A Full-Visible-Spectrum Invisibility Cloak for Mesoscopic Metal WiresSang-Woo Kim,†,§ Byeong Wan An,‡,§ Eunjin Cho,‡ Byung Gwan Hyun,‡ Yoon-Jong Moon,†

Sun-Kyung Kim,*,† and Jang-Ung Park*,‡

†Department of Applied Physics, Kyung Hee University, Gyeonggi-do 17104, Republic of Korea‡School of Materials Science and Engineering, Ulsan National Institute of Science and Technology (UNIST), Ulsan MetropolitanCity 44919, Republic of Korea

*S Supporting Information

ABSTRACT: Structured metals can sustain a very largescattering cross-section that is induced by localized surfaceplasmons, which often has an adverse effect on their use astransparent electrodes in displays, touch screens, and smartwindows due to an issue of low clarity. Here, we report abroadband optical cloaking strategy for the network ofmesoscopic metal wires with submicrometer to micrometerdiameters, which is exploited for manufacturing andapplication of high-clarity metal-wires-based transparentelectrodes. We prepare electrospun Ag wires with 300−1800nm in diameter and perform a facile surface oxidation processto form Ag/Ag2O core/shell heterogeneous structures. Theabsorptive Ag2O shell, together with the coating of a dielectric cover, leads to the cancellation of electric multipole moments inAg wires, thereby drastically suppressing plasmon-mediated scattering over the full visible spectrum and rendering Ag wires to beinvisible. Simultaneously with the effect of invisibility, the transmittance of Ag/Ag2O wires is significantly improved compared tobare Ag wires, despite the formation of an absorptive Ag2O shell. As an application example, we demonstrate that these invisibleAg wires serve as a high-clarity, high-transmittance, and high-speed defroster for automotive windshields.

KEYWORDS: Invisibility cloak, scattering suppression, multipole cancelation, Ag fibers, transparent electrodes

The negative permittivity (ε) of metals at opticalfrequencies enables them to function as a broadband

high-reflectivity mirror.1 More interestingly, structuring metalsat the nanometer to micrometer scale can induce localizedsurface plasmon resonances at specific wavelengths, whichconvert the energy of free propagating radiation into the energyof subwavelength-volume electromagnetic modes and viceversa; both phenomena are referred to as plasmonic scattering.2

The concept of plasmonic scattering has been utilized toimprove the performance of a diverse range of photonic devicesincluding optical sensors, lithography systems, solar cells, andlight-emitting diodes.3−6 However, plasmon-mediated scatter-ing often causes an unwanted situation, particularly for imagingapplications such as displays, touch screens, and smart windowsthat use metal wires as a transparent electrode; metal wires canbe vividly observed by the naked eye, thereby lowering theclarity of visibility,7 as illustrated in Figure 1a. The electric-fieldprofiles show that a bare metal wire significantly disturbs thephase of an incoming plane wave, whereas a cloaked metal wiremarginally interacts with an incoming plane wave. Suchdegraded clarity has impeded the extensive use of metal wiresthat are made of periodic grids, nanowires (NWs),8−13

nanofibers (NFs),14,15 or nanotroughs (NTs)16−19 of metalsas transparent electrodes, despite a high figure of merit (i.e., aratio of transmittance to sheet resistance). Therefore, there is a

need for the development of an economically viable strategy forconcealing such two-dimensional metal networks over the fullvisible spectrum while simultaneously retaining their hightransmittance and low sheet resistance.To date, optical cloaks prepared from rationally designed

metamaterials and metasurfaces have attracted significantattention for various industrial and military purposes.20−29

For example, the Zhang group demonstrated a metasurface filmusing phase delaying Au nanobars that cover up the informationon the curvature of an arbitrarily shaped surface.20 The Kocabasgroup demonstrated a graphene-based tunable absorberoperating at microwave frequencies to avoid radar detection.21

The Brongersma group demonstrated a 50 nm diameter Si-NW-based photodetector featuring reduced scattering bycoating a thin (<20 nm) conformal Au shell.22 Thesedemonstrations are based on the manipulation of thepropagation wave vector, absorption, or scattering of incominglight. Although the optical cloaks that have been developed thusfar have successfully opened up a new paradigm of nanoma-terial-enabled photonics, they have limitations such asoperation at a limited range of wavelengths and specific

Received: March 22, 2018Revised: April 29, 2018Published: May 16, 2018

Letter

pubs.acs.org/NanoLettCite This: Nano Lett. 2018, 18, 3865−3872

© 2018 American Chemical Society 3865 DOI: 10.1021/acs.nanolett.8b01157Nano Lett. 2018, 18, 3865−3872

