/IOPs-Mid-NANO-pdf/IOPs-Mid-NANO-pdf.jpg/1?Minimizing residues and strain in 2Dmaterials transferred from PDMSAchint Jain1 , Palash Bharadwaj2 , Sebastian Heeg1 ,Markus Parzefall1 , Takashi Taniguchi3, Kenji Watanabe3and Lukas Novotny11 Photonics Laboratory, ETH Zürich, 8093 Zürich, Switzerland2Department of Electrical and Computer Engineering, Rice University, Houston, TX 77005, United Statesof America3National Institute for Material Science, 1-1 Namiki, Tsukuba 305-0044, JapanE-mail: lnovotny@ethz.chReceived 9 February 2018, revised 29 March 2018Accepted for publication 12 April 2018Published 3 May 2018AbstractIntegrating layered two-dimensional (2D) materials into 3D heterostructures offers opportunitiesfor novel material functionalities and applications in electronics and photonics. In order to buildthe highest quality heterostructures, it is crucial to preserve the cleanliness and morphology of2D material surfaces that come in contact with polymers such as PDMS during transfer. Here wereport that substantial residues and up to ∼0.22% compressive strain can be present inmonolayer MoS2 transferred using PDMS. We show that a UV-ozone pre-cleaning of the PDMSsurface before exfoliation significantly reduces organic residues on transferred MoS2 flakes. Anadditional 200 ◦C vacuum anneal after transfer efficiently removes interfacial bubbles andwrinkles as well as accumulated strain, thereby restoring the surface morphology of transferredflakes to their native state. Our recipe is important for building clean heterostructures of 2Dmaterials and increasing the reproducibility and reliability of devices based on them.Supplementary material for this article is available onlineKeywords: TMDCs, van der Waals heterostructures, transfer techniques, material synthesisstrain(Some figures may appear in colour only in the online journal)1. IntroductionTwo-dimensional (2D) materials like graphene, hexagonal boronnitride (hBN), transition metal dichalcogenides (TMDCs), etc.,have gained immense attention in recent years due to the wealthof novel fundamental properties and fascinating physical phe-nomena exhibited by them [1, 2]. A unique possibility existingwith these materials is that of assembling them into 3D hetero-structures to create artificial materials which do not exist naturallyand thus give rise to new functionalities which seem promisingfor many future applications [3, 4]. These heterostructures can bebuilt by a variety of methods [5] like chemical vapor deposition(CVD) [6], epitaxial growth [7, 8], inkjet printing [9] or moregenerally by mechanical stacking of individual layers [10, 11].Among the many available techniques for stacking 2D materials[11–16], one approach that has become common recently is tomechanically exfoliate bulk 2D crystals onto a stamp made of aviscoelastic material, such as poly-dimethylsiloxane (PDMS),bring them in contact with a desired substrate and then slowlydetach the stamp to leave the 2D flakes behind [17, 18].Although this procedure is very versatile, deterministic and fairlysimple to perform, not much attention has been paid to the sur-face cleanliness of the flakes transferred this way.PDMS is a widely used polymer for contact printing [19],microfluidics [20] as well as stretchable electronics [21], andNanotechnologyNanotechnology 29 (2018) 265203 (9pp) https://doi.org/10.1088/1361-6528/aabd90Original content from this work may be used under the termsof the Creative Commons Attribution 3.0 licence. Anyfurther distribution of this work must maintain attribution to the author(s) andthe title of the work, journal citation and DOI.0957-4484/18/265203+09$33.00 © 2018 IOP Publishing Ltd Printed in the UK1https://orcid.org/0000-0002-3495-5935https://orcid.org/0000-0002-3495-5935https://orcid.org/0000-0003-0722-6697https://orcid.org/0000-0003-0722-6697https://orcid.org/0000-0002-6485-3083https://orcid.org/0000-0002-6485-3083https://orcid.org/0000-0003-2930-6698https://orcid.org/0000-0003-2930-6698https://orcid.org/0000-0002-9970-8345https://orcid.org/0000-0002-9970-8345mailto:lnovotny@ethz.chhttps://doi.org/10.1088/1361-6528/aabd90https://doi.org/10.1088/1361-6528/aabd90http://crossmark.crossref.org/dialog/?doi=10.1088/1361-6528/aabd90&domain=pdf&date_stamp=2018-05-03http://crossmark.crossref.org/dialog/?doi=10.1088/1361-6528/aabd90&domain=pdf&date_stamp=2018-05-03http://creativecommons.org/licenses/by/3.0/its chemistry has been studied extensively [22]. PDMS iscomposed of a network of cross-linked dimethylsiloxaneoligomers which do not get fully cross-linked even afterextensive curing [23] and depending on the curing time andtemperature, up to 5% of oligomers can remain uncrosslinkedwithin the PDMS bulk [23, 24]. It is well-known that theseuncrosslinked species are even present on the surface ofPDMS stamps and get transferred to the target substrateduring contact printing [25, 26], thereafter acting as a surfacecontamination layer. PDMS oligomer residues have beencharacterized previously by many techniques such as non-linear spectroscopy [27], x-ray photoelectron spectroscopy[25, 28], atomic force microscopy (AFM) and ToF-SIMS[28], confirming that PDMS can indeed degrade the surfacecleanliness of transferred materials. Residues were also foundto occur on 2D materials, e.g.graphene [29] and TMDCs[30], transferred using PDMS. Owing to their 2D structure,the properties of 2D materials are quite sensitive to surfacecontaminants and residues trapped at the interfaces within 3Dheterostructures [31]. Hence, it is essential to stack thesematerials with minimum residues.In this work, the cleanliness of 2D materials transferredusing PDMS was investigated with the