# Fileset

[ncomms10783.pdf](https://mdr.nims.go.jp/filesets/798a3d4f-a861-4594-8c75-59d4bc66bf2d/download)

## Creator

Achim Woessner, Pablo Alonso-González, Mark B. Lundeberg, Yuanda Gao, Jose E. Barrios-Vargas, Gabriele Navickaite, Qiong Ma, Davide Janner, [Kenji Watanabe](https://orcid.org/0000-0003-3701-8119), Aron W. Cummings, [Takashi Taniguchi](https://orcid.org/0000-0002-1467-3105), Valerio Pruneri, Stephan Roche, Pablo Jarillo-Herrero, James Hone, Rainer Hillenbrand, Frank H. L. Koppens

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Near-field photocurrent nanoscopy on bare and encapsulated graphene](https://mdr.nims.go.jp/datasets/a1f37478-216d-404e-918b-d547b44b0c6b)

## Fulltext

Near-field photocurrent nanoscopy on bare and encapsulated grapheneARTICLEReceived 27 Oct 2015 | Accepted 20 Jan 2016 | Published 26 Feb 2016Near-field photocurrent nanoscopy on bare andencapsulated grapheneAchim Woessner1, Pablo Alonso-González2,3,*, Mark B. Lundeberg1,*, Yuanda Gao4, Jose E. Barrios-Vargas5,Gabriele Navickaite1, Qiong Ma6, Davide Janner1, Kenji Watanabe7, Aron W. Cummings5, Takashi Taniguchi7,Valerio Pruneri1,8, Stephan Roche5,8, Pablo Jarillo-Herrero6, James Hone4, Rainer Hillenbrand9,10& Frank H.L. Koppens1,8Optoelectronic devices utilizing graphene have demonstrated unique capabilities andperformances beyond state-of-the-art technologies. However, requirements in terms ofdevice quality and uniformity are demanding. A major roadblock towards high-performancedevices are nanoscale variations of the graphene device properties, impacting theirmacroscopic behaviour. Here we present and apply non-invasive optoelectronic nanoscopy tomeasure the optical and electronic properties of graphene devices locally. This is achieved bycombining scanning near-field infrared nanoscopy with electrical read-out, allowing infraredphotocurrent mapping at length scales of tens of nanometres. Using this technique, we studythe impact of edges and grain boundaries on the spatial carrier density profiles and localthermoelectric properties. Moreover, we show that the technique can readily be applied toencapsulated graphene devices. We observe charge build-up near the edges and demonstratea solution to this issue.DOI: 10.1038/ncomms10783 OPEN1 ICFO—Institut de Ciencies Fotoniques, The Barcelona Institute of Science and Technology, 08860 Barcelona, Spain. 2 CIC nanoGUNE, 20018 Donostia-SanSebastian, Spain. 3 Institute of Physics, Chinese Academy of Science, Beijing 100190, China. 4 Department of Mechanical Engineering, Columbia University,New York, New York 10027, USA. 5 Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and The Barcelona Institute of Science andTechnology, Campus UAB, 08193 Barcelona, Spain. 6 Department of Physics, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA.7 National Institute for Materials Science, 1-1 Namiki, Tsukuba 305-0044, Japan. 8 ICREA-Institució Catalana de Recerca i Estudis Avançats, 08010Barcelona, Spain. 9 CIC nanoGUNE and UPV/EHU, 20018 Donostia-San Sebastian, Spain. 10 IKERBASQUE, Basque Foundation for Science, 48011 Bilbao,Spain. * These authors contributed equally to this work. Correspondence and requests for materials should be addressed to F.H.L.K. (email:frank.koppens@icfo.eu).NATURE COMMUNICATIONS | 7:10783 | DOI: 10.1038/ncomms10783 | www.nature.com/naturecommunications 1mailto:frank.koppens@icfo.euhttp://www.nature.com/naturecommunicationsAs large scale integration and wafer scale device processingcapabilities of graphene have become available1–8,technological implementations of electronic andoptoelectronic graphene devices are within reach9–11. At thesame time, to achieve high device performance, any imperfectionsat the nanometer or even atomic scale need to be minimizedor even eliminated. For example, in large area graphene, grownby chemical vapour deposition (CVD), grain boundaries arethe stitching regions between different monocrystalline partsof graphene and act as carrier scatterers, limiting the graphenemobility and uniformity12,13. In addition, even perfectlymonocrystalline graphene is still highly sensitive to itsenvironment, and on typical substrates charge–densityinhomogeneities (charge puddles)14–19 and additional dopingnear contacts, defects and edges arise, which reduce the deviceperformance as well. Therefore, it is important to efficiently probethe nanoscale optoelectronic properties of graphene devices andto understand their microscopic physical behaviour.A major challenge is that many of the available characterizationtechniques are invasive20, need specifically designed devicestructures13,21,22, image only very small areas14,15,21,23–25, relyon high doping of the graphene,26 require unhindered electricalaccess of the probe to the graphene14,15,21,23,24 or lack the desirednanometer resolution27 and are expensive and difficult toimplement. For the direct quality control of graphene devices, amethod that can image electrical and optical properties ofgraphene devices, at nanoscale resolution, without any specialpreparation and without modifying the devices is required.Here we demonstrate fully non-invasive room-temperaturescanning near-field photocurrent nanoscopy28–38 for the firsttime applied on graphene with infrared frequencies and use it tostudy the nanoscale optoelectronic properties of graphenedevices that can later be used for real applications. Thistechnique is based on electrical probing of the photoresponsedue to strongly localized heating. We apply this techniqueto study the microscopic physics of grain