Dow

nloa

ded

via

KY

UN

G H

EE

UN

IV o

n A

ugus

t 1, 2

018

at 0

0:16

:02

(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.

polarization. A few studies on broadband optical cloaks basedon the principle of transformation optics have beenreported;23−26 however, for the implementation of trans-formation optics, a macroscopic and sophisticated opticaldesign is required, which would not be appropriate for practicaldevices. Furthermore, for implementation in optoelectronicdevices, there is a lingering issue of whether cloaking structurescan preserve device performance; the aforementioned invisibleSi-NW photodetector would sacrifice the absorption efficiencyowing to the presence of a thin Au shell as a cloakingstructure.22 In this study, we report a facile cloaking strategythat conceals mesoscopic (i.e., subwavelength to wavelengthscale) metal wires over the full visible spectrum for twoorthogonal polarizations. It was believed that such giant metalstructures cannot be readily cloaked because of their large

electric multipole moments.28,29 Metal wires on the invisibilitycloak serve as an ultrahigh figure-of-merit transparent electrode,simultaneously maintaining the level of clarity as good as apristine glass substrate. Furthermore, we highlight that theinvisibility cloak studied herein is generalized for a wide rangeof sizes (a few hundred nanometers to a few micrometers) anddifferent shapes (solid NFs and hollow NTs) in metal wires. Asan application example, we demonstrate a high-clarity,transparent, and ultrafast defroster for automotive windshieldsby using the network of invisible metal wires.To prepare the mesoscopic metal wires used for our

invisibility cloak experiments, continuous networks of ultralongAg NFs were directly electrospun on a glass substrate by using asuspension of Ag nanoparticles as an ink (Figure 1b). Asubsequent thermal annealing process at 150 °C for 30 min in

Figure 1. Fabrication and optoelectronic characterization of Ag-NFs-based electrodes. (a) Schematics of the network of metal wires with low (left)or high (right) clarity. Typical profiles of the electric field of a bare Ag NF (left) and a cloaked Ag/Ag2O NF (right) are shown together. (b)Schematic illustration of the formation of Ag2O shells by exposure to UVO and the coating of an IML layer. SEM cross-sectional images of thesesamples. Scale bars, 500 nm. (c) Photographs of samples with random networks of bare Ag NFs, Ag/Ag2O NFs, and Ag/Ag2O NFs/IML on a glasssubstrate in typical room light (upper images) and focused high-intensity light with 251 mW/cm2 (lower images). The structural parameters of (D,t) are (300, 150) nm. Scale bars, 2 cm. (d) Measured transmittance (λ = 550 nm) and sheet resistance of bare Ag NFs, Ag/Ag2O NFs, and Ag/Ag2ONFs/IML electrodes as a function of D. The oxidized samples have the same t value of 150 nm.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.8b01157Nano Lett. 2018, 18, 3865−3872

3866

air coalesced the Ag nanoparticles, thereby transforming theminto electrically conductive Ag NFs. These electrospun Ag NFswere adequate for these experiments because they wereuniformly dispersed on a substrate without bunching orstacking. In addition, the diameter of the Ag NFs was readilytunable by adjusting the flow rate of the Ag nanoparticle ink.We successfully obtained Ag NFs with diameters (D) rangingbetween 300 and 1800 nm (Supporting Information FigureS1). This direct electrospinning method allows relatively rapidproduction of large-area (potentially larger than 1 m2) metal-wire-based transparent electrodes. Fabricated ultralong Ag NFsminimize the number of interconnections between NFs, leadingto a decrease of the sheet resistance despite small Ag areafractions; this geometrical factor also promises high trans-mittance. Although Ag NFs serve as a high figure-of-merittransparent electrode, which outperforms conductive oxidefilms, conducting polymers, graphene, metal grids, and evenrandom networks of Ag NWs,14 the large amount of scatteringinduced by individual NFs degrades the clarity of visibilitythrough a substrate, as shown in the left images of Figure 1c.To enhance the clarity of randomly networked Ag NFs

through a glass substrate, we propose Ag/oxide core/shell NFsthat were covered with an optically thick (∼7 μm) index-matching layer (IML) composed of SiO2 and binding polymers,as shown in Figure 1b. The surface of the Ag core was gentlyoxidized under an ultraviolet-ozone (UVO) illumination togenerate a Ag2O shell, as evident from the scanning electron

microscopy (SEM) images in Figure 1b and X-ray photo-electron spectroscopy (XPS) spectra in Supporting InformationFigure S2a,b. It is well-known that a Ag2O material has a largeextinction coefficient (k) over the full visible spectrum, aspreviously reported in ref 30 (Supporting Information FigureS2c). In general, a dielectric shell coating on metal orsemiconductor nanostructures can be leveraged to amplifytheir optical antenna effect (i.e., scattering cross section).31

