HSU ET AL. VOL. 6 ’ NO. 6 ’ 5150–5156 ’ 2012

www.acsnano.org

5150

May 01, 2012

C 2012 American Chemical Society

Passivation Coating on ElectrospunCopper Nanofibers for StableTransparent ElectrodesPo-ChunHsu,†HuiWu,† Thomas J. Carney,†MatthewT.McDowell,† YuanYang,† Erik C. Garnett,†Michael Li,†

Liangbing Hu,†,§ and Yi Cui†,‡,*

†Department of Materials Science and Engineering, Stanford University, Stanford, California 94305, United States and ‡Stanford Institute for Materials and EnergySciences, SLAC National Accelerator Laboratory, 2575 Sand Hill Road, Menlo Park, California 94025, United States. §Present address: Department of Materials Scienceand Engineering, University of Maryland, College Park, MD 20742.

Transparent electrodes are widely usedin optical-electronic devices such assolar cells, displays, organic light-

emitting diodes (OLED), and touch screens.1�3

While the performance of a transparentelectrode is generally determined by thesheet resistance (Rs) and the transparency(T), other properties also profoundly affectthe applicability of the material as a trans-parent electrode. For instance, durabilityagainst chemical corrosion and mechanicalstress are key characteristics for practicalapplications. Although the conventionaltransparent electrode material, tin-dopedindiumoxide (ITO), possesses high transpar-ency and conductivity, ITO is inherentlybrittle and therefore not stable undermechanical stress. Indeed, the sheet resis-tance of ITO thin film increases dramaticallywhen the strain exceeds 0.03 due to crackformation.4 This strain-induced failure notonly compromises the reliability of ITO butalso limits its applicability on flexible sub-strates. Moreover, the cost of manufactur-ing ITO-based transparent electrodes isdirectly proportional to the cost and scarcityof indium. In 2010, the price of elementalindium reached $550 per kilogram, andgiven the abundance of transparent elec-trodes inmodern optical-electronic devices,the cost of indium is only expected toincrease due to both supply and demandconsequences.5 Thus, it is of vital impor-tance to find an alternative material andengineer its nanostructure to improve itsperformance and reduce its cost in an effortto replace traditional ITO. Recently, re-searchers fabricated transparent electrodesfrom conductive polymers,6 carbon nano-tubes,7 graphenes,8 metal nanowires,9 andnanofibers10 and evaluated their performance.Conductive polymers, carbon nanotubes,

and graphene demonstrate better flexibilitythan ITO, but with sheet resistance typicallymore than 100 Ω/sq at 80% transparency,3

which is inferior to conventional ITO inmoderndevices such as solar cells andOLEDs.Although the conductivity can be furtherenhanced by proper doping, the subse-quent degradation of sheet resistance re-veals the instability of this doping method.11

Metal nanowires, generally made of silver,have low Rs/T ratio comparable with ITO

* Address correspondence toyicui@stanford.edu.

Received for review February 25, 2012and accepted May 1, 2012.

Published online10.1021/nn300844g

ABSTRACT

Copper nanofiber networks, which possess the advantages of low cost, moderate flexibility,

small sheet resistance, and high transmittance, are one of the most promising candidates to

replace indium tin oxide films as the premier transparent electrode. However, the chemical

activity of copper nanofibers causes a substantial increase in the sheet resistance after thermal

oxidation or chemical corrosion of the nanofibers. In this work, we utilize atomic layer

deposition to coat a passivation layer of aluminum-doped zinc oxide (AZO) and aluminum

oxide onto electrospun copper nanofibers and remarkably enhance their durability. Our

AZO�copper nanofibers show resistance increase of remarkably only 10% after thermal

oxidation at 160 �C in dry air and 80 �C in humid air with 80% relative humidity, whereas bare

copper nanofibers quickly become insulating. In addition, the coating and baking of the acidic

PEDOT:PSS layer on our fibers increases the sheet resistance of bare copper nanofibers by 6

orders of magnitude, while the AZO�Cu nanofibers show an 18% increase.

KEYWORDS: transparent electrodes . electrospinning . metal nanofibers .atomic layer deposition . surface passivation

ARTIC

LE

Dow

nloa

ded

via

CO

LU

MB

IA U

NIV

on

Aug

ust 2

1, 2

018

at 2

0:39

:12

(UT

C).

Se

e ht

tps:

//pub

s.ac

s.or

g/sh

arin

ggui

delin

es f

or o

ptio

ns o

n ho

w to

legi

timat

ely

shar

e pu

blis

hed

artic

les.