help of AFM andphotoluminescence (PL) measurements. Using monolayer(1L) MoS2 as a test layer, we show evidence for substantialresidues being present on MoS2 transferred onto hBN fromPDMS. For transferring MoS2 in a cleaner way, we developedan ultraviolet-ozone (UV-O3) treatment recipe to clean thePDMS surface before exfoliating MoS2 on it and found asignificant reduction in residues compared to transfer fromuntreated PDMS. Using PL and Raman spectroscopy, wefurther reveal that a small compressive strain can be present inMoS2 after transfer from PDMS. We found that subsequentvacuum annealing leads to an almost pristine MoS2 surface onhBN, mostly free from interfacial bubbles, wrinkles andstrain. Although here we chose 1L-MoS2 for demonstrationsince its sensitive PL and Raman signals allow for systematicoptical characterization, our cleaning recipe is general and canbe used with other 2D materials as well.2. Results and discussionNaturally occurring MoS2 crystals purchased from SPI Sup-plies were exfoliated on commercial PDMS films (Gel-Film®PF-40-X4 sold by Gel-Pak) using a blue tape (BT-150E-KL).Figure 1(a) is an optical microscope image of a monolayerMoS2 flake on PDMS. Bulk hBN crystals were separatelyexfoliated on O2 plasma cleaned p+Si/SiO2 (285 nm) sub-strates. The PDMS stamp with MoS2 was placed on a trans-parent quartz plate and aligned on top of a suitable hBN flakeon SiO2 using a mask aligner as depicted schematically infigure 1(b). Upon slowly bringing MoS2 in contact with hBNat room temperature, the entire stack was heated to ∼65 ◦Cfor two minutes using a Peltier module kept underneath theSi/SiO2 substrate. After allowing the stack to cool down, thePDMS stamp was slowly detached as described in [18],leaving behind the MoS2 flake on hBN (figure 1(c)).To characterize the surface of the flakes after transfer, weperformed topography mapping using an AFM operating intapping mode. Figure 1(d) shows an AFM topography map ofthe region outlined in black in figure 1(c). The MoS2 flakeappears mostly ‘clean’ on this large scale apart from the usualwrinkles and bubbles which are frequently seen in PDMStransferred flakes [18]. However, if we examine the regionoutlined in red more closely in the higher resolution map infigure 1(e), a dense network of residue islands is clearlyvisible on the entire MoS2 surface, making it difficult to evenobtain an accurate estimate of the 1L-MoS2 thickness from ahorizontal cross-section (figure 1(h)). Note that these ±0.6 nm(rms) variations in topography cannot arise from substrateroughness alone as the hBN layer below is atomically smooth.Moreover, on the narrower 1L-MoS2 flake on the right infigure 1(d), a thick residue layer can be clearly identifiedwhich is unmistakably distinct from the MoS2 flake itself.Better visible in the higher resolution AFM map in figure 1(f),this additional layer left-over by PDMS can be as thick as∼2.5 nm in some areas (magenta profile in figure 1(i)).Together with the topography map in figure 1(e), we alsorecorded the corresponding AFM phase map as shown infigure 1(g) and observed a poor phase contrast between theMoS2 and hBN surfaces which also indicates the presence ofPDMS residues everywhere.We tested several flakes and similar residues were foundon all of them (see supplementary figure S1 available onlineat stacks.iop.org/NANO/29/265203/mmedia for residuespresent on another flake). The amount of residues transferredwas also affected by the temperature and pressure appliedduring the transfer. In case of transfers done without applyingany heat, residues were visible even in an optical microscopedue a change in the color of the SiO2 (285 nm) layer whicharises from interference and is sensitive to a few nanometersthick transparent organic layer on top (see supplementaryfigure S2). These results unambiguously demonstrate, inagreement with previous reports [25–30], that PDMS canindeed leave a significant amount of residues behind and anefficient method for eliminating them is genuinely needed.Although a high temperature (400 ◦C) vacuum annealafter transfer has been reported to reduce PDMS residues ongraphene [29], such high temperatures are known to introduceadditional defects in TMDCs [32, 33] which are more fragilecompared to graphene. Moreover, we found that annealing byitself is not sufficient to fully restore the transferred flakes totheir pristine state. The MoS2 flake shown in figure 1(c) wasannealed at 200 ◦C in vacuum for 3 h. In the AFM maps infigures 2(a), (b) it can be noticed that although bubbles andwrinkles are mostly gone, the MoS2 surface is still smearedwith residues. This is also evident from the AFM cross-section in figure 2(b) revealing the total thickness of the flaketo be 1.6 nm, which is significantly higher than the expectedvalue of 0.7 nm for monolayer MoS2. An additional vacuumanneal using the same parameters did somewhat reduce thethickness to ∼0.9 nm as depicted in figure 2(c). However, itcan be easily seen that the surface topography is quite inho-mogeneous and does not approach the cleanliness level of afreshly exfoliated MoS2 flake.2Nanotechnology 29 (2018) 265203 A Jain et alhttp://stacks.iop.org/NANO/29/265203/mmediaAlternatively, dissolving PDMS residues in organic sol-vents such as dichloromethane and toluene [24] is also notvery effective as the solvent molecules themselves tend to getadsorbed at exposed 2D material surfaces/edges and can evenchemically dope TMDCs [34]. This brings us to the questionof how to eliminate PDMS residues in a way that does notcompromise the 2D material being transferred in any way. Inthis regard, it appears more reasonable to clean the PDMSsurface