boundaries andcharge–density inhomogeneities. In the case of grainboundaries, we were able to identify the magnitude of theirSeebeck coefficient, while for charge–density inhomogeneities, weshow how they influence the global charge neutrality point ofgraphene devices. In addition, we study encapsulated graphenedevices39,40, where the encapsulation would prevent many otherscanning probe techniques from accessing local properties ofgraphene. In these devices, we find a charge build-up near theedges and show that using local metal gates instead of a globalbackgate effectively suppresses this type of edge doping.In general, this technique operates most effectively withmid-infrared light because it does not lead to photodoping41and it is more stable in operation, compared with visible light.ResultsMeasurement principle. The measurement principle is sketchedin Fig. 1a. The setup is based on a scattering-type scanningnear-field optical microscope (s-SNOM)26,42 augmented withelectrical contact to the sample to measure currents in situ28–38.In contrast to conventional s-SNOM, we do not need to measurethe outscattered light but rather directly measure current inducedby the near-field as explained in the following. A 10.6-mmmid-infrared laser illuminates a metallized atomic forcemicroscope probe, tapping at its mechanical resonancefrequency. Part of the incoming light, polarized parallel to theshaft of the probe, excites a strong electric field at the tip apex dueto an antenna effect43. The spatial extent of this near-field is onthe order of 25 nm, limited only by the tip radius and muchsmaller than the free space wavelength of the impinging light43.The near and far fields impinging on the device induce chargeflows in the device (by mechanisms discussed below), and drivecurrents into an external current amplifier via contacts on thedevice. We isolate the part of the current that is induced by nearfields by demodulating the current at the second harmonic of thetip tapping frequency43. This demodulated current is denotedIPC and referred to as near-field photocurrent and is obtainedtogether with near-field optical and topography information.A typical map of IPC, obtained by scanning the tip over a CVDgraphene device, is shown in Fig. 1b.We can assess the spatial resolution of the photocurrent mapsby comparing a region near the edge (Fig. 1d) with a topographicimage from the same region (Fig. 1c). As can be seen, IPC falls tozero for tip locations away from the graphene on a similar lengthscale as the topography, demonstrating the successful isolation ofnear-field contributions. In Fig. 1e, we quantify the resolution byobserving the change in IPC as the tip is moved over the edge ofgraphene. The full-width at half maximum of the photocurrentpeak at this location is B100 nm, matching the rise distance inthe topographic signal. This resolution is far below any limitsrelating to the 10.6 mm free space light wavelength.Photothermoelectric photocurrent generation mechanism. Asto the physical mechanism of the photocurrent, we considerthe photothermoelectric effect that has been shown to dominatethe photoresponse of graphene11,44–48: the light (in this case, thetip-enhanced near-field) locally heats the graphene, and this heatacts via non-uniformities in Seebeck coefficient S to drive chargecurrents within the device and into the contacts (see Methodssection). Therefore, we interpret the variations of IPC in terms ofmicroscopic variations in S. The Seebeck coefficient, whichdepends on material properties such as carrier density andmobility, is a measure of the electromotive force driven by atemperature difference in a material. A complete description ofIPC needs to take into account the carrier cooling length45,46and overall sample geometry49. The carrier cooling lengthlcool ¼ffiffiffiffiffiffiffiffik=gp, where k the sheet thermal conductivity inplane and g the interfacial thermal conductivity out of plane tothe heat sinking substrate, describes how far heat propagatesthrough the charge carriers, before dissipating to the environment(see Supplementary Fig. 2)46. A quantitative model ofthe thermoelectric photocurrent mechanism can be found inthe Methods section.Grain boundary characterization. We first discuss theapplication of this infrared near-field photocurrent technique tograin boundaries. They are not visible in the simultaneouslyacquired topography, and are responsible for some of theline-shaped features in the photocurrent map in Fig. 1b. Some ofthe other features stem from large scale inhomogeneities of thesample. The region within the green frame is shown with higherresolution in Fig. 1d, exhibiting a strong photocurrent signalthat changes sign along a sharp boundary, yet the graphene istopographically flat in the vicinity of this boundary (Fig. 1c). Weshow now that this type of feature indicates a grain boundary.Figure 2a shows a line profile of IPC across the boundaryfeature identified in Fig. 1d. This antisymmetric IPC can beexplained by a localized deviation in S at the boundary, that is, aline defect within an otherwise uniform thermoelectricmedium (Supplementary Figs 3,4 and Supplementary Note 1).Indeed, grain boundaries behave as localized lines of stronglymodified electronic properties, within otherwise uniformgraphene20,21,26,50,51. We remark that the decay of thephotocurrent away from the boundary extends over more thanARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms107832 NATURE COMMUNICATIONS | 7:10783 | DOI: 10.1038/ncomms10783 | www.nature.com/naturecommunicationshttp://www.nature.com/naturecommunications100 nm, which is due to a larger hot carrier cooling. We find inthis case lcool¼ 140 nm.To gain more insight in the Seebeck coefficient at the grainboundary, we tune the carrier density by a global gate (Fig. 2d).We observe that the antisymmetric spatial photocurrent profilechanges sign as the backgate voltage VBG passes the peak inresistance, that is, the global charge neutrality point VD.The Seebeck coefficient SG of graphene itself changes sign atthe charge neutrality point44,45,52,53 (Fig. 2c). Thus, aftercalibrating the sign to the known sign of the contactphotocurrent, our data implies that the Seebeck coefficient ofthe grain boundary SGB is always smaller in magnitude than SG,since IPC(VBG)pSG(VBG)� SGB(VBG).Using a polycrystalline graphene model, we compute theresistance due to grain boundaries using a Kubo transportformalism and real space simulations54. SGB is the ratio of thefirst- and zero-order Onsager coefficients (SupplementaryNote 2). Indeed, we find that SGB is always smaller inmagnitude and has a similar lineshape as SG in the carrierdensity range measured (Fig. 2c). Figure 2e shows a simulation ofthe photocurrent for the calculated Seebeck coefficients, which isin agreement with the measurements (Supplementary Figs 5,6).Charge puddle characterization. We next examine near-fieldphotocurrent in a typical two-probe exfoliated graphene device(Fig. 3). A strong photocurrent is obtained with the tip near themetal contacts, similar to previous near- and far-field measure-ments34,47,55. In addition, an apparently random pattern ofphotocurrent is present throughout the device, as in high-resolution far-field measurements55 but at a much finer scale.The random photocurrent pattern between the contacts inFig. 3a indicates random variations in Seebeck coefficient overshort length scales (Supplementary Fig. 7). Random variations ofthe Seebeck coefficient are indeed expected since it depends oncarrier density52, which in turn has fine-scaled inhomogeneities(charge puddles)14–18. The photocurrent variations can thus beused to gain insight in the charge puddle distribution. A moredetailed view of the photocurrent due to charge puddles in Fig. 3bshows that the length scale that can be resolved is on the order ofhundreds of nanometres.Quantitatively, from the autocorrelation of the photocurrent incomparison with a photothermoelectric model taking intoaccount the size of the charge puddles in Fig. 3c we extractlcoolB200 nm. The charge puddles are modelled to have a size ofB20 nm, in accordance with measurements of graphene onsilicon oxide (SiO2; refs 15–18).By changing the gate voltage we study the carrier densityprofile with high spatial resolution (Fig. 4) and highlight thepossibility of spatially resolving the charge neutrality point for alarge device. IPC from charge puddles is largest around the chargeneutrality point and varies with position. This is consistent withthe very high sensitivity of the Seebeck coefficient to changes incarrier density, near-zero density (Fig. 4b). The magnitude ofphotocurrent from charge puddles depends on the difference ofSeebeck coefficient between two adjacent charge puddles or inother words the strongest photocurrent from charge puddleappears at the position of highest Seebeck gradient. This allows usto map the local carrier density offset (charge inhomogeneity)throughout the device, as indicated by the extremum of photo-current in a scan of photocurrent versus gate voltage (Fig. 4c). Thephotocurrent from adjacent charge puddles with a given chargecarrier density offset does not change sign when sweeping throughthe charge neutrality point. This is because the difference inSeebeck coefficient between these puddles does not change sign.a−0.20.00.2I PC (nA)b051015Topography (nm)−0.20.00.2I PC (nA)Position (nm)0.00.10.20.3−I PC (nA)05Height (nm)c d e0 100 200 300 400 500Figure 1 | Near-field photocurrent working principle and photocurrent from grain boundaries. (a) Sketch of the scattering-type scanning near-fieldoptical microscope setup. A mid-infrared laser illuminates the atomic force microscope tip, which generates a locally concentrated optical field, which isabsorbed by the graphene generating a position dependent photocurrent. The blue region in the graphene lattice represents a grain boundary with amodified Seebeck coefficient. The arrows sketch the photocurrent flow pattern. For each position only the magnitude and direction of the current aremeasured. The sketch is not to scale. (b) IPC map at at backgate voltage VBG¼0 V of a single layer CVD graphene device (Supplementary Fig. 1) with threecontacts: top left (drain), right (source) and bottom left (ground). Both grain boundaries and wrinkles show characteristic photocurrent patterns. (scale bar,5 mm) The green box indicates the measurement region in c,d. (c) Topography of etched CVD graphene does not show grain boundary but only wrinklesand other inhomogeneities due to the transfer process. (scale bar, 500 nm) (d) IPC at VBG¼0 V clearly shows a grain boundary and the expected signchange around it. The black dashed line indicates the measurement positions in Fig. 2a,d. (scale bar, 500 nm) (e) Topography (orange) and IPC (red)measured at the orange line in c and the red line in d, respectively.NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10783 ARTICLENATURE COMMUNICATIONS | 7:10783 | DOI: 10.1038/ncomms10783 | www.nature.com/naturecommunications 3http://www.nature.com/naturecommunicationsWe can thus resolve the local charge neutrality