HSU ET AL. VOL. 6 ’ NO. 6 ’ 5150–5156 ’ 2012

www.acsnano.org

5151

thin films,12 but the high price of silver remains anobstacle for future large-scale use. Recently, we de-monstrated the novel idea of electrospinning coppernanofibers as a transparent electrode to achieve ex-cellent transparency and sheet resistance with rela-tively cheap material and processing. This techniqueshows great potential in replacing ITO as the conven-tional transparent electrode; however, due to the largesurface area of nanofibers, the lack of stability ofcopper against thermal oxidation and chemical corro-sion remains a concern. For example, in our early study,the performance of electrospun Cu nanofibers de-grades after it is employed in an organic photovoltaiccell, mainly due to the acidic poly(3,4-ethylenedioxy-thiophene):poly(styrenesulfonate) (PEDOT:PSS), acommonly used polymeric hole conductor in organicphotovoltaic cells. The corrosion and oxidation candecrease the effective cross section area of Cu nanofi-bers and cause the conductivity to diminish.10 Giventhe high performance and low cost of electrospun Cunanofibers as a transparent electrode, it is crucial toenhance the chemical stability of copper nanofibers tofurther strengthen its application in practical devices.In this paper, we sought to address this problem byconstructing a conformal passivation coating on theoutside of copper nanofibers. We expect that thepassivation coating developed here can also be ex-tended to other nanowire network technology.Atomic layer deposition (ALD) features in its con-

formal coating in comparison with other depositiontechniques.13 By purging precursors and adsorbingmolecular monolayers in sequence, the growth rateof a thin film can be precisely limited to only onemolecular layer at one time. Owing to this self-limitingmechanism, ALD can achieve a highly conformal anddense metal oxide coating without pinholes even onsurfaces with high aspect ratio14 or high curvature.15

These features make ALD an ideal tool for buildingpassivation layers on nanostructuredmaterials. Herein,we introduce the ALD coatings on electrospun Cunanofibers for the purpose of enhancing the chemicalresistance against corrosion and oxidation. In ourstudy, the profound distinction of stability betweenbare Cu nanofibers and AZO�Cu nanofibers clearlyproves the effectiveness of our ALD-based protectiontechnique, which can extend the applicability of anelectrospun Cu nanofiber transparent electrode.

RESULTS AND DISCUSSION

Cu Nanofiber Synthesis and ALD Coating. Electrospinningis one of the most facile and cost-effective techniquesto synthesize ultralong one-dimensional nanomaterialsdue to its large yield and relatively cheap equipment.16

Typically, a viscous polymer solution is loaded insidea syringe and pushed out by a syringe pump. As adroplet of the solution appears at the needle tip, anelectrical field is applied between the needle and the

grounded substrate. When the electrical force exertedon the solution overcomes the surface tension, thespherical droplet become cone-shaped, which isknown as Taylor cone, and nanofibers are expelledfrom the tip of the needle.17,18 In addition to fabricatingpolymer nanofiber networks, electrospinning has beenalso employed as a method to produce carbon,19

metal,20 ceramics,21 and composite nanofibers.22 Re-cently, we introduced electrospinning to fabricate Cunanofiber webs as a transparent electrode material.10

The high ratio of transmittance to sheet resistance,excellent flexibility, and simple experimental setup areadvantageous for numerous transparent electrodeapplications. However, the sheet resistance of Cunanofibers gradually increases after being exposed inair.10 Unlike aluminum oxide which is capable ofpassivating the underlying aluminum, copper oxidegrows as an island and then coalesces into the filmwhich tends to have pinholes and microcracks.23 Thusthe oxidation can often spread through the entirecopper nanofiber and cause a significant drop inconductivity. Furthermore, the large surface area ofcopper nanofibers compounds the problem of con-ductivity loss as compared to a bulk or thin film of Cubecause the large surface area provides more sites forCuxO island formation. To solve this reliability problem,we proposed three routes of metal passivation: noblemetal coating, alloying, and oxide layer coating.Among those routes, noble metals, such as gold orplatinum, are too expensive, and therefore, the advan-tage of making a cheap transparent electrode solelyfrom cheap, earth-abundant copper is sacrificed forperformance. Alloying employs a different metal toform a dense and continuous oxide layer for surfacepassivation; however, transparent electrodes requirehigh conductivity, which limits the alloying concentra-tion and therefore cannot guarantee continuous coat-ing at nanoscale. Consequently, coating a uniform,dense, and inert oxide layer to protect Cu nanofibersis the most feasible and effective method for corrosionand oxidation prevention.