itself before it comes in contact with the 2D material.To achieve this, we developed a UV-O3 treatment recipe tomodify the PDMS surface prior to MoS2 exfoliation. Thisrecipe was found to significantly reduce transfer residueswithout having any negative effect on the MoS2. UV-O3cleaning of PDMS has been proposed in the past [25] and themechanism behind it can be understood as follows. Oxygen freeradicals and ozone (O3) created from atmospheric oxygen in thepresence of UV-radiation, break down organic species on thePDMS surface into CO2, H2O and simpler volatile organicproducts. At the same time, the silicon present in PDMS getsoxidized and forms a thin (20–30 nm) layer of silicon oxide(SiOx) on PDMS [35]. This SiOx surface layer besides having alow carbon content, also acts as a diffusion barrier for uncros-slinked oligomers still present within the PDMS bulk.To optimize the treatment time, we exposed severalPDMS stamps to UV-O3 in a Bioforce Nanosciences UVOzone ProCleaner for varying times, and found a duration of30–40 min to be the optimum. Shorter times did not fullyclean the PDMS and longer exposure led to a poorer coverageof flakes on the PDMS upon exfoliation. We also observedthat if the exfoliation was done immediately after UV-O3Figure 1. Characterization of PDMS residues. (a) Optical microscope image of a 1L-MoS2 flake exfoliated on PDMS. (b) Schematicillustration of an inverted PDMS stamp with MoS2 being transferred onto hBN. (c) The same MoS2 flake as in (a) after transfer to hBN. (d)Large area AFM topography map of the region outlined in (c) displaying an ‘apparently clean’ surface together with some bubbles andwrinkles. (e) Higher resolution AFM map of the red-outlined region in (d) exhibiting a dense distribution of residues over the entire MoS2flake. (f) Higher resolution AFM map of the green-outlined region in (d) with a thick layer of residues covering a majority of the area. (g)AFM phase map recorded together with the topography in (e) depicting a poor phase contrast between MoS2 and hBN due to PDMS residueson both surfaces. (h) Height profile across the MoS2 edge along the dashed line marked in (e). An accurate estimation of the thickness ofmonolayer MoS2 is hindered by topography variations due to surface residues. (i) Height profile across the PDMS residue layer and PDMS +MoS2 along the dashed lines marked in (f). The thickness of the residue layer is ∼2.5 nm.3Nanotechnology 29 (2018) 265203 A Jain et altreatment, bonding often occurred between the PDMS and theblue tape used for exfoliation or in rare cases even betweenthe PDMS and O2 plasma treated SiO2 surface during thetransfer process. However, this bonding could be avoided byleaving the PDMS exposed to ambient air for a few hours tolet the surface undergo a partial hydrophobic recovery[36, 37]. This led us to the following optimized process flow,the results of which are presented below.The PDMS stamp was exposed to UV-O3 for 30 min andthen after a wait interval of 2 h in air, MoS2 was exfoliated onPDMS as usual. Upon optical identification of suitable flakes,transfer was carried out the same way as described earlier.Figure 3(a) shows an optical microscope image of a largeMoS2 flake exfoliated on 30 min UV-O3 treated PDMS whichwas subsequently transferred to hBN (figure 3(b)). In theAFM topography map of the 1L region shown in figure 3(c),one can notice the absence of dense islands or thick layers ofPDMS residues unlike in figures 1(e) or (f). Although smallamounts of residues can still be detected, this MoS2 flake issignificantly cleaner than the one in figure 1 which wastransferred from untreated PDMS (also see supplementaryfigure S3 for better resolved AFM images of another cleanMoS2 flake).The large number of bright spots in figure 3(c) arewrinkles and bubbles filled mainly with air molecules andorganic adsorbates that become trapped between MoS2 andhBN during transfer and get squeezed into small pockets via aself-cleaning effect [13, 38]. The amount of trapped bubblescan be reduced by performing the exfoliation and transferinside an Ar filled glove box to exclude molecular adsorbates[39] or even fully eliminated by stacking in vacuum as shownrecently [40]. Alternatively, we found that bubbles andwrinkles can be very efficiently removed by vacuumannealing at 200 ◦C, similar to another recent report [41]. Atthis temperature and under low pressure, the trapped speciesinside the bubbles become mobile and coalesce into biggerbubbles to minimize the total surface energy [16]. They alsotend to migrate towards the edge from where they eventuallyescape the interface, thereby lowering the overall density ofbubbles. Figure 3(d) is an AFM map of the same region asfigure 3(c) after 3 h vacuum annealing at 200 ◦C exhibiting acomplete removal of wrinkles and a remarkable reduction inbubbles. The higher resolution AFM map in figure 3(e) of thered-outlined region features a nearly pristine surface with aclean step of 7.2Å (figure 3(f)) and a surface roughness of±1.4Å (rms) over the entire flake. This is in striking contrastto the MoS2 flake transferred from untreated PDMS where thesurface quality remained compromised by PDMS residueseven after prolonged annealing as shown in figures 2(b), (c).The corresponding AFM phase map in figure 3(f) (inset)shows a clear phase contrast between MoS2 and hBN (unlikefigure 1(g)) providing further evidence for the decrease inresidues by UV-O3 treatment of PDMS. Detailed AFM ana-lysis of two additional 1L-MoS2 flakes transferred fromUV-O3 treated PDMS can be found in supplementary figuresS3–S4 showing similarly clean surfaces.Thus our recipe provides a new method to obtain veryclean 