Alternatively, metal/dielectric (or dielectric/metal) core/shellnanostructures can sustain reduced scattering cross sectionowing to their ε’s with opposite signs (ε < 0 for metal and ε > 0for dielectric).22,28,29 Particularly for mesoscopic objectsbeyond the quasistatic approximation, which is indeed ourcase, the design rules established by previous theoretical studiesindicate that a low-ε dielectric shell and a high-ε backgroundmedium are desirable to eliminate high-order scatteringcoefficients.28,29 These conditions are met well with the useof a highly absorptive Ag2O shell and a dielectric (n = 1.5) IML.The surfaces of as-grown Ag NFs and the following AgO2 shellare fairly rough; therefore, an IML can additionally reducerough surface scattering.The thickness (t) of a Ag2O shell can be precisely controlled

by the duration of the UVO illumination, and the formation ofa Ag2O shell reduces the diameter of the Ag core at the sametime (Supporting Information Figure S3). Notably, a thin voidlayer is observed after the oxidation, separating the core (Ag)and shell (Ag2O) (Figure 1b) because the diffusion velocity of

Figure 2.Wavelength-resolved scattering distribution of cloaked Ag NFs. (a) Dark-field microscopic images of bare Ag NFs, Ag/Ag2O NFs, and Ag/Ag2O NFs/IML with D’s of 1800 (left groups) and 750 nm (right groups). Scale bars, 10 μm. (b) Schematic of the measurement setup for thebackward scattering distribution. (c−e) Measured backward scattering distributions of bare Ag NFs, Ag/Ag2O NFs, and Ag/Ag2O NFs/IML with D= 300 (c), 750 (d), 1800 nm (e), acquired at λ = 450, 532, and 660 nm. The incident angle was set to 60° to the specimen. The two white dashedlines indicate the angular boundaries of ±40°. The oxidized samples have the same t value of 150 nm.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.8b01157Nano Lett. 2018, 18, 3865−3872

3867

the oxygen atoms from the shell is relatively slower than that ofthe Ag atoms from the core, which is known as the Kirkendalleffect.32 Randomly networked Ag/Ag2O NFs with (D, t) =(300, 150) nm, covered with an IML, displayed dramaticallyenhanced clarity through a glass substrate in typical room lightand even focused high-intensity light with 251 mW/cm2 (rightimages in Figure 1c, Supporting Information Figure S4, andSupporting Information Movie S1). The level of clarityprogressively improved with each fabrication step, indicatingthat both the oxidation and coating processes were importantfor rendering the Ag NFs to be near invisible to the naked eye.Remarkably, a thin void layer in the oxidized Ag NFs only had amarginal effect on the level of clarity through a glass substrate,which was supported by electromagnetic simulations, as shownin Supporting Information Figure S5.To probe whether the cloaked Ag/Ag2O NFs preserved their

pristine performance as a transparent electrode, we obtained

the transmittance (T), sheet resistance (Rs), and haze for threetypes of samples (bare Ag NFs, Ag/Ag2O NFs, and Ag/Ag2ONFs/IML) for a wide range (300−1800 nm) of initial Ag NFdiameters (D), as shown in Figures 1d and SupportingInformation Figure S6. For all of the samples, the area fractionsof the Ag NFs were initially the same at each Ag NF diameter.The oxidized samples had a 150 nm thick Ag2O shell. Theresults showed that the bare Ag NFs exhibited a high figure ofmerit over the entire range of Ag NF diameters, characterizedby T > 92% and Rs < 8.6 Ω/sq. More importantly, the cloakedsamples (i.e., Ag/Ag2O NFs/IML) had greater transmittanceand lower haze than the bare Ag NFs, despite the addition of anabsorptive Ag2O shell. The improved transmittance was partlyattributed to the shrinkage of the Ag core. However, thecomparison of the transmittance data from the Ag/Ag2O NFsbefore and after the IML indicated that the improvement intransmittance was mostly owing to the invisibility phenomenon