The principle of our experiment is illustrated inFigure 1a. We employ the passivation layer on electro-spun Cu nanofibers and compare the reliability withbare Cu nanofibers. This passivation layer can shieldthe Cu nanofibers from oxidation and corrosion andthus maintain its electrical and optical performance inharsh environments.We choose theALDprocess as ourcoating method for its capability of creating a con-formal layer, as compared with other coating techni-ques, such as sputtering, evaporation, and chemicalvapor deposition (CVD). The self-limiting, surface-controlled growth mechanism of ALD allows only onemolecular layer to coat onto Cu nanofibers at anysingle time; thus, the film growth rate is identical acrossthe surface area of the Cu nanofibers. For transparentelectrode applications, we chose AZO as our protecting

ARTIC

LE

HSU ET AL. VOL. 6 ’ NO. 6 ’ 5150–5156 ’ 2012

www.acsnano.org

5152

oxide layer due to its transparency, conductivity, andlow cost. However, one possible drawback of AZOcoating is its low chemical stability against acid corro-sion. Anticipating this potential problem, a 1 nm Al2O3

layer was coated onto AZO to improve its acid resis-tance without sacrificing its previously mentionedadvantages. The core�shell structure is illustrated inFigure 1b. Althoughwe expect someof the Al to diffuseinto the AZO film, nevertheless, pure Al2O3 is provedto exist on the outside by our experimental chemicalcorrosion analysis. It is also worth noting that Al2O3

and AZO share the same precursors and experimentalsetup during ALD, which simplifies this dual-layerprocess. Moreover, the hydroxyl groups on the surfaceof the AZO layer are reactivewith the organoaluminumprecursors, namely, trimethylaluminum. These hydro-xyl groups are able to provide efficient nucleation andensure that the coalescence of nuclei is good enoughfor a conformal coating, which is not the case ifthe ultrathin Al2O3 layer is deposited directly onto Cunanofibers.24

Morphology and Chemical Analysis. The morphology ofCu nanofibers and AZO�Cu nanofibers was character-ized by scanning electron microscopy (Figure 2a,b).The fusion at the junctions, which remarkably de-creases the resistance, can be observed in both sam-ples. In AZO�Cu nanofibers, the junctions are rein-forced by the AZO/Al2O3 layer and hence may furtherimprove the conductivity. Furthermore, the coatingholds Cu nanofibers to the glass substrate strongerthan bare Cu nanofibers and therefore improves themechanical durability. To demonstrate this phenom-enon, we applied Scotch adhesive tape on both

samples and then peeled off the tape. All of the bareCu nanofibers were removed by the tape, and thesample became insulating (>20 MΩ), whereas AZO�Cunanofiber sheet resistance only increased by a factor of 6.TEM and energy-dispersive X-ray spectroscopy (EDS)

Figure 1. (a) Schematic of AZO�Cu nanofiber preparation and comparison with bare Cu nanofiber in terms of oxidationresistance. Because of the passivation of the AZO/Al2O3 layer, the core Cu nanofiber is not subjected to oxidation andtherefore can maintain its high conductivity. (b) Cartoon (not to scale) of the Cu@AZO/Al2O3 nanofiber.

Figure 2. Characterization of bare Cu and AZO�Cu nanofi-bers: SEM images of bare Cu (a) and AZO�Cu (b) nanofibersexhibit fusion at junctions. TEM images and diffractionpatterns (insets) of Cu nanofibers (c) and AZO�Cu (d)nanofibers. EDS line scan profile and the correspondingdark-field image of Cu/AZO/Al2O3 nanofiber (e).

ARTIC

LE

HSU ET AL. VOL. 6 ’ NO. 6 ’ 5150–5156 ’ 2012

www.acsnano.org

5153

were employed to characterize the phases, crystal-linity, and chemical composition of the nanofibers.Figure 2c shows the polycrystalline nature of Cu nano-fibers. The diffraction pattern indicates that the nano-fibers are composed of not only Cu but also Cu2O,which means the native oxide has already formedduring the exposure in air. It is also the formation ofCu2Owhich explains the irregular grains on the surface,as shown in Figure 2c. This result microscopicallydemonstrates that Cu nanofibers are inherently vulner-able to atmospheric oxidation. As for AZO�Cu, onlyface-centered cubic (fcc) Cu and ZnO are found in thesample (Figure 2d inset). Taking the chemical activity ofbare Cu nanofibers into account, it is reasonable toinfer that, if Cu nanofibers are stored in air for too longbefore ALD passivation, Cu2O should be found in thesample, as well. Consequently, it is critical to reduce theexposure time of Cu nanofibers with air in order tomaximize the conductivity of the fibers.