2D material flakes on hBN. We would like to stress thata combination of UV-O3 pre-cleaning followed by vacuumannealing is crucial, and vacuum annealing by itself does notresult in a pristine surface as evident from figure 2. Note thathere we transferred MoS2 onto atomically smooth hBN flakesto decouple the SiO2 substrate roughness from the surfacetopography of MoS2 which allowed us to better resolve sur-face residues using AFM. Although the surface cleanlinessdoes not depend on the substrate used, we have observed thatthe removal of bubbles and wrinkles during vacuum anneal-ing is not very effective for MoS2 transferred directly on SiO2(see discussion in supplementary section S4).3. Optical characterizationTo further characterize the effect of PDMS transfer on theoptical properties of MoS2, we performed PL and Ramanspectroscopy. Experimental details can be found in section 5(methods). Spatial maps of the spectrally integrated PL countswere recorded from the MoS2 flake in figure 1(d) transferredFigure 2. Limited effect of annealing on PDMS residues. (a) AFM topography map of the dirty 1L-MoS2 flake shown in figure 1(d) after 3 hannealing in vacuum. (b) Higher resolution AFM map of the white-outlined region in (a). Even though the surface appears mostly free frombubbles and more homogeneous than before (see figure 1(e)), a thin layer of PDMS residues is still present over the entire MoS2 flake. Inset:height profile along the red-dashed line revealing a total thickness of 1.6 nm for the monolayer flake. (d) AFM map of the same region afterannealing again for 3 h. Although the total amount of residues does decrease after prolonged annealing, the surface looks far from pristine.Inset: height profile along the green dashed line.4Nanotechnology 29 (2018) 265203 A Jain et alusing untreated PDMS (labelled as ‘untreated’ in figure 4) aswell as from the clean MoS2 flake transferred using UV-O3treated PDMS and subsequently vacuum annealed(figure 3(d)). PL maps of the two flakes depicted infigures 4(a), (b) show no apparent differences and the countrates were comparable for similar excitation powers. Thisindicates that PDMS residues (and/or vacuum annealing) donot seem to affect the overall PL quantum yield which couldbe one reason why residues have been overlooked in the past.The PL spectra of the two flakes, however, did show someinteresting differences.Besides the well-known A, B-exciton and trion (X−)peaks, a new broad feature around 1.68 eV can be noticed infigure 4(c) in the PL spectra of untreated MoS2. This low-energy peak can be better visualized at the locations markedby green and yellow dots in figure 4(a) where thick PDMSlayers were present on MoS2 (as shown previously in theAFM map in figure 1(f)). At these spots, the main excitonicpeaks were weaker which made it easier to resolve the newpeak (see figure 4(c) inset). Surprisingly, this peak could notbe detected in MoS2 flakes transferred from UV-O3 treatedPDMS (light blue and orange curves in figure 4(d) inset) aswell as in the fluorescence spectra of an untreated PDMSstamp without MoS2 (data not shown here). It appeared in thePL spectra of MoS2 only when PDMS residues were presenton MoS2 but its true origin is unclear at present. Such a broadlow-energy emission could possibly be attributed to impuritybound excitons at room temperature [33, 42] whose creation,however, is still not well understood in literature and a moredetailed investigation is needed to elucidate the exact mech-anism that gives rise to this additional peak.Apart from this, it can be clearly observed in figure 4(d)that the A-exciton peaks of as-transferred MoS2 (light anddark blue curves) are slightly blue-shifted with respect to thatof vacuum annealed MoS2 (orange curve) and lie at1.887± 0.002 eV which agrees very well with the value of1.89 eV measured previously on MoS2 transferred fromPDMS onto various substrates [43]. On the contrary, forvacuum annealed MoS2 the A-exciton peak lies at1.876±0.002 eV, very close to that of MoS2 exfoliateddirectly on SiO2 [44, 45]. By comparing the PL spectra atvarious locations on the two flakes in figures 4(a), (b) andperforming multi-Lorentzian fitting to decouple the trion peakfrom the exciton peak, we estimated a blue-shift of11±3 meV for the A-exciton peak of as-transferred MoS2.Moreover, one can also notice a blue-shift in the B-excitonpeak as highlighted by the Lorentzian fits in figure 4(e). Thecomplete set of fits for the entire spectra can be found insupplementary figure S6.Figure 3. Clean transfer from UV-ozone cleaned PDMS. (a) Optical microscope image of 1L-MoS2 exfoliated on 30 min UV-O3 cleanedPDMS. (b) The same MoS2 flake after transfer to hBN with the monolayer segment demarcated. (c) AFM topography map of the regionoutlined in (b). Areas densely covered with residues like in figures 1(e) or (f) could no longer be detected although a considerable amount ofbubbles do occur which could also be efficiently removed. (d) AFM map of the same region showing a drastic reduction in bubbles andwrinkles from the entire flake after 200 ◦C vacuum annealing. (e) Higher resolution AFM map of the red-outlined region in (d) displaying avery clean and smooth MoS2 surface. The faint streaks correspond to migration paths of residual bubbles which could not reach the MoS2edge and get released. (f) Height profile along the red-dashed line in (e) exhibiting a clear monolayer MoS2 step of 7.2 Å unlike infigures 1(h) or 2(b), (c). Inset: AFM phase map recorded together with the topography in (e) revealing a clear phase contrast between MoS2and hBN which further indicates the absence of residues (compare with figure 1(g)).5Nanotechnology 