Figure 3. Electromagnetic analysis of the optical cloaking effects. (a) Schematic of a Ag/Ag2O NF on a glass substrate for simulation. The orthogonalpolarizations of TE and TM were defined as shown in the schematic. (b) Profiles of the electric-field intensity of a bare Ag NF, a Ag/Ag2O NF, and aAg/Ag2O NF/IML, acquired at λ = 530 nm for TE polarized light. The structural parameters of (D, t) are (800, 160) nm. (c) Surface plotspresenting the wavelength-dependent scattering efficiencies of a bare Ag NF, a Ag/Ag2O NF, and a Ag/Ag2O NF/IML with the variation of the D’sfor TE polarized light. (d) Profiles of the electric-field intensity of the structures with a Ag/Ag2O NF/IML (D = 800 nm) with a different t of 0, 20,or 100 nm, acquired at λ = 530 nm for TE polarized light. (e) Surface plots presenting the wavelength-dependent scattering efficiencies of Ag/Ag2ONFs/IML (D = 800 nm) with a variation of the t’s for TE polarized light.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.8b01157Nano Lett. 2018, 18, 3865−3872

3868

by which the optical loss within the Ag2O shell was fullycompensated. Although the Ag2O shell slightly increased Rsowing to its relatively low carrier density, the cloaked samplesstill had a high figure of merit as a transparent electrode; T =99% and Rs = 4.8 ± 1.1 Ω/sq were recorded at a Ag NFdiameter of 750 nm. The same trend was observed forrandomly networked Ag/Ag2O NTs covered with an IML, asshown in Supporting Information Figures S7 and S8, which hadhigher clarity and improved transmittance compared to bareNTs. These results suggested that the cloaking effect couldconceal any forms of mesoscopic metallic wires by introducingan absorptive conformal shell and a dielectric cover.To delineate the principle of the observed invisibility, we

obtained optical images of the three types of samples by dark-field microscopy with a 100× magnification lens (Figure 2a);the diameters of the Ag NFs were 750 (left group) and 1800nm (right group). For the oxidized samples, the thickness ofthe AgO2 shell was 150 nm. On the basis of the microscopyimages, both the submicrometer- and micrometer-scale Ag NFsgradually dimmed as each fabrication step (i.e., oxidation andcoating) proceeded, consistent with the camera images inFigure 1c and Supporting Information Figure S4. At the finalstep, the Ag-based NFs were near invisible in the dark-fieldmicroscopy images, which was indicative of suppressedbackward scattering. The level of backward scattering wasaccurately evaluated by scattering distribution measurements,for which a photodetector was programmed to move along thepolar (θ) and azimuthal (ϕ) directions, as shown in Figure2b.33 For the examination of the broadband invisibility over thefull visible spectrum, blue- (λ = 450 nm), green- (λ = 532 nm),red-emitting (λ = 660 nm) lasers were individually used as anincoming light source. The polarization of the incoming lighton such random network configuration is irrelevant. Themeasured backward scattering distributions revealed two keyfeatures related to the invisibility (Figure 2c−e and SupportingInformation Figure S9). First, for a fixed Ag NF diameter andlight wavelength, the scattering distribution becomes signifi-cantly localized after each fabrication step; the final cloakedsamples behaved as a specular surface that was characterized bya point-like scattering distribution. Second, such broadbandinvisibility was observed for all the considered diameters (300,750, and 1800 nm), although the level of scattering was slightlyaugmented with increasing Ag NF diameter. The scatteringcross section of a metal wire tends to increase with increasingits diameter because a larger metal wire excites more numbersof electric multipoles. Therefore, the level of scatteringsuppression is slightly reduced for relatively large Ag NFs(e.g., D = 750 and 1800 nm). Taken together, our cloakingstrategy, which concealed mesoscopic metallic wires via thesuppression of scattering, was reliable for a broad range of wirediameters and light wavelengths.As discussed in Figure 1b, the formation of an Ag2O shell was

accompanied by the reduction of the diameter of the Ag core.To rationalize that the invisibility observed herein occursmarginally by the reduced Ag core and dominantly by thepresence of the thin Ag2O shell, we prepared a series ofmicrometer-scale (D = 1800 nm) Ag NFs samples with agradual variation of the Ag2O shell thicknesses (t = 0−150 nm),as shown in Supporting Information Figure S10. Such giant AgNFs can be relatively free from the argument related to theshrinkage of a Ag core. The results showed that the scatteringdistributions became significantly localized after oxidation for 5min only for the Ag/Ag2O NFs/IML samples, as shown in