In the TEM image of AZO�Cu nanofibers (Figure 2d),the contrast line along the axial direction is the bound-ary between Cu and AZO/Al2O3 coating. The thicknessof the coating is approximately 25 nm, and the di-ameter of core Cu nanofiber is around 100 nm. EDS linescan (Figure 2e) is also used to examine the elementaldistribution. The two peaks at the edges of the Znconcentration profile clearly show the shell structure ofthe AZO/Al2O3 layer. By measuring the width of thepeak, we are able to confirm the thickness of the AZO/Al2O3 layer to be the same as observed in microscopeimages.

Optical and Electronic Property. A major concern sur-rounding our method is how the resistance of AZO/Al2O3 influences the transmittance and sheet resis-tance of Cu nanofiber electrodes. To examine this, wepatterned electrical contacts by gold evaporation onthe samples for electrical measurement. Upon compar-ing the sheet resistance both before and after ALD, wefind the resistance to be the same at 47.1 Ω/sq. Thisindicates the AZO/Al2O3 layer does not block theelectrical current, and the ALD process itself does notdegrade the conductivity of Cu nanofibers. It is worthnoting that the conductance of nanostructure trans-parent electrodes is typically attributed to percolationtheory, which assumes that only nanofibers whichform a continuous pathway between electrical con-tacts contribute to the conductance, while dead-endnanofibers do not.25 Although the sheet resistancedoes not change here, it is reasonable to argue thatthe implementation of AZO can connect those dead-end Cu nanofibers together and enable them to com-plete the electrical circuit and possibly increase thetotal conductance. The thickness of Al2O3 layer is only1 nm, which is thin enough for carrier tunneling, andthe AZO layer is also thin and inherently conductive, socarrier transport from the active layer to the transpar-ent electrode will also not be blocked. Moreover,

optical measurements (Figure 3a) show less than a1% decrease in transmittance throughout the entirevisible light wavelength range. Thus, we conclude thechoice of transparent AZO and ultrathin Al2O3 as ourcoating has no effect on the performance of Cunanofibers.

Thermal Oxidation Stability. AZO/Al2O3-coated Cu na-nofibers exhibit significant enhancement of thermaloxidation resistance. Cu nanofibers were tested at 80,120, and 160 �C. To show the effect of degradation, thesheet resistance was normalized using its originalvalue. Figure 3b shows that all bare Cu nanofibersbecome more insulating than their AZO-coated coun-terpart. In particular, the sheet resistance of bare Cunanofibers increases dramatically at 120 and 160 �C.The AZO�Cu nanofibers, on the other hand, showincreases of sheet resistance of only 10%, even attemperatures as high as 160 �C for 8 h, as comparedto the bare Cu nanofibers which become insulating(>20MΩ) after only 40min. This profound difference inthermal oxidation stability demonstrates the effective-ness of the AZO/Al2O3 layer in enhancing the durabilityof Cu nanofiber transparent electrodes for practicalapplication. As shown in Figure 3c, after baking thefibers at 120 �C for 7 h, the bare Cu nanofibers becomelight brown, suggesting the oxidation of Cu, while theAZO�Cu nanofibers remain red, indicating a lack ofoxidation. At ambient temperature, the electrical per-formance of Cu nanofibers degrades relatively slowly(Figure 3d), but the sheet resistance still increases 53%after 25 days. The two sudden increases of sheetresistance might correspond to the change in ambienthumidity because water vapor can greatly acceleratethe oxidation of copper metal.26 Transmission electronmicroscopy characterization (Figure 3e) shows that Cunanofibers are oxidized into Cu2O after baking at 120 �Cfor more than 3 h, during which the sheet resistancerises to more than 20 MΩ. We find no fcc Cu inside thesample, although we do observe unidentified diffrac-tion peaks which are probably due to the mixture ofcopper carbonate and copper hydroxide, the twomainproducts of copper oxidation in air. In contrast, forAZO�Cu nanofibers, the electron image and diffrac-tion pattern remain the same as before heating(Figure 3f), indicating our thermal oxidation does notaffect the AZO/Al2O3-protected Cu nanofibers. EDSscanning profile (Figure S1 in Supporting Information)verifies that the core�shell structure does not changeafter thermal treatment. Therefore, we conclude thatCu nanofibers are preserved inside the AZO/Al2O3 layerand maintain their high conductivity.