point at a givenposition of the device (green curve, Fig. 4a), which can bedifferent from the global charge neutrality point VD, the backgatevoltage VBG at which the resistance is maximum (blue curve,Fig. 4a). We show that the global charge neutrality point (bluecurve, Fig. 4a) is determined by an average of the gate voltages atwhich the local charge neutrality points appear (red curve,Fig. 4a). Spatially resolved puddle photocurrent can be muchnarrower (green curve, Fig. 4a) than the average of all possiblecurrent paths (red curve, Fig. 4a) This indicates that the graphenelocally has less inhomogeneity. Thus the technique gives insightnot only in the global but also in the local behaviour of the device.Characterization of encapsulated devices. Finally, we apply thistechnique to a graphene device encapsulated between two layersof hexagonal boron nitride (h-BN), using the polymer-free vander Waals assembly technique39,40 as sketched in Fig. 5e. Thisdevice lies on top of an oxidized silicon wafer, used as a backgate.The stack is etched into a triangle and electrically side-contactedby two metal contacts39. Recent studies in vacuum and lowtemperatures56,57 have shown that the edges affect where currentflows in the device, in particular near charge neutrality. In thefollowing we study the build-up of edge doping in ambientconditions and provide a solution to this.While monitoring the photocurrent of such encapsulateddevices, we observe indications of strong carrier density variationsnear the edges over micrometre scales. These variations areinfluenced by lighting conditions (Supplementary Fig. 8), gatevoltages, and temperature, and evolve over timescales rangingfrom minutes to weeks. As an example, Fig. 5a–d shows aprogression of photocurrent maps, taken after annealing thedevice at 200 �C for 30 min in ambient conditions to temporarilyremove charge density variations near the edges. Initially inFig. 5a we see very small photocurrents indicating a flat carrierdensity landscape. After some time (hours), in the dark with onlygate voltages smaller than 3 V applied, a small doping gradientbetween the contacts builds up. This gradient leads to thestronger photocurrent shown in Fig. 5b. The local chargeneutrality point, indicated by the maximum of photocurrent, isat the same position close to the edge of the device as furtherinside the bulk. After keeping the device for 3 h in ambientconditions we can see a change of the local charge neutrality pointat the edge of the graphene compared to the bulk in Fig. 5c.The edge is slightly more p-type compared to the bulk. Finally, weapply high gate voltages, of in this case 50 V for B20 h, toincrease the edge doping. A strong p-doping at the edge and ann-doping in the bulk of the graphene is induced in Fig. 5d. Thisindicates that electric field accelerates the speed and increases thistype of edge doping.We exploit the observed edge doping to create a naturalp–n junction along the edge of the device. For this we0 100 200 300 400 500 600Position (nm)0Norm. IPC−20020S (μVK−1 )a0 2R (kΩ)10305070VBG (V)−50 0 50S (μ VK−1)−200 0 200Position (nm)IPC (a.u.)−200 0 200Position (nm)10305070VBG (V)−0.5 0.0 0.5IPC (nA)b c d eFigure 2 | Photocurrent profile at a grain boundary and its gate voltagedependence. (a) Photocurrent profile, measured at the black dashed line inFig. 1d, perpendicular to the grain boundary at VBG¼0 V shows goodagreement with the photothermoelectric model with lcool¼ 140 nm. Themodelled spatial Seebeck profile (with FWHM 20 nm) is shown in black.(b) Two-probe device resistance as a function of VBG. (c) SimulatedSeebeck coefficient SG for pristine graphene (solid line) and SGB forpolycrystalline graphene with an average grain size of 25 nm (dashed line;Supplementary Note 2). (d) Backgate dependent photocurrent profileperpendicular to the grain boundary shows that the grain boundary changesits sign at the charge neutrality point. (e) Simulated backgate dependentphotocurrent profile based on the Seebeck profiles in c normalized to thesimulated photocurrent maximum.Au AuSiO2SiO2−101I PC (nA)−101I PC (nA)−500 0 500Δx (nm)0.00.51.0Norm.Autocorr.abcFigure 3 | Photocurrent from charge puddles. (a) Near-field photocurrentmap of an exfoliated graphene device on 300 nm SiO2 at VBG¼ 20 V. Thedashed lines indicate the position of the contacts and solid lines thegraphene edges. The green box indicates the measurement region in b(scale bar, 5 mm). (b) Detailed photocurrent map at the charge neutralitypoint of the device (VBG¼ 7 V) reveals the charge puddles and the highspatial resolution of the technique. (scale bar, 1 mm) (c) Autocorrelation ofthe photocurrent from charge puddles at VD (data points) compared tophotocurrent expected from a random charge puddle distribution andlcool¼ 200 nm (blue curve). Autocorrelation is taken along the source draincurrent path.ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms107834 NATURE COMMUNICATIONS | 7:10783 | DOI: 10.1038/ncomms10783 | www.nature.com/naturecommunicationshttp://www.nature.com/naturecommunicationsapply a backgate voltage at which the edge of the grapheneis p-type and the bulk n-type. This is similar to the situationin Fig. 5d at VBG¼VD. We observe photocurrent at the junctionin Fig. 5f around the whole device, indicating that theedge doping is uniform around the graphene. The photocurrentdecays gradually towards the midline between the electrodesas a result of how the triangular geometry modifies theability of the contacts to capture photocurrents49. The signchange in the middle of the device is because in the currentdirection the junction changes from a p–n junction to ann–p junction (Supplementary Fig. 9 and SupplementaryNote 2). We are able to temporarily reset the edge doping byannealing the device on a hotplate at 200 �C for 30 min as weshow in Fig. 5a.0 1Norm. IPC0 5 10 15 20Position (μm)−101I PC (nA)−1 0 1Norm. S0 5R (kΩ)−10−505101520VBG (V)a b cFigure 4 | Dependence of photocurrent profiles on backgate voltage reveals doping inhomogeneities. (a) Backgate dependence of the resistance of thedevice measured simultaneously to the photocurrent in blue. The red curve shows the normalized root mean square of the photocurrent across the device.The green curve shows a single normalized photocurrent backgate trace, corresponding to the green dashed dotted line in c. (b) Backgate dependentSeebeck coefficient of graphene, calculated from the gate dependent resistance in a using the Mott formula52. (c) Backgate dependence of thephotocurrent across the device. Graphene is between the black dashed lines, which indicate the edges of the metal contacts.−2 −1 0VBG − VD (V) VBG − VD (V) VBG − VD (V) VBG − VD (V)0123Distance from edge (μm)After annealingEdge−0.1 0.0 0.1−1 0 1 2After 3 h−0.5 0.0 0.5−4 −2 0 2After 6 h−1.5 0.0 1.5−10 0 10After 20 h at 50 V1.0 × 1012 cm−20.1 × 1012 cm−2−1.5 0.0 1.5IPC (nA)IPC (nA)IPC (nA)IPC (nA)a b c dp-type p-typep-typen-typeContact Contact−20 −10 0 10 20IPC (nA)e fAuh-BNSiO2SiFigure 5 | Near-field photocurrent maps revealing edge doping in encapsulated graphene. (a) Spatial photocurrent profile versus backgate voltage VBG(minus voltage of the resistance maximum VD) near the edge of encapsulated graphene. These data are taken directly after annealing the device.(VD¼ �0.3 V). (b) The same scan on the same device after three hours in air, (VD¼ �0.5 V) and c, after annealing and applying VBG up to 3 V (VD¼ 3 V).(d) The same scan after approximately 20 h in air and after applying VBG up to 50 V (VD¼ 30 V). (e) Sketch of the device, a stack of h-BN(46 nm)/Graphene/h-BN(7 nm) on a Si/SiO2(300 nm) wafer used as global backgate. In this sketch we only show one of the two contacts used for electricalmeasurements. (f) Photocurrent close to the resistance maximum at VBG¼ � 28 V shows a triangular photocurrent pattern, due to edge p-doping. Thedashed dotted line indicates where the measurements in a–d were taken. (scale bar, 2 mm).NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10783 ARTICLENATURE COMMUNICATIONS | 7:10783 | DOI: 10.1038/ncomms10783 | www.nature.com/naturecommunications 5http://www.nature.com/naturecommunicationsLocal gates prevent charge build-up at the device edges. Whilewe have not been able to precisely identify the origin of theedge doping, we suspect that water molecules are able topenetrate between the boron nitride and the SiO2 due to thesurface of the h-BN being not completely conformal with theunderlying SiO2 substrate. This water molecule penetrationthen leads to trapped charges is responsible for the observededge doping. We also present here a technique to completelyeliminate the edge doping. We place encapsulated graphene ontop of a local conductive gate, such as a 15 nm AuPd alloysketched in Fig. 6a. As the photocurrent measurement in Fig. 6bshows, we find that edge doping is efficiently suppressedeven after extended periods of time at ambient conditions andhigh gate voltages. Furthermore, such devices efficientlysuppress the photodoping effect observed for devices where theh-BN is in contact with SiO2 (ref. 41; Supplementary Fig. 10 andSupplementary Note 1).In the device with a local metal gate used to suppress bothedge- and photodoping, we find small features due to chargepuddles on top of a slowly varying background photocurrent, dueto large-scale carrier density inhomogeneities. The size of thefeatures due to charge puddles determined by autocorrelation isB800 nm. The long length scale of those features is either due tothe longer cooling length of the encapsulated graphene comparedto the graphene on SiO2 or due to larger charge puddle size in theencapsulated devices. Further work is required to clearlydistinguish these effects.DiscussionTo conclude, we have demonstrated that scanning near-fieldphotocurrent nanoscopy is a versatile technique to characterizethe electronic and optoelectronic and even previously inaccessibleproperties of relevant graphene devices. This technique is highlypromising for spatially resolved quality control of regulargraphene devices without the need for special device structuresand can therefore be readily applied.MethodsPhotocurrent model. Photocurrent IPC in graphene as generated by thephotothermoelectric effect and is described as:45–47IPC ¼1RWZ Z@T@xS dxdy ð1Þwhere R is the total resistance including graphene, contacts and circuitry, W thedevice width and x the current flow direction. This is valid for rectangular graphenedevices and special care needs to be taken for arbitrary shapes, such as in Fig. 5(ref. 49). For the temperature profile T(x) we consider that the heat spreads in twodimensions with heat sinking to lattice and substrate, producing a T(x) profiledescribed by a modified Bessel function of the second kind, with a finite tip sizecorrection (Supplementary Note 2). A 25-nm finite tip-size correction was used forall simulations.Measurement details. The s-SNOM used was a NeaSNOM from Neaspec GmbH,equipped with a CO2 laser operated at 10.6 mm, away from the phonon resonanceof SiO2, which can lead to strong substrate contributions to the photocurrent48.The laser power used