29 (2018) 265203 A Jain et alIn order to understand the origin of the A, B-excitonblue-shift upon transfer, we performed Raman spectroscopyto gain further insight from the E2g1 mode which is sensitive tostrain in MoS2. In exfoliated, unstrained 1L-MoS2 at roomtemperature, the E2g1 peak should lie at 385 cm−1 [46] whereasfor our as-transferred MoS2 on hBN it lies at ∼386 cm−1(blue curve in figure 4(f)). An E2g1 peak up-shift signifies anincrease in built-in compressive strain, thus implying thatPDMS not only leaves residues behind, but can also compressthe MoS2 during transfer. An up-shift of ∼1 cm−1 corre-sponds to an accumulation of ∼0.22% compressive strain inMoS2 after transfer [47], similar to estimate made by Bus-cema et al [43]. This induced strain causes an increase in thedirect bandgap at the K-point which leads to blue-shifted A,B-exciton emission [48, 49]. According to previously reportedDFT calculations [50] as well as experimental results [47], a0.22% strain would induce an A-exciton shift of 10± 1.5 meVwhich is in good agreement with the shift of 11± 3meVestimated from our PL measurements. For the sake of com-pleteness, one can also compare the A1g peaks in the twoRaman spectra to characterize the effect of residues on doping.The strong electron–phonon coupling of the out-of-plane A1gmode in MoS2 causes it to down-shift with increasing doping[51]. In figure 4f, the two A1g peaks lying at ∼406.5 cm−1imply a low n-doping in both MoS2 flakes on hBN (comparedto 403–404 cm−1 for 1L-MoS2 on SiO2) [43, 46].The origin of strain can possibly be attributed to theinherent lack of stiffness of PDMS. As discussed previouslyFigure 4. Optical characterization. (a, b) Spatial maps of the spectrally integrated PL counts from the 1L-MoS2 flakes in figures 1(d), 3(d)displaying similar emission intensities. The dark spots and lines in (a) are interfacial bubbles and wrinkles respectively (see figure 1(d)) wherethe photon counts were somewhat decreased. (c) PL spectra of MoS2 transferred from untreated PDMS recorded at points marked bymatching colored dots in (a). Besides the indicated excitonic peaks, all spectra also show a low-energy feature (marked by arrow) around1.68 eV. (d) PL spectra of MoS2 transferred from UV-O3 treated PDMS showing the absence of any low-energy feature in comparison withMoS2 from untreated PDMS. This indicates that the new feature arises only in the presence of PDMS residues on MoS2. Apart from this, thePL spectra in light and dark blue show blue-shifted A, B excitonic peaks which reveals the presence of compressive strain in as-transferredMoS2 (irrespective of PDMS pre-treatment). This strain got released from the flake in (b) upon annealing. The PL spectra in light blue wasrecorded from the flake in figure S3a before annealing while the annealed MoS2 spectra (orange) was taken from the flake in (b). Insets:magnified plots of the spectra around 1.68 eV. (e) A magnified view of the B-exciton emission in (d). The filled area plots are Lorentzian fitsto the B-exciton peaks and display a clear shift in the center positions. (f) Raman spectra of the two flakes (vertically offset for clarity). Theempty circles are measured data points while the smooth lines are Lorentzian fits to estimate the center position of each peak. The E2g1 peak(in blue) is up-shifted by ∼1 cm−1 which also indicates a build-up of compressive strain in as-transferred MoS2, in agreement with the PLspectra. All spectra have been normalized to the Raman peak of silicon for a better comparison.6Nanotechnology 29 (2018) 265203 A Jain et alby Gomez et al [18], PDMS being soft can get slightlydeformed during transfer by the pressure exerted on PDMSupon coming in contact with the target substrate. It is quitelikely that this deformation of PDMS induces a compressivestrain in the flake being transferred as measured for MoS2 inour case. The induced strain eventually gets released uponvacuum annealing, down-shifting the E2g1 peak to 385 cm−1and at the same time red-shifting the A, B-exciton peaks tounstrained values close to that of directly exfoliated MoS2.The release of transfer induced compression in MoS2 can alsobe evidenced in the AFM scans in figures 3(c), (d) before andafter annealing where an increase in the total surface areaupon annealing can be clearly noticed.4. ConclusionsTo summarize, we have demonstrated that PDMS residues aswell as compressive strain can be present in 2D materialsexfoliated onto and transferred from PDMS stamps. Using1L-MoS2 as an example, we have observed evidence of sur-face contamination in both AFM maps and PL spectra.UV-O3 treatment of PDMS prior to exfoliation significantlyreduces the amount of unwanted surface oligomers whichresults in a cleaner transfer of MoS2 in comparison withuntreated PDMS. We showed that a 1L-MoS2 surface of avery high-quality can be obtained by a combination of UV-O3pre-cleaning followed by vacuum annealing after transfer.AFM topography of annealed MoS2 flakes on hBN displayeda homogeneously smooth surface with a substantial reductionin interfacial bubbles and wrinkles. PL spectroscopy of as-transferred MoS2 revealed blue-shifted A, B-exciton peaksdue to an accumulation of compressive strain during thetransfer. This induced compression could be released by apost-transfer vacuum anneal.It is advantageous to exfoliate MoS2 on PDMS forobtaining large area (>1000 μm2) monolayer MoS2 flakeswith a high yield unlike direct exfoliation on SiO2 whichresults in relatively smaller flakes. The recipe we providednow makes it possible to integrate these large area flakesexfoliated on PDMS into clean heterostructures for highperformance