Supporting Information Figure S11. The transmission electronmicroscopy (TEM) images (Supporting Information FigureS10a) revealed that a very thin (approximately 20 nm) Ag2Oshell abruptly formed after an oxidation time of 5 min. Thisobservation confirmed that the presence of a Ag2O shelldominated the invisibility, in conjunction with the IML coating.Furthermore, the cloaked Ag NF samples dramaticallysuppressed the forward scattering as much as the backwardscattering, as shown in Supporting Information Figure S12,which suggests their generic use for both external and internalilluminations.To support the experimental findings discussed thus far, we

conducted numerical scattering analyses using finite-differencetime-domain (FDTD) simulations. In the simulation model, asingle hemispherical Ag NF with a conformal Ag2O shell wasplaced on a glass substrate, embedded within a backgroundmedium of air (n = 1.0) or a dielectric layer (n = 1.5), as shownin Figure 3a. To clearly verify the effect of the Ag2O shell, thediameter of the Ag core was not changed by the introduction ofan Ag2O shell. Typical snapshots of the electric-field intensityfor a Ag NF, a Ag/AgO2 NF, and a Ag/AgO2 NF/IML showthat the scattering was significantly suppressed by bothoxidation and coating steps, as shown in Figure 3b. For amore quantitative investigation, we obtained the scatteringcross sections of the three structures with a variation of the AgNF diameters (D = 260−1800 nm) for a normally incidentbroadband (λ = 400−700 nm) plane wave with transverse-electric (TE) or transverse-magnetic (TM) polarization (Figure3c for TE data and Supporting Information Figure S13a for TMdata). These surface plots represent a size-independentscattering suppression that was preserved over the full visiblespectrum for both TE and TM polarizations, which wasconsistent with the measured scattering distributions shown inFigure 2c−e and Supporting Information Figure S9. Bare AgNFs sustained discrete high-amplitude bands that wereidentified as localized surface plasmon resonances; however,the Ag/Ag2O NF, covered with a dielectric layer, significantlymitigated the plasmon-mediated scattering.In addition to the studies of cloaked Ag NFs with a wide

range of sizes, we investigated the effect of the Ag2O shellthickness. Figure 3d and Supporting Information Figure S13bshow the electric-field intensity snapshots of the 800 nmdiameter Ag NFs with different Ag2O shell thicknesses (0, 20,and 100 nm) for TE and TM polarizations, respectively. Thesimulated profiles indicated that the scattering was drasticallysuppressed by the addition of an ultrathin Ag2O (20 nm) shell;a thicker Ag2O shell slightly boosted the suppression ofscattering. A surface plot of the scattering efficiencies with thevariation of the Ag2O shell thicknesses and wavelengthsrevealed a cutoff Ag2O shell thickness (approximately 20 nmin this simulation) at which the scattering behavior wasdramatically altered (Figure 3e for TE data and SupportingInformation Figure S13c for TM data). This simulated resultaccounted for the measured scattering distributions with a fineincrement in the Ag2O shell thickness shown in SupportingInformation Figure S11. As a final analysis, we obtained thescattering efficiencies of the Ag NFs with a strongly absorptive(k = 0.5), weakly absorptive (k = 0.1), or transparent (k = 0)dielectric shell to verify the effectiveness of an absorptive Ag2Oshell, as shown in Supporting Information Figure S14. Acomparison of these surface plots indicated that the plasmonicscattering induced by a Ag NF vanished only when thedielectric shell was absorptive. Interestingly, the cloaked Ag

Nano Letters Letter

DOI: 10.1021/acs.nanolett.8b01157Nano Lett. 2018, 18, 3865−3872

3869

NFs with an absorptive Ag2O shell samples have a significantlyhigher transmittance than their counterpart bare Ag NFssamples (Figure 1c). In general, scattering and absorptionprocesses are strongly correlated and thus, a cloaked structurecan reduce an absorption cross section as much as a scatteringone.31 While the diameter of Ag NFs increases, the level ofscattering suppression becomes reduced, as shown in Figure2c−e, because larger Ag NFs are capable of sustaining morenumbers of high-order electric multipoles. This accounts for themore improved transmittance for relatively small Ag NFs (e.g.,D = 300 nm), as shown in Figure 1d.As an application example, we developed a defroster