was B20 mW. The probes were commercially availablemetallized atomic force microscopy probes with an apex radius of approximately25 nm. The tip height was modulated at a frequency of approximately 250 kHz witha 60–80 nm amplitude. A Femto DLPCA-200 current pre-amplifier was used. Theprobe tip was electrically grounded. Because of the different device geometries andthe fact that all the measurements were taken at different times and slightlydifferent device conditions the absolute values of the photocurrents are notcomparable.Device fabrication. The CVD graphene, grown on copper, was transferred onto aself-assembled monolayer58 on 285 nm of SiO2 to stabilize the charge neutralitypoint. The contacts were defined using optical lithography with Ti (5 nm)/Pd(35 nm). The graphene was transferred onto deposited contacts.The exfoliated graphene device was fabricated on a Si/SiO2(300 nm) wafer, usedas backgate. The Cr(0.8 nm)/Au(80 nm) contacts were defined using electron beamlithography.The Si/SiO2(300 nm)/h-BN(46 nm)/Graphene/h-BN(7 nm) and theSi/SiO2(300 nm)/AuPd(15 nm)/h-BN(42 nm)/Graphene/h-BN(13 nm) stacks, werefabricated using the polymer-free van der Waals assembling technique39.References1. Li, X. et al. Large-area synthesis of high-quality and uniform graphene films oncopper foils. Science 324, 1312–1314 (2009).2. Reina, A. et al. Large area, few-layer graphene films on arbitrary substrates bychemical vapor deposition. Nano Lett. 9, 30–35 (2009).3. Bae, S. et al. Roll-to-roll production of 30-inch graphene films for transparentelectrodes. Nat. Nanotech. 5, 574–578 (2010).4. Bonaccorso, F. et al. Production and processing of graphene and 2d crystals.Mater. Today 15, 564–589 (2012).5. Ren, W. & Cheng, H.-M. The global growth of graphene. Nat. Nanotech. 9,726–730 (2014).6. Lee, J.-H. et al. Wafer-scale growth of single-crystal monolayer graphene onreusable hydrogen-terminated germanium. Science 344, 286–289 (2014).7. Gao, L. et al. Face-to-face transfer of wafer-scale graphene films. Nature 505,190–194 (2014).8. de Heer, W. A. et al. Large area and structured epitaxial graphene produced byconfinement controlled sublimation of silicon carbide. Proc. Natl Acad. Sci.USA 108, 16900–16905 (2011).9. Ferrari, A. C. et al. Science and technology roadmap for graphene, relatedtwo-dimensional crystals, and hybrid systems. Nanoscale 7, 4598–4810 (2014).10. Akinwande, D., Petrone, N. & Hone, J. Two-dimensional flexiblenanoelectronics. Nat. Commun. 5, 5678 (2014).11. Koppens, F. H. L. et al. Photodetectors based on graphene, other two-dimensional materials and hybrid systems. Nat. Nanotech. 9, 780–793 (2014).12. Yazyev, O. V. & Louie, S. G. Electronic transport in polycrystalline graphene.Nat. Mater. 9, 806–809 (2010).13. Yu, Q. et al. Control and characterization of individual grains and grainboundaries in graphene grown by chemical vapour deposition. Nat. Mater. 10,443–449 (2011).14. Martin, J. et al. Observation of electron-hole puddles in graphene using ascanning single-electron transistor. Nat. Phys. 4, 144–148 (2008).15. Chen, J.-H., Jang, C., Xiao, S., Ishigami, M. & Fuhrer, M. S. Intrinsic andextrinsic performance limits of graphene devices on SiO2. Nat. Nanotech. 3,206–209 (2008).16. Zhang, Y., Brar, V. W., Girit, C., Zettl, A. & Crommie, M. F. Origin of spatialcharge inhomogeneity in graphene. Nat. Phys. 5, 722–726 (2009).17. Decker, R. et al. Local electronic properties of graphene on a BN substrate viascanning tunneling microscopy. Nano Lett. 11, 2291–2295 (2011).18. Xue, J. et al. Scanning tunnelling microscopy and spectroscopy of ultra-flatgraphene on hexagonal boron nitride. Nat. Mater. 10, 282–285 (2011).19. Burson, K. M. et al. Direct imaging of charged impurity density in commongraphene substrates. Nano Lett. 13, 3576–3580 (2013).20. Duong, D. L. et al. Probing graphene grain boundaries with optical microscopy.Nature 490, 235–239 (2012).21. Huang, P. Y. et al. Grains and grain boundaries in single-layer graphene atomicpatchwork quilts. Nature 469, 389–392 (2011).−10 0 10IPC (pA)a bFigure 6 | Edge doping is efficiently suppressed using devices with localmetal gates. (a) Sketch of the device with local metal gate with the twoh-BN layers in light blue and the local metal gate in gold. The layers arefrom bottom to top SiO2(300 nm)/AuPd(15 nm)/h-BN(42 nm)/Graphene/h-BN(13 nm). (b) Photocurrent from charge puddles in encapsulatedgraphene on a metal gate close to the charge neutrality point. The electricalcontacts are on the left and right outside of this figure. (scale bar, 2mm).ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms107836 NATURE COMMUNICATIONS | 7:10783 | DOI: 10.1038/ncomms10783 | www.nature.com/naturecommunicationshttp://www.nature.com/naturecommunications22. Yasaei, P. et al. Bimodal phonon scattering in graphene grain boundaries. NanoLett. 15, 4532–4540 (2015).23. Deshpande, A., Bao, W., Miao, F., Lau, C. & LeRoy, B. Spatially resolvedspectroscopy of monolayer graphene on SiO2. Phys. Rev. B 79, 205411 (2009).24. Gibertini, M., Tomadin, A., Guinea, F., Katsnelson, M. I. & Polini, M.Electron-hole puddles in the absence of charged impurities. Phys. Rev. B 85,201405 (2012).25. Cho, S. et al. Thermoelectric imaging of structural disorder in epitaxialgraphene. Nat. Mater. 12, 913–918 (2013).26. Fei, Z. et al. Electronic and plasmonic phenomena at graphene grainboundaries. Nat. Nanotech. 8, 821–825 (2013).27. Ferrari, A. C. & Basko, D. M. Raman spectroscopy as a versatile tool forstudying the properties of graphene. Nat. Nanotech. 