BURGHARD AND MEWS VOL. 6 ’ NO. 7 ’ 5752–5756 ’ 2012

www.acsnano.org

5752

July 10, 2012

C 2012 American Chemical Society

High-Resolution PhotocurrentMapping of Carbon NanostructuresMarko Burghard† and Alf Mews‡,*

†Max Planck Institute for Solid State Research, Heisenbergstrasse 1, D-70569 Stuttgart, Germany and ‡Institute of Physical Chemistry, University of Hamburg,Grindelallee 117, D-20146 Hamburg, Germany

Photocurrent generation is a key pro-cess in a range of technologically rele-vant devices, in particular, solar cells

and photodetectors. One method used toinvestigate the solar cells' spatial uniformityin terms of electrical performance is a tech-nique called laser-beam-induced current(LBIC), which was developed in the early1980s.1 In this technique, a focused laserbeam performs a two-dimensional scan ofthe photoactive surface of the device,while measuring the generated short cir-cuit photocurrent. Thus, images of thephotoconversion efficiency of the scan-ned surface are obtained, with spatialresolution governed by the size of thelaser spot. These images can be directlycorrelated with structural informationprovided by other image-capturing tech-niques such as optical or electronicmicroscopy.2

With the advent of efficient chemicalsynthesis methods for carbon or inorganicsemiconductor-based nanostructures, aslight variation of the LBIC technique, scan-ning photocurrent microscopy (SPCM),emerged as a versatile diagnostic tool toexplore the electronic properties of indivi-dual nanotubes or nanowires. The pioneer-ing experiments revealed the presence ofphotocurrent signals at the metal contactsto individual semiconducting single-wallcarbon nanotubes (SWCNTs),3 whose mag-nitude is controllable by the application of aback gate voltage,4�6 which is known tomodulate the height of the Schottky bar-riers at these locations. Further SPCM stud-ies focused on determining built-in electricfields at the metal contacts to metallicSWCNTs7 or SWCNT interconnections.8

In parallel, the application of SPCM wasexpanded to inorganic semiconductornanowires,9 providing useful informationsuch as the electric field dependenceof the photocurrent decay length.10 Besides

the one-dimensional carbon nanotubes,SPCM was then also applied to their two-dimensional counterpart, graphene. The firstexperiments along this line examined thepotential steps formedat themetal contacts.11

However, the spatial resolution in the above-described SPCM experiments was alwaysrestricted to the diffraction-limited spot sizeof the laser (gλ/2), which prevented theevaluation of the electric potential variationson the nanoscale. As a first step towardincreased spatial resolution, Avouris and co-workers reported in 2009 scanning near-fieldopticalmicroscope (SNOM)-based SPCMmea-surements on graphene.12 In this manner,spatial resolutionof∼50nmcouldbe reached,which opened the possibility to probe theinterfaces between single-layer and multi-layered graphene regions.A further leap in terms of spatial resolu-

tion of SPCM has now been attained byRauhut et al. for individual carbonnanotubes,as described in this issue of ACS Nano.13

* Address correspondence tomews@chemie.uni-hamburg.de.

Published online10.1021/nn3029088

ABSTRACT

The spatial resolution of photocurrent measurements on carbon nanostructures has reached

20 nm, as demonstrated by Hartschuh and co-workers for individual carbon nanotubes in this

issue of ACS Nano. In this Perspective, we provide a brief overview of the applications of

scanning photocurrent microscopy to various one- and two-dimensional nanostructures and

highlight the importance of the optical antenna concept for future studies of the

optoelectronic properties of hybrid nanostructures.

PERSPECTIV

E

BURGHARD AND MEWS VOL. 6 ’ NO. 7 ’ 5752–5756 ’ 2012

www.acsnano.org

5753

They utilize the so-called aperture-less near-field technique, wherea sharp metal tip acting as ananoscopic antenna is brought intoclose proximity to the sample. Theantenna effect is based on the col-lective motion of the free electronswithin the metallic tip, which areresonantly excited by laser irradia-tion. These plasmonic oscillationsstrongly increase the electromag-netic field strength up to a fewnanometers away from the tip sur-face. Accordingly, strong light inter-action is limited to the proximity ofthe tip, enabling spatial resolutionfor optical excitation that is gov-erned mainly by the size of the tip.This is superior by approximately 1order of magnitude in comparisonto confocal microscopy methodssuch as scanning confocal Ramanmicroscopy14 or fluorescence micro-scopy and spectroscopy,15 whichhave been successfully employedto distinguish between metallicand semiconducting nanotubesand even to determine the helicalindex of individual carbon nano-tubes.16 As early as 2003, Hartschuhet al. used tip-enhanced near-fieldoptical microscopy (TENOM) to in-crease the spatial resolution of fluo-rescence and Raman microscopies.In these experiments, they couldobserve fluorescing segments ofnanotubes, which were only a fewtens of nanometers in length andwere attributed to chirality changesalong the tube axis.15 Now, they