electronic and photonic devices. Our results arevaluable for future experimental studies and practical appli-cations utilizing clean 2D material heterostructures.One must keep in mind that even though PDMS residuesdo not significantly influence the optical properties of MoS2,surface residues could still affect electrical transport byscattering charge carriers and thereby reduce carrier mobility.Moreover, trapped residues within TMDC heterostructurescould also weaken interlayer coupling [31] and adverselyaffect physical phenomena such as interlayer charge recom-bination or separation, interlayer electron–phonon coupling,polariton formation, out-of-plane tunneling, etc., which relyon high-quality interfaces. Hence, the importance of elim-inating residues while stacking 2D materials using PDMS, orany other technique in general, cannot be overstated. In thisdirection, our results highlight the significance of a carefulcharacterization and optimization of any transfer procedureand subsequent processing steps in order to preserve interfacequality and obtain unperturbed crystal structures with well-defined physical properties.5. MethodsAll transfers were done with a SÜSS MicroTec MJB4 maskaligner in air. Annealing was performed in a quartz tubefurnace (Carbolite Gero) at 200 ◦C for 3 h in a low vacuum of5×10−3 mbar (limited by our rotary pump). Before heating,the quartz tube was flushed several times with argon gas toremove any residual water or oxygen molecules. After 3 h, thefurnace was left for a few hours to cool down naturally toroom temperature before taking out the samples.All measurements were carried out at room temperatureunder ambient conditions. For PL measurements, sampleswere mounted on a piezoelectric stage and excited with a532 nm Nd:YAG laser (attenuated to 15 μW power) using a50x air objective (0.8 NA) in a scanning confocal microscopysetup. The MoS2 flakes were raster scanned across the laserfocus and photon counts at each xy-position were recordedwith a single photon counting module (Perkin Elmer SPCM-AQRH-14) after passing through a 532 nm RazorEdge long-pass filter. PL spectra were measured with a PrincetonInstruments Acton SP300i spectrometer at 100 μW excitationpower (30 s integration). For Raman spectroscopy, excitationwas done with the 530.8 nm line of an Ar-Kr laser (Coherent)at 100 μW power and the back-scattered light was dispersedonto a grating with 1200 grooves mm–1. The grating wascalibrated with a Neon lamp before recording the Ramanspectra. Lorentzian fits were obtained using the freely avail-able peak-o-mat software [52]. All spectra were normalized tothe Raman peak of p+Si lying at 521.7 cm−1 for a bettercomparison.AcknowledgmentsThis research was supported by the Swiss National ScienceFoundation (grant no. 200021_165841) and ETH Zürich(ETH-32 15-1). TT and KW acknowledge support from theElemental Strategy Initiative conducted by the MEXT, Japanand JSPS KAKENHI (grant no. JP15K21722).Author contributionsThe overall project was conceived and supervised by PB andLN. AJ developed the cleaning procedure, fabricated thesamples, performed the measurements and wrote the manu-script. PB built the confocal optical microscopy setup used forPL spectroscopy of MoS2. Raman characterization was car-ried out together with SH. The vacuum tube furnace wassetup by MP who also wrote the LabVIEW script for per-forming vacuum annealing. TT and KW synthesized the hBNcrystals used in this study. PB, SH, MP and LN also7Nanotechnology 29 (2018) 265203 A Jain et alparticipated in discussion of results and contributed to writingthe manuscript.ORCID iDsAchint Jain https://orcid.org/0000-0002-3495-5935Palash Bharadwaj https://orcid.org/0000-0003-0722-6697Sebastian Heeg https://orcid.org/0000-0002-6485-3083Markus Parzefall https://orcid.org/0000-0003-2930-6698Lukas Novotny https://orcid.org/0000-0002-9970-8345References[1] Xu M, Liang T, Shi M and Chen H 2013 Graphene-like two-dimensional materials Chem. Rev. 113 3766–98[2] Bhimanapati G R et al 2015 Recent advances in two-dimensionalmaterials beyond graphene ACS Nano 9 11509–39[3] Geim A K and Grigorieva I V 2013 Van der Waalsheterostructures Nature 499 419–25[4] Novoselov K, Mishchenko A, Carvalho A and Neto A C 20162D materials and van der Waals heterostructures Science353 461[5] Lim H, Yoon S I, Kim G, Jang A-R and Shin H S 2014Stacking of two-dimensional materials in lateral and verticaldirections Chem. Mater. 26 4891–903[6] Gong Y et al 2014 Vertical and in-plane heterostructures fromWS2/MoS2 monolayers Nat. Mater. 13 1135–42[7] Koma A, Sunouchi K and Miyajima T 1985 Fabrication ofultrathin heterostructures with van der Waals epitaxy J. VacSci. Technol. B 3 724[8] Lin Y-C et al 2015 Atomically thin resonant tunnel diodes builtfrom synthetic van der Waals heterostructures Nat.Commun. 6 7311[9] McManus D et al 2017 Water-based and biocompatible 2Dcrystal inks for all-inkjet-printed heterostructures Nat.Nanotechnol. 12 343–50[10] Dean C R et al 2010 Boron nitride substrates for high-qualitygraphene electronics Nat. Nanotechnol. 5 722–6[11] Wang L et al 2013 One-dimensional electrical contact to a two-dimensional material Science 342 614–7[12] Zomer P J, Guimarães M H D, Brant J C, Tombros N andvan Wees B J 2014 Fast pick up technique for high qualityheterostructures of bilayer graphene and hexagonal boronnitride Appl. Phys. Lett. 105 013101[13] Kretinin A V et al 2014 Electronic properties of grapheneencapsulated with different two-dimensional atomic crystalsNano Lett. 14 3270–6[14] Parzefall M et al 2015 Antenna-coupled photon emission fromhexagonal boron nitride tunnel junctions Nat. Nanotechnol.10 1058–63[15] Pizzocchero F et al 2016 The hot pick-up technique for batchassembly of van der Waals heterostructures Nat. Commun. 