8, 235–246 (2013).28. Hsu, J. W. P., Fitzgerald, E. A., Xie, Y. H. & Silverman, P. J. Near-field scanningoptical microscopy imaging of individual threading dislocations on relaxedGexSi1-x films. Appl. Phys. Lett. 65, 344–346 (1994).29. Buratto, S. K. et al. Imaging InGaAsP quantum-well lasers using near-fieldscanning optical microscopy. J. Appl. Phys. 76, 7720 (1994).30. Goldberg, B. B., Ünlü, M. S., Herzog, W. D., Ghaemi, H. F. & Towe, E.Near-field optical studies of semiconductor heterostructures and laser diodes.IEEE J. Sel. Top. Quantum Electron. 1, 1073–1081 (1995).31. Richter, A., Tomm, J. W., Lienau, C. & Luft, J. Optical near-field photocurrentspectroscopy: A new technique for analyzing microscopic aging processes inoptoelectronic devices. Appl. Phys. Lett. 69, 3981 (1996).32. McNeill, C. R. et al. Direct photocurrent mapping of organic solar cells using anear-field scanning optical microscope. Nano Lett. 4, 219–223 (2004).33. Gu, Y. et al. Near-field scanning photocurrent microscopy of a nanowirephotodetector. Appl. Phys. Lett. 87, 043111 (2005).34. Mueller, T., Xia, F., Freitag, M., Tsang, J. & Avouris, P. Role of contacts in graphenetransistors: a scanning photocurrent study. Phys. Rev. B 79, 245430 (2009).35. Rauhut, N. et al. Antenna-enhanced photocurrent microscopy on single-walledcarbon nanotubes at 30 nm resolution. ACS Nano 6, 6416–6421 (2012).36. Mauser, N. & Hartschuh, A. Tip-enhanced near-field optical microscopy.Chem. Soc. Rev. 43, 1248–1262 (2014).37. Mauser, N. et al. Antenna-enhanced optoelectronic probing of carbonnanotubes. Nano Lett. 14, 3773–3778 (2014).38. Grover, S., Dubey, S., Mathew, J. P. & Deshmukh, M. M. Limits on thebolometric response of graphene due to flicker noise. Appl. Phys. Lett. 106,051113 (2015).39. Wang, L. et al. One-dimensional electrical contact to a two-dimensionalmaterial. Science 342, 614–617 (2013).40. Kretinin, A. V. et al. Electronic properties of graphene encapsulated withdifferent two-dimensional atomic crystals. Nano Lett. 14, 3270–3276 (2014).41. Ju, L. et al. Photoinduced doping in heterostructures of graphene and boronnitride. Nat. Nanotech. 9, 348–352 (2014).42. Fei, Z. et al. Infrared nanoscopy of dirac plasmons at the graphene-SiO2interface. Nano Lett. 11, 4701–4705 (2011).43. Keilmann, F. & Hillenbrand, R. Near-field microscopy by elastic light scatteringfrom a tip. Phil. Trans. R Soc. A 362, 787–805 (2004).44. Lemme, M. C. et al. Gate-activated photoresponse in a graphene p-n junction.Nano Lett. 11, 4134–4137 (2011).45. Gabor, N. M. et al. Hot carrier-assisted intrinsic photoresponse in graphene.Science 334, 648–652 (2011).46. Song, J. C. W., Rudner, M. S., Marcus, C. M. & Levitov, L. S. Hot carriertransport and photocurrent response in graphene. Nano Lett. 11, 4688–4692(2011).47. Tielrooij, K. J. et al. Hot-carrier photocurrent effects at graphene’ metalinterfaces. J. Phys. Condens. Matter 27, 164207 (2015).48. Badioli, M. et al. Phonon-mediated mid-infrared photoresponse of graphene.Nano Lett. 14, 6374–6381 (2014).49. Song, J. C. W. & Levitov, L. S. Shockley-Ramo theorem and long-rangephotocurrent response in gapless materials. Phys. Rev. B 90, 075415 (2014).50. Tsen, A. W. et al. Tailoring electrical transport across grain boundaries inpolycrystalline graphene. Science 336, 1143–1146 (2012).51. Van Tuan, D. et al. Scaling properties of charge transport in polycrystallinegraphene. Nano Lett. 13, 1730–1735 (2013).52. Zuev, Y., Chang, W. & Kim, P. Thermoelectric and magnetothermoelectrictransport measurements of graphene. Phys. Rev. Lett. 102, 096807 (2009).53. Wei, P., Bao, W., Pu, Y., Lau, C. N. & Shi, J. Anomalous thermoelectrictransport of dirac particles in graphene. Phys. Rev. Lett. 102, 166808 (2009).54. Cummings, A. W. et al. Charge transport in polycrystalline graphene:challenges and opportunities. Adv. Mater. 26, 5079–5094 (2014).55. Lee, E. J. H., Balasubramanian, K., Weitz, R. T., Burghard, M. & Kern, K.Contact and edge effects in graphene devices. Nat. Nanotech. 3, 486–490(2008).56. Ben Shalom, M. et al. Quantum oscillations of the critical current andhigh-field superconducting proximity in ballistic graphene. Nat. Phys.doi:10.1038/nphys3592 (2015).57. Allen, M. T. et al. Spatially resolved edge currents and guided-wave electronicstates in graphene. Nat. Phys. 12, 128–133 (2016).58. Chen, S. Y., Ho, P. H., Shiue, R. J., Chen, C. W. & Wang, W. H. Transport/magnetotransport of high-performance graphene transistors on organicmolecule-functionalized substrates. Nano Lett. 12, 964–969 (2012).AcknowledgementsWe thank Stijn Goossens, Klaas-Jan Tielrooij and Misha Fogler for many fruitfuldiscussions and Fabien Vialla for the assistance in rendering the graphics in Fig. 1a.Open source software was used (www.matplotlib.org, www.python.org, www.povray.org).F.H.L.K. acknowledges financial support from the Spanish Ministry of Economy andCompetitiveness, through the ‘Severo Ochoa’ Programme for Centres of Excellence inR&D (SEV-2015-0522), support by Fundacio Cellex Barcelona, the ERC Careerintegration grant (294056, GRANOP), the ERC starting grant (307806, CarbonLight), theGovernment of Catalonia trough the SGR grant (2014-SGR-1535), the Mineco grantsRamón y Cajal (RYC-2012-12281) and Plan Nacional (FIS2013-47161-P), and projectGRASP (FP7-ICT-2013-613024-GRASP). F.H.L.K. and R.H. acknowledge support by theE.C. (European Commission) under Graphene Flagship (contract no. CNECT-ICT-604391). D.J. acknowledges support from the Ramón y Cajal Fellowship Program. Y.G.and J.H. acknowledge support from the US Office of Naval Research N00014-13-1-0662.We acknowledge financial support from the Spanish Ministry of Economy and Compe-titiveness and ‘Fondo Europeo de Desarrollo Regional’ through Grant TEC2013-46168-R.Q.M. and P.J.