have combined tip-enhancedRaman and photocurrent micro-scopy, which can even bemeasuredsimultaneously (Figure 1).13 This en-ables direct comparisons betweenthe signal intensity and spatial reso-lution in the Raman and SPCMmea-surements, which provides usefulinformation about the underlyingphysical enhancement mechanisms.Interestingly, they observed thespatial width of the Raman signals(∼20 nm) to be narrower than thewidth of the respective SPCM signals(∼30 nm). To explain this difference,Rauhut et al. suggest that, in Ramanscattering, the antenna effect en-hances both the excitation and alsothe emission rate, whereas in SPCM,only the excitation rate is enhanced.Accordingly, the width of theGaussian excitation cross sectionshould theoretically be smaller bya factor of

√2 for the Raman signal,

which is in perfect agreement withthe experimental observations.The tip-enhanced SPCM enabled

Rauhut et al. to gain deeper insightinto the photocurrent generationmechanism in carbon nanostruc-tures. Recently, it has been dis-cussed that, besides the built-inelectric field, the thermoelectric ef-fect represents another mechanismof photocurrent generation, basedon SPCM studies on graphene.17

The latter involves the formationof a temperature gradient ΔT dueto local heating by the laser, whichproduces a thermoelectric voltagegiven by V = S � ΔT (where S is theSeebeck coefficient). While thus far,no general consensus has been

attained as to which of the twomechanisms dominates, the dataobtained by Rauhut et al. supportthe built-in field mechanism for tworeasons. First, with increased opticalresolution, they were able to showthat the energy band profile ob-tained by integrating the photocur-rent signal along the nanotubedecays exponentially along thelength of the nanotube, in closeagreement with theoretical predic-tions. Second, they observed a signchange in the signal far away fromthemetal contacts that could not beexplained by local laser heating ofthe nanotube nor by (indirect) heat-ing of the electrodes. However, itshould be noted that the situationmay well be different for the photo-response close to the metal con-tacts. Indeed, according to a recentconfocal-microscopy-based studyon networks of SWCNTs, the con-tact-related photoresponse mainlyarises from temperature variationsat the metal�nanotube interface.18

In addition, the photoresponsehas been found to be influencedby the device operation regime,the type of substrate, as wellas postgrowth treatment of thenanotubes.Another interesting aspect of tip-

enhanced SPCM is the ability toexplore the effects of local pertur-bations along the nanotube axis.For example Rauhut et al. observedfor some samples that the sign ofthe SPCM signal changed on lengthscales of only a few tens of nano-meters. Possible origins for thisbehavior involve local structural

A further leap in terms

of spatial resolution of

scanning photocurrent

microscopy has now

been attained by

Rauhut et al. for

individual carbon

nanotubes, as described

in this issue of ACS

Nano.

Figure 1. Scheme of the tip-enhanced photocurrent microscopy setup as used byRauhut et al. Reprinted from ref 13. Copyright 2012 American Chemical Society.The photocurrent through contacted carbonnanotubes ismonitored as a functionof the illumination position. By using a sharpmetallic tip as an antenna to enhancethe light field in a small volume, changes in the electric field along the tube withextensions of just a few tens of nanometers can be resolved.

PERSPECTIV

E

BURGHARD AND MEWS VOL. 6 ’ NO. 7 ’ 5752–5756 ’ 2012

www.acsnano.org

5754

defects of the CNTs or electrostaticperturbations from the environ-ment of the nanotubes. By takingadvantage of the combined micro-scopic information, they couldidentify charged particles in closeproximity to the nanotube as themost likely cause of the photo-current features. This conclusionis based on the finding that thephotocurrent signatures alwayscoincide with the presence of asmall particle at the same positionin the acquired topography image.Further support could be gainedfrom the simultaneously recordedintensity of the so-called defectRaman mode (D-band), which isa measure of the local structuraldefect density. In particular, theabsence of correlations betweenthe local defect density and thespatial photocurrent modulationsuggests that the latter is due toband bending resulting from elec-trostatic interactions with theattached particles.