After oxidation, most of the oxidized Cu nanofibersretain their polycrystallinity and morphology ratherthan transforming into Cu2O nanotubes as in the caseof Cu single-crystalline nanowires.27 The formationof CuO or Cu2O nanotubes is primarily due to theKirkendall effect, which arises from the difference in

ARTIC

LE

HSU ET AL. VOL. 6 ’ NO. 6 ’ 5150–5156 ’ 2012

www.acsnano.org

5154

diffusivity. Since copper diffuses faster than oxygen inits oxide phase, voids will form in the copper region. Ina one-dimensional nanostructure, the accumulationand coalescence of voids eventually forms a hollowstructure inside and becomes a CuO nanotube. Thereason why Kirkendall effect is not observed in electro-spun Cu nanofibers lies in the crystallinity of our fibers.For the electrospun polycrystalline Cu nanofibers, oxi-dation begins at the grain boundary rather than thesurface of nanofibers because of the dominating diffu-sivity along the grain boundary. SEM and TEM imagesshow that the grain size of the Cu nanofiber is less than40 nm. As a result, vacancies tend to diffuse to the grainboundaries, and the vacancy concentration inside thegrain will not be high enough to form voids or at leastnot experimentally distinguishable under TEM.

The influence of humidity on sheet resistance in-crease was also tested. As shown in Figure 3g, thetemperature and relative humidity were controlled at80 �C and 80 ( 10%, respectively. Under this environ-ment, bare Cu nanofibers became insulating in lessthan 60 min, due to the fact that a large amount ofwater vapor molecules is adsorbed onto the surfaceand accelerates the oxidation.28 In comparison with

testing in 80 �C dry air (Figure 3b) in which the sheetresistance of bare Cu nanofibers increases 22% after700 min, humidity does play an important role in Cuthermal oxidation. As for AZO/Al2O3-protected Cunanofibers, the sheet resistance increases by only4.6% in humid air and 1% in dry air. Although theAZO�Cu nanofibers are still affected by the humid air,the passivation layer still demonstrates a significantimprovement in performance as compared to theextremely high oxidation rate of bare Cu nanofibersin a humid environment.

Chemical Corrosion Stability. For transparent electrodeapplications, the chemical stability against an acidicenvironment is of great interest. In particular, thecorrosion of a transparent electrode by acidic PEDOT:PSS can degrade the interface and thus reduce theperformance of the transparent electrode.29 This phe-nomenon is compounded in Cu nanofibers as com-pared to ITO because of the greater chemical activity ofCu. Moreover, the baking process can accelerate theoxidation of the nanofibers, which results in a dramaticincrease of sheet resistance. To test the corrosionresistance of our fibers, we immersed the bare Cuand AZO�Cu nanofibers into aqueous solution of

Figure 3. (a) Transmittance of electrospun Cu nanofibers before and after ALD. Note that the transmittance is measured fromthe same sample, and sheet resistance does not change after the ALD process. (b) Normalized sheet resistance versus bakingtime at 80, 120, and 160 �C. Solid circles refer to bare Cu nanofibers, and hollow circles represent AZO�Cu nanofibers. In thecase of bare Cu nanofibers at 120 and 160 �C, the resistance exceedsmultimeter limit (20MΩ) in the end. (c) Photos of Cu andAZO�Cu nanofibers after baking in dry air. The color change reveals the oxidation of Cu nanofibers. The edge length of onegrid cell is 0.2 in. (d) Sheet resistance change versus time under ambient temperature and atmosphere sheet resistance. (e,f)TEM images and diffraction patterns of Cu and AZO�Cu nanofibers, respectively. (g) Humidity test data.

ARTIC

LE

HSU ET AL. VOL. 6 ’ NO. 6 ’ 5150–5156 ’ 2012

www.acsnano.org

5155

various pH values for 1 min (Figure 4a). In the acidicsolutions, the AZO�Cu nanofibers exhibit remarkableresistance against corrosion as compared to the poorperformance of the bare Cu fibers; while in DI waterand basic solution, the sheet resistance of both Cu andAZO�Cu nanofibers does not change significantly.SEM images (Figure 4b,c) also demonstrate that bareCu nanofibers suffer severe etching, which shrinks theeffective cross section area on surface active sites oreven breaks the entire nanofiber, whereas AZO�Cunanofibers do not show morphology change afteretching, further confirming the stability of the AZO/Al2O3 passivation layer.