(namely, a heater) using random networks of cloaked AgNFs, functioning as a transparent electrode for Joule heating.For a thermal test, a sample of cloaked Ag NFs (D = 750 nm)characterized by Rs of 4.8 ± 1.1 Ω/sq and T of 99% wasprepared. Both ends of this sample were connected to a voltagesupplier by using a silver epoxy paste (Figures 4a, 4b, andSupporting Information Figure S15). The fabricated sampleexhibited a uniform temperature distribution at a constantdirect current (dc) bias, as evident from an infrared (IR) imageshown in Figure 4c. Figure 4d shows the change of temperature

with increasing dc bias; the input dc bias was elevated to 6 Vwith a step increase of 1 V every 30 s, which finally yielded anaverage temperature of 248.2 ± 0.8 °C. A heating cycle test wasconducted by repeatedly applying a dc bias (6 V) to the samesample, as shown in Supporting Information Figure S16; nosignificant degradation in thermal performance was observedthroughout this test. To highlight the effectiveness of the AgNFs developed in this study as transparent electrodes, weprepared two samples containing Ag NWs (Rs = 15 Ω/sq) oran ITO film (Rs = 60 Ω/sq) and conducted a thermal test forall these samples at a dc bias of 2 V, as shown in Figure 4e. Forcomparison, the transmittance of these three samples wasadjusted to the same value (94% in this experiment). To obtainthis identical T value, random networks of Ag/Ag2O NFs(average diameter: 1139 nm) with Rs of 1.9 ± 0.5 Ω/sq and Tof 94% at 550 nm were used. A dramatic increase in the thermalequilibrium temperature was observed for the Ag NFs, incontrast to that of the Ag NWs and ITO film, which stemmedfrom their different Rs’s.Figure 4f and Supporting Information Figure S17 represent

the heating and cooling rates of these three samples for whichdifferent dc biases (2 V for the cloaked Ag NFs, 5.5 V for the

Figure 4. Thermal characteristics of cloaked Ag NFs and demonstration of a defroster for automobile windshields. (a) Schematic illustration of athermal test experiment using Ag NFs. (b) Photograph of the fabricated Ag/Ag2O NFs/IML heater. Scale bar, 2 cm. (c) The temperaturedistribution of the fabricated Ag/Ag2O NFs/IML heater at a dc bias of 6 V obtained with an IR camera. Scale bar, 2 cm. (d) Measured temporalchange of the average temperatures of the fabricated Ag/Ag2O NFs/IML heater with increasing dc bias. The dc bias was elevated to 6 V with a stepincrease of 1 V every 30 s. (e) Measured temporal change of the average temperatures of the heaters containing Ag/Ag2O NFs/IML (Rs = 1.9 Ω/sq),Ag NWs (Rs = 15 Ω/sq), and ITO film (Rs = 60 Ω/sq) at a dc bias of 2 V. All the heaters exhibited the same transmittance of 94% at λ = 550 nm. (f)Comparison of the heating and cooling rates of the same heaters as shown in (e) with different applied dc biases (2 V for the Ag/Ag2O NFs/IML,5.5 V for the Ag NWs, and 11 V for the ITO film) to obtain the same thermal equilibrium temperature of 120 °C. (g) Measured defogging time ofthe Ag/Ag2O NFs/IML heater with variation of the dc biases. (h) Photographs of the Ag/Ag2O-NF/IML heater integrated automobile windshield atturn-off (left) and turn-on (right) modes. The applied dc bias was 6 V. Scale bars, 5 cm. The insets show a driver’s view through the windshield(lower) and an image captured with an IR camera displaying the temperature distribution of the windshield (upper).

Nano Letters Letter

DOI: 10.1021/acs.nanolett.8b01157Nano Lett. 2018, 18, 3865−3872

3870

Ag NWs, and 11 V for the ITO film) were applied to reach thesame thermal equilibrium temperature of 120 °C. The cloakedAg NFs sample had faster heating and cooling rates than theother samples, as shown in Figure 4f, because the Ag NFs hadgreater thermal conductivity that results from their higherelectrical conductivity by the Wiedemann−Franz law. Inaddition, we conducted a heat spreading test for the samesamples; one end of these samples was connected to a hot platewith a constant temperature of 140 °C, while the opposite endwas placed on a cooling pad with a constant temperature of 23°C, as schematically shown in Supporting Information FigureS18a. The Ag NFs sample featured relatively fast heatspreading, as shown in Supporting Information Figures S18band S18c. These great thermal and optical capabilities ofcloaked Ag NFs can be appropriately harnessed for themanufacturing of a defroster that requires high clarity, hightransmittance, and high defogging speed. A defroster fabricatedfrom cloaked Ag NFs (Rs = 4.8 ± 1.1 Ω/sq and T = 99%)exhibited a linear relation between defogging time and anapplied dc bias, as shown in Figure 4g; defogging time was lessthan 1 s at a dc bias of 6 V. Such rapid defogging is desirable forautomotive windshields that require a clear view while driving.To do this, we fabricated a windshield using the cloaked AgNFs for an automobile model and conducted a defogging test,as shown in Figure 4h. Intentionally formed frost covering theentire surface of the windshield was completely evaporatedwithin 1 s at a dc bias of 6 V (Supporting Information MovieS2).The broadband invisibility cloak discussed herein is not