OUTLOOK AND FUTURECHALLENGES

The use of antenna effects repre-sents a highly attractive approachthat is applicable more broadly toother nanostructures. Along theselines, it has been demonstratedthat, by attaching plasmonic nano-structures (e.g., lithographically de-fined gold dot patterns) to gra-phene, the efficiency of graphenep�n junction photodetectors canbe increased up to 20-fold.19 Further-more, with the aid of nanostructures

of different geometries, it has beenpossible to impart wavelength andpolarization selectivity to such de-vices. On a similar basis, photocur-rent enhancement has recentlybeen achieved for inorganic nano-wires. Here, the polarization sensi-tivity and charge carrier generationin a 50 nm Si nanowire could betailored by decorating its surfacewith plasmonic Au nanoparticles.20

The local photocurrent enhance-ment observed in such sampleshas been assigned to the plasmonicnear-field response of the individualnanoparticles, complemented bybroad-band enhancement due tosurface-enhanced optical absorption.As an extension of separated

metal dots or particles, so-calledplasmonic clusters21 emergeas idealcandidate structures for amplifyingthe optical response of graphene.Such clusters exhibit a transparencywindow where scattering is sup-pressed and hot electron�hole pair

generation is the dominant lightabsorption mechanism. The asso-ciated strong enhancement of thenear-field22 has recently beenexploited for the direct excitationof electron�hole pairs in a gra-phene monolayer in direct proxi-mity to the plasmonic clusters.23

By collecting the light-inducedconduction electrons in graphene,efficient photodetectors can be ob-tained. The reported photodetectordevices, consisting of an array ofgold nanoparticle heptamers sand-wiched between two graphenelayers, have enabled the conversionof visible and near-infrared photonsinto electrons with an 8-fold en-hancement of the photocurrent, ascompared to graphene deviceswithout antennas. Moreover, thespectral sensitivity of the devicescould be tuned by the geometryof the plasmonic antenna as wellas the application of a gate bias.A similar strategy to control the

Another interesting

aspect of tip-enhanced

scanning photocurrent

microscopy is the

ability to explore the

effects of local

perturbations along the

nanotube axis.

Figure 2. Spatially resolved photocurrent response of a gold-coated CdS nano-wire segment. Reprinted from ref 25. Copyright 2012 American Chemical Society.(a) Atomic force microscopy image of the nanowire (diameter of ∼50 nm)contacted by one gold and one titanium electrode, separated by a fewnanometers. The black frame marks the area where the photocurrent map inpanel (b) was recorded (λlaser = 488 nm). The electrode edges are highlighted by ablack solid line. (c) Photocurrent profile along the dotted line in panel (b). In thefirst regime (I), the photocurrent is almost constant, whereas in the secondregime (II), an approximately linear decrease in photocurrent with increasingdistance from the electrode edge is observed. Regimes I and II are attributable todrift and diffusion currents, respectively.

PERSPECTIV

E

BURGHARD AND MEWS VOL. 6 ’ NO. 7 ’ 5752–5756 ’ 2012

www.acsnano.org

5755

light�matter interactions in gra-phene involves its implementationinto an optical microcavity.24 Themicrocavity-induced optical con-finement yields a 20-fold enhance-ment of photocurrent, combinedwith the capability to control boththe thermal light emission fromgraphene and the electrical trans-port characteristics of an integratedgraphene transistor.

In addition to antenna-based de-vices, the SPCM response of inor-ganic nanowires or carbon nano-structures below the metal contactsdeserves further attention. For in-stance, it has recently been foundthat the photoresponse of theSchottky-type contact betweengold and a CdS nanowire con-tains spatially separate contribu-tions of diffusion and drift currents(Figure 2).25 This has enabled the

determination of the electron diffu-sion length and recombination ratein the metal-coated nanowire sec-tion. Moreover, evidence has beengained that the charge carriers canbe efficiently extracted out of thecontacts, which has implications forthe further development of nano-wire-based solar cells or photode-tectors. In the case of gold-coatedgraphene, the photoresponse hasprovided evidence that the elec-tronic structure, the electron�pho-non coupling, and the doping levelare largely preserved.26 Further, thetransfer lengths for electrons andholes at the graphene�gold con-tact could be determined to be ashigh as 1.6 μm, which poses limita-tions on the scaling of gold contactswithin graphene devices.The aforementioned examples

highlight a wide variety of physicaleffects that are observable in hybridnanostructures. These effects canlead to a wide range of new materi-als properties, opening numerouspossible applications. In this re-spect, the work of Rauhut et al. un-derscores that detailed knowledgeof the physical effects is required todevelop new analytical tools for thenanoscale. At the same time, it be-comes apparent that only combina-tions of several different techniquessuch as confocal optical microscopyand spectroscopy enhanced by nano-scopic antennas, as well as topogra-phy measurements and electronic