Besides the testing under various pH environments,the impact of the PEDOT:PSS coating process is alsoinvestigated. PEDOT:PSS aqueous solution (pH∼3)wasspin-coated onto the nanofibers on glass substrates

followed by 120 �C baking in air for 15 min. The sheetresistance of Cu nanofibers increased by a factor of5.26 � 106, whereas AZO�Cu nanofibers only showedan 18% increase. It is important to note that thenanofibers were subjected to a relatively harsh envi-ronment through this test. In practice, this degradationcaused by PEDOT:PSS can be mitigated by choosing apH-neutral PEDOT:PSS30 or baking in vacuum to avoidsubsequent thermal oxidation, and thus AZO�Cu nano-fibers should be able to withstand the corrosion by thePEDOT:PSS coating process to an even higher level.

CONCLUSIONS

In summary, we demonstrated that using an ALDtechnique to coat an AZO/Al2O3 passivation layer ontoelectrospun Cu nanofibers can significantly enhancethe fibers' stability against oxidation and corrosion.This approach hinges on the chemical stability of theouter Al2O3 ultrathin film and the conductivity andtransparency of AZO to solve the reliability problem ofa Cu nanostructure-based transparent electrode. Thispassivation layer has a remarkable effect in preventingCu nanofibers from thermal oxidation at high tempera-ture of up to 160 �C in dry air and 80 �C in humid air.After coating the nanofibers with PEDOT:PSS, the sheetresistance of AZO�Cu nanofibers increases only 18%,while bare Cu nanofibers increase by 5.26� 106. Giventhe extraordinary performance and cost-effectivenessof electrospun Cu nanofibers as transparent electro-des, this passivation method provides a critical advan-tage for Cu nanofibers to replace ITO in industry andcommercial use. In addition, our study shows thatbuilding an ALD passivation layer on material surfacesis also a general and effective route to improve thechemical stabilities of other nanostructured metals.

METHODSElectrospinning. The polymer solution was prepared by dis-

solving polyvinyl alcohol (PVA) (avg. Mw = 95 000 g/mol, Acros)in distilled water with a concentration of 10 wt %. The PVAsolution was stirred and heated at 85 �C for 3 h until a clear anduniform solution was formed. Copper(II) acetate monohydrate(CuAc2) (Aldrich) was added into the PVA solution to a concen-tration of 10 wt %. The solution was again stirred and heated at85 �C for 3 h forming a turquoise gel solution. After the CuAc2dissolved, the saturated solution was cooled to room tempera-ture and held still for 1 day. The CuAc2/PVA solution was elec-trospun using a 1 mL syringe with a 22 gauge, blunt, stainlesssteel needle, flow rate of 0.1 mL/h (KDS200, KD Scientific Inc.),and applied voltage of 0.6 kV/cm (Gamma High VoltageResearch). The collector used for the electrospinning processwas aluminum foil with a glass slide on top. These parametersproduced a Taylor cone at the needle tip, although slight flowrate adjustment was needed to maintain a stable Taylor cone.

Calcination and Reduction of Nanofibers. The CuAc2/PVA nanofi-bers were calcinated using the tube furnace (Lindberg/Blue M)under a nitrogen/oxygen gas mixture at 1 atm. The flow rate ofnitrogen and oxygen was 40 and 6 sccm, respectively. Thenanofibers were then heated to 500 �C at a rate of 5 �C/min.Next, the nanofibers were held at 500 �C for 2 h and then cooled

to ambient temperature. In this step, the CuAc2/PVA nanofiberswere calcinated into CuxO nanofibers. To reduce the CuxOnanofibers to Cu nanofibers, the nanofibers were then heatedin an atmosphere of hydrogen at a pressure of 40 Torr and flowrate of 200 sccm to 300 �C at 2 �C/min and held at 300 �C for30 min.

Atomic Layer Deposition. The conformal AZO layers werecoated at 150 �C using the Cambridge Nanotech Savannahwith trimethyl aluminum (TMA), diethyl zinc (DEZ), and DI wateras precursors. The pulse times for DEZ, TMA, and DI water wereall 15ms. A 20 nm thick AZO thin filmwas created by depositing25 cycles of ZnO followed by 1 cycle of Al2O3 and then repeatingthis process four times. The molecular monolayers of Al pre-cursors dope the ZnO to form a uniform AZO film. After the AZOfilm was deposited, another 10 cycles of Al2O3 was deposited tocreate a 1 nm layer of Al2O3.

Characterization. Gold contacts (180 nm thick) were coatedonto the Cu nanofibers' transparent electrode using a thermalevaporator (MBraun). To reinforce the adhesion of the goldcontact, a 7 nm layer of chromium was deposited before thegold deposition process. The pattern of gold contacts wasdesigned to produce a square area of Cu nanofibers for sheetresistance measurement. Electrical measurements were per-formed using a multimeter, and the transmittance measurements

Figure 4. (a) Normalized sheet resistance after acid or baseetching. (b,c) SEM imagesof bareCu andAZO�Cunanofiberafter etching in acidic (pH 3) solution for 1 min.