limited to Ag wires but can be generalized to mesoscopicparticles and flakes of any metal, including Cu as a cost-effectiveelectrode material. The effect of the scattering suppression canbe further enhanced by tailored design of the shell and coverlayer (e.g., appropriately designed double shells combined witha high-refractive-index cover layer). These experimental andtheoretical results will be utilized for the development ofvarious imaging applications that require a high-clarity view anduse metal structures.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.nano-lett.8b01157.

Detailed description of experimental and numericalmethods and additional figures (PDF)Clarity test of bare and cloaked Ag wires in high-intensityvisible light (AVI)Defogging test of a cloaked-Ag-wires-based defroster onan automotive windshield (AVI)

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: jangung@unist.ac.kr (J.-U.P).*E-mail: sunkim@khu.ac.kr (S.-K.K.).ORCIDSun-Kyung Kim: 0000-0002-0715-0066Author Contributions§S.W.K. and B.W.A. contributed equally to this work.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

J.-U.P. was supported by the Ministry of Science & ICT and theMinistry of Trade, Industry and Energy (MOTIE) of Koreat h r o u g h t h e N a t i o n a l R e s e a r c h F o u n d a t i o n(2016R1A2B3013592 and 2016R1A5A1009926), the NanoMa t e r i a l T e c h n o l o g y D e v e l o pm e n t P r o g r am(2015M3A7B4050308 and 2016M3A7B4910635), the Con-vergence Technology Development Program for Bionic Arm(NRF-2017M3C1B2085316), the Industrial Technology In-novation Program (10080577), and the Pioneer ResearchCenter Program (NRF-2014M3C1A3001208). S.-K.K. wassupported by the Basic Science Research Program throughthe National Research Foundation of Korea (NRF-2017R1A2B4005480) funded by the Ministry of Science,ICT, and Future Planning.