711894[16] Frisenda R et al 2018 Recent progress in the assembly ofnanodevices and van der Waals heterostructures bydeterministic placement of 2D materials Chem. Soc. Rev. 4753–68[17] Meitl M A et al 2006 Transfer printing by kinetic control ofadhesion to an elastomeric stamp Nat. Mater. 5 33–8[18] Castellanos-Gomez A et al 2014 Deterministic transfer of two-dimensional materials by all-dry viscoelastic stamping 2DMater. 1 011002[19] Kumar A and Whitesides G M 1993 Features of gold havingmicrometer to centimeter dimensions can be formed througha combination of stamping with an elastomeric stamp and analkanethiol ink followed by chemical etching Appl. Phys.Lett. 63 2002–4[20] Duffy D C, McDonald J C, Schueller O J A andWhitesides G M 1998 Rapid prototyping of microfluidicsystems in poly(dimethylsiloxane) Anal. Chem. 704974–84[21] Rogers J A, Someya T and Huang Y 2010 Materials andmechanics for stretchable electronics Science 327 1603–7[22] Xia Y and Whitesides G M 1998 Soft lithography Angew.Chem., Int. Ed. Engl. 37 550–75[23] Regehr K J et al 2009 Biological implications ofpolydimethylsiloxane-based microfluidic cell culture LabChip 9 2132–9[24] Lee J N, Park C and Whitesides G M 2003 Solventcompatibility of poly(dimethylsiloxane)-based microfluidicdevices Anal. Chem. 75 6544–54[25] Glasmästar K, Gold J, Andersson A-S, Sutherland D S andKasemo B 2003 Silicone transfer during microcontactprinting Langmuir 19 5475–83[26] Briseno A L et al 2006 Patterning organic semiconductorsusing dry poly(dimethylsiloxane) elastomeric stamps for thinfilm transistors J. Am. Chem. Soc. 128 3880–1[27] Böhm I, Lampert A, Buck M, Eisert F and Grunze M 1999 Aspectroscopic study of thiol layers prepared by contactprinting Appl. Surf. Sci. 141 237–43[28] Yunus S, de Looringhe C d C, Poleunis C and Delcorte A 2007Diffusion of oligomers from polydimethylsiloxane stamps inmicrocontact printing: surface analysis and possibleapplication Surf. Int. Anal. 39 922–5[29] Allen M J et al 2009 Soft transfer printing of chemicallyconverted graphene Adv. Mat. 21 2098–102[30] Liu K et al 2014 Elastic properties of chemical-vapor-deposited monolayer MoS2, WS2, and their bilayerheterostructures Nano Lett. 14 5097–103[31] Tongay S et al 2014 Tuning interlayer coupling in large-areaheterostructures with CVD-grown MoS2 and WS2monolayers Nano Lett. 14 3185–90[32] Tongay S et al 2013 Broad-range modulation of light emissionin two-dimensional semiconductors by molecularphysisorption gating Nano Lett. 13 2831–6[33] Tongay S et al 2013 Defects activated photoluminescence intwo-dimensional semiconductors: interplay between bound,charged, and free excitons Sci. Rep. 3 2657[34] Yang L et al 2014 Chloride molecular doping technique on 2Dmaterials: WS2 and MoS2 Nano Lett. 14 6275–80[35] Ouyang M, Yuan C, Muisener R J, Boulares A andKoberstein J T 2000 Conversion of some siloxane polymersto silicon oxide by UV/ozone photochemical processesChem. Mater. 12 1591–6[36] Bodas D and Khan-Malek C 2007 Hydrophilization andhydrophobic recovery of PDMS by oxygen plasma andchemical treatment - an SEM investigation SensorsActuators B 123 368–73[37] Morent R et al 2007 Adhesion enhancement by a dielectricbarrier discharge of PDMS used for flexible and stretchableelectronics J. Phys. D: Appl. Phys. 40 7392–401[38] Haigh S J et al 2012 Cross-sectional imaging of individuallayers and buried interfaces of graphene-basedheterostructures and superlattices Nat. Mater. 11 764–7[39] Cao Y et al 2015 Quality heterostructures from two-dimensional crystals unstable in air by their assembly ininert atmosphere Nano Lett. 15 4914–21[40] Kang K et al 2017 Layer-by-layer assembly of two-dimensional materials into wafer-scale heterostructuresNature 550 229–338Nanotechnology 29 (2018) 265203 A Jain et alhttps://orcid.org/0000-0002-3495-5935https://orcid.org/0000-0002-3495-5935https://orcid.org/0000-0002-3495-5935https://orcid.org/0000-0002-3495-5935https://orcid.org/0000-0003-0722-6697https://orcid.org/0000-0003-0722-6697https://orcid.org/0000-0003-0722-6697https://orcid.org/0000-0003-0722-6697https://orcid.org/0000-0002-6485-3083https://orcid.org/0000-0002-6485-3083https://orcid.org/0000-0002-6485-3083https://orcid.org/0000-0002-6485-3083https://orcid.org/0000-0003-2930-6698https://orcid.org/0000-0003-2930-6698https://orcid.org/0000-0003-2930-6698https://orcid.org/0000-0003-2930-6698https://orcid.org/0000-0002-9970-8345https://orcid.org/0000-0002-9970-8345https://orcid.org/0000-0002-9970-8345https://orcid.org/0000-0002-9970-8345https://doi.org/10.1021/cr300263ahttps://doi.org/10.1021/cr300263ahttps://doi.org/10.1021/cr300263ahttps://doi.org/10.1021/acsnano.5b05556https://doi.org/10.1021/acsnano.5b05556https://doi.org/10.1021/acsnano.5b05556https://doi.org/10.1038/nature12385https://doi.org/10.1038/nature12385https://doi.org/10.1038/nature12385https://doi.org/10.1126/science.aac9439https://doi.org/10.1021/cm502170qhttps://doi.org/10.1021/cm502170qhttps://doi.org/10.1021/cm502170qhttps://doi.org/10.1038/nmat4091https://doi.org/10.1038/nmat4091https://doi.org/10.1038/nmat4091https://doi.org/10.1116/1.583125https://doi.org/10.1038/ncomms8311https://doi.org/10.1038/nnano.2016.281https://doi.org/10.1038/nnano.2016.281https://doi.org/10.1038/nnano.2016.281https://doi.org/10.1038/nnano.2010.172https://doi.org/10.1038/nnano.2010.172https://doi.org/10.1038/nnano.2010.172https://doi.org/10.1126/science.1244358https://doi.org/10.1126/science.1244358https://doi.org/10.1126/science.1244358https://doi.org/10.1063/1.4886096https://doi.org/10.1021/nl5006542https://doi.org/10.1