ARTIC

LE

HSU ET AL. VOL. 6 ’ NO. 6 ’ 5150–5156 ’ 2012

www.acsnano.org

5156

were conductedusing an integrating sphere (Newport) and xenonlamp coupled with a monochromator. All transmittance measure-ments were referenced to a clear glass slide. Cu nanofibers werecharacterized by a SEM (FEI XL30 Sirion SEM, operating voltage5 kV) and a TEM (FEI Tecnai G2 F20 X-TWIN with EDS, operatingvoltage 200 kV) equipped with EDS for chemical analysis.

Oxidation and Corrosion Testing. The thermal oxidation testswere performed in air using an oven (Fisher Scientific). Eachsample was placed in the oven for various temperatures andtime, removed from the oven, and allowed to equilibrate withthe ambient temperature for 10 min. For the humidity testing,the humidity was measured and controlled by the hygrometer(Fisher Scientific) and oven (Barnstead Lab-line). For inorganicacid corrosion tests, the nanofibers were dipped into solutionswith various pH values adjusted by KOH or HCl for 1 min andthen rinsed with DI water twice to remove the residual etchant.The nanofibers were subsequently vacuum-dried for electricalmeasurement.

Conflict of Interest: The authors declare no competingfinancial interest.

Acknowledgment. P.-C.H. acknowledges the support fromInternational Fulbright Science and Technology Award.

Supporting Information Available: Additional experimentaldetails and figures. This material is available free of charge viathe Internet at http://pubs.acs.org.

REFERENCES AND NOTES1. Granqvist, C. G. Transparent Conductors as Solar Energy

Materials: A Panoramic Review. Sol. Energy Mater. Sol. Cells2007, 91, 1529–1598.

2. Hilsum, C. Flat-Panel Electronic Displays: A Triumph ofPhysics, Chemistry and Engineering. Philos. Trans. R. Soc., A2010, 368, 1027–1082.

3. Hecht, D. S.; Hu, L. B.; Irvin, G. Emerging TransparentElectrodes Based on Thin Films of Carbon Nanotubes,Graphene, and Metallic Nanostructures. Adv. Mater.2011, 23, 1482–1513.

4. Cairns, D. R.; Witte, R. P.; Sparacin, D. K.; Sachsman, S. M.;Paine, D. C.; Crawford, G. P.; Newton, R. R. Strain-Depen-dent Electrical Resistance of Tin-Doped Indium Oxide onPolymer Substrates. Appl. Phys. Lett. 2000, 76, 1425–1427.

5. Mineral Commodity Summaries. U.S. Geological Society,2011; p 74.

6. Yang, Y.; Heeger, A. J. Polyaniline as a Transparent Elec-trode for Polymer Light-Emitting-Diodes - Lower Operat-ing Voltage and Higher Efficiency. Appl. Phys. Lett. 1994,64, 1245–1247.

7. Wu, Z. C.; Chen, Z. H.; Du, X.; Logan, J. M.; Sippel, J.; Nikolou,M.; Kamaras, K.; Reynolds, J. R.; Tanner, D. B.; Hebard, A. F.;et al. Transparent, Conductive Carbon Nanotube Films.Science 2004, 305, 1273–1276.

8. Kim, K. S.; Zhao, Y.; Jang, H.; Lee, S. Y.; Kim, J. M.; Kim, K. S.;Ahn, J. H.; Kim, P.; Choi, J. Y.; Hong, B. H. Large-Scale PatternGrowth of Graphene Films for Stretchable TransparentElectrodes. Nature 2009, 457, 706–710.

9. Lee, J. Y.; Connor, S. T.; Cui, Y.; Peumans, P. Solution-Processed Metal Nanowire Mesh Transparent Electrodes.Nano Lett. 2008, 8, 689–692.

10. Wu, H.; Hu, L. B.; Rowell, M. W.; Kong, D. S.; Cha, J. J.;McDonough, J. R.; Zhu, J.; Yang, Y. A.; McGehee, M. D.; Cui,Y. Electrospun Metal Nanofiber Webs as High-Perfor-mance Transparent Electrode. Nano Lett. 2010, 10,4242–4248.

11. Kim, K. K.; Reina, A.; Shi, Y. M.; Park, H.; Li, L. J.; Lee, Y. H.;Kong, J. Enhancing the Conductivity of Transparent Gra-phene Films via Doping. Nanotechnology 2010, 21,285205.