-H. have been supported by AFOSR Grant No. FA9550-11-1-0225 and thePackard Fellowship program. This work made use of the Materials Research Science andEngineering Center Shared Experimental Facilities supported by the National ScienceFoundation (NSF) (Grant No. DMR-0819762) and of Harvard’s Center for NanoscaleSystems, supported by the NSF (Grant No. ECS-0335765). S.R. acknowledges the SpanishMinistry of Economy and Competitiveness for funding (MAT2012-33911), the Secretariade Universidades e Investigacion del Departamento de Economia y Conocimiento de laGeneralidad de Catalunya and the Severo Ochoa Program (MINECO SEV-2013-0295).J.E.B.-V. acknowledges support from SECITI (Mexico, D.F.).Author contributionsA.W. performed the experiments, analysed the data and wrote the manuscript. P.A.-G.helped with measurements, interpretation and discussion of the results. M.B.L. helpedwith data analysis, measurements, interpretation, discussion of the results andmanuscript writing. Y.G. fabricated the h-BN/graphene/h-BN devices. G.N. and D.J.fabricated the CVD graphene devices. Q.M. fabricated the exfoliated graphene devices.K.W. and T.T. synthesized the h-BN. J.E.B.-V., A.W.C. and S.R. performed thesimulations of the Seebeck coefficient at grain boundaries. R.H. and F.H.L.K. supervisedthe work, discussed the results and co-wrote the manuscript. All authors contributed tothe scientific discussion and manuscript revisions.Additional informationSupplementary Information accompanies this paper at http://www.nature.com/naturecommunicationsCompeting financial interests: R.H. is co-founder of Neaspec GmbH, a companyproducing scattering-type scanning near-field optical microscope systems such as theones used in this study. All other authors declare no competing financial interests.Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/How to cite this article: Woessner, A. et al. Near-field photocurrent nanoscopy on bareand encapsulated graphene. Nat. Commun. 7:10783 doi: 10.1038/ncomms10783 (2016).This work is licensed under a Creative Commons Attribution 4.0International License. The images or other third party material in thisarticle are included in the article’s Creative Commons license, unless indicated otherwisein the credit line; if the material is not included under the Creative Commons license,users will need to obtain permission from the license holder to reproduce the material.To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/NATURE COMMUNICATIONS | DOI: 10.1038/ncomms10783 ARTICLENATURE COMMUNICATIONS | 7:10783 | DOI: 10.1038/ncomms10783 | www.nature.com/naturecommunications 7http://dx.doi.org/10.1038/nphys3592www.matplotlib.orgwww.python.orgwww.povray.orghttp://www.nature.com/naturecommunicationshttp://www.nature.com/naturecommunicationshttp://npg.nature.com/reprintsandpermissions/http://npg.nature.com/reprintsandpermissions/http://creativecommons.org/licenses/by/4.0/http://www.nature.com/naturecommunications title_link Results Measurement principle Photothermoelectric photocurrent generation mechanism Grain boundary characterization Charge puddle characterization Figure™1Near-field photocurrent working principle and photocurrent from grain boundaries.(a) Sketch of the scattering-type scanning near-field optical microscope setup. A mid-infrared laser illuminates the atomic force microscope tip, which generates a lo Characterization of encapsulated devices Figure™2Photocurrent profile at a grain boundary and its gate voltage dependence.(a) Photocurrent profile, measured at the black dashed line in Fig.™1d, perpendicular to the grain boundary at VBG=0thinspV shows good agreement with the photothermoelectric  Figure™3Photocurrent from charge puddles.(a) Near-field photocurrent map of an exfoliated graphene device on 300thinspnm SiO2 at VBG=20thinspV. The dashed lines indicate the position of the contacts and solid lines the graphene edges. The green box indica Figure™4Dependence of photocurrent profiles on backgate voltage reveals doping inhomogeneities.(a) Backgate dependence of the resistance of the device measured simultaneously to the photocurrent in blue. The red curve shows the normalized root mean square Figure™5Near-field photocurrent maps revealing edge doping in encapsulated graphene.(a) Spatial photocurrent profile versus backgate voltage VBG (minus voltage of the resistance maximum VD) near the edge of encapsulated graphene. These data are taken dire Local gates prevent charge build-up at the device edges Discussion Methods Photocurrent model Measurement details Device fabrication LiX.Large-area synthesis of high-quality and uniform graphene films on copper foilsScience324131213142009ReinaA.Large area, few-layer graphene films on arbitrary substrates by chemical vapor depositionNano Lett.930352009BaeS.Roll-to-roll production of 30- Figure™6Edge doping is efficiently suppressed using devices with local metal gates.(a) Sketch of the device with local metal gate with &!QJ;the two h-—BN layers in light blue and the local metal gate in gold. The layers are from bottom to top SiO2(300thin We thank Stijn Goossens, Klaas-Jan Tielrooij and Misha Fogler for many fruitful discussions and Fabien Vialla for the assistance in rendering the graphics in Fig.™1a. Open source software was used (www.matplotlib.org, www.python.org, www.povray.org). F.H ACKNOWLEDGEMENTS Author contributions Additional information