021/nl5006542https://doi.org/10.1021/nl5006542https://doi.org/10.1038/nnano.2015.203https://doi.org/10.1038/nnano.2015.203https://doi.org/10.1038/nnano.2015.203https://doi.org/10.1038/ncomms11894https://doi.org/10.1038/ncomms11894https://doi.org/10.1039/C7CS00556Chttps://doi.org/10.1039/C7CS00556Chttps://doi.org/10.1039/C7CS00556Chttps://doi.org/10.1039/C7CS00556Chttps://doi.org/10.1038/nmat1532https://doi.org/10.1038/nmat1532https://doi.org/10.1038/nmat1532https://doi.org/10.1088/2053-1583/1/1/011002https://doi.org/10.1063/1.110628https://doi.org/10.1063/1.110628https://doi.org/10.1063/1.110628https://doi.org/10.1021/ac980656zhttps://doi.org/10.1021/ac980656zhttps://doi.org/10.1021/ac980656zhttps://doi.org/10.1021/ac980656zhttps://doi.org/10.1126/science.1182383https://doi.org/10.1126/science.1182383https://doi.org/10.1126/science.1182383https://doi.org/10.1002/(SICI)1521-3773(19980316)37:5<550::AID-ANIE550>3.0.CO;2-Ghttps://doi.org/10.1002/(SICI)1521-3773(19980316)37:5<550::AID-ANIE550>3.0.CO;2-Ghttps://doi.org/10.1002/(SICI)1521-3773(19980316)37:5<550::AID-ANIE550>3.0.CO;2-Ghttps://doi.org/10.1039/b903043chttps://doi.org/10.1039/b903043chttps://doi.org/10.1039/b903043chttps://doi.org/10.1021/ac0346712https://doi.org/10.1021/ac0346712https://doi.org/10.1021/ac0346712https://doi.org/10.1021/la026558xhttps://doi.org/10.1021/la026558xhttps://doi.org/10.1021/la026558xhttps://doi.org/10.1021/ja058226vhttps://doi.org/10.1021/ja058226vhttps://doi.org/10.1021/ja058226vhttps://doi.org/10.1016/S0169-4332(98)00510-8https://doi.org/10.1016/S0169-4332(98)00510-8https://doi.org/10.1016/S0169-4332(98)00510-8https://doi.org/10.1002/sia.2623https://doi.org/10.1002/sia.2623https://doi.org/10.1002/sia.2623https://doi.org/10.1002/adma.200803000https://doi.org/10.1002/adma.200803000https://doi.org/10.1002/adma.200803000https://doi.org/10.1021/nl501793ahttps://doi.org/10.1021/nl501793ahttps://doi.org/10.1021/nl501793ahttps://doi.org/10.1021/nl500515qhttps://doi.org/10.1021/nl500515qhttps://doi.org/10.1021/nl500515qhttps://doi.org/10.1021/nl4011172https://doi.org/10.1021/nl4011172https://doi.org/10.1021/nl4011172https://doi.org/10.1038/srep02657https://doi.org/10.1021/nl502603dhttps://doi.org/10.1021/nl502603dhttps://doi.org/10.1021/nl502603dhttps://doi.org/10.1021/cm990770dhttps://doi.org/10.1021/cm990770dhttps://doi.org/10.1021/cm990770dhttps://doi.org/10.1016/j.snb.2006.08.037https://doi.org/10.1016/j.snb.2006.08.037https://doi.org/10.1016/j.snb.2006.08.037https://doi.org/10.1088/0022-3727/40/23/021https://doi.org/10.1088/0022-3727/40/23/021https://doi.org/10.1088/0022-3727/40/23/021https://doi.org/10.1038/nmat3386https://doi.org/10.1038/nmat3386https://doi.org/10.1038/nmat3386https://doi.org/10.1021/acs.nanolett.5b00648https://doi.org/10.1021/acs.nanolett.5b00648https://doi.org/10.1021/acs.nanolett.5b00648https://doi.org/10.1038/nature23905https://doi.org/10.1038/nature23905https://doi.org/10.1038/nature23905[41] Wierzbowski J et al 2017 Direct exciton emission fromatomically thin transition metal dichalcogenideheterostructures near the lifetime limit Sci. Rep. 7 12383[42] Chow P K et al 2015 Defect-induced photoluminescence inmonolayer semiconducting transition metal dichalcogenidesACS Nano 9 1520–7[43] Buscema M, Steele G A, van der Zant H S J andCastellanos-Gomez A 2014 The effect of the substrate on theRaman and photoluminescence emission of single-layerMoS2 Nano Res. 7 561–71[44] Schuller J A et al 2013 Orientation of luminescent excitons inlayered nanomaterials Nat. Nanotechnol. 8 271–6[45] Amani M et al 2015 Near-unity photoluminescence quantumyield in MoS2 Science 350 1065–8[46] Li H et al 2012 From bulk to monolayer MoS2: evolution oframan scattering Adv. Funct. Mater. 22 1385–90[47] Conley H J et al 2013 Bandgap engineering of strainedmonolayer and bilayer MoS2 Nano Lett. 13 3626–30[48] Nayak A P et al 2015 Pressure-dependent optical andvibrational properties of monolayer molybdenum disulfideNano Lett. 15 346–53[49] Dou X, Ding K, Jiang D, Fan X and Sun B 2016 Probing spin–orbit coupling and interlayer coupling in atomically thinmolybdenum disulfide using hydrostatic pressure ACS Nano10 1619–24[50] Liu Z et al 2014 Strain and structure heterogeneity in MoS2atomic layers grown by chemical vapour deposition Nat.Commun. 5 5246[51] Chakraborty B et al 2012 Symmetry-dependent phononrenormalization in monolayer MoS2 transistor Phys. Rev. B85 161403[52] peak-o-mat http://lorentz.sourceforge.net/9Nanotechnology 29 (2018) 265203 A Jain et alhttps://doi.org/10.1038/s41598-017-09739-4https://doi.org/10.1021/nn5073495https://doi.org/10.1021/nn5073495https://doi.org/10.1021/nn5073495https://doi.org/10.1007/s12274-014-0424-0https://doi.org/10.1007/s12274-014-0424-0https://doi.org/10.1007/s12274-014-0424-0https://doi.org/10.1038/nnano.2013.20https://doi.org/10.1038/nnano.2013.20https://doi.org/10.1038/nnano.2013.20https://doi.org/10.1126/science.aad2114https://doi.org/10.1126/science.aad2114https://doi.org/10.1126/science.aad2114https://doi.org/10.1002/adfm.201102111https://doi.org/10.1002/adfm.201102111https://doi.org/10.1002/adfm.201102111https://doi.org/10.1021/nl4014748https://doi.org/10.1021/nl4014748https://doi.org/10.1021/nl4014748https://doi.org/10.1021/nl5036397https://doi.org/10.1021/nl5036397https://doi.org/10.1021/nl5036397https://doi.org/10.1021/acsnano.5b07273https://doi.org/10.1021/acsnano.5b07273https://doi.org/10.1021/acsnano.5b07273https://doi.org/10.1038/ncomms6246https://doi.org/10.1103/PhysRevB.85.161403http://lorentz.sourceforge.net/ 1. Introduction 2. Results and discussion 3. Optical characterization 4. Conclusions 5. Methods Acknowledgments Author contributions References