12. Cui, Y.; Hu, L. B.; Kim, H. S.; Lee, J. Y.; Peumans, P. ScalableCoating and Properties of Transparent, Flexible, SilverNanowire Electrodes. ACS Nano 2010, 4, 2955–2963.

13. Leskela, M.; Ritala, M. Atomic Layer Deposition Chemistry:Recent Developments and Future Challenges. Angew.Chem., Int. Ed. 2003, 42, 5548–5554.

14. Elam, J. W.; Routkevitch, D.; Mardilovich, P. P.; George, S. M.Conformal Coating on Ultrahigh-Aspect-Ratio Nanoporesof Anodic Alumina by Atomic Layer Deposition. Chem.Mater. 2003, 15, 3507–3517.

15. Hong, S. S.; Cha, J. J.; Cui, Y. One Nanometer ResolutionElectrical Probe via Atomic Metal Filament Formation.Nano Lett. 2011, 11, 231–235.

16. Greiner, A.; Wendorff, J. H. Electrospinning: A FascinatingMethod for the Preparation of Ultrathin Fibres. Angew.Chem., Int. Ed. 2007, 46, 5670–5703.

17. Taylor, G. Disintegration of Water Drops in Electric Field.Proc. R. Soc. London, Ser. A 1964, 280, 383–397.

18. Deitzel, J. M.; Kleinmeyer, J.; Harris, D.; Tan, N. C. B. TheEffect of Processing Variables on the Morphology ofElectrospun Nanofibers and Textiles. Polymer 2001, 42,261–272.

19. Hou, H. Q.; Reneker, D. H. Carbon Nanotubes on CarbonNanofibers: A Novel Structure Based on Electrospun Poly-mer Nanofibers. Adv. Mater. 2004, 16, 69–73.

20. Wu, H.; Zhang, R.; Liu, X. X.; Lin, D. D.; Pan, W. Electrospinn-ing of Fe, Co, and Ni Nanofibers: Synthesis, Assembly, andMagnetic Properties. Chem. Mater. 2007, 19, 3506–3511.

21. Dai, H. Q.; Gong, J.; Kim, H.; Lee, D. A Novel Method forPreparing Ultra-fine Alumina-Borate Oxide Fibres via anElectrospinning Technique. Nanotechnology 2002, 13,674–677.

22. Sen, R.; Zhao, B.; Perea, D.; Itkis, M. E.; Hu, H.; Love, J.;Bekyarova, E.; Haddon, R. C. Preparation of Single-WalledCarbon Nanotube Reinforced Polystyrene and Polyur-ethane Nanofibers and Membranes by Electrospinning.Nano Lett. 2004, 4, 459–464.

23. Yang, J. C.; Kolasa, B.; Gibson, J. M.; Yeadon,M. Self-LimitingOxidation of Copper. Appl. Phys. Lett. 1998, 73, 2841–2843.

24. George, S. M. Atomic Layer Deposition: An Overview.Chem. Rev. 2010, 110, 111–131.

25. Hu, L.; Hecht, D. S.; Gruner, G. Percolation in Transparentand Conducting Carbon Nanotube Networks. Nano Lett.2004, 4, 2513–2517.

26. Haugsrud, R. The Influence of Water Vapor on the Oxida-tion of Copper at Intermediate Temperatures. J. Electro-chem. Soc. 2002, 149, B14–B21.

27. Chang, Y.; Lye, M. L.; Zeng, H. C. Large-Scale Synthesis ofHigh-Quality Ultralong Copper Nanowires. Langmuir2005, 21, 3746–3748.

28. Sharma, S. P. Adsorption of Water on Copper and CuprousOxide. J. Vac. Sci. Technol. 1979, 16, 1557–1559.

29. Ni, J.; Yan, H.; Wang, A. C.; Yang, Y.; Stern, C. L.; Metz, A. W.;Jin, S.; Wang, L.; Marks, T. J.; Ireland, J. R.; et al. MOCVD-Derived Highly Transparent, Conductive Zinc- and Tin-Doped Indium Oxide Thin Films: Precursor Synthesis,Metastable Phase Film Growth and Characterization, andApplication as Anodes in Polymer Light-Emitting Diodes.J. Am. Chem. Soc. 2005, 127, 5613–5624.

30. Wu, S.; Han, S. H.; Zheng, Y. N.; Zheng, H.; Liu, N. L.; Wang, L.;Cao, Y.; Wang, J. A. pH-Neutral PEDOT:PSS as Hole InjectionLayer in Polymer Light Emitting Diodes. Org. Electron.2011, 12, 504–508.

ARTIC

LE