■ REFERENCES(1) Johnson, P. B.; Christy, R.-W. Phys. Rev. B 1972, 6, 4370−4379.(2) Novotny, L.; Van Hulst, N. Nat. Photonics 2011, 5, 83−90.(3) Atwater, H. A.; Polman, A. Nat. Mater. 2010, 9, 205−213.(4) Kim, N. Y.; Hong, S. H.; Kang, J. W.; Myoung, N.; Yim, S. Y.;Jung, S.; Lee, K.; Tu, C. W.; Park, S. J. RSC Adv. 2015, 5, 19624−19629.(5) Park, K. D.; Jiang, T.; Clark, G.; Xu, X. D.; Raschke, M. B. Nat.Nanotechnol. 2018, 13, 59−64.(6) Pan, L.; Park, Y.; Xiong, Y.; Ulin-Avila, E.; Wang, Y.; Zeng, L.;Xiong, S. M.; Rho, J.; Sun, C.; Bogy, D. B.; Zhang, X. Sci. Rep. 2011, 1,175.(7) Fang, Y. S.; Wu, Z. C.; Li, J.; Jiang, F. Y.; Zhang, K.; Zhang, Y. L.;Zhou, Y. H.; Zhou, J.; Hu, B. Adv. Funct. Mater. 2018, 28, 1705409.(8) Lee, M. S.; Lee, K.; Kim, S. Y.; Lee, H.; Park, J.; Choi, K. H.; Kim,H. K.; Kim, D. G.; Lee, D. Y.; Nam, S.; Park, J. U. Nano Lett. 2013, 13,2814−2821.(9) Kim, J.; Kim, M.; Lee, M. S.; Kim, K.; Ji, S.; Kim, Y. T.; Park, J.;Na, K.; Bae, K. H.; Kim, H. K.; Bien, F.; Lee, C. Y.; Park, J. U. Nat.Commun. 2017, 8, 8.(10) Kim, J.; Lee, M. S.; Jeon, S.; Kim, M.; Kim, S.; Kim, K.; Bien, F.;Hong, S. Y.; Park, J. U. Adv. Mater. 2015, 27, 3292−3297.(11) Yu, G. H.; Cao, A. Y.; Lieber, C. M. Nat. Nanotechnol. 2007, 2,372−377.(12) Wu, Y.; Xiang, J.; Yang, C.; Lu, W.; Lieber, C. M. Nature 2004,430, 61−65.(13) Park, J.; Kim, J.; Kim, S. Y.; Cheong, W. H.; Jang, J.; Park, Y. G.;Na, K.; Kim, Y. T.; Heo, J. H.; Lee, C. Y.; Lee, J. H.; Bien, F.; Park, J.U. Sci. Adv. 2018, 4, eaap9841.(14) Jang, J.; Hyun, B. G.; Ji, S.; Cho, E.; An, B. W.; Cheong, W. H.;Park, J. U. NPG Asia Mater. 2017, 9, e432.(15) 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. NanoLett. 2010, 10, 4242−4248.(16) An, B. W.; Gwak, E. J.; Kim, K.; Kim, Y. C.; Jang, J.; Kim, J. Y.;Park, J. U. Nano Lett. 2016, 16, 471−478.(17) An, B. W.; Hyun, B. G.; Kim, S. Y.; Kim, M.; Lee, M. S.; Lee, K.;Koo, J. B.; Chu, H. Y.; Bae, B. S.; Park, J. U. Nano Lett. 2014, 14,6322−6328.(18) Im, H. G.; An, B. W.; Jin, J.; Jang, J.; Park, Y. G.; Park, J. U.; Bae,B. S. Nanoscale 2016, 8, 3916−3922.(19) Wu, H.; Kong, D. S.; Ruan, Z. C.; Hsu, P. C.; Wang, S.; Yu, Z.F.; Carney, T. J.; Hu, L. B.; Fan, S. H.; Cui, Y. Nat. Nanotechnol. 2013,8, 421−425.(20) Ni, X. J.; Wong, Z. J.; Mrejen, M.; Wang, Y.; Zhang, X. Science2015, 349, 1310−1314.(21) Balci, O.; Polat, E. O.; Kakenov, N.; Kocabas, C. Nat. Commun.2015, 6, 6628.(22) Fan, P. Y.; Chettiar, U. K.; Cao, L. Y.; Afshinmanesh, F.;Engheta, N.; Brongersma, M. L. Nat. Photonics 2012, 6, 380−385.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.8b01157Nano Lett. 2018, 18, 3865−3872

3871

(23) Chen, X. Z.; Luo, Y.; Zhang, J. J.; Jiang, K.; Pendry, J. B.; Zhang,S. A. Nat. Commun. 2011, 2, 176.(24) Gabrielli, L. H.; Cardenas, J.; Poitras, C. B.; Lipson, M. Nat.Photonics 2009, 3, 461−463.(25) Valentine, J.; Li, J. S.; Zentgraf, T.; Bartal, G.; Zhang, X. Nat.Mater. 2009, 8, 568−571.(26) Liu, R.; Ji, C.; Mock, J. J.; Chin, J. Y.; Cui, T. J.; Smith, D. R.Science 2009, 323, 366−369.(27) Kim, K. H.; No, Y. S.; Chang, S.; Choi, J. H.; Park, H. G. Sci.Rep. 2015, 5, 16027.(28) Alu, A.; Engheta, N. Phys. Rev. E 2005, 72, 016623.(29) Alu, A.; Engheta, N. Phys. Rev. Lett. 2008, 100, 113901.(30) Gao, X. Y.; Wang, S. Y.; Li, J.; Zheng, Y. X.; Zhang, R. J.; Zhou,P.; Yang, Y. M.; Chen, L. Y. Thin Solid Films 2004, 455-456, 438−442.(31) Kim, S. K.; Zhang, X.; Hill, D. J.; Song, K. D.; Park, J. S.; Park,H. G.; Cahoon, J. F. Nano Lett. 2015, 15, 753−758.(32) Yu, L.; Yan, Z. Y.; Cai, Z. H.; Zhang, D. T.; Han, P.; Cheng, X.M.; Sun, Y. G. Nano Lett. 2016, 16, 6555−6559.(33) Moon, Y.-J.; Na, J.-Y.; Park, Y. H.; Kim, S. B.; Lee, S. W.; Kim,S.-K. ACS Photonics 2018, 5, 752.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.8b01157Nano Lett. 2018, 18, 3865−3872

3872