transport experiments, are ade-quate to explore the details of com-plex nanoscopic devices. An addi-tional direction is to use the tip as atool to measure the local electronicproperties. To mention just two ex-amples, electrostatic force micro-scopy (EFM) is well-suited to probethe charge distribution along semi-conductor nanowires (Figure 3),27

and Kelvin probe force microscopy(KPFM) enables directmonitoring ofthe potential profile along nano-structures.28 Additionally, the localprobes could be used tomanipulatethe electronic properties of indivi-dual nanostructure, either by localinjection of charges or by applyinglocal electrostatic field.

Conflict of Interest: The authors declareno competing financial interest.

REFERENCES AND NOTES1. Marek, J. Light-Beam Induced Cur-

rent Characterization of GrainBoundaries. J. Appl. Phys. 1984, 55,318–326.

2. Vorster, F. J.; van Dyk, E. E. Bias-Dependent High Saturation SolarLBIC Scanning of Solar Cells. Sol.Energy Mater. Sol. Cells 2007, 91,871–876.

3. Balasubramanian, K.; Fan, Y. W.; Bur-ghard, M.; Kern, K.; Friedrich, M.;Wannek, U.; Mews, A. Photoelectro-nic Transport Imaging of IndividualSemiconducting Carbon Nano-tubes. Appl. Phys. Lett. 2004, 84,2400–2402.

4. Balasubramanian, K.; Burghard, M.Charge Transport through CarbonNanotubes Interacting with Light.Semicond. Sci. Technol. 2006, 21,S22–S32.

Figure 3. Measurement of the charge distribution in CdSe semiconductor nanowires upon local laser illumination. Reprintedfrom ref 27. Copyright 2011 American Chemical Society. The nanowires are excited with a focused laser spot (λlaser = 488 nm)in a confocal microscope, and the resulting charge imbalance is imaged by electrostatic force microscopy (EFM) at the sametime. In the respective EFM images (upper graphs) and EFM simulations (lower graphs), a bright color corresponds to positivecharging while a dark color corresponds to negative charging. Because photogenerated electrons and holes have differentmobilities within the one-dimensional nanowire, the electrons spread almost uniformly along thewire, whereas the holes aremore localized at the excitation spot. This leads to positive excess charges at the excitation spot.

In addition to antenna-

based devices, the

scanning photocurrent

microscopy response of

inorganic nanowires or

carbon nanostructures

below the metal

contacts deserves

further attention.

PERSPECTIV

E

BURGHARD AND MEWS VOL. 6 ’ NO. 7 ’ 5752–5756 ’ 2012

www.acsnano.org

5756

5. Freitag, M.; Tsang, J. C.; Bol, A.; Yuan,D. N.; Liu, J.; Avouris, P. Imaging ofSchottky Barriers and Charge Deple-tion in Carbon Nanotube Transis-tors. Nano Lett. 2007, 7, 2037–2042.

6. Lee, E. J. H.; Balasubramanian, K.;Dorfmüller, J.; Vogelgesang, R.; Fu,N.; Mews, A.; Burghard, M.; Kern, K.Electronic-Band-Structure Mappingof Nanotube Transistors by ScanningPhotocurrent Microscopy. Small2007, 3, 2038–2042.

7. Balasubramanian, K.; Burghard, M.;Kern, K.; Scolari, M.; Mews, A. Photo-current Imaging of Charge Trans-port Barriers in Carbon NanotubeDevices. Nano Lett. 2005, 5, 507–510.

8. Lee, E. J. H.; Balasubramanian, K.;Burghard, M.; Kern, K. Spatially Re-solvedPotentialDistribution inCarbonNanotube Cross-Junction Devices.Adv. Mater. 2009, 21, 2720–2724.

9. Ahn, Y.; Dunning, J.; Park, J. ScanningPhotocurrent Imaging and Elec-tronic Band Studies in Silicon Nano-wire Field Effect Transistors. NanoLett. 2005, 5, 1367–1370.

10. Graham, R.; Miller, C.; Oh, E.; Yu, D.Electric Field Dependent Photocur-rent Decay Length in Single LeadSulfide Nanowire Field Effect Transis-tors. Nano Lett. 2011, 11, 717–722.

11. Lee, E. J. H.; Balasubramanian, K.;Weitz, R. T.; Burghard, M.; Kern, K.Contact and Edge Effects in Gra-phene Devices. Nat. Nanotechnol.2008, 3, 486–490.

12. Mueller, T.; Xia, F.; Freitag, M.; Tsang,J.; Avouris, P. Role of Contacts inGraphene Transistors: A ScanningPhotocurrent Study. Phys. Rev. B2009, 79, 245430.

13. Rauhut, N.; Engel, M.; Steiner, M.;Krupke, R.; Avouris, P.; Hartschuh,A. Antenna-Enhanced Photocur-rent Microscopy on Single-WalledCarbon Nanotubes at 30 nm Re-solution. ACS Nano 2012, 10.1021/nn301979c.

14. Mews, A.; Koberling, F.; Basche, T.;Philipp, G.; Duesberg, G. S.; Roth, S.;Burghard, M. Raman Imaging of Sin-gle Carbon Nanotubes. Adv. Mater.2000, 12, 1210–1214.

15. Hartschuh, A.; Pedrosa, H. N.; No-votny, L.; Krauss, T. D. SimultaneousFluorescence and Raman Scatteringfrom Single Carbon Nanotubes.Science 2003, 301, 1354–1356.

16. Hartschuh, A.; Qian, H.; Meixner, A. J;Anderson, N.; Novotny, L. NanoscaleOptical Imaging of Excitons in Sin-gle-Walled Carbon Nanotubes.Nano Lett. 2005, 5, 2310–2313.

17. Park, J.; Ahn, Y. H.; Ruiz-Vargas, C.Imaging of Photocurrent Generationand Collection in Single-Layer Gra-phene.Nano Lett. 2009, 9, 1742–1746.

18. St-Antoine, B. C.; Ménard, D.;Martel, R.Photothermoelectric Effects in Single-Walled Carbon Nanotube Films: Re-interpreting Scanning PhotocurrentExperiments. Nano Res. 2012, 5,73–81.

19. Echtermeyer, T. J.; Britnell, L.; Jasnos,P. K.; Lombardo, A.; Gorbachev, R .V.;Grigorenko, A. N.; Geim, A. K.; Ferrari,A. C.; Novoselov, K. S. Strong Plas-monic Enhancement of Photovol-tage in Graphene. Nat. Commun.2011, 2, 458.

20. Hyun, J. K.; Lauhon, L. J. SpatiallyResolved Plasmonically EnhancedPhotocurrent from Au Nanoparti-cles on a Si Nanowire. Nano Lett.2011, 11, 2731–2734.

21. Lukyanchuk, B.; Zheludev, N. I.;Maier, S. A.; Halas, N. J.; Nordlander,P.; Giessen, H.; Chong, C. T. The FanoResonance in Plasmonic Nanostruc-tures andMetamaterials. Nat. Mater.2010, 9, 707–715.

22. Stockman, M. I. Nanoscience: Dark-Hot Resonances. Nature 2010, 467,541–542.

23. Fang, Z.; Liu, Z.; Wang, Y.; Ajayan,P. M.; Nordlander, P.; Halas, N. J.Graphene-Antenna Sandwich Photo-detector. Nano Lett. 2012, 10.1021.nl301774e.

24. Engel, M.; Steiner, M.; Lombardo, A.;Ferrari, A. C.; von Löhneysen, H.;Avouris, P.; Krupke, R. Light�MatterInteraction in a Microcavity-Controlled Graphene Transistor.Nat. Commun. 2012, 10.1038/ncomms1911.

25. Dufaux, T.; Burghard, M.; Kern, K.Efficient Charge Extraction out ofNanoscale Schottky Contacts toCdS Nanowires. Nano Lett. 2012,12, 2705–2709.

26. Sundaram, R. S.; Steiner, M.; Chiu,H.-Y.; Engel, M.; Bo, A. A.; Krupke, R.;Burghard, M.; Kern, K.; Avouris, P.The Graphene�Gold Interface andIts Implications for Nanoelectronics.Nano Lett. 2011, 11, 3833–3837.

27. Schäfer, S.; Wang, Z.; Zierold, R.;Kipp, T.; Mews, A. Laser-InducedCharge Separation in CdSe Nano-wires. Nano Lett. 2011, 11, 2672–2677.

28. Koren, E.; Rosenwaks, Y.; Allen, J. E.;Hemesath, E. R.; Lauhon, L. J. Non-uniform Doping Distribution AlongSilicon Nanowires Measured by Kel-vin Probe Force Microscopy andScanning Photocurrent Microscopy.Appl. Phys. Lett. 2009, 95, 